JP2004192461A - Information processor, information processing method, information processing system, recording medium, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To generate user's taste information on the basis of attribute information of contents meeting prescribed conditions. <P>SOLUTION: Program vectors of programs being user model generation objects are acquired in a step S231, and a topic inputted by the user is set as a filtering condition in a step S232. Program vectors are filtered on the basis of the filtering condition to extract program vectors meeting the topic in a step S233, and a user model vector is generated from the sum total of extracted program vectors in a step S234. Cosine distances between the user model vector and program vectors of candidate programs for obtaining recommended information are calculated in a step S235, and the prescribed number of programs out of candidate programs of which the program vectors have higher degrees of similarity to the user model vector are extracted in order from the highest degree as recommended information in a step S236. This invention is applicable to program recommendation apparatus . <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

【０２３３】
なお、式（５）においては、番組ベクトルPP＝（ｐ₁、ｐ₂、・・・）、正履歴ベクトルUP＝（ｕ₁，ｕ₂，・・・）、番組側効き目ベクトルEfPP＝（ｅｐｄ₁，ｅｐｄ₂，・・・）、ユーザ側効き目ベクトルEfUP＝（ｅｕｄ₁，ｅｕｄ₂，・・・）であるものとする。また、式（５）においては、番組側効き目ベクトルEfPPおよびユーザ側効き目ベクトルEfUPのいずれをも利用するものとして説明しているが、設定により、番組側効き目ベクトルEfPPおよびユーザ側効き目ベクトルEfUPのうちのいずれかが利用されない場合、利用されないベクトルに代わって数値「１」が代入されて計算される。
【０２３４】
また、ユーザ側効き目ベクトルEfUPは、ユーザにより設定可能であっても、ユーザの初期設定を基に設定されるものであっても良いし、ユーザ情報登録部６３において生成されるものであっても良い。ユーザ側効き目ベクトルEfUPの生成の詳細については、図２２乃至図２５を用いて後述する。
【０２３５】
ステップＳ１２３において、ベクトル演算部６２は、ステップＳ１２２において算出された、番組ベクトルPPと正履歴ベクトルUPとの類似度SimUPを比較し、類似度の上位から、例えば、１０などの所定の数の番組ベクトルPPを抽出する。
【０２３６】
ステップＳ１２４において、ベクトル演算部６２は、必要に応じて、ユーザ情報登録部６３から、ユーザ側反効き目ベクトルEfMUPを読み込み、次の式（６）を用いて、ステップＳ１２３において抽出された番組ベクトルPPと負履歴ベクトルMUPのコサイン距離を算出する。
【０２３７】
【数２】

【０２３８】
なお、式（６）においては、番組ベクトルPP＝（ｐ₁、ｐ₂、・・・）、負履歴ベクトルＭUP＝（ｍ₁，ｍ₂，・・・）、番組側効き目ベクトルEfPP＝（ｅｐｄ₁，ｅｐｄ₂，・・・）、ユーザ側反効き目ベクトルEfMUP＝（ｅｍｄ₁，ｅｍｄ₂，・・・）であるものとする。また、式（６）においては、番組側効き目ベクトルEfPPおよびユーザ側反効き目ベクトルEfMUPのいずれをも利用するものとして説明しているが、設定により、番組側効き目ベクトルEfPPおよびユーザ側反効き目ベクトルEfMUPのうちのいずれかが利用されない場合、利用されないベクトルに代わって、数値「１」が代入されて計算される。
【０２３９】
また、ユーザ側反効き目ベクトルEfMUPは、ユーザにより設定可能であっても、ユーザの初期設定を基に設定されるものであっても良いし、ユーザ情報登録部６３において生成されるものであっても良い。ユーザ側反効き目ベクトルEfMUPの生成の詳細については、図２６または図２７を用いて後述する。
【０２４０】
ステップＳ１２５において、ベクトル演算部６２は、番組ベクトルPPと正履歴ベクトルUPとの類似度SimUP−番組ベクトルPPと負履歴ベクトルとの類似度SimMUPを算出し、その上位となる所定数（例えば３つ）の番組の番組ベクトルPPまたはＥＰＧデータを推薦情報として抽出し、推薦情報出力部４９に出力して、推薦番組リスト５０に登録させるとともに、テレビジョン表示装置１１および録画再生装置１２に出力して、処理が終了される。
【０２４１】
このような処理により、設定に応じて、番組側効き目ベクトルEfPP、ユーザ側効き目ベクトルEfUP、または、ユーザ側反効き目ベクトルEfMUPが用いられて、推薦情報が抽出されるので、ユーザの嗜好を正しく反映した番組を推薦するようにすることができる。
【０２４２】
図１９を用いて説明した処理では、番組ベクトルPP、正履歴ベクトルUPおよび負履歴ベクトルMUPが、詳細項目の全要素を１列に並べたベクトルで表されている場合の処理について説明したが、番組ベクトルPP、正履歴ベクトルUPおよび負履歴ベクトルMUPを、それぞれ、大項目ごとに演算するようにしても良い。
【０２４３】
次に、図２０のフローチャートを参照して、大項目ごとに、番組側効き目ベクトルEfPP、ユーザ側効き目ベクトルEfUP、または、ユーザ側反効き目ベクトルEfMUPを反映させることができるようにした、マッチング処理４について説明する。
【０２４４】
ステップＳ１３１において、図１９のステップＳ１２１と同様の処理が実行されて、効き目ベクトルの利用設定内容が取得される。
【０２４５】
ステップＳ１３２において、ベクトル演算部６２は、供給された番組ベクトルPPと、正履歴保存部４７から読み出した正履歴ベクトルUPのそれぞれの大項目について、番組ベクトルPPと正履歴ベクトルUPのコサイン距離を算出する。ここでは、効き目ベクトルは演算に利用されない。
【０２４６】
ステップＳ１３３において、ベクトル演算部６２は、次の式（７）を用いて、項目ごとに算出されたコサイン距離に、必要に応じて効き目ベクトルを乗算し、得られた値を合計して、類似度SimUPを算出する。
【０２４７】
SimUP＝epd_t・eud_t・cosθu_t+ epd_g・eud_g・cosθu_g+ epd_p・eud_p・cosθu_p
+ epd_a・eud_a・cosθu_a+ epd_k・eud_k・cosθu_k・・・（７）
【０２４８】
なお、式（７）においては、番組ベクトルPP＝（ｐ_t，ｐ_g，ｐ_p，ｐ_a，ｐ_k）と正履歴ベクトルUP＝（ｕ_t，ｕ_g，u_p，u_a，u_k）との大項目ごとのコサイン距離が、（cosθu_t，cosθu_g，cosθu_p ，cosθu_a，cosθu_k）であり、番組側効き目ベクトルEfPP＝（epd_t，epd_g，epd_p，epd_a，epd_k）、ユーザ側効き目ベクトルEfUP＝（eud_t，eud_g，eud_p，eud_a，eud_k）であるものとする。また、式（７）においては、番組側効き目ベクトルEfPPおよびユーザ側効き目ベクトルEfUPのいずれをも利用するものとして説明しているが、設定により、番組側効き目ベクトルEfPPおよびユーザ側効き目ベクトルEfUPのうちのいずれかが利用されない場合、利用されないベクトルに代わって、数値「１」が代入されて計算される。
【０２４９】
ステップＳ１３４において、ベクトル演算部６２は、ステップＳ１３３において算出された、番組ベクトルPPと正履歴ベクトルUPとの類似度SimUPを比較し、類似度の上位から、例えば、１０などの所定の数の番組ベクトルPPを抽出する。
【０２５０】
ステップＳ１３５において、ベクトル演算部６２は、ステップＳ１３４の処理により抽出された番組ベクトルPPと、負履歴保存部４８から読み出した負履歴ベクトルMUPのそれぞれの大項目について、番組ベクトルPPと負履歴ベクトルMUPのコサイン距離を算出する。ここでは、効き目ベクトルは演算に利用されない。
【０２５１】
ステップＳ１３６において、ベクトル演算部６２は、次の式（８）を用いて、項目ごとに算出されたコサイン距離に、必要に応じて効き目ベクトルを乗算し、得られた値を合計して、類似度SimMUPを算出する。
【０２５２】
SimMUP＝epd_t・emd_t・cosθm_t+ epd_g・emd_g・cosθm_g+ epd_p・emd_p・cosθm_p
+ epd_a・emd_a・cosθm_a+ epd_k・emd_k・cosθm_k・・・（８）
【０２５３】
なお、式（８）においては、番組ベクトルPP＝（ｐ_t，ｐ_g，ｐ_p，ｐ_a，ｐ_k）と負履歴ベクトルUP＝（ｍ_t，ｍ_g，ｍ_p，ｍ_a，ｍ_k）との大項目ごとのコサイン距離が、（cosθm_t，cosθm_g，cosθm_p ，cosθm_a，cosθm_k）であり、番組側効き目ベクトルEfPP＝（epd_t，epd_g，epd_p，epd_a，epd_k、）、ユーザ側反効き目ベクトルEfMUP＝（emd_t，emd_g，emd_p，emd_a，emd_k）であるものとする。また、式（８）においては、番組側効き目ベクトルEfPPおよびユーザ側反効き目ベクトルEfMUPのいずれをも利用するものとして説明しているが、設定により、番組側効き目ベクトルEfPPおよびユーザ側反効き目ベクトルEfMUPのうちのいずれかが利用されない場合、利用されないベクトルに代わって、数値「１」が代入されて計算される。
【０２５４】
ステップＳ１３７において、ベクトル演算部６２は、｛番組ベクトルPPと正履歴ベクトルUPとのコサイン距離である類似度SimUP｝−｛番組ベクトルPPと負履歴ベクトルとのコサイン距離である類似度SimMUP｝を算出し、その上位となる所定数（例えば３つ）の番組の番組ベクトルPPまたはＥＰＧデータを推薦情報として抽出し、推薦情報出力部４９に出力して、推薦番組リスト５０に登録させるとともに、テレビジョン表示装置１１および録画再生装置１２に出力して、処理が終了される。
【０２５５】
このような処理により、大項目ごとに、効き目ベクトルが利用されて重み付けが行われるので、ユーザの嗜好に詳細に合致した推薦情報を生成することが可能となる。
【０２５６】
次に、図２１のフローチャートを参照して、図１６を用いて説明した正履歴ベクトルおよび負履歴ベクトル生成処理２の処理により生成された、ジャンル別の正履歴ベクトルUPおよび負履歴ベクトルMUP、並びに、ジャンル別のユーザ側効き目ベクトルEfUPおよびユーザ側反効き目ベクトルEfMUPを用いて、マッチング処理を実行するマッチング処理５について説明する。
【０２５７】
ステップＳ１４１において、図１９のステップＳ１２１と同様の処理が実行されて、効き目ベクトルの利用設定内容が取得される。
【０２５８】
ステップＳ１４２において、ベクトル演算部６２は、供給された番組ベクトルPPのジャンルを抽出する。ここでは、例えば、供給された番組ベクトルPPのジャンルが「ドラマ」であったものとして説明する。
【０２５９】
ステップＳ１４３において、ベクトル演算部６２は、供給された番組ベクトルPPと、正履歴保存部４７から読み出した、ジャンル「ドラマ」に対応する正履歴ベクトルUPのそれぞれの大項目について、番組ベクトルPPと正履歴ベクトルUPのコサイン距離を算出する。ここでは、効き目ベクトルは演算に利用されない。
【０２６０】
ステップＳ１４４において、ベクトル演算部６２は、次の式（９）を用いて、項目ごとに算出されたコサイン距離に、必要に応じて、ジャンル「ドラマ」に対応するユーザ側効き目ベクトルを乗算し、得られた値を合計して、類似度SimUPを算出する。
【０２６１】
SimUP＝epd_t・eud_td・cosθu_td+ epd_g・eud_gd・cosθu_gd+ epd_p・eud_pd・cosθu_pd+ epd_a・eud_ad・cosθu_ad+ epd_k・eud_kd・cosθu_kd・・・（９）
【０２６２】
なお、式（９）においては、番組ベクトルPP＝（ｐ_t，ｐ_g，ｐ_p，ｐ_a，ｐ_k）と、ジャンル「ドラマ」に対応する正履歴ベクトルUP＝（ｕ_t，ｕ_g，u_p，u_a，u_k）との大項目ごとのコサイン距離が、（cosθu_td，cosθu_gd，cosθu_pd ，cosθu_ad，cosθu_kd）であり、番組側効き目ベクトルEfPP＝（epd_t，epd_g，epd_p，epd_a，epd_k）、ジャンル「ドラマ」に対応するユーザ側効き目ベクトルEfUP＝（eud_td，eud_gd，eud_pd，eud_ad，eud_kd）であるものとする。また、式（９）においては、番組側効き目ベクトルEfPPおよびユーザ側効き目ベクトルEfUPのいずれをも利用するものとして説明しているが、設定により、番組側効き目ベクトルEfPPおよびユーザ側効き目ベクトルEfUPのうちのいずれかが利用されない場合、利用されないベクトルに代わって、数値「１」が代入されて計算される。
【０２６３】
ステップＳ１４５において、ベクトル演算部６２は、ステップＳ１４３において算出された、番組ベクトルPPと正履歴ベクトルUPとの類似度SimUPを比較し、類似度の上位から、例えば、１０などの所定の数の番組ベクトルPPを抽出する。
【０２６４】
ステップＳ１４６において、ベクトル演算部６２は、ステップＳ１４５の処理により抽出された番組ベクトルPPと、負履歴保存部４８から読み出したジャンル「ドラマ」に対応する負履歴ベクトルMUPのそれぞれの大項目について、番組ベクトルPPと負履歴ベクトルMUPのコサイン距離を算出する。ここでは、効き目ベクトルは演算に利用されない。
【０２６５】
ステップＳ１４７において、ベクトル演算部６２は、次の式（１０）を用いて、項目ごとに算出されたコサイン距離に、必要に応じて、ジャンル「ドラマ」に対応する効き目ベクトルを乗算し、得られた値を合計して、類似度SimMUPを算出する。
【０２６６】
SimMUP＝epd_t・emd_td・cosθm_td+ epd_g・emd_gd・cosθm_gd+ epd_p・emd_pd・cosθm_pd+ epd_a・emd_ad・cosθm_ad+ epd_k・emd_kd・cosθm_kd・・・（１０）
【０２６７】
なお、式（１０）においては、番組ベクトルPP＝（ｐ_t，ｐ_g，ｐ_p，ｐ_a，ｐ_k）と負履歴ベクトルＭUP＝（ｍ_t，ｍ_g，ｍ_p，ｍ_a，ｍ_k）との大項目ごとのコサイン距離が、（cosθm_td，cosθm_gd，cosθm_pd ，cosθm_ad，cosθm_kd）であり、番組側効き目ベクトルEfPP＝（epd_t，epd_g，epd_p，epd_a，epd_k）、ジャンル「ドラマ」に対応するユーザ側反効き目ベクトルEfMUP＝（emd_td，emd_gd，emd_pd，emd_ad，emd_kd）であるものとする。また、式（１０）においては、番組側効き目ベクトルEfPPおよびユーザ側反効き目ベクトルEfMUPのいずれをも利用するものとして説明しているが、設定により、番組側効き目ベクトルEfPPおよびユーザ側反効き目ベクトルEfMUPのうちのいずれかが利用されない場合、利用されないベクトルに代わって、数値「１」が代入されて計算される。
【０２６８】
ステップＳ１４８において、ベクトル演算部６２は、｛番組ベクトルPPと正履歴ベクトルUPとのコサイン距離である類似度SimUP｝−｛番組ベクトルPPと負履歴ベクトルとのコサイン距離である類似度SimMUP｝を算出し、その上位となる所定数（例えば３つ）の番組の番組ベクトルPPまたはＥＰＧデータを推薦情報として抽出し、推薦情報出力部４９に出力して、推薦番組リスト５０に登録させるとともに、テレビジョン表示装置１１および録画再生装置１２に出力して、処理が終了される。
【０２６９】
このような処理により、大項目ごとに、ジャンル別の正履歴ベクトルUPおよび負履歴ベクトルMUPと番組ベクトルPPとのコサイン距離が求められ、ジャンルに対応した効き目ベクトルが利用されて重み付けが行われて類似度が算出されるので、ユーザの嗜好に詳細に合致した推薦情報を生成することが可能となる。
【０２７０】
また、ユーザ側効き目ベクトルEfUPおよび反効き目ベクトルEfMUPは、上述したように、初期登録保存部４５に登録されている、ユーザの初期登録内容を基に生成されるようにしても良いし、正履歴ベクトルUPまたは負履歴ベクトルMUP、あるいは、一定期間にユーザが視聴した番組をカウントすることにより、ユーザ固有のユーザ側効き目ベクトルEfUPおよび反効き目ベクトルEfMUPが生成されるようにしても良い。
【０２７１】
図２２のフローチャートを参照して、一定期間にユーザが視聴した番組をカウントすることによりユーザ側効き目ベクトルEfUPを生成する、ユーザ側効き目ベクトル生成処理１について説明する。
【０２７２】
ステップＳ１５１において、マッチング処理部４３のユーザ情報登録部６３は、未処理の大項目のうちのいずれかを選択する。
【０２７３】
ステップＳ１５２において、ユーザ情報登録部６３は、正履歴保存部４７に保存されている正履歴を参照して、例えば、１週間、１ヶ月、または３ヶ月などの一定期間中にユーザが視聴した番組を検出し、番組ベクトル抽出部４２に、一定期間中にユーザが視聴した番組に対応する番組ベクトルPPをデータ取得部４１から抽出させて、ステップＳ１５１において選択された大項目に含まれる詳細項目ごとに番組数をカウントする。
【０２７４】
具体的には、例えば、ステップＳ１５１において選択された大項目が、大項目ジャンルＧｍ＝｛ドラマ，バラエティ，スポーツ，映画，音楽，子供向け／教育，教養／ドキュメント，ニュース／報道，その他｝であれば、ユーザ情報登録部６３は、一定期間中にユーザが視聴した番組の番組ベクトルPPの構成要素を、それぞれに対応する項目に分類してカウントする。一定期間中にユーザが視聴した番組が、例えば、５０番組であった場合、番組数のカウント結果は、例えば、ジャンルＧｍ＝（１０，１８，５，２，８，１，０，１，５）となる。
【０２７５】
ステップＳ１５３において、ユーザ情報登録部６３は、番組ベクトル抽出部４２に、同一期間における全ての番組に対応する番組ベクトルPPをデータ取得部４１から抽出させて、ステップＳ１５１において選択された大項目に含まれる詳細項目ごとに番組数をカウントする。
【０２７６】
具体的には、例えば、ステップＳ１５１において選択された大項目が、大項目ジャンルＧｍ＝｛ドラマ，バラエティ，スポーツ，映画，音楽，子供向け／教育，教養／ドキュメント，ニュース／報道，その他｝であれば、ユーザ情報登録部６３は、同一期間における全ての番組の番組ベクトルPPの構成要素を、それぞれに対応する項目に分類してカウントする。同一期間における全ての番組が、例えば、１０００番組であった場合、番組数のカウント結果は、例えば、ジャンルＧｍ＝（１０４，２３９，６８，２５，７８，９１，６０，２５４，８１）となる。
【０２７７】
ステップＳ１５４において、ユーザ情報登録部６３は、ステップＳ１５２およびステップＳ１５３のカウント結果を基に、ユーザの視聴実績のカウント数／全番組のカウント数を算出する。
【０２７８】
番組編成は、視聴率競争の影響により、大衆の嗜好を反映していると考えられる。すなわち、ユーザの視聴実績のカウント数／全番組のカウント数の演算は、換言すれば、ユーザの視聴実績のカウント数を、標準モデルとしての全番組のカウント数で正規化することと同義である。ステップＳ１５４において算出される正規化ベクトルを正規化ベクトルＤと称するものとする。
【０２７９】
例えば、大項目ジャンルＧｍ＝｛ドラマ，バラエティ，スポーツ，映画，音楽，子供向け／教育，教養／ドキュメント，ニュース／報道，その他｝において、１週間の全番組のカウント数が、（８，１２，３，７，６，４，２，８，１０）であり、対してユーザが視聴した番組のカウント数が、（４，０，１，２，３，４，５，５，２)であった場合、正規化ベクトルＤは、以下のようになる。
Ｄ＝（４／８，０／１２，１／３ ，２／７，３／６，４／４，
１／２，２／８，２／１０）
＝（０．５，０，０．３３，０．２８，０．５，１．０，
０．５，０．１３，０．２）
すなわち、正規化ベクトルＤの構成要素が１．０である場合は、設定された期間中に、該当する項目の全番組を視聴したこと、正規化ベクトルＤの構成要素が０である場合は、設定された期間中に、該当する番組は一つも視聴しなかったことを意味する。
【０２８０】
ステップＳ１５５において、ユーザ情報登録部６３は、ステップＳ１５４の算出結果を基に、対応する大項目の効き目ベクトルを生成する。
【０２８１】
効き目ベクトルを生成するために、大項目ジャンルＧｍ＝｛ドラマ，バラエティ，スポーツ，映画，音楽，子供向け／教育，教養／ドキュメント，ニュース／報道，その他｝のうちのいずれかの項目を用いて、標準値を設定する。例えば、ジャンル：「ドラマ」の番組1週間分において、一般的なユーザは、その２割程度を視聴しているとして、標準値を０．２に設定すればよい。大項目の効き目ベクトルは、相対値として算出されるため、設定値は０乃至１のいずれの値であってもかまわない。そして、ユーザ側効き目ベクトルは、ステップＳ１５４において算出された正規化ベクトルＤと、設定値との相対値である。
【０２８２】
従って、ユーザの興味ジャンルを示す、大項目ジャンルＧｍの効き目ベクトルＥは、
Ｅ＝（０．３，−０．２，０．１３，０．０８，０．３，
０．８，０．３，−０．０７，０．０）
と算出されるので、対応するユーザは、「子供向け／教育」のジャンルを好み、「バラエティ」のジャンルを好まないと判断することができる。
【０２８３】
ステップＳ１５６において、ユーザ情報登録部６３は、全ての大項目の効き目ベクトルが生成されたか否かを判断する。ステップＳ１５６において、全ての大項目の効き目ベクトルが生成されていないと判断された場合、処理は、ステップＳ１５１に戻り、それ以降の処理が繰り返される。
【０２８４】
ステップＳ１５６において、全ての大項目の効き目ベクトルが生成されたと判断された場合、ステップＳ１５７において、ユーザ情報登録部６３は、全ての大項目の効き目ベクトルを保存して、処理が終了される。
【０２８５】
このような処理により、一般的な嗜好と、ユーザ固有の嗜好との差を求めることができる。また、ユーザ側効き目ベクトルEfUPを、例えば、３ヶ月、半年などの所定の期間ごとに再演算するようにすることにより、ユーザの嗜好をリアルタイムに反映した番組を推薦することができる。
【０２８６】
また、図２２の処理においては、例えば、１週間、１ヶ月、または３ヶ月などの一定期間中にユーザが視聴した番組を基にユーザ側効き目ベクトルEfUPを求めるものとして説明したが、複数の期間を対象として、例えば、短期、中期、長期に対応するユーザ側効き目ベクトルEfUPを算出し、これら複数の効き目ベクトルを用いて、推薦情報を決定するようにしても良い。
【０２８７】
ここでは、ユーザ固有の嗜好を、ユーザ側効き目ベクトルEfUPとして利用する場合について説明したが、ユーザ固有の嗜好を、正履歴ベクトルUPとしてマッチング処理に用いるようにしても良い。
【０２８８】
また、放送されている全番組に代わって、視聴者が最も番組を視聴する所定の時間帯（例えば、１８時乃至２２時の、いわゆる、ゴールデンタイム）に放送される全番組をカウントするようにしてもよい。このようにすることにより、一般的な嗜好を求めるための演算処理を大幅に少なくすることができる。
【０２８９】
次に、図２３のフローチャートを参照して、正履歴ベクトルUPと、大衆の嗜好との類似度を示すコサイン距離の演算によって、ユーザ固有の嗜好と、大衆の嗜好との隔たりをマッチングに利用するためのユーザ側効き目ベクトルEfUPを求める、ユーザ側効き目ベクトル算出処理２について説明する。
【０２９０】
ステップＳ１６１において、マッチング処理部４３のユーザ情報登録部６３は、正履歴保存部４７に保存されている正履歴ベクトルUPを取得する。
【０２９１】
ステップＳ１６２において、ユーザ情報登録部６３は、一般的な嗜好を示す、標準嗜好ベクトルAPPを取得する。
【０２９２】
標準嗜好ベクトルAPPは、配信サーバ５から供給されるものとしても良いし、または、番組編成は、視聴率競争の影響により、大衆の嗜好を反映していると考えられるので、図２２を用いて説明したユーザ側効き目ベクトル算出処理１と同様にして、一定期間に放送された全番組の内容をカウントし、必要に応じて正規化して、標準嗜好ベクトルAPPとするようにしても良い。
【０２９３】
配信サーバ５においては、例えば、一般的な視聴率調査や、その他の方法を用いて、一般的な嗜好を示す標準嗜好ベクトルAPPを生成するようにしても良い。
【０２９４】
ステップＳ１６３において、ユーザ情報登録部６３は、大項目ごとに、標準嗜好ベクトルAPPと、正履歴ベクトルUPとのコサイン距離を算出する。コサイン距離が大きいほど、標準嗜好ベクトルAPPと、正履歴ベクトルUPとの類似度が高い。
【０２９５】
ステップＳ１６４において、ユーザ情報登録部６３は、ステップＳ１６３において算出されたコサイン距離を基に、大項目ごとに、コサイン距離を逆数にして、効き目ベクトルEfUPを生成して、処理が終了される。コサイン距離の逆数が大きいほど、標準嗜好ベクトルAPPと、正履歴ベクトルUPとの類似度は低い。
【０２９６】
このような処理により、一般的な嗜好と、対応するユーザ固有の嗜好との差を反映したユーザ側効き目ベクトルEfUPを求めることができる。このユーザ側効き目ベクトルEfUPを用いて番組推薦処理を行った場合、ユーザの嗜好と一般的な嗜好との違いが強調されて、推薦される番組が決定される。
【０２９７】
なお、ここでは、番組ベクトルPP、および正履歴ベクトルUPが、大項目ごとにベクトルで表されているものとして説明したが、番組ベクトルPP、および正履歴ベクトルUPが、詳細項目の全要素を１列に並べたベクトルで表されている場合においても、同様の処理が実行可能であることは言うまでもない。
【０２９８】
また、標準嗜好ベクトルAPPと、正履歴ベクトルUPとの類似度は、効き目ベクトルを算出するのみならず、そのユーザのユニーク度を示す指数として、番組の推薦に直接的に用いるようにしても良い。例えば、標準嗜好ベクトルAPPと、正履歴ベクトルUPとの類似度が高い場合、一般向けの流行に即した新番組などを優先的に推薦するようにしても良い。
【０２９９】
図２２および図２３を用いて説明したように、ユーザ側効き目ベクトルEfUPは、ユーザの操作履歴に基づいて、学習的に求められると好適であるが、ユーザ側効き目ベクトルは、予め初期登録として登録されていても良いし、経験などにより求められる、予め設定されている値を用いるようにしても良い。
【０３００】
なお、大項目に着目して、ユーザ側効き目ベクトルEfUPを生成するのみならず、大項目を構成する構成要素に着目して、ユーザ側効き目ベクトルEfUPを生成するようにしても良い。例えば、大項目を構成する構成要素「出演者Ｐｍ」において、主役と脇役を区別可能にしておき、ドラマや映画において、主役よりも脇役の配役を優先するユーザは、主役に対して、脇役の重み付けを大きくするようにユーザ側効き目ベクトルEfUPを設定可能としたり、大項目を構成する構成要素「脚本／原作／演出Am」において、監督、演出家、原作者、カメラマンなどを区別可能にしておき、監督や演出家より、カメラマンを重視するユーザは、カメラマンの重み付けを大きくするようにユーザ側効き目ベクトルEfUPを設定可能なようにしても良い。
【０３０１】
また、ユーザ側効き目ベクトルEfUPをジャンルごとに生成し、図２１を用いて説明したマッチング処理５のように、対応するジャンルの正履歴ベクトルUPと番組ベクトルPPとのマッチングのときに作用させるようにしても良い。
【０３０２】
次に、図２４のフローチャートを参照して、一定期間にユーザが視聴した番組を、ジャンル別にカウントすることによりユーザ側効き目ベクトルEfUPを生成する、ユーザ側効き目ベクトル生成処理３について説明する。
【０３０３】
ステップＳ１７１において、マッチング処理部４３のユーザ情報登録部６３は、一定期間にユーザが視聴した番組を、ジャンル別にカウントするために、いずれかのジャンルを選択する。
【０３０４】
ステップＳ１７２において、ユーザ情報登録部６３は、未処理の大項目のうちのいずれかを選択する。
【０３０５】
ステップＳ１７３において、ユーザ情報登録部６３は、正履歴保存部４７に保存されている正履歴を参照して、例えば、１週間、１ヶ月、または３ヶ月などの一定期間中にユーザが視聴した番組のうち、選択されたジャンルのものを検出し、番組ベクトル抽出部４２に、一定期間中にユーザが視聴したもののうち、選択されたジャンルの番組に対応する番組ベクトルPPをデータ取得部４１から抽出させて、ステップＳ１７２において選択された大項目に含まれる詳細項目ごとに番組数をカウントする。
【０３０６】
ステップＳ１７４において、ユーザ情報登録部６３は、番組ベクトル抽出部４２に、同一期間における全ての番組のうち、選択されたジャンルに対応する番組ベクトルPPをデータ取得部４１から抽出させて、ステップＳ１７２において選択された大項目に含まれる詳細項目ごとに番組数をカウントする。
【０３０７】
ステップＳ１７５において、ユーザ情報登録部６３は、ステップＳ１７３およびステップＳ１７４のカウント結果を基に、選択されたジャンルにおける、ユーザの視聴実績のカウント数／全番組のカウント数を算出する。
【０３０８】
上述したように、番組編成は、視聴率競争の影響により、大衆の嗜好を反映していると考えられる。すなわち、選択されたジャンルにおける、ユーザの視聴実績のカウント数／全番組のカウント数の演算は、換言すれば、対応するジャンルにおけるユーザの視聴実績のカウント数を、標準モデルとしての対応するジャンルの全番組のカウント数で正規化することと同義である。ステップＳ１７５において算出される正規化ベクトルを正規化ベクトルＤ´と称するものとする。
