WO2014118975A1 - Recommendation creation system - Google Patents

Recommendation creation system Download PDF

Info

Publication number
WO2014118975A1
WO2014118975A1 PCT/JP2013/052396 JP2013052396W WO2014118975A1 WO 2014118975 A1 WO2014118975 A1 WO 2014118975A1 JP 2013052396 W JP2013052396 W JP 2013052396W WO 2014118975 A1 WO2014118975 A1 WO 2014118975A1
Authority
WO
WIPO (PCT)
Prior art keywords
category
recommendation
user
question
item
Prior art date
Application number
PCT/JP2013/052396
Other languages
French (fr)
Japanese (ja)
Inventor
智也 石田
彰宏 西山
直晴 ▲高▼多
庸介 牧
幸裕 堀北
倫太郎 鷲尾
Original Assignee
株式会社電通
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 株式会社電通 filed Critical 株式会社電通
Priority to GB1421558.6A priority Critical patent/GB2517358A/en
Priority to PCT/JP2013/052396 priority patent/WO2014118975A1/en
Priority to JP2014510321A priority patent/JP5782562B2/en
Publication of WO2014118975A1 publication Critical patent/WO2014118975A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2457—Query processing with adaptation to user needs
    • G06F16/24575—Query processing with adaptation to user needs using context
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/335—Filtering based on additional data, e.g. user or group profiles
    • G06F16/337—Profile generation, learning or modification

