JP2013250672A - Interest analysis method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately present information desired by a user at that point of time.SOLUTION: A feature score calculation part 120 acquires a first condition list in which a plurality of concepts as the selection candidates of narrowing-down conditions in a concept system are browsed as a list and a second condition list including the concept selected by a user from this, and searches analysis parameters S', N', a' and n' from the first condition list and the second condition list, and calculates first probability that the number of the concepts which appear in the second condition list becomes n' or more and second probability that the number of the concepts which appear in the second condition list becomes n' or less under the conditions of the analysis parameters S', N' and a', and calculates feature scores from the inverse functions of the cumulative distribution functions of standard normal distribution on the basis of the first probability and the second probability. A concept system update processing part 130 updates user interest scores with respect to the concept by using the calculated feature scores.

Description

  The present invention relates to an interest analysis method used in an information recommendation service in which an information provider selects and presents information suitable for each information user on a computer network.

  Meta information that summarizes the content is added to each content as a method of displaying information by selecting information suitable for each information user or rearranging in a suitable order for each information user. As a premise, a method for estimating the user's interest from the frequency of the concept or the like appearing in the user history is a content-based filtering method (Content Based Filtering: CBF), and research is being conducted particularly as a memory-based method.

  Specifically, the content-based filtering technique presents products of the same brand (holding the same concept tag) when browsing products of a specific brand (holding information indicating the brand as a concept tag). The memory-based method in this case presents a product of a specific brand if the specific brand is frequently browsed from the history of browsing in the past. In a simple method, products related to concept tags that appear more frequently in the browsing history are presented. As a method using relevance feedback and statistical hypothesis testing, for example, Non-Patent Document 1 “Proposal of News Recommendation Method Using Relevance Feedback and Statistical Hypothesis Test” is available.

“Proposal of News Recommendation Method Using Relevance Feedback and Statistical Hypothesis Testing”, Type 2 Study Group Material, WI2-2012-19, pp.53-58, 2012

  However, even if a large amount of detailed history is obtained in the above-described conventional technique, the user's preference may change depending on the mood at that time. Information that the user does not want at that time may be presented.

  The present invention has been made paying attention to the above circumstances, and an object thereof is to provide an interest analysis method for accurately presenting information desired by a user at that time.

  A first aspect of the present invention is a method for analyzing a user's interest using a concept system in which user interest scores for a plurality of concepts are systematized by a computer, and a plurality of candidates for selection of narrowing conditions in the concept system. A first condition list obtained by browsing the concepts as a list, a second condition list including the concept selected by the user from the first condition list, and included in the first condition list The total number of concepts to be generated is the first total number, the number of occurrences of the concept selected by the user in the first condition list is the first appearance number, and the total number of concepts included in the second condition list is When the second total number and the number of occurrences of the concept selected by the user in the second condition list are the second number of appearances, the first total number and the first number of appearances And under the condition of the second total number, the number of occurrences of the concept in the second condition list is a first probability that the number of occurrences is equal to or greater than the second number of occurrences, and a number that is equal to or less than the second number of occurrences. A calculation step of calculating a feature score by an inverse function of a cumulative distribution function of a standard normal distribution based on the first probability and the second probability, and the concept using the feature score And updating the user interest score for.

  According to the first aspect, when the user selects a narrowing-down condition from the selection candidates in the first condition list, the specification of the narrowing-down condition is comparatively “selected” or “not selected” compared to chance. The feature score is obtained, and the user's interest is estimated using the feature score. Even if the user designates a narrowing condition based on the use of the feature score together with the interest estimation based on the content selection history, The user's interests can be analyzed reasonably and accurately.

  A second aspect of the present invention is a method for analyzing user interests using a concept system in which user interest scores for a plurality of concepts are systematized by a computer, and a plurality of candidates for selection of exclusion conditions in the concept system. A first condition list obtained by browsing the concepts as a list, a second condition list selected by the user from the first condition list, and a concept included in the first condition list The total number is the first total number, the number of occurrences of the concept selected by the user in the first condition list is the first appearance number, and the total number of concepts included in the second condition list is the second number. When the total number and the number of appearances of the concept selected by the user in the second condition list are defined as the second number of appearances, the first total number, the first number of appearances, and the second number A first probability that the number of occurrences of the concept in the second condition list is greater than or equal to the second occurrence number and a second probability that is less than or equal to the second occurrence number under the condition of the total number A calculation step of calculating a feature score by an inverse function of a cumulative distribution function of a standard normal distribution based on the first probability and the second probability, and the user interest score for the concept using the feature score And an update step for updating.

  According to the second aspect, when the user selects an exclusion condition from the selection candidates in the first condition list, the user's interest is utilized by utilizing the feature of “select” or “not select” relatively compared to chance. The user's interest can be analyzed reasonably and accurately even when the user designates the narrowing-down condition by using it together with the interest estimation based on the content selection history.

  According to a third aspect of the present invention, in the first or second aspect, the calculating step specifies a selection path in the concept system based on a concept included in the second condition list, and the selection The feature score is calculated for each concept corresponding to each selection occurring in the route.

  According to the third aspect, it is possible to increase the accuracy of interest estimation by specifying the selection path of the narrowing-down condition or exclusion condition from the superordinate concept in the concept system and using each of those selections for the interest estimation. Become.

According to a fourth aspect of the present invention, in any one of the first to third aspects, the updating step updates a user interest score of a subordinate concept of the concept using the feature score.
According to the fourth aspect, by updating the user interest score of the subordinate concept of the concept using the feature score, not only the selected concept but also the subordinate concept can be analyzed reasonably and accurately. .

According to a fifth aspect of the present invention, in any one of the first to fourth aspects described above, for the content in which one or more concepts appear, the user interest score of each concept that appears in the content is used. The method further includes an evaluation step of calculating a user evaluation score for the content.
According to the fifth aspect, it is possible to recommend content that matches the user's interest by calculating the user's evaluation score for the content using the calculated user interest score.

  That is, according to the present invention, it is possible to provide an interest analysis method for accurately presenting information desired by a user at that time.