【０３０９】
例えば、大項目ジャンルＧｍ＝｛ドラマ，バラエティ，スポーツ，映画，音楽，子供向け／教育，教養／ドキュメント，ニュース／報道，その他｝のうちの「ドラマ」のジャンルに対応する番組ベクトルPPの、大項目時間帯Ｔｍ＝｛朝、昼、夕方、ゴールデン、深夜｝において、１週間の全番組のカウント数が、（１０，３５，７，５３，１７）であり、対してユーザが視聴した番組のカウント数が、（５，０，０，８，４)であった場合、正規化ベクトルＤ´は、以下のようになる。
Ｄ´＝（５／１０，０／３５，０／７，８／５３，４／１７）
＝（０．５，０，０，０．２８，０．１５，０．２４）
すなわち、正規化ベクトルＤ´の構成要素が１．０である場合は、設定された期間中に、該当する項目の全番組を視聴したことを意味し、正規化ベクトルＤ´の構成要素が０である場合は、設定された期間中に、該当する番組は一つも視聴しなかったことを意味する。
【０３１０】
ステップＳ１７６において、ユーザ情報登録部６３は、ステップＳ１７５の算出結果を基に、選択されたジャンルにおける、対応する大項目の効き目ベクトルを生成する。
【０３１１】
効き目ベクトルを生成するために、大項目時間帯Ｔｍ＝｛朝、昼、夕方、ゴールデン、深夜｝のうちのいずれかの項目を用いて、標準値を設定する。例えば、時間帯：「ゴールデン」の番組1週間分において、一般的なユーザは、ドラマの２割程度を視聴しているとして、標準値を０．２に設定すればよい。大項目の効き目ベクトルは、相対値として算出されるため、設定値は０乃至１のいずれの値であってもかまわない。そして、ユーザ側効き目ベクトルは、ステップＳ１７５において算出された正規化ベクトルＤ´と、設定値との相対値である。
【０３１２】
従って、ユーザの興味ジャンルを示す、大項目ジャンルＧｍの効き目ベクトルＥ´は、
Ｅ´＝（０．３，−０．２，−０．２，−０．０５，０．０４）
と算出されるので、対応するユーザは、朝の時間帯のドラマを好み、昼や夕方の時間帯のドラマを好まないと判断することができる。
【０３１３】
ステップＳ１７７において、ユーザ情報登録部６３は、選択されたジャンルにおいて、全ての大項目の効き目ベクトルが生成されたか否かを判断する。ステップＳ１７７において、全ての大項目の効き目ベクトルが生成されていないと判断された場合、処理は、ステップＳ１７２に戻り、それ以降の処理が繰り返される。
【０３１４】
ステップＳ１７７において、全ての大項目の効き目ベクトルが生成されたと判断された場合、ステップＳ１７８において、ユーザ情報登録部６３は、全てのジャンルの処理が終了したか否かを判断する。ステップＳ１７８において、全てのジャンルが終了していないと判断された場合、処理は、ステップＳ１７１に戻り、それ以降の処理が繰り返される。
【０３１５】
ステップＳ１７８において、全てのジャンルが終了したと判断された場合、ステップＳ１７９において、ユーザ情報登録部６３は、全ての大項目の効き目ベクトルを保存して、処理が終了される。
【０３１６】
このような処理により、一般的な嗜好と、ユーザ固有の嗜好との差をジャンル別に求めることができる。また、図２２を用いて説明した場合と同様にして、ユーザ側効き目ベクトルEfUPを、例えば、３ヶ月、半年などの所定の期間ごとに再演算するようにすることにより、ユーザの嗜好をリアルタイムに反映した番組を推薦することができる。
【０３１７】
また、図２４の処理においても、例えば、１週間、１ヶ月、または３ヶ月などの一定期間中にユーザが視聴した番組を基にユーザ側効き目ベクトルEfUPを求めるものとして説明したが、図２２を用いて説明した場合と同様にして、複数の期間を対象として、例えば、短期、中期、長期に対応するユーザ側効き目ベクトルEfUPを算出し、これら複数の効き目ベクトルを用いて、推薦情報を決定するようにしても良い。
【０３１８】
また、図２４の処理においても、放送されている全番組に代わって、視聴者が最も番組を視聴する所定の時間帯（例えば、１８時乃至２２時の、いわゆる、ゴールデンタイム）に放送される全番組をカウントするようにしてもよい。
【０３１９】
次に、図２５のフローチャートを参照して、正履歴ベクトルUPと、大衆の嗜好との類似度を示すコサイン距離の演算を、ジャンル別に実行することによって、ユーザ固有の嗜好と大衆の嗜好との隔たりをマッチングに利用するためのユーザ側効き目ベクトルEfUPを求める、ユーザ側効き目ベクトル生成処理４について説明する。
【０３２０】
ステップＳ１９１において、マッチング処理部４３のユーザ情報登録部６３は、処理を実行するジャンルを指定するために、いずれかのジャンルを選択する。
【０３２１】
ステップＳ１９２において、ユーザ情報登録部６３は、正履歴保存部４７に保存されている正履歴ベクトルUPのうち、選択されたジャンルの正履歴ベクトルUPを取得する。
【０３２２】
ステップＳ１９３において、ユーザ情報登録部６３は、一般的な嗜好を示す、標準嗜好ベクトルAPPのうち、選択されたジャンルの標準嗜好ベクトルAPPを取得する。
【０３２３】
標準嗜好ベクトルAPPは、上述したように、配信サーバ５から供給されるものとしても良いし、または、番組編成は、視聴率競争の影響により、大衆の嗜好を反映していると考えられるので、図２４を用いて説明したユーザ側効き目ベクトル算出処理３と同様にして、一定期間に放送された全番組の内容をジャンル別にカウントし、必要に応じて正規化して、ジャンル別標準嗜好ベクトルAPPとするようにしても良い。
【０３２４】
配信サーバ５においては、例えば、一般的な視聴率調査や、その他の方法を用いて、一般的な嗜好を示す標準嗜好ベクトルAPPをジャンル別に生成するようにしても良い。
【０３２５】
ステップＳ１９４において、ユーザ情報登録部６３は、選択されたジャンルの正履歴ベクトルUPと、選択されたジャンルの標準嗜好ベクトルAPPを用いて、大項目ごとに、標準嗜好ベクトルAPPと、正履歴ベクトルUPとのコサイン距離を算出する。コサイン距離が大きいほど、標準嗜好ベクトルAPPと、正履歴ベクトルUPとの類似度が高い。
【０３２６】
ステップＳ１９５において、ユーザ情報登録部６３は、ステップＳ１９４において算出されたコサイン距離を基に、大項目ごとに、コサイン距離を逆数にして、選択されたジャンルの効き目ベクトルEfUPを生成する。コサイン距離の逆数が大きいほど、標準嗜好ベクトルAPPと、正履歴ベクトルUPとの類似度は低い。
【０３２７】
ステップＳ１９６において、ユーザ情報登録部６３は、全てのジャンルの処理が終了したか否かを判断する。ステップＳ１７８において、全てのジャンルが終了していないと判断された場合、処理は、ステップＳ１９１に戻り、それ以降の処理が繰り返される。ステップＳ１９６において、全てのジャンルが終了したと判断された場合、処理が終了される。
【０３２８】
このような処理により、一般的な嗜好と、対応するユーザ固有の嗜好との差を反映した、ジャンル別のユーザ側効き目ベクトルEfUPを求めることができる。
【０３２９】
また、図２３を用いて説明した処理と同様にして、標準嗜好ベクトルAPPと、負履歴ベクトルMUPとの類似度を求め、その逆数を、反効き目ベクトルEfMUPとして算出するようにしてもよい。
【０３３０】
次に、図２６のフローチャートを参照して、負履歴ベクトルMUPを、大衆の嗜好と比較することにより、反効き目ベクトルEfMUPを求めるユーザ側反効き目ベクトル算出処理１について説明する。
【０３３１】
ステップＳ２０１において、マッチング処理部４３のユーザ情報登録部６３は、負履歴保存部４８に保存されている負履歴ベクトルMUPを取得する。
【０３３２】
ステップＳ２０２において、ユーザ情報登録部６３は、一般的な嗜好を示す、標準嗜好ベクトルAPPを取得する。
【０３３３】
標準嗜好ベクトルAPPは、配信サーバ５から供給されるものとしても良い。あるいは、番組編成は、視聴率競争の影響により、大衆の嗜好を反映していると考えられるので、図２２を用いて説明したユーザ側効き目ベクトル算出処理１と同様にして、一定期間に放送された全番組の内容をカウントし、必要に応じて正規化して、標準嗜好ベクトルAPPとするようにしても良い。
【０３３４】
ステップＳ２０３において、ユーザ情報登録部６３は、大項目ごとに、標準嗜好ベクトルAPPと、負履歴ベクトルMUPとのコサイン距離を算出する。コサイン距離が大きいほど、標準嗜好ベクトルAPPと、負履歴ベクトルMUPとの類似度が高い。
【０３３５】
ステップＳ２０４において、ユーザ情報登録部６３は、ステップＳ２０３において算出されたコサイン距離を基に、大項目ごとに、コサイン距離を逆数にして、反効き目ベクトルEfMUPを生成して、処理が終了される。
【０３３６】
このような処理により、反効き目ベクトルEfMUPを生成することができるので、ユーザが好まない番組を、推薦する番組から効果的に省くことができる。
【０３３７】
次に、図２７のフローチャートを参照して、ジャンル別に実行されるユーザ側反効き目ベクトル生成処理２について説明する。
【０３３８】
ステップＳ２１１において、マッチング処理部４３のユーザ情報登録部６３は、処理を実行するジャンルを指定するために、いずれかのジャンルを選択する。
【０３３９】
ステップＳ２１２において、ユーザ情報登録部６３は、負履歴保存部４８に保存されている負履歴ベクトルMUPのうち、選択されたジャンルの負履歴ベクトルMUPを取得する。
【０３４０】
ステップＳ２１３において、ユーザ情報登録部６３は、一般的な嗜好を示す、標準嗜好ベクトルAPPのうち、選択されたジャンルの標準嗜好ベクトルAPPを取得する。
【０３４１】
ステップＳ２１４において、ユーザ情報登録部６３は、選択されたジャンルの負履歴ベクトルMUPと選択されたジャンルの標準嗜好ベクトルAPPとを基に、大項目ごとに、標準嗜好ベクトルAPPと、負履歴ベクトルMUPとのコサイン距離を算出する。コサイン距離が大きいほど、標準嗜好ベクトルAPPと、負履歴ベクトルMUPとの類似度が高い。
【０３４２】
ステップＳ２１５において、ユーザ情報登録部６３は、ステップＳ２１４において算出されたコサイン距離を基に、大項目ごとに、コサイン距離を逆数にして、反効き目ベクトルEfMUPを生成する。コサイン距離の逆数が大きいほど、標準嗜好ベクトルAPPと、負履歴ベクトルMUPとの類似度が低い。
【０３４３】
ステップＳ２１６において、ユーザ情報登録部６３は、全てのジャンルの処理が終了したか否かを判断する。ステップＳ２１６において、全てのジャンルが終了していないと判断された場合、処理は、ステップＳ２１１に戻り、それ以降の処理が繰り返される。ステップＳ２１６において、全てのジャンルが終了したと判断された場合、処理が終了される。
【０３４４】
このような処理により、反効き目ベクトルEfMUPを、ジャンル別に生成することができるので、ユーザが好まない番組を、推薦する番組から効果的に省くことができる。
【０３４５】
なお、ユーザ側効き目ベクトルEfUPと、反効き目ベクトルEfMUPは、図２３、および、図２５乃至図２７を用いて説明した大項目ごとのコサイン距離の逆数をｎ倍した値、あるいは、所定の桁で四捨五入した値を用いるものとしてもかまわないし、１からコサイン距離の逆数を減算した値、あるいは、その値をｎ倍した値を用いるものとしてもかまわない。
【０３４６】
ここでは、番組ベクトルPP、および負履歴ベクトルMUPが、大項目ごとにベクトルで表されているものとして説明したが、番組ベクトルPP、および負履歴ベクトルMUPが、詳細項目の全要素を１列に並べたベクトルで表されている場合においても、同様の処理が実行可能であることは言うまでもない。
【０３４７】
また、番組ベクトル抽出部４２によって抽出される番組ベクトルには、図７または図８を用いて説明した処理により生成された、クループＩＤ、またはクラスタコードなどのグループを示す情報が付加されている場合がある。
【０３４８】
例えば、ユーザが好んで視聴する番組が、連続ドラマであった場合、該当する連続ドラマは、全ての回において推薦されるようにすれば、推薦のための演算処理を削減することができる。また、ユーザに非常に好まれている番組と同グループの番組を優先的に推薦するようにすることによっても、推薦のための演算処理を削減することができる。
【０３４９】
次に、図２８のフローチャートを参照して、クループ推薦を含むマッチング処理について説明する。
【０３５０】
ステップＳ２２１において、マッチング処理部４３のベクトル演算部６２は、番組ベクトル抽出部４２から供給された番組ベクトルPPに、クループＩＤ、またはクラスタコードなどのグループを示す情報が付加されているか否かを基に、対応する番組は、グループ化されているか否かを判断する。
【０３５１】
ステップＳ２２１において、対応する番組はグループ化されていると判断された場合、ステップＳ２２２において、ベクトル演算部６２は、クループＩＤ、またはクラスタコードで示されるグループは、推薦優先グループとして、ユーザ情報登録部６３に登録されているか否かを判断する。
【０３５２】
ステップＳ２２２において、推薦優先グループとして登録されていると判断された場合、ステップＳ２２３において、ベクトル演算部６２は、対応する番組の情報を、推薦情報として、推薦情報出力部４９に供給する。推薦情報出力部４９は、推薦番組リスト５０に推薦する番組を登録するとともに、テレビジョン表示装置１１、または、録画再生装置１２に、推薦する番組の情報を出力して、処理は、後述するステップＳ２２７に進む。
【０３５３】
ステップＳ２２１において、対応する番組はグループ化されていないと判断された場合、または、ステップＳ２２２において、推薦優先グループとして登録されていないと判断された場合、ステップＳ２２４において、図１７乃至図２１を用いて説明した、マッチング処理１乃至マッチング処理５のいずれかの処理が実行される。
【０３５４】
ステップＳ２２５において、ベクトル演算部６２は、ステップＳ２２４において実行されたマッチング処理１乃至マッチング処理３のいずれかの処理において推薦された番組の番組ベクトルに、クループＩＤ、またはクラスタコードなどのグループを示す情報が付加されているか否かを基に、推薦された番組は、グループ化されているか否かを判断する。ステップＳ２２５において、推薦された番組は、グループ化されていないと判断された場合、処理は終了される。
【０３５５】
ステップＳ２２５において、推薦された番組は、グループ化されていると判断された場合、ステップＳ２２６において、ベクトル演算部６２は、番組ベクトルに付加しているグループＩＤ、またはクラスタコードを、推薦優先グループとして、ユーザ情報登録部６３に登録して保存する。
【０３５６】
ステップＳ２２３、または、ステップＳ２２６の処理の終了後、ステップＳ２２７において、ベクトル演算部６２は、負履歴保存部４８に保存されている負履歴の操作ログを参照して、ステップＳ２２３の処理により推薦された番組、または、ステップＳ２２４の処理により推薦された番組のうち、グループ化されているので、ステップＳ２２６の処理において推薦優先グループとして登録された番組に対して、例えば、推薦した番組が受け入れられずにその他の番組の視聴の視聴または録画を指令する操作入力や、自動録画された番組の再生前の消去を指令する操作入力など、負の履歴となる操作入力を受けたか否かを判断する。ステップＳ２２７において、負の履歴となる操作入力を受けていないと判断された場合、処理が終了される。
【０３５７】
ステップＳ２２７において、負の履歴となる操作入力を受けたと判断された場合、ステップＳ２２８において、ベクトル演算部６２は、ユーザ情報登録部６３の推薦優先グループの登録から、対応するグループをはずして、処理が終了される。
【０３５８】
なお、図９を用いて説明したタイトルグループ化処理１により、１つの番組ベクトルPPに複数のグループＩＤが関連付けられている場合もある。特にこのような場合には、ステップＳ２２８において、負の履歴となる操作入力により、ユーザ情報登録部６３の推薦優先グループの登録から、全てのグループＩＤがはずされるようにしても良いし、負の履歴となる操作入力の回数を蓄積するようにし、所定の数だけ負の履歴となる操作入力が行われた場合に、ユーザ情報登録部６３の推薦優先グループの登録から、対応するグループＩＤがはずされるようにしても良い。また、１つの番組ベクトルPPに1つののグループＩＤが関連付けられている場合においても同様である。
【０３５９】
このような処理により、ユーザが好んで視聴している連続ドラマを全ての回において推薦されるようにしたり、ユーザに非常に好まれている番組と同グループの番組を推薦するようにすることができるので、推薦のための演算処理を削減することができる。
【０３６０】
このように、グループ推薦を含むマッチング処理が行われている場合、テレビジョン表示装置１１または録画再生装置１２から取得されるユーザの操作履歴を基に生成される正履歴ベクトルUPおよび負履歴ベクトルMUPは、対応するグループごとに生成されるようにしてもよい。
【０３６１】
また、テレビジョン表示装置１１または録画再生装置１２から取得されるユーザの操作履歴のうちの正履歴を基に、同一の連続ドラマにおける視聴または録画予約の回数をカウントし、所定回数以上になった場合に、マッチング処理を特に行うことなく、優先的に推薦するようにしてもかまわない。
【０３６２】
また、ユーザ履歴を用いて番組の推薦処理を行うのではなく、ユーザが予め指定したトピックを基にユーザモデルを設定し、ユーザモデルを基に番組の推薦処理を行うようにしてもよい。
【０３６３】
ユーザモデルは、所定の番組群から、ユーザが予め指定したトピックを用いてフィルタリングを行うことにより、トピックに対応する番組を抽出することで得られる。初期登録保存部４５は、ユーザにより入力されたトピックを保存する。登録されるトピックは複数であってもかまわないし、ユーザの操作入力によって、適宜、更新される。ベクトル演算部６２は、番組ベクトル抽出部４２から供給される、ユーザモデル作成対象の番組の番組ベクトルにおいて、初期登録保存部４５に保存されているトピックを含む番組ベクトルを抽出し、抽出された番組ベクトルの総和を、必要に応じて正規化することにより、ユーザモデルベクトルを生成し、ユーザ情報登録部６３に登録する。
【０３６４】
これにより、例えば、トピックとして、時間帯Ｈｍ＝「２３時以降」およびジャンルＧｍ＝「バラエティ」が指定された場合、これらのトピックを用いてフィルタリングが行われ、ユーザモデル「深夜のバラエティ」が生成される。ユーザモデル「深夜のバラエティ」には、２３時以降の深夜枠に放送されるバラエティ番組に出演したコメディアンが構成要素として含まれるので、例えば、ドラマなどの、バラエティとは異なるジャンルの番組において、対応するコメディアンが出演する番組がマッチング処理において抽出され、推薦されるので、項目ごとにマッチング処理を行う場合と比較して、ユーザの嗜好を、ジャンルなどを超越して適用することが可能となる。
【０３６５】
なお、ユーザモデル作成対象の番組は、例えば、所定期間の全番組であっても、所定の時間帯（例えば、いわゆるゴールデンタイム）に放送された番組群であってもよい。
【０３６６】
また、例えば、同じフィルタリング条件を用いて、異なる番組集合を対象にフィルタリングを行うことによって、例えば、異なる時期、あるいは、異なる時間帯などの、異なる番組編成に詳細に対応したユーザモデルベクトルを生成することができる。
【０３６７】
具体的には、同じ「ジャンルＧｍ＝音楽」という条件でも、現在の番組編成と、１０年前の番組編成では、放送時刻や出演者に違いがあるので、「現在の音楽ファン」「１０年前の音楽ファン」という異なるユーザモデルを生成することができる。これにより、「１０年前の音楽ファン」が好んで聞いていたポップス歌手が、現在、俳優として出演しているドラマや映画などを推薦したり、または、「１０年前の音楽ファン」が好んで聞いていた楽曲を、現在の歌手が歌うような番組を推薦することが可能となる。
【０３６８】
また、例えば、主に社会人が視聴するような、２０時乃至２４時などの時間帯と、主に子供が視聴するような、１５時乃至２０時などの時間帯とでは、同じ「ジャンルＧｍ＝音楽」という条件でも、フィルタリングにより異なる番組群を得ることができるので、異なるユーザモデルを生成することができる。
【０３６９】
そして、ベクトル演算部６２は、番組ベクトル抽出部４２から供給される番組の番組ベクトルPPと、ユーザ情報登録部６３に登録されているユーザモデルベクトルとの類似度を演算し、その結果を基に、推薦情報を生成し、推薦情報出力部４９に供給する。推薦情報出力部４９は、推薦情報を、推薦番組リスト５０に登録するとともに、テレビジョン表示装置１１、または、録画再生装置１２に供給する。
【０３７０】
次に、図２９のフローチャートを参照して、ユーザモデルを用いたマッチング処理について説明する。
【０３７１】
ステップＳ２３１において、マッチング処理部４３のベクトル演算部６２は、番組ベクトル抽出部４２において抽出された、ユーザモデル作成対象の番組の番組ベクトルを取得する。ユーザモデル作成対象の番組とは、例えば、１ヶ月または３ヶ月などの所定の期間に放送された番組であっても、それら所定の期間中の所定の時間帯（例えば、ゴールデンタイム）に放送された番組であっても良いし、あるいは、１０年前、２０年前などの、過去の所定の期間に放送された番組であっても良い。
【０３７２】
ステップＳ２３２において、初期登録保存部４５は、操作入力部４４より、ユーザにより入力されたトピックを取得し、保存する。ベクトル演算部６２は、初期登録保存部４５からトピックを読み出して、フィルタリング条件として設定する。フィルタリング条件は、例えば、ハッシュテーブルとして記載されるようにしても良い。
【０３７３】
ステップＳ２３３において、ベクトル演算部６２は、ステップＳ２３１において取得された番組ベクトルPPを、ステップＳ２３２において設定されたフィルタリング条件を基にフィルタリングし、トピックに合致する番組ベクトルPPを抽出する。例えば、ユーザモデル作成対象の番組が、過去３ヶ月間の全番組である場合、ベクトル演算部６２は、過去３ヶ月間の全番組ベクトルPPに対して「タイトルＴｍ、または、ジャンルＧｍ＝サッカー」という条件でフィルタリングを施す。
【０３７４】
ステップＳ２３４において、ベクトル演算部６２は、ステップＳ２３３においてフィルタリングによって抽出された番組ベクトルPPの総和を求めて、ユーザモデルベクトルとする。例えば、「タイトルＴｍ、または、ジャンルＧｍ＝サッカー」という条件でフィルタリングが施された場合、「サッカーファン」というユーザモデルのユーザモデルベクトルが生成される。
【０３７５】
ステップＳ２３５において、ベクトル演算部６２は、ステップＳ２３４において生成されたユーザモデルベクトルと、番組ベクトル抽出部４２において抽出された、推薦情報を求めるための候補となる（この後放送される予定である）番組の番組ベクトルPPとのコサイン距離を算出する。
【０３７６】
ステップＳ２３６において、ベクトル演算部６２は、ステップＳ２３５において算出されたコサイン距離を基に、ユーザモデルベクトルと推薦情報を求めるための候補となる番組の番組ベクトルPPとの類似度を比較し、比較結果を基に、類似度が高いもののうち、上位から所定の数の番組を推薦情報として抽出し、推薦情報出力部４９に供給する。推薦情報出力部４９は、推薦情報を、推薦番組リスト５０に登録するとともに、テレビジョン表示装置１１、または、録画再生装置１２に供給して、処理が終了される。
【０３７７】
このような処理により、ユーザの操作履歴がない場合においても、ユーザが設定したトピックに合致した番組を推薦することができる。また、従来においては、例えば、「タイトルＴｍ、または、ジャンルＧｍ＝サッカー」という条件を設定してマッチング処理を行った場合、サッカー選手が出演するバラエティやドラマなどを抽出することができなかった。これに対して、図２９を用いて説明した処理においては、「タイトルＴｍ、または、ジャンルＧｍ＝サッカー」というトピックでユーザモデルを設定することにより、ユーザモデルベクトルの出演者や内容などの項目にも、サッカー選手名が多く含まれるために、たとえ「タイトルＴｍ、または、ジャンルＧｍ＝サッカー」という条件に合致していない番組であっても、サッカー選手が出演するバラエティやドラマなどを抽出して推薦することが可能となる。
【０３７８】
ここでは、ユーザモデルベクトルと、番組ベクトルとのコサイン距離を算出することにより、類似度を求めるものとして説明しているが、大項目ごとに、個別にコサイン距離を算出し、その合計を用いて類似度を求めるようにしても良い。
【０３７９】
なお、ユーザモデルベクトルの生成処理は、配信サーバ５において実行されるようにしてもよい。その場合、例えば、図２を用いて説明した番組ベクトル生成部２３が生成した番組ベクトルを用いて、図２９のステップＳ２３１乃至ステップＳ２３４の処理が実行されるようにすれば良い。
【０３８０】
また、ユーザの嗜好にあった番組を推薦するばかりでなく、正履歴ベクトルUPおよび負履歴ベクトルMUPのいずれとも類似度が低いものを選ぶと、好きでも嫌いでもない、すなわち、これまでユーザが視聴したことのない特徴を有する番組、換言すれば、食わず嫌いの番組が推薦される可能性が高くなる。このように、ユーザの嗜好を抽出して学習を進めるためには、これまでユーザが視聴したことのない特徴を有する番組も評価してもらうことが重要である。
【０３８１】
これにより、ユーザに対して提供する推薦結果に意外性を与え、ユーザの興味の拡大につなげることが可能になるばかりでなく、よりユーザの嗜好に合致した番組を推薦するために、非常に重要な履歴情報を取得することが可能となる。
【０３８２】
次に、図３０のフローチャートを参照して、例外推薦番組選択処理について説明する。
【０３８３】
ステップＳ２４１において、マッチング処理部４３のベクトル演算部６２は、番組ベクトル抽出部４２から供給された番組ベクトルPPと、正履歴保存部４７に保存されている正履歴ベクトルUP、および、負履歴保存部４８に保存されている負履歴ベクトルMUPとを用いて、大項目ごとに、正履歴ベクトルUPと番組ベクトルPPとのコサイン距離、および、負履歴ベクトルMUPと番組ベクトルPPとのコサイン距離を、それぞれ算出する。
【０３８４】
ステップＳ２４２において、ベクトル演算部６２は、項目ごとに算出されたコサイン距離の値を、正履歴側と負履歴側で、それぞれ、合計する。すなわち、ステップＳ２４１およびステップＳ２４２の処理により、上述した、正履歴ベクトルUPと番組ベクトルPPとの類似度SimUPおよび負履歴ベクトルMUPと番組ベクトルPPとの類似度SimMUPが算出される。
【０３８５】
ステップＳ２４３において、ベクトル演算部６２は、正履歴ベクトルUP、および、負履歴ベクトルMUPとの類似度の低さを示す例外推薦値を算出する。
【０３８６】
具体的には、例外推薦値は、（１−SimUP）×（１−SimMUP）または（１／SimUP）×（１／SimMUP）によって求めることが可能である。
【０３８７】
ステップＳ２４４において、ベクトル演算部６２は、ステップＳ２４３の算出結果を基に、例外推薦値の高い番組を求め、推薦情報として抽出し、処理が終了される。
【０３８８】
このような処理により、これまでユーザが視聴したことのない特徴を有する番組を抽出して推薦することができるので、ユーザへの推薦番組の選択に、意外性を与え、ユーザの興味の拡大につなげることが可能になるばかりでなく、よりユーザの嗜好に合致した番組を推薦するために、非常に重要な履歴情報を取得することが可能となる。
【０３８９】
図１３乃至図３０を用いて説明した処理により、番組推薦処理装置１０において、推薦情報が生成されて、テレビジョン表示装置１１、または、録画再生装置１２に供給される。
【０３９０】
テレビジョン表示装置１１、または、録画再生装置１２には、テレビジョン受信装置４において受信されて復号された衛星波または地上波の放送信号が供給される。
【０３９１】
テレビジョン表示装置１１は、ユーザの操作入力を基に、テレビジョン受信装置４から供給された放送信号、または、録画再生装置１２から供給された再生データを表示したり、番組推薦処理装置１０から供給された推薦情報を基に、推薦番組情報を表示したり、チャンネルの自動設定を実行する。更に、テレビジョン表示装置１１は、操作ログを番組推薦処理装置１０に供給する。
【０３９２】
また、録画再生装置１２は、ユーザの操作入力を基に、テレビジョン受信装置４から供給された放送信号を録画または録画予約したり、番組推薦処理装置１０から供給された推薦情報を基に、番組を自動録画する。また、録画再生装置１２は、装着された記録媒体、または、内蔵する記録媒体に録画されている番組を再生し、テレビジョン表示装置１１に出力して表示させる。更に、録画再生装置１２は、操作ログを番組推薦処理装置１０に供給する。
【０３９３】
図３１は、テレビジョン受信装置４の構成を示すブロック図である。テレビジョン受信装置４は、デジタル放送用受信装置の標準規格に対応する一般的な受信装置であるものとして説明する。
【０３９４】
衛星波検波部９１は、衛星２を介して送信され、アンテナ３によって受信された衛星波を、テレビジョン表示装置１１、または、録画再生装置１２から供給されたチャンネルを選択する信号を基に選局して検波し、伝送モードに関する制御信号をＴＭＣＣ（Transmission and Multiplexing Configuration Control）復号部９２に、放送信号を復調・復号処理部９３に供給する。
【０３９５】
ＴＭＣＣ復号部９２は、伝送多重制御信号における伝送モード（変調方式、符号化率など）やスロット等の情報の入力を受け、これらの情報を復号し、復調・復号処理部９３に供給する。
【０３９６】
復調・復号処理部９３は、ＴＭＣＣ復号部９２から供給された、伝送モードに関する情報に基づいて、供給された放送信号を、例えば、ＱＰＳＫ（quadri-phase shift keying：４位相変調方式、または、４相ＰＳＫとも称する）方式や、８相ＰＳＫ方式などの方法を用いて復調および復号し、デインタリーバ９４に供給する。
【０３９７】
デインタリーバ９４は、供給された信号をデインタリーブして、誤り訂正処理部９５に供給する。また、デインタリーバ９４は、供給された信号に対して、更に、フレーム分離やデスクランブルの処理を施すようにしても良い。
【０３９８】
誤り訂正処理部９５は、例えば、リードソロモン符号などを用いた誤り訂正処理を行い、ＣＡ（Conditional Access：限定受信）デスクランブル部１０１に供給する。
【０３９９】
地上波検波部９６は、アンテナ３によって受信された地上波を、テレビジョン表示装置１１または録画再生装置１２から供給された、チャンネルを選択する制御信号を基に選局して検波し、伝送モードに関する制御信号をＴＭＣＣ復号部９７に、放送信号を復調・復号処理部９８に供給する。
【０４００】