Definitions

  • the present invention relates to a recommendation creation system that provides an optimal answer to a user's question.
  • Patent Document 1 discloses a portable information terminal having an agent function that enables a search process of information required for a user's action request content. Yes.
  • the information retrieval service as described above provides answers based on the keywords included in the user's inquiry and the user's attribute information, an answer that is not expected by the user or an appropriate recommendation for the user can be obtained. I can't.
  • the present invention has been made in view of such points, and an object thereof is to provide a recommendation creation system that can provide an optimal recommendation such as an answer that the user does not expect or an appropriate answer to the user. To do.
  • the recommendation creating system includes a database storing user information in a state of being classified into a plurality of categories each including at least one item, and the user for acquiring user information regarding uncollected items in the category.
  • Question control means for generating a question to the user
  • recommendation creation means for creating a recommendation for the user's inquiry from user information regarding items in a plurality of categories stored in the database.
  • the question control means estimates one category from the keywords included in the question, sets the estimated category as a central category, and relates to the question in a category other than the central category. It is preferable to set a block matrix with the central category and the related category.
  • each item included in each category has a degree of relevance with respect to the other category in advance
  • the question control unit is configured to correspond to the central category in the related category. It is preferable to generate a question for acquiring user information regarding an uncollected item having a high degree of association.
  • the question control means may ask a question about an item for which user information has not been collected when registration rates of items in the central category and the related category constituting the block matrix are less than a predetermined value. Is preferably generated.
  • a recommendation including information on related categories can be provided in response to a user's inquiry, so that an unexpected and useful answer that the user himself / herself is not aware of can be obtained. Can do.
  • the recommendation creating means creates a recommendation for the user's inquiry using a user's history log.
  • the database stores a question for obtaining user information regarding each item in each category.
  • the database stores a block matrix set for each central category.
  • the portable terminal device of the present invention is characterized in that the above-mentioned recommendation creation system is mounted.
  • an item having a high correlation with the inquiry is preferentially used. Since the question for collecting items is asked to the user, it is possible to provide an optimal recommendation such as an answer that the user does not expect or an appropriate answer to the user.
  • FIG. 1 It is a figure which shows the system configuration
  • FIG. 1 is a diagram showing a system configuration including a recommendation creation system according to an embodiment of the present invention.
  • FIG. 1 shows a case where the recommendation creation system is in a server connected to a network.
  • the present invention is not limited to the system configuration shown in FIG. 1 and can also be applied to a case where a recommendation creation system is mounted on a mobile terminal device.
  • the system shown in FIG. 1 includes a mobile terminal device 1 such as a mobile phone, a search engine 4 connected to the mobile terminal device 1 through a mobile network 2 and a network 3 such as the Internet, a mobile network 2 and the Internet.
  • the system is mainly composed of a recommendation creation system (Web server) 5 connected via such a network 3.
  • Web server recommendation creation system
  • the mobile terminal device 1 includes a terminal device such as a mobile phone or a tablet-type terminal, can connect to at least the search engine 4 to search for information, and can connect to a server equipped with a recommendation creation system to obtain recommendations. It has a function. Further, the mobile terminal device 1 has a function of analyzing the voice of the user who has input the voice and acquiring information requested by the user from the search engine 4. Moreover, the portable terminal device 1 may be provided with other call functions, e-mail functions, contactless card functions, and the like.
  • the mobile network 2 is a network between the mobile terminal device 1 and the Internet 3.
  • the mobile network 2 includes a mobile packet communication network in addition to a normal mobile communication network.
  • the mobile network 2 has a gateway function for connecting to the network 3 such as the Internet.
  • the network 3 includes the Internet and other networks such as LAN and WAN.
  • the recommendation creation system 5 mainly includes a dialogue control unit 51, a recommendation creation unit 52, a question control unit 53, and a personal information database (DB) 54.
  • DB personal information database
  • the recommendation creation system 5 in providing a recommendation for a question from a user, a recommendation is provided by preferentially using items having a high correlation with the question. If there is an uncollected item even though the correlation is high, a question for collecting such an item is made to the user.
  • the dialogue control unit 51 outputs a recommendation to the user in response to an inquiry input from the user, and outputs a question for acquiring user information regarding an uncollected item to the user.
  • “question” means a question that the user inputs to the recommendation creation system in order to request a recommendation
  • “recommendation” means a proposal provided by the recommendation creation system for the “question” from the user. means. For example, as a “question”, if the user inputs “Recommendation of purchasing a house, but how about?” To the recommendation creation system, “Recommendation” is “There are many debts against the deposit amount. "What if the remaining car is sold?"
  • the “question” means a question corresponding to an uncollected item in user information stored in the personal information DB 54 described later and classified by category including at least one item. For example, in the category of financial information, if the deposit amount item has not been collected, the question “How much is the deposit amount?” Is output from the recommendation creation system to the user.
  • the “category” is for classifying user information, and examples thereof include meal, health, human relations, dreams / goals, basic information, hobbies, exercise, work, finance, and the like.
  • “Item” means specific information included in a category.
  • the items included in the meal category include favorite foods and the like
  • the items included in the financial category include deposit amount, periodic expenditure, debt amount, and the like.
  • Basic information is basic information about the user such as name, height, weight, address, age, occupation, family structure, etc. This basic information may be input in advance by the user, and other applications Information input by the user during operation may be used, or may be collected by a question.
  • the categories and items are not limited to these and can be set as appropriate.
  • the dialogue control unit 51 estimates a category for classifying user information from keywords included in the question input by the user. For example, when the question from the user is “I am considering purchasing a house, but what about?”, “Finance category” is estimated as a central category from “house” and “purchase” that can be keywords.
  • the dialogue control unit 51 outputs the estimated central category to the question control unit 53.
  • a user may input with a voice
  • the dialogue control unit 51 outputs the above question and the above in accordance with the collection status (registration rate) of user information about items in the block matrix set with the central category and related categories other than the central category. Switch the output of the recommendation. That is, if the registration rate in the block matrix is a predetermined value (for example, 80%) or more, the dialogue control unit 51 outputs a recommendation for the question from the recommendation creation unit 52 described later. Note that the question control unit 53, which will be described later, notifies that the registration rate is equal to or greater than a predetermined value. In addition, when the registration rate in the block matrix is less than a predetermined value (for example, 80%), the dialogue control unit 51 outputs a question for an uncollected item from the question control unit 53.
  • a predetermined value for example, 80%
  • the question control unit 53 generates a question to the user in order to acquire user information regarding uncollected items in the central category and related categories (block matrix).
  • the question control unit 53 estimates one category from the keywords included in the question, sets the estimated category as the central category, sets the related category as a related category other than the central category for the question, Set the block matrix.
  • the block matrix By using the block matrix in this way, user information can be used for items about categories other than the category estimated from the user's question, so the answer that the user does not expect, an appropriate answer to the user, etc. It is possible to provide a more optimal recommendation.
  • the central category is notified from the dialogue control unit 51. Therefore, as shown in FIG. A block matrix in which category B is arranged around is set.
  • the central category A is a financial category
  • the related category B is a meal category, a hobby category, a sports category, and a health category.
  • the block matrix is set in advance for each type of the central category, and the block matrix is determined when the central category is determined.
  • the block matrix set for each central category is stored in the personal information DB 54.
  • the question control unit 53 generates a question for acquiring user information regarding an uncollected item having a high degree of association with the central category in the related category.
  • the degree of association is set for each block matrix. Since each item in the central category and the related category has a degree of relevance (distance) from the central item in the central category set in advance, the question control unit 53 asks questions regarding items with a high relevance (short distance). Is generated with priority. For example, as shown in FIG. 4, the distance from the central item of the central category A (the central item (distance 0) in the central category) to other items is set in the block matrix.
  • the question control unit 53 asks questions about items for which user information has not been collected and whose distance from the central item of the central category A is small (relevance to the central item is high). Generate with priority. As a result, uncollected items with high relevance to the central category can be collected in preference to uncollected items with low relevance to the central category. It is possible to provide more optimal recommendations such as simple answers.
  • the question to be generated is stored in advance in the personal information DB 54 for each category and for each item as a question for acquiring user information regarding each item in each category, and can be read by designating the item and the category.
  • questions are managed by data records [items, categories, questions], and may be stored in the personal information DB 54 as a list (table) in which the items are sorted in the order of the first key and the category as the second key. it can.
  • a flag indicating whether the user information has been collected or not collected for each item is attached to this list, and the collection status of the user information in the item is also associated. For example, if the user information has been collected, the flag “1” is input, and if the user information has not been collected, “0” is input.
  • the present invention is not limited to this, and the block in which the question control unit is stored in the personal information DB 54 is described. The same applies to the case of generating a question for an uncollected item of a matrix.
  • the question control unit 53 searches the list by a binary search or a linear search using the item determined by the distance from the center item as the first key and the center category as the second key as described above.
  • a question is generated as a question to be output to the user.
  • the contents of the question are not particularly limited as long as the user information for uncollected items can be collected.
  • the question control unit 53 generates a question about an item for which user information has not been collected when the registration rate of items in the central category and related categories constituting the block matrix is less than a predetermined value.
  • the registration rate is the ratio of the number of items of user information collected (the number of items of flag “1”) to the total number of items (45 items in FIG. 3) in the central category and related categories constituting the block matrix.
  • a recommendation is created when the registration rate of user information exceeds a predetermined value.
  • the present invention is not limited to this, and weights items in categories.
  • a recommendation may be created when user information is collected for all items with high weighting (high importance). Even if the registration rate is less than 80%, when the registration rate is within a certain range (for example, 60% to 80%), it is set to output a question about uncollected items to the user and a recommendation. You may do it.
  • the question control unit 53 generates a question and receives an answer from the user.
  • the question control unit 53 calculates the registration rate.
  • the question control unit 53 It outputs to the control part 51 and the recommendation preparation part 52. This means that the recommendation is not output to the user until the registration rate reaches 80% or more.
  • the threshold of the registration rate is not limited to this and can be set as appropriate.
  • the personal information DB 54 stores user information in a state of being classified into a plurality of categories each including at least one item.
  • the personal information DB 54 stores a block matrix set for each central category. Further, the personal information DB 54 stores a question for obtaining user information regarding each item in each category.
  • the personal information DB 54 stores recommendation patterns corresponding to combinations of items for which user information has been collected. This recommendation pattern is predetermined for each set block matrix. As shown in the table of FIG. 6, the recommendation pattern is associated with an item for which user information has been collected.
  • the recommendation pattern 1 includes a deposit amount item, a deposit amount history item, a periodic expenditure item (financial category), This is a pattern for creating a recommendation from the items of car item (hobby category), favorite food item, meal history item (meal category), and recommendation pattern 2 is a deposit amount item, car item (hobby category), favorite food item. This is a pattern for creating a recommendation from other search information.
  • the recommendation pattern is not limited to that shown in FIG. 6 and can be set as appropriate.
  • the recommendation creation unit 52 creates a recommendation for the user's inquiry from user information regarding items in a plurality of categories stored in the personal information DB 54.
  • the recommendation creating unit 52 creates a recommendation for the user's inquiry.
  • the recommendation creating unit 52 creates a recommendation for the user using the user information regarding the items for which user information has been collected in the block matrix.
  • a recommendation may be created only with items for which user information has been collected in the block matrix.
  • other search information for example, user information collection
  • Recommendations may be created using information related to completed items).
  • the recommendation creation unit 52 selects a recommendation pattern from the items for which user information has been collected, and creates a recommendation using the recommendation pattern.
  • information necessary for the recommendation other than the collected user information is collected using the search engine 4.
  • the recommendation creating unit 52 creates a recommendation using such user information and search information.
  • the recommendation creating unit 52 creates a recommendation for the user's inquiry using the user's history log.
  • this user history log shows the passage of time for the items in the category (in FIG. 7, the passage of time of the deposit amount is shown).
  • a category for classifying user information is estimated from keywords included in the question inputted by the user (ST11). For example, when the question from the user is “I am considering purchasing a house, but what about?”, “Finance category” is estimated as a central category from “house” and “purchase” that can be keywords. Information on the estimated category is output to the question control unit 53.
  • the question control unit 53 sets a block matrix having the financial category as a central category (ST12). This block matrix is set by selecting a preset block matrix stored in the personal information DB 54. Next, the registration rate of the set block matrix is obtained. The registration rate is obtained by calculating the ratio of the number of items for which user information has been collected with respect to the total number of items in the central category and related categories constituting the block matrix. Then, the question control unit 53 determines whether or not the registration rate is equal to or higher than a predetermined value (80% here) (ST13).
  • a predetermined value 80% here
  • the question control unit 53 When the registration rate is less than 80% (N), the question control unit 53 generates a question to the user in order to obtain user information regarding an uncollected item in the block matrix.
  • an item for which user information has not yet been collected and an item having a small distance from the central item of the central category (highly related to the central item) is selected (ST14).
  • the question control part 53 produces
  • the question is extracted using the item and category as a key.
  • the extracted question is output to the user via the dialogue control unit 51 (ST15).
  • the answer is sent to the personal information DB 54 via the dialogue control unit 51 and registered as user information in the block matrix item (selected uncollected item). (ST16). At this time, in the personal information DB 54, the flag is rewritten for items for which user information is registered (not collected ⁇ collected).
  • the question control unit 53 obtains the registration rate of the block matrix after registering the user information, and determines whether or not the registration rate is a predetermined value (80% in this case) or more (ST13). When the registration rate is less than 80% (N), the procedures of ST14 to ST16 are repeated. On the other hand, if the registration rate is 80% or more (Y), the question control unit 53 notifies the recommendation creating unit 52 to that effect, and the recommendation creating unit 52 creates a recommendation for the user's question (ST17).
  • the recommendation creation unit 52 refers to the table stored in the personal information DB 54 and selects a recommendation pattern from the items for which user information has been collected. Then, a recommendation is created along this recommendation pattern. The created recommendation is output to the user via the dialogue control unit 51.
  • a case is described in which a question is repeated to the user until the registration rate reaches a predetermined value or more.
  • the dialog control unit 51 obtains a search result other than the question. May be output.
  • search results may be output.
  • the dialogue control unit 51 may combine the question to the user and the search result and output it to the user.
  • this question is sent to the recommendation creation system via the network.
  • the block matrix is determined by estimating the category from the user's question.
  • the recommendation creation system estimates a financial category and determines a block matrix having the financial category as a central category (FIG. 3).
  • the recommendation creation system while monitoring the registration rate, the user is asked questions about items highly relevant to the central item in the central category, and user information on uncollected items is collected. This operation is continued until the registration rate exceeds a predetermined value. Thereafter, in the recommendation creation system, when the registration rate becomes equal to or higher than a predetermined value, a recommendation is created based on the collected user information and search information.
  • a recommendation pattern is selected from the collected items, and a recommendation is created according to the recommendation pattern.
  • This recommendation pattern uses at least the items of deposit amount item, deposit amount history item, periodic expenditure item (financial category), car item (hobby category), favorite food item, meal history item (meal category).
  • the deposit amount history is as shown in FIG.
  • the search engine 4 searches for the house price and the like.
  • the recommendation created in this way is as follows, and this recommendation is output to the user. “The price of the house is about X yen, but the deposit amount is Y yen, so it ’s not enough, but be careful because the expenses are increasing. It ’s costing you a Z yen every month, and your favorite food is expensive at O , so you should refrain from buying a house now! ”
  • an unexpected answer that suppresses the purchase of a house is obtained instead of making a recommendation for introducing a house in response to a request to purchase a house.
  • the recommendation creation system while monitoring the registration rate, the user is asked questions about items highly relevant to the central item in the central category, and user information on uncollected items is collected. This operation is continued until the registration rate exceeds a predetermined value. Thereafter, in the recommendation creation system, when the registration rate becomes equal to or higher than a predetermined value, a recommendation is created based on the collected user information and search information.
  • a recommendation pattern is selected from the collected items, and a recommendation is created according to the recommendation pattern.
  • This recommendation pattern uses at least blood pressure items, tobacco items (health category), TOEIC items (work category), jogging items (sports category), and alcohol items (meal category) (recommendation pattern 3 in FIG. 10). ).
  • the recommendation created in this way is as follows, and this recommendation is output to the user. "I don't exercise like jogging, but I like alcohol when I smoke, but it's hard because you have a lot of alcohol on your days off, so you can use that time to study TOEIC. if you do!"
  • the present invention is not limited to the above embodiment, and can be implemented with various modifications.
  • the block matrix, category, item, question, question, etc. can be appropriately changed and implemented.
  • Other modifications can be made without departing from the scope of the present invention.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Finance (AREA)
  • Strategic Management (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • General Engineering & Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Computational Linguistics (AREA)
  • Game Theory and Decision Science (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The objective of the present invention is to provide a recommendation creation system capable of providing responses not expected by the user or optimal recommendations such as responses appropriate for the user. This recommendation creation system is characterized in being equipped with: a database which stores user information sorted according to a plurality of categories which each include at least one item; a question control means for generating a question to the user for acquiring user information related to uncollected items in the categories; and a recommendation creation means for creating recommendations for queries by the user on the basis of the user information related to items in the plurality of categories stored in the database.

