JPS62281059A - Parallel phrase processor - Google Patents
Parallel phrase processorInfo
- Publication number
- JPS62281059A JPS62281059A JP61124821A JP12482186A JPS62281059A JP S62281059 A JPS62281059 A JP S62281059A JP 61124821 A JP61124821 A JP 61124821A JP 12482186 A JP12482186 A JP 12482186A JP S62281059 A JPS62281059 A JP S62281059A
- Authority
- JP
- Japan
- Prior art keywords
- parallel
- parallel phrase
- priority
- phrase
- language
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004364 calculation method Methods 0.000 claims description 13
- 238000004458 analytical method Methods 0.000 abstract description 21
- 235000013305 food Nutrition 0.000 description 17
- 235000013351 cheese Nutrition 0.000 description 14
- 235000013336 milk Nutrition 0.000 description 13
- 239000008267 milk Substances 0.000 description 13
- 210000004080 milk Anatomy 0.000 description 13
- 235000013405 beer Nutrition 0.000 description 12
- 238000010586 diagram Methods 0.000 description 10
- 235000013361 beverage Nutrition 0.000 description 7
- 235000014121 butter Nutrition 0.000 description 3
- 235000008429 bread Nutrition 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 235000011389 fruit/vegetable juice Nutrition 0.000 description 2
- 238000000034 method Methods 0.000 description 2
- 235000015041 whisky Nutrition 0.000 description 2
- 241000538132 Moya Species 0.000 description 1
- 235000013334 alcoholic beverage Nutrition 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 235000013365 dairy product Nutrition 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 235000016709 nutrition Nutrition 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
Landscapes
- Machine Translation (AREA)
Abstract
Description
【発明の詳細な説明】
3、発明の詳細な説明の欄
〔概要〕
本発明は、自然言語を処理する際の並列句処理装置にお
いて、並列句解析のため言語概念体系モデル及びその検
索部と優先度演算部とを備えることにより、複数の解析
結果から最も適した並列句の処理を出力するようにした
ものである。[Detailed Description of the Invention] 3. Detailed Description of the Invention [Summary] The present invention provides a parallel phrase processing device for processing natural language, which uses a language concept system model and its search unit for parallel phrase analysis. By including a priority calculation unit, the most suitable parallel phrase processing is output from a plurality of analysis results.
本発明は、自然言語を処理する際の並列句処理装置に関
し、特に、複数の処理結果が存在する場合にそれらの優
先度を判断する並列句処理装置に関する。The present invention relates to a parallel phrase processing device for processing natural language, and particularly to a parallel phrase processing device that determines the priority of a plurality of processing results when they exist.
データベース照会や自動翻訳機などで、機械語もしくは
コートでなく、自然言語を使用する場合には、当然その
自然言語を処理する必要が生じ、その自然言語処理に際
して、複数の単語が並列に提示されると、それらの並列
句をどのように扱うかは解析と検討により判断される。When natural language is used instead of machine language or code in database queries or automatic translators, it is naturally necessary to process that natural language, and during natural language processing, multiple words are presented in parallel. Then, how to handle those parallel phrases is determined through analysis and consideration.
例えば、自動処理式の食堂で”チーズとバターとパン”
とい・う並列句が使用された場合、チーズとバターを乳
製品として同一に扱うか、バターとパンを英語式に”ト
ースト”として同一に扱うかの判断が必要となるが、そ
の判断には、自然言語に伴う人間的な意味や概念の解析
と検討が必要で、専用の並列句処理装置を付設されてい
る。For example, ``cheese, butter, and bread'' in an automated cafeteria.
When parallel phrases such as and are used, it is necessary to judge whether cheese and butter should be treated the same as dairy products, or whether butter and bread should be treated the same as "toast" in the English style. , it is necessary to analyze and study the human meanings and concepts associated with natural language, and is equipped with a dedicated parallel phrase processing device.
並列句の解析では、一般的には複数の解析結果が出力さ
れ、それら複数の解析結果から最も適したものを選択す
ることが必要とされるが、従来は、複数の解析結果が得
られるとき、最初のものだけを採用していた。In the analysis of parallel clauses, multiple analysis results are generally output, and it is necessary to select the most suitable one from those multiple analysis results. , only the first one was adopted.