ＴＭＣＣ復号部９７は、伝送多重制御信号における伝送モード（変調方式、符号化率など）やスロット、ＴＳ等の情報の入力を受け、これらの情報を復号し、復調・復号処理部９８に供給する。
【０４０１】
復調・復号処理部９８は、ＴＭＣＣ復号部９７から供給された、伝送モードに関する情報に基づいて、供給された放送信号を、例えば、ＱＡＭ（quadrature amplitude modulation：直交振幅変調）方式などの方法を用いて復調および復号し、デインタリーバ９９に供給する。
【０４０２】
デインタリーバ９９は、供給された信号をデインタリーブして、ＴＳ（Transport Stream；トランスポートストリーム）再生部１００に供給する。また、デインタリーバ９９は、供給された信号に対して、更に、フレーム分離やデスクランブルの処理を施すようにしても良い。
【０４０３】
ＴＳ再生部１００は、供給された信号を基に、トランスポートストリームを再生し、ＣＡデスクランブル部１０１に供給する。
【０４０４】
ＣＡデスクランブル部１０１は、誤り訂正処理部９５、または、ＴＳ再生部１００から供給された信号を基に、限定受信信号をデスクランブルし、デマルチプレクサ１０２に供給する。
【０４０５】
データ入力部１０３は、ＥＰＧ受信装置９から、ＥＰＧデータの入力を受けたり、ネットワーク８を介して、配信サーバ５より、ストリーミングデータの供給を受け、デマルチプレクサ１０２に供給する。
【０４０６】
デマルチプレクサ１０２は、ＣＡデスクランブル部１０１、または、データ入力部１０３から供給された信号をデマルチプレクスし、音声信号は音声信号復号部１０４に、映像信号は、映像信号復号部１０５に、そして、制御信号やＥＰＧなどのデータは、データ復号部１０６にそれぞれ供給する。
【０４０７】
音声信号復号部１０４は、供給された音声信号を復号し、テレビジョン表示装置１１、または、録画再生装置１２に供給する。映像信号復号部１０５は、供給された映像信号を復号し、テレビジョン表示装置１１、または、録画再生装置１２に供給する。データ復号部１０６は、供給された制御信号やＥＰＧなどのデータを復号し、テレビジョン表示装置１１、または、録画再生装置１２に供給する。
【０４０８】
テレビジョン受信装置４の処理により、受信された衛星波、または地上波、あるいは、配信されたストリーミングデータが、所定の方式により復調および復号されて、テレビジョン表示装置１１、または、録画再生装置１２に供給される。
【０４０９】
図３２は、テレビジョン表示装置１１の構成を示すブロック図である。
【０４１０】
操作入力部１２１は、ユーザからの操作入力を受け、テレビジョン表示装置１１の各部にユーザの操作入力に対応する制御信号を供給するとともに、ユーザの操作内容を操作ログリスト１２２に供給して保存させる。操作ログリスト１２２に保存されているユーザの操作ログは、図１３を用いて説明した番組推薦処理装置１０の操作ログ取得部４６により取得される。また、ユーザの操作入力が、チャンネルの選択であった場合、操作入力部１２１は、入力されたユーザの操作をチャンネル設定部１２３に供給する。
【０４１１】
チャンネル設定部１２３は、操作入力部１２１から供給されたユーザの操作入力を示す信号に基づいて、チャンネルの選択を示す制御信号を生成し、テレビジョン受信装置４に供給する。また、チャンネル設定部１２３は、後述する推薦番組リスト１２８に登録されている推薦情報に基づいて、自動的にチャンネルを設定するために、チャンネルの選択を示す制御信号を生成し、テレビジョン受信装置４に供給する。テレビジョン受信装置４は、制御信号に基づいて、指定されたチャンネルの放送信号を受信する。
【０４１２】
データ入力部１２４は、テレビジョン受信装置４から放送信号の入力を受け、画像処理部１２５に供給する。画像処理部１２５は、供給された放送信号に対して、出力部１２６の画像表示方法に基づいて、画像処理を施し、出力部１２６に供給する。出力部１２６は、例えば、ＣＲＴ（Cathode Ray Tube）、または、ＬＣＤ（Liquid Crystal Display）などの表示デバイスと、スピーカなどの音声出力デバイスで構成され、供給された画像処理後の放送信号の画像信号を表示デバイスに表示し、音声出力デバイスから音声信号を出力する。
【０４１３】
推薦番組リスト取得部１２７は、番組推薦処理装置１０から、推薦情報を取得し、推薦番組リスト１２８に供給する。推薦番組リスト１２８は、供給された推薦情報を登録する。推薦番組リスト１２８に登録された推薦情報は、チャンネル設定部１２３に読み出されたり、推薦番組情報表示制御部１２９に読み出される。
【０４１４】
推薦番組情報表示制御部１２９は、番組の推薦情報をユーザに提示するために、推薦番組リスト１２８から読み出した推薦情報を、画像処理部１２５に出力する。画像処理部１２５は、推薦番組情報表示制御部１２９から供給された推薦情報を、単独で、または、データ入力部１２４から供給された放送信号の画像に重畳させて、出力部１２６に出力して、表示デバイスに表示させる。
【０４１５】
なお、番組推薦処理装置１０から供給される推薦情報を基に、出力部１２６に推薦情報を表示させるか、自動的にチャンネルを設定させるようにするかは、ユーザの設定によって決定されるようにしても良い。
【０４１６】
図３３のフローチャートを参照して、番組推薦処理装置１０から供給される推薦情報を基に、ユーザに番組推薦情報を提示するための推薦情報表示処理について説明する。
【０４１７】
ステップＳ２５１において、推薦番組リスト取得部１２７は、番組推薦処理装置１０の推薦情報出力部４９から出力された推薦情報を取得する。
【０４１８】
ステップＳ２５２において、推薦番組リスト取得部１２７は、取得した推薦情報を、推薦番組リスト１２８に登録する。
【０４１９】
ステップＳ２５３において、推薦番組情報表示制御部１２９は、推薦番組リスト１２８から、例えば、３時間や１日など、現在時刻よりも所定の時間以内に放送される番組の推薦情報を読み出し、推薦する番組のタイトル、内容、放送時間、および、放送チャンネルなどを表示させるための、推薦情報表示用のデータを生成し、画像処理部１２５に供給する。
【０４２０】
ステップＳ２５４において、画像処理部１２５は、供給された推薦情報表示用のデータを、出力部１２６に表示させるための画像処理を実行し、出力部１２６に供給する。推薦情報は、単独で、または、データ入力部１２４から供給された放送信号の画像に重畳されて、画像処理される。
【０４２１】
ステップＳ２５５において、出力部１２６は、画像処理部１２５から供給された推薦情報を表示し、処理が終了される。
【０４２２】
このような処理により、出力部１２６に推薦情報が表示されるので、ユーザは、表示された推薦情報を参照して、視聴する番組を選択することが可能となる。
【０４２３】
次に、図３４のフローチャートを参照して、番組推薦処理装置１０から供給される推薦情報を基に、自動的にチャンネルを設定して、ユーザの嗜好に合致した番組を表示させるための自動チャンネル設定処理について説明する。
【０４２４】
ステップＳ２７１およびステップＳ２７２において、図３３を用いて説明したステップＳ２５１およびステップＳ２５２と同様の処理が実行される。すなわち、番組推薦処理装置１０の推薦情報出力部４９から出力された推薦情報が取得され、取得された推薦情報が、推薦番組リスト１２８に登録される。
【０４２５】
チャンネル設定部１２３は、ステップＳ２７３において、番組推薦リスト９８から、現在時刻に対応する推薦番組の情報を取得し、ステップＳ２７４において、推薦番組の情報を基に、チャンネル設定情報を生成し、テレビジョン受信装置４に出力する。テレビジョン受信装置４は、制御信号に基づいて、指定されたチャンネルの放送信号を受信する。
【０４２６】
ステップＳ２７５において、データ入力部１２４は、テレビジョン受信装置４から指定されたチャンネルの放送信号を取得し、画像処理部１２５に供給する。
【０４２７】
ステップＳ２７６において、画像処理部１２５は、供給された放送信号に、出力部１２６に表示させるための画像処理を実行し、出力部１２６に供給する。
【０４２８】
ステップＳ２７７において、出力部１２６は、画像処理部１２５から供給された、推薦された番組の映像を表示して、音声を出力し、処理が終了される。
【０４２９】
このような処理により、自動的に、ユーザの嗜好に合致した番組が放送されているチャンネルにチャンネル設定される。図３４を用いて説明した自動チャンネル設定処理は、例えば、ユーザが指令したときに実行されるようにしても良い。これにより、ユーザが、どの番組を視聴するかを悩んだ場合などに、適切な番組が放送されている番組にチャンネルを自動的に設定することが可能となる。
【０４３０】
また、図３４を用いて説明した自動チャンネル設定処理は、例えば、２時間など、特に意識してチャンネルを設定せず、ただ放置されていると判断可能なような所定の時間ごとに、あるいは、放置されていると判断可能なような所定の時間、ユーザからの操作入力がなかった場合などに実行されるようにしても良い。
【０４３１】
なお、ユーザが意識して番組を視聴している途中で、自動的にチャンネルが切り替わってしまうことがないように、自動チャンネル設定処理の実行不可のモードを用意して、ユーザが指令することができるようにしても良い。
【０４３２】
図３５は、録画再生装置１２の構成を示すブロック図である。
【０４３３】
操作入力部１４１は、ユーザからの操作入力を受け、録画再生装置１２の各部にユーザの操作入力に対応する制御信号を供給するとともに、ユーザの操作内容を操作ログリスト１４２に供給して保存させる。操作ログリスト１４２に保存されているユーザの操作ログは、図１３を用いて説明した番組推薦処理装置１０の操作ログ取得部４６により取得される。
【０４３４】
録画設定部１４３は、操作入力部１４１から供給されたユーザの操作入力を示す信号に基づいて、または、後述する推薦番組リスト１４９に登録されている推薦情報から、例えば、録画する番組の放送開始時間および放送終了時間、並びに、放送されるチャンネルなど、録画処理を行うために必要な情報を抽出する。録画設定部１４３は、操作入力部１４１から供給されたユーザの操作入力が録画予約である場合、録画処理を行うために必要な情報を録画予約リスト１４４に登録し、操作入力部１４１から供給されたユーザの操作入力が、現在放送されている番組の録画処理である場合、または、後述する推薦番組リスト１４９に登録されている推薦情報を用いて自動録画処理を行う場合、録画処理を行うために必要な情報を、録画制御部１４５に供給する。
【０４３５】
録画制御部１４５は、録画設定部１４３から供給された、録画処理を行うために必要な情報を基に、または、録画予約リスト１４４に登録されている録画予約情報のうち現在時刻に対応する録画予約情報を抽出することにより、録画する番組の放送チャンネルを示す制御信号を生成して、テレビジョン受信装置４に供給するとともに、録画処理を実行するための制御信号を生成し、録画再生処理部１４７に供給する。テレビジョン受信装置４は、制御信号に基づいて、指定されたチャンネルの放送信号を受信する。
【０４３６】
データ入力部１４６は、テレビジョン受信装置４から放送信号の入力を受け、録画再生処理部１４７に供給する。録画再生処理部１４７は、例えば、磁気テープ、光ディスク、磁気ディスク、光磁気ディスク、または、半導体メモリなどの記録媒体が装着可能なように構成されるか、その内部に、例えば、ハードディスク、または、半導体メモリなどの記録媒体を有し、それらの記録媒体に情報を記録したり、それらの記録媒体から情報を再生可能なようになされている。具体的には、録画再生処理部１４７に装着可能な記録媒体が磁気テープである場合、録画再生処理部１４７は、磁気ヘッドを有し、磁気テープにデータ入力部１４６から供給された放送信号を記録（すなわち、録画）したり、磁気テープに記録されている情報を再生して、テレビジョン表示装置１１などに供給して再生出力させる。
【０４３７】
推薦番組リスト取得部１４８は、番組推薦処理装置１０から、推薦情報を取得し、推薦番組リスト１４９に供給する。推薦番組リスト１４９は、供給された推薦情報を登録する。推薦番組リスト１４９に登録された推薦情報は、録画設定部１４３に読み出されて、自動的に録画処理が実行される。
【０４３８】
次に、図３６のフローチャートを参照して、自動録画処理について説明する。
【０４３９】
ステップＳ２９１において、推薦番組リスト取得部１４８は、番組推薦処理装置１０の推薦情報出力部４９から出力された推薦情報を取得する。
【０４４０】
ステップＳ２９２において、推薦番組リスト取得部１４８は、取得した推薦情報を、推薦番組リスト１４９に登録する。
【０４４１】
ステップＳ２９３において、録画設定部１４３は、番組推薦リスト１０９から、現在時刻に対応する推薦番組の情報を抽出し、放送開始時間および放送終了時間、並びに、放送チャンネルなどの、録画処理のために必要な情報を取得して、録画制御部１４５に供給する。
【０４４２】
ステップＳ２９４において、録画制御部１４５は、録画する番組の放送信号の供給を受けるためのチャンネル設定情報を生成し、テレビジョン受信装置４に出力する。テレビジョン受信装置４は、制御信号に基づいて、指定されたチャンネルの放送信号を受信する。
【０４４３】
ステップＳ２９５において、データ入力部１４６は、テレビジョン受信装置４から、指定されたチャンネルの放送信号を取得し、録画再生処理部１４７に供給する。
【０４４４】
ステップＳ２９６において、録画再生処理部１４７は、供給された放送信号を、装着されている、または、内蔵されている記録媒体に記録して、処理が終了される。
【０４４５】
このような処理により、ユーザの嗜好に合致した番組を、自動的に録画することが可能となる。図３６を用いて説明した自動録画処理は、ユーザが指令した録画処理、または、録画予約処理による録画動作中など、すでに録画動作が行われている間には実行されない。
【０４４６】
ここでは、現在時刻に対応する推薦番組を基に、自動的に録画を行い場合について説明したが、例えば、現在時刻よりも先の所定の時間分の推薦情報を取得して、同様の処理により、自動的に録画予約を設定することができるようにしてもよいことは言うまでもない。
【０４４７】
図１乃至図３６を用いて説明した処理では、番組ベクトルPPが、配信サーバ５において生成されるものとして説明したが、配信サーバが番組ベクトルPPを生成するのではなく、ネットワーク８を介して、ＥＰＧデータを番組推薦処理装置に供給し、番組推薦処理装置において、番組ベクトルPPを生成するようにしても良い。
【０４４８】
配信サーバが番組ベクトルPPを生成するのではなく、ネットワーク８を介して、ＥＰＧデータを番組推薦処理装置に供給し、番組推薦処理装置において、番組ベクトルPPを生成する場合の配信サーバ１７１の構成を図３７に示し、番組推薦処理装置１９１の構成を図３８に示す。
【０４４９】
なお、図２の配信サーバ５、および、図１３の番組推薦処理装置１０における場合と対応する部分には同一の符号を付してあり、その説明は適宜省略する。
【０４５０】
すなわち、配信サーバ１７１は、図２を用いて説明した配信サーバ５のデータ取得部２１およびデータ送信部２５で構成され、ストリーミングデータデータベース６またはメタデータデータベース７からストリーミングデータ、または、メタデータで構成されるＥＰＧデータを取得して、ネットワーク８を介して、ＥＰＧ受信装置８またはテレビジョン受信装置４に送信する。
【０４５１】
そして、番組推薦処理装置１９１は、図２の配信サーバ５に設けられていたものと同様のメタデータ取得部２２および番組ベクトル生成部２３が新たに設けられている以外は、図１３を用いて説明した番組推薦処理装置１０と同様の構成を有するものであり、番組推薦処理装置１０が実行する処理に加えて、図３を用いて説明した番組ベクトル生成処理１、および、図６を用いて説明した番組ベクトル生成処理２、並びに、図７を用いて説明したグループ化処理１、および、図８を用いて説明したグループ化処理を実行する。
【０４５２】
配信サーバが番組ベクトルPPを生成するのではなく、ネットワーク８を介して、ＥＰＧデータを番組推薦処理装置に供給し、番組推薦処理装置において、番組ベクトルPPを生成する場合においても、図１乃至図３６を用いて説明した場合と同様の効果を奏することが可能である。
【０４５３】
また、更に、ＥＰＧ受信装置９が、テレビジョンジョン表示装置１１および録画再生装置１２から、ユーザの操作履歴や設定情報などを収集して、ネットワーク８を介して、配信サーバに供給するようにし、配信サーバが、番組ベクトルPPを生成するのみでなく、マッチング処理を実行し、ネットワーク８を介して、マッチング結果をＥＰＧ受信装置９に供給するようにしてもよい。この場合のネットワーク構成を図３９に示し、配信サーバ２０１の構成を示すブロック図を図４０に示す。
【０４５４】
なお、図１、図３７、および図３８における場合と対応する部分には同一の符号を付してあり、その説明は適宜省略する。
【０４５５】
すなわち、配信サーバ２０１は、図３７を用いて説明した配信サーバ１７１に、図３８を用いて説明した番組推薦処理装置１９１の機能を加えたものであり、ユーザは、番組推薦処理装置を保有する必要がない。
【０４５６】
図３９および図４０に示される構成においては、番組ベクトルPP、番組側効き目ベクトルEfPP、正履歴ベクトルUP、負履歴ベクトルMUP、もしくは、標準嗜好ベクトルAPPなどの生成処理や、番組ベクトルのグループ化処理、マッチング処理、または、例外推薦番組選択処理などの、図３乃至図１２、および、図１４乃至図３０を用いて説明した全ての処理が、配信サーバ２０１で実行される。
【０４５７】
なお、このとき、ＥＰＧ受信装置９が、テレビジョンジョン表示装置１１および録画再生装置１２から収集して、ネットワーク８を介して送信する、ユーザの操作履歴や設定情報などには、たとえば、ユーザＩＤなどが付けられて、ユーザごとに区別可能なようになされる。また、配信サーバ２０１の番組推薦処理装置１９１において、初期登録保存部４５、正履歴保存部４７、負履歴保存部４８などは、供給された情報のユーザＩＤを基に、それぞれの情報をユーザ別に保存する。
【０４５８】
以上の説明においては、テレビジョン放送信号のＥＰＧデータを用いて、ユーザの嗜好にあった番組を推薦する場合について説明したが、本発明は、ラジオ放送、あるいは、ストリーミングデータなどの各種デジタルコンテンツに属性情報を付加し、ユーザの嗜好にあったものを推薦する場合についても適用可能である。
【０４５９】
上述した一連の処理は、ソフトウェアにより実行することもできる。そのソフトウェアは、そのソフトウェアを構成するプログラムが、専用のハードウェアに組み込まれているコンピュータ、または、各種のプログラムをインストールすることで、各種の機能を実行することが可能な、例えば汎用のパーソナルコンピュータなどに、記録媒体からインストールされる。
【０４６０】
この記録媒体は、図２、図１３、または、図３８に示すように、コンピュータとは別に、ユーザにプログラムを提供するために配布される、プログラムが記録されている磁気ディスク３１または７１（フレキシブルディスクを含む）、光ディスク３２または７２（CD-ROM（Compact Disk-Read Only Memory），DVD（Digital Versatile Disk）を含む）、光磁気ディスク３３または７３（ＭＤ（Mini-Disk)（商標）を含む）、もしくは半導体メモリ３４または７４などよりなるパッケージメディアなどにより構成される。
【０４６１】
また、本明細書において、記録媒体に記録されるプログラムを記述するステップは、記載された順序に沿って時系列的に行われる処理はもちろん、必ずしも時系列的に処理されなくとも、並列的あるいは個別に実行される処理をも含むものである。
【０４６２】
なお、本明細書において、システムとは、複数の装置により構成される装置全体を表すものである。
【０４６３】
【発明の効果】
このように、本発明によれば、コンテンツの属性情報を生成することができる。特に、コンテンツの分類に基づいて、属性情報を生成することができる。
また、他の本発明によれば、ユーザの嗜好と合致したコンテンツを選択することができるほか、コンテンツの分類に基づいて、コンテンツが選択されるので、ユーザの嗜好に正しく合致したコンテンツを選択することができる。
更に、他の本発明によれば、ユーザの嗜好と合致したコンテンツを選択することができるほか、所定の条件に合致したコンテンツの属性情報を基に、ユーザの嗜好情報が生成されるので、ユーザの嗜好に正しく合致したコンテンツを選択することができる。
【図面の簡単な説明】
【図１】テレビジョン番組放送、およびストリームデータの配信について説明する図である。
【図２】図１の配信サーバの構成を示すブロック図である。
【図３】番組ベクトル生成処理１について説明するフローチャートである。
【図４】ＥＰＧデータを説明する図である。
【図５】番組ベクトルについて説明する図である。
【図６】番組ベクトル生成処理２について説明するフローチャートである。
【図７】グループ化処理１について説明するフローチャートである。
【図８】グループ化処理２について説明するフローチャートである。
【図９】タイトルグループ化処理１について説明するフローチャートである。
【図１０】タイトルグループ化処理２について説明するフローチャートである。
【図１１】タイトルグループ化処理３について説明するフローチャートである。
【図１２】タイトルグループ化処理４について説明するフローチャートである。
【図１３】図１の番組推薦処理装置の構成を示すブロック図である。
【図１４】正履歴ベクトルおよび負履歴ベクトル生成処理１について説明するフローチャートである。
【図１５】正履歴ベクトルを説明する図である。
【図１６】正履歴ベクトルおよび負履歴ベクトル生成処理２について説明するフローチャートである。
【図１７】マッチング処理１について説明するフローチャートである。
【図１８】マッチング処理２について説明するフローチャートである。
【図１９】マッチング処理３について説明するフローチャートである。
【図２０】マッチング処理４について説明するフローチャートである。
【図２１】マッチング処理５について説明するフローチャートである。
【図２２】ユーザ側効き目ベクトル生成処理１について説明するフローチャートである。
【図２３】ユーザ側効き目ベクトル生成処理２について説明するフローチャートである。
【図２４】ユーザ側効き目ベクトル生成処理３について説明するフローチャートである。
【図２５】ユーザ側効き目ベクトル生成処理４について説明するフローチャートである。
【図２６】ユーザ側反効き目ベクトル生成処理１について説明するフローチャートである。
【図２７】ユーザ側反効き目ベクトル生成処理２について説明するフローチャートである。
【図２８】グループ推薦を含むマッチング処理について説明するフローチャートである。
【図２９】ユーザモデルを用いたマッチング処理について説明するフローチャートである。
【図３０】例外推薦処理について説明するフローチャートである。
【図３１】図１のテレビジョン受信装置の構成を示すブロック図である。
【図３２】図１のテレビジョン表示装置の構成を示すブロック図である。
【図３３】推薦情報表示処理について説明するフローチャートである。
【図３４】自動チャンネル設定処理について説明するフローチャートである。
【図３５】図１の録画再生装置の構成を示すブロック図である。
【図３６】自動録画処理について説明するフローチャートである。
【図３７】配信サーバの異なる構成例を示すブロック図である。
【図３８】番組推薦処理装置の異なる構成例を示すブロック図である。
【図３９】テレビジョン番組放送、およびストリームデータの配信のためのネットワークの異なる例について説明する図である。
【図４０】配信サーバの異なる構成例を示すブロック図である。
【符号の説明】
５ 配信サーバ， １０ 番組推薦処理装置， ２１ データ取得部， ２２メタデータ抽出部， ２３ 番組ベクトル生成部， ２４ データ記憶部， ２５ データ送信部， ４１ データ取得部， ４２ 番組ベクトル抽出部， ４３ マッチング処理部， ４４ 操作入力部， ４５ 初期登録保存部， ４６ 操作ログ取得部， ４７ 正履歴保存部， ４８ 負履歴保存部， ４９ 推薦情報出力部， ５０ 推薦番組リスト， ６１ 正規化処理部， ６２ ベクトル演算部， ６３ ユーザ情報登録部， １７１ 配信サーバ， １９１ 番組推薦処理装置， ２０１ 配信サーバ[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an information processing apparatus and an information processing method, an information processing system, a recording medium, and a program. In particular, for a user who watches a program such as a television broadcast or a radio broadcast, or streaming data, for example, The present invention relates to an information processing apparatus, an information processing method, an information processing system, a recording medium, and a program that are suitable for use in recommending a recording or a program.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, for example, when recommending a program to a user in a television broadcast, a radio broadcast, or the like, a user's program is recommended based on program information (program metadata) such as an electronic program guide (EPG). A program matching the preference information was selected. The method of recommending a program to a user differs depending on the method of acquiring user preference data. For example, an initial interest registration method in which information related to a user's interest is initially registered, and a program is recommended based on the information, There is a method of using a viewing history using a history of programs viewed in the past to recommend a program, or a collaborative filtering method of recommending a program using viewing histories of other users.
[0003]
In the initial interest registration method, at the start of use, for example, a user registers a favorite program category (for example, drama, variety, etc.), a favorite genre (inference, comedy, etc.), or a favorite talent name to the user. The name of the program to be recommended is obtained by performing matching with the program metadata using these pieces of information as keywords.
[0004]
Further, in the viewing history utilization method, each time a user views a program, the viewed program metadata is accumulated, and when the history metadata is accumulated to some extent, the metadata is analyzed and, for example, Information such as a program category, a favorite genre, or a favorite talent name is acquired. Then, by using these pieces of information as keywords and performing matching with the program metadata, a program name to be recommended is obtained.
[0005]
For example, in a recording device using an HDD (Hard Disk Drive), not only the viewing history, but also a user operation such as recording reservation or recording is accumulated as history information and may be used to acquire preference information. . In this case, the user is not watching the program with particular interest, but is watching the program because he / she has a television receiver or a radio. , It is possible to partially distinguish programs that have been consciously viewed, and it is possible to acquire information that more reflects the user's preferences.