Description

レコメンド作成システムRecommendation creation system
　本発明は、ユーザの問いかけに対して最適な回答を提供するレコメンド作成システムに関する。 The present invention relates to a recommendation creation system that provides an optimal answer to a user's question.
　近年、携帯電話機等から入力されたユーザの音声を解析し、ユーザが要求する情報を検索して取得するサービスがある。これらのサービスでは、正確な回答、ユーザに最適な回答を提供することを目的として、ユーザ特性に応じた情報の提供等が行われている。このようなサービスを実現する技術として、例えば、特許文献１には、ユーザの行動要求内容に対して必要とされる情報の検索処理を可能にするエージェント機能を備えた携帯情報端末が開示されている。 In recent years, there is a service that analyzes a user's voice input from a mobile phone or the like and searches for and acquires information requested by the user. In these services, for the purpose of providing an accurate answer and an optimum answer to the user, provision of information according to user characteristics is performed. As a technique for realizing such a service, for example, Patent Document 1 discloses a portable information terminal having an agent function that enables a search process of information required for a user's action request content. Yes.
特開平１０－２８３３６２号公報JP-A-10-283362
　しかしながら、上記のような情報検索サービスにおいては、ユーザの問いかけに含まれるキーワード及びユーザの属性情報から回答を提供しているので、ユーザが予期していない回答やユーザに対して適切なレコメンドは得られない。 However, since the information retrieval service as described above provides answers based on the keywords included in the user's inquiry and the user's attribute information, an answer that is not expected by the user or an appropriate recommendation for the user can be obtained. I can't.
　本発明はかかる点に鑑みてなされたものであり、ユーザが予期していない回答やユーザに対して適切な回答等の最適なレコメンドを提供することができるレコメンド作成システムを提供することを目的とする。 The present invention has been made in view of such points, and an object thereof is to provide a recommendation creation system that can provide an optimal recommendation such as an answer that the user does not expect or an appropriate answer to the user. To do.
　本発明のレコメンド作成システムは、少なくとも一つの項目をそれぞれ含む複数のカテゴリで分類された状態でユーザの情報を格納したデータベースと、前記カテゴリにおける未収集の項目に関するユーザ情報を取得するために前記ユーザへの質問を生成する質問制御手段と、前記データベースに格納されている複数のカテゴリにおける項目に関するユーザ情報から前記ユーザの問いかけに対するレコメンドを作成するレコメンド作成手段と、を具備することを特徴とする。 The recommendation creating system according to the present invention includes a database storing user information in a state of being classified into a plurality of categories each including at least one item, and the user for acquiring user information regarding uncollected items in the category. Question control means for generating a question to the user, and recommendation creation means for creating a recommendation for the user's inquiry from user information regarding items in a plurality of categories stored in the database.
　この構成によれば、ユーザが予期していない回答やユーザに対して適切な回答等の最適なレコメンドを提供することができる。 According to this configuration, it is possible to provide an optimal recommendation such as an answer that the user does not expect or an appropriate answer to the user.
　本発明のレコメンド作成システムにおいては、前記質問制御手段は、前記問いかけに含まれるキーワードから一つのカテゴリを推定し、推定されたカテゴリを中心カテゴリとし、前記問いかけに対して中心カテゴリ以外で関連するカテゴリを関連カテゴリとし、前記中心カテゴリ及び前記関連カテゴリでブロックマトリクスを設定することが好ましい。 In the recommendation creating system according to the present invention, the question control means estimates one category from the keywords included in the question, sets the estimated category as a central category, and relates to the question in a category other than the central category. It is preferable to set a block matrix with the central category and the related category.
　この構成によれば、ユーザの問いかけから推定されたカテゴリ以外のカテゴリについての項目に対するユーザ情報を用いることができるので、ユーザが予期していない回答やユーザに対して適切な回答等のより最適なレコメンドを提供することができる。 According to this configuration, it is possible to use user information for items about categories other than the category estimated from the user's inquiry, and thus more optimal responses such as an answer that the user does not expect or an appropriate answer to the user. Recommendations can be provided.
　本発明のレコメンド作成システムにおいては、各カテゴリに含まれるそれぞれの項目には、他のカテゴリに対して関連度が予め設定されており、前記質問制御手段は、前記関連カテゴリにおける、前記中心カテゴリに対する関連度が高い未収集の項目に関するユーザ情報を取得するための質問を生成することが好ましい。 In the recommendation creating system of the present invention, each item included in each category has a degree of relevance with respect to the other category in advance, and the question control unit is configured to correspond to the central category in the related category. It is preferable to generate a question for acquiring user information regarding an uncollected item having a high degree of association.
　この構成によれば、中心カテゴリに対する関連度が高い未収集の項目を優先して収集することができるので、ユーザが予期していない回答やユーザに対して適切な回答等のより最適なレコメンドを提供することができる。 According to this configuration, it is possible to preferentially collect uncollected items that have a high degree of relevance to the central category, so that more optimal recommendations such as answers that are not anticipated by the user and answers that are appropriate for the user can be provided. Can be provided.
　本発明のレコメンド作成システムにおいては、前記質問制御手段は、前記ブロックマトリクスを構成する中心カテゴリ及び関連カテゴリにおける項目の登録率が所定値未満であるときに、ユーザ情報が未収集の項目についての質問を生成することが好ましい。 In the recommendation creating system according to the present invention, the question control means may ask a question about an item for which user information has not been collected when registration rates of items in the central category and the related category constituting the block matrix are less than a predetermined value. Is preferably generated.
　この構成によれば、ユーザの問いかけに対して、関連カテゴリの情報を含めたレコメンドを提供することができるので、ユーザ自身も気がついていないような、意外で、ユーザのためになる回答を得ることができる。 According to this configuration, a recommendation including information on related categories can be provided in response to a user's inquiry, so that an unexpected and useful answer that the user himself / herself is not aware of can be obtained. Can do.
　本発明のレコメンド作成システムにおいては、前記レコメンド作成手段は、ユーザの履歴ログも用いて前記ユーザの問いかけに対するレコメンドを作成することが好ましい。 In the recommendation creating system of the present invention, it is preferable that the recommendation creating means creates a recommendation for the user's inquiry using a user's history log.
　この構成によれば、ユーザ情報としてユーザの履歴ログを用いるので、ユーザ自身も気がついていないような、意外で、ユーザのためになる回答を得ることができ、ユーザが予期していない回答やユーザに対して適切な回答等のより最適なレコメンドを提供することができる。 According to this configuration, since the user's history log is used as the user information, it is possible to obtain an unexpected and useful answer that the user himself / herself does not notice. It is possible to provide more optimal recommendations such as appropriate answers.
　本発明のレコメンド作成システムにおいては、前記データベースは、各カテゴリにおける各項目に関するユーザ情報を取得するための質問を格納していることが好ましい。 In the recommendation creation system of the present invention, it is preferable that the database stores a question for obtaining user information regarding each item in each category.
　本発明のレコメンド作成システムにおいては、前記データベースは、前記中心カテゴリ毎に設定されたブロックマトリクスを格納していることが好ましい。 In the recommendation creating system according to the present invention, it is preferable that the database stores a block matrix set for each central category.
　本発明の携帯端末装置は、上記レコメンド作成システムを搭載したことを特徴とする。 The portable terminal device of the present invention is characterized in that the above-mentioned recommendation creation system is mounted.
　本発明によれば、ユーザからの問いかけに対するレコメンドを提供するにあたり、問いかけと相関関係が高い項目を優先的に使用し、相関関係が高いにもかかわらず未収集の項目があれば、そのような項目を収集する質問をユーザに対して行うので、ユーザが予期していない回答やユーザに対して適切な回答等の最適なレコメンドを提供することができる。 According to the present invention, in providing a recommendation for an inquiry from a user, an item having a high correlation with the inquiry is preferentially used. Since the question for collecting items is asked to the user, it is possible to provide an optimal recommendation such as an answer that the user does not expect or an appropriate answer to the user.
本発明の実施の形態に係るレコメンド作成システムを含むシステム構成を示す図である。It is a figure which shows the system configuration | structure containing the recommendation production system which concerns on embodiment of this invention. 本発明の実施の形態に係るレコメンド作成システムの概略構成を示すブロック図である。It is a block diagram which shows schematic structure of the recommendation production system which concerns on embodiment of this invention. ブロックマトリクスの一例を示す図である。It is a figure which shows an example of a block matrix. ブロックマトリクスにおける中心項目からの関連度の例を示す図である。It is a figure which shows the example of the relevance from the center item in a block matrix. 履歴ログを含めたブロックマトリクスの例を示す図である。It is a figure which shows the example of the block matrix containing a log | history log. レコメンドパターンの例を示す図である。It is a figure which shows the example of a recommendation pattern. 預金額の時間的推移を示す図である。It is a figure which shows the time transition of deposit amount. 本発明の実施の形態に係るレコメンド作成システムにおけるレコメンド作成動作を示すフローチャートである。It is a flowchart which shows the recommendation preparation operation | movement in the recommendation preparation system which concerns on embodiment of this invention. ブロックマトリクスの他の例を示す図である。It is a figure which shows the other example of a block matrix. レコメンドパターンの例を示す図である。It is a figure which shows the example of a recommendation pattern.
　以下、本発明の実施の形態について、添付図面を参照して詳細に説明する。
　図１は、本発明の実施の形態に係るレコメンド作成システムを含むシステム構成を示す図である。図１においては、レコメンド作成システムがネットワークに接続されたサーバにある場合について示している。本発明においては、図１に示すシステム構成に限定されず、レコメンド作成システムが携帯端末装置に搭載されている場合についても適用することができる。 Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
FIG. 1 is a diagram showing a system configuration including a recommendation creation system according to an embodiment of the present invention. FIG. 1 shows a case where the recommendation creation system is in a server connected to a network. The present invention is not limited to the system configuration shown in FIG. 1 and can also be applied to a case where a recommendation creation system is mounted on a mobile terminal device.
　図１に示すシステムは、携帯電話等の携帯端末装置１と、この携帯端末装置１とモバイルネットワーク２及びインターネットのようなネットワーク３を介して接続された検索エンジン４と、モバイルネットワーク２及びインターネットのようなネットワーク３を介して接続されたレコメンド作成システム（Ｗｅｂサーバ）５とから主に構成されている。 The system shown in FIG. 1 includes a mobile terminal device 1 such as a mobile phone, a search engine 4 connected to the mobile terminal device 1 through a mobile network 2 and a network 3 such as the Internet, a mobile network 2 and the Internet. The system is mainly composed of a recommendation creation system (Web server) 5 connected via such a network 3.
　携帯端末装置１は、携帯電話やタブレット型端末等の端末装置を含み、少なくとも検索エンジン４と接続して情報検索をすることができ、レコメンド作成システムを備えたサーバに接続してレコメンドを取得できる機能を備えている。また、携帯端末装置１は、音声入力したユーザの音声を解析し、ユーザが要求する情報を検索エンジン４から取得する機能を備えている。また、携帯端末装置１は、その他の通話機能、電子メール機能、非接触カード機能等を備えていても良い。 The mobile terminal device 1 includes a terminal device such as a mobile phone or a tablet-type terminal, can connect to at least the search engine 4 to search for information, and can connect to a server equipped with a recommendation creation system to obtain recommendations. It has a function. Further, the mobile terminal device 1 has a function of analyzing the voice of the user who has input the voice and acquiring information requested by the user from the search engine 4. Moreover, the portable terminal device 1 may be provided with other call functions, e-mail functions, contactless card functions, and the like.