1 is an overall configuration diagram of an information recommendation system according to the present invention. The figure which shows the subject of the information recommendation using only an information selection log | history. The figure which shows the concept of the information recommendation using information selection history and explicit condition designation. The block diagram which shows the function structure of each apparatus of the information recommendation system shown in FIG. The figure which shows an example of content request data. The figure which shows an example of content browsing operation on a client terminal. The figure which shows the data structural example of a list browsing content list. The figure which shows the data structural example of detailed browsing content. The figure which shows an example of condition list request data. The figure which shows the collection method of condition selection log | history. The figure which shows the data structural example of a presentation content list. The figure which shows the process outline in the case of estimating a user's interest from browsing history. The figure which shows the condition setting method using a conceptual system. The figure which shows the specific method of the target content at the time of condition setting. The figure which shows an example of a content database. The figure which shows an example of a concept system / user interest score database. The figure which shows the processing flow of a log | history information receiving part. The figure which shows the processing flow of a characteristic score calculation part. The figure which shows the data structural example of an analysis parameter list. The schematic diagram for demonstrating operation | movement of a characteristic score calculation part. The figure which shows the detail of a characteristic score calculation process. The figure which shows the interest estimation method using the scarcity at the time of narrowing-down condition selection. The figure which shows the identification method of a narrowing-down condition selection path | route. The figure which shows the selection list | wrist performed when narrowing-down conditions were designated. The figure which shows the calculation result of the characteristic score at the time of narrowing-down condition selection. The figure which shows the interest estimation method using the rarity at the time of exclusion condition selection. The figure which shows the identification method of an exclusion condition selection path | route. The figure which shows the selection list | wrist performed when exclusion conditions were designated. The figure which shows the calculation result of the characteristic score at the time of exclusion condition selection. The figure which shows the calculation method of the characteristic score at the time of exclusion condition selection. The figure which shows the processing flow of a concept system update process part. The figure which shows the detail of a concept system update process. The figure which shows the propagation process of the interest score using a concept system. The figure which shows the characteristic score which carried out the probability coupling | bonding of all the selections regarding a narrowing-down condition. The figure which shows the characteristic score when not selecting the selection path | route from a high-order concept. The figure which shows the characteristic score which stochastically combined all the selections regarding exclusion conditions. The figure which shows the characteristic score when not selecting the selection path | route from a high-order concept. The figure which shows the characteristic score which carried out the probability coupling | bonding of all the selections regarding a narrowing-down condition. The figure which shows the characteristic score when not selecting the selection path | route from a high-order concept. The figure which shows the processing flow of a content evaluation process part. The figure which shows an example of a content score list. The figure which shows the detail of a content evaluation process. The figure which shows the subject regarding the interest learning at the time of condition setting.

Embodiments according to the present invention will be described below with reference to the drawings.
FIG. 1 is an overall configuration diagram of an information recommendation system according to the present invention. This system includes a client terminal 200, a content server 300, and an interest analysis device 100. The client terminal 200 and the content server 300, and the content server 300 and the interest analysis device 100 are connected to each other via a communication network. The user can acquire desired content from the content server 300 by browsing operation and specifying a condition on the client terminal 200, and can display the acquired content on the screen of the client terminal 200 for browsing.

  The client terminal 200 collects a content browsing history by a user operation, browses a plurality of content as a list, a list browsing content list (first content list), and a detailed browsing content list that browses the content body from the content list (Second content list) is transmitted to the content server 300. The content server 300 transfers the list browsing content list and the detailed browsing content list to the interest analysis device 100 via the communication network.

  Furthermore, the client terminal 200 collects a narrowing condition or exclusion condition designation history by a user operation, and selects a condition from the list browsing condition list (first condition list) browsed as an option when the user designates the condition and the list browsing condition list. The selected condition selection list (second condition list) is transmitted to the content server 300. The content server 300 transfers the list browsing condition list and the condition selection list to the interest analysis apparatus 100 via the communication network.

  The interest analysis apparatus 100 calculates the feature score for each concept appearing in the content and updates the user interest score based on the list browsing content list and the detailed browsing content list, and estimates the user's interest. If there is a condition designation, the interest analysis device 100 calculates the feature score for each concept selected as the condition and updates the user interest score based on the list browsing condition list and the condition selection list, and the user interest score is updated. Estimate the interest.

  The interest analysis device 100 generates a list of sorted contents (sorted presented content list) from the “presented content list” received from the content server 300 based on the user interest score, according to the user's interest. To the content server 300.

  For example, in the example shown in FIG. 2, when a user who frequently selects a ramen shop has a desire to drink alcohol only on that day, the information providing side uses the user's daily information selection history. I would recommend a ramen shop because I would presume that I am interested in ramen. This recommendation result does not match the user's interest to “drink”. That is, there is a possibility that the information desired by the user (in this example, a bar) cannot be presented.

  Therefore, in the present invention, not only the information selection history of the user but also the explicit condition specification indicated by the user is taken into consideration so that the information desired by the user can be presented with high accuracy. Further, the history of condition designation explicitly performed by the user is also a useful history for estimating the user's interest. Therefore, the recommendation accuracy is improved by combining the interest estimation using the user browsing history and the interest estimation using the condition designation history performed by the user.

  For example, as shown in FIG. 3, the information providing side “has a high ratio of selecting“ ramen ”among“ all genres ”” and “a ratio of selecting“ beer ”among“ alcohol ”for a specific user. Is “high”, “among“ beer ”,“ the rate of selecting “ale” is high ”,” and so on. "I like" beer "" and "I like" ale "in" beer "". As a result, in consideration of the explicit condition specification of “I want to drink alcohol” by the user, the recommendation result is limited to only “alcohol”, and “beer” in “alcohol” is especially “beer” Since “Ale” can be recommended in particular, information desired by the user (in this example, “beer”, in particular, a bar where “Ale” is provided) can be presented.

  FIG. 4 is a block diagram showing a functional configuration of each device of the information recommendation system shown in FIG. Each unit in FIG. 4 can be realized by, for example, a CPU (Central Processing Unit) of each device and a control program executed on a memory. Details of each device will be described below.

[Client terminal]
4, the client terminal 200 includes a history collection unit 210, a history information transmission unit 220, a content presentation unit 230, a content request transmission unit 240, a list browsing condition list request transmission unit 250, a list browsing condition list presentation unit 260, a browsing. A condition selection history collection unit 270 and a condition selection history transmission unit 280 are provided.

  The content request transmission unit 240 makes a content presentation request to the content server 300 in accordance with a user instruction (input). Specifically, content request data as shown in FIG. 5 is transmitted to the content server 300. For example, the content request data has a user ID (or client terminal ID) and a request time. The request time may be added in the content server 300. The user ID (or client terminal ID) is a number or character string that is uniquely assigned to each terminal (or user), and is a user ID in a user interest score table of the conceptual system / user interest score database 140 described later. The matching ID.

  Based on the sorted presentation content list received from the content server 300, the content presentation unit 230 displays a list from the top of the sort order within the range allowed by the display screen size of the client terminal 200. FIG. 6 shows an example of a content browsing operation by the user on the client terminal 200.

  In the example of FIG. 6, ten contents (contents 1 to 10) are displayed in a list. The content in the lower order of the sort order can be displayed by the list by the user's flick, scroll bar operation or the like. The list of contents actually displayed on the client terminal 200 in this way is referred to as a list browsing content list. That is, not all the contents in the sorted presentation content list are displayed on the client terminal 200, and thus are not necessarily included in the list browsing content list. When the user selects a title of each content from this list by clicking or the like, the main body (details) of the content of the selected title (contents 3, 5, and 6 in FIG. 6) can be viewed. The content whose details are browsed is included in the detailed browsing content list.

  As described above, the history collection unit 210 collects user operation histories and creates a list browsing content list and a detailed browsing content list. The history information transmission unit 220 transmits the list browsing content list and the detailed browsing content list created by the history collection unit 210 to the content server 300 together with the user ID (or client terminal ID).