[0006]
Then, the collaborative filtering method matches the viewing history of the first user with the viewing history of the other user by matching the viewing history of the first user with the viewing history of the other user. By searching for the user and acquiring the viewing / operation history data, a program that has been viewed by the second user and has not yet been viewed by the first user is extracted and recommended.
[0007]
Also, an n-dimensional attribute vector is added to the broadcast program in advance as program attribute information, and the average value of each attribute item of the user's initial registration contents and the attribute vector of the program reproduced or recorded by the user is set. There is a technique in which a program to be recorded or a program to be reproduced can be selected by comparing a selection vector generated based on the attribute vector with an attribute vector (for example, Patent Document 1).
[0008]
[Patent Document 1]
JP 2001-16095A
[0009]
[Problems to be solved by the invention]
However, when selecting a program based on the initial registration method, only a fixed interest of the user at the time of the initial registration can be reflected, and the registration operation of the user becomes complicated in order to obtain detailed information. Would. On the other hand, if the number of registered information is reduced to simplify the initial registration information input operation, the user can only make recommendations based on rough user preference information. The accuracy with which a program matching the preference is selected is low.
[0010]
On the other hand, when a program is selected by simply using the total or average of metadata collected based on the user's viewing history or the like, such as a history utilization method, if the history is not accumulated to some extent, A program that correctly matches the user's preference cannot be recommended. Furthermore, in the history using method, the relationship between the metadata is lost, and sufficient personalization cannot be performed. Further, by stacking the histories, for example, items that are likely to be overlapped as histories (items that are easily detected as being preferred by the user), such as genres, and items that have elements that are likely to be spread as histories, such as performers (detected as being preferred by the user) (Items that are difficult to perform), there is a case where the weighting is biased.
[0011]
Specifically, for example, since the user is a fan of the commentator A, the genre is “genre” if the user likes to watch “live broadcast of the game of the team B explained by the commentator A”. The information "live broadcast of baseball" tends to overlap as a history (the information "live broadcast of baseball" is likely to be detected as a preference of the user), but the information "commentator A" who is a performer overlaps as a history. Difficult (the information “explanator A” is hard to be detected as a preference of the user). For this reason, a case may occur in which a live broadcast of the game of the baseball team B being commented on by another commentator is recommended, and the variety program in which the commentator A appears is not recommended.
[0012]
Further, as shown in Patent Document 1, an attribute vector of a program is added to a broadcast program in advance, and the average content of the attribute items of the user's initial registration contents and the attribute vector of a program reproduced or recorded by the user is added. By comparing the selection vector generated based on the value with the attribute vector to select a program to be recorded or a program to be reproduced, the operation history of the user is used. In some cases, weighting may be biased between items that are likely to overlap as histories and items whose elements are easily spread as histories, such as performers.
[0013]
Furthermore, for example, in a selection vector generated for a user who likes a drama and likes only those in which comedian A who rarely appears in the drama and who has a variety-to-drama viewing ratio of 2: 8, However, a performer B who frequently appears in a drama, even though he is not a favorite actor more than a comedian A who rarely appears in a drama, overlaps as a history. Therefore, the documentary in which the performer B who frequently appears in the drama appears in preference to the variety in which the comedian A appears.
[0014]
In addition, even when important items for selecting a program differ depending on the user (for example, emphasis on performers or content), all items are calculated in the same manner, so that user-specific preferences are not reflected. was there.
[0015]
Further, since only the preference information of other users is used in the collaborative filtering method, it has been difficult to extract information indicating the preference of each user in detail.
[0016]
The present invention has been made in view of such a situation, and it is an object of the present invention to be able to select a program that matches a user's preference.
[0017]
[Means for Solving the Problems]
A first information processing apparatus according to the present invention includes: an obtaining unit that obtains information about a content; a classifying unit that classifies the content based on the information about the content obtained by the obtaining unit; and an attribute based on the information about the content. And an attribute information generating means for generating information and associating the classification result by the classification means.
[0018]
The information processing apparatus may further include a transmission unit configured to transmit the attribute information generated by the attribute information generation unit and the classification result associated with the attribute information to another information processing apparatus.
[0019]
The classifying unit can detect the content that matches a predetermined condition based on the information on the content, and the attribute information generating unit includes unique information for distinguishing the classification by the classifying unit. , Attribute information.
[0020]
The predetermined condition may include a condition that a difference between start times of broadcasting or distribution of the content is within a predetermined time.
[0021]
The information related to the content can be constituted by a plurality of items, and the classification unit is configured to detect a content including at least one identical word in a predetermined item of the information related to the content. The attribute information generating means can associate unique information for distinguishing the classification by the classifying means with the attribute information.
[0022]
The information relating to the content may be constituted by a plurality of items, and the classifying unit may be configured to detect the content in which the matching rate of words constituting the predetermined item of the information relating to the content is greater than a predetermined value. And the attribute information generating means can associate the unique information for distinguishing the classification by the classifying means with the attribute information.
[0023]
The attribute information generating means may further associate the attribute information already associated with the second content of the same classification as the first content with the first content.
[0024]
The content can be a broadcast program, and the classifying means can classify continuously broadcast programs as one classification based on information on the content, and generate attribute information. The means can generate attribute information of a broadcast program classified as a broadcast program to be continuously broadcast by the classifying means using information on contents of the first broadcast.
[0025]
According to a first information processing method of the present invention, there are provided a classification step of classifying a content based on information related to the content, an attribute information generating step of generating attribute information based on the information related to the content, and a process of the attribute information generating step And an associating step of associating the attribute information generated by the above with the classification result by the processing of the classification step.
[0026]
The program recorded on the first recording medium according to the present invention includes: a classification step of classifying the content based on the information on the content; an attribute information generation step of generating attribute information based on the information on the content; It is characterized by including an associating step of associating the attribute information generated by the processing of the information generating step with the classification result by the processing of the classification step.
[0027]
A first program according to the present invention is generated by processing of a classification step of classifying content based on information about content, an attribute information generation step of generating attribute information based on information about content, and an attribute information generation step. And an associating step of associating the attribute information with the classification result obtained by the processing of the classification step.
[0028]
In the first information processing apparatus, the information processing method, and the program according to the present invention, contents are classified based on information about the contents, and attribute information is generated based on the information about the contents. And the classification result.
[0029]
The second information processing apparatus of the present invention includes: an acquiring unit for acquiring attribute information of a content; a storing unit for storing user preference information; a preference information stored by the storage unit; Detecting means for detecting the first content that matches the user's preference by comparing the attribute information with the attribute information, wherein the attribute information of the content includes unique information indicating a classification of the content; , The second content of the same classification as the first content that matches the user's preference is further detected as the content that matches the user's preference.
[0030]
According to a second information processing method of the present invention, a storage control step of controlling storage of unique information indicating a classification of a first content that matches a user's preference, and the attribute information of the second content includes unique information. If it is determined by the first determination step of determining whether or not the attribute information includes the unique information, the storage is controlled by the processing of the storage control step. Based on the unique information of the first content, the second content is subjected to a second determination step of determining whether or not the second content is in the same classification as the first content, and a second determination step is performed. Is determined to be of the same classification as the first content, a detecting step of detecting the second content as content that matches the user's preference.
[0031]
The program recorded on the second recording medium according to the present invention includes a storage control step of controlling storage of unique information indicating a classification of the first content that matches the user's preference; A first determining step of determining whether or not the attribute information includes the unique information; and determining that the attribute information includes the unique information by the processing of the first determining step. A second determining step of determining whether the second content has the same classification as the first content based on the unique information of the controlled first content; and a second determining step. A process of detecting the second content as content that matches the user's preference when the second content is determined to be in the same classification as the first content by the processing. And wherein the Mukoto.
[0032]
A second program according to the present invention includes a storage control step of controlling storage of unique information indicating a classification of the first content that matches a user's preference, and whether the attribute information of the second content includes unique information. When it is determined that the attribute information includes the unique information by the first determination step of determining whether the attribute information includes the unique information, the first storage control is performed by the storage control step. Based on the unique information of the content, the second content is determined by the second determination step of determining whether the second content is in the same classification as the first content and the second determination step. Is characterized by including a detecting step of detecting the second content as the content that matches the user's preference when it is determined that the classification is the same as the first content.
[0033]
In the second information processing apparatus, the second information processing method, and the program according to the present invention, it is determined whether the classification of the first content that matches the user's preference and the classification of the second content are the same. If it is determined that the second content has the same classification as the first content, the second content is detected as content that matches the user's preference.
[0034]
According to an information processing system of the present invention, a first information processing apparatus obtains information related to a content, and a classification that classifies the content based on the information related to the content obtained by the first obtaining means. Means, attribute information generating means for generating attribute information based on information relating to the content, attribute information generated by the attribute information generating means, and the classification result by the classifying means associated with each other and transmitted to another information processing apparatus. A second obtaining means for obtaining attribute information transmitted from the first information processing apparatus, a storing means for storing user preference information, and a storing means. Is compared with the attribute information acquired by the second acquisition unit, thereby detecting the first content that matches the user's preference. Output means, and the attribute information of the content includes unique information indicating a classification of the content, and the detecting means determines, based on the unique information, a second content of the same category as the first content that matches the user's preference. The content is further detected as content that matches the user's preference.
[0035]
In the information processing system according to the present invention, the first information processing apparatus acquires information about the content, classifies the content based on the acquired information about the content, and generates attribute information based on the information about the content. Then, the generated attribute information is associated with the classification result by the classifying unit and transmitted to another information processing device, and the attribute information transmitted from the first information processing device is transmitted by the second information processing device. The acquired preference information of the user is stored, the stored preference information is compared with the acquired attribute information, and a first content that matches the user's preference is detected. And the second content having the same classification as the first content that matches the user's preference based on the unique information. It is further detected as contents matching with THE preferences.
[0036]
A third information processing apparatus according to the present invention includes: an obtaining unit that obtains attribute information of a content; a storage unit that stores a predetermined condition; and a predetermined condition stored in the storage unit of a predetermined content group. Generating means for generating user preference information based on attribute information of the content corresponding to the matched content.
[0037]
Calculating a similarity between the preference information generated by the generation unit and the attribute information of the content acquired by the acquisition unit, thereby generating recommendation information generation unit for generating recommendation information indicating content recommended for viewing by the user; It can be provided.
[0038]
Operation input means for receiving an operation input by the user may be further provided, and the storage means may store a predetermined condition input by the user through the operation input means.
[0039]
The predetermined content group may be a set of contents broadcast or distributed during a predetermined period.
[0040]
The predetermined content group may be a set of contents broadcast or distributed in a predetermined time zone.
[0041]
According to a third information processing method of the present invention, an acquisition step of acquiring attribute information of a content, a setting step of setting a predetermined condition, and a predetermined condition set by the processing of the setting step in a predetermined content group Extracting the content that matches the predetermined condition, and generating the preference information of the user based on the attribute information of the content corresponding to the content that matches the predetermined condition extracted by the processing of the extracting step. It is characterized by.
[0042]
The program recorded on the third recording medium of the present invention includes an acquisition step of acquiring attribute information of a content, a setting step of setting a predetermined condition, and a setting step of setting a predetermined content group. Extracting the content meeting the predetermined condition, and generating the user preference information based on the attribute information of the content corresponding to the content meeting the predetermined condition extracted by the process of the extracting step. And a step.
[0043]
A third program according to the present invention includes: an acquiring step of acquiring attribute information of a content; a setting step of setting a predetermined condition; and a predetermined group of contents which satisfy a predetermined condition set by the processing of the setting step. An extraction step of extracting the extracted content, and a generation step of generating user preference information based on attribute information of the content corresponding to the content meeting predetermined conditions extracted by the processing of the extraction step. And
[0044]
According to the third information processing apparatus, the information processing method, and the program of the present invention, content attribute information is acquired, and a content meeting a predetermined condition is extracted from a predetermined content group. Is generated based on the attribute information of the content corresponding to the content that satisfies the above condition.
[0045]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0046]
With reference to FIG. 1, television program broadcasting and distribution of stream data will be described.
[0047]
The broadcast station 1 transmits a terrestrial program broadcast or a satellite program broadcast via the satellite 2. In FIG. 1, only one broadcasting station 1 is shown and described, but it goes without saying that a plurality of broadcasting stations 1 exist. The antenna 3 of the television receiver 4 receives a terrestrial or satellite program broadcast. The broadcast signal includes an EPG (Electronic Program Guide) as necessary.
[0048]
The distribution server 5 reads the streaming data from the streaming data database 6 and distributes the streaming data to the television receiver 4 via a network 8 including the Internet and other networks, and broadcasts from the broadcasting station 1 from the metadata database 7. EPG or metadata including more detailed information than the EPG is read out, a program vector PP is generated for each program, and the program vector PP is sent to the EPG receiving device 9 via the network 8 together with the EPG data. To deliver.
[0049]
If the information amount of the EPG superimposed on the general broadcast signal is sufficient information for the processing described below, the processing may use the same EPG as the EPG superimposed on the general broadcast signal. You may do it. If the amount of information of the EPG superimposed on the general broadcast signal is not sufficient information for the processing described below, the processing includes the addition of the EPG superimposed on the general broadcast signal. Alternatively, or independently, metadata may be used. Hereinafter, since an EPG superimposed on a general broadcast signal does not have a sufficient amount of information for processing, information including metadata is used in this embodiment, and this information is collectively referred to as EPG data. Shall be.
[0050]
The EPG receiving device 9 supplies the distributed EPG data to the television receiving device 4. Further, the EPG receiving device 9 supplies the program recommendation processing device 10 with the program vector PP distributed together with the EPG data.
[0051]
The television receiver 4 having a tuner receives a terrestrial wave via the antenna 3 based on a control signal indicating selection of a channel supplied from a television display device 11 having an operation unit or a recording / reproducing device 12. Alternatively, a broadcast signal of a satellite wave is selected and received, and streaming data is received from the delivery server 5 via the network 8. The television receiver 4 receives the supply of the EPG data from the EPG receiver 9 and supplies the EPG data to the television display device 11 or the recording / reproducing device 12. When the received broadcast wave includes an EPG, the television receiver 4 separates the broadcast wave from a program signal and supplies the signal to the television display device 11 or the recording / reproducing device 12, respectively.
[0052]
The program recommendation processing device 10 obtains a program vector PP from the EPG receiving device 9 and obtains operation logs from the television display device 11 and the recording / reproducing device 12, and transmits these information or user operation inputs. Based on the information, recommendation information for recommending a program that matches the user's preference is generated and supplied to the television display device 11 and the recording / reproducing device 12.
[0053]
The television display device 11 displays a broadcast signal supplied from the television receiving device 4 or a reproduced signal supplied from the recording / reproducing device 12 based on a user's operation input, Based on the supplied recommendation information, a channel is automatically set or recommended program information is displayed. The television display device 11 supplies an operation log, which is an operation history of the user, to the program recommendation processing device 10.
[0054]
The recording / reproducing apparatus 12 records a broadcast signal supplied from the television receiver 4 on a mounted recording medium or a built-in recording medium (for example, a hard disk) based on a user's operation input, Based on the recommendation information supplied from the recommendation processing device 10, the broadcast signal supplied from the television receiver 4 is automatically recorded on a mounted recording medium or a built-in recording medium. Further, the recording / reproducing apparatus 12 reproduces a program recorded on a mounted recording medium or a built-in recording medium, and outputs the program to the television display device 11 for display. Further, the recording / reproducing device 12 supplies an operation log, which is an operation history of the user, to the program recommendation processing device 10.
[0055]
Here, the EPG receiving device 9, the television receiving device 4, the program recommendation processing device 10, the television display device 11, and the recording / reproducing device 12 have been described as different devices, but these devices are individually configured. It does not have to be done. For example, it goes without saying that the EPG receiving device 9, the television receiving device 4, and the television display device 11 may be integrally configured as a television receiver 15-1 having a built-in tuner function. No. Further, the recording / reproducing apparatus 12 may be integrally configured to be configured as a television receiver 15-2 having a recording function. It goes without saying that the recording / reproducing apparatus 12 may be a so-called hard disk recorder having a large-capacity hard disk as a recording medium. In addition, the program recommendation processing device 10 is built in a television receiver 15-1 having a built-in tuner function, and may be used as a television receiver 15-3 or a television receiver 15-2 having a recording function. The television receiver 15-4 may be incorporated therein.
[0056]
FIG. 2 is a block diagram illustrating a configuration of the distribution server 5.
[0057]
The data acquisition unit 21 acquires data from the metadata database 7 and the streaming data database 6 and supplies the data to the data transmission unit 25 and supplies EPG data to the metadata extraction unit 22. Further, the data acquisition unit 21 performs a process of grouping the EPG data registered in the metadata database 7 according to the contents.
[0058]
The metadata extraction unit 22 extracts data necessary for generating the program vector PP from the EPG data supplied from the data acquisition unit 21 and supplies the data to the program vector generation unit 23. The program vector generating unit 23 generates a program vector PP based on the metadata, and supplies the program vector PP to the data transmitting unit 25 in association with the program side effect vector EfPP stored in the data storage unit 24 as necessary. I do.
[0059]
The data storage unit 24 stores the program side effect vector EfPP, which is information necessary for generating the program vector PP, as necessary.
[0060]
Based on the viewpoint of the production side and the broadcast side of how to gain audience rating, for example, in the case of educational programs, the characteristic that the genre and content elements are strong, and in the case of variety programs, the characteristic that the performer element is strong , Dramas, etc., have the characteristic that performers and screenwriters have strong elements. In order to correctly extract the features of the program, it is necessary to use these characteristics. That is, in the matching processing executed for recommending a program that matches the user's preference among the elements constituting the program vector PP, which element is important depends on the genre of the program.
[0061]
That is, when the genre of the program is “cultural / documentation”, the important items are not the performers but the contents and titles. On the other hand, when the genre of the program is “variety”, the important item is the cast, and when the genre of the program is “drama”, the important items are the cast and the writer. When these are used to generate the program vector PP, a program-side effect vector that defines the degree of contribution in the matching process for each item is set for each program genre, and stored in the data storage unit 24.
[0062]
The data transmission unit 25 transmits the EPG data and the streaming data supplied from the data acquisition unit 21, the program vector PP supplied from the program vector generation unit 23, and the program side effect vector EfPP via the network 8. , EPG receiving device 9 or television receiving device 4.
[0063]
A drive 26 is connected to the program vector generation unit 23 as needed. A magnetic disk 31, an optical disk 32, a magneto-optical disk 33, and a semiconductor memory 34 are mounted on the drive 26 as necessary to exchange data.
[0064]
Next, the program vector generation processing 1 executed by the distribution server 5 will be described with reference to the flowchart of FIG.
[0065]
In step S1, the data acquisition unit 21 receives supply of EPG data composed of metadata from the metadata database 7.
[0066]
In step S2, the metadata extraction unit 22 receives the supply of the EPG data from the data acquisition unit 21, extracts the metadata necessary for generating the program vector PP, and outputs the metadata to the program vector generation unit 23.
[0067]
FIG. 4 shows an example of the metadata. The metadata includes, for example, “movie-Japanese movie” as the genre, “Tokaido Mitani Kaidan” as the title of the movie, data of the broadcast date and distribution source, broadcast date and time, broadcast station name to broadcast, and broadcast The time is included. Further, the metadata includes data such as a director's name, a scriptwriter's name, a photographer (photographer), a musician, a cast, and the like, and a description of a movie as the content of the program.
[0068]
In step S3, the program vector generation unit 23 performs morphological analysis on the titles, contents, and the like included in the metadata as necessary, and decomposes the words into words. More specifically, the program vector generation unit 23 decomposes the movie title included in the metadata into three words of “Tokaido”, “Mitani”, and “Kaidan” as a title. In addition, as shown in FIG. 4, the program vector generation unit 23 includes, in the metadata, as a commentary on the movie, a masterpiece of a Japanese film horror depicting the world of “'Mitani Kaidan” famous for its '59 Shinseiho style beauty' . "Is extracted, the words included in the commentary are extracted, and" Seiho "," Style "," Beauty "," Plenty "," Famous "," Mitani "," Kaidan " , "World", "draw", "Japanese", "horror", "masterpiece" as the content information.
[0069]
In step S4, the program vector generation unit 23 converts each item included in the metadata into a vector, generates a program vector PP, and ends the processing. The generated program vector PP is output to the data transmitting unit 25 and transmitted to the EPG receiving device 9 via the network 8. The vectorization of the items may be performed by arranging all the detailed elements as one column, or may be separated into large items and then vectorized for each large item.
[0070]
When vectorization is performed for each major item, for example, as shown in FIG. 5, the major items are divided into seven items of a title, a genre, a time zone, a broadcast station, a performer, a script / original / production, and a content, and the program is defined as seven items. A vector PP = (Tm, Gm, Pm, Am, Km) is generated. The content of each item is title Tm = {title1, title2,...}, Genre (Genre) Gm = {drama, variety, sports, movie, music, children / education, education / documentation, News / Information, Others｝, Hour Hm = ｛Morning, Lunch, Evening, Golden, Late Night｝, TV Station Sm = ｛NNK General, NNK Education, Asia TV, TTS, Buji, Teleday , Toto, NNK Satellite No. 1, NNK Satellite No. 2, WOWO}, Performer (Person) Pm = {person A, person B, ...} Script / Original / Direction, etc. (Author) Am = {person a, person b,... {Content (Keyword) Km = {kw1, kw2,.
[0071]
Of the seven major items described above, the major items “genre”, “broadcast station”, and “time zone” can be fixed in type, so that detailed elements can be represented by numerical vectors. In the case of a broadcasting station, for example, the broadcasting station Sm = {NNK General, NNK Education, Asia Television, TTS, Buji, TV Japan, Toto, NNK Satellite 1, NNK Satellite 2, WOWO} Is WOWO, the broadcast station vector can be represented by broadcast station Sm = {0, 0, 0, 0, 0, 0, 0, 0, 0, 1}. In the case of the genre, for example, in the genre Gm = {drama, variety, sports, movie, music, children / education, culture / documentation, news / reporting, etc.}, the genre of the corresponding program was culture / documentation. In this case, the genre vector can be represented by the genre Gm = {0, 0, 0, 0, 0, 0, 1, 0, 0}.
[0072]
On the other hand, for items that cannot be represented by numerical vectors, such as the major items “title”, “performers”, “screenplay / original / directed”, and “contents”, the item values are determined by the frequency of words included in the vectors. (Word) and weight (numerical value) are associated with each other. For example, the major item “title” is represented by an element composed of a word and a numerical value, such as title Tm = {Tokaido-1, Mitani-1, Kaidan-1}. In this example, for example, “Tokaido-1” means that the frequency of the word “Tokaido” is “1”.
[0073]
By such processing, the program vector PP is generated and transmitted to the EPG receiving device 9. Specifically, when a program vector PP is generated from the program metadata described with reference to FIG. 4, the program vector PP = {title Tm = {Tokaido-1, Mitani-1, Kaidan-1}, genre Gm = {0,0,0,1,0,0,0,0,0}, time zone Hm = {0,0,0,0,1}, broadcasting station Sm = {0,0,0,0,0} , 0,0,0,0,1}, Performer Pm = {Katsumi Wakasugi-1, Shigeru Amami-1, Toshihiko Emi-1, Ryujiro Nakamura-1, Noriko Nishizawa-1}, Script / Original / Direction Am = {Nobuo Nakagawa-1, Shoichi Onuki-1, Yoshihiro Ishida-1, Shoji Nishimoto-1, Shota Watanabe-1} Content Km = {Seiho-1, Style-1, Beauty-1, Rich-1, Famous-1, Mitani-1, Kaidan-1, World-1, Drawing-1, Japanese film 1, Horror-1, Masterpiece-1 are generated and transmitted to the EPG receiving device 9.
[0074]
In the program vector PP generated in this way, for example, of the components of the performer Pm, the leading role is weighted, and of the components of the script / original / directing Am, the director is weighted, and the program vector PP is It may be generated. In such a case, the components of the program vector PP are, for example, performers Pm = {Kasumi Wakasugi-3, Shigeru Amami-2, Toshihiko Emi-1, Ryujiro Nakamura-1, Noriko Nishizawa-1}, script / original / Direction Am = {Nobuo Nakagawa-3, Shoichi Onuki-1, Yoshihiro Ishida-1, Shoji Nishimoto-1, Shota Watanabe-1}.
[0075]
In addition, which element is important for generating the program vector PP depends on the genre of the program, as described above. When a program side effect vector EfPP indicating what elements are important for each genre is stored in the data storage unit 24 as program side effect vector information, the program side effect vector EfPP is stored in the program vector PP May be transmitted. The program side effect vector EfPP is set corresponding to the major items of the program vector PP.
[0076]
In the case of the program vector PP = {title Tm, genre Gm, time zone Hm, broadcasting station Sm, performer Pm, script / original / directing Am, content Km}, for example, when the genre of the program is “culture / document”, The important items are their contents and titles (when the weight value is 3), so that the program side effect vector EfPP = {3,1,1,1,1,1,3}. On the other hand, when the genre of the program is “variety”, the important item is the performer (when the weight value is 5), so the program-side effect vector EfPP = ｛1,1,1,1,5, 1,1｝. When the genre of the program is "drama", the important items are the cast (when the weight value is 2) and the writer, and especially when the writer is important (when the weight value is 3). , The program side effect vector EfPP = {1, 1, 1, 1, 2, 3, 1}.
[0077]
Next, with reference to the flowchart of FIG. 6, a description will be given of the program vector generation processing 2 when the program side effect vector EfPP is transmitted in association with the program vector PP.
[0078]
In steps S11 to S14, processing similar to the processing in steps S1 to S4 described with reference to FIG. 3 is executed. That is, the EPG data is supplied from the metadata database 7, metadata required for generating the program vector PP is extracted from the EPG data, and the title, contents, etc. included in the metadata are required. Is subjected to morphological analysis and is decomposed into words. Then, each item included in the metadata is vectorized, and a program vector PP is generated.
[0079]
In step S15, the program vector generation unit 23 extracts a corresponding effect vector from the effect vector information stored in the data storage unit 24, based on the genre of the program of the supplied metadata. For example, when the genre of the program is “variety”, an effect vector EfPP = {1,1,1,1,5,1,1} for weighting the cast is extracted from the data storage unit 24.
[0080]
In step S16, the program vector generation unit 23 associates the effect vector EfPP extracted in step S15 with the program vector PP generated in step S14, and ends the processing.
[0081]
By such processing, the program vector PP is generated, and the program side effect vector EfPP for weighting important items is associated with the EPG receiving device via the network 8 based on the genre of the program. 9 is transmitted.
[0082]
As described above, the program vector PP is generated. For example, by grouping programs by their attributes, the program vector PP can be accurately created with a small amount of calculation. .
[0083]
The program grouping can be generated, for example, by a series of drama, a group of programs of the same genre in one-week units, and a group of programs for each performer in a unit of program organization (13 weeks). As a specific example, FIG. The grouping process 1 in the case of serial drama grouping will be described with reference to the flowchart of FIG.
[0084]
In step S31, the data acquisition unit 21 determines from the EPG data registered in the metadata database 7 that, for example, the broadcast time is the same time on weekdays or the same time on the same day of the week for the same title and the same broadcasting station. The program which meets the predetermined condition (the condition for grouping) is extracted.
[0085]
In step S32, the data acquisition unit 21 groups the extracted programs and attaches a group ID to the EPG data of the corresponding program.
[0086]
In step S33, the metadata extraction unit 22 extracts metadata having the same group ID, that is, metadata necessary for generating a program vector of the first broadcast of a program recognized as a serial drama.
[0087]
In step S34, the program vector generation processing described with reference to FIG. 3 or FIG. 6 is executed.
[0088]
In step S35, the program vector generation unit 33 fixes and sets the program vector PP of the corresponding group ID to the program vector for the first broadcast, and the process ends.
[0089]
By such processing, the serial drama is grouped, and the program vector is unified. Further, the group ID and the program vector PP may be stored in the data storage unit 24 in association with each other.
[0090]
In particular, in the EPG data of a serial drama, the contents of the first time are commentary for the whole program, but the contents other than the first time are often commentary for that time, and represent the contents of the whole program. Not. In addition, since the metadata other than the content is almost the same every time, the number of program vector generation processes can be reduced only by generating the program vector PP by using the first-time EPG data. Instead, it is possible to more accurately generate the program vector PP that matches the program characteristics.
[0091]
In FIG. 7, the process of adding the same ID to the EPG data of the serial drama has been described. However, if the EPG data includes information for distinguishing the serial drama in advance, the process of step S31 is omitted. The same ID may be added to the EPG of the serial drama with reference to the information for identifying the serial drama included in the EPG data.
[0092]
Next, with reference to a flowchart of FIG. 8, a grouping process 2 that can perform grouping by elements other than the serial drama will be described. Here, description will be made assuming that grouping is performed by genre and performer.
[0093]
In step S51, the data acquisition unit 21 adds a cluster code corresponding to the content of the metadata to the EPG data of the program with reference to the EPG data registered in the metadata database 7.