　モバイルネットワーク２は、携帯端末装置１とインターネット３との間のネットワークである。このモバイルネットワーク２には、通常の移動通信網に加えて移動パケット通信網も含まれる。モバイルネットワーク２は、インターネットのようなネットワーク３とを結ぶゲートウェイ機能を担っている。 The mobile network 2 is a network between the mobile terminal device 1 and the Internet 3. The mobile network 2 includes a mobile packet communication network in addition to a normal mobile communication network. The mobile network 2 has a gateway function for connecting to the network 3 such as the Internet.
　ネットワーク３には、インターネットをはじめその他のネットワーク、例えばＬＡＮやＷＡＮなども含まれる。 The network 3 includes the Internet and other networks such as LAN and WAN.
　レコメンド作成システム５は、図２に示すように、対話制御部５１と、レコメンド作成部５２と、質問制御部５３と、個人情報データベース（ＤＢ）５４とから主に構成されている。レコメンド作成システム５においては、ユーザからの問いかけに対するレコメンドを提供するにあたり、問いかけと相関関係が高い項目を優先的に使用してレコメンドを提供する。また、相関関係が高いにもかかわらず未収集の項目があれば、そのような項目を収集する質問をユーザに対して行う。 As shown in FIG. 2, the recommendation creation system 5 mainly includes a dialogue control unit 51, a recommendation creation unit 52, a question control unit 53, and a personal information database (DB) 54. In the recommendation creation system 5, in providing a recommendation for a question from a user, a recommendation is provided by preferentially using items having a high correlation with the question. If there is an uncollected item even though the correlation is high, a question for collecting such an item is made to the user.
　対話制御部５１は、ユーザから入力された問いかけに対してレコメンドをユーザに出力すると共に、未収集の項目に関するユーザ情報を取得するための質問をユーザに出力する。ここで、「問いかけ」とは、ユーザがレコメンドを求めるためにレコメンド作成システムに入力する問いを意味し、「レコメンド」とは、ユーザからの「問いかけ」に対してレコメンド作成システムが提供する提案を意味する。例えば、「問いかけ」として、「家の購入を検討しているけど、どうかな？」とユーザがレコメンド作成システムに入力すると、「レコメンド」として、「預金額に対して負債が多い。まずはローンが残っている車を売却したら？」が出力される。 The dialogue control unit 51 outputs a recommendation to the user in response to an inquiry input from the user, and outputs a question for acquiring user information regarding an uncollected item to the user. Here, “question” means a question that the user inputs to the recommendation creation system in order to request a recommendation, and “recommendation” means a proposal provided by the recommendation creation system for the “question” from the user. means. For example, as a “question”, if the user inputs “Recommendation of purchasing a house, but how about?” To the recommendation creation system, “Recommendation” is “There are many debts against the deposit amount. "What if the remaining car is sold?"
　また、「質問」とは、後述する個人情報ＤＢ５４に格納され、少なくとも一つの項目を含むカテゴリ別に分類されたユーザ情報において、未収集である項目に対応する質問を意味する。例えば、金融情報のカテゴリにおいて、預金額の項目が未収集であれば、「預金額はいくら？」という質問がレコメンド作成システムからユーザに出力される。 The “question” means a question corresponding to an uncollected item in user information stored in the personal information DB 54 described later and classified by category including at least one item. For example, in the category of financial information, if the deposit amount item has not been collected, the question “How much is the deposit amount?” Is output from the recommendation creation system to the user.
　また、「カテゴリ」とは、ユーザ情報を分類するためのものであり、例えば、食事、健康、人間関係、夢・目標、基本情報、趣味、運動、仕事、金融等が挙げられる。「項目」とは、カテゴリに含まれる具体的な情報を意味する。例えば、食事カテゴリに含まれる項目としては、好きな食べ物等が挙げられ、金融カテゴリに含まれる項目としては、預金額、定期的支出、負債額等が挙げられる。基本情報とは、氏名、身長、体重、住所、年齢、職業、家族構成等のユーザについての基本的な情報であり、この基本情報は、ユーザが予め入力しておいても良く、他のアプリケーション操作時にユーザが入力した情報を用いても良く、質問により収集しても良い。なお、このカテゴリや項目は、これらに限定されず、適宜設定することができる。 Further, the “category” is for classifying user information, and examples thereof include meal, health, human relations, dreams / goals, basic information, hobbies, exercise, work, finance, and the like. “Item” means specific information included in a category. For example, the items included in the meal category include favorite foods and the like, and the items included in the financial category include deposit amount, periodic expenditure, debt amount, and the like. Basic information is basic information about the user such as name, height, weight, address, age, occupation, family structure, etc. This basic information may be input in advance by the user, and other applications Information input by the user during operation may be used, or may be collected by a question. The categories and items are not limited to these and can be set as appropriate.
　対話制御部５１は、ユーザから入力された問いかけに含まれるキーワードからユーザ情報を分類するカテゴリを推定する。例えば、ユーザからの問いかけが上記「家の購入を検討しているけど、どうかな？」である場合、キーワードとなり得る「家」や「購入」から中心カテゴリとして「金融カテゴリ」を推定する。対話制御部５１は、推定した中心カテゴリを質問制御部５３に出力する。なお、対話制御部５１にユーザが入力する場合、ユーザが音声で入力しても良く、ユーザが手でデータ入力しても良い。 The dialogue control unit 51 estimates a category for classifying user information from keywords included in the question input by the user. For example, when the question from the user is “I am considering purchasing a house, but what about?”, “Finance category” is estimated as a central category from “house” and “purchase” that can be keywords. The dialogue control unit 51 outputs the estimated central category to the question control unit 53. In addition, when a user inputs into the dialog control part 51, a user may input with a voice | voice and a user may input data by hand.
　対話制御部５１は、上記中心カテゴリと、この中心カテゴリ以外で関連する関連カテゴリとで設定されたブロックマトリクスにおける項目についてのユーザ情報の収集状況（登録率）に応じて、上記質問の出力と上記レコメンドの出力を切り替える。すなわち、対話制御部５１は、ブロックマトリクスにおける登録率が所定値（例えば、８０％）以上であれば、後述するレコメンド作成部５２から、問いかけに対するレコメンドを出力する。なお、登録率が所定値以上である旨は、後述する質問制御部５３から通知される。また、対話制御部５１は、ブロックマトリクスにおける登録率が所定値（例えば、８０％）未満であれば、質問制御部５３から未収集の項目に対する質問を出力する。 The dialogue control unit 51 outputs the above question and the above in accordance with the collection status (registration rate) of user information about items in the block matrix set with the central category and related categories other than the central category. Switch the output of the recommendation. That is, if the registration rate in the block matrix is a predetermined value (for example, 80%) or more, the dialogue control unit 51 outputs a recommendation for the question from the recommendation creation unit 52 described later. Note that the question control unit 53, which will be described later, notifies that the registration rate is equal to or greater than a predetermined value. In addition, when the registration rate in the block matrix is less than a predetermined value (for example, 80%), the dialogue control unit 51 outputs a question for an uncollected item from the question control unit 53.
　質問制御部５３は、中心カテゴリ及び関連カテゴリ（ブロックマトリクス）における未収集の項目に関するユーザ情報を取得するためにユーザへの質問を生成する。また、質問制御部５３は、問いかけに含まれるキーワードから一つのカテゴリを推定し、推定されたカテゴリを中心カテゴリとし、問いかけに対して中心カテゴリ以外で関連する関連カテゴリとし、中心カテゴリ及び関連カテゴリでブロックマトリクスを設定する。このようにブロックマトリクスを用いることにより、ユーザの問いかけから推定されたカテゴリ以外のカテゴリについての項目に対するユーザ情報を用いることができるので、ユーザが予期していない回答やユーザに対して適切な回答等のより最適なレコメンドを提供することができる。 The question control unit 53 generates a question to the user in order to acquire user information regarding uncollected items in the central category and related categories (block matrix). In addition, the question control unit 53 estimates one category from the keywords included in the question, sets the estimated category as the central category, sets the related category as a related category other than the central category for the question, Set the block matrix. By using the block matrix in this way, user information can be used for items about categories other than the category estimated from the user's question, so the answer that the user does not expect, an appropriate answer to the user, etc. It is possible to provide a more optimal recommendation.
　例えば、質問制御部５３では、上述したように、対話制御部５１から中心カテゴリが通知されるので、図３に示すように、この中心カテゴリＡを中心とし、この中心カテゴリＡ以外で関連する関連カテゴリＢを周囲に配置してなるブロックマトリクスを設定する。図３においては、中心カテゴリＡが金融カテゴリであり、関連カテゴリＢが食事カテゴリ、趣味カテゴリ、スポーツカテゴリ及び健康カテゴリである。なお、このブロックマトリクスは、中心カテゴリの種類毎に予めそれぞれ設定されており、中心カテゴリが決定したときに、ブロックマトリクスが決まるようになっている。この中心カテゴリ毎に設定されたブロックマトリクスは、個人情報ＤＢ５４に格納されている。 For example, in the question control unit 53, as described above, the central category is notified from the dialogue control unit 51. Therefore, as shown in FIG. A block matrix in which category B is arranged around is set. In FIG. 3, the central category A is a financial category, and the related category B is a meal category, a hobby category, a sports category, and a health category. The block matrix is set in advance for each type of the central category, and the block matrix is determined when the central category is determined. The block matrix set for each central category is stored in the personal information DB 54.
　また、質問制御部５３は、関連カテゴリにおける、中心カテゴリに対する関連度が高い未収集の項目に関するユーザ情報を取得するための質問を生成する。なお、関連度については、ブロックマトリクス毎に設定されている。中心カテゴリ及び関連カテゴリにおける各項目は、中心カテゴリの中心項目からの関連度（距離）が予め設定されているので、質問制御部５３は、その関連度が高い（距離が短い）項目についての質問を優先して生成する。例えば、図４に示すように、ブロックマトリクスには、中心カテゴリＡの中心項目（中心カテゴリにおける中心の項目（距離０））から他の項目までの距離が設定されている。質問制御部５３は、選択されたブロックマトリクスにおいて、ユーザ情報が未収集である項目であって、中心カテゴリＡの中心項目からの距離が小さい（中心項目に関連度が高い）項目についての質問を優先して生成する。これにより、中心カテゴリに対する関連度が高い未収集の項目を中心カテゴリに対する関連度が低い未収集の項目より優先して収集することができるので、ユーザが予期していない回答やユーザに対して適切な回答等のより最適なレコメンドを提供することができる。 Also, the question control unit 53 generates a question for acquiring user information regarding an uncollected item having a high degree of association with the central category in the related category. Note that the degree of association is set for each block matrix. Since each item in the central category and the related category has a degree of relevance (distance) from the central item in the central category set in advance, the question control unit 53 asks questions regarding items with a high relevance (short distance). Is generated with priority. For example, as shown in FIG. 4, the distance from the central item of the central category A (the central item (distance 0) in the central category) to other items is set in the block matrix. In the selected block matrix, the question control unit 53 asks questions about items for which user information has not been collected and whose distance from the central item of the central category A is small (relevance to the central item is high). Generate with priority. As a result, uncollected items with high relevance to the central category can be collected in preference to uncollected items with low relevance to the central category. It is possible to provide more optimal recommendations such as simple answers.
　生成すべき質問は、各カテゴリにおける各項目に関するユーザ情報を取得するための質問として、カテゴリ別、項目毎に予め個人情報ＤＢ５４に記憶されており、項目及びカテゴリを指定して読み出すことができる。例えば、質問は、データレコード［項目，カテゴリ，質問］で管理されており、項目を第１キー、カテゴリを第２キーとしてそれらの順にソートしたリスト（表）として個人情報ＤＢ５４に格納することができる。また、このリストには、項目毎にユーザ情報が収集済であるか未収集であるかを示すフラグが付けられており、項目におけるユーザ情報の収集状況も関連付けられている。例えば、ユーザ情報が収集済であれば、フラグ”１”が入力されており、ユーザ情報が未収集であれば”０”が入力されている。 The question to be generated is stored in advance in the personal information DB 54 for each category and for each item as a question for acquiring user information regarding each item in each category, and can be read by designating the item and the category. For example, questions are managed by data records [items, categories, questions], and may be stored in the personal information DB 54 as a list (table) in which the items are sorted in the order of the first key and the category as the second key. it can. In addition, a flag indicating whether the user information has been collected or not collected for each item is attached to this list, and the collection status of the user information in the item is also associated. For example, if the user information has been collected, the flag “1” is input, and if the user information has not been collected, “0” is input.