  FIG. 7 shows a data configuration example of the list browsing content list in the case of FIG. The list browsing content list has a cluster ID, a content ID, and a browsing time. The cluster is an identifier ("1" in FIG. 7) uniquely assigned to the list browsing content list and the detailed browsing content list. When the user browses the list browsing content displayed at another time (time zone), another cluster ID is given. The conditions for newly assigning the cluster ID under conditions other than the time include when there is no operation for a certain period of time while the list browsing content list is displayed, or when the browsing user (user ID) is switched, the list browsing content list On the other hand, when the content genre or the like is narrowed down by specifying narrowing conditions or exclusion conditions, the browsing mode may be switched in other browsing applications. The content ID is an identifier uniquely assigned to each content of the list browsing content, and is assumed to match a value held in a content database 160 described later.

  FIG. 8 shows an example of the data structure of the detailed browsing content list in the case of FIG. The detailed browsing content list has a cluster ID, a content ID, and a browsing time, like the list browsing content list. The cluster ID is the same value as the list browsing content list (“1” in FIG. 8). In the detailed browsing content list, the content ID and browsing time are the identifier of the content (contents 3, 5, and 6 in FIG. 8) that the user has selected from the browsing content and browsed the content and the time when the content was browsed.

  The list browsing condition list request transmission unit 250 requests the content server 300 for a list browsing condition list in accordance with a user instruction (input). Specifically, condition list request data as shown in FIG. 9 is transmitted to the content server 300. For example, the condition list request data has a type, a parameter name, and a parameter value. As shown in FIG. 9A, the condition list request data when a condition is specified from a keyword search result has a type (= search), a parameter name (= keyword), and a parameter value (= sport). As shown in FIG. 9B, the condition list request data when a condition is specified from the list has a type (= list), a parameter name (= display rule), and a parameter value (= ascending order). Further, as shown in FIG. 9C, the condition list request data when the condition is specified from the concept system has a type (= concept), a parameter name (= parent concept), and a parameter value (= sport). . The list browsing list request data is sent to the list browsing condition list creation unit 185 by the list browsing condition list request transfer unit 370.

  The list browsing condition list presenting unit 260 displays a list of presentation list browsing condition lists sent from the interest analysis apparatus 100 via the content server 300 within a range that the display screen size of the client terminal 200 allows. FIG. 10 shows an example of a condition browsing operation by the user on the client terminal 200. FIG. 10A shows a case where conditions are designated from the keyword search result, and FIG. 10B shows a case where conditions are designated from the list. The user selects a narrowing condition or an exclusion condition from the presented list browsing condition list. In addition, designation | designated of selecting a narrowing-down condition or an exclusion condition distinguishes by changing the button at the time of selection, for example. When a condition is specified from a parent concept, the condition is specified from a list browsing condition list obtained by extracting a direct subordinate concept of the parent concept as described later.

  The browsing condition selection history collection unit 270 collects the presented list browsing condition list (first condition list) and a condition selection list (second condition list) including one or more conditions selected from the list of conditions. To do. If there are many choices in the presentation list browsing condition list sent from the interest analysis device 100, all options may not be displayed on the screen of the client terminal 200 as a list browsing condition list. In such a case, only what is displayed on the screen can be regarded as a list browsing condition list.

When the condition has a hierarchical structure, a plurality of selections may occur depending on the selected route with respect to setting the condition. In that case, a condition selection list is transmitted for each selection that occurs. Alternatively, as shown in FIG. 10, the condition selection list may be selected from a list browsing condition list arranged according to the result of keyword search or alphabetical order without the hierarchical structure of the condition. In such a case, a single condition setting list is transmitted.
The condition selection history transmission unit 280 transmits the condition selection history collected by the browsing condition selection history collection unit 270 to the interest analysis device 100 via the content server 300.

[Content Server]
4, the content server 300 includes a content transmission processing unit 310, a sorted presentation content list reception unit 320, a presentation content list transmission unit 330, a presentation content list input unit 340, a history information transfer unit 350, and a content request transfer unit 360. A list browsing condition list request transfer unit 370, a list browsing condition list transfer unit 380, and a condition selection history transfer unit 390.

The history information transfer unit 350 transfers the list browsing content list and the detailed browsing content list received from the client terminal 200 to the interest analysis device 100 via the communication network.
The presentation content list input unit 340 receives a presentation content list that lists contents to be presented to the client terminal 200 used by the user by the service operator. The presented content list transmission unit 330 transmits the input presented content list to the interest analysis apparatus 100 via the communication network.

  FIG. 11 shows a data configuration example of the presented content list. The presented content list has a content ID, a concept ID / relevance list, a content body, and a content registration time. The content ID is a unique ID assigned by the content server 300 to each content. The concept ID / relationship degree list stores a concept ID of a concept that appears in the content and a set of values indicating the degree of relevance between the concept and the content. The concept ID / relevance degree list is set in advance for each content. As a specific example, content 1 (sports article) includes {“baseball” concept ID = 1, relevance = 0.5}, { A set of concept ID and degree of association is stored as “soccer” concept ID = 2, degree of association = 0.8}, {“golf” concept ID = 3, degree of association = 0.6}. .

  The concept ID matches the value stored in the concept system / user interest score database 140. For example, the relevance is a value from 0 to 1, and the larger the relevance, the stronger the relevance. As the relevance, a value set by the service operator at the time of content registration or a value calculated by another system is used.

  The sorted presentation content list receiving unit 320 receives a sorted presentation content list and a user ID (or client terminal ID) obtained by sorting a part or all of the presentation content list from the interest analysis device 100. The content transmission processing unit 310 transmits the sorted presentation content list to the client terminal 200 corresponding to the user ID (or client terminal ID).

The content request transfer unit 360 transfers the content request data (FIG. 5) transmitted from the content request transmission unit 240 of the client terminal 200 to the interest analysis device 100.
Further, the list browsing condition list request transfer unit 370 transfers the condition list request data (FIG. 9) transmitted from the list browsing condition list request transmission unit 250 of the client terminal 200 to the interest analysis device 100.

The list browsing condition list transfer unit 380 transfers the list browsing condition list for presentation created by the list browsing condition list creation unit 185 of the interest analysis apparatus 100 described later to the client terminal 200.
The condition selection history transfer unit 390 transfers the condition selection history sent from the condition selection history transmission unit 280 of the client terminal 200 to the interest analysis device 100.

[Interest analysis device]
The interest analysis apparatus 100 includes a history information reception unit 110, a feature score calculation unit 120, a concept system update processing unit 130, a concept system / user interest score database 140, a presented content list reception unit 150, a content database 160, and a content evaluation processing unit 170. The sorted content score list transmission unit 180, the list browsing condition list creation unit 185, and the condition selection history reception unit 190 are provided.

  FIG. 12 shows an outline of processing when estimating the user's interest from the content browsing history. The history information receiving unit 110 receives the list browsing content list and the detailed browsing content list from the client terminal 200 via the content server 300. The list browsing content list is, for example, a list of content in which the user browses only the content titles in a list. The detailed browsing content list is a list of content that the user has viewed the content (details) of the content body. For example, in FIG. 12, the list browsing content list includes contents 1 to 8, and the detailed browsing content list includes contents 1, 3, and 4. In FIG. 12, the content indicated by the hatched pattern indicates that the concept B appears in the contents 1, 6, 7, and 8.

