JP2013205875A - Recommend item detection program, device and method for detecting item to be reviewed by reviewer - Google Patents

Abstract

A recommendation item detection program capable of detecting an item for which a description of a criticism comment is desired according to a critic's specialty, while a criticism comment is posted by a large number of critics. I will provide a.
A reviewer goodness accumulating means for storing an estimated goodness of each item for each reviewer identifier, an item attention degree accumulating means for storing an estimated attention degree of the entire user for each item, and a reviewer identifier to be recommended The reviewer applicability calculation means for calculating the applicability of the item to the reviewer of the reviewer identifier from the estimated goodness of the item and the estimated attention level of each item, and the reviewer identifier is selected based on the applicability of the item The computer is caused to function as item selection means.
[Selection] Figure 6

Description

  The present invention relates to a technology of an EC (Electronic Commerce) site service for disclosing criticism comments posted from critics to items to an unspecified number of third parties.

  FIG. 1 is a system configuration diagram including an EC site server in the prior art.

  According to FIG. 1, the EC site server 3 is connected to the Internet. The “EC site server” is a site that sells items to users, and specifically may be a site that downloads an application to a smartphone, for example. According to the EC site, a web beacon is embedded in an item purchase completion page. For example, a display history of a purchase completion page displayed when the user 1 purchases an item is accumulated in the database as a purchase history.

  In recent years, according to the EC site, an unspecified number of users who are actually purchasing or considering an item can post a review comment (review) on the product. A user who posts a review comment is generally referred to as a “reviewer”. The criticism comment is disclosed to an unspecified number of third parties, and a user who is considering purchasing the product can freely view the criticism comment and use it as a judgment material for purchase.

  Here, the “user” refers to an entity that performs a selection action based on a preference such as purchase / viewing an item. The “item” may be a product / service, and includes, for example, a product such as a home appliance, a smartphone application, content of an electronic book, viewing content such as a web page or a television.

  Also, according to FIG. 1, an SNS (Social Networking Service) site server 5 is connected to the Internet. The “SNS site server” publishes posted text (which can be regarded as a critical comment for a specific item) posted by one user in a group of a plurality of users. For example, there are facebook (registered trademark), twitter (registered trademark), google + (registered trademark), and mixi (registered trademark), which are also generally referred to as miniblog sites. The critical comments include, for example, tweets about the app on the SNS site server and articles summarizing how to use the app on the blog site server.

  For example, according to the EC site server 3, for example, a “Like” button of facebook is displayed on the display part for each comment comment posted by a critic. This button simply means that the user has an interest card. General readers who have a good appreciation for the comment can click the “Like” button. Here, a general reader who performs evaluation (stage evaluation, rating) on a critical comment is hereinafter referred to as a “critical reviewer (rater)”. When the button is clicked by a critique evaluator, information indicating that the critique evaluator has received a critique of the critique comment is collected in the SNS site server 5. Then, the information is disclosed to other users belonging to the group of the SNS site server site. These other users will be able to see critical comments for which the critical reviewer has a critique.

  Furthermore, according to FIG. 1, it can be assumed that the recommendation device 7 is provided on the Internet. 2. Description of the Related Art Conventionally, there is a recommendation service technology that recommends an item based on a user's preference using the user's implicit action history information. Examples of the action history information include an item purchase history and a usage browsing history for the user. As an algorithm for estimating items to be recommended, there is typically “Collaborative Filtering” (for example, see Non-Patent Documents 1 and 2). According to this algorithm, it is estimated that an item with a high degree of relevance should be recommended to the user based on the action history for the user's item and the action history for another user's item. As a result, the item recommended to the user is highly likely to meet the user's preference. The recommendation device 7 in FIG. 1 recommends an item to each user in cooperation with, for example, the EC site server 3.

  2. Description of the Related Art Conventionally, there is a technique for recommending a conversation partner who can communicate smoothly using history information during synchronous communication such as conversation or chat (see, for example, Patent Document 1). According to this technique, the attribute of the request source is compared with the attribute of the information provider, and the information provider who will provide information according to the preference of the request source can be selected. Thereafter, the activity of information exchange between them can be further collected, and the attribute of the information provider can be further updated according to the activity.

  In addition, the client sends a search request for an individual requesting a criticism comment to the server, and the server detects the document linked to the criticism comment by the similarity of the contents, and identifies the critic who created the document. There is a technique for extracting (see, for example, Patent Document 2). According to this technique, it is possible to calculate an index value of criticism experience in the same knowledge field and rank candidate individuals.