[0094]
For example, a program that matches the condition of a serial drama, such as the same title, the same broadcasting station, and the broadcast time is the same time on weekdays or the same time on the same day of the week, is extracted and extracted. For example, according to the broadcast time of a serial drama, code 1 for the same time on weekdays, code 2 for the same time on the same day of the week, code 2 for every other week, every other day, etc. If the broadcast time is other than, the first cluster code of the smallest digit such as code 3 is determined. If it is not a serial drama, the first cluster code of the minimum digit is 0.
[0095]
Next, by referring to data relating to the genre of metadata registered in the metadata database 7, a numerical value is assigned to the second digit (10's place) from the lower end of each genre, for example, code 20, code 30, etc. Is determined.
[0096]
Finally, with reference to the performer data of the metadata registered in the metadata database 7, the data is represented using the second digit or more from the lowermost position (100 or more places) such as code 2300 or code 800. A third cluster code is determined.
[0097]
Then, the total value of the first to third cluster codes becomes the cluster code added to the EPG of the program.
[0098]
In step S52, the data acquisition unit 21 determines whether or not the corresponding EPG is a serial drama based on whether or not the first digit of the cluster code is 0.
[0099]
If it is determined in step S52 that the drama is a serial drama, in steps S53 and S54, the same processing as in steps S33 and S34 in FIG. 7 is performed. That is, the EPG of the first broadcast is extracted, and the program vector generation processing described with reference to FIG. 3 or FIG. 6 is executed.
[0100]
In step S55, the program vector generation unit 33 sets the program vector PP of the serial drama to be fixed to the program vector PP for the first broadcast, and associates the generated cluster code with the program vector PP. After the setting, the process ends.
[0101]
If it is determined in step S52 that the program is not a serial drama, the program vector generation process described with reference to FIG. 3 or FIG. 6 is executed in step S56.
[0102]
In step S57, the program vector generation unit 33 associates the generated cluster code with the program vector PP, and the process ends.
[0103]
By such processing, even if the program is not a serial drama, the program vectors PP can be grouped by genre or performer, and the cluster code indicating the group can be associated with the program vector PP.
[0104]
Since the distribution server 5 described with reference to FIG. 2 generates the program vector PP as described with reference to FIGS. 3 to 8, it is necessary to always generate the program vector PP corresponding to a new term or genre. Becomes possible. The generated program vector PP is received by the EPG receiving device 9 via the network 8 together with the EPG data, and supplied to the program recommendation processing device 10.
[0105]
The grouping process may be performed by morphologically analyzing the title, decomposing the title into words, and attaching a group ID to each word.
[0106]
Next, the title grouping process 1 will be described with reference to the flowchart in FIG.
[0107]
In step S61, the data acquisition unit 21 extracts a title from the registered metadata with reference to the EPG data registered in the metadata database 7, and supplies the title to the program vector generation unit 23.
[0108]
In step S62, the program vector generation unit 23 performs a morphological analysis on the title and breaks it down into words. Specifically, if the title of the movie included in the metadata is “Tokaido Mitani Kaidan”, this will be morphologically analyzed as a title, so the three words “Tokaido”, “Mitani”, and “Kaidan” Is obtained.
[0109]
In step S63, the program vector generation unit 23 extracts one of the analyzed word or a word group composed of a plurality of words, and extracts the extracted word or word from the data storage unit 24. Extract the group ID corresponding to the group.
[0110]
Here, the word group composed of a plurality of words is a word group generated by a combination of words obtained by morphological analysis. For example, words obtained by morphological analysis are “Tokaido”, “Mitani” If the word is "Kaidan", the words are "Tokaido / Kitani", "Tokaido / Kaidan", and "Mitani / Kaidan".
[0111]
In step S64, the program vector generation unit 23 determines whether the corresponding group ID has been extracted from the data storage unit 24.
[0112]
If it is determined in step S64 that the corresponding group ID has not been extracted, the extracted word or a group of words composed of a plurality of words has not yet been assigned a group ID. The program vector generation unit 23 associates a new group ID with the extracted word or a word group including a plurality of words. In addition, the program vector generation unit 23 causes the data storage unit 24 to store a word or a word group including a plurality of words and a corresponding group ID.
[0113]
If it is determined in step S64 that the corresponding group ID has been extracted, or after the processing in step S65 is completed, in step S66, the program vector generation unit 23 sets all the words constituting the title or a plurality of words. It is determined whether a group ID has been extracted for a word group composed of words.
[0114]
If it is determined in step S66 that the group ID has not been extracted for all words constituting the title or for a word group composed of a plurality of words, the process returns to step S63, and the subsequent processes are performed. Repeated.
[0115]
In step S66, when it is determined that the group ID has been extracted for all the words constituting the title or for a word group composed of a plurality of words, in step S67, the program vector generating unit 23 The process is ended by associating the extracted group ID.
[0116]
According to such processing, the program vector is associated with the words constituting the title or the group ID corresponding to the word group, and the data transmitting unit 25 transmits the television receiver 4 or the EPG via the network 8. It is transmitted to the receiving device 9.
[0117]
Also, programs having similar titles may be set to the same group. For example, the title is configured so that the serial drama of the title “Two Years A Gin Ginpachi-sensei” and the special program of the title “Two Years A Gumi Ginpachi-sensei Special” can be grouped as the same group. Based on the result of the morphological analysis of the word, for example, a word match rate is calculated on a round robin basis for a program title for a predetermined period such as two weeks, one month, and six months, and when the word match rate is equal to or more than a predetermined value. In some cases, they may be in the same group.
[0118]
Next, a title grouping process 2 for executing grouping based on the matching rate of words constituting a title will be described with reference to the flowchart of FIG.
[0119]
In steps S401 and S402, the same processing as steps S61 and S62 described with reference to FIG. 9 is executed. That is, the data acquisition unit 21 refers to the EPG data registered in the metadata database 7, extracts a title from the registered metadata, supplies the title to the program vector generation unit 23, and 23 morphologically analyzes the title and breaks it down into words.
[0120]
In step S403, the program vector generation unit 23 calculates a degree of matching between words between titles, that is, a matching rate indicating a rate of matching between words, based on the analyzed words.
[0121]
More specifically, the title “Two-year A-Gumi Ginpachi-sensei” and the title “Two-year A-Gumi Ginpachi-sensei Special” are “2”, “Year”, “A”, “Gumi”, “Ginpachi”, and “ When the morphological analysis is performed on “teacher” and “2” “year” “A” “gumi” “ginpachi” “teacher” “special”, the matching rate of the words constituting the titles of these two programs is 6 / 7 is 85.7%.
[0122]
In step S404, the program vector generation unit 23 determines whether the words match at least a predetermined value such as 70%. It goes without saying that the threshold value of the coincidence rate may be any numerical value other than 70%.
[0123]
If it is determined in step S404 that the words match at least a predetermined value such as 70%, in step S405, the program vector generation unit 23 associates those programs with the same group ID. In addition, the program vector generation unit 23 causes the data storage unit 24 to store the matched word or word group and the corresponding group ID.
[0124]
If it is determined in step S404 that the matching rate is equal to or less than a predetermined value such as 70%, or after the processing in step S405 is completed, in step S406, the program vector generation unit 23 ends the brute force of the title. It is determined whether or not it has been done.
[0125]
If it is determined in step S406 that the round robin of the title has not been completed, the process returns to step S403, and the subsequent processes are repeated.
[0126]
If it is determined in step S406 that the round robin of the title has been completed, the process is terminated.
[0127]
By such processing, the program vector is associated with the group ID based on the matching rate of the words constituting the title, and the data transmitting unit 25 transmits the program to the television receiver 4 or the EPG via the network 8. Since the program is transmitted to the receiving device 9, for example, programs having similar titles such as a serial drama and a special program can be processed as the same group.
[0128]
Furthermore, by determining a group based on the matching rate of words constituting a title, for example, in the EPG or metadata, a half-width and full-width of a number, a half-width and a full-width of an alphabet, or a capital Even if there is a variation in the notation such as lowercase letters, it is possible to detect programs with the same title as the same group.
[0129]
Further, in addition to the word matching rate, for example, a broadcast station, a program genre, or a broadcast start time may be added to the grouping condition. For example, in a news program or the like, the title is composed of a small number of words including “news”. Therefore, in the processing described with reference to FIG. May be detected as the same group. Therefore, when a broadcast station also matches in addition to the word matching rate, the same group may be used.
[0130]
With reference to the flowchart of FIG. 11, a description will be given of a title grouping process 3 for performing grouping based on the matching rate of words constituting a title, in addition to the condition of matching broadcasting stations.
[0131]
In steps S421 to S424, the same processing as steps S401 to S404 described with reference to FIG. 10 is executed. That is, the data acquisition unit 21 refers to the EPG data registered in the metadata database 7, extracts a title from the registered metadata, supplies the title to the program vector generation unit 23, and 23 morphologically analyzes the title and breaks it down into words. Then, the program vector generation unit 23 calculates the degree of matching between the titles based on the analyzed words, and determines whether the words match at least a predetermined value such as 70%, for example. .
[0132]
When it is determined in step S424 that the words match at least a predetermined value such as 70%, in step S425, the program vector generation unit 23 determines whether the broadcasting stations of those programs match. to decide.
[0133]
If it is determined in step S425 that the broadcasting stations of these programs match, in step S426, the program vector generation unit 23 associates the programs with the same group ID. In addition, the program vector generation unit 23 causes the data storage unit 24 to store the matched word or word group and the corresponding broadcast station and group ID.
[0134]
If it is determined in step S424 that the matching rate is equal to or less than a predetermined value such as 70%, if it is determined in step S425 that the broadcasting stations of these programs do not match, or if the processing in step S426 is performed. Is completed, in step S427, the program vector generation unit 23 determines whether or not the round robin of the title has been completed.
[0135]
If it is determined in step S427 that the round robin of the title has not been completed, the process returns to step S423, and the subsequent processes are repeated.
[0136]
If it is determined in step S427 that the brute force of the title has been completed, the process is terminated.
[0137]
By such processing, the program vector is associated with the group ID based on the match of the broadcast station and the match rate of the words constituting the title. Since the program is transmitted to the device 4 or the EPG receiving device 9, for example, when programs having similar titles are in the same group, it is possible to prevent news programs of other stations from being in the same group. .
[0138]
Note that, in FIG. 11, the description has been made assuming that the grouping is performed on the condition that they are the same broadcasting station, in addition to the coincidence rate of the words constituting the title. It goes without saying that the grouping may be performed using the band, the genre, and the like as conditions other than the matching rate of the words constituting the title.
[0139]
Furthermore, for example, even when the broadcast start time of a serial drama or a band program is shifted due to a sports broadcast or a special program, conditions other than the matching rate of the words constituting the title can be detected as the same group. May be determined based on whether or not the broadcast times coincide with each other within a predetermined time range such as one hour.
[0140]
Referring to the flowchart of FIG. 12, a title grouping is performed in which grouping is performed based on a matching rate of words forming a title, in addition to a condition that broadcast times match with a shift within a predetermined time range. Processing 4 will be described.
[0141]
In steps S441 to S444, the same processing as steps S401 to S404 described with reference to FIG. 10 is executed. That is, the data acquisition unit 21 refers to the EPG data registered in the metadata database 7, extracts a title from the registered metadata, supplies the title to the program vector generation unit 23, and 23 morphologically analyzes the title and breaks it down into words. Then, the program vector generation unit 23 calculates the degree of matching between the titles based on the analyzed words, and determines whether the words match at least a predetermined value such as 70%, for example. .
[0142]
If it is determined in step S444 that the words match at least a predetermined value such as 70%, in step S445, the program vector generation unit 23 sets the broadcast start time of those programs to, for example, one hour. It is determined whether or not they coincide with each other within a predetermined range.
[0143]
If it is determined in step S445 that the broadcast start times of those programs match with a deviation within a predetermined range, in step S446, the program vector generation unit 23 assigns the same group ID to those programs. Correspond. In addition, the program vector generation unit 23 causes the data storage unit 24 to store the matched word or word group, the corresponding broadcast start time range, and the group ID.
[0144]
If it is determined in step S444 that the coincidence rate is equal to or less than a predetermined value such as 70%, if it is determined in step S445 that the broadcast start times of those programs are out of the predetermined range, or After the process of step S446 is completed, in step S447, the program vector generation unit 23 determines whether or not the round robin of the title has been completed.
[0145]
If it is determined in step S447 that the round robin of the title has not been completed, the process returns to step S443, and the subsequent processes are repeated.
[0146]
If it is determined in step S447 that the round robin of the title has been completed, the process is terminated.
[0147]
By such processing, the program vector is associated with a match including a deviation of the broadcast start time within a predetermined range and a group ID based on the match rate of words constituting the title, and the data transmission unit 25 Is transmitted to the television receiver 4 or the EPG receiver 9 via the network 8, for example, when a program having a similar title is in the same group, the broadcast time of a special program is changed. Furthermore, it is possible to prevent programs that should be detected as being in the same group from being detected as being in the same group.
[0148]
FIG. 13 is a block diagram showing the configuration of the program recommendation processing device 10.
[0149]
The data acquisition unit 41 acquires the program vector PP transmitted from the distribution server 5 and the program-side effect vector EfPP corresponding to the program vector PP. The program vector extraction unit 42 extracts, from the program vector PP acquired by the data acquisition unit 41, a program vector PP required for matching processing or a program vector PP required for user model generation. The program-side effect vector EfPP corresponding to the program vector PP is supplied to the matching processing unit 43.
[0150]
The operation input unit 44 includes, for example, an input device such as a keyboard, a touch pad, and a mouse, and receives input of initial registration information input by a user and a topic for generating a user model, and outputs the information to the initial registration storage unit 45. I do. The initial registration storage unit 45 registers the initial registration content supplied from the operation input unit 44 and a topic for generating a user model, and supplies the topic to the operation log acquisition unit 46 or the matching processing unit 43 as necessary. I do. The contents stored in the initial registration storage unit 45 are sequentially updated based on a user operation input from the operation input unit 44. The contents to be initially registered include, for example, information indicating a program that the user does not like, such as disliked genre, disliked keyword, disliked performer, and favorite genre, favorite keyword, favorite performer, etc. There is information indicating programs that the user prefers.
[0151]
The operation log acquisition unit 46 acquires an operation log from the television display device 11 or the recording / reproducing device 12, classifies the information into a positive history and a negative history, and stores the information in the initial registration storage unit 45 as necessary. The program vector PP corresponding to the positive history and the negative history is read out of the program vectors PP acquired by the data acquisition unit 41 with reference to the information which has been acquired, and supplied to the positive history storage unit 47 and the negative history storage unit 48. And save. The main history storage unit 47 stores the supplied main history and generates a main history vector UP. The negative history storage unit 48 stores the supplied negative history and generates a negative history vector MUP. The generated positive history vector UP and negative history vector MUP are supplied to the matching processing unit 43.
[0152]
Here, the correct history is information used to extract a candidate for a program that the user intends to actively view, in other words, a preferred program. For example, the user can view the program, When the program is recorded, or when a program proposed as a recommended program list to the user by the processing described later is accepted and viewed or recorded by the user, the metadata of the program is stored as a positive impression metadata as a good history metadata. Stored in the unit 47. The main history storage unit 47 calculates the sum of the main histories for each detailed item or each large item, and generates a main history vector UP.
[0153]
On the other hand, the negative history is information used for excluding a program that the user is reluctant to watch, in other words, a program that he or she considers unfavorable from recommended programs, for example, dislikes initial registration. If the item, the program that was deleted without being watched after recording, or the program suggested as a recommended program list to the user by the processing described below was not accepted by the user and was not viewed or recorded, the program The metadata is stored in the negative history storage unit 48 as negative impression metadata. For example, if the user dislikes “sports” in the initial registration, the genre Gmup = {0,0,5,0,0,0,0,0,0} is negative with a weight of 5 as a negative impression. It is additionally stored in the history storage unit 48. The negative history storage unit 48 calculates the sum of the negative histories for each detailed item or for each large item, and generates a negative history vector MUP.
[0154]
The matching processing unit 43 verifies matching between the program vector extracted from the program vector extraction unit 42 and the positive history vector UP or the negative history vector MUP supplied from the positive history storage unit 47 or the negative history storage unit 48. I do.
[0155]
When the program vector PP, the positive history vector UP, or the negative history vector MUP is vectorized by arranging all the detailed elements in one column, the title or keyword is composed of a plurality of words. One word and an item such as a genre have the same weight in the vector. Thus, the normalization processing unit 61 of the matching processing unit 43 normalizes the title and the contents, which are the items constituted by the words, by dividing the words by the number of words for each program. For example, if the title Tm = {Tokaido-1, Mitani-1, Kaidan-1}, the title Tm = {Tokaido: 0.33, Mitani: 0.33, Kaidan: 0.33} is normalized. As a result, both the title and the content have a weight of 1 in total of the words in the item, so that no inconvenience occurs in the matching processing.
[0156]
The vector calculation unit 62 executes a matching process between the program vector PP and the positive history vector UP or the negative history vector MUP.
[0157]
When the program vector PP, the positive history vector UP, and the negative history vector MUP are represented by a vector in which all the elements of the detailed items are arranged in one column, the vector operation unit 62 , The similarity SimUP between the program vector PP and the positive history vector UP is determined by the cosine distance (cos θ) between the vector and the program vector PP, and expressed by the following equation (2). To obtain the similarity SimMUP between the program vector PP and the negative history vector MUP. The cosine distance is a value obtained by dividing the inner product of two vectors by the product of the absolute values of the vectors, as shown in Expressions (1) and (2).
[0158]
SimUP = cosθu = UP ・ PP / | UP | × | PP | ・ ・ ・ (1)
SimMUP = cosθm = MUP · PP / | MUP | × | PP | (2)
[0159]
However, in Equations (1) and (2), PP indicates the program vector PP, UP indicates the positive history vector UP, and MUP indicates the negative history vector MUP. "." Indicates an inner product, and "x" indicates multiplication of elements (scalar operation).
[0160]
Further, when the program vector PP, the positive history vector UP, and the negative history vector MUP are vectorized for each large item, the vector calculation unit 62 determines the similarity between the program vector PP and the positive history vector UP for each large item. Then, the similarity between the program vector PP and the negative history vector MUP is obtained, and the sum of the similarities is calculated for each major item, so that the similarity SimUP and the similarity SimMUP can be calculated. For example, in a large item "title" which is broken down into words, the correct history vector UP is the title Tup = {school-1, ghost story-1, toilet-1}, and the supplied program vector PP is the title Tm = {Tokaido-1, Mitani-1, Kaidan-1}, assuming that the unit vector length is 1, the vector length (absolute value) is the square root of the sum of squares of the sum of the elements. The cosine distance cosθt, which is the degree of similarity between titles, is calculated as in the following equation (3).
cosθt = (1.1) / (√3 × √3) = 1/3 ... (3)
[0161]
In the equation (3), “·” indicates a dot product, and “×” indicates a scalar operation. The degree of similarity between the program vector PP and the negative history vector MUP for each major item is also calculated by the same calculation as Expression (3).
[0162]
For example, a positive history vector UP, a positive history vector UP = {title Tup, genre Gup, performer Pup, script / original / production Aup, content (keyword) Kup} and a negative history vector MUP, a negative history vector MUP = ｛ When the title Tmup, the genre Gmup, the performer Pmup, the script / original / directing Amup, and the content (keyword) Kmup｝, the similarity is calculated for each of the major items by the same calculation as in the equation (3). Therefore, the similarity SimUP and the similarity SimMUP are both calculated as the similarity Sim of the following equation (4).
[0163]
Sim = cosθ_t+ cosθ_g+ cosθ_p+ cosθ_a+ cosθ_k... (4)
[0164]
Where cosθ_tIs the cosine distance between the program vector PP and the positive history vector UP or the negative history vector MUP in the large item “title”, and cos θ_gIs the cosine distance between the program vector PP and the positive history vector UP or the negative history vector MUP in the large item "genre", and cosθ_pIs the cosine distance between the program vector PP and the positive history vector UP or the negative history vector MUP in the large item “performers”, and cos θ_aIs the cosine distance between the program vector PP and the positive history vector UP or the negative history vector MUP in the major item "Screenplay / Original / Direction"._kIs the cosine distance between the program vector PP and the positive history vector UP or the negative history vector MUP in the large item "content".
[0165]
When the similarity SimUP and the similarity SimMUP are calculated as the sum of the calculation results for each large item, there is no bias in the weight between items, so it is moral to perform normalization. Therefore, unlike a vector in which all elements are arranged in one column, the normalization processing by the normalization processing unit 61 need not be performed.
[0166]
That is, if a vector in which all elements are arranged in one column without normalization is used, the frequency tends to accumulate. For example, items such as titles and contents are compared with items such as broadcast stations and genres. In, the number of words increases as the history increases, but the frequency of each word does not easily increase.
[0167]
For this reason, when the total is calculated for each element using all the elements, the influence of items such as the broadcasting station and the genre, which are likely to overlap frequently, becomes large. For example, the user is a fan of the commentator A. So, if you prefer to watch “live commentary on the game of team B commented on by commentator A”, the genre “live commentary on baseball” is likely to overlap as a history, Is difficult to overlap as a history. Accordingly, a case may occur where a live broadcast of the game of the team B being commented on by another commentator is recommended and a variety program in which the commentator A appears is not recommended.
[0168]
On the other hand, by performing normalization, or by calculating the sum of the operation results for each large item as the similarity SimUP and the similarity SimMUP, without being affected by the magnitude of the history frequency, The variety program on which the commentator A appears can be recommended, and the user's preferences can be more accurately reflected.
[0169]
In addition, the vector calculation unit 62 generates the user's initial registration information stored in the initial registration storage unit 45, the program side effect vector EfPP transmitted in association with the program vector PP, or the user information registration unit 63. The similarity SimUP and the similarity SimMUP may be calculated by performing weighting using the registered and effective user side effect vector EfUP (described later) or the registered user side effect vector EfMUP (described later).
[0170]
Based on the similarity SimUP calculated by the above-described method, the vector calculation unit 62 further adds, for example, a history vector to a predetermined number (for example, 10) of programs having a high similarity with the main history vector. The similarity SimMUP with the MUP is calculated, and SimUP-SimMUP is calculated, and a predetermined number (for example, 3) of programs having higher calculation results are output to the recommended information output unit 49 as recommended programs.
[0171]
When the program vectors are grouped, the vector calculation unit 62 registers the recommended priority group in the user information registration unit 63 based on the information of the recommended program, and registers the program corresponding to the recommended priority group. , Preferentially recommend.
[0172]
Further, the vector calculation unit 62 generates a user model vector by filtering the program vector PP using the topic stored in the initial registration storage unit 45, registers the user model vector in the user information registration unit 63, and performs matching. Processing can be performed. Details of the user model will be described later.
[0173]
The user information registration unit 63 is based on the initial registration contents of the user supplied from the initial registration storage unit 45 or the positive history vector UP or the negative history vector MUP supplied from the positive history storage unit 47 or the negative history storage unit 48. Then, a user-side effect vector EfUP and a counter-effect vector EfMUP are generated and stored. The user side effect vector EfUP is, for the user, a vector indicating which element of the large items is important for the selection of the program and is an item that is weighted for the selection of the program, or , A vector indicating the user's preference in each item. The reaction vector EfMUP is a vector indicating which element of the large item is an element that is not important for the selection of the program for the user and is an item that is not weighted for the selection of the program, Alternatively, in each item, it is a vector indicating an item that the user does not like.
[0174]
In other words, the user-side effect vector EfUP and the counter-effect vector EfMUP define which item greatly contributes in matching the program vector PP with the positive history vector UP or the negative history vector MUP.
[0175]
The user-side effect vector EfUP and the counter-effect vector EfMUP may be settable by a user or may use a predetermined value. May be generated.
[0176]
Specifically, if the user-side effect vector EfUP is information indicating which element of the large items is important for the user to select a program, the program vector PP = {title Tm, genre Gm , Time zone Hm, broadcast station Sm, performer Pm, script / original / directing Am, content Km}, when the genre is important to the user, for example, the effect vector EfUP = (1,5,1,1,1) , 1,1). On the other hand, when the performer and the genre are important to the user, for example, the effect vector EfUP is set to (1, 3, 1, 1, 5, 1, 1).
[0177]
When the user side effect vector EfUP is a vector indicating the user's preference in each item, a large item genre Gm = ｛drama, variety, sports, movie, music, for children / education, education / documentation, news / If the genre of the program that the user considers preferable in the news report or other｝ is culture / document, the user effect vector EfUP is, for example, genre Gm = {0, 0, 0, 0, 0, 0, 5, 0,0}.
[0178]
Further, the user-side effect vector EfUP and the countereffect vector EfMUP may be generated based on the positive history vector UP or the negative history vector MUP, or by counting programs watched by the user during a certain period. good. Further, the user-side effect vector EfUP and the counter-effect vector EfMUP can be generated for each genre. A method of generating the user side effect vector EfUP or the counter effect vector EfMUP will be described later with reference to FIGS.
[0179]
Further, the user information registration unit 63 registers information of the recommendation priority group, a user model vector, and the like generated by the processing of the vector calculation unit 62 as necessary.
[0180]
In the matching processing unit 43 described above, from programs having a high degree of similarity to the positive history vector UP (programs that the user actively views), programs that the user does not like (user Even without performing the process of removing the program that is reluctant to view), for example, it is possible to determine the recommended program using only the main history.
[0181]
The recommended information output unit 49 registers the recommended program information supplied from the matching processing unit 43 in the recommended program list 50 and supplies the information to the television display device 11 or the recording / reproducing device 12. The recommended program list 50 is configured to be detachable from the program recommendation processing device 10, and registers the recommended program information output from the recommended program output unit 49. By storing the recommended program information in the recommended program list 50, for example, even when a different television receiver 4, the television display device 11, or the recording / reproducing device 12 is used, the history accumulated so far can be obtained. Using the information, it is possible to execute program recommendation, automatic recording, and the like.
[0182]
Further, the matching processing unit 43 is connected to the drive 51 as necessary. A magnetic disk 71, an optical disk 72, a magneto-optical disk 73, and a semiconductor memory 74 are mounted on the drive 51 as necessary, and exchange data.
[0183]
The positive history vector and negative history vector generation processing 1 executed by the program recommendation processing device 10 will be described with reference to the flowchart of FIG.
[0184]
In step S71, the operation log acquisition unit 46 supplies the initial registration contents read from the initial registration storage unit 45 to the negative history storage unit 48. The negative history storage unit 48 generates a negative history vector MUP with reference to the supplied initial registration contents.
[0185]
In step S72, the operation log obtaining unit 46 determines whether the initial registration content has been changed based on the registration content stored in the initial registration storage unit 45. If it is determined in step S72 that the initial registration content has been changed, the process returns to step S71, and the subsequent processes are repeated.
[0186]
If it is determined in step S72 that the initial registration contents have not been changed, in step S73, the operation log acquisition unit 46 determines whether an operation log has been supplied from the television display device 11 or the recording / reproducing device 12. to decide. If it is determined in step S73 that the operation log has not been supplied, the process returns to step S72, and the subsequent processes are repeated.
[0187]
When it is determined in step S73 that the operation log has been supplied, in step S74, the operation log acquisition unit 46 determines whether or not the supplied operation log is a normal history. For example, if the operation log is a recording operation, the program vector PP of the program corresponding to the operation is a normal history, and if the operation log is erasure of recorded data that has not been reproduced, the program corresponding to the operation is deleted. Is a negative history.
[0188]
If it is determined in step S74 that the supplied operation log has the correct history, in step S75, the operation log obtaining unit 46 stores the program vector PP corresponding to the operation log determined to be the correct history in the data. It is extracted from the acquisition unit 41 and supplied to the main history storage unit 47. The main history storage unit 47 additionally stores the supplied program vector PP as a main history.
[0189]
In step S76, the main history storage unit 47 obtains the sum of the vectors of the main history program vector PP for each detailed item or each large item, and generates the main history vector UP. After the process in step S76 ends, the process returns to step S72, and the subsequent processes are repeated.
[0190]
If it is determined in step S74 that the supplied operation log is not a positive history, the supplied operation log is a negative history. Therefore, in step S77, the operation log acquisition unit 46 determines that the supplied operation log is a negative history. The program vector PP corresponding to the determined operation log is extracted from the data acquisition unit 41 and supplied to the negative history storage unit 48. The negative history storage unit 48 additionally stores the supplied program vector PP as a negative history.
[0191]
In step S78, the negative history storage unit 48 calculates the sum of the vectors of the negative history program vector PP for each detailed item or for each large item, and generates a negative history vector MUP. After the end of the process in the step S78, the process returns to the step S72, and the subsequent processes are repeated.