例えば”ビールと、ミルクと、チーズ”という並列句に
対する解析としては、ビールもミルクもチーズも食料で
あるとして一括する扱い方もあれば、ビールとミルクは
飲料であるとして一括する扱い方もあり、ビールはアル
コール飲料だがミルクとチーズは栄養食品であるとする
扱い方もあり得る。このように複数の解析結果が得られ
るとき、従来は、例えば”ビールと、ミルク”という最
初の2語の共通概念である”飲料”が発見されると、そ
の”飲料”を採用して処理を終了していた。For example, when analyzing the parallel phrase "beer, milk, and cheese," there are ways to treat beer, milk, and cheese as food, and there are ways to treat beer and milk as drinks. It is also possible to treat beer as an alcoholic beverage, but milk and cheese as nutritional foods. When multiple analysis results are obtained in this way, conventionally, for example, if a "beverage" is found that is a common concept between the first two words "beer and milk", that "beverage" is adopted and processed. had finished.
従来の並列句の解析では、複数個の解析結果がある場合
でも、それらを正当に評価せず、最初に出力されたもの
を採用していたので、最も適切に並列句を処理した結果
が2番目以降に出力される場合は採用されないという問
題を生していた。In conventional parallel phrase analysis, even if there are multiple analysis results, the first output is adopted without properly evaluating them, so the most appropriate result of parallel phrase processing is 2. This caused the problem that if it was output after the number, it would not be adopted.
本発明は、このような問題点を鑑み、並列句の可能な解
析結果すべてについて優先度を比較し、最も適切な解析
結果を選択できる並列句処理装置を提供することを目的
とする。In view of these problems, it is an object of the present invention to provide a parallel phrase processing device that can compare the priorities of all possible analysis results of parallel phrases and select the most appropriate analysis result.
本発明において、上記の問題点を解決するための手段は
、第1図に示すように、各自然言語を所望の分野別に体
系付けた言語概念体系モデル1と、複数の所望の自然言
語を並列句に構成する並列句組立装置2と、その所要の
自然言語を前記モデル1内から検索する言語概念体系検
索部3と、検索された自然言語の体系内での順位から並
列句の優先度を計算する優先度演算部4と、並列句に付
与された優先度を比較する並列句比較部5とを備える並
列句処理装置とするものである。In the present invention, the means for solving the above problems is as shown in FIG. A parallel phrase assembling device 2 that composes a phrase into a phrase, a language concept system search unit 3 that searches the model 1 for the required natural language, and a priority of the parallel phrase based on the ranking within the system of the searched natural language. The parallel phrase processing device includes a priority calculating section 4 that calculates priorities, and a parallel phrase comparing section 5 that compares priorities given to parallel phrases.
本発明では、自然言語を所望の分野別に体系付けた言語
概念体系モデル1を用意し、言語概念体系検索部3によ
って、各並列句の構成要素に対応するモデル中の対応概
念が検索され、更に優先度演算部4によって並列句の体
系内順位の優先度が演算され、並列句比較部5で最適な
ものが選択されることになる。In the present invention, a language concept system model 1 in which natural language is organized into desired fields is prepared, and a language concept system search unit 3 searches for corresponding concepts in the model corresponding to the constituent elements of each parallel phrase. The priority calculation section 4 calculates the priority of the system ranking of the parallel phrases, and the parallel phrase comparison section 5 selects the optimal one.
以下、本発明を、実施例及び図面を参照して、詳細に説
明する。Hereinafter, the present invention will be explained in detail with reference to examples and drawings.
並列句処理装置の構成(第1図)
第1図は、本発明の原理とともに本発明を実施した並列
句処理装置の1構成例を示すブロック図である。第1図
において、並列句処理装置は、各自然言語を所望の分野
別に体系付けたデータベースである言語概念体系モデル
1と、複数の自然言語を入力され、それらを並列句に合
成する並列句組立装置2とを主部分として構成され、言
語概念体系モデル1と並列句組立装置2とを連結し、並
列句組立装置2に入力された自然言語を前記モデル1の
体系内から検索する言語概念体系検索部3と、検索され
た体系内での順位から並列句の優先度を演算し、その優
先度を付与して並列句組立装置2に送り返す優先度演算
部4と、並列句に付与された優先度を比較する並列句比
較部5とで構成されている。Configuration of Parallel Phrase Processing Device (FIG. 1) FIG. 1 is a block diagram showing an example of the configuration of a parallel phrase processing device implementing the present invention together with the principle of the present invention. In FIG. 1, the parallel phrase processing device has a language concept system model 1, which is a database that organizes each natural language into a desired field, and a parallel phrase assembly that receives multiple natural languages and synthesizes them into parallel phrases. A language concept system that connects a language concept system model 1 and a parallel phrase assembly device 2, and searches the natural language input to the parallel phrase assembly device 2 from within the system of the model 1. A search unit 3, a priority calculation unit 4 which calculates the priority of parallel phrases based on the ranking within the searched system, assigns the priority, and sends it back to the parallel phrase assembling device 2; It is composed of a parallel phrase comparison section 5 that compares priorities.
言語概念体系モデルの構成ツリー(第2図)第2図は、
上記データベースとして格納される言語概念体系モデル
の1例を示す部分構成ツリー図で、図中左方はど上位概
念を示している。このモデルは、あくまでデータベース
として作成されるものであって、もちろん自然言語の通
常概念に従うことが望ましいが、用途に応じて所望の体
系を作成して差し支えない。第2図は食品サービス業種
用に作成されたモデルの1例を示すツリー図で、従って
、左方より4番目の”食料”項までは属性関係を示し、
”食料”項より右方は1項目毎に体系内の上下順位を示
す。Structure tree of language concept system model (Figure 2)Figure 2 shows the following:
This is a partial structure tree diagram showing an example of a language concept system model stored as the database, and the left side of the diagram shows superordinate concepts. This model is created as a database to the last, and of course it is desirable to follow the normal concepts of natural language, but any desired system may be created depending on the purpose. Figure 2 is a tree diagram showing an example of a model created for the food service industry. Therefore, up to the fourth "food" term from the left, attribute relationships are shown.
The area to the right of the "Food" section shows the ranking of each item in the system.
並列句の処理方式(第3図)
第3図(a)〜(e)は、並列句処理の1例を示す説明
図である。Parallel phrase processing method (FIG. 3) FIGS. 3(a) to 3(e) are explanatory diagrams showing an example of parallel phrase processing.
1例として°ビールとミルクとチーズ”という並列句に
ついて説明する。この並列句が第1図の並列句組立装置
2に入力されると、その並列句を構成する3つの自然言
語の組み合わせに対して、下記の各解析結果が考えられ
る。As an example, we will explain the parallel phrase ``beer, milk, and cheese.'' When this parallel phrase is input to the parallel phrase assembly device 2 shown in Fig. 1, the combination of the three natural languages that make up the parallel phrase is Therefore, the following analysis results can be considered.
まず”ビール”と”′ミルク”と”チーズ”とを第3図
(a)に示すように1つにまとめる解析で、並列句組立
装置2がこられの3語を言語概念体系検索部3へ入力す
ると、該検索部3は言語概念体系モデル1を検索して、
共通項”食料”を発見すると共に、各自然言語から共通
項”食料”までの順位差を優先度演算部4へ読め出す。First, in an analysis that combines "beer,""milk," and "cheese" into one as shown in FIG. , the search unit 3 searches the language concept system model 1, and
The common term "food" is discovered, and the rank difference between each natural language and the common term "food" is read out to the priority calculation unit 4.
体系内での順位差は、第2r!lに示すように、ビール
と食料。The difference in ranking within the system is 2nd place! Beer and food as shown in l.
ミルクと食料は3段であり、−f〜−ズと食料は2段で
ある。優先度演(り部4ば、これら順位差の最大として
演出し、Iii+記検索部3から送られた組み合わせに
付与して、並列句絹立装置2に送り返す。Milk and food are on three levels, and -f~-'s and food are on two levels. The priority rendering section 4 performs the maximum of these ranking differences, assigns it to the combination sent from the Iiii+ record search section 3, and sends it back to the parallel phrase making device 2.
各自然言語に共通項が発見されない場合は、優先度演算
部4は、第3図(e)に示すように、自然言語のそれぞ
れからツリーを遡って枝を辿り、各言語間相互の段数距
離を加算するものとする。If a common term is not found in each natural language, the priority calculation unit 4 traces the tree back from each natural language, as shown in FIG. shall be added.
次に、第3図(b)に示すように、”ビール”と”ミル
ク゛を1つにまとめ、これに”チーズ゛′を加える解析
で、並列句組立装N2がこの組み合わせで3語を言語概
念体系検索部3へ入力すると、該検索部3は言語概念体
系モデル1を検索して、まず”ビール”と”ミルク”の
共通項”飲料”を発見し、更に”飲料”と゛°チーズ”
の共通項”食料”を発見すると共に、”ビール”と゛ミ
ルク”からその共通項”飲料”までの順位差2.”チー
ズ゛から共通項”食料”までの順位差2及び”飲料”か
ら”食料”までの順位差1を前記優先度演算部4へ読み
出す。優先度演算部4は、これら順位差の最大値2を第
3図(b)に示す組み合わせの優先度Pとして算出し、
前記検索部3から送られた組み合わせに付与して、並列
句組立装置2に送り返す。Next, as shown in Figure 3(b), in an analysis that combines ``beer'' and ``milk'' into one and adds ``cheese'' to it, the parallel phrase assembly system N2 converts the three words into a language using this combination. When the input is input to the concept system search section 3, the search section 3 searches the language concept system model 1 and first finds the common term "beverage" between "beer" and "milk", and then finds the common term "beverage" between "beverage" and "cheese".
In addition to discovering the common term "food", the rank difference between "beer" and "milk" and their common term "beverage" is 2. The rank difference between "cheese" and the common term "food" is 2, and from "beverage" The rank difference 1 up to "food" is read out to the priority calculation section 4. The priority calculation unit 4 calculates the maximum value 2 of these ranking differences as the priority P of the combination shown in FIG. 3(b),
It is added to the combination sent from the search section 3 and sent back to the parallel phrase assembly device 2.
本実施例では更に、第3図(C)に示すように、”ミル
ク”と”チーズ”を1つにまとめ、これに”ビール”を
加える解析が考えられる。並列句組立装置2が此の組み
合わせで3語を言語概念体系検索部3へ入力Jると、該
検索部3は前記モデル1を検索して、まず”ミルク”と
”チーズ”との共通項”食料”を発見し、更に”ビール
”もその”食料”以外に共通項がないことを発見する。In this embodiment, as shown in FIG. 3(C), an analysis can be considered in which "milk" and "cheese" are combined into one and "beer" is added to this. When the parallel phrase assembly device 2 inputs the three words in this combination to the language concept system search unit 3, the search unit 3 searches the model 1 and first finds the common term between “milk” and “cheese”. He discovers ``food'' and also discovers that ``beer'' has nothing in common other than ``food.''
ここで、第2図に示す如く、′ミクル”から”食料”ま
での順位差は3であり、”チーズ”から”食料”までの
順位差は2である。”ビール”から”食料”までの順位
差は3である。これらを読み出された優先度演算部4は
、更に次のような演算を行う。Here, as shown in Figure 2, the difference in rank from 'Mikuru' to 'Food' is 3, and the difference in rank from 'Cheese' to 'Food' is 2. From 'Beer' to 'Food'. The difference in the rankings is 3. The priority calculation unit 4 which has read these data further performs the following calculations.
即ち、′ビール”は、”ミルク”又は”チーズ”との共
通項”食料”を介して自らの共通項”食料”に体系付け
られるので、”ミルク”又は”チーズ”からその共通項
”食料”までの順位差の最大値3と自らの共通項”食料
”までの順位差3とを加算された順位差6が、両者を隔
てる正確な順位差となる。優先度演算部4は、これら順
位差の最大値6を第3図(C)に示す組み合わせの優先
度Pとして算出し、前記検索部3から送られた組み合わ
せに付与して、並列句組立装置2に送り返す。In other words, ``beer'' is organized into its own common term ``food'' through the common term ``food'' with ``milk'' or ``cheese.'' The rank difference 6 which is the sum of the maximum rank difference 3 up to ``and the rank difference 3 up to their common item ``food'' becomes the accurate rank difference that separates the two.The priority calculation unit 4 The maximum value 6 of the rank difference is calculated as the priority P of the combination shown in FIG.
並列句組立装置2には、これらの組み合わせを優先度付
きで並列句比較部5へ送付し、P−3゜P=2.P=6
を比較し、P値の最小なものほど優先度が高いとして、
第3図(b)の例を最適な解析結果として採用し、本並
列句は、”ヒールとミルク”と”チーズ”という処理が
行われる。The parallel phrase assembling device 2 sends these combinations with priorities to the parallel phrase comparison section 5, and calculates P-3゜P=2. P=6
Compare them, and assume that the one with the smallest P value has a higher priority,
The example in FIG. 3(b) is adopted as the optimal analysis result, and the parallel phrases are processed as "heels and milk" and "cheese".
第3図(d)は別な実施例を示す図であって、”ビール
とウィスキーとジュースとチーズとハム”という5個の
自然言語で成る1例を示す。FIG. 3(d) is a diagram showing another embodiment, and shows one example consisting of five natural languages: "beer, whiskey, juice, cheese, and ham."