  Further, for an item selection support device that recommends a plurality of items that are suitable for the user, a similar representative recommender similar to the representative recommender is extracted, and the recommended item of the similar representative recommender is also recommended to the user. (For example, refer to Patent Document 3). The user is displayed so that recommended items can be selected.

JP 2011-013786 A JP 2010-061183 A JP 2007-200099 A

P. Resnick, N. Iacovou, M. Suchak, P. Bergstrom, and J. Riedl. “GroupLens: Open Architecture for Collaborative Filtering of Netnews” In Conference on Computer Supported Cooperative Work, pp. 175-186 (1994) Badrul Sarwar, George Karypis, Joseph Konstan, and John Reidl. 2001. Item-based collaborative filteringrecommendation algorithms.In Proceedings of the 10th international conferenceon World 'Wide Web (WWW '01) .ACM, New York, NY, USA, 285- 295.

  According to the above-mentioned prior art, the level of critic's comment comment is graded (like a simple assessment! 5 stars (★★★ ☆☆)), Bad / Some of them are expressed as “Good”, and the information is certainly useful for users who are interested in the item. In addition, it is possible to carefully select useful criticisms such as displaying / not displaying criticism comments according to the classification of the critique evaluation levels by critics, and it is easy for the user to browse.

  However, many critique comments were based on a volunteer spirit where critics actively searched for items and wrote critique comments on those items. In addition, it is only expected that other users' responses (returning actions such as gratitude comments, evaluation points, retweets, etc.) and returns (affiliates, point reductions, etc.) to the comment comment described. Therefore, from the viewpoint of the EC site server, it does not increase the efficiency with respect to the act in which critics actively write criticism comments. In addition, each critic has a different field of expertise regarding insight into items.

  Here, the inventors thought that it would be beneficial to recommend (recommend) an item for which a criticism comment is desired depending on the field of criticism. In other words, it is beneficial for a large number of unspecified third parties to write criticism comments on items in their field of expertise as much as possible for each critic. On the other hand, for items that are not of interest to a large number of unspecified third parties, it is the responsibility of the critic to request that critics write comments about criticism, and critique comments are desired as much as possible. It would be beneficial to have preferential comments written on items that are being considered for purchase by many users.

  In addition, according to the conventional recommendation service technology, only a preference item with respect to the user is estimated, and there is an inorganic impression, which often does not lead to a user's purchase (use) willingness (marketing). In particular, the recommendation information itself has a strong image as a commercial (advertisement information, advertisement information), and is often less interesting for users with a low willingness to purchase. In addition, there are many cases where some users do not desire distribution of recommendation information such as advertisement information.

  In addition, if a critic's “Like” is given to a critic's critique comment through the SNS site server, other users will be forced to view it. It does not necessarily lead to purchase incentives. This is because the preferences of the members (followers and friends) of the group belonging to the SNS site server are not necessarily the same.

  In addition, according to the technique described in Patent Document 1, only an appropriate one-to-one communication (conversation) partner is found. Moreover, according to the technique described in Patent Document 2, the user himself / herself searches consciously, and does not recommend implicit information. In addition, critics' evaluations are expressed in terms of activity, such as the number of document creation experiences, and are not based on the user who receives the review. Furthermore, according to the technique described in Patent Document 3, the user actively selects an item, and automatically detects a critic who describes a criticism comment based on the user's preference. Absent.

  Therefore, according to the present invention, a recommendation that can detect an item for which a description of a criticism comment is desired according to a critic's specialty field, in which criticism comments are posted by a large number of critics on a large number of items. An object is to provide an item detection program, apparatus, and method.

According to the present invention, in a recommended item detection program for causing a computer mounted on a device to function to recommend an item to which a comment is to be posted to a reviewer in a reviewer identifier,
Critics strength storage means that stores the estimated strength of each item for each reviewer identifier,
Item attention storage means for storing the estimated attention of the entire user for each item;
A critic applicability calculating means for calculating the applicability of the item to the critic of the reviewer identifier from the estimated goodness in the reviewer identifier to be recommended and the estimated attention level in each item;
For the item, the computer is caused to function as item selection means for selecting a reviewer identifier based on the degree of application.

According to another embodiment of the recommended item detection program of the present invention,
It is preferable to further include recommendation means for transmitting the item selected by the item selection means to a terminal operated by a reviewer of the selected reviewer identifier.

According to another embodiment of the recommended item detection program of the present invention,
For each reviewer identifier, a reviewer posting table storage means for storing a reviewer posting table indicating whether or not the reviewer has posted for each item;
A computer as a collaborative filtering means for calculating the estimated goodness of each unposted item for each reviewer identifier from the reviewer posting table accumulating means using collaborative filtering, and outputting the estimated goodness to the critic goodness accumulating means It is also preferable to further function.

According to another embodiment of the recommended item detection program of the present invention,
For each user list, an item interest list table accumulating unit that accumulates a table indicating the interest of each item,
As an item attention level calculation means for calculating an estimated attention level by averaging (or accumulating, normalizing, maximum value) the value of presence / absence of interest for each item and outputting the estimated attention level to the item attention level storage means It is also preferable to make the computer further function.

According to another embodiment of the recommended item detection program of the present invention,
It is also preferable to make the computer function so that the user list is a wish list, a bookmark registration list, a wish list or a favorites list.

According to another embodiment of the recommended item detection program of the present invention,
When the user U does not belong to the communication group of the critic R, the critic applicability calculation means calculates the value of the interest of each item in the user list of the user U with respect to the item interest list table storage means. It is also preferable to make the computer function so that it becomes zero.

According to another embodiment of the recommended item detection program of the present invention,
The item interest list table storage means is given a weighting coefficient W (U) for each user list,
The item attention level calculating means calculates an estimated attention level by averaging (or accumulating, normalizing, and maximum value) a value obtained by multiplying the value of presence / absence of interest in each item for each user list by a weighting factor. It is also preferable to make the computer function.

According to another embodiment of the recommended item detection program of the present invention,
The critic applicability calculation means functions the computer to reset the weight coefficient W (R, U) of the user U based on the user list for the critic R of the reviewer identifier to the item interest list table storage means. It is also preferable that

According to another embodiment of the recommended item detection program of the present invention,
For each item, there is further a criticism comment accumulating means for accumulating criticism comments for each reviewer identifier,
Recommendation means are
Search for reviewer identifiers of other critics whose estimated goodness is equal to or greater than a predetermined threshold for the item to be sent from the critic goodness accumulation means,
Search for comment comments based on reviewer identifiers of other reviewers from the comment comment storage means,
It is also preferable to cause the computer to function so as to transmit the comment to the terminal operated by the reviewer of the reviewer identifier selected by the item selection means.

According to another embodiment of the recommended item detection program of the present invention,
Recommendation means are
Search for reviewer identifiers of other critics whose estimated goodness is equal to or higher than a predetermined threshold for the item to be sent from the critic goodness accumulating means,
Search for other reviewers' reviewer identifiers and review comments based on the item from other servers,
It is also preferable to cause the computer to function so as to transmit the comment to the terminal operated by the reviewer of the reviewer identifier selected by the item selection means.

According to another embodiment of the recommended item detection program of the present invention,
The reviewer submission table of the reviewer posting table storage means that, for each reviewer identifier, when there are more than a predetermined threshold number of review posts by a plurality of reviewers for a review comment posted for each item by the reviewer. It is also preferable to express “posted”.

According to the present invention, in the recommended item detection device that functions to detect an item for which a review comment is to be posted to a reviewer in a reviewer identifier,
Critics strength storage means that stores the estimated strength of each item for each reviewer identifier,
Item attention storage means for storing the estimated attention of the entire user for each item;
A critic applicability calculating means for calculating the applicability of the item to the critic of the reviewer identifier from the estimated goodness in the reviewer identifier to be recommended and the estimated attention level in each item;
Item selection means for selecting a reviewer identifier based on the applicability of the item.

According to the present invention, in the recommended item detection method for causing a critic in a reviewer identifier to function to detect an item to which a comment is to be posted, using the apparatus,
The critic strength storage section that stores the estimated strength of each item for each reviewer identifier,
An item attention storage unit that stores the estimated attention of the entire user for each item,
A first step of calculating the applicability of the item to the reviewer of the reviewer identifier from the estimated goodness of the reviewer identifier to be recommended and the estimated attention level of each item;
The item includes a second step of selecting a reviewer identifier based on the applicability.

  According to the recommended item detection program, apparatus, and method of the present invention, it is desired to describe a criticism comment according to a critic's field of expertise, while a criticism comment is posted by a large number of critics on a large number of items. Items to be detected can be detected. By recommending a detected item to a critic, the critic can post a criticism comment that makes the most of his ability as a critic.

It is a system block diagram containing EC site server in a prior art. It is a system configuration diagram in the present invention. It is a sequence diagram which collects a critical comment and recommends an item. It is explanatory drawing showing a critics contribution table. It is explanatory drawing showing an item interest list table. It is a functional block diagram of the recommendation item detection apparatus in this invention. It is explanatory drawing showing the presumption ability calculated by collaborative filtering. It is explanatory drawing showing the item attention degree calculated by the item attention degree calculation part. It is explanatory drawing showing the application degree for every critic with respect to each user. It is a sequence diagram which collects a critical comment and recommends an item. It is a sequence diagram which collects the critics contribution table in other embodiments.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 2 is a system configuration diagram according to the present invention.
FIG. 3 is a sequence diagram for collecting critical comments and recommending items.

  According to FIG. 2, a recommended item detection device (server) 1 is connected to the Internet, as in FIG. Here, the recommended item detection apparatus 1 according to the present invention recommends an item having a high degree of strength of a critic to a critic who posts a criticism comment. The reviewer then expects to post a review comment on the recommended item.

  The “item” may be a link (URL) to a web page describing the details of the product / service or a hash tag related to the product / service (to the user, Recommend URLs and hashtags). Below, it demonstrates as application software installed in a smart phone, for example.

  “Critical comment” means a sentence written by a critic for an item. For example, it may be a comment published on the product detail page of the EC site, or a comment text posted on an external SNS site or blog site. When using a critical comment from an external SNS site or the like, it is also desirable to determine which item is a comment by using a commonly defined item identifier or by using a hash tag associated with the item.

(S301) First, the reviewer uses the terminal 4 to click the “comment” button for the desired item on the item page of the EC site server 3 to describe the comment. The critic's terminal 4 transmits a criticism comment associated with the item to the EC site server 3. The EC site server 3 transfers the critical comment to the SNS site server 5 of, for example, facebook (registered trademark) corresponding to the “Like!” Button. As a result, the SNS site server 5 publishes the transferred critical comment in the group.

(S302) The user purchases or uses various items displayed on the EC site server 3 or browses them with interest. In the process, the EC site server 3 registers the wish list (or bookmark list, wish list, favorite, purchase, use, browsing, etc.) of the user. The “wish list” is a list in which items (or URL (Uniform Resource Locator)) selected by a visitor (user) for purchase / use are registered as a list. For the operator of the EC site server 3, by analyzing a huge wish list collected from a large number of users, it is possible to extract items of interest to consumers in general.

(S303) The recommended item detection device 1 acquires the “critician posting table” and the “item interest list table” from the EC site server 3. The recommended item detection apparatus 1 acquires a reviewer posting table and an item interest list table from the EC site server 3 via an API (Application Programming Interface). The API is a rule interface for using the function of the application service. This API is prepared differently for each server.

  In addition, according to FIG. 3, although the recommendation item detection apparatus 1 acquires these information from EC site server 3, it is not restricted to these, You may acquire from the terminal 2 based on user operation. It may be acquired from another server.

[Critics submission table]
The “critician posting table” indicates whether or not the reviewer has posted each item for each reviewer identifier (binary value: 0 or 1). The “reviewer identifier” is an identifier assigned to each critic.

  FIG. 4 is an explanatory diagram showing a reviewer posting table.

  The table in FIG. 4 includes reviewer identifiers (columns) and items (rows). According to FIG. 4, the critic 1 indicates that a critical comment has been posted to the application 1 (the critical comment has been described = 1 / not yet described = 0). In addition, the critic 4 indicates that a critical comment is not posted to the application 1. The “application” is assumed to be application software that can be installed in, for example, a mobile phone or a smartphone.

[Item interest list table]
The “item interest list table” represents the interest of each item for each user list. For example, the usage history and interest history of each item are represented by binary values (0, 1). The user list is, for example, a wish list, a bookmark registration list (book marker), or a wish list.

  FIG. 5 is an explanatory diagram showing an item interest list table.

  According to FIG. 5, for each book marker (column), “1” is assigned only to items (rows) that the user has purchased or used in the past. For example, according to FIG. 5, it can be understood that “the user of the book marker 2 is interested in the app 1” and “the user of the book marker 4 is not interested in the app 1”. For example, according to FIG. 5, it can be understood that “the user of the book marker 1 is interested in the applications 1, 2, 3, and 5”.

  And the recommended item detection apparatus 1 of this invention accumulate | stores a critics contribution table and an item interest list table, and criticism comment is posted by many critics with respect to many items. The user's specialty field and the entire user's field of interest are analyzed, an item for which a comment is desired to be described is selected, and the item is recommended to a critic.

FIG. 6 is a functional configuration diagram of the recommended item detection device according to the present invention.
In the following, each functional component of FIG. 6 will be described with reference to the sequence of FIG. 3 described above.

  According to FIG. 6, the recommended item detection device 1 includes a communication interface 10 connected to the Internet, a critic strength storage unit 11, an item attention level storage unit 12, a critic applicability calculation unit 13, and an item selection. Section 14, recommendation section 15, and critical comment storage section 16. These functional components excluding the communication interface 10 are realized by executing a recommended item detection program that causes a computer installed in the apparatus to function.

[Critic Experts Accumulation Unit 11]
The critic strength accumulator 11 stores the estimated strength of each item for each reviewer identifier. Here, according to FIG. 6, the recommended item detection device 1 includes a reviewer posting table storage unit 111 and a collaborative filtering unit 112.

  The reviewer posting table storage unit 111 receives and stores the reviewer posting table (see FIG. 4 described above) from, for example, the EC site server 3 (or a user terminal or another server) (see S311 in FIG. 3).

  The collaborative filtering unit 112 calculates the estimated goodness of each unposted item for each reviewer identifier from the critic post table of the critic post table storage unit 111 using collaborative filtering (see S312 in FIG. 3). . The estimated goodness is obtained by quantifying (0-1) the goodness of an item (application) that has not been described for the critic. Items that have been described by other critics similar to the critic's posting tendency of criticism comments and that have not been described by the critic are high. The calculated estimated strength is accumulated by the critic strength accumulation unit 11 (see S313 in FIG. 3).

Collaborative filtering is executed in the following two steps using a reviewer posting table (see, for example, Non-Patent Documents 1 and 2). However, the following steps are examples of collaborative filtering, and are not limited to this method.
(S1) Analyzing the relationship between items, for example, by counting the number of combinations of items posted simultaneously by critics. For example, according to FIG. 4, three critics have posted comments about the apps 1 and 3. That is, the second item purchased together with the first item is cross tabulated.
(S2) Next, an estimated ratio posted as a set is calculated from the sum of the combinations. This estimated ratio is used as the estimated goodness.

  FIG. 7 is an explanatory diagram showing the estimated goodness calculated by the collaborative filtering.

According to FIG. 7, for example, the critic 1 has posted critical comments about the applications 1, 2, 3 and 5 in the past. Here, the critic's strength in the application to which the critical comment has been posted gives a minimum value (0) that defines collaborative filtering. Define critics' strength from 0 to 1.
Pref (R, A): Estimated goodness that critic R posts about app A
Pref (R, A) = 0: Critic R has posted a critical comment on app A

  Next, the estimated goodness that the critic 1 criticizes the unposted application 4 is 0.73. Further, the estimated goodness that the critic 1 criticizes the unposted application 6 is 0.99. Here, it is estimated that the critic 1 is highly skilled in the criticism comment regarding the application 6. The collaborative filtering algorithm is implemented in advance as a standard function in, for example, spreadsheet software.

[Item attention storage unit 12]
The item attention level accumulating unit 12 stores the estimated attention level of the entire user for each item. Here, according to FIG. 6, the recommended item detection device 1 further includes an item interest list table storage unit 121 and an item attention level calculation unit 122.

The item interest list table storage unit 121 stores the interest of each item for each user list (see FIG. 5 and S311 in FIG. 3). Those who are interested are assigned “1”.
φ (U, A) = 1: App A checked in user list U (interested)
= 0: Not checked in app A in user list U (not interested)

The item attention level calculation unit 122 calculates, for each item, an estimated attention level by averaging (or accumulating, normalizing, and maximum values) the value of interest (see S314 in FIG. 3). For example, the degree of attention for each item can be calculated by simply summing up the check for interest (1/0) for each item.
Fun (A): Estimated attention level of app A
Fun (A) = Σφ (U, A)
In addition, when the presence or absence of interest has a value range, the attention degree for each item may be calculated by normalization, for example. Note that normalization means a value between 0 and 1 in terms of (primary value−minimum value) / (maximum value−minimum value).

  FIG. 8 is an explanatory diagram showing the first item attention level.

  According to FIG. 8, for each reviewer identifier, the estimated attention level obtained by accumulating the evaluation levels from the respective critics for each item is represented. This estimated attention level is accumulated by the item attention level accumulation unit 12 (see S315 in FIG. 3). According to FIG. 8, the estimated attention level of the entire user with respect to the application 4 is “3”. Further, the estimated attention degree of the entire user with respect to the application 6 is “2”. In this case, it can be said that the estimated attention level of the entire user is higher in the application 4.

  FIG. 9 is an explanatory diagram showing the second item attention degree.

As another embodiment, according to FIG. 9, a weight coefficient W (U) is assigned to each user list U. Then, the item attention level calculation unit 122 calculates an average of the estimated attention level of each item using a value obtained by multiplying the value of presence / absence of interest by a weighting coefficient.
W (U): Weighting factor in user list U
Fun (A) = Σ (W (U) * φ (U, A))

According to FIG. 9, the estimated attention level of the entire user with respect to the application 4 is expressed as follows.
1 * 0.8 + 1 * 0.3 + 1 * 0.2 = 1.3
Moreover, the estimated attention degree of the whole user with respect to the application 6 is expressed as follows.
1 * 0.5 + 1 * 0.2 = 0.7
In this case, it can be said that the app 4 has a higher overall estimated user attention.

[Critic Applicability Calculation Unit 13]
The critic applicator calculation unit 13 applies the item to the critic of the reviewer identifier from the estimated goodness Pref (R, A) in the reviewer identifier R to be recommended and the estimated attention degree Fun (A) in each item. The degree Point (R, A) is calculated (see S316 in FIG. 3). Specifically, as shown in the following formula, the estimated goodness Pref (R, A) is multiplied by the estimated attention degree Fun (A).
Point (R, A): Applicability of item A by critic R
Point (R, A) = Fun (A) * Pref (R, A)

  FIG. 10 is an explanatory diagram showing the applicability of each critic for each user.

According to FIG. 10, the following is calculated for the app 4 that has not been posted by the critic 1.
Point (R, A) = 0.73 * 1.3 = 0.219
Further, the app 6 for which the critic 1 has not posted is calculated as follows.
Point (R, A) = 0.99 * 0.7 = 0.693
In this case, it can be said that the app 6 is more highly recommended for the critic 1.

As another embodiment, the critic applicability calculation unit 13 preferably considers whether a critic based on a reviewer identifier and a user based on a user list belong to a communication group. For example, this is a follow relationship for twitter, and a friend relationship for facebook. If the user U does not belong to the communication group of the critic R, the critic applicability calculation unit 13 determines whether the item interest list table storage unit 121 is interested in each item in the user list of the user U. Set the value to 0. In this case, the item attention level is expressed by the following formula.
Fun (R, A): Estimated attention level of item A by critic R
Fun (R, A) = Σφ (U, A) * Ψ (U, R)
Ψ (U, R) = 1: User U follows critic R
= 0: User U is not following critic R In this case, the critic applicability is expressed by the following equation. .
Point (R, A) = Fun (R, A) * Pref (R, A)

  In this way, by focusing on popular items in a communication group (community), “motivation to comment (enhance joy)” and “activate communication” (comment) It is thought that the potential effect of “discovering a theme to be solved” is also obtained.

As another embodiment, it is also preferable to consider the weight of the user U in the critic R for the item attention degree. The critic applicability calculation unit 13 resets the weighting factor W (R, U) of the user U based on the user list for the critic R of the reviewer identifier in the item interest list table storage unit 121. In this case, the item attention level is expressed by the following formula.
W (R, U): Weight of user U in critic R
Fun (R, A) = Σ (W (R, U) * φ (U, A))

Here, the weighting factor W (U) or W (R, U) evaluates the foresight of the app checked by the user list U. For example, an application checked by the user U for a certain period is recorded. Then, for each user list U, when an item with a high degree of item interest (for example, an app with a lot of sales) is checked in advance, a weighting factor is additionally fed back.
W '(U): Previous W (U)
α: Constant from 0 to 1 W (U) = αW ′ (U) + (1−α) φ (U, A)

[Item selection unit 14]
The item selection unit 14 selects a reviewer identifier for the item based on the applicability. Specifically, for each item, the reviewer identifier having the highest applicability Point (R, A) is selected (see S317 in FIG. 3). In addition, one or more reviewer identifiers whose applicability point (R, A) is equal to or greater than a predetermined threshold may be selected.

[Recommendation 15]
The recommendation unit 15 transmits the item to the terminal operated by the reviewer of the selected reviewer identifier (see S318 in FIG. 3).

[Critical Comment Accumulation Unit 16]
The critical comment accumulating unit 16 accumulates critical comments for each reviewer identifier for each item. The critical comment is acquired from, for example, the EC site server 3.

  On the other hand, the recommendation unit 15 first searches the reviewer identifiers of other critics whose estimated goodness is equal to or higher than a predetermined threshold for the item to be transmitted, from the critic goodness accumulation unit 11. Next, the recommendation unit 15 searches the review comment storage unit 16 for a review comment based on a reviewer identifier of another reviewer. Then, the recommendation unit 15 also transmits a review comment to the terminal operated by the reviewer of the reviewer identifier selected by the item selection unit 14. This series of processing makes sense for reviewers who recommend items to view review comments posted by other reviewers. Since the critic is highly skilled at the item, it is expected that he will post more appropriate criticism comments by browsing the criticism comments of other critics.

  As another embodiment, it is also preferable to obtain a criticism comment of another critic from another server (for example, an SNS site server) without providing the criticism comment accumulation unit 16. The recommendation unit 15 first searches the reviewer identifiers of other critics whose estimated goodness is equal to or greater than a predetermined threshold for the item to be transmitted from the critics goodness accumulating unit 11. Next, the recommender 15 retrieves a reviewer identifier of another reviewer and a review comment based on the item from another server (for example, an SNS site server). Then, the recommendation unit 15 also transmits a review comment to the terminal operated by the reviewer of the reviewer identifier selected by the item selection unit 14.

  FIG. 11 is a sequence diagram for collecting a reviewer posting table according to another embodiment.

(S301) As in S301 of FIG. 3 described above, first, the critic uses the terminal 4 and clicks the “Comment” button on the item page of the EC site server 3 to comment on the comment. Is described. The critic's terminal 4 transmits a criticism comment associated with the item to the EC site server 3.

(S321) Thereafter, the critique evaluator can browse the item in the EC site server 3 using the terminal 6, and can visually recognize the critique comment. At this time, it is assumed that the critique evaluator clicks a “Like” button for the critique comment, for example. The terminal 6 of the critic evaluator transmits this evaluation level to the EC site server 3.

(S322) The EC site server 3 transfers the critical comment to the SNS site server 5 of, for example, facebook (registered trademark) corresponding to the “Like” button. As a result, the SNS site server 5 publishes the transferred critical comment in the group.

(S323) The EC site server 3 counts the number of critical comments transferred for the item. Then, the EC site server 3 transmits the collected count number to the recommended item detection device 1 according to the present invention.

(S324) The recommended item detection device 1 determines whether the count received from the EC site server 3 is equal to or greater than a predetermined threshold, and reflects the result in the reviewer posting table. For example, when “Like!” (Evaluation posting) is posted to a critical comment number or more by a critic 1 with respect to the app 1, “posted (1)” is indicated in the app 1 for the critic 1. . In other words, in the reviewer posting table, for each reviewer identifier, when there are more than a predetermined threshold number of reviews by a plurality of reviewers for a review comment posted for each item by the reviewer, "Done".

  As described above in detail, according to the recommended item detection program, apparatus, and method of the present invention, criticism comments are posted by a large number of critics for a large number of items (products / services). Depending on the critic's field of expertise, it is possible to select an item for which a criticism comment is desired and recommend the item to the critic. In other words, it is possible to preferentially find an item to be criticized that is worth the critics' time. In addition, for EC site servers, we build a targeted and efficient affiliate business model that encourages critics who are highly adaptable to critics to write comments about items and reduce affiliate rewards. It becomes possible to do. In other words, when viewed from the EC site server, it is possible to increase the efficiency with respect to an act in which a critic is actively writing a criticism comment.

  Essentially, critics need to have a deep insight into the field of the item and the ability to write sentences that are easy for general consumers to understand. However, it is extremely difficult for critics to find a specific item that is an item in their field of expertise and that requires other users to comment on the criticism. By using the recommended item detection device of the present invention and the like, even for critics who are not very good at predicting items to be noticed in the future, such information is supported by supporting the act of searching for items. The discovery effort can be reduced.

  Various changes, modifications, and omissions of the above-described various embodiments of the present invention can be easily made by those skilled in the art. The above description is merely an example, and is not intended to be restrictive. The invention is limited only as defined in the following claims and the equivalents thereto.

DESCRIPTION OF SYMBOLS 1 Recommendation item detection apparatus 10 Communication interface 11 Reviewer's specialty storage part 111 Reviewer contribution table storage part 112 Collaborative filtering part 12 Item attention degree storage part 121 Item interest list table storage part 122 Item attention degree calculation part 13 Critic applicability degree Calculation unit 14 Item selection unit 15 Recommendation unit 16 Critic comment storage unit 2 User terminal 3 EC site server 4 Critic terminal 5 SNS site server 6 Critic evaluator terminal 7 Recommendation device

Claims (13)

In a recommended item detection program for causing a computer mounted on a device to function to recommend an item for which a comment is to be posted to a reviewer in a reviewer identifier,
Critics strength storage means that stores the estimated strength of each item for each reviewer identifier,
Item attention storage means for storing the estimated attention of the entire user for each item;
A critic applicability calculating means for calculating the applicability of the item to the critic of the reviewer identifier from the estimated goodness in the reviewer identifier to be recommended and the estimated attention level in each item;
A recommended item detection program for causing a computer to function as item selection means for selecting a reviewer identifier based on the applicability of the item.   The recommended item detection program according to claim 1, further comprising recommendation means for transmitting the item selected by the item selection means to a terminal operated by a reviewer of the selected reviewer identifier. For each reviewer identifier, a reviewer posting table storage means for storing a reviewer posting table indicating whether or not the reviewer has posted for each item;
Collaborative filtering that calculates the estimated goodness of each unposted item for each reviewer identifier from the critic post table storage means, and outputs the estimated goodness to the critic strength storage means The recommended item detection program according to claim 1, further comprising a computer functioning as means. For each user list, an item interest list table accumulating unit that accumulates a table indicating the interest of each item,
Item attention level calculating means for calculating an estimated attention level by averaging (or accumulating, normalizing, maximum value) the value of interest for each item and outputting the estimated attention level to the item attention level storage means The recommended item detection program according to any one of claims 1 to 3, wherein the computer is further functioned as a computer program.   5. The recommended item detection program according to claim 4, wherein the user list causes the computer to function as a wish list, a bookmark registration list, a wish list, or a favorite list.   If the user U does not belong to the critic R's communication group, the critic applicability calculation means determines whether each item in the user U's user list is interested in the item interest list table storage means. 6. The recommended item detection program according to claim 4, wherein the computer is caused to function so as to set the value to zero. The item interest list table storage means is given a weighting coefficient W (U) for each user list,
The item attention level calculation means calculates an estimated attention level by averaging (or accumulating, normalizing, and maximum value) a value obtained by multiplying the value of presence / absence of interest in each item for each user list by the weight coefficient. The recommended item detection program according to claim 4, wherein the computer is caused to function as described above.   The critic applicability calculating means resets the weighting factor W (R, U) of the user U based on the user list for the reviewer identifier critic R to the item interest list table accumulating means. The recommended item detection program according to claim 7, wherein: For each item, there is further a criticism comment accumulating means for accumulating criticism comments for each reviewer identifier,
The recommendation means is:
For the item to be transmitted, the reviewer identifiers of other critics whose estimated goodness is equal to or higher than a predetermined threshold are retrieved from the critic goodness accumulating means,
From the review comment accumulating means, search for a review comment based on the reviewer identifier of the other reviewer,
The recommendation according to any one of claims 1 to 8, wherein the computer is caused to function so as to transmit the review comment to a terminal operated by a reviewer of the reviewer identifier selected by the item selection unit. Item detection program. The recommendation means is:
For the item to be transmitted, the reviewer identifiers of other critics whose estimated goodness is equal to or higher than a predetermined threshold are retrieved from the critic goodness accumulating means,
Retrieve other reviewers' reviewer identifiers and review comments based on the item from other servers,
The recommendation according to any one of claims 1 to 8, wherein the computer is caused to function so as to transmit the review comment to a terminal operated by a reviewer of the reviewer identifier selected by the item selection unit. Item detection program. The reviewer posting table of the reviewer posting table accumulating means includes, for each reviewer identifier, a review comment of a plurality of review evaluators exceeding a predetermined threshold number for a review comment posted for each item by the reviewer. The recommended item detection program according to any one of claims 1 to 10, characterized in that "Posted" is displayed. In a recommended item detection device that functions to detect an item for which a review comment is to be posted to a reviewer in a reviewer identifier,
Critics strength storage means that stores the estimated strength of each item for each reviewer identifier,
Item attention storage means for storing the estimated attention of the entire user for each item;
A critic applicability calculating means for calculating the applicability of the item to the critic of the reviewer identifier from the estimated goodness in the reviewer identifier to be recommended and the estimated attention level in each item;
A recommended item detection apparatus comprising: an item selection unit that selects a reviewer identifier based on the applicability of the item. In a recommended item detection method, which uses a device to make a reviewer in a reviewer identifier function to detect an item for which a comment is to be posted,
The critic strength storage section that stores the estimated strength of each item for each reviewer identifier,
An item attention storage unit that stores the estimated attention of the entire user for each item,
A first step of calculating the applicability of the item to the reviewer of the reviewer identifier from the estimated goodness of the reviewer identifier to be recommended and the estimated attention level of each item;
And a second step of selecting a reviewer identifier for the item based on the applicability.

Images