[0192]
The major item is the main history vector UP = {title Tup, genre Gup, performer Pup, script / original / directing Aup, content (keyword) Kup}, and when detailed items are described in the major item, In the positive history vector UP, a numerical value indicating the vector sum is described after each detailed item. For example, as shown in FIG. 15, regarding the major item “genre”, the genre Gup = ｛(drama-25), (variety-34), (sports-42), (movie-37), (music-73). ), (For children / education-120), (education / document-3), (news / reports-5), (others-23)}, the numerical value indicating the sum of the positive history for each detailed item is described. Is done. In a vector indicated by a word, such as a major item “title”, for example, a title Tup = {(title1-12), (title2-3),. A numerical value indicating the sum of the correct histories for each word is described. In the negative history vector MUP, similarly to the positive history vector UP, a numerical value indicating the sum is described following the item.
[0193]
In FIG. 15, the major items of the positive history vector UP (and the negative history vector MUP) are the title, genre, cast, script / original / production, and content (keyword), and the program vector described with reference to FIG. Although the description has been made on the assumption that the number of items is smaller than that of the PP, it goes without saying that it may be a large item similar to the program vector PP.
[0194]
In FIG. 14, it is described that the negative history vector MUP is generated before the operation log is input, based on the contents of the initial registration. However, at the time of the initial registration, the user selects a program that he / she prefers. May be registered, and the normal history vector UP may be generated before the operation log is input. Alternatively, the positive history vector UP or the negative history vector MUP may be generated using only the operation log without generating the positive history vector UP or the negative history vector MUP based on the initial registration.
[0195]
In this way, by independently generating and holding the positive history vector UP and the negative history vector MUP, it is possible to more accurately perform the matching process with the user's preference.
[0196]
At this time, the positive history and the negative history may be determined more precisely. For example, in the processing described with reference to FIG. 14, the positive history vector UP and the negative history vector MUP are generated using the sum of the program vectors PP corresponding to the positive history and the negative history in all items. For example, the total sum of the program vectors PP corresponding to the negative history may be accumulated for each genre, and the positive history vector UP and the negative history vector MUP may be generated for each genre.
[0197]
Performers are often biased depending on the genre of the broadcasted program, so that the user's preferences may not be correctly reflected. Specifically, a user who likes a drama and prefers only those in which comedian A who hardly appears in the drama appears, and who has a variety-to-drama viewing ratio of 2: 8, performs without distinguishing the genre Performer B who frequently appears in the drama despite being not a favorite actor more frequently than the comedian A who rarely appears in the drama, May be lost. In such a case, for example, a documentary in which the performer B who frequently appears in the drama will be recommended, rather than the variety in which the comedian A appears. To prevent this, the positive history and the negative history may be accumulated for each genre, and based on this, the positive history vector UP and the negative history vector MUP may be generated for each genre.
[0198]
Further, for example, since the user is a fan of the commentator A, if the user likes to watch “live broadcast of the game of the team B explained by the commentator A”, the genre is called “sports”. Although the information is likely to overlap as a history, the information of the performer "commentator A" is unlikely to overlap as a history, so a live commentary on the game of team B being described by another commentator is recommended, and commentator A There is a case where a variety program in which is appeared is not recommended. To prevent this, for example, a positive history and a negative history may be accumulated for each performer, and based on this, a positive history vector UP and a negative history vector MUP may be generated for each performer.
[0199]
In this way, by accumulating the history for each specific element, it is possible to more precisely reflect the user's preferences without dulling the user's preferences.
[0200]
Then, the matching processing unit 43 verifies the matching between the positive history vector UP and the negative history vector MUP thus generated and the supplied program vector PP, thereby recommending recommended program information that correctly reflects the user's preference. Can be generated.
[0201]
Next, the positive history vector and negative history vector generation processing 2 in which histories are accumulated for each genre will be described with reference to the flowchart of FIG.
[0202]
In steps S81 to S84, the same processing as in steps S71 to S74 in FIG. 14 is performed. That is, the initial registration is referred to, a negative history vector MUP is generated, and it is determined whether or not the initial registration content has been changed. If the initial registration content has not been changed, whether or not the supplied operation log has the correct history is determined. Is determined.
[0203]
If it is determined in step S84 that the supplied operation log has the correct history, in step S85, the operation log acquisition unit 46 acquires the program vector PP corresponding to the operation log determined to be the correct history. Extracted from the section 41 and supplied to the main history storage section 47. The correct history storage unit 47 extracts the genre of the supplied program vector PP.
[0204]
In step S86, the main history storage unit 47 additionally stores the program vector PP extracted from the data acquisition unit 41 as a main history for each genre.
[0205]
In step S87, the main history storage unit 47 obtains the sum total of the main history program vectors PP in the genre in which the program vector is additionally stored for each detailed item or each large item, and Generate the positive history vector UP. After the processing in step S87 ends, the processing returns to step S82, and the subsequent processing is repeated.
[0206]
If it is determined in step S84 that the supplied operation log is not a positive history, the supplied operation log is a negative history. Therefore, in step S88, the operation log acquisition unit 46 determines that the supplied operation log is a negative history. The program vector PP corresponding to the operation log is extracted from the data acquisition unit 41 and supplied to the negative history storage unit 48. The negative history storage unit 48 extracts the genre of the supplied program vector PP.
[0207]
In step S89, the negative history storage unit 48 additionally stores the program vector PP extracted from the data acquisition unit 41 as a negative history for each genre.
[0208]
In step S90, the negative history storage unit 48 obtains the sum total of the negative history program vectors PP in the genre in which the program vector is additionally stored for each detailed item or for each large item, and Generate a negative history vector MUP. After the processing in step S90 ends, the processing returns to step S82, and the subsequent processing is repeated.
[0209]
By such processing, the positive history vector UP and the negative history vector MUP are generated for each genre, so that it is possible to more accurately reflect the user's preferences without dulling the user's preferences, It is possible to generate recommended program information that is correctly reflected.
[0210]
With reference to the flowchart of FIG. 17, a description will be given of the matching process 1 when the program vector PP, the positive history vector UP, and the negative history vector MUP are represented by a vector in which all the elements of the detailed items are arranged in one column. .
[0211]
In step S101, the program vector extraction unit 42 extracts the program vectors PP of a plurality of programs (for example, programs broadcasted in a predetermined time zone) from the data acquisition unit 41, and normalizes the program vectors PP of the matching processing unit 43. To supply. The normalization processing unit 61 normalizes the supplied program vector PP and the titles and contents composed of words among the components of the main history vector UP read from the main history storage unit 47, and The result of the normalization is supplied to the calculation unit 62.
[0212]
Specifically, if the supplied program vector PP is, for example, the title Tm = {Tokaido-1, Mitani-1, Kaidan-1}, the title Tm = {Tokaido: 0. 33, Mitani: 0.33, Kaidan: 0.33｝, and normalize the program so that the total weight of words in the item is 1 for each program.
[0213]
In step S102, the vector calculation unit 62 of the matching processing unit 43 calculates the similarity SimUP, which is the cosine distance between the program vector PP of a plurality of programs and the positive history vector UP, using the above-described equation (1).
[0214]
In step S103, the vector calculation unit 62 compares the similarity SimUP indicating the similarity between the program vector PP and the main history vector UP calculated in step S102, and from the highest similarity, a predetermined value such as 10, for example. Are extracted as many as the program vector PP.
[0215]
In step S104, the vector calculation unit 62 calculates the similarity SimMUP, which is the cosine distance between the program vector PP extracted in step S103 and the negative history vector MUP read from the negative history storage unit 48, using the above-described equation (2). Is calculated using
[0216]
In step S105, the vector calculation unit 62 calculates the similarity (ie, the cosine distance) SimMUP with the positive history vector UP−the similarity (ie, the cosine distance) SimMUP with the negative history vector, and sets a higher rank, the predetermined value. A program vector or EPG data of a number (for example, one) of programs is extracted as recommendation information, output to the recommendation information output unit 49, registered in the recommended program list 50, and the television display device 11 and the recording / reproducing device. 12 and the process ends.
[0217]
By such processing, when the program vector PP, the positive history vector UP, and the negative history vector MUP are represented by a vector in which all the elements of the detailed items are arranged in one column, the program vector PP and the positive history vector UP , And the similarity between the program vector PP and the negative history vector, it is possible to determine a recommended program that matches the user's preference.
[0218]
Next, referring to the flowchart of FIG. 18, instead of normalizing the item indicated by the word, the cosine distance is calculated for each large item, and the sum thereof is calculated as the similarity SimUP and the similarity SimMUP. The following describes a matching process 2 for determining a recommended program.
[0219]
In step S111, the program vector extraction unit 42 extracts a program vector PP of a plurality of programs (for example, a program broadcasted in a predetermined time zone) from the data acquisition unit 41 and sends the program vector PP to the vector calculation unit 62 of the matching processing unit 43. Supply. The vector calculation unit 62 calculates the cosine distance between the program vector PP and the main history vector UP for each of the large items of the supplied program vector PP and the main history vector UP read from the main history storage unit 47.
[0220]
In step S112, the vector calculation unit 62 sums the values of the cosine distance calculated for each item in step S111 using the above-described equation (4), and calculates the similarity SimUP.
[0221]
In step S113, the vector calculation unit 62 compares the similarity SimUP between the program vector PP and the normal history vector UP calculated in step S112, and determines a predetermined number of programs such as 10 from the highest similarity. Extract the vector PP.
[0222]
In step S114, the vector calculation unit 62 determines the program vector PP and the negative history vector MUP for each of the large items of the program vector PP extracted by the process of step S113 and the negative history vector MUP read from the negative history storage unit 48. Is calculated.
[0223]
In step S115, the vector calculation unit 62 calculates the similarity SimMUP by summing the values of the cosine distance calculated for each item in step S114 using Expression (4) described above.
[0224]
In step S116, the vector calculation unit 62 calculates {similarity SimUP which is a cosine distance between the program vector PP and the positive history vector UP} − {similarity SimMUP} which is a cosine distance between the program vector PP and the negative history vector. Then, program vectors PP or EPG data of a predetermined number (for example, three) of higher-ranked programs are extracted as recommendation information, output to the recommendation information output unit 49, registered in the recommended program list 50, and The data is output to the display device 11 and the recording / reproducing device 12, and the process ends.
[0225]
By such processing, the sum of the calculation results is calculated as the similarity SimUP and the similarity SimMUP for each large item without normalizing the item indicated by the word. A recommendation that matches the user's preference based on the similarity between the program vector PP and the positive history vector UP and the similarity between the program vector PP and the negative history vector MUP without being affected by the bias of the history overlap. A program can be determined.
[0226]
In the matching process 1 and the matching process 2 described with reference to FIGS. 17 and 18, from programs having a high degree of similarity to the user's positive history vector UP, programs that the user does not like using the negative history are used. Although it has been described that the program can be removed, for example, the recommended program may be determined using only the main history.
[0227]
In addition, some users prefer news and news programs in selecting a program, or performers are more important than genres, and they do not care about performers, but the content is important. In some cases, items that are weighted to select the item and items that are not weighted, in other words, important items and items that are not are fixed.
[0228]
Therefore, the matching process may be performed using the above-described program side effect vector EfPP, user side effect vector EfUP, or user side effect vector EfMUP. Whether to use the program side effect vector EfPP, the user side effect vector EfUP, or the user side effect vector EfMUP may be set by the user.
[0229]
Next, referring to the flowchart of FIG. 19, when the program vector PP, the positive history vector UP, and the negative history vector MUP are represented by a vector in which all the elements of the detailed items are arranged in one column, In the following, a description will be given of a matching process 3 in which the matching process is performed using the program side effect vector EfPP or the user side effect vector EfUP.
[0230]
In step S121, the vector operation unit 62 inputs the program side effect vector EfPP and the user side effect vector EfUP or the user side reaction effect vector EfMUP input by the user using the operation input unit 44 and registered in the initial registration storage unit 45. Get the usage setting contents of. The effect vector usage setting content is information indicating whether or not weighting is performed using the program side effect vector EfPP, the user side effect vector EfUP or the user side effect vector EfMUP in the matching process.
[0231]
In step S122, the vector calculation unit 62 reads the user-side effect vector EfUP from the user information registration unit 63 as necessary, and uses the following equation (5) to cosine the program vector PP and the positive history vector UP. The distance is calculated and set as the similarity SimUP.
[0232]
(Equation 1)

[0233]
In equation (5), the program vector PP = (p₁, P_Two, ...), the positive history vector UP = (u₁, U_Two, ...), the program side effect vector EfPP = (epd₁, Epd_Two,...), The user-side effect vector EfUP = (eud₁, Eud_Two, ...). In addition, in the equation (5), both the program side effect vector EfPP and the user side effect vector EfUP are described. However, depending on the setting, the program side effect vector EfPP and the user side effect vector EfUP are used. Is not used, the numerical value “1” is substituted for the vector not used and calculated.
[0234]
Further, the user side effect vector EfUP may be set by the user, may be set based on the user's initial setting, or may be generated in the user information registration unit 63. good. Details of generation of the user-side effect vector EfUP will be described later with reference to FIGS.
[0235]
In step S123, the vector calculation unit 62 compares the similarity SimUP between the program vector PP and the main history vector UP calculated in step S122, and determines a predetermined number of programs such as 10 from the highest similarity. Extract the vector PP.
[0236]
In step S124, the vector calculation unit 62 reads the user side reaction vector EfMUP from the user information registration unit 63 as necessary, and uses the following equation (6) to extract the program vector PP extracted in step S123. And the cosine distance of the negative history vector MUP.
[0237]
(Equation 2)

[0238]
In equation (6), the program vector PP = (p₁, P_Two, ...), the negative history vector MUP = (m₁, M_Two, ...), the program side effect vector EfPP = (epd₁, Epd_Two,...), The user side reaction vector EfMUP = (emd₁, Emd_Two, ...). In addition, in equation (6), both the program side effect vector EfPP and the user side effect vector EfMUP are described. However, depending on the setting, the program side effect vector EfPP and the user side effect vector EfMUP are used. If any of is not used, the numerical value “1” is substituted for the unused vector and calculated.
[0239]
The user-side effect vector EfMUP may be set by the user, may be set based on the user's initial setting, or may be generated by the user information registration unit 63. Is also good. Details of generation of the user-side reaction vector EfMUP will be described later with reference to FIG. 26 or FIG.
[0240]
In step S125, the vector calculation unit 62 calculates the similarity SimUP between the program vector PP and the positive history vector UP-the similarity SimMUP between the program vector PP and the negative history vector, and a predetermined number (for example, three The program vector PP or EPG data of the program is extracted as recommendation information, output to the recommendation information output unit 49, registered in the recommended program list 50, and output to the television display device 11 and the recording / reproducing device 12. , The process ends.
[0241]
According to such processing, the recommendation information is extracted using the program side effect vector EfPP, the user side effect vector EfUP, or the user side effect vector EfMUP according to the setting, so that the user's preference is correctly reflected. The recommended program can be recommended.
[0242]
In the processing described with reference to FIG. 19, the processing when the program vector PP, the positive history vector UP, and the negative history vector MUP are represented by a vector in which all the elements of the detailed items are arranged in one column has been described. The program vector PP, the positive history vector UP, and the negative history vector MUP may be calculated for each large item.
[0243]
Next, referring to the flowchart of FIG. 20, a matching process 4 in which a program-side effect vector EfPP, a user-side effect vector EfUP, or a user-side countereffect vector EfMUP can be reflected for each large item. Will be described.
[0244]
In step S131, the same processing as in step S121 in FIG. 19 is performed, and the usage setting content of the effect vector is obtained.
[0245]
In step S132, the vector calculation unit 62 calculates the cosine distance between the program vector PP and the main history vector UP for each of the large items of the supplied program vector PP and the main history vector UP read from the main history storage unit 47. I do. Here, the effect vector is not used for the calculation.
[0246]
In step S133, the vector calculation unit 62 multiplies the cosine distance calculated for each item by the effect vector as necessary using the following equation (7), sums the obtained values, Calculate the degree SimUP.
[0247]
SimUP = epd_t・ Eud_t・ Cosθu_t+ epd_g・ Eud_g・ Cosθu_g+ epd_p・ Eud_p・ Cosθu_p
+ epd_a・ Eud_a・ Cosθu_a+ epd_k・ Eud_k・ Cosθu_k... (7)
[0248]
In equation (7), the program vector PP = (p_t, P_g, P_p, P_a, P_k) And the positive history vector UP = (u_t, U_g, U_p, U_a, U_k) And the cosine distance of each major item is (cosθu_t, Cosθu_g, Cosθu_p , Cosθu_a, Cosθu_k), And the program side effect vector EfPP = (epd_t, Epd_g, Epd_p, Epd_a, Epd_k), User side effect vector EfUP = (eud_t, Eud_g, Eud_p, Eud_a, Eud_k). In addition, in the equation (7), both the program side effect vector EfPP and the user side effect vector EfUP are described. However, depending on the setting, the program side effect vector EfPP and the user side effect vector EfUP are used. Is not used, the numerical value “1” is substituted for the vector not used and calculated.
[0249]
In step S134, the vector calculation unit 62 compares the similarity SimUP between the program vector PP and the normal history vector UP calculated in step S133, and determines a predetermined number of programs such as 10 from the highest similarity. Extract the vector PP.
[0250]
In step S135, the vector calculation unit 62 determines the program vector PP and the negative history vector MUP for each of the large items of the program vector PP extracted by the process of step S134 and the negative history vector MUP read from the negative history storage unit 48. Is calculated. Here, the effect vector is not used for the calculation.
[0251]
In step S136, the vector calculation unit 62 multiplies the cosine distance calculated for each item by an effect vector as necessary using the following equation (8), sums the obtained values, Calculate the degree SimMUP.
[0252]
SimMUP = epd_t・ Emd_t・ Cosθm_t+ epd_g・ Emd_g・ Cosθm_g+ epd_p・ Emd_p・ Cosθm_p
+ epd_a・ Emd_a・ Cosθm_a+ epd_k・ Emd_k・ Cosθm_k... (8)
[0253]
In equation (8), the program vector PP = (p_t, P_g, P_p, P_a, P_k) And negative history vector UP = (m_t, M_g, M_p, M_a, M_k) And the cosine distance of each major item is (cosθm_t, Cosθm_g, Cosθm_p , Cosθm_a, Cosθm_k), And the program side effect vector EfPP = (epd_t, Epd_g, Epd_p, Epd_a, Epd_k,), User side reaction vector EfMUP = (emd_t, Emd_g, Emd_p, Emd_a, Emd_k). Further, in the equation (8), both the program side effect vector EfPP and the user side effect vector EfMUP have been described. However, depending on the setting, the program side effect vector EfPP and the user side effect vector EfMUP If any of is not used, the numerical value “1” is substituted for the unused vector and calculated.
[0254]
In step S137, the vector calculation unit 62 calculates {similarity SimUP which is a cosine distance between the program vector PP and the positive history vector UP}-{similarity SimMUP} which is a cosine distance between the program vector PP and the negative history vector. Then, program vectors PP or EPG data of a predetermined number (for example, three) of higher-ranked programs are extracted as recommendation information, output to the recommendation information output unit 49, registered in the recommended program list 50, and The data is output to the display device 11 and the recording / reproducing device 12, and the process ends.
[0255]
By such processing, the weighting is performed using the effect vector for each large item, so that it is possible to generate recommendation information that matches the user's preference in detail.
[0256]
Next, with reference to the flowchart of FIG. 21, the positive history vector UP and the negative history vector MUP for each genre generated by the processing of the positive history vector and the negative history vector generation processing 2 described with reference to FIG. The matching process 5 for executing the matching process using the genre-specific user side effect vector EfUP and the user side countereffect vector EfMUP will be described.
[0257]
In step S141, the same processing as in step S121 in FIG. 19 is executed, and the usage setting contents of the effect vector are obtained.
[0258]
In step S142, the vector calculation unit 62 extracts the genre of the supplied program vector PP. Here, for example, it is assumed that the genre of the supplied program vector PP is “drama”.
[0259]
In step S143, the vector calculation unit 62 determines the program vector PP and the positive vector of the supplied program vector PP and the major items of the primary history vector UP read from the primary history storage unit 47 and corresponding to the genre “drama”. Calculate the cosine distance of the history vector UP. Here, the effect vector is not used for the calculation.
[0260]
In step S144, the vector calculation unit 62 multiplies the cosine distance calculated for each item by the user-side effect vector corresponding to the genre “drama” as necessary using the following equation (9). The obtained values are summed to calculate the similarity SimUP.
[0261]
SimUP = epd_t・ Eud_td・ Cosθu_td+ epd_g・ Eud_gd・ Cosθu_gd+ epd_p・ Eud_pd・ Cosθu_pd+ epd_a・ Eud_ad・ Cosθu_ad+ epd_k・ Eud_kd・ Cosθu_kd... (9)
[0262]
In equation (9), the program vector PP = (p_t, P_g, P_p, P_a, P_k) And the positive history vector UP = (u) corresponding to the genre “drama”_t, U_g, U_p, U_a, U_k) And the cosine distance of each major item is (cosθu_td, Cosθu_gd, Cosθu_pd , Cosθu_ad, Cosθu_kd), And the program side effect vector EfPP = (epd_t, Epd_g, Epd_p, Epd_a, Epd_k), User side effect vector EfUP corresponding to the genre “drama” = (eud_td, Eud_gd, Eud_pd, Eud_ad, Eud_kd). In addition, in the equation (9), both the program side effect vector EfPP and the user side effect vector EfUP are described. However, depending on the setting, the program side effect vector EfPP and the user side effect vector EfUP are used. Is not used, the numerical value “1” is substituted for the vector not used and calculated.
[0263]
In step S145, the vector calculation unit 62 compares the similarity SimUP between the program vector PP and the normal history vector UP calculated in step S143, and determines a predetermined number of programs such as 10 from the highest similarity. Extract the vector PP.
[0264]
In step S146, the vector calculation unit 62 determines whether the program vector PP extracted by the processing in step S145 and the negative history vector MUP corresponding to the genre “drama” read from the negative history storage unit 48 have the program Calculate the cosine distance between the vector PP and the negative history vector MUP. Here, the effect vector is not used for the calculation.
[0265]
In step S147, the vector operation unit 62 multiplies the cosine distance calculated for each item by the effect vector corresponding to the genre “drama” as necessary using the following equation (10), and The similarity SimMUP is calculated by summing the calculated values.
[0266]
SimMUP = epd_t・ Emd_td・ Cosθm_td+ epd_g・ Emd_gd・ Cosθm_gd+ epd_p・ Emd_pd・ Cosθm_pd+ epd_a・ Emd_ad・ Cosθm_ad+ epd_k・ Emd_kd・ Cosθm_kd... (10)
[0267]
In equation (10), the program vector PP = (p_t, P_g, P_p, P_a, P_k) And the negative history vector MUP = (m_t, M_g, M_p, M_a, M_k) And the cosine distance of each major item is (cosθm_td, Cosθm_gd, Cosθm_pd , Cosθm_ad, Cosθm_kd), And the program side effect vector EfPP = (epd_t, Epd_g, Epd_p, Epd_a, Epd_k), The user side reaction vector EfMUP corresponding to the genre "drama" = (emd_td, Emd_gd, Emd_pd, Emd_ad, Emd_kd). Further, in the equation (10), both the program side effect vector EfPP and the user side effect vector EfMUP are described. However, depending on the setting, the program side effect vector EfPP and the user side effect vector EfMUP are used. If any of is not used, the numerical value "1" is substituted for the vector not used and calculated.
[0268]
In step S148, the vector calculation unit 62 calculates {similarity SimUP, which is the cosine distance between the program vector PP and the positive history vector UP}-{similarity SimMUP, which is the cosine distance between the program vector PP and the negative history vector}. Then, program vectors PP or EPG data of a predetermined number (for example, three) of higher-ranked programs are extracted as recommendation information, output to the recommendation information output unit 49, registered in the recommended program list 50, and The data is output to the display device 11 and the recording / reproducing device 12, and the process ends.
[0269]
By such processing, the cosine distance between the positive history vector UP and the negative history vector MUP and the program vector PP for each genre is obtained for each large item, and weighting is performed using the effect vector corresponding to the genre. Since the similarity is calculated, it is possible to generate recommendation information that matches the user's preference in detail.
[0270]
As described above, the user-side effect vector EfUP and the counter-effect vector EfMUP may be generated based on the initial registration contents of the user registered in the initial registration storage unit 45, The user-side effect vector EfUP and the countereffect vector EfMUP unique to the user may be generated by counting the vector UP or the negative history vector MUP, or the programs watched by the user during a certain period.
[0271]
The user-side effect vector generation processing 1 for generating the user-side effect vector EfUP by counting the programs watched by the user during a certain period will be described with reference to the flowchart of FIG.
[0272]
In step S151, the user information registration unit 63 of the matching processing unit 43 selects one of the unprocessed large items.
[0273]
In step S152, the user information registration unit 63 refers to the main history stored in the main history storage unit 47, and refers to the program viewed by the user during a certain period such as one week, one month, or three months. And causes the program vector extraction unit 42 to extract, from the data acquisition unit 41, the program vector PP corresponding to the program watched by the user during the certain period, and to output the detailed program included in the large items selected in step S151. Count the number of programs.
[0274]
Specifically, for example, the large item selected in step S151 is a large item genre Gm = {drama, variety, sports, movie, music, children / education, education / documentation, news / report, etc.}. For example, the user information registration unit 63 classifies the components of the program vector PP of the program watched by the user during a certain period into items corresponding to each, and counts them. If the number of programs viewed by the user during a certain period is, for example, 50, the count result of the number of programs is, for example, genre Gm = (10, 18, 5, 2, 8, 1, 0, 1, 5, 5). ).
[0275]
In step S153, the user information registration unit 63 causes the program vector extraction unit 42 to extract the program vectors PP corresponding to all programs in the same period from the data acquisition unit 41, and includes the program vectors PP in the large item selected in step S151. The number of programs is counted for each detailed item.
[0276]
Specifically, for example, the large item selected in step S151 is a large item genre Gm = {drama, variety, sports, movie, music, children / education, education / documentation, news / report, etc.}. For example, the user information registration unit 63 classifies the components of the program vector PP of all programs in the same period into items corresponding to each, and counts them. If all programs in the same period are, for example, 1000 programs, the count result of the number of programs is, for example, genre Gm = (104, 239, 68, 25, 78, 91, 60, 254, 81). .
[0277]
In step S154, the user information registration unit 63 calculates the count of the user's viewing performance / the count of all programs based on the count results of steps S152 and S153.
[0278]
It is considered that the programming reflects the public preference due to the influence of audience competition. That is, the calculation of the count of the user's viewing results / the count of all programs is, in other words, synonymous with normalizing the count of the user's viewing results by the count of all programs as a standard model. . The normalized vector calculated in step S154 is referred to as a normalized vector D.
[0279]
For example, in the major item genre Gm = {drama, variety, sports, movie, music, children / education, culture / documentation, news / reporting, etc.}, the count number of all programs in one week is (8, 12, 3, 7, 6, 4, 2, 8, 10), and the count number of the program viewed by the user is (4, 0, 1, 2, 3, 4, 5, 5, 2). In this case, the normalized vector D is as follows.
D = (4/8, 0/12, 1/3, 2/7, 3/6, 4/4,
1/2, 2/8, 2/10)
= (0.5,0,0.33,0.28,0.5,1.0,
0.5, 0.13, 0.2)
That is, when the component of the normalized vector D is 1.0, all the programs of the corresponding item have been viewed during the set period, and when the component of the normalized vector D is 0, This means that no corresponding program was viewed during the set period.
[0280]
In step S155, the user information registration unit 63 generates an effect vector of the corresponding large item based on the calculation result of step S154.
[0281]
In order to generate an effect vector, using any one of the major item genre Gm = {drama, variety, sports, movie, music, children / education, culture / documentation, news / reporting, etc.} Set the standard value. For example, in a week of the program of the genre: “drama”, it is assumed that a general user views about 20% of the program, and the standard value may be set to 0.2. Since the effect vector of the large item is calculated as a relative value, the set value may be any value from 0 to 1. The user-side effect vector is a relative value between the normalized vector D calculated in step S154 and the set value.
[0282]
Therefore, the effect vector E of the large item genre Gm, which indicates the user's genre of interest, is
E = (0.3, -0.2, 0.13, 0.08, 0.3,
0.8, 0.3, -0.07, 0.0)
Therefore, it can be determined that the corresponding user likes the genre of “for children / education” and does not like the genre of “variety”.
[0283]
In step S156, the user information registration unit 63 determines whether the effect vectors of all large items have been generated. If it is determined in step S156 that the effect vectors of all large items have not been generated, the process returns to step S151, and the subsequent processes are repeated.
[0284]
If it is determined in step S156 that the effect vectors of all large items have been generated, in step S157, the user information registration unit 63 stores the effect vectors of all large items, and the process ends.
[0285]
Through such processing, a difference between general preferences and preferences unique to the user can be obtained. In addition, by re-calculating the user-side effect vector EfUP every predetermined period, for example, three months or half a year, it is possible to recommend a program reflecting the user's preferences in real time.
[0286]
Further, in the processing of FIG. 22, for example, it has been described that the user side effect vector EfUP is obtained based on a program watched by the user during a certain period such as one week, one month, or three months. For example, the user-side effect vector EfUP corresponding to the short-term, middle-term, and long-term may be calculated, and the recommendation information may be determined using the plurality of effect vectors.