　なお、本実施の形態においては、質問が予め個人情報ＤＢ５４に格納されている場合について説明しているが、本発明はこれに限定されず、質問制御部が個人情報ＤＢ５４に格納されているブロックマトリクスの未収集項目に対する質問を生成する場合についても同様に適用することができる。 In the present embodiment, the case where a question is stored in advance in the personal information DB 54 is described. However, the present invention is not limited to this, and the block in which the question control unit is stored in the personal information DB 54 is described. The same applies to the case of generating a question for an uncollected item of a matrix.
　質問制御部５３は、上記のように中心項目からの距離で決定された項目を第１キーとし、中心カテゴリを第２キーとして、二分探索又は線形探索によってこのリストを探索し、一致したレコードの質問を、ユーザに出力すべき質問として生成する。なお、質問の内容としては、未収集の項目に対するユーザ情報が収集できる内容であれば特に制限はない。 The question control unit 53 searches the list by a binary search or a linear search using the item determined by the distance from the center item as the first key and the center category as the second key as described above. A question is generated as a question to be output to the user. The contents of the question are not particularly limited as long as the user information for uncollected items can be collected.
　また、質問制御部５３は、ブロックマトリクスを構成する中心カテゴリ及び関連カテゴリにおける項目の登録率が所定値未満であるときに、ユーザ情報が未収集の項目についての質問を生成する。ここで、登録率とは、ブロックマトリクスを構成する中心カテゴリ及び関連カテゴリにおける全項目数（図３においては４５項目）に対するユーザ情報収集済の項目数（フラグ”１”の項目数）の割合をいう。例えば、質問制御部５３において、登録率が８０％未満である場合に、ユーザ情報が未収集の項目についての質問を生成すると設定されたときには、登録率が８０％以上になるまで、ユーザ情報が未収集の項目についての質問を生成し続ける。図５においては、ユーザ情報が収集済の項目（例えば、金融カテゴリの預金額項目、負債額項目、定期的支出項目）を網掛けしている。したがって、ユーザ情報が未収集の項目（図５において網掛けのない項目）について登録率が８０％以上になるまで順次質問を行う。 Also, the question control unit 53 generates a question about an item for which user information has not been collected when the registration rate of items in the central category and related categories constituting the block matrix is less than a predetermined value. Here, the registration rate is the ratio of the number of items of user information collected (the number of items of flag “1”) to the total number of items (45 items in FIG. 3) in the central category and related categories constituting the block matrix. Say. For example, in the question control unit 53, when the registration rate is less than 80% and the user information is set to generate a question about an uncollected item, the user information is updated until the registration rate reaches 80% or more. Continue to generate questions about uncollected items. In FIG. 5, items (for example, deposit amount items, liability amount items, and periodic expenditure items in the financial category) that have been collected by the user information are shaded. Therefore, questions are sequentially asked until the registration rate reaches 80% or more for items for which user information has not been collected (items not shaded in FIG. 5).
　本実施の形態においては、ユーザ情報の登録率が所定値以上になったときにレコメンドを作成する場合について説明しているが、本発明はこれに限定されず、カテゴリにおける項目に重み付けをしておき、重み付けが高い（重要度が高い）項目すべてについてユーザ情報が収集されたときにレコメンドを作成するようにしても良い。また、登録率が８０％未満であっても登録率がある範囲内（例えば、６０％～８０％）にあるときには、ユーザに未収集項目についての質問を出力すると共にレコメンドを出力するように設定しても良い。 In the present embodiment, a case has been described in which a recommendation is created when the registration rate of user information exceeds a predetermined value. However, the present invention is not limited to this, and weights items in categories. Alternatively, a recommendation may be created when user information is collected for all items with high weighting (high importance). Even if the registration rate is less than 80%, when the registration rate is within a certain range (for example, 60% to 80%), it is set to output a question about uncollected items to the user and a recommendation. You may do it.
　質問制御部５３は、質問を生成してユーザからの回答があり、項目に対するユーザ情報が収集された際に登録率を計算し、登録率が８０％以上になったときに、その旨を対話制御部５１及びレコメンド作成部５２に出力する。これは、登録率が８０％以上になるまではレコメンドをユーザに出力しないことを意味する。なお、登録率の閾値については、これに限定されず、適宜設定することができる。 The question control unit 53 generates a question and receives an answer from the user. When the user information for the item is collected, the question control unit 53 calculates the registration rate. When the registration rate reaches 80% or more, the question control unit 53 It outputs to the control part 51 and the recommendation preparation part 52. This means that the recommendation is not output to the user until the registration rate reaches 80% or more. Note that the threshold of the registration rate is not limited to this and can be set as appropriate.
　このように、登録率が所定値以上になったときにレコメンドを出力することにより、ユーザの問いかけに対して、関連カテゴリの情報を含めたレコメンドを提供することができるので、ユーザ自身も気がついていないような、意外で、ユーザのためになる回答を得ることができる。 In this way, by outputting a recommendation when the registration rate exceeds a predetermined value, it is possible to provide a recommendation including related category information in response to the user's inquiry. You can get answers that are surprising and helpful for the user.
　個人情報ＤＢ５４は、少なくとも一つの項目をそれぞれ含む複数のカテゴリで分類された状態でユーザの情報を格納する。また、個人情報ＤＢ５４は、中心カテゴリ毎に設定されたブロックマトリクスを格納する。また、個人情報ＤＢ５４は、各カテゴリにおける各項目に関するユーザ情報を取得するための質問を格納する。 The personal information DB 54 stores user information in a state of being classified into a plurality of categories each including at least one item. The personal information DB 54 stores a block matrix set for each central category. Further, the personal information DB 54 stores a question for obtaining user information regarding each item in each category.
　また、個人情報ＤＢ５４は、ユーザ情報が収集済である項目の組み合わせに応じたレコメンドパターンを格納する。このレコメンドパターンは、設定されたブロックマトリクス毎に予め決められている。レコメンドパターンは、図６に示すテーブルのように、ユーザ情報収集済の項目と関連付けられており、例えば、レコメンドパターン１は、預金額項目、預金額履歴項目、定期的支出項目（金融カテゴリ）、車項目（趣味カテゴリ）、好きな食べ物項目、食事履歴項目（食事カテゴリ）の項目からレコメンドを作成するパターンであり、レコメンドパターン２は、預金額項目、車項目（趣味カテゴリ）、好きな食べ物項目、他の検索情報からレコメンドを作成するパターンである。なお、このレコメンドパターンは図６に限定されず適宜設定することができる。 Also, the personal information DB 54 stores recommendation patterns corresponding to combinations of items for which user information has been collected. This recommendation pattern is predetermined for each set block matrix. As shown in the table of FIG. 6, the recommendation pattern is associated with an item for which user information has been collected. For example, the recommendation pattern 1 includes a deposit amount item, a deposit amount history item, a periodic expenditure item (financial category), This is a pattern for creating a recommendation from the items of car item (hobby category), favorite food item, meal history item (meal category), and recommendation pattern 2 is a deposit amount item, car item (hobby category), favorite food item. This is a pattern for creating a recommendation from other search information. The recommendation pattern is not limited to that shown in FIG. 6 and can be set as appropriate.
　レコメンド作成部５２は、個人情報ＤＢ５４に格納されている複数のカテゴリにおける項目に関するユーザ情報からユーザの問いかけに対するレコメンドを作成する。レコメンド作成部５２は、質問制御部５３から登録率が所定値以上になった旨の通知を受けると、ユーザの問いかけに対するレコメンドを作成する。この場合において、レコメンド作成部５２は、ブロックマトリクスにおいてユーザ情報収集済の項目についてのユーザ情報を用いてユーザに対するレコメンドを作成する。レコメンドを作成する際には、ブロックマトリクスにおいてユーザ情報収集済の項目のみでレコメンドを作成しても良く、ブロックマトリクスにおいてユーザ情報収集済の項目に加えて、他の検索情報（例えば、ユーザ情報収集済の項目に関連する情報）を用いてレコメンドを作成しても良い。 The recommendation creation unit 52 creates a recommendation for the user's inquiry from user information regarding items in a plurality of categories stored in the personal information DB 54. When receiving a notification from the question control unit 53 that the registration rate has exceeded a predetermined value, the recommendation creating unit 52 creates a recommendation for the user's inquiry. In this case, the recommendation creating unit 52 creates a recommendation for the user using the user information regarding the items for which user information has been collected in the block matrix. When creating a recommendation, a recommendation may be created only with items for which user information has been collected in the block matrix. In addition to items for which user information has been collected in the block matrix, other search information (for example, user information collection) Recommendations may be created using information related to completed items).
　例えば、レコメンド作成部５２は、ユーザ情報が収集済である項目からレコメンドパターンを選択し、そのレコメンドパターンを用いてレコメンドを作成する。この場合において、収集済のユーザ情報以外の情報でレコメンドに必要な情報については検索エンジン４を用いて情報収集する。レコメンド作成部５２は、このようなユーザ情報や検索情報を用いてレコメンドを作成する。 For example, the recommendation creation unit 52 selects a recommendation pattern from the items for which user information has been collected, and creates a recommendation using the recommendation pattern. In this case, information necessary for the recommendation other than the collected user information is collected using the search engine 4. The recommendation creating unit 52 creates a recommendation using such user information and search information.
　レコメンド作成部５２においては、図５及び図６に示すように、ユーザの履歴ログも用いてユーザの問いかけに対するレコメンドを作成する。このユーザの履歴ログは、例えば、図７に示すように、カテゴリにおける項目についての時間経過を示すものである（図７では預金額の時間経過を示している）。このようにユーザ情報としてユーザの履歴ログを用いることにより、ユーザ自身も気がついていないような、意外で、ユーザのためになる回答を得ることができ、ユーザが予期していない回答やユーザに対して適切な回答等のより最適なレコメンドを提供することができる。 As shown in FIGS. 5 and 6, the recommendation creating unit 52 creates a recommendation for the user's inquiry using the user's history log. For example, as shown in FIG. 7, this user history log shows the passage of time for the items in the category (in FIG. 7, the passage of time of the deposit amount is shown). By using the user's history log as user information in this way, it is possible to obtain unexpected and useful answers that the user himself / herself is not aware of. It is possible to provide more optimal recommendations such as appropriate answers.
　上記構成を有するレコメンド作成システムのレコメンド作成手順について図８を用いて説明する。 The recommendation creation procedure of the recommendation creation system having the above configuration will be described with reference to FIG.
　まず、対話制御部５１において、ユーザから入力された問いかけに含まれるキーワードからユーザ情報を分類するカテゴリを推定する（ＳＴ１１）。例えば、ユーザからの問いかけが上記「家の購入を検討しているけど、どうかな？」である場合、キーワードとなり得る「家」や「購入」から中心カテゴリとして「金融カテゴリ」を推定する。推定されたカテゴリの情報は、質問制御部５３に出力される。 First, in the dialogue control unit 51, a category for classifying user information is estimated from keywords included in the question inputted by the user (ST11). For example, when the question from the user is “I am considering purchasing a house, but what about?”, “Finance category” is estimated as a central category from “house” and “purchase” that can be keywords. Information on the estimated category is output to the question control unit 53.
　次いで、質問制御部５３において、金融カテゴリを中心カテゴリとするブロックマトリクスを設定する（ＳＴ１２）。このブロックマトリクスの設定は、個人情報ＤＢ５４に格納されている、予め設定されたブロックマトリクスを選択することにより行う。次いで、設定したブロックマトリクスの登録率を求める。登録率は、ブロックマトリクスを構成する中心カテゴリ及び関連カテゴリにおける全項目数に対するユーザ情報収集済の項目数の割合を演算することにより求める。そして、質問制御部５３において、登録率が所定値（ここでは８０％）以上であるかどうかを判断する（ＳＴ１３）。 Next, the question control unit 53 sets a block matrix having the financial category as a central category (ST12). This block matrix is set by selecting a preset block matrix stored in the personal information DB 54. Next, the registration rate of the set block matrix is obtained. The registration rate is obtained by calculating the ratio of the number of items for which user information has been collected with respect to the total number of items in the central category and related categories constituting the block matrix. Then, the question control unit 53 determines whether or not the registration rate is equal to or higher than a predetermined value (80% here) (ST13).