The feature score calculation unit 120 calculates a feature score (Z value to be described later) of each concept by a statistical model of concept selection using the list browsing content list and the detailed browsing content list.
The concept system update processing unit 130 updates the user interest score for each concept based on the relationship information (superordinate concept and subordinate concept) in the concept system using the feature score. Nodes included in the graph of the concept system represent concepts, and links represent relationships between concepts. The user interest score is held as a value of a node corresponding to each concept in the concept system. In the concept system, the nodes located at the higher level represent the abstract concept, and the nodes located at the lower level represent the specific concept. The concept system and concept ID (identifier assigned to each node) are designed and defined in advance by a service operator or the like.

  The content evaluation processing unit 170 calculates the user's evaluation score for the content by probability combining using the user interest score of each concept appearing in the evaluation content. In the example of FIG. 12, the evaluation score of the evaluation content 1 is obtained using the user interest scores of the concepts E, F, and D that appear in the content 1.

  Furthermore, when conditions are specified, for example, in FIG. 13, specifying “sake” as a narrowing condition or specifying “wine” as a narrowing condition specifies the narrowing condition that is the user's explicit interest. For example, when “alcohol” is designated as a refinement condition in FIG. 14, the concept is regarded as a refinement condition, and contents B, C, D, E having at least one link are determined based on the refinement condition. Is the content to be evaluated. At this time, feature scores other than the narrowing-down conditions are also used for evaluation. For example, in FIG. 14, the characteristic score of “ramen” is used for the evaluation of the content B.

  Further, for example, in FIG. 13, by selecting “whiskey” as the exclusion condition from the candidate exclusion condition list (“wine”, “beer”, “whiskey”), the exclusion condition can be specified. In FIG. 14, when “whiskey” is designated as an exclusion condition, the concept is regarded as an exclusion condition, and the link from the narrowing-down condition is not used for evaluation. For example, in FIG. 14, a link from whiskey to content E, which is an exclusion condition, is not used for evaluation, and content E is excluded from the evaluation target because it does not have any other links from the refinement concept. Alternatively, content having a link from the exclusion condition may be excluded from the evaluation target. In this case, the content D is also excluded from the evaluation target.

Hereinafter, the detail of each part of the interest analysis apparatus 100 is demonstrated.
(Content database 160)
FIG. 15 shows an example of the data structure of the content database 160. The content database 160 has a content table and a user history table. The content table has a content ID, a concept ID / relevance list, a content body, and a content registration time, and stores each value received by the presented content list receiving unit 150.

  The user history table stores content ID, user ID (client terminal ID), detailed browsing total number, detailed browsing time, list browsing total number, list browsing time, and list non-display flag. The detailed browsing time is null when the total number of detailed browsing is 0, and if it is 1 or more, a list of browsing times in a time series of each browsing is stored. The list browsing time is null when the total number of browsing the list is 0, and if it is 1 or more, a list of browsing times according to the time series of each browsing is stored. The list non-display flag stores false when the user has not yet displayed / viewed as a list on the screen of the client terminal, and stores true when the list has been viewed even once. In the user history table, all content ID values are held for each user ID. As the detailed browsing total number and the list browsing total number, when a large number of list browsing content lists indicated by the cluster ID are received, a cumulative total of past histories is stored.

  For example, by using the data of the user history table, the evaluation score of the content is lowered in the future evaluation (during processing in the content evaluation processing unit 170) according to the number of times the user browses. As a method for reducing the evaluation score, when the number of browsing for a certain content is k, the evaluation score of the content is divided by k + 1, or the evaluation score is multiplied by a weight (for example, 0.9) to the kth power. By this process, it is possible to reduce the repeated presentation of the same content in consideration of the degree of coincidence with the interest, so that the satisfaction with the user's recommendation can be improved.

(Conceptual system / user interest score 140)
FIG. 16 shows an example of the data structure of the conceptual system / user interest score database 140. The concept system / user interest score 140 includes a root concept node ID, a concept system table, and a user interest score table. The root concept node ID is a concept node ID at the highest level in the concept system structure. There is only one in the system. The concept system table stores a self-concept ID, a parent concept ID list, and a child concept ID list. All the self-concept IDs in the concept system are stored in association with a parent concept ID and a child concept ID (however, if the self-concept is the lowest, there is no child concept ID). Defined. The user interest score table stores values of concept ID, user ID (or client terminal ID), TotalZ (user interest score), X, and Y. The definition and calculation method of TotalZ, X, and Y will be described later.

(Presentation content list receiving unit 150)
The presented content list receiving unit 150 receives the presented content list as shown in FIG. 11 from the content server 300 and stores it in the content database 160.
(History information receiving unit 110)
FIG. 17 shows a processing flow of the history information receiving unit 110. In step S11, the history information receiving unit 110 receives the user ID (or client terminal ID), the list browsing content list, and the detailed browsing content list transmitted from the client terminal 200 via the content server 300, and calculates the characteristic score. To the unit 120.

(Condition selection history receiving unit 190)
The condition selection history reception unit 190 receives the list browsing condition list and the condition selection list transmitted from the condition selection history transmission unit 280 of the client terminal 200 via the content server 300 and outputs them to the feature score calculation unit 120. Further, the condition selection history receiving unit 190 outputs the condition selection list to the list browsing condition list creating unit 185 and the content evaluation processing unit 170.

(List browsing condition list creation unit 185)
The list browsing condition list creation unit 185 creates a list browsing condition list for presentation based on the list browsing condition request data sent from the list browsing condition list request transfer unit 370 of the content server 300. For example, when a conditional search keyword as shown in FIG. 9A is sent, a presentation list browsing condition list indicating a search result that matches the keyword is created. When a condition list request as shown in FIG. 9B is sent, a presentation list browsing condition list is created by rearranging the condition list in, for example, name order. When a parent concept as shown in FIG. 9C is sent, a presentation list browsing condition list obtained by extracting a list of direct subordinate concepts of the parent concept with reference to the concept system / user interest score database 140 Create In addition, when the condition selection list is input from the condition selection history receiving unit 190, the list browsing condition list creation unit 185 refers to the concept system / user interest score database 140 and corresponds to the condition specified in the condition selection list. A list browsing condition list for presentation in which subordinate concepts to be extracted are extracted is created. The created list browsing condition list for presentation is transmitted to the content server 300 via the communication network.

(Feature score calculator 120)
The feature score calculation unit 120 uses the list browsing content list and the detailed browsing content list to calculate a feature score (Z value to be described later) of each concept using a statistical model of concept selection. In addition, the feature score calculation unit 120 calculates a feature score (Z ′ value described later) for each concept selected as a condition using the list browsing condition list and the condition selection list input from the condition selection history reception unit 190. To do.

(Feature score calculation processing from content browsing history)
FIG. 18 shows a processing flow of the feature score calculation unit 120. The feature score calculation unit 120 receives the user ID (or client terminal ID), the list browsing content list, and the detailed browsing content list from the history information receiving unit 110.