[0287]
Here, the case where the user-specific preference is used as the user-side effect vector EfUP has been described, but the user-specific preference may be used as the normal history vector UP in the matching process.
[0288]
Also, instead of all programs being broadcast, all programs broadcast in a predetermined time zone (for example, so-called golden time from 18:00 to 22:00) in which the viewer views the program most are counted. You may. By doing so, it is possible to greatly reduce the number of calculation processes for obtaining general preferences.
[0289]
Next, with reference to the flowchart of FIG. 23, the difference between the user-specific preference and the public preference is used for matching by calculating a cosine distance indicating the similarity between the positive history vector UP and the public preference. User-side effect vector calculation processing 2 for obtaining the user-side effect vector EfUP for this purpose will be described.
[0290]
In step S161, the user information registration unit 63 of the matching processing unit 43 acquires the main history vector UP stored in the main history storage unit 47.
[0291]
In step S162, the user information registration unit 63 acquires a standard preference vector APP indicating a general preference.
[0292]
The standard preference vector APP may be supplied from the distribution server 5, or the programming is considered to reflect the preference of the general public due to the influence of the audience rating competition. In the same manner as the user-side effect vector calculation process 1 described above, the contents of all programs broadcast in a certain period may be counted, and if necessary, may be normalized to obtain a standard preference vector APP.
[0293]
The distribution server 5 may generate the standard preference vector APP indicating the general preference using, for example, a general audience rating survey or another method.
[0294]
In step S163, the user information registration unit 63 calculates a cosine distance between the standard preference vector APP and the normal history vector UP for each large item. The larger the cosine distance, the higher the similarity between the standard preference vector APP and the normal history vector UP.
[0295]
In step S164, based on the cosine distance calculated in step S163, the user information registration unit 63 generates an effect vector EfUP by setting the cosine distance to the reciprocal for each large item, and the process ends. The larger the reciprocal of the cosine distance, the lower the similarity between the standard preference vector APP and the normal history vector UP.
[0296]
Through such processing, the user-side effect vector EfUP that reflects the difference between the general preference and the corresponding user-specific preference can be obtained. When the program recommendation process is performed using the user side effect vector EfUP, the difference between the user's preference and the general preference is emphasized, and the recommended program is determined.
[0297]
Here, the program vector PP and the main history vector UP have been described as being represented by vectors for each major item. However, the program vector PP and the main history vector UP have one It is needless to say that the same processing can be executed even in the case of being represented by vectors arranged in columns.
[0298]
In addition, the similarity between the standard preference vector APP and the normal history vector UP may be used not only to calculate the effect vector, but also as an index indicating the uniqueness of the user directly to recommend a program. . For example, when the degree of similarity between the standard preference vector APP and the normal history vector UP is high, a new program or the like that is popular with the general public may be preferentially recommended.
[0299]
As described with reference to FIGS. 22 and 23, it is preferable that the user-side effect vector EfUP be obtained by learning based on the operation history of the user, but the user-side effect vector is registered in advance as initial registration. May be used, or a preset value obtained by experience or the like may be used.
[0300]
It should be noted that the user-side effect vector EfUP may be generated not only by generating the user-side effect vector EfUP by paying attention to the large item but also by paying attention to the constituent elements constituting the large item. For example, in the constituent element “Performer Pm” constituting the major item, the leading role and the supporting role can be distinguished, and in a drama or a movie, a user who gives priority to the casting of the supporting role over the leading role is assigned to the leading role by the supporting role The user side effect vector EfUP can be set so as to increase the weighting, and the director, director, original creator, photographer, etc. can be distinguished in the component "Screenplay / Original / Direction Am" that constitutes a major item. Alternatively, a user who attaches importance to a photographer rather than a director or a director may set the user-side effect vector EfUP so as to increase the weight of the photographer.
[0301]
In addition, the user-side effect vector EfUP is generated for each genre, and is used when matching the correct history vector UP and the program vector PP of the corresponding genre, as in the matching process 5 described with reference to FIG. May be.
[0302]
Next, a user-side effect vector generation process 3 for generating a user-side effect vector EfUP by counting programs watched by the user during a certain period by genre will be described with reference to the flowchart of FIG.
[0303]
In step S171, the user information registration unit 63 of the matching processing unit 43 selects one of the genres in order to count the programs watched by the user during a certain period by genre.
[0304]
In step S172, the user information registration unit 63 selects one of the unprocessed large items.
[0305]
In step S173, the user information registration unit 63 refers to the main history stored in the main history storage unit 47, and refers to the program viewed by the user during a certain period, such as one week, one month, or three months. Of the selected genre, the program vector extracting unit 42 extracts, from the data acquiring unit 41, the program vector PP corresponding to the program of the selected genre, from those watched by the user during a certain period. Then, the number of programs is counted for each detailed item included in the large item selected in step S172.
[0306]
In step S174, the user information registration unit 63 causes the program vector extraction unit 42 to extract the program vector PP corresponding to the selected genre from the data acquisition unit 41 among all programs in the same period. The number of programs is counted for each detailed item included in the selected large item.
[0307]
In step S175, the user information registration unit 63 calculates the count of the user's viewing performance / the count of all programs in the selected genre based on the count results in steps S173 and S174.
[0308]
As described above, it is considered that the programming reflects the public preference due to the influence of audience rating competition. That is, the calculation of the count of the user's viewing performance / the count of all programs in the selected genre is, in other words, the count of the user's viewing performance in the corresponding genre is calculated as the standard model's corresponding genre. This is synonymous with normalizing by the count number of all programs. The normalized vector calculated in step S175 is referred to as a normalized vector D '.
[0309]
For example, the large item genre Gm = {drama, variety, sports, movie, music, children / education, culture / documentation, news / reporting, etc.} In the item time zone Tm = {morning, noon, evening, golden, midnight}, the count number of all programs in one week is (10, 35, 7, 53, 17), and the When the count number is (5, 0, 0, 8, 4), the normalized vector D 'is as follows.
D '= (5/10, 0/35, 0/7, 8/53, 4/17)
= (0.5,0,0,0.28,0.15,0.24)
That is, if the component of the normalized vector D 'is 1.0, it means that all programs of the corresponding item have been watched during the set period, and the component of the normalized vector D' is 0. In the case of, it means that no corresponding program was viewed during the set period.
[0310]
In step S176, the user information registration unit 63 generates an effect vector of the corresponding large item in the selected genre based on the calculation result in step S175.
[0311]
In order to generate an effect vector, a standard value is set using one of the items of the large item time zone Tm = {morning, noon, evening, golden, midnight}. For example, in a week of the program “time period:“ Golden ”, it is assumed that a general user watches about 20% of the drama, and the standard value may be set to 0.2. Since the effect vector of the large item is calculated as a relative value, the set value may be any value from 0 to 1. The user-side effect vector is a relative value between the normalization vector D ′ calculated in step S175 and the set value.
[0312]
Therefore, the effect vector E ′ of the large item genre Gm indicating the user's genre of interest is
E '= (0.3, -0.2, -0.2, -0.05, 0.04)
Therefore, it can be determined that the corresponding user likes the drama in the morning time zone and does not like the drama in the daytime or the evening time zone.
[0313]
In step S177, the user information registration unit 63 determines whether or not the effect vectors of all the large items have been generated in the selected genre. If it is determined in step S177 that the effect vectors of all the large items have not been generated, the process returns to step S172, and the subsequent processes are repeated.
[0314]
When it is determined in step S177 that the effect vectors of all large items have been generated, in step S178, the user information registration unit 63 determines whether or not the processing of all genres has been completed. If it is determined in step S178 that not all genres have been completed, the process returns to step S171, and the subsequent processes are repeated.
[0315]
If it is determined in step S178 that all genres have been completed, in step S179, the user information registration unit 63 stores the effect vectors of all large items, and the process ends.
[0316]
By such a process, a difference between a general preference and a user-specific preference can be obtained for each genre. Further, in the same manner as described with reference to FIG. 22, the user-side effect vector EfUP is recalculated every predetermined period, for example, three months, six months, etc. The reflected program can be recommended.
[0317]
Also, in the processing of FIG. 24, for example, the user side effect vector EfUP is determined based on the program watched by the user during a certain period such as one week, one month, or three months. In the same manner as described above, for a plurality of periods, for example, a short-term, middle-term, and long-term effect vector EfUP corresponding to a long term is calculated, and the recommendation information is determined using the plurality of effect vectors. You may do it.
[0318]
Also in the processing of FIG. 24, instead of all the programs being broadcast, the programs are broadcast in a predetermined time zone (for example, so-called golden time from 18:00 to 22:00) when the viewer views the program most. All programs may be counted.
[0319]
Next, with reference to the flowchart of FIG. 25, the calculation of the cosine distance indicating the similarity between the normal history vector UP and the public preference is performed for each genre, so that the user-specific preference and the public preference are compared. The user-side effect vector generation processing 4 for obtaining the user-side effect vector EfUP for using the gap for matching will be described.
[0320]
In step S191, the user information registration unit 63 of the matching processing unit 43 selects one of the genres to specify the genre in which the processing is performed.
[0321]
In step S192, the user information registration unit 63 acquires the correct history vector UP of the selected genre from the correct history vectors UP stored in the correct history storage unit 47.
[0322]
In step S193, the user information registration unit 63 acquires the standard preference vector APP of the selected genre from the standard preference vectors APP indicating general preferences.
[0323]
As described above, the standard preference vector APP may be supplied from the distribution server 5, or the program composition is considered to reflect the public preference due to the influence of audience rating competition. In the same manner as the user-side effect vector calculation processing 3 described with reference to FIG. 24, the contents of all programs broadcasted in a certain period are counted by genre, normalized as necessary, and used as a genre-specific standard preference vector APP. You may do it.
[0324]
The distribution server 5 may generate a standard preference vector APP indicating a general preference for each genre using, for example, a general audience rating survey or another method.
[0325]
In step S194, the user information registration unit 63 uses the main history vector UP of the selected genre and the standard preference vector APP of the selected genre, and for each major item, the standard preference vector APP and the main history vector UP. Calculate the cosine distance between The larger the cosine distance, the higher the similarity between the standard preference vector APP and the normal history vector UP.
[0326]
In step S195, based on the cosine distance calculated in step S194, the user information registration unit 63 generates an effect vector EfUP of the selected genre for each of the large items by reversing the cosine distance. The larger the reciprocal of the cosine distance, the lower the similarity between the standard preference vector APP and the normal history vector UP.
[0327]
In step S196, the user information registration unit 63 determines whether or not processing for all genres has been completed. If it is determined in step S178 that not all genres have been completed, the process returns to step S191, and the subsequent processes are repeated. If it is determined in step S196 that all genres have been completed, the process is terminated.
[0328]
Through such processing, the user-side effect vector EfUP for each genre, which reflects the difference between the general preference and the corresponding user-specific preference, can be obtained.
[0329]
Further, in the same manner as the processing described with reference to FIG. 23, the similarity between the standard preference vector APP and the negative history vector MUP may be obtained, and the reciprocal thereof may be calculated as the counter effect vector EfMUP.
[0330]
Next, a user-side counter-effect vector calculation process 1 for obtaining a counter-effect vector EfMUP by comparing the negative history vector MUP with the public preference will be described with reference to the flowchart of FIG.
[0331]
In step S201, the user information registration unit 63 of the matching processing unit 43 acquires the negative history vector MUP stored in the negative history storage unit 48.
[0332]
In step S202, the user information registration unit 63 acquires a standard preference vector APP indicating a general preference.
[0333]
The standard preference vector APP may be supplied from the distribution server 5. Alternatively, it is considered that the programming reflects the public's preference due to the influence of the audience rating competition. Therefore, the program is broadcast for a certain period in the same manner as the user side effect vector calculation processing 1 described with reference to FIG. Alternatively, the contents of all the programs may be counted and, if necessary, normalized to obtain a standard preference vector APP.
[0334]
In step S203, the user information registration unit 63 calculates a cosine distance between the standard preference vector APP and the negative history vector MUP for each large item. The larger the cosine distance, the higher the similarity between the standard preference vector APP and the negative history vector MUP.
[0335]
In step S204, the user information registration unit 63 generates a counter-effect vector EfMUP by setting the cosine distance to the reciprocal for each large item based on the cosine distance calculated in step S203, and the process ends.
[0336]
By such processing, the reaction vector EfMUP can be generated, so that programs that the user does not like can be effectively omitted from recommended programs.
[0337]
Next, with reference to the flowchart of FIG. 27, the user-side counter-effect vector generation processing 2 executed for each genre will be described.
[0338]
In step S211, the user information registration unit 63 of the matching processing unit 43 selects one of the genres in order to specify the genre in which the processing is performed.
[0339]
In step S212, the user information registration unit 63 acquires the negative history vector MUP of the selected genre from the negative history vectors MUP stored in the negative history storage unit 48.
[0340]
In step S213, the user information registration unit 63 acquires the standard preference vector APP of the selected genre from the standard preference vectors APP indicating general preferences.
[0341]
In step S214, based on the negative history vector MUP of the selected genre and the standard preference vector APP of the selected genre, the user information registration unit 63 sets a standard preference vector APP and a negative history vector MUP for each large item. Calculate the cosine distance between The larger the cosine distance, the higher the similarity between the standard preference vector APP and the negative history vector MUP.
[0342]
In step S215, the user information registration unit 63 generates a counter-effect vector EfMUP based on the cosine distance calculated in step S214, by reversing the cosine distance for each large item. The larger the reciprocal of the cosine distance, the lower the similarity between the standard preference vector APP and the negative history vector MUP.
[0343]
In step S216, the user information registration unit 63 determines whether or not processing for all genres has been completed. If it is determined in step S216 that all genres have not been completed, the process returns to step S211 and the subsequent processes are repeated. If it is determined in step S216 that all genres have been completed, the process is terminated.
[0344]
By such processing, the reaction vector EfMUP can be generated for each genre, so that programs that the user does not like can be effectively omitted from recommended programs.
[0345]
The user-side effect vector EfUP and the counter-effect vector EfMUP are obtained by multiplying the reciprocal of the cosine distance for each large item described with reference to FIG. 23 and FIGS. 25 to 27 by n, or a predetermined digit. A rounded value may be used, or a value obtained by subtracting the reciprocal of the cosine distance from 1 or a value obtained by multiplying the value by n may be used.
[0346]
Here, it has been described that the program vector PP and the negative history vector MUP are represented by vectors for each large item. However, the program vector PP and the negative history vector MUP are arranged such that all elements of the detailed items are arranged in one column. It is needless to say that the same processing can be executed even in the case of being represented by arranged vectors.
[0347]
Also, a case where information indicating a group such as a group ID or a cluster code generated by the processing described with reference to FIG. 7 or 8 is added to the program vector extracted by the program vector extracting unit 42. There is.
[0348]
For example, if the program that the user prefers to watch is a serial drama, the corresponding serial drama can be recommended in all episodes, so that the arithmetic processing for the recommendation can be reduced. Also, by preferentially recommending a program in the same group as a program that is highly liked by the user, it is possible to reduce the number of calculation processes for recommendation.
[0349]
Next, a matching process including group recommendation will be described with reference to the flowchart in FIG.
[0350]
In step S221, the vector operation unit 62 of the matching processing unit 43 determines whether or not information indicating a group such as a group ID or a cluster code is added to the program vector PP supplied from the program vector extraction unit 42. Then, it is determined whether or not the corresponding program is grouped.
[0351]
If it is determined in step S221 that the corresponding programs are grouped, in step S222, the vector calculation unit 62 sets the group indicated by the group ID or the cluster code as a recommendation priority group in the user information registration unit. It is determined whether it is registered in 63.
[0352]
If it is determined in step S222 that the program is registered as a recommendation priority group, in step S223, the vector calculation unit 62 supplies information of the corresponding program to the recommendation information output unit 49 as recommendation information. The recommended information output unit 49 registers the recommended program in the recommended program list 50 and outputs information of the recommended program to the television display device 11 or the recording / reproducing device 12. It proceeds to S227.
[0353]
If it is determined in step S221 that the corresponding program is not grouped, or if it is determined in step S222 that the corresponding program is not registered as a recommendation priority group, in step S224, FIG. One of the matching processes 1 to 5 described above is executed.
[0354]
In step S225, the vector calculation unit 62 adds information indicating a group such as a group ID or a cluster code to the program vector of the program recommended in any one of the matching processes 1 to 3 performed in step S224. It is determined whether or not the recommended programs are grouped based on whether or not is added. If it is determined in step S225 that the recommended programs are not grouped, the process ends.
[0355]
If it is determined in step S225 that the recommended programs are grouped, in step S226, the vector calculation unit 62 sets the group ID or the cluster code added to the program vector as the recommended priority group. Registered in the user information registration unit 63 and stored.
[0356]
After the processing in step S223 or S226 is completed, in step S227, the vector calculation unit 62 refers to the operation log of the negative history stored in the negative history storage unit 48 and is recommended by the processing in step S223. Since the programs registered as the recommended priority group in the processing of step S226 are not grouped, the recommended programs are not accepted. Then, it is determined whether or not an operation input for giving a negative history, such as an operation input for instructing viewing or recording of other programs, or an instruction for erasing an automatically recorded program before reproduction, is received. If it is determined in step S227 that an operation input having a negative history has not been received, the process ends.
[0357]
If it is determined in step S227 that an operation input having a negative history has been received, in step S228, the vector calculation unit 62 removes the corresponding group from the registration of the recommended priority group of the user information registration unit 63, and performs processing. Is terminated.
[0358]
Note that a plurality of group IDs may be associated with one program vector PP by the title grouping process 1 described with reference to FIG. In such a case, in particular, in step S228, all group IDs may be removed from the registration of the recommended priority group in the user information registration unit 63 by an operation input that results in a negative history. In the case where a predetermined number of operation inputs having a negative history are performed, the user ID registration unit 63 registers the recommended priority group, and the corresponding group ID becomes You may make it take off. The same applies to the case where one group ID is associated with one program vector PP.
[0359]
By such processing, it is possible to recommend a series drama that the user prefers to watch in all episodes or to recommend a program in the same group as a program that is highly liked by the user. Therefore, the number of calculation processes for recommendation can be reduced.
[0360]
As described above, when the matching process including the group recommendation is performed, the positive history vector UP and the negative history vector MUP generated based on the user's operation history acquired from the television display device 11 or the recording / reproducing device 12 May be generated for each corresponding group.
[0361]
Also, based on the correct history of the user's operation history acquired from the television display device 11 or the recording / reproducing device 12, the number of viewing or recording reservations in the same drama is counted, and the number has reached a predetermined number or more. In this case, the recommendation may be preferentially performed without performing the matching process.
[0362]
Instead of using the user history to perform program recommendation processing, a user model may be set based on a topic specified in advance by the user, and program recommendation processing may be performed based on the user model.
[0363]
The user model is obtained by extracting a program corresponding to a topic from a predetermined program group by performing filtering using a topic specified in advance by the user. The initial registration storage unit 45 stores the topic input by the user. There may be a plurality of registered topics, and the topics are updated as appropriate according to a user's operation input. The vector calculation unit 62 extracts a program vector including a topic stored in the initial registration storage unit 45 from among the program vectors of the programs to be created by the user model, which are supplied from the program vector extraction unit 42. A user model vector is generated by normalizing the sum of the vectors as necessary, and registered in the user information registration unit 63.
[0364]
Thus, for example, when the time zone Hm = “after 23:00” and the genre Gm = “variety” are specified as the topics, the filtering is performed using these topics, and the user model “late night variety” is generated. Is done. The user model “Midnight Variety” includes, as a component, a comedian who has appeared in a variety program broadcasted in the late night frame after 23:00. Since the program in which the comedian performs is extracted and recommended in the matching process, it is possible to apply the user's preference beyond the genre and the like as compared with the case where the matching process is performed for each item.
[0365]
The program for which the user model is to be created may be, for example, all programs in a predetermined period or a group of programs broadcast in a predetermined time period (for example, so-called golden time).
[0366]
Also, for example, by using the same filtering conditions to perform filtering on different sets of programs, user model vectors corresponding to different programming in detail, for example, at different times or different time periods, are generated. be able to.
[0367]
Specifically, even under the same condition "Genre Gm = Music", there is a difference in broadcast time and performers between the current programming and the programming 10 years ago. A different user model of "previous music fan" can be generated. As a result, a pop singer who has been listening to “music fans 10 years ago” recommends dramas and movies currently appearing as actors, or “music fans 10 years ago” likes this. It is possible to recommend a program in which the current singer sings the song that was being listened to.
[0368]
In addition, for example, the same “genre Gm” is used in a time zone such as 20:00 to 24:00, which is mainly watched by members of society, and in a time zone such as 15:00 to 20:00, which is mainly watched by children. Even under the condition of “= music”, different program groups can be obtained by filtering, so that different user models can be generated.
[0369]
Then, the vector calculation unit 62 calculates the similarity between the program vector PP of the program supplied from the program vector extraction unit 42 and the user model vector registered in the user information registration unit 63, and based on the result. , Recommendation information is generated and supplied to the recommendation information output unit 49. The recommended information output unit 49 registers the recommended information in the recommended program list 50 and supplies the recommended information to the television display device 11 or the recording / reproducing device 12.
[0370]
Next, the matching processing using the user model will be described with reference to the flowchart in FIG.
[0371]
In step S231, the vector operation unit 62 of the matching processing unit 43 acquires the program vector of the program for which a user model is to be created, extracted by the program vector extraction unit 42. The program for which the user model is to be created is, for example, a program broadcast during a predetermined period such as one month or three months, and is broadcast during a predetermined time period (eg, golden time) during the predetermined period. The program may be a program broadcasted during a predetermined period in the past, such as 10 years ago, 20 years ago, or the like.
[0372]
In step S232, the initial registration storage unit 45 acquires the topic input by the user from the operation input unit 44, and stores the topic. The vector calculation unit 62 reads a topic from the initial registration storage unit 45 and sets it as a filtering condition. The filtering condition may be described as, for example, a hash table.
[0373]
In step S233, the vector calculation unit 62 filters the program vector PP acquired in step S231 based on the filtering conditions set in step S232, and extracts a program vector PP matching the topic. For example, if the program for which the user model is to be created is all the programs for the past three months, the vector calculation unit 62 determines that “the title Tm or the genre Gm = soccer” for all the program vectors PP for the past three months. Filtering is performed under the condition.
[0374]
In step S234, the vector calculation unit 62 obtains the sum total of the program vectors PP extracted by the filtering in step S233, and sets the sum as a user model vector. For example, when filtering is performed under the condition of “title Tm or genre Gm = soccer”, a user model vector of a user model “soccer fan” is generated.
[0375]
In step S235, the vector calculation unit 62 becomes a candidate for obtaining the recommendation information extracted in the program vector extraction unit 42 and the user model vector generated in step S234 (will be broadcast later). The cosine distance between the program and the program vector PP is calculated.
[0376]
In step S236, the vector calculation unit 62 compares the similarity between the user model vector and the program vector PP of the candidate program for obtaining recommendation information based on the cosine distance calculated in step S235, and Based on the above, a predetermined number of programs from the highest similarity are extracted as recommendation information from those having a high degree of similarity, and supplied to the recommendation information output unit 49. The recommended information output unit 49 registers the recommended information in the recommended program list 50 and supplies the recommended information to the television display device 11 or the recording / reproducing device 12, and the process ends.
[0377]
By such processing, even when there is no operation history of the user, a program matching the topic set by the user can be recommended. In the related art, for example, when a condition of “title Tm or genre Gm = soccer” is set and a matching process is performed, it is not possible to extract a variety or drama in which a soccer player appears. On the other hand, in the processing described with reference to FIG. 29, by setting a user model on the topic “Title Tm or genre Gm = soccer”, items such as performers and contents of the user model vector are set. Since many soccer player names are included, even if the program does not meet the condition of "title Tm or genre Gm = soccer", it extracts variety or drama in which soccer players appear. It becomes possible to recommend.
[0378]
Here, the description has been made assuming that the similarity is obtained by calculating the cosine distance between the user model vector and the program vector. However, the cosine distance is calculated individually for each large item, and the sum is used. The similarity may be obtained.
[0379]
Note that the generation process of the user model vector may be executed in the distribution server 5. In this case, for example, the processing of steps S231 to S234 of FIG. 29 may be performed using the program vector generated by the program vector generation unit 23 described with reference to FIG.
[0380]
In addition to recommending a program that suits the user's preference, selecting a program with low similarity to both the positive history vector UP and the negative history vector MUP means that the user does not like or dislike it. There is a high possibility that a program having a feature that has not been performed, in other words, a program that the user does not like to eat is recommended. As described above, in order to advance the learning by extracting the user's preference, it is important that the user evaluate a program having a feature that the user has not watched before.
[0381]
As a result, it is possible to give unexpectedness to the recommendation result provided to the user and to increase the interest of the user, and it is very important to recommend a program that more closely matches the user's preference. History information can be obtained.
[0382]
Next, the exception recommended program selection processing will be described with reference to the flowchart in FIG.
[0383]
In step S241, the vector operation unit 62 of the matching processing unit 43 includes the program vector PP supplied from the program vector extraction unit 42, the positive history vector UP stored in the positive history storage unit 47, and the negative history storage unit. 48, the cosine distance between the positive history vector UP and the program vector PP, and the cosine distance between the negative history vector MUP and the program vector PP, for each major item, calculate.
[0384]
In step S242, the vector calculation unit 62 sums the values of the cosine distance calculated for each item on the positive history side and the negative history side. That is, the similarity SimUP between the positive history vector UP and the program vector PP and the similarity SimMUP between the negative history vector MUP and the program vector PP are calculated by the processes of steps S241 and S242.
[0385]
In step S243, the vector calculation unit 62 calculates an exception recommendation value indicating a low degree of similarity with the positive history vector UP and the negative history vector MUP.
[0386]
Specifically, the exception recommendation value can be obtained by (1−SimUP) × (1−SimMUP) or (1 / SimUP) × (1 / SimMUP).
[0387]
In step S244, the vector calculation unit 62 obtains a program with a high exceptional recommendation value based on the calculation result in step S243, extracts the program as recommendation information, and ends the process.
[0388]
By such a process, it is possible to extract and recommend programs having characteristics that the user has never watched before, so that unexpectedness is given to the selection of a recommended program to the user, and expansion of the user's interest is provided. Not only can it be connected, but also very important history information can be obtained in order to recommend a program that more closely matches the user's preference.
[0389]
Through the processing described with reference to FIGS. 13 to 30, recommendation information is generated in the program recommendation processing device 10 and supplied to the television display device 11 or the recording / reproducing device 12.
[0390]
The television display device 11 or the recording / reproducing device 12 is supplied with a satellite wave or terrestrial broadcast signal received and decoded by the television receiver 4.
[0391]
The television display device 11 displays a broadcast signal supplied from the television reception device 4 or reproduction data supplied from the recording / reproduction device 12 based on a user's operation input, Based on the supplied recommendation information, recommended program information is displayed and automatic channel setting is executed. Further, the television display device 11 supplies the operation log to the program recommendation processing device 10.
[0392]
Further, the recording / reproducing device 12 records or reserves a broadcast signal supplied from the television receiving device 4 based on a user's operation input, or, based on recommendation information supplied from the program recommendation processing device 10, Automatically record programs. Further, the recording / reproducing apparatus 12 reproduces a program recorded on a mounted recording medium or a built-in recording medium, and outputs the program to the television display device 11 for display. Further, the recording / reproducing device 12 supplies the operation log to the program recommendation processing device 10.
[0393]
FIG. 31 is a block diagram illustrating a configuration of the television receiver 4. The television receiver 4 will be described as a general receiver that complies with the standard for digital broadcast receivers.
[0394]
The satellite wave detector 91 selects a satellite wave transmitted via the satellite 2 and received by the antenna 3 based on a signal for selecting a channel supplied from the television display device 11 or the recording / reproducing device 12. The signal is transmitted to a transmission and multiplexing configuration control (TMCC) decoding unit 92 and a demodulation / decoding processing unit 93 is supplied to the broadcast signal.
[0395]
The TMCC decoding unit 92 receives input of information such as a transmission mode (modulation method, coding rate, etc.) and a slot in the transmission multiplex control signal, decodes the information, and supplies the information to the demodulation / decoding processing unit 93.
[0396]
The demodulation / decoding processing unit 93 converts the supplied broadcast signal into, for example, QPSK (quadri-phase shift keying: four-phase modulation scheme or four-phase modulation scheme) based on the information on the transmission mode supplied from the TMCC decoding unit 92. The signal is demodulated and decoded by using a method such as a phase PSK method or an 8-phase PSK method, and is supplied to the deinterleaver 94.
[0397]
The deinterleaver 94 deinterleaves the supplied signal and supplies it to the error correction processing unit 95. Further, the deinterleaver 94 may further perform frame separation and descrambling processing on the supplied signal.
[0398]
The error correction processing unit 95 performs an error correction process using, for example, a Reed-Solomon code and supplies the result to a CA (Conditional Access: conditional access) descramble unit 101.
[0399]
The terrestrial wave detection unit 96 selects and detects the terrestrial wave received by the antenna 3 based on a control signal for selecting a channel, which is supplied from the television display device 11 or the recording / reproducing device 12, and performs a transmission mode. The control signal is supplied to the TMCC decoding unit 97 and the broadcast signal is supplied to the demodulation / decoding processing unit 98.
[0400]
The TMCC decoding unit 97 receives input of information such as a transmission mode (modulation scheme, coding rate, etc.), a slot, and a TS in the transmission multiplex control signal, decodes the information, and supplies the information to the demodulation and decoding processing unit 98. .
[0401]
The demodulation / decoding processing unit 98 converts the supplied broadcast signal based on the information on the transmission mode supplied from the TMCC decoding unit 97 using a method such as a QAM (quadrature amplitude modulation) method. Demodulated and decoded, and supplied to a deinterleaver 99.
[0402]
The deinterleaver 99 deinterleaves the supplied signal and supplies the signal to a TS (Transport Stream) reproducing unit 100. Further, the deinterleaver 99 may further perform frame separation and descrambling processing on the supplied signal.
[0403]
The TS reproducing unit 100 reproduces the transport stream based on the supplied signal and supplies the transport stream to the CA descramble unit 101.
[0404]
The CA descrambler 101 descrambles the limited reception signal based on the signal supplied from the error correction processor 95 or the TS reproducer 100 and supplies the descrambled signal to the demultiplexer 102.
[0405]
The data input unit 103 receives input of EPG data from the EPG receiving device 9, receives streaming data from the distribution server 5 via the network 8, and supplies the streaming data to the demultiplexer 102.
[0406]
The demultiplexer 102 demultiplexes the signal supplied from the CA descramble unit 101 or the data input unit 103, the audio signal to the audio signal decoding unit 104, the video signal to the video signal decoding unit 105, and , Control signals and data such as EPG are supplied to the data decoding unit 106.
[0407]
The audio signal decoding unit 104 decodes the supplied audio signal and supplies it to the television display device 11 or the recording / reproducing device 12. The video signal decoding unit 105 decodes the supplied video signal and supplies the decoded video signal to the television display device 11 or the recording / reproducing device 12. The data decoding unit 106 decodes the supplied control signal and data such as the EPG and supplies the decoded data to the television display device 11 or the recording / reproducing device 12.
[0408]
By the processing of the television receiver 4, the received satellite wave or terrestrial wave or the distributed streaming data is demodulated and decoded by a predetermined method, and the television display device 11 or the recording / reproducing device 12 Supplied to
[0409]
FIG. 32 is a block diagram illustrating a configuration of the television display device 11.
[0410]
The operation input unit 121 receives an operation input from the user, supplies a control signal corresponding to the user's operation input to each unit of the television display device 11, and supplies and saves the operation contents of the user to the operation log list 122. Let it. The operation log of the user stored in the operation log list 122 is acquired by the operation log acquisition unit 46 of the program recommendation processing device 10 described with reference to FIG. When the user's operation input is a channel selection, the operation input unit 121 supplies the input user operation to the channel setting unit 123.
[0411]
The channel setting unit 123 generates a control signal indicating a channel selection based on the signal indicating the user's operation input supplied from the operation input unit 121 and supplies the control signal to the television receiver 4. Further, the channel setting unit 123 generates a control signal indicating channel selection in order to automatically set a channel based on recommendation information registered in a recommended program list 128 to be described later. 4 The television receiver 4 receives a broadcast signal of a specified channel based on the control signal.
[0412]
The data input unit 124 receives an input of a broadcast signal from the television receiver 4 and supplies the broadcast signal to the image processing unit 125. The image processing unit 125 performs image processing on the supplied broadcast signal based on an image display method of the output unit 126, and supplies the broadcast signal to the output unit 126. The output unit 126 includes, for example, a display device such as a CRT (Cathode Ray Tube) or an LCD (Liquid Crystal Display), and an audio output device such as a speaker. Is displayed on the display device, and an audio signal is output from the audio output device.
[0413]
The recommended program list acquisition unit 127 acquires recommendation information from the program recommendation processing device 10 and supplies the recommended information to the recommended program list 128. The recommended program list 128 registers the supplied recommendation information. The recommended information registered in the recommended program list 128 is read by the channel setting unit 123 or by the recommended program information display control unit 129.
[0414]
The recommended program information display control unit 129 outputs the recommended information read from the recommended program list 128 to the image processing unit 125 in order to present the recommended information of the program to the user. The image processing unit 125 outputs the recommendation information supplied from the recommended program information display control unit 129 to the output unit 126 alone or by superimposing the recommendation information on the image of the broadcast signal supplied from the data input unit 124. Is displayed on the display device.
[0415]
Whether the recommended information is displayed on the output unit 126 or the channel is automatically set based on the recommended information supplied from the program recommendation processing device 10 is determined by the setting of the user. May be.
[0416]
With reference to the flowchart in FIG. 33, a recommendation information display process for presenting program recommendation information to a user based on recommendation information supplied from the program recommendation processing device 10 will be described.
[0417]
In step S251, the recommended program list acquisition unit 127 acquires the recommendation information output from the recommendation information output unit 49 of the program recommendation processing device 10.
[0418]
In step S252, the recommended program list acquisition section 127 registers the acquired recommendation information in the recommended program list 128.
[0419]
In step S253, the recommended program information display control unit 129 reads, from the recommended program list 128, recommendation information of a program broadcast within a predetermined time from the current time, for example, three hours or one day, and recommends the program to be recommended. , Generates recommendation information display data for displaying the title, content, broadcast time, broadcast channel, and the like, and supplies the data to the image processing unit 125.
[0420]
In step S254, the image processing unit 125 executes image processing for displaying the supplied recommendation information display data on the output unit 126, and supplies the data to the output unit 126. The recommendation information is subjected to image processing alone or by being superimposed on the image of the broadcast signal supplied from the data input unit 124.
[0421]
In step S255, the output unit 126 displays the recommendation information supplied from the image processing unit 125, and the process ends.
[0422]
By such processing, the recommendation information is displayed on the output unit 126, so that the user can select a program to view by referring to the displayed recommendation information.
[0423]
Next, referring to the flowchart of FIG. 34, an automatic channel for automatically setting a channel based on the recommendation information supplied from the program recommendation processing device 10 and displaying a program matching the user's preference. The setting process will be described.
[0424]
In steps S271 and S272, the same processing as in steps S251 and S252 described with reference to FIG. 33 is performed. That is, the recommended information output from the recommended information output unit 49 of the program recommendation processing device 10 is acquired, and the acquired recommended information is registered in the recommended program list 128.
[0425]
In step S273, the channel setting unit 123 obtains information on the recommended program corresponding to the current time from the program recommendation list 98. In step S274, the channel setting unit 123 generates channel setting information based on the information on the recommended program. Output to the receiving device 4. The television receiver 4 receives a broadcast signal of a specified channel based on the control signal.
[0426]
In step S275, the data input unit 124 acquires a broadcast signal of the designated channel from the television receiver 4, and supplies the broadcast signal to the image processing unit 125.
[0427]
In step S276, the image processing unit 125 performs image processing for displaying the supplied broadcast signal on the output unit 126, and supplies the broadcast signal to the output unit 126.
[0428]
In step S277, the output unit 126 displays the video of the recommended program supplied from the image processing unit 125, outputs sound, and the process ends.
[0429]
By such processing, a channel is automatically set to a channel on which a program matching the user's preference is being broadcast. The automatic channel setting process described with reference to FIG. 34 may be executed, for example, when the user instructs. This makes it possible to automatically set a channel to a program on which an appropriate program is being broadcast, for example, when the user is worried about which program to watch.
[0430]
In addition, the automatic channel setting process described with reference to FIG. 34 is performed, for example, at predetermined time intervals, such as 2 hours, in which a channel is not specifically set and it can be determined that the channel is left alone, or It may be executed when there is no operation input from the user for a predetermined period of time when it can be determined that the user has been left unattended.
[0431]
It should be noted that, in order to prevent the channel from being automatically switched while the user is consciously watching the program, a mode in which the automatic channel setting process cannot be executed is prepared, and the user may give an instruction. You may be able to.
[0432]
FIG. 35 is a block diagram illustrating a configuration of the recording / reproducing device 12.
[0433]
The operation input unit 141 receives an operation input from the user, supplies a control signal corresponding to the user's operation input to each unit of the recording / reproducing apparatus 12, and supplies and saves the operation contents of the user to the operation log list 142. . The user operation logs stored in the operation log list 142 are acquired by the operation log acquisition unit 46 of the program recommendation processing device 10 described with reference to FIG.
[0434]
The recording setting unit 143 starts broadcast of a program to be recorded, for example, based on a signal indicating a user operation input supplied from the operation input unit 141 or from recommendation information registered in a recommended program list 149 described later. Information necessary for performing a recording process, such as a time, a broadcast end time, and a broadcast channel, is extracted. When the user's operation input supplied from the operation input unit 141 is a recording reservation, the recording setting unit 143 registers information necessary for performing a recording process in the recording reservation list 144 and is supplied from the operation input unit 141. If the user's operation input is a recording process of a currently broadcasted program or if an automatic recording process is performed using recommendation information registered in a recommended program list 149 described later, the recording process is performed. Is supplied to the recording control unit 145.
[0435]
The recording control unit 145 performs recording based on information necessary for performing a recording process supplied from the recording setting unit 143 or recording corresponding to the current time in the recording reservation information registered in the recording reservation list 144. By extracting the reservation information, a control signal indicating the broadcast channel of the program to be recorded is generated and supplied to the television receiver 4, and a control signal for executing the recording process is generated. 147. The television receiver 4 receives a broadcast signal of a specified channel based on the control signal.
[0436]
The data input unit 146 receives an input of a broadcast signal from the television receiver 4 and supplies the broadcast signal to the recording / playback processing unit 147. The recording / playback processing unit 147 is configured such that, for example, a recording medium such as a magnetic tape, an optical disk, a magnetic disk, a magneto-optical disk, or a semiconductor memory can be mounted therein, or a hard disk or It has a recording medium such as a semiconductor memory so that information can be recorded on the recording medium and the information can be reproduced from the recording medium. Specifically, when the recording medium attachable to the recording / reproducing processing unit 147 is a magnetic tape, the recording / reproducing processing unit 147 has a magnetic head and transmits the broadcast signal supplied from the data input unit 146 to the magnetic tape. The information is recorded (that is, recorded) or the information recorded on the magnetic tape is reproduced and supplied to the television display device 11 or the like for reproduction and output.
[0437]
The recommended program list acquisition unit 148 acquires recommendation information from the program recommendation processing device 10 and supplies it to the recommended program list 149. The recommended program list 149 registers the supplied recommended information. The recommendation information registered in the recommended program list 149 is read out to the recording setting unit 143, and the recording processing is automatically executed.
[0438]
Next, the automatic recording process will be described with reference to the flowchart in FIG.
[0439]
In step S291, the recommended program list acquisition unit 148 acquires the recommendation information output from the recommendation information output unit 49 of the program recommendation processing device 10.
[0440]
In step S292, the recommended program list acquisition unit 148 registers the acquired recommendation information in the recommended program list 149.
[0441]
In step S293, the recording setting unit 143 extracts the information of the recommended program corresponding to the current time from the program recommendation list 109, and determines the information necessary for the recording process such as the broadcast start time and the broadcast end time, and the broadcast channel. Information is obtained and supplied to the recording control unit 145.
[0442]
In step S294, the recording control unit 145 generates channel setting information for receiving the supply of the broadcast signal of the program to be recorded, and outputs it to the television receiver 4. The television receiver 4 receives a broadcast signal of a specified channel based on the control signal.
[0443]
In step S295, the data input unit 146 acquires a broadcast signal of the specified channel from the television receiver 4, and supplies the broadcast signal to the recording / playback processing unit 147.
[0444]
In step S296, the recording / playback processing unit 147 records the supplied broadcast signal on a mounted or built-in recording medium, and the process ends.
[0445]
Through such processing, it is possible to automatically record a program that matches the user's preference. The automatic recording process described with reference to FIG. 36 is not executed while a recording operation is already being performed, such as during a recording operation instructed by a user or a recording operation by a recording reservation process.
[0446]
Here, a case has been described in which recording is automatically performed based on a recommended program corresponding to the current time.However, for example, recommendation information for a predetermined time earlier than the current time is obtained, and the same processing is performed. Needless to say, a recording reservation may be automatically set.
[0447]
In the processing described with reference to FIGS. 1 to 36, the program vector PP is described as being generated in the distribution server 5, but the distribution server does not generate the program vector PP, but via the network 8, The EPG data may be supplied to the program recommendation processing device, and the program recommendation processing device may generate the program vector PP.
[0448]
The configuration of the distribution server 171 in the case where the distribution server supplies the EPG data to the program recommendation processing device via the network 8 and generates the program vector PP in the program recommendation processing device instead of generating the program vector PP. FIG. 37 shows the configuration of the program recommendation processing device 191 shown in FIG.
[0449]
Parts corresponding to those in the distribution server 5 in FIG. 2 and the program recommendation processing device 10 in FIG. 13 are denoted by the same reference numerals, and description thereof will be omitted as appropriate.
[0450]
That is, the distribution server 171 includes the data acquisition unit 21 and the data transmission unit 25 of the distribution server 5 described with reference to FIG. 2, and includes streaming data or metadata from the streaming data database 6 or the metadata database 7. The acquired EPG data is acquired and transmitted to the EPG receiving device 8 or the television receiving device 4 via the network 8.
[0451]
The program recommendation processing device 191 uses FIG. 13 except that a metadata acquisition unit 22 and a program vector generation unit 23 similar to those provided in the distribution server 5 of FIG. 2 are newly provided. It has a configuration similar to that of the program recommendation processing apparatus 10 described above. In addition to the processing executed by the program recommendation processing apparatus 10, the program vector generation processing 1 described with reference to FIG. The program vector generation processing 2 described above, the grouping processing 1 described using FIG. 7, and the grouping processing described using FIG. 8 are executed.
[0452]
Even when the distribution server does not generate the program vector PP, but supplies the EPG data to the program recommendation processing device via the network 8 and the program recommendation processing device generates the program vector PP, FIGS. It is possible to achieve the same effect as the case described using FIG.
[0453]
Further, the EPG receiving device 9 collects a user's operation history and setting information from the television display device 11 and the recording / reproducing device 12, and supplies the collected operation history and setting information to the distribution server via the network 8. The distribution server may not only generate the program vector PP but also execute a matching process and supply the matching result to the EPG receiving device 9 via the network 8. The network configuration in this case is shown in FIG. 39, and a block diagram showing the configuration of the distribution server 201 is shown in FIG.
[0454]
Parts corresponding to those in FIGS. 1, 37, and 38 are denoted by the same reference numerals, and description thereof will be omitted as appropriate.
[0455]
That is, the distribution server 201 is obtained by adding the function of the program recommendation processing device 191 described with reference to FIG. 38 to the distribution server 171 described with reference to FIG. 37, and the user owns the program recommendation processing device. No need.
[0456]
In the configuration shown in FIG. 39 and FIG. 40, generation processing of a program vector PP, a program side effect vector EfPP, a positive history vector UP, a negative history vector MUP, a standard preference vector APP, and the like, and grouping of program vectors The distribution server 201 executes all the processes described with reference to FIGS. 3 to 12 and FIGS. 14 to 30, such as the matching process, the exception recommended program selection process, and the like.
[0457]
At this time, the user's operation history and setting information collected by the EPG receiving device 9 from the television display device 11 and the recording / reproducing device 12 and transmitted via the network 8 include, for example, a user ID. And so on so that it can be distinguished for each user. Further, in the program recommendation processing device 191 of the distribution server 201, the initial registration storage unit 45, the positive history storage unit 47, the negative history storage unit 48, and the like separate each information on a user basis based on the user ID of the supplied information. save.
[0458]
In the above description, a case has been described in which the EPG data of the television broadcast signal is used to recommend a program that suits the user's preference. However, the present invention is applicable to various digital contents such as radio broadcasts and streaming data. The present invention can also be applied to a case where attribute information is added and a user's preference is recommended.
[0459]
The series of processes described above can also be executed by software. The software is a computer in which a program constituting the software is built in dedicated hardware, or a general-purpose personal computer capable of executing various functions by installing various programs. For example, it is installed from a recording medium.
[0460]
As shown in FIG. 2, FIG. 13, or FIG. 38, this recording medium is a magnetic disk 31 or 71 (flexible) on which the program is recorded, which is distributed to provide the program to the user separately from the computer. Discs), optical disks 32 or 72 (including CD-ROM (Compact Disk-Read Only Memory), DVD (Digital Versatile Disk)), and magneto-optical disks 33 or 73 (including MD (Mini-Disk) (trademark)) ), Or a package medium such as a semiconductor memory 34 or 74.
[0461]
In this specification, the step of describing a program recorded on a recording medium may be performed in chronological order according to the described order, or may be performed in parallel or even if not necessarily performed in chronological order. This also includes processes executed individually.
[0462]
In the present specification, the system represents the entire device including a plurality of devices.
[0463]
【The invention's effect】
As described above, according to the present invention, it is possible to generate content attribute information. In particular, attribute information can be generated based on the content classification.
Further, according to another aspect of the present invention, it is possible to select a content that matches the user's preference, and to select a content that correctly matches the user's preference because the content is selected based on the classification of the content. be able to.
Further, according to another aspect of the present invention, it is possible to select a content that matches the user's preference, and to generate the user's preference information based on the attribute information of the content that matches a predetermined condition. Content that correctly matches the user's preference can be selected.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating television program broadcasting and distribution of stream data.
FIG. 2 is a block diagram illustrating a configuration of a distribution server in FIG. 1;
FIG. 3 is a flowchart illustrating a program vector generation process 1;
FIG. 4 is a diagram illustrating EPG data.
FIG. 5 is a diagram illustrating a program vector.
FIG. 6 is a flowchart illustrating a program vector generation process 2;
FIG. 7 is a flowchart illustrating grouping processing 1;
FIG. 8 is a flowchart illustrating a grouping process 2;
FIG. 9 is a flowchart illustrating title grouping processing 1;
FIG. 10 is a flowchart illustrating title grouping processing 2;
FIG. 11 is a flowchart illustrating a title grouping process 3;
FIG. 12 is a flowchart illustrating title grouping processing 4;
FIG. 13 is a block diagram showing a configuration of a program recommendation processing device of FIG. 1;
FIG. 14 is a flowchart illustrating a positive history vector and negative history vector generation process 1.
FIG. 15 is a diagram illustrating a normal history vector.
FIG. 16 is a flowchart illustrating a positive history vector and negative history vector generation process 2.
FIG. 17 is a flowchart illustrating matching processing 1;
FIG. 18 is a flowchart illustrating a matching process 2;
FIG. 19 is a flowchart illustrating a matching process 3;
FIG. 20 is a flowchart illustrating a matching process 4;
FIG. 21 is a flowchart illustrating a matching process 5;
FIG. 22 is a flowchart illustrating a user-side effect vector generation process 1;
FIG. 23 is a flowchart illustrating a user-side effect vector generation process 2;
FIG. 24 is a flowchart illustrating a user-side effect vector generation process 3;
FIG. 25 is a flowchart illustrating a user-side effect vector generation process 4;
FIG. 26 is a flowchart illustrating a user-side reaction vector generation process 1;
FIG. 27 is a flowchart illustrating a user-side reaction vector generation process 2;
FIG. 28 is a flowchart illustrating a matching process including group recommendation.
FIG. 29 is a flowchart illustrating a matching process using a user model.
FIG. 30 is a flowchart illustrating exception recommendation processing.
FIG. 31 is a block diagram illustrating a configuration of the television receiver in FIG. 1.
32 is a block diagram illustrating a configuration of the television display device in FIG.
FIG. 33 is a flowchart illustrating a recommendation information display process.
FIG. 34 is a flowchart illustrating automatic channel setting processing.
FIG. 35 is a block diagram illustrating a configuration of the recording / reproducing apparatus of FIG. 1;
FIG. 36 is a flowchart illustrating automatic recording processing.
FIG. 37 is a block diagram illustrating a different configuration example of a distribution server.
FIG. 38 is a block diagram illustrating a different configuration example of the program recommendation processing device.
[Fig. 39] Fig. 39 is a diagram illustrating a different example of a network for television program broadcasting and stream data distribution.
FIG. 40 is a block diagram illustrating a different configuration example of the distribution server.
[Explanation of symbols]
5 distribution server, 10 program recommendation processing device, 21 data acquisition unit, 22 metadata extraction unit, 23 program vector generation unit, 24 data storage unit, 25 data transmission unit, 41 data acquisition unit, 42 program vector extraction unit, 43 matching Processing section, 44 operation input section, 45 initial registration storage section, 46 operation log acquisition section, 47 positive history storage section, 48 negative history storage section, 49 recommended information output section, 50 recommended program list, 61 normalization processing section, 62 Vector operation unit, 63 user information registration unit, 171 distribution server, 191 program recommendation processing device, 201 distribution server

Claims (24)

In an information processing apparatus that generates attribute information of the content based on information about the content,
Acquisition means for acquiring information on the content,
Classifying means for classifying the content based on the information on the content obtained by the obtaining means,
An information processing apparatus, comprising: an attribute information generating unit that generates the attribute information based on the information regarding the content and associates a result of the classification by the classifying unit. The method according to claim 1, further comprising: a transmission unit configured to transmit the attribute information generated by the attribute information generation unit and the classification result associated with the attribute information to another information processing apparatus. An information processing apparatus according to claim 1. The classification unit detects the content that matches a predetermined condition based on the information about the content,
The information processing apparatus according to claim 1, wherein the attribute information generation unit associates unique information for distinguishing a classification by the classification unit with the attribute information. The information processing apparatus according to claim 3, wherein the predetermined condition includes a condition that a difference between start times of broadcasting or distribution of the content is within a predetermined time. The information on the content is composed of a plurality of items,
The classification means detects the content including at least one and the same word in a predetermined item of the information on the content,
The information processing apparatus according to claim 1, wherein the attribute information generation unit associates unique information for distinguishing a classification by the classification unit with the attribute information. The information on the content is composed of a plurality of items,
The classification means detects the content in which the matching rate of words constituting a predetermined item of the information on the content is larger than a predetermined value,
The information processing apparatus according to claim 1, wherein the attribute information generation unit associates unique information for distinguishing a classification by the classification unit with the attribute information. The attribute information generating unit further associates the attribute information already associated with the second content of the same classification as the first content with the first content. An information processing apparatus according to claim 1. The content is a broadcast program,
The classifying means classifies the broadcast programs that are continuously broadcast as one classification based on the information regarding the content,
The attribute information generating unit generates the attribute information of the broadcast program classified as a continuously broadcasted broadcast program by the classifying unit using information on the content for a first broadcast. The information processing apparatus according to claim 7, wherein: An information processing method of an information processing apparatus that generates attribute information of the content based on information about the content,
A classification step of classifying the content based on the information about the content,
An attribute information generating step of generating the attribute information based on the information on the content;
An information processing method, comprising: associating the attribute information generated by the processing of the attribute information generating step with a classification result obtained by the processing of the classification step. A program that causes a computer to execute a process of generating attribute information of the content based on information about the content,
A classification step of classifying the content based on the information about the content,
An attribute information generating step of generating the attribute information based on the information on the content;
A recording medium in which a computer-readable program is recorded, comprising: an association step of associating the attribute information generated by the processing of the attribute information generation step with a classification result obtained by the processing of the classification step. . A program that causes a computer to execute a process of generating attribute information of the content based on information about the content,
A classification step of classifying the content based on the information about the content,
An attribute information generating step of generating the attribute information based on the information on the content;
A program comprising: an associating step of associating the attribute information generated by the processing of the attribute information generating step with a classification result by the processing of the classification step. In an information processing apparatus that executes a process of detecting content that matches a user's preference,
Acquiring means for acquiring the attribute information of the content;
Storage means for storing the user preference information;
A detection unit that detects the first content that matches the user's preference by comparing the preference information stored by the storage unit with the attribute information acquired by the acquisition unit;
The attribute information of the content includes unique information indicating a classification of the content,
The detection unit further detects, based on the unique information, a second content of the same classification as the first content that matches the user's preference as a content that matches the user's preference. Information processing device. In an information processing method of an information processing apparatus that executes a process of detecting content that matches a user's preference,
A storage control step of controlling storage of unique information indicating a classification of the first content that matches the user's preference;
A first determining step of determining whether or not the attribute information of the second content includes the unique information;
When it is determined by the processing of the first determination step that the attribute information includes the unique information, the unique information of the first content whose storage is controlled by the processing of the storage control step is determined. A second determining step of determining whether the second content has the same classification as the first content,
When the second content is determined to have the same classification as the first content by the processing of the second determination step, the second content is detected as content that matches the user's preference. An information processing method, comprising: A computer-executable program that executes a process of detecting content that matches a user's preference,
A storage control step of controlling storage of unique information indicating a classification of the first content that matches the user's preference;
A first determining step of determining whether or not the attribute information of the second content includes the unique information;
When it is determined by the processing of the first determination step that the attribute information includes the unique information, the unique information of the first content whose storage is controlled by the processing of the storage control step is determined. A second determining step of determining whether the second content has the same classification as the first content,
When the second content is determined to have the same classification as the first content by the processing of the second determination step, the second content is detected as content that matches the user's preference. A recording medium on which a computer-readable program is recorded. A computer-executable program that executes a process of detecting content that matches a user's preference,
A storage control step of controlling storage of unique information indicating a classification of the first content that matches the user's preference;
A first determining step of determining whether or not the attribute information of the second content includes the unique information;
When it is determined by the processing of the first determination step that the attribute information includes the unique information, the unique information of the first content whose storage is controlled by the processing of the storage control step is determined. A second determining step of determining whether the second content has the same classification as the first content,
When the second content is determined to have the same classification as the first content by the processing of the second determination step, the second content is detected as content that matches the user's preference. And a detecting step of performing the following. A first information processing device that generates attribute information of the content based on information about the content,
An information processing system configured to execute a process of detecting the content that matches the user's preference based on the attribute information of the content supplied from the first information processing device; ,
The first information processing device includes:
First acquisition means for acquiring information on the content,
Classifying means for classifying the content based on the information on the content obtained by the first obtaining means;
Attribute information generating means for generating the attribute information based on the information regarding the content,
And transmitting the attribute information generated by the attribute information generating unit and the classification result by the classification unit, and transmitting the result to another information processing apparatus,
The second information processing device includes:
Second acquisition means for acquiring the attribute information transmitted from the first information processing device;
Storage means for storing the user preference information;
Detecting means for detecting the first content that matches the user's preference by comparing the preference information stored by the storage means with the attribute information obtained by the second obtaining means; With
The attribute information of the content includes unique information indicating a classification of the content,
The detection means further detects, based on the unique information, a second content of the same classification as the first content that matches the preference of the user as a content that matches the preference of the user. Information processing system. In an information processing apparatus that executes a process of selecting content that matches a user's preference,
Acquiring means for acquiring the attribute information of the content;
Storage means for storing predetermined conditions;
Generating means for generating preference information of the user based on attribute information of the content corresponding to the content meeting the predetermined condition stored in the storage means, in a predetermined content group. Characteristic information processing device. A recommendation for generating recommendation information indicating content recommended for viewing by the user by calculating a similarity between the preference information generated by the generation unit and the attribute information of the content acquired by the acquisition unit. The information processing apparatus according to claim 17, further comprising an information generation unit. Further comprising an operation input means for receiving an operation input of the user,
18. The information processing apparatus according to claim 17, wherein the storage unit stores the predetermined condition input from the user by the operation input unit. 18. The information processing apparatus according to claim 17, wherein the predetermined content group is a set of contents broadcast or distributed during a predetermined period. 18. The information processing apparatus according to claim 17, wherein the predetermined content group is a set of contents broadcast or distributed in a predetermined time zone. In an information processing method of an information processing device that executes a process of selecting content that matches a user's preference,
An obtaining step of obtaining attribute information of the content;
A setting step of setting a predetermined condition;
An extracting step of extracting, from a predetermined content group, content that matches the predetermined condition set by the processing of the setting step;
A generating step of generating preference information of the user based on attribute information of the content corresponding to the content meeting the predetermined condition extracted by the processing of the extracting step. . A computer-executable program that executes a process of selecting content that matches a user's preference,
An obtaining step of obtaining attribute information of the content;
A setting step of setting a predetermined condition;
An extracting step of extracting, from a predetermined content group, content that matches the predetermined condition set by the processing of the setting step;
Generating a preference information of the user based on attribute information of the content corresponding to the content meeting the predetermined condition extracted by the processing of the extraction step. A recording medium on which a possible program is recorded. A computer-executable program that executes a process of selecting content that matches a user's preference,
An obtaining step of obtaining attribute information of the content;
A setting step of setting a predetermined condition;
An extracting step of extracting, from a predetermined content group, content that matches the predetermined condition set by the processing of the setting step;
A generating step of generating preference information of the user based on attribute information of the content corresponding to the content that meets the predetermined condition extracted by the processing of the extracting step.

Images