　登録率が８０％未満である場合（Ｎ）、質問制御部５３において、ブロックマトリクスにおける未収集の項目に関するユーザ情報を取得するためにユーザへの質問を生成する。まず、ブロックマトリクスにおいて、ユーザ情報が未収集である項目であって、中心カテゴリの中心項目からの距離が小さい（中心項目に関連度が高い）項目を選択する（ＳＴ１４）。そして、質問制御部５３は、選択した未収集項目についての質問を生成する。具体的には、データレコード［項目，カテゴリ，質問］で管理されたリストにおいて、項目及びカテゴリをキーとして質問を抽出する。抽出された質問は、対話制御部５１を介してユーザに出力される（ＳＴ１５）。 When the registration rate is less than 80% (N), the question control unit 53 generates a question to the user in order to obtain user information regarding an uncollected item in the block matrix. First, in the block matrix, an item for which user information has not yet been collected and an item having a small distance from the central item of the central category (highly related to the central item) is selected (ST14). And the question control part 53 produces | generates the question about the selected uncollected item. Specifically, in the list managed by the data record [item, category, question], the question is extracted using the item and category as a key. The extracted question is output to the user via the dialogue control unit 51 (ST15).
　次いで、ユーザが質問に対して回答したときに、その回答は、対話制御部５１を介して個人情報ＤＢ５４に送られ、ブロックマトリクスの項目（選択された未収集項目）にユーザ情報として登録される（ＳＴ１６）。このとき、個人情報ＤＢ５４において、ユーザ情報が登録された項目についてはフラグが書き換えられる（未収集→収集済）。 Next, when the user answers the question, the answer is sent to the personal information DB 54 via the dialogue control unit 51 and registered as user information in the block matrix item (selected uncollected item). (ST16). At this time, in the personal information DB 54, the flag is rewritten for items for which user information is registered (not collected → collected).
　次いで、質問制御部５３において、ユーザ情報を登録した後のブロックマトリクスの登録率を求め、登録率が所定値（ここでは８０％）以上であるかどうかを判断する（ＳＴ１３）。登録率が８０％未満である場合（Ｎ）、ＳＴ１４～ＳＴ１６の手順を繰り返す。一方、登録率が８０％以上である場合（Ｙ）、その旨が質問制御部５３からレコメンド作成部５２に通知され、レコメンド作成部５２でユーザの問いかけに対するレコメンドを作成する（ＳＴ１７）。この場合、レコメンド作成部５２は、個人情報ＤＢ５４に格納されているテーブルを参照して、ユーザ情報収集済の項目からレコメンドパターンを選択する。そして、このレコメンドパターンに沿ってレコメンドを作成する。作成されたレコメンドは、対話制御部５１を介してユーザに出力される。 Next, the question control unit 53 obtains the registration rate of the block matrix after registering the user information, and determines whether or not the registration rate is a predetermined value (80% in this case) or more (ST13). When the registration rate is less than 80% (N), the procedures of ST14 to ST16 are repeated. On the other hand, if the registration rate is 80% or more (Y), the question control unit 53 notifies the recommendation creating unit 52 to that effect, and the recommendation creating unit 52 creates a recommendation for the user's question (ST17). In this case, the recommendation creation unit 52 refers to the table stored in the personal information DB 54 and selects a recommendation pattern from the items for which user information has been collected. Then, a recommendation is created along this recommendation pattern. The created recommendation is output to the user via the dialogue control unit 51.
　本実施の形態においては、登録率が所定値以上になるまでユーザに質問を繰り返す場合について説明しているが、本発明はこれに限定されず、対話制御部５１が質問以外の検索結果をユーザに対して出力しても良い。例えば、ユーザがシステムに「問いかけ」をした後に、システム側からユーザに対して質問を出力して未収集項目のユーザ情報を収集しながら、ユーザからの通常の検索要求に対して検索エンジンを用いて検索結果を出力しても良い。この場合、対話制御部５１がユーザへの質問と検索結果を組み合わせてユーザに出力するようにしても良い。 In the present embodiment, a case is described in which a question is repeated to the user until the registration rate reaches a predetermined value or more. However, the present invention is not limited to this, and the dialog control unit 51 obtains a search result other than the question. May be output. For example, after a user asks the system, he / she uses a search engine in response to a normal search request from the user while outputting a question to the user from the system side and collecting user information of uncollected items. Search results may be output. In this case, the dialogue control unit 51 may combine the question to the user and the search result and output it to the user.
　以下、本発明の効果を明確にするために実施した実施例について説明する。なお、本発明は、以下の実施例によって何ら限定されるものではない。 Hereinafter, examples carried out in order to clarify the effects of the present invention will be described. In addition, this invention is not limited at all by the following examples.
＜住宅購入を考えているユーザの例＞
　ユーザが「家の購入を検討しているけど、どうかな？」と携帯端末装置に入力すると、ネットワークを介してこの問いかけはレコメンド作成システムに送られる。レコメンド作成システムにおいては、ユーザの問いかけからカテゴリを推定して、ブロックマトリクスを決定する。ここでは、レコメンド作成システムが金融カテゴリと推定して、金融カテゴリを中心カテゴリとするブロックマトリクスを決定する（図３）。 <Examples of users considering home purchase>
When the user inputs “I am considering buying a house, how about?” To the mobile terminal device, this question is sent to the recommendation creation system via the network. In the recommendation creation system, the block matrix is determined by estimating the category from the user's question. Here, the recommendation creation system estimates a financial category and determines a block matrix having the financial category as a central category (FIG. 3).
　そして、レコメンド作成システムにおいて、登録率を監視しながら、中心カテゴリの中心項目に関連度の高い項目についての質問をユーザに行って、未収集項目のユーザ情報を収集する。この動作を、登録率が所定値以上になるまで続ける。その後、レコメンド作成システムにおいては、登録率が所定値以上になったときに、収集済のユーザ情報と、検索情報とによりレコメンドを作成する。 In the recommendation creation system, while monitoring the registration rate, the user is asked questions about items highly relevant to the central item in the central category, and user information on uncollected items is collected. This operation is continued until the registration rate exceeds a predetermined value. Thereafter, in the recommendation creation system, when the registration rate becomes equal to or higher than a predetermined value, a recommendation is created based on the collected user information and search information.
　このとき、レコメンド作成システムにおいては、収集済の項目からレコメンドパターンを選択し、そのレコメンドパターンに沿ってレコメンドを作成する。このレコメンドパターンは、少なくとも預金額項目、預金額履歴項目、定期的支出項目（金融カテゴリ）、車項目（趣味カテゴリ）、好きな食べ物項目、食事履歴項目（食事カテゴリ）の項目を用いるものである（図６のレコメンドパターン１）。ここで、預金額履歴については、図７に示すようになっているとする。また、レコメンド作成システムにおいて、家の値段等については検索エンジン４で検索する。 At this time, in the recommendation creation system, a recommendation pattern is selected from the collected items, and a recommendation is created according to the recommendation pattern. This recommendation pattern uses at least the items of deposit amount item, deposit amount history item, periodic expenditure item (financial category), car item (hobby category), favorite food item, meal history item (meal category). (Recommendation pattern 1 in FIG. 6). Here, it is assumed that the deposit amount history is as shown in FIG. In the recommendation creation system, the search engine 4 searches for the house price and the like.
　このようにして作成されたレコメンドは以下のようなものであり、このレコメンドをユーザに出力する。
「家の値段は　　Ｘ　　円ぐらいだけど、預金額が　　Ｙ　　円なので不足しているよ。ただし、出費が増加しているので注意が必要だよ。ところで、あなたは車を持っているので、一般的に　　Ｚ　　円が毎月かかっているし、あなたの好きな食べ物が　　Ｏ　　で高くつくから今家を買うのは控えた方がいいよ！」 The recommendation created in this way is as follows, and this recommendation is output to the user.
“The price of the house is about X yen, but the deposit amount is Y yen, so it ’s not enough, but be careful because the expenses are increasing. It ’s costing you a Z yen every month, and your favorite food is expensive at O , so you should refrain from buying a house now! ”
　このように本発明のレコメンド作成システムによれば、家を購入したいという問いかけに対して家を紹介するレコメンドをするのではなく、家を購入することを抑制するという予期しない回答が得られる。 As described above, according to the recommendation creation system of the present invention, an unexpected answer that suppresses the purchase of a house is obtained instead of making a recommendation for introducing a house in response to a request to purchase a house.
＜高血圧を改善したいユーザの例＞
　ユーザが「血圧を１４５ｍｍＨｇ以下にしたい、どうしたらいいかな？」と携帯端末装置に入力すると、ネットワークを介してこの問いかけはレコメンド作成システムに送られる。レコメンド作成システムにおいては、ユーザの問いかけからカテゴリを推定して、ブロックマトリクスを決定する。ここでは、レコメンド作成システムが健康カテゴリと推定して、健康カテゴリを中心カテゴリとするブロックマトリクスを決定する（図９）。 <Examples of users who want to improve hypertension>
When the user inputs to the portable terminal device “What should I do if I want to lower my blood pressure to 145 mmHg or less?”, This inquiry is sent to the recommendation creation system via the network. In the recommendation creation system, the block matrix is determined by estimating the category from the user's question. Here, the recommendation creating system estimates a health category and determines a block matrix having the health category as a central category (FIG. 9).
　そして、レコメンド作成システムにおいて、登録率を監視しながら、中心カテゴリの中心項目に関連度の高い項目についての質問をユーザに行って、未収集項目のユーザ情報を収集する。この動作を、登録率が所定値以上になるまで続ける。その後、レコメンド作成システムにおいては、登録率が所定値以上になったときに、収集済のユーザ情報と、検索情報とによりレコメンドを作成する。 In the recommendation creation system, while monitoring the registration rate, the user is asked questions about items highly relevant to the central item in the central category, and user information on uncollected items is collected. This operation is continued until the registration rate exceeds a predetermined value. Thereafter, in the recommendation creation system, when the registration rate becomes equal to or higher than a predetermined value, a recommendation is created based on the collected user information and search information.
　このとき、レコメンド作成システムにおいては、収集済の項目からレコメンドパターンを選択し、そのレコメンドパターンに沿ってレコメンドを作成する。このレコメンドパターンは、少なくとも血圧項目、たばこ項目（健康カテゴリ）、ＴＯＥＩＣ項目（仕事カテゴリ）、ジョギング項目（スポーツカテゴリ）、アルコール項目（食事カテゴリ）の項目を用いるものである（図１０のレコメンドパターン３）。 At this time, in the recommendation creation system, a recommendation pattern is selected from the collected items, and a recommendation is created according to the recommendation pattern. This recommendation pattern uses at least blood pressure items, tobacco items (health category), TOEIC items (work category), jogging items (sports category), and alcohol items (meal category) (recommendation pattern 3 in FIG. 10). ).
　このようにして作成されたレコメンドは以下のようなものであり、このレコメンドをユーザに出力する。
「たばこは吸うのに、ジョギング等の運動はしないし、アルコールが好きみたいだから大変だよ。ところで、あなたは休みの日にアルコールの量が多いみたいだから、その時間を利用してＴＯＥＩＣの勉強でもしたら！」 The recommendation created in this way is as follows, and this recommendation is output to the user.
"I don't exercise like jogging, but I like alcohol when I smoke, but it's hard because you have a lot of alcohol on your days off, so you can use that time to study TOEIC. if you do!"
　このように本発明のレコメンド作成システムによれば、血圧を下げたいという問いかけに対してＴＯＥＩＣの勉強をしろという予期しない回答が得られる。 Thus, according to the recommendation creation system of the present invention, an unexpected answer to study TOEIC is obtained in response to the question of lowering blood pressure.
　本発明は上記実施の形態に限定されず、種々変更して実施することが可能である。例えば、本発明の範囲を逸脱しない限りにおいて、ブロックマトリクス、カテゴリ、項目、問いかけ、質問等については適宜変更して実施することが可能である。その他、本発明の範囲を逸脱しないで適宜変更して実施することが可能である。 The present invention is not limited to the above embodiment, and can be implemented with various modifications. For example, without departing from the scope of the present invention, the block matrix, category, item, question, question, etc. can be appropriately changed and implemented. Other modifications can be made without departing from the scope of the present invention.

Claims (8)

  1. 　少なくとも一つの項目をそれぞれ含む複数のカテゴリで分類された状態でユーザの情報を格納したデータベースと、前記カテゴリにおける未収集の項目に関するユーザ情報を取得するために前記ユーザへの質問を生成する質問制御手段と、前記データベースに格納されている複数のカテゴリにおける項目に関するユーザ情報から前記ユーザの問いかけに対するレコメンドを作成するレコメンド作成手段と、を具備することを特徴とするレコメンド作成システム。 A database storing user information in a state of being classified into a plurality of categories each including at least one item, and question control for generating a question to the user in order to acquire user information regarding uncollected items in the category And a recommendation creating means for creating a recommendation for the user's inquiry from user information relating to items in a plurality of categories stored in the database.
  2. 　前記質問制御手段は、前記問いかけに含まれるキーワードから一つのカテゴリを推定し、推定されたカテゴリを中心カテゴリとし、前記問いかけに対して中心カテゴリ以外で関連する関連カテゴリとし、前記中心カテゴリ及び前記関連カテゴリでブロックマトリクスを設定することを特徴とする請求項１記載のレコメンド作成システム。 The question control means estimates one category from keywords included in the question, sets the estimated category as a central category, sets the related category as a related category other than the central category for the question, the central category and the related The recommendation creation system according to claim 1, wherein a block matrix is set by category.
  3. 　各カテゴリに含まれるそれぞれの項目には、他のカテゴリに対して関連度が予め設定されており、前記質問制御手段は、前記関連カテゴリにおける、前記中心カテゴリに対する関連度が高い未収集の項目に関するユーザ情報を取得するための質問を生成することを特徴とする請求項２記載のレコメンド作成システム。 Each item included in each category has a degree of relevance with respect to another category in advance, and the question control unit relates to an uncollected item having a high degree of relevance with respect to the central category in the related category. The recommendation creation system according to claim 2, wherein a question for obtaining user information is generated.
  4. 　前記質問制御手段は、前記ブロックマトリクスを構成する中心カテゴリ及び関連カテゴリにおける項目の登録率が所定値未満であるときに、ユーザ情報が未収集の項目についての質問を生成することを特徴とする請求項２又は請求項３に記載のレコメンド作成システム。 The said question control means produces | generates the question about the item which user information is not collected when the registration rate of the item in the center category and related category which comprises the said block matrix is less than predetermined value. Item 4. The recommendation creating system according to item 2 or item 3.
  5. 　前記レコメンド作成手段は、ユーザの履歴ログも用いて前記ユーザの問いかけに対するレコメンドを作成することを特徴とする請求項１から請求項４のいずれかに記載のレコメンド作成システム。 The recommendation creation system according to any one of claims 1 to 4, wherein the recommendation creation means creates a recommendation for the user's inquiry using a user's history log.
  6. 　前記データベースは、各カテゴリにおける各項目に関するユーザ情報を取得するための質問を格納していることを特徴とする請求項１から請求項５のいずれかに記載のレコメンド作成システム。 The recommendation creation system according to any one of claims 1 to 5, wherein the database stores a question for obtaining user information regarding each item in each category.
  7. 　前記データベースは、前記中心カテゴリ毎に設定されたブロックマトリクスを格納していることを特徴とする請求項１から請求項６のいずれかに記載のレコメンド作成システム。 The recommendation creation system according to any one of claims 1 to 6, wherein the database stores a block matrix set for each central category.
  8. 　請求項１から請求項７のいずれかに記載のレコメンド作成システムを搭載したことを特徴とする携帯端末装置。 A portable terminal device comprising the recommendation creating system according to any one of claims 1 to 7.
PCT/JP2013/052396 2013-02-01 2013-02-01 Recommendation creation system WO2014118975A1 (en)

Priority Applications (3)

Application Number Priority Date Filing Date Title
GB1421558.6A GB2517358A (en) 2013-02-01 2013-02-01 Recommendation creation system
PCT/JP2013/052396 WO2014118975A1 (en) 2013-02-01 2013-02-01 Recommendation creation system
JP2014510321A JP5782562B2 (en) 2013-02-01 2013-02-01 Recommendation creation system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2013/052396 WO2014118975A1 (en) 2013-02-01 2013-02-01 Recommendation creation system

Publications (1)

Publication Number Publication Date
WO2014118975A1 true WO2014118975A1 (en) 2014-08-07

Family

ID=51261717

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2013/052396 WO2014118975A1 (en) 2013-02-01 2013-02-01 Recommendation creation system

Country Status (3)

Country Link
JP (1) JP5782562B2 (en)
GB (1) GB2517358A (en)
WO (1) WO2014118975A1 (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2016038898A (en) * 2014-08-08 2016-03-22 バイドゥ オンライン ネットワーク テクノロジー （ベイジン） カンパニー リミテッド Search recommendation method and device
JP2018128747A (en) * 2017-02-06 2018-08-16 ソフトバンク株式会社 Data processing device, data processing method and program
US20190026813A1 (en) * 2017-07-18 2019-01-24 International Business Machines Corporation Elicit user demands for item recommendation
JP6482703B1 (en) * 2018-03-20 2019-03-13 ヤフー株式会社 Estimation apparatus, estimation method, and estimation program

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005044292A (en) * 2003-07-25 2005-02-17 Sony Corp Information processing system, information processor and method, recording medium, and program
JP2011065627A (en) * 2009-08-21 2011-03-31 Yahoo Japan Corp Information display device, display control method, and program

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005044292A (en) * 2003-07-25 2005-02-17 Sony Corp Information processing system, information processor and method, recording medium, and program
JP2011065627A (en) * 2009-08-21 2011-03-31 Yahoo Japan Corp Information display device, display control method, and program

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
RYOSUKE INOUE ET AL.: "Onsei Taiwa System ni Okeru Shitsumon Oto Database no Bunrui to sono Bunseki", PROCEEDINGS OF THE 18TH ANNUAL MEETING OF THE ASSOCIATION FOR NATURAL LANGUAGE PROCESSING TUTORIAL HONKAIGI (CD-ROM), 13 March 2012 (2012-03-13), pages 642 - 645 *
TERUHISA MISU ET AL.: "Speech-based Interactive Information Guidance System Using Question- Answering and Information Recommendation", TRANSACTIONS OF INFORMATION PROCESSING SOCIETY OF JAPAN, vol. 48, no. 12, 15 December 2007 (2007-12-15), pages 3602 - 3611 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2016038898A (en) * 2014-08-08 2016-03-22 バイドゥ オンライン ネットワーク テクノロジー （ベイジン） カンパニー リミテッド Search recommendation method and device
JP2018128747A (en) * 2017-02-06 2018-08-16 ソフトバンク株式会社 Data processing device, data processing method and program
US20190026813A1 (en) * 2017-07-18 2019-01-24 International Business Machines Corporation Elicit user demands for item recommendation
US10529001B2 (en) * 2017-07-18 2020-01-07 International Business Machines Corporation Elicit user demands for item recommendation
JP6482703B1 (en) * 2018-03-20 2019-03-13 ヤフー株式会社 Estimation apparatus, estimation method, and estimation program
JP2019164314A (en) * 2018-03-20 2019-09-26 ヤフー株式会社 Estimation device, estimation method, and estimation program

Also Published As

Publication number Publication date
GB2517358A (en) 2015-02-18
JPWO2014118975A1 (en) 2017-01-26
GB201421558D0 (en) 2015-01-21
JP5782562B2 (en) 2015-09-24

Similar Documents

Publication Publication Date Title
US11513666B2 (en) Matching process system and method
US10348586B2 (en) Parallel computatonal framework and application server for determining path connectivity
US8566327B2 (en) Matching process system and method
Koivumäki et al. The effects of information quality of mobile information services on user satisfaction and service acceptance–empirical evidence from Finland
JP5902274B2 (en) Ranking and selection entities based on calculated reputation or impact scores
Zhang et al. How the spread of user-generated contents (UGC) shapes international tourism distribution: Using agent-based modeling to inform strategic UGC marketing
US20170192994A1 (en) Systems and methods concerning tracking models for digital interactions
CN105930540A (en) Data processing system
US20140129551A1 (en) Collections of linked databases
KR101579376B1 (en) Personalized place recommendation system and method by using subjectivity analysis for user classification
US20150302306A1 (en) Systems and methods of processing personality information
JP5782562B2 (en) Recommendation creation system
KR102415084B1 (en) Insurance agent matching server and matching method using persona model
US20220076140A1 (en) Prioritization of electronic communications
JP6042188B2 (en) Group extraction apparatus, group extraction method and program
US20190333176A1 (en) Recording recommendation method, device, apparatus and computer-readable storage medium
AU2008200194A1 (en) Street Quality Assessment Method and System
JP2012078768A (en) Person matching device, method and program
JP5373670B2 (en) Question recommendation device, method and program
Weng et al. Lining up for quick service—The business impact of express lines on fast-food restaurant operations
US12105941B2 (en) Matching process system and method
JP7082008B2 (en) Information processing equipment, information processing methods, and information processing programs
KR101480714B1 (en) A system, an apparatus, a method and a storage means for searching based on topic and intention
JP6901139B2 (en) Lifelog image generator, lifelog image generation method, and program
Bolinder et al. Exploring the customer journey: An exploratory study investigating the customer journey

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2014510321

Country of ref document: JP

Kind code of ref document: A

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 13873467

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 1421558

Country of ref document: GB

Kind code of ref document: A

Free format text: PCT FILING DATE = 20130201

WWE Wipo information: entry into national phase

Ref document number: 1421558.6

Country of ref document: GB

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 13873467

Country of ref document: EP

Kind code of ref document: A1

ENPC Correction to former announcement of entry into national phase, pct application did not enter into the national phase

Ref country code: GB