  In step S <b> 12, the feature score calculation unit 120 extracts a concept ID that appears in each content in the detailed browsing content list from the content database 160. Specifically, the “concept ID” associated with each content ID in the detailed browsing content list of FIG. 8 is searched from the content table of the content database 160 of FIG. The feature score calculation unit 120 includes cluster data {cluster ID, list browsing content list, detailed browsing content list} and content ID / concept ID association list {{content ID, {related concept ID, ...}}, ... } And the appearance concept list {concept ID}. The “content ID / concept ID association list” is a list of concept IDs searched based on the content ID. The “appearance concept list” is a list of all concept IDs of concepts that appear in each content included in the list browsing content list and the detailed browsing content list.

  In step S13, the feature score calculation unit 120 extracts, for each concept ID in the “appearance concept list”, a superordinate concept from the concept system / user interest score database 140 of FIG. And “content ID / concept ID association list”.

  Specifically, the feature score calculation unit 120 searches the concept system table in FIG. 16 for the concept ID of the “appearance concept list” that matches the “own concept ID”, and extracts the “parent concept ID”. To do. Further, the “parent concept ID” that is extracted matches the “self concept ID” in the concept system table of FIG. 16, and the “parent concept ID” is also extracted as a superordinate concept, and this parent concept ID list The process of extracting the parent concept list with reference to the concept system table with respect to the concept ID is repeated. Then, the feature score calculation unit 120 associates the concept ID of the superordinate concept with the content ID having the concept ID of the appearance concept that is the source of extraction. That is, the extracted “superior concept” is regarded as having been assigned a superordinate concept with respect to “content ID having the original concept ID”, and “appearance concept list” “content ID / concept ID association” Add to list. The extraction of the root concept in the concept system hierarchy is excluded.

In step S14, the feature score calculation unit 120 calculates the number of appearances for each concept in the “appearance concept list”, extracts analysis parameters necessary for calculating the feature score, and generates an analysis parameter list.
FIG. 19 shows a data configuration example of the analysis parameter list. The analysis parameter list includes, for each cluster ID, the total content S (first total) of the list browsing content list, the total content a (second total) of the detailed browsing content list, and the appearance concept list associated with the cluster ID. N and n are calculated for each concept ID. N (first appearance number) is the number of contents to which the concept ID is assigned in the list browsing content list. n (second appearance number) is the number of contents to which the concept ID is assigned in the detailed browsing content list. Note that N and n are calculated for all concept IDs in the appearance concept list including the superordinate concept added in step S13.

  FIG. 20A shows a schematic diagram of the analysis parameter extraction process. For example, the case where the user browses the details of 10 (= a) contents from a list of 50 (= S) contents is shown. Here, 15 articles (= N) containing the concept of “baseball” among the 50 contents displayed in a list are displayed, and the concept of “baseball” is included in 10 contents viewed by the user. This indicates that there are five (= n) contents including “”.

  In step S15, the feature score calculation unit 120 calculates a feature score Z for each concept ID using the analysis parameters S, a, N, and n. FIG. 21 shows details of the feature score calculation process. In FIG. 21, i indicates a concept identifier, and j indicates a cluster ID. When H1 (first probability) is the total number S of the list browsing contents included in the list browsing content list and the number N of the contents of the list browsing content where the concept i appears, a detailed browsing content is randomly selected and viewed. In this case, the cumulative probability that the number of detailed browsing contents in which the concept i appears is n or more. When H2 (second probability) is the total number S of the list browsing contents included in the list browsing content list and the number of contents N in which the concept i appears in the list browsing contents, a detailed browsing content is randomly selected and viewed. In this case, the cumulative probability that the number of detailed browsing contents in which the concept i appears is n or less. In the present embodiment, the cumulative probabilities H1 and H2 are obtained by the hypergeometric distribution, but are not limited to this method. Examples of other distributions include a binomial distribution and a normal distribution.

  As shown in FIG. 20B, for example, there are 5 or more contents including the concept of “baseball” among the 10 contents viewed by the user using the analysis parameters S, N, a, and n described above. It is shown that the probability of being “0.12”. Here, “0.12” corresponds to the value of the cumulative probability H1.

As an example of using the value of H2, consider the case where n is 0 in the above analysis parameters. In this case, the probability when the number of appearances is 0 or less is calculated. Specifically, in FIG. 20B, since the horizontal axis is the value of the item of 0, “0.02”.
Then, as shown in FIG. 21, the feature score calculation unit 120 uses the calculated cumulative probabilities H1 and H2 to calculate the feature score Z by the inverse function of the standard normal distribution cumulative distribution function. As shown in FIG. 20C, the feature score Z is obtained by the inverse function of the cumulative distribution function of the standard normal distribution with H1 as the cumulative probability. When H2 is used as the cumulative probability, the feature score Z is obtained with the sign of the return value of the inverse function of the standard normal distribution cumulative distribution function being negative. Using this feature score Z, a concept system update processing unit 130, which will be described later, obtains a user interest score (TotalZ) for the concept of “baseball”.

  The feature score calculation unit 120 generates an update target concept list and outputs it to the concept system update processing unit 130. The “update target concept list” is a set of a concept ID, the characteristic score Z calculated above, and a weight w. The concept ID appearing in the update target concept list becomes a node (concept) to be updated in the next concept system update process. The feature score Z and the weight w are calculated for all the concept IDs in the appearance concept list to which the superordinate concept is added. The weight w is a value set for each concept in each cluster ID.

  Note that the weight w can be changed as follows when the initial value w = 1 and a user's characteristic operation is performed. For example, with respect to the content presented to the user at the client terminal 200, the user can register as a favorite content, recommend to other users, or input an evaluation for the content. When the client terminal 200 can transmit an operation history other than the browsing operation to the interest analysis apparatus 100, the following processing is performed.

  For example, when the content is registered as a favorite, the feature score calculation unit 120 increases the weight w such that w = 1.5 for all the concept IDs included in the content. In addition to the content browsing time, browsing weather, temperature, humidity, season, day of the week, holiday, leisure time, user location information at browsing, scheduler, diary, etc., the value of weight w Can also be changed.

(Feature score calculation processing for each concept selected as a condition)
In addition, the feature score calculation unit 120 calculates a feature score (Z ′ value) for each concept selected as a condition using the list browsing condition list and the condition selection list input from the condition selection history reception unit 190.

(When filtering conditions are specified)
For example, in the conceptual system of FIG. 22, when beer or red wine is specified as a narrowing condition, “red wine” is narrowed down from the choices under “wine” (“red wine”, “white wine”, “rose”). How rare it is to select as a condition, or “beer” from the choices under “Western sake” (“wine”, “beer”, “whiskey”) The user's interest score can be estimated using the degree to which the selection is unusual compared to chance.

  As shown in FIG. 23, when the narrowing selection result is “red wine”, the selection route is specified as “alcohol” → “western sake” → “wine” → “red wine”. When the narrowing selection result is “beer”, the selection route is specified as “alcohol” → “western sake” → “beer”. The feature score calculation unit 120 extracts all the selections that have occurred according to the selected selection path as shown in FIG. As shown in FIG. 24, seven selection histories are obtained. Selection is made from “alcohol” → “western sake” → “wine” → “red wine” 1-4, “alcohol” → “western sake” → “beer” 5-7 are extracted. In FIG. 24, selection 1 indicates that “alcohol” is selected as a selection result from selection candidates “alcohol” and “noodles”.

  As shown in FIG. 25, the feature score calculation unit 120 extracts the analysis parameters N ′, S ′, n ′, and a ′ for each of the selections 1 to 7, and determines the rarity of the concept appearance when selecting the narrowing conditions. The used feature score Z ′ is calculated. For example, in the selection 1 of FIG. 24, the selection candidates “alcohol” and “noodles” (S ′ = 2) are presented as the list browsing condition list, and the user selects “alcohol” (a ′ = 1). Is selected as the condition selection list. Here, “alcohol” (N ′ = 1) is included in the list browsing condition list, and “alcohol” (n ′ = 1) is included in the condition selection list. The characteristic score Z ′ can be calculated by replacing N, S, n, a with N ′, S ′, n ′, a ′ in the calculation formula shown in FIG.

(When an exclusion condition is specified)
For example, in the conceptual system of FIG. 26, when whiskey is specified as an exclusion condition, “whiskey” is selected as an exclusion condition from the options under “Western sake” (“wine”, “beer”, “whiskey”). The user's interest score can be estimated using the degree to which things are unusual compared to chance.

  As shown in FIG. 27, when the exclusion selection result is “whiskey”, the selection path of the exclusion condition is specified as “alcohol” → “western sake” → “whiskey”. The feature score calculation unit 120 extracts all selections that have occurred when selecting a lower concept from a higher concept according to the selected route specified as shown in FIG. As shown in FIG. 28, three exclusion condition selection histories (selections 8 to 10) are obtained.

  The feature score calculation unit 120 extracts the analysis parameters N ′, S ′, n ′, and a ′ for each of the selections 8 to 10 as shown in FIG. 29, and uses the rarity of the concept appearance when the exclusion condition is selected. The feature score Z ′ obtained is calculated. However, when the given history is the exclusion condition selection history, the feature score Z ′ is calculated using the formula shown in FIG. That is, the feature score calculated using the content selection history and the feature score calculated using the condition selection history can be integrated into the same interest score.

(Concept system update processing unit 130)
FIG. 31 shows a processing flow of the conceptual system update processing unit 130. The concept system update processing unit 130 receives the user ID (or client terminal ID) and the update target concept list {cluster ID, {concept ID, feature score = Z, weight = w},. Is entered.

In step S <b> 16, the concept system update processing unit 130 updates the node value of each concept ID in the “update target concept list”. FIG. 32 shows details of the processing of the conceptual system update processing unit 130. The concept system update processing unit 130 uses the user interest score update formula for each concept i shown in FIG. 32 for the concept (appearance concept) that appeared in the content and the concept ID of the superordinate concept of this appearance concept, and the user interest score. TotalZ _in and the values of X _{i (n−1)} and Y _{i (n−1)} are obtained. Then, in the user interest score table of the conceptual system / user interest score database 140 in FIG. 16, each value stored in the column corresponding to the concept ID and the user ID (client terminal ID) input in step S12 in FIG. (TotalZ, X, Y) is updated.

Here, X _{i (n−1)} is the sum of the squares of the weights w of the update target concept list in the past (up to the previous time) for each concept ID (represented by the identifier i here). Similarly, Y _{i (n−1)} is the sum of multiplication of the weight w of the past update target concept list and the feature score Z for each concept ID (represented by identifier i here).

  X and Y hold intermediate results in the user interest score (TotalZ) calculation process. When prioritizing memory saving / storage, at least three real values of TotalZ, X, and Y are used as variables of each node. It can be realized by holding. When priority is not given to memory saving / storage, all the calculated concept scores and feature scores Z of the respective clusters are stored. In this case, storage of X and Y is not necessary.

In FIG. 32, n is an identifier indicating how many times the concept system update process is performed. A series of processes for obtaining the user interest score TotalZ is performed in units of cluster IDs. When the unit in which this series of processes is performed is counted once, n is the number of times this series of processes has been performed. Is an identifier. i is an identifier of a concept ID. Z _in is a Z value used for each update process of concept i. Note that Z _ij is a Z value for each of the list browsing content list and the detailed browsing content list, and has a relationship of Z _ij εZ _in . Weight w _in is the weight to be used in each process of updating the concept i. The weight w is the same as that set by the feature score calculation unit 120.

For example, w _in the favorite registration, if the user a special operation other than the inspection of the public, such as with the other user has performed, and (interval to the end from the viewing start) content viewing time, related the degree of content and concepts, The value is changed according to the content browsing time, whether it is weather / temperature / humidity / season / day of the week / holiday / leisure at the time of browsing, user location information at the time of browsing, event information collected from scheduler / diary, etc. It is also changed when specified by other service users and service operators. Alternatively, the value can be changed for each concept (or concept group). Thereby, for example, a setting that places importance on learning of a specific concept can be made based on the judgment of the service provider.

Note that the forgetting of the user interest score, such as to reduce the influence of the history after a certain period, attenuates TotalZ, X, and Y, respectively, when the threshold of the time interval from the last update time to the current time is exceeded. It will be realized.
An example of an attenuation calculation formula is shown. For example, k can be set as an attenuation rate (for example, k = 0 · 8), and can be calculated as follows.
TotalZ (after attenuation) = k × TotalZ (current)
X (after attenuation) = k ² × X (current)
Y (after attenuation) = k ² × Y (current)
Further, in step S <b> 17, the concept system update processing unit 130 extracts a subordinate concept of each concept ID (appearance concept and superordinate concept) in the “update target concept list”, and updates the node value of the subordinate concept. In the extraction of subordinate concepts, for each concept ID of the “update target concept list”, the concept system table of the concept system / interest degree database 140 in FIG. 16 is referred to, and a list of concept IDs is extracted from the child concept ID list. The process of extracting the child concept list with reference to the concept system table for the concept ID of each child concept ID list is repeated.

The feature score Z used to update the interest level of the lower concept is, for example, the value having the largest absolute value of the feature score among the adjacent parent nodes, the value of the closest higher node, the average of the values of the parent nodes Or a probability-coupled value. In the “update target concept list”, the user interest scores of the concepts updated in step S16 (concepts that appear in the content and higher-level concepts) are not updated.
In addition, when the narrowing-down condition or the exclusion condition is designated, the interested concept system update processing unit 130 applies the feature score Z ′ to each concept based on the relationship information (superordinate concept and subordinate concept) between the concepts in the conceptual system. Update user interest score.

(When filtering conditions are specified)
The interest concept system update processing unit 130 propagates the feature score for each selection calculated as shown in FIG. 25 to the descendant concept as shown in FIG. For example, the feature score (Z ′ = 0.67) calculated for “Western Sake” is also added to the child node of “Western Sake” by stochastic coupling. When the feature score is propagated to the descendant concept, and the selection of the narrowing condition selection path is considered and all selections regarding the narrowing condition are stochastically combined, the result shown in FIG. 34 is obtained. Alternatively, when the condition selection route as shown in FIG. 23 is not specified, the result shown in FIG. 35 is obtained. The values in FIGS. 34 and 35 represent the values of (Z, X, Y) used to calculate the user interest score shown in FIG.

(When an exclusion condition is specified)
The interest concept system update processing unit 130 propagates the feature score for each selection calculated as shown in FIG. 29 to the descendant concept as shown in FIG. For example, the feature score (Z ′ = − 0.67) calculated for “Western Sake” is also added to the child node of “Western Sake” by stochastic coupling. If the feature score is propagated to the descendant concept and the selection of the exclusion condition selection path is considered, and all the selections regarding the exclusion condition are stochastically combined, the result shown in FIG. 36 is obtained. Alternatively, when the exclusion condition selection route as shown in FIG. 27 is not specified, the result shown in FIG. 37 is obtained. The values in FIGS. 36 and 37 represent the values of (Z, X, Y) used to calculate the user interest score shown in FIG.
Furthermore, when the result shown in FIG. 34 and the result shown in FIG. 36 are integrated, the result shown in FIG. 38 is obtained. When the result shown in FIG. 35 and the result shown in FIG. 37 are integrated, the result shown in FIG. 39 can be obtained.

(Content Evaluation Processing Unit 170)
FIG. 40 shows a processing flow of the content evaluation processing unit 170. The content evaluation processing unit 170 receives the notification from the content request transfer unit 360 of the content server 300, reads a presentation content list in the format shown in FIG. 11 from the content table of the content database 160, and performs the following content evaluation processing. Do. The content request transfer unit 360 receives content request data including a user ID (or client terminal ID) as shown in FIG. In addition, with regard to the above-mentioned presented content list, whether the content registered within the past number of days is to be evaluated based on the prior setting of the service operator or service user (client terminal user) (whether it is included in the presented content list) ) Can be set.

  In the analysis target concept filtering in step S21, when the service operator or the service user designates a concept ID that is particularly an analysis target or a concept ID that is excluded from the analysis target in the condition selection history receiving unit 190 or presetting, With reference to the concept system / interest degree database 140, as shown in FIG. 14, only the concept ID and the lower-level concept ID of the specified analysis target are evaluated. Alternatively, the concept ID and lower concept ID specified in the exclusion condition are excluded from the evaluation target. The content evaluation processing unit 170 performs filtering according to the conditions set in advance by the service operator or service user with respect to the concept ID held by the input presentation content list, and generates a “filtered content list”.

  For example, when the user requests a recommendation for content related to baseball, the concept system table in FIG. 16 is referred to, and only the subordinate concepts of the concept ID corresponding to baseball are analyzed. The “filtered content list” is a content list obtained by filtering the concept ID linked to each content ID according to the conditions set in advance by the service operator or the user. The “filtered content list” has the same data configuration as the presented content list of FIG.

  In step S22, the content evaluation processing unit 170 calculates an evaluation score of the content included in the “filtered content list”, and generates a content score list as shown in FIG. The content score list has a content ID, an evaluation score, a content body, and a content registration time.

FIG. 42 shows an example of a method for calculating the evaluation score. For example, according to the content evaluation formula shown in FIG. 42, the evaluation score EntityZ _x for the content x is changed to the user interest score TotalZ _{i of} the concept i, the relevance w _i between the content x and the concept _i (or the importance of the concept i), and It can be calculated using a set p of concept IDs appearing in the content x. The concept identifier i corresponds to the concept ID in the set p.

The user interest score (TotalZ) used in the calculation of FIG. 42 is obtained from the user interest score table (FIG. 16) of the concept system / user interest score database 140 for the user ID (or client terminal) for the concept ID related to each content. Read out based on (ID). In FIG. 42, when the evaluation content 1 in which the concept K, the concept B, and the concept D appear is set as the evaluation content, the evaluation score EntityZ _{evaluation content 1} using the TotalZ, w of the concept K, the concept B, and the concept D = 0. 18 can be calculated. On the other hand, when the content 2 in which only the concept B appears is set as the evaluation content, the evaluation score EntityZ _{evaluation content 2} = −0.3 can be calculated using the TotalZ, w of the concept B. Content 1 with a large value of evaluation score EntityZ _x is displayed preferentially.

In addition, the evaluation score EntityZ _x can be obtained by the following methods 1 to 3.
As a first modification, it is obtained by EntityZ _x = MAX (TotalZ _i * w _i ). MAX (TotalZ _i * w _i ) is a function that returns the maximum value of TotalZ _i * w _i for i∈p.

The second modification, the value of EntityZ _x When the value of the MAX (TotalZ _i * _{w i)} exceeds the threshold value, the return value of _{MAX (TotalZ i * w i)} . MAX (TotalZ _i * w _i ) is a function that returns the maximum value of TotalZ _i * w _i for i∈p. If the threshold is not exceeded, the result of the content evaluation formula in FIG. 42 is set to EntityZ _x . MAX () is, is a function that returns the maximum value in TotalZ _i * _{w i} of i∈p. The threshold is a value set by the service operator.

As a third modification example, only the value iεp where TotalZ _i is a positive value is extracted, and the value integrated by the content evaluation formula of FIG. 42 is EntityZ _x .
In step S23, the content evaluation processing unit 170 sorts the content included in the content score list in descending order of the evaluation score EntityZ _x , and outputs the sorted content score list to the sorted content score list transmission unit 180.

(Sorted content score list transmission unit 180)
The sorted content score list transmission unit 180 transmits the sorted content score list and the user ID (or client terminal ID) input from the content evaluation processing unit 170 to the content server 300 via the communication network.

  As described above, according to the above configuration, the list of candidates for user selection is defined, and the number of appearances of concepts in content selection from there is analyzed, thereby taking into consideration the rare occurrence of each concept. In addition, since it is possible to use history features of concepts not selected from the list, it becomes possible to estimate the user's interest with high accuracy.

  Further, as shown in steps S15, S16, and S17, when calculating the feature score and the user interest score, the user's situation at the time of browsing, the characteristics of the browsing operation (favorite registration, long-time browsing, etc.), etc. Can be reflected via the weighting coefficient (weight w), so that the user interest score can be obtained with higher accuracy.

  In addition, in the content browsing history analysis with the concept defined by taxonomies (ontologies) as meta tags, it was difficult to rationally reflect the rareness of concept appearances in the analysis. Although it was necessary to adjust the ontology structure side such as unifying the abstraction level of the concept to be given to the content by fixing the thickness, etc., in this embodiment, the user interest score of the ontology structure considering the rarity of concept appearance Since it is possible to analyze the content with the higher concept by the update process and the browsing history of the content with the lower concept, the restriction on the taxonomy (ontology) used for analysis is reduced, and the taxonomy (ontology) is reduced. ) Maintenance, operation, and management costs can be reduced.

Furthermore, by calculating the user's evaluation score for content using the user interest score, it is possible to recommend content that matches the user's interest.
Moreover, in this embodiment, the interest estimation using the user's history and the interest estimation using the condition designation history performed by the user can be combined, and the information desired by the user at that time can be accurately presented.