結果のみを説明すると、”ヒールとウィスキーとジュー
ス”及び”チーズとハム”でまとめれば、その共通項”
飲料”に対する順位差2が゛食料”でまとめる場合の順
位差4よりも少なく、優先度の高い解析結果となる。To explain only the results, if you summarize them as "heels, whiskey, and juice" and "cheese and ham," they will have something in common.
The ranking difference 2 for "drinks" is smaller than the ranking difference 4 for "food", which is an analysis result with a high priority.
処理動作のフローチャート(第4図)
第4図は、上記並列句処理のフローチャートである。第
4図において、フローが開始されると、第の段として、
複数の自然言語から成る並列句が並列句組立装置2に入
力され、並列句組立装W2はフローの第0段として、そ
の並列句を自然言語に分解し、それらの種々な組み合わ
せを作成して、組み合わせ毎に言語概念体系検索部3へ
出力する。Flowchart of Processing Operation (FIG. 4) FIG. 4 is a flowchart of the above-mentioned parallel phrase processing. In FIG. 4, when the flow is started, as a stage,
Parallel phrases consisting of a plurality of natural languages are input to the parallel phrase assembling device 2, and as the 0th stage of the flow, the parallel phrase assembling device W2 decomposes the parallel phrases into natural languages and creates various combinations of them. , is output to the language concept system search unit 3 for each combination.
該検索部3はフローの第0段として、その組み合わせを
データベースである言語概念体系モデル1内に検索し、
フローの第0段で共通項を求める。As the 0th stage of the flow, the search unit 3 searches for the combination in the language concept system model 1 which is a database,
Find the common denominator in the 0th stage of the flow.
共通項が発見された場合は、フローの第0段として共通
項までの順位差を読み出し、共通項が発見されない場合
は、各自然言語間もしくは途中共通項間の段数距離を読
み出す。優先度演算部4は、読み出された順位差から、
フローの第0段として、その組み合わセの優先度を算出
し、並列句が並列句組立装置2に返信する。並列句組立
装W2は、すべての組み合わゼについて優先度が得られ
たならば、それらを並列句比較部5へ送出し、フローの
第0段として、優先度を比較し、最適な解析結果を決定
かつ処理する。If a common term is found, the rank difference up to the common term is read as the 0th stage of the flow, and if a common term is not found, the stage number distance between each natural language or between intermediate common terms is read. The priority calculation unit 4 calculates, from the read rank difference,
As the 0th stage of the flow, the priority of the combination is calculated, and the parallel phrases are sent back to the parallel phrase assembling device 2. When the parallel phrase assembling device W2 has obtained the priorities for all combinations, it sends them to the parallel phrase comparison unit 5, and as the 0th stage of the flow, the priorities are compared and the optimal analysis result is determined. Decide and process.
以上述べてきたように、本発明によれば、並列句の可能
な解析結果すべてについて優先度を比較し、最も適切な
解析結果を選択できる並列句処理装置を提供することが
できる。As described above, according to the present invention, it is possible to provide a parallel phrase processing device that can compare the priorities of all possible analysis results of parallel phrases and select the most appropriate analysis result.
第1図は本発明の原理とともに並列句処理装置の基本構
成の一実施例を示すブロック図、第2図は言語概念体系
モデルの説明図、第3図は並列句処理の説明図、第4図
はその処理動作のフローチャートである。
1・・・言語概念体系モデル、
2・・・並列句組立装置、
3・・・言語概念体系検索部、
4・・・優先度演算部、
5・・・並列句比較部。
“〜二ノ″
モヤ:
コロ【
r+既へイ本斧モデルが連成・・刈−
P=3
P=2
(C)
ヒール ヒ ミノLグ と 牛−ズ゛P
、6
第2図
(d)
(e)
並列旬処理の真梵明日
第3図
特開BRG2−281059(6)
曲ジ
盛列句の幻理動咋
第4図
一2瓜n−Fig. 1 is a block diagram showing the principle of the present invention as well as an embodiment of the basic configuration of a parallel phrase processing device, Fig. 2 is an explanatory diagram of a linguistic conceptual system model, Fig. 3 is an explanatory diagram of parallel phrase processing, and Fig. 4 is an explanatory diagram of a language concept system model. The figure is a flowchart of the processing operation. DESCRIPTION OF SYMBOLS 1...Language concept system model, 2...Parallel phrase assembly device, 3...Language concept system search unit, 4...Priority calculation unit, 5...Parallel phrase comparison unit. “~Nino” Moya: Koro [r + already real ax model combined...Kari- P = 3 P = 2 (C) Heel Himino Lg and Ushi-Z゛P
, 6 Figure 2 (d) (e) Parallel processing of Shinbon tomorrow Figure 3 Unexamined Japanese Patent Publication BRG2-281059 (6) Genri Dokui of Quji Shengrenku Figure 4 12 瓜n-
Claims (1)
デル(1)と、複数の所望の自然言語を並列句に構成す
る並列句組立装置(2)と、その所要の自然言語を前記
モデル(1)内から検索する言語概念体系検索部(3)
と、検索された自然言語の体系内での順位から並列句の
優先度を計算する優先度演算部(4)と、並列句に付与
された優先度を比較する並列句比較部(5)とを備える
ことを特徴とする並列句処理装置。A language concept system model (1) that organizes each natural language into desired fields; a parallel phrase assembly device (2) that constructs a plurality of desired natural languages into parallel phrases; 1) Language concept system search section that searches from within (3)
, a priority calculation unit (4) that calculates the priority of the parallel phrase from the ranking within the searched natural language system, and a parallel phrase comparison unit (5) that compares the priorities given to the parallel phrases. A parallel phrase processing device comprising:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP61124821A JPS62281059A (en) | 1986-05-30 | 1986-05-30 | Parallel phrase processor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP61124821A JPS62281059A (en) | 1986-05-30 | 1986-05-30 | Parallel phrase processor |
Publications (1)
Publication Number | Publication Date |
---|---|
JPS62281059A true JPS62281059A (en) | 1987-12-05 |
Family
ID=14894941
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
JP61124821A Pending JPS62281059A (en) | 1986-05-30 | 1986-05-30 | Parallel phrase processor |
Country Status (1)
Country | Link |
---|---|
JP (1) | JPS62281059A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018180935A (en) * | 2017-04-13 | 2018-11-15 | 日本電信電話株式会社 | Parallel phrase analysis device, parallel phrase analysis model learning device, method, and program |
-
1986
- 1986-05-30 JP JP61124821A patent/JPS62281059A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018180935A (en) * | 2017-04-13 | 2018-11-15 | 日本電信電話株式会社 | Parallel phrase analysis device, parallel phrase analysis model learning device, method, and program |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9754508B2 (en) | Computerized method and system for analyzing and processing a food recipe | |
Slimani et al. | The EPIC nutrient database project (ENDB): a first attempt to standardize nutrient databases across the 10 European countries participating in the EPIC study | |
Rand et al. | Compiling data for food composition data bases | |
US8647121B1 (en) | Food item grading | |
Jermsurawong et al. | Predicting the structure of cooking recipes | |
US20170046980A1 (en) | Nutrition system | |
US20130144875A1 (en) | Set expansion processing device, set expansion processing method, program and non-transitory memory medium | |
US20070233663A1 (en) | Method, apparatus, and computer program product for searching information | |
Symoneaux et al. | 12 Open-Ended Questions | |
Park et al. | Kitchenette: Predicting and recommending food ingredient pairings using siamese neural networks | |
Adeyonu et al. | Determinants of sweet potato value addition among smallholder farming households in Kwara state, Nigeria | |
Diwan et al. | A named entity based approach to model recipes | |
Amac et al. | Procedural reasoning networks for understanding multimodal procedures | |
JPS62281059A (en) | Parallel phrase processor | |
US20160335343A1 (en) | Method and apparatus for utilizing agro-food product hierarchical taxonomy | |
Hanai et al. | Clustering for closely similar recipes to extract spam recipes in user-generated recipe sites | |
Pochmann et al. | Multi-objective bilevel recommender system for food diets | |
KR20210057867A (en) | Food recommandation system based on wine information | |
Moore et al. | Hierarchical representations of market structures and choice processes through preference trees | |
Ratisoontorn | Recipe Recommendations for Toddlers Using Integrated Nutritional and Ingredient Similarity Measures | |
Isokawa et al. | Performances in GA-based menu production for hospital meals | |
Pugsee et al. | Suggestion analysis for food recipe improvement | |
JP6918695B2 (en) | Information processing equipment and programs | |
JP2002175318A (en) | Correspondence system for food material name, and recording medium | |
CN117541359B (en) | Dining recommendation method and system based on preference analysis |