Information processing device, information processing method, and program

Abstract

To provide a technique for realizing targeting in consideration of a relation between a plurality of users and a plurality of items to be advertised.SOLUTION: An information processing device is configured to: select a model user group and an object user group different from the model user group, from a plurality of users; generate user representation representing a relation between the plurality of users and a plurality of items, with respect to each user of the object user group; and estimates, as a similar user group, one or more users having similar features as users included in the model user group out of the object user group based on the user representation.SELECTED DRAWING: Figure 10

Description

[Technical field]

[0001] The present invention relates to an information processing device, an information processing method, and a program, and in particular to a technique related to advertisement provision. [Background technology]

[0002] In recent years, in the distribution of advertisements via the Internet, a group of users to which the advertisements are to be distributed (targeting) is being selected. As a targeting technique, for example, the technique disclosed in Patent Document 1 is known. This document describes that even if there is no purchase history of the product corresponding to the advertisement, a group of consumers who are estimated to have purchased the product corresponding to the advertisement based on the purchase history of other products are determined as the target of the advertisement. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2017-097717 A Summary of the Invention [Problem to be solved by the invention]

[0004] In the technology of Patent Document 1, advertisement delivery targets are determined based on a database that indicates characteristics related to the consumption behavior of each consumer belonging to multiple consumer groups that have been set in advance. However, this technology does not take into account the relationships between multiple users and multiple advertised items that may affect the consumption behavior of consumers (users), and there is a problem in that effective targeting is not achieved.

[0005] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a technique for realizing targeting that takes into account relationships between multiple users and multiple advertised items. [Means for solving the problem]

[0006] In order to solve the above problem, one aspect of an information processing device according to the present invention includes a selection means for selecting a model user group and a target user group different from the model user group from a plurality of users, a generation means for generating a user representation for each user in the target user group that represents a relationship between the plurality of users and a plurality of items, and a prediction means for predicting, based on the user representation, one or more users from the target user group that have characteristics similar to those of users included in the model user group as a similar user group.

[0007] The information processing device further has a user feature acquisition means for acquiring factual features of each of the multiple users as user features, and the prediction means can predict the similar user group from the user features of the target user group and the user expressions.

[0008] The information processing device may further include a construction means for constructing a graph representing mutual relationships between the plurality of users and the plurality of items, and the generation means may generate the user representation from the graph.

[0009] The information processing device may further have an item feature acquisition means for acquiring features related to a plurality of items as item features, and the construction means may construct the graph based on the user features of the plurality of users and the item features.

[0010] The construction means may construct a first relationship that is a relationship between the multiple users, a second relationship that is a relationship between the multiple items, and a third relationship that is a relationship between the multiple users and the multiple items based on the user characteristics and the item characteristics of the multiple users, and may construct the graph using the first relationship, the second relationship, and the third relationship.

[0011] The information processing device further has a task setting means for setting a plurality of tasks representing a classification of the plurality of items from the item features, and the construction means can construct a graph network representing the mutual relationships among the plurality of users, the plurality of items, and the plurality of tasks as the graph based on the user features of the plurality of users and the item features.

[0012] Each of the plurality of tasks may be a brand name for each of the plurality of items.

[0013] The prediction means may predict the similar user group using a learning model for machine learning configured to input the user characteristics and user expressions of the target user group and output the possibility of having characteristics similar to those of a user included in the model user group.

[0014] The prediction means can predict one or more users whose likelihood is greater than a predetermined threshold as the similar user group.

[0015] The information processing device may further include a distribution means for distributing an advertisement to the similar user group.

[0016] In order to solve the above problem, one aspect of an information processing method according to the present invention includes a selection step of selecting a model user group and a target user group different from the model user group from a plurality of users, a generation step of generating a user representation for each user in the target user group that represents a relationship between the plurality of users and a plurality of items, and a prediction step of predicting, based on the user representation, one or more users from the target user group that have characteristics similar to those of users included in the model user group as a similar user group.