  For example, the example shown in FIG. 43 is a case where the user designates “beer” as a narrowing condition. In this case, both the first content list and the second content list are “beer” content. Naturally, the second content list is all about “beer”, and the user chooses relatively “ale” compared to chance, or does not choose “lager” relatively compared to chance. From the facts, it is possible to estimate the interest of “I like ale” or “I don't like lager”, but I am interested in “beer” even if the user specifies “beer” as the filtering condition. Cannot learn.

  In response to this problem, in the present embodiment, when the user selects a narrowing condition from the narrowing condition selection candidates, the specification of the narrowing condition is analyzed to the extent that it is relatively “selected” or “not selected” compared to chance. , Which estimates the user's interests, and can be used in conjunction with the interest estimation based on the content selection history, so that even if the user specifies the narrowing conditions, the user's interests can be analyzed rationally and accurately. It becomes.

  For example, in the conceptual system of FIG. 22, when beer or red wine is specified as a narrowing condition, “red wine” is narrowed down from the choices under “wine” (“red wine”, “white wine”, “rose”). How rare it is to select as a condition compared to chance, or “Beer” as a narrowing condition from the choices under “Western Sake” (“Wine”, “Beer”, “Whisky”) The user's interest score can be estimated using the degree to which the selection is rare compared to chance, and the interest estimation accuracy is improved by using it together with the interest estimation result using the content browsing history. I can do things.

  Furthermore, with regard to the narrowing selection result, it is intended to increase the interest estimation accuracy by specifying the narrowing condition selection path from the superordinate concept in the concept system and using the selection for the interest estimation, For example, in FIG. 23, when selection is made from the upper concepts in the order of “alcohol”, “western sake”, “wine”, “red wine”, the options under “wine” (“red wine”, “white wine”, “rose” )), Not only the fact that “red wine” was selected as a narrowing condition, but also the choices under “Sake” (“Japanese liquor”, “Western liquor”, “Korean liquor”, “Chinese liquor” The interest estimation accuracy can be improved by further analyzing the fact that “Western Sake” was selected from “Sake”).

  Further, in the present embodiment, when the user selects an exclusion condition from the list browsing condition list, the user's interest is estimated by utilizing the feature of “select” or “not select” relatively compared to chance. Thus, the user's interest can be analyzed reasonably and accurately even when the user designates an exclusion condition by using it together with the interest estimation based on the content selection history.

  For example, in the conceptual system of FIG. 26, when whiskey is specified as an exclusion condition, “whiskey” is selected as an exclusion condition from the options under “Western sake” (“wine”, “beer”, “whiskey”). It is possible to estimate the user's interest score by using how rare it is compared with chance, and to improve the accuracy of interest estimation by using it together with the result of interest estimation using content browsing history. it can.

  Furthermore, the exclusion condition selection result is intended to improve the accuracy of interest estimation by specifying the exclusion condition selection route from the superordinate concept in the concept system and using the selection for the interest estimation as well. For example, in FIG. 27, when the selection is made in the order of “alcohol”, “western sake”, and “whiskey”, from the options under “Western sake” (“wine”, “beer”, “whiskey”) In addition to the fact that “Whisky” was selected as an exclusion condition, among the choices under “Sake” (“Japanese Sake”, “Western Sake”, “Korean Sake”, “Chinese Sake”) Therefore, the interest estimation accuracy can be improved by further analyzing the fact that “Western sake” was selected.

  In the present embodiment, the feature score calculated for each selection is also applied to the child concept group. For example, in the example of FIG. 33, when the feature score for Western liquor is calculated to be 0.67, interest estimation accuracy is also increased by adding the feature scores for “wine” and “red wine” which are the concept of descendants of Western liquor. Can be increased.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. Further, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, you may combine suitably the component covering different embodiment.

  DESCRIPTION OF SYMBOLS 100 ... Interest analysis apparatus, 200 ... Client terminal, 300 ... Content server, 110 ... History information receiving part, 120 ... Feature score calculation part, 130 ... Concept system update processing part, 140 ... Concept system / user interest score database, 150 ... Presented content list receiving unit, 160 ... content database, 170 ... content evaluation processing unit, 180 ... sorted content score list transmitting unit, 185 ... list browsing condition list creating unit, 190 ... condition selection history receiving unit, 210 ... history collecting unit , 220 ... history information transmission unit, 230 ... content presentation unit, 240 ... content request transmission unit, 250 ... list browsing condition list request transmission unit, 260 ... list browsing condition list presentation unit, 270 ... browsing condition selection history collection unit, 280 ... Condition selection history transmission unit, 310 ... Content transmission processing 320 ... Sorted presentation content list reception unit 330 ... Presentation content list transmission unit 340 ... Presentation content list input unit 350 ... History information transfer unit 360 ... Content request transfer unit 370 ... List browsing condition list request transfer unit 380 ... List browsing condition list transfer unit, 390 ... Condition selection history transfer unit.

Claims (5)

A method of analyzing a user's interest using a concept system in which user interest scores for a plurality of concepts are systematized by a computer,
Obtaining a first condition list in which a plurality of concepts as candidates for selection of narrowing conditions in the concept system are browsed as a list, and a second condition list including the concept selected by the user from the first condition list And steps to
The total number of concepts included in the first condition list is a first total number, the number of occurrences of the concept selected by the user in the first condition list is a first appearance number, and the second condition When the total number of concepts included in the list is the second total number and the number of occurrences of the concept selected by the user in the second condition list is the second appearance number, the first total number, the first number A first probability that the number of occurrences of the concept in the second condition list is greater than or equal to the second occurrence number under the condition of the occurrence number of 1 and the second total number, and the second occurrence Calculating a second probability that is less than or equal to a number, and calculating a feature score by an inverse function of a cumulative distribution function of a standard normal distribution based on the first probability and the second probability;
An updating step of updating the user interest score for the concept using the feature score. A method of analyzing a user's interest using a concept system in which user interest scores for a plurality of concepts are systematized by a computer,
Obtaining a first condition list obtained by browsing a plurality of concepts as candidates for selection of exclusion conditions in the concept system, and a second condition list selected by the user from the first condition list; ,
The total number of concepts included in the first condition list is a first total number, the number of occurrences of the concept selected by the user in the first condition list is a first appearance number, and the second condition When the total number of concepts included in the list is the second total number and the number of occurrences of the concept selected by the user in the second condition list is the second appearance number, the first total number, the first number A first probability that the number of occurrences of the concept in the second condition list is greater than or equal to the second occurrence number under the condition of the occurrence number of 1 and the second total number, and the second occurrence Calculating a second probability that is less than or equal to a number, and calculating a feature score by an inverse function of a cumulative distribution function of a standard normal distribution based on the first probability and the second probability;
An updating step of updating the user interest score for the concept using the feature score.   The calculating step specifies a selection path in the concept system based on a concept included in the second condition list, and calculates the feature score for each concept corresponding to each selection occurring in the selection path. The interest analysis method according to claim 1, wherein the method is an interest analysis method.   The interest analysis method according to any one of claims 1 to 3, wherein the updating step updates a user interest score of a subordinate concept of the concept using the feature score.   The content further includes an evaluation step of calculating a user evaluation score for the content using the user interest score of each concept that appears in the content for the content in which one or more concepts appear. 5. The interest analysis method according to any one of items 1 to 4.

Images