[0017] In order to solve the above problems, one aspect of the program according to the present invention is an information processing program for causing a computer to execute information processing, the program causing the computer to execute processes including a selection process for selecting a model user group and a target user group different from the model user group from a plurality of users, a generation process for generating a user representation for each user in the target user group that represents the relationship between the plurality of users and a plurality of items, and a prediction process for predicting, based on the user representation, one or more users from the target user group that have characteristics similar to those of users included in the model user group as a similar user group. Effect of the Invention

[0018] The present invention enables targeting that takes into account relationships among multiple users and multiple advertised items. The above-mentioned objects, aspects, and advantages of the present invention, as well as objects, aspects, and advantages of the present invention not described above, will be understood by those skilled in the art from the following detailed description of the invention by referring to the accompanying drawings and the claims. [Brief description of the drawings]

[0019] [Figure 1] FIG. 1 shows an example of the configuration of an information processing system. [Diagram 2] FIG. 2 shows an example of the functional configuration of the information processing device 10 according to the first embodiment. [Diagram 3] Figure 3 shows a flowchart of the knowledge graph creation procedure. [Figure 4A] FIG. 4A is a diagram for explaining an explicit link. [Figure 4B] FIG. 4B is a diagram for explaining an implicit link. [Diagram 5] FIG. 5 is a diagram for explaining the process of inferring the relationship between links. [Figure 6A] FIG. 6A shows a conceptual diagram of a score (closeness score) based on the closeness of the relationship between a pair of users. [Figure 6B]FIG. 6B shows a schematic architecture of the score prediction model 112. [Figure 6C] FIG. 6C shows a conceptual diagram of a user relationship graph. [Figure 7] FIG. 7 shows a conceptual diagram of an item relationship graph. [Figure 8] FIG. 8 shows a conceptual diagram of a user-item relationship graph. [Figure 9] Figure 9 shows a conceptual diagram of the knowledge graph and user representation. [Figure 10] FIG. 10 is a diagram for explaining the prospective user prediction process according to the first embodiment. [Figure 11] FIG. 11 shows an example of the hardware configuration of the information processing device 10 and the user device 11. As shown in FIG. [Figure 12] FIG. 12 shows a flowchart of the process executed by the information processing device 10 according to the first embodiment. [Figure 13] FIG. 13 shows an example of the functional configuration of an information processing device 10 according to the second embodiment. [Figure 14] FIG. 14 shows a flowchart of the procedure for creating a GNN. [Figure 15] Figure 15 shows a conceptual diagram of the GNN and user representation. [Figure 16] FIG. 16 is a diagram for explaining the prospective user prediction process according to the second embodiment. [Figure 17] FIG. 17 shows a flowchart of the process executed by the information processing device 10 according to the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0020] Hereinafter, with reference to the attached drawings, an embodiment for carrying out the present invention will be described in detail. Among the components disclosed below, those having the same functions are given the same reference numerals, and their description will be omitted. Note that the embodiment disclosed below is an example of a means for realizing the present invention, and should be appropriately modified or changed depending on the configuration of the device to which the present invention is applied and various conditions, and the present invention is not limited to the following embodiment. Furthermore, not all of the combinations of features described in the present embodiment are necessarily essential to the solution of the present invention.

[0021] (First embodiment) [Information processing system configuration] Fig. 1 shows an example of the configuration of an information processing system according to this embodiment. As an example, as shown in Fig. 1, this information processing system includes an information processing device 10 and multiple user devices 11-1 to 11-N (N>1) used by multiple arbitrary users 1 to N. In the following description, unless otherwise specified, user devices 11-1 to 11-N may be collectively referred to as user device 11. In the following description, the terms user device and user may be used synonymously.

[0022] The user device 11 is, for example, a device such as a smartphone or a tablet, and is configured to be able to communicate with the information processing device 10 via a public network such as LTE (Long Term Evolution) or a wireless communication network such as a wireless LAN (Local Area Network). The user device 11 has a display unit (display surface) such as a liquid crystal display, and each user can perform various operations using a GUI (Graphic User Interface) equipped on the liquid crystal display. The operations include various operations on content such as images displayed on the screen, such as tapping, sliding, and scrolling using a finger or a stylus. Note that user device 11 is not limited to the device shown in Fig. 1, and may be a device such as a desktop PC (Personal Computer) or a notebook PC. In this case, each user may perform operations using input devices such as a mouse and a keyboard. User device 11 may also be provided with a separate display screen.

[0023] The user device 11 can log in to a web service (internet-related service) provided from the information processing device 10 or from another device (not shown) via the information processing device 10, and can use the service. The web service can include an online mall, an online supermarket, or services related to communications, finance, real estate, sports, and travel, which are provided via the Internet. By using such a web service, the user device 11 can transmit information about the user of the user device 11 to the information processing device 10.

[0024] For example, the user device 11 can transmit to the information processing device 10 information on characteristics of the user device or the user, such as the IP (Internet Protocol) address of the user device 11, the user's address, and the user's name. In addition, the user device 11 performs positioning calculations based on signals received from GPS (Global Positioning System) satellites (not shown), generates the information obtained by the calculations as location information of the user device 11, and transmits it to the information processing device 10. The information processing device 10 acquires various information from the user device 11, acquires features related to items from a predetermined database, and creates a graph network (graph) based on the information. In this embodiment, the information processing device 10 creates a knowledge graph as a graph network. A knowledge graph is a directed graph that expresses knowledge in the real world in a structured fact structure. In this embodiment, the knowledge graph is composed of an inter-user relationship (interaction) graph, an inter-item relationship graph, and a user-item relationship graph. Then, the information processing device 10 extracts a user expression (user feature vector, embedded expression, vector expression in a directed graph such as a knowledge graph) for an arbitrary user from the knowledge graph.

[0025] [Functional configuration of information processing device 10] The information processing device 10 according to the present embodiment first acquires various user characteristics from the user devices 11-1 to 11-N, and also acquires characteristics related to items from a predetermined database. In the present embodiment, an item can be a tangible or intangible thing (Thing) that can be provided for various services. For example, for financial (fintech) services, there are items such as bank accounts, financial products such as stocks, investment trusts, and insurance products, cryptocurrency, and smartphone app payments. For digital content services, there are items such as video content such as movies and animations, and still image content such as photos, illustrations, and text. For e-commerce services, there are items such as intangible or tangible products handled in online shopping. For travel services, there are items such as information and reservations related to hotels, package tours, and transportation facilities. For mobile services, there are items such as mobile devices, public network / Internet connections, and communication usage fees. For advertising and media services, there are items such as offline and offline advertising products, direct mail, and advertising via broadcasting and the Internet. For card services, there are items such as credit card payments and point transactions. In addition, with regard to sports and cultural services, there are events such as sporting events and concerts, and items such as merchandise sold at the events.

[0026] The information processing device 10 constructs a knowledge graph from the acquired characteristic user features and items, and extracts a user expression for an arbitrary user from the knowledge graph. The knowledge graph is composed of a user-to-user relationship graph, an item-to-item relationship graph, and a user-item relationship graph, as described later. Furthermore, the information processing device 10 uses the user expression to predict a prospective user who has similar characteristics to a predetermined (given) seed user (model user) (e.g., who is likely to purchase a predetermined item with the seed user).

[0027] FIG. 2 shows an example of a functional configuration of the information processing device 10 according to the present embodiment. The information processing device 10 shown in Fig. 2 includes a user feature acquisition unit 101, an item feature acquisition unit 102, a graph construction unit 103, an expression extraction unit 104, a potential user prediction unit 105, a learning unit 106, an output unit 107, a learning model storage unit 110, and a feature storage unit 120. The learning model storage unit 110 stores a potential user prediction model 111 and a score prediction model 112. The various learning models will be described later. In addition, the feature storage unit 120 is configured to store a user feature 121 and an item feature 122.

[0028] The user feature acquisition unit 101 acquires factual features (factual information) (hereinafter, user features) about each of the user devices 11-1 to 11-N. The user features are factual features (information) that are actually or objectively obtained from the user device or user. The user feature acquisition unit 101 can acquire the user features directly from the user device 11, for example. The user feature acquisition unit 101 can also acquire the user features as information registered in a predetermined web service by the user of the user device 11.

[0029] The user characteristics include the IP address of the user device, the user's address and name, the number of the credit card held by the user, and the user's demographic information (demographic user attributes such as gender, age, residential area, occupation, and family structure). The user characteristics may also include a registration number and a registered name when using a specified web service. The user characteristics may also include a call history, a delivery address other than the user's address for the product when using a specified web service, a usage status when using a specified web service, a usage history (including a purchase history and a sales history), a search history, a browsing history (including a click history), and information on points that can be accumulated by using a service. In this way, the user characteristics can include any information, including information related to the user device or the user himself / herself, and information on the use of a specified service via communication. The user characteristic acquiring unit 101 stores the acquired user characteristic in the characteristic storage unit 120 as a user characteristic 121 .

[0030] The item feature acquisition unit 102 acquires item features (attributes) based on registration information and transaction history in various web services from a predetermined database (not shown). The item features include information identifying the item (hereinafter, item ID), information identifying the genre (higher classification) of the item (hereinafter, genre ID), information identifying the shop where the item is sold (hereinafter, shop ID), and the like. The item features can also include transaction information (number of transactions, etc.) between the item ID and genre ID, and between the item ID and shop ID, depending on the transaction history. The item feature acquisition unit 102 stores the acquired item features in the feature storage unit 120 as item features 122.

[0031] The graph construction unit 103 constructs a knowledge graph based on the various features acquired from the user feature acquisition unit 101 and the item feature acquisition unit 102. The knowledge graph will be described later.

[0032] The expression extraction unit 104 extracts a user expression for any user from the knowledge graph constructed by the graph construction unit 103. The expression extraction unit 104 may also extract an item expression for any item from the knowledge graph. The process of extracting user expressions (or item expressions) will be described later. The expression extraction unit 104 may also extract an embedded expression (vector expression) for any node in the constructed knowledge graph, for example, as a shop expression or a genre expression.

[0033] The prospective user prediction unit 105 predicts a user who is predicted to have similar characteristics to a predetermined seed user as a prospective user (similar user). The seed user is one or more users who have purchased and / or used a given product or service through a web service and / or positively evaluated the product or service through the web service. The seed user is one or more users selected and set from the user devices 11-1 to 11-N. The seed user can be selected and set by the prospective user prediction unit 105. In addition, the seed user may be set by an input operation by an operator using an input unit (input unit 205 in FIG. 12), may be set in advance in the system, or may be set by an arbitrary program stored in a storage unit (ROM 202 or RAM 203 in FIG. 11). In this embodiment, the prospective user is predicted using a prospective user prediction model 111 that has been learned by the learning unit 106. The prediction process of the prospective user will be described later.

[0034] The learning unit 106 trains the potential user prediction model 111 and the score prediction model 112, and stores the trained potential user prediction model 111 and score prediction model 112 in the learning model storage unit 110. The learning process of each learning model will be described later.

[0035] The output unit 107 outputs the user expressions extracted by the expression extraction unit 104 and the information of the prospective users predicted by the prospective user prediction unit 105. The output may be any output process, may be an output to an external device via a communication I / F (communication I / F 207 in FIG. 11), or may be a display on a display unit (display unit 206 in FIG. 11).

[0036] [Procedure for constructing knowledge graph] Next, the procedure for creating the knowledge graph according to this embodiment will be described. The knowledge graph is composed of a user relationship graph, an item relationship graph, and a user-item relationship graph. FIG. 3 shows a flowchart of the procedure for constructing the knowledge graph executed by the graph construction unit 103 according to this embodiment. Hereinafter, first, the creation procedures for the user relationship graph, the item relationship graph, and the user-item relationship graph (corresponding to the process of S30 in FIG. 3) will be described.

[0037] (1) Procedure for creating the user relationship graph The procedure for creating the user relationship graph will be described. In the following description, users A to E are users referred to for the purpose of explanation and can be users of the user device 11. The user relationship graph is composed of the connections of each user node (node having user identification information) surrounded by circles in FIGS. 4A and 4B. In the following description, the user node will be simply referred to as a user. Hereinafter, each step of the process of S30 in FIG. 3 regarding the user relationship graph will be described.

[0038] <S31: Creation of links> In S31, the graph construction unit 103 predicts and creates links between a plurality of users. The link creation process will be described with reference to FIGS. 4A and 4B. FIG. 4A is a diagram for explaining an explicit link, and FIG. 4B is a diagram for explaining an implicit link. An explicit link is a link created based on an explicit common feature between two users (user pair). An implicit link is a link created as an indirect relationship using an already created explicit link, although the existence of an explicit common feature of the user pair is unclear. Thus, the links between users are identified by explicit links and implicit links.

[0039] Fig. 4A shows an example of creating an explicit link using the IP addresses of the users' user devices as a common feature. Fig. 4A shows an example in which there are an online mall 41, a golf course reservation service 42, a travel-related reservation service 43, and a card management system 44 as web services available to users A to C. Fig. 4A shows these four web services, but the number of web services is not limited to a specific number.

[0040] The online mall 41 is a shopping mall available online (using the Internet). The online mall 41 can provide a wide variety of products and services, such as fashion, books, food, concert tickets, and real estate. The golf course reservation service 42 is operated as a website that provides online services related to golf courses, and can provide, for example, golf course search and reservation and lesson information. The travel-related reservation service 43 is operated as a website that provides various travel services available online. For example, the travel-related reservation service 43 can provide hotel and travel tour reservations, airline and rental car reservations, tourist information, and information about hotels and the surrounding areas. The card management system 44 is operated on a website that provides services related to credit cards issued and managed by a specific card management company. The card management system 44 may provide services in association with at least one of an online mall 41, a golf course reservation service 42, and a travel-related reservation service 43.

[0041] 4A, users A to C each use the same IP address (=198.45.66.xx) to access an online mall 41, a golf course reservation service 42, and a travel-related reservation service 43. The IP address information can be acquired by the user characteristic acquisition unit 101. In such a case, as shown in the link state 45, the graph construction unit 103 causes users A to C to create explicit links (for example, link L1 between user A and user C) with each other based on the feature of the same IP address. The explicit links are represented by solid lines.

[0042] In addition to FIG. 4A, as common features, features such as the feature of the user's address and the feature of the credit card number used by the user can be used as common features to create explicit links.

[0043] FIG. 4B shows an example of creating an implicit link between users. In the example of FIG. 4B, for user A, users C, D, and E are connected by explicit links, and for user B, users C, D, and E are connected by explicit links. Such link features (features indicating the relationship between links) are embedded into a common feature space, and links inferred as having an implicit relationship constructed between each user (each node) are created (established) as implicit links. In the example of FIG. 4B, although user A and user B are not connected by an explicit link, as a result of being inferred to have a relationship in the common feature space, an implicit link L2 shown by a dashed line is created. Note that the graph construction unit 103 predicts and creates an implicit link between users by performing learning (representation learning, relational learning, embedding learning, knowledge graph embedding) of a user relationship graph composed of nodes (users) connected by explicit links. At this time, the graph construction unit 103 may perform the learning based on a known embedding model or an extension thereof as appropriate.

[0044] <S32: Inference of Relationship between Links> In S32, the graph construction unit 103 infers the relationship between the links predicted and created in S31. The inference process of the relationship between links will be described with reference to FIG. 5. FIG. 5 is a diagram for explaining the inference process of the relationship between links, and shows an example of inferring the relationship of a link between user A and user B connected by an explicit link.

[0045] The graph construction unit 103 treats pairs of users connected by the links created in S31 as data points, and groups the pairs (data points) into clusters representing a common type using various information acquired by the user feature acquisition unit 101. The various information may be information such as IP address, address, credit card, age, gender, and friends. Also, each cluster may be a cluster having relationships such as spouse, parent-child, neighbor, same household, colleague, friend, same-sex siblings, opposite-sex siblings, etc. In the example of FIG. 5, pairs of users are indicated by crosses, and as clusters into which the pairs can be grouped, a parent-child cluster 51, a spouse cluster 52, a same-sex sibling cluster 53, a friend cluster 54, and a colleague cluster are shown. Note that although FIG. 5 shows five clusters, the number of clusters is not limited to a specific number.

[0046] For example, if user A and user B have (share) the feature 50 of having the same surname, an age difference of less than 10 years, opposite genders, and the same address, the graph construction unit 103 can group the pair of user A and user B into a cluster (spouse cluster 52) representing the relationship of husband and wife (spouse).

[0047] <S33: Score Assignment Based on Proximity of Relationships> In S33, the graph construction unit 103 predicts a score based on the proximity of the relationship for the pairs inferred in S32, and assigns the score to the pairs. In the present embodiment, the score is a numerical value between 0 and 1, but there is no specific limitation on the numerical values that the score can take. FIG. 6A shows a conceptual diagram of a score (hereinafter, proximity score) based on the proximity of the relationship of user pairs.

[0048] In the example of FIG. 6A, the closeness of the relationship between the user pair changes depending on the characteristics owned (shared) by user A and user B connected by an explicit link. In the upper part of FIG. 6A, if the relationship between user A and user B has characteristics 60 of same-gender siblings, same address, 1200 call histories, and 50 gift exchanges, the closeness of the relationship between the user pair (i.e., closeness score) is high. On the other hand, in the lower part of FIG. 6A, if the relationship between user A and user B has characteristics 61 of same-gender siblings, different addresses, 30 call histories, and 2 gift exchanges, the closeness of the relationship between the user pair (i.e., closeness score) is low. Thus, as in the example of FIG. 6A, even if user A and user B are siblings of the same gender, the closeness of the relationship between the pair differs depending on other characteristics shared by the pair of users. It is observed that pairs with high relationship closeness have a close social distance from each other and have high influence. On the other hand, pairs with low relationship closeness are observed to have a large social distance from each other and are not closely related.

[0049] In this embodiment, the closeness score for a user pair is predicted using a score prediction model 112. Fig. 6B shows a schematic architecture of the score prediction model 112. The score prediction model 112 is a learning model that predicts a closeness score 64 for a feature 63 of a user pair using the feature 63 as an input.

[0050] The score prediction model 112 is, for example, a learning model that performs weak supervised learning, for example, a learning model using a convolutional neural network (CNN). In this embodiment, the score prediction model 112 is a learning model that is learned using the closeness scores (0 to 1) assigned to a plurality of features for a user pair as teacher data, as shown in FIG. 6A. For example, in the learning stage, combined data such as a closeness score close to 1 set for feature 60 in FIG. 6A and a closeness score close to 0 set for feature 61 is used as teacher data. The learning process is performed by the learning unit 108.

[0051] In this embodiment, the proximity score for a user pair is predicted using the score prediction model 112. However, the graph construction unit 103 may be configured to predict the score by other methods.

[0052] Through the above processing, explicit or implicit links are formed among multiple users, a proximity score is assigned to each link, and a user relationship graph is created. A conceptual diagram of the user relationship graph is shown in FIG. 6C. The proximity score predicted as described above is assigned to each pair of users.

[0053] (2) Procedure for creating an item relationship graph Next, the procedure for creating an item relationship graph will be described. Similar to the procedure for creating a user relationship graph, the graph construction unit 103 creates an item relationship graph according to the flowchart of the relationship graph creation process in S30 in FIG. 3. Note that in the creation of the item relationship graph, the step of S32 is not performed.

[0054] <S31: Creation of links> In S31, the graph construction unit 103 creates links among multiple items based on the item features 122 stored in the feature storage unit 120. As described above, the item features according to this embodiment include an item ID, a genre ID, and a shop ID. That is, at least one genre ID and / or shop ID is associated with one item ID. Note that the genre ID and shop ID may each be hierarchically configured. For example, the genres may be hierarchically configured, and each may have a genre ID. Further, the item features are not limited to the item ID, genre ID, and shop ID, and may include other information (attributes) such as the brand, color, and property-related information of the item.

[0055] The graph construction unit 103 links the genre ID and shop ID associated with any item ID to the item ID. Fig. 7 shows a conceptual diagram of the inter-item relationship graph. The inter-item relationship graph is composed of the connections of each node indicating the item ID, genre ID, or shop ID enclosed by a circle in Fig. 7. In the following description, the node will be simply referred to as an item, genre, or shop.

[0056] In Fig. 7, the link connected from the item feature is an explicit link, which is shown by a solid line (for example, the link L1 between item A and shop A). Also, when the colors and properties between items are similar (when the similarity degree is higher than a predetermined threshold), the items can be connected by an explicit link (for example, the link between item A and item B). On the other hand, for example, there may be cases where different items are sold in the same shop. In Fig. 7, since both item A and item C are sold in shop A, they are linked to shop A. From this, the graph construction unit 103 can connect item A and item C with an implicit link shown by a dashed line (in the example of Fig. 7, the link L2 between item A and item C). Note that the graph construction unit 103 predicts and creates the implicit link between items by performing learning (representation learning, relational learning, embedding learning, knowledge graph embedding) of the inter-item relationship graph composed of nodes (users) connected by explicit links. At this time, the graph construction unit 103 may perform the learning based on a known embedding model or its extension as appropriate.

[0057] <S33: Score Assignment Based on Proximity of Relationships> In S33, the graph construction unit 103 predicts a score (proximity score) based on the proximity of the relationship for each pair in the link created in S31, and assigns the score to the pair. In FIG. 7, for example, in genre A, if item A sells better (is traded more frequently) than item B, a high score is assigned to the pair of item A and genre A. Also, for item A, if it is determined that the probability of belonging to genre A is high, a score based on the probability is highly assigned to the pair of item A and genre A. Further, if the similarity between any items is higher, a high score is assigned to the pair of the items.

[0058] The graph construction unit 103 may predict the proximity score for each pair using the above-described score prediction model 112. When using the score prediction model 112, the number of transactions for any pair among items, genres, and shops, and the proximity score (0 to 1) assigned to the characteristics of the similarity between items are learned as teacher data (see FIG. 6B). For example, in the learning stage, as teacher data, combined data such as a proximity score close to 1 set for the characteristic of a large number of transactions or high similarity, and a proximity score close to 0 set for the characteristic of a small number of transactions or low similarity is used. The learning process is performed by the learning unit 108. By score assignment, a score can be represented as a numerical value for each pair, similar to the user relationship graph shown in FIG. 6C.

[0059] (2) Procedure for creating user-item relationship graph Next, the procedure for creating the user-item relationship graph will be described. Similar to the procedure for creating the user relationship graph, the graph construction unit 103 creates a user-item relationship graph according to the flowchart of the S30 relationship graph creation process in FIG. 3. Note that in the creation of the user-item relationship graph, the step of S32 is not performed.

[0060] <S31: Creation of link> In S31, the graph construction unit 103 creates a link between an arbitrary user and one or more items based on the user characteristics 121 stored in the feature memory unit 120. First, the graph construction unit 103 obtains user characteristics related to items for each user, such as the purchase history, search history, or browsing history (including click history) of each user, from the user characteristics 121 stored in the feature memory unit 120. The graph construction unit 103 creates a user-item relationship graph for each user using the user characteristics related to the item for each user.

[0061] FIG. 8 shows a conceptual diagram of a user-item relationship graph. The item relationship graph is composed of connections between user nodes surrounded by circles and nodes indicating item IDs in FIG. 7. In the following description, the user nodes and the nodes indicating item IDs are simply referred to as users and items.

[0062] In FIG. 8, from the click history of item A by user A, user A and item A are linked. The click history of the item A includes clicking on a predetermined screen area (e.g., button, photo, icon, etc.) related to the item A. Also, from the purchase history of item B by user A, user A and item B are linked. Further, from the sales history of item C by user A, user A and item C are linked. In FIG. 8, although user A and each item are connected by explicit links, an indirect connection from user A to any item may be represented by an implicit link. Also, a user and an item may be linked from the advertisement delivery history related to the item to the user. Note that the graph construction unit 103 predicts and creates an implicit link between a user and an item by performing learning (representation learning, relational learning, embedding learning, knowledge graph embedding) of a user-item relationship graph composed of nodes (users) connected by explicit links. At this time, the graph construction unit 103 may perform the learning based on a known embedding model or an extension thereof as appropriate.

[0063] <S33: Score Assignment Based on Proximity of Relationships> In S33, the graph construction unit 103 predicts a score (closeness score) based on the closeness of the relationship for each pair in the link created in S31, and assigns the score to the pair. The graph construction unit 103 may predict a closeness score for a pair of a user and an item based on history related to item transactions, such as the above-mentioned click history, purchase history, sales history, and advertisement delivery history. For example, it can be said that the distance between a user and an item is closer when the user actually purchases an item than when the user only clicks on the item. Therefore, in the example of FIG. 8, a higher score is assigned to the pair of user A and item B than to the pair of user A and item A. In addition, a user selling an item is considered to be closer in relationship than a click action, and in the example of FIG. 8, a higher score is assigned to the pair of user A and item C than to the pair of user A and item A. In addition, for example, a low score may be assigned to the pair of user and item when the user does not click or purchase the item even though advertisements related to the item have been delivered to the user multiple times.

[0064] The graph construction unit 103 may predict the closeness score for each pair using the score prediction model 112 described above. When the score prediction model 112 is used, the closeness scores (0 to 1) assigned to the features of the user and item pair are used as training data for learning (see FIG. 6B). For example, in the learning stage, combined data such as a closeness score close to 1 set for the feature of purchase and a closeness score close to 0 set for the feature of click are used as training data. The learning process is performed by the learning unit 108. By assigning scores, the scores can be expressed as numerical values ​​for each pair, similar to the user relationship graph shown in FIG. 6C.

[0065] In Figure 8, the user-item relationship graph is defined as a pair of users and items, but if a user has a history of clicking, purchasing, or selling a specific item, the pair of users and genres (genre IDs) is linked. Then, a closeness score is assigned to the features shared by the pair.

[0066] (4) Procedure for constructing the entire knowledge graph After the user-to-user relationship graph, user-to-item relationship graph, and item-to-item relationship graph are created by the above steps (1) to (3), in S34 of FIG. 3, the graph construction unit 103 connects these graphs to construct (create) the entire knowledge graph. A conceptual diagram of the knowledge graph (knowledge graph 90) is shown in the upper part of FIG. 9. The graph construction unit 103 uses common nodes as connection points in the created user-to-user relationship graph, user-to-item relationship graph, and item-to-item relationship graph to connect all the graphs. Next, the graph construction unit 103 organizes the graphs by deleting duplicate links, etc., and constructs the knowledge graph. Although not shown in FIG. 9, a closeness score indicating the closeness between the nodes is shown between each node (user, item, genre, etc.). The closeness score may be represented by the length of the arrow. Also, although not shown in FIG. 9, each user includes a user feature, or the user feature is connected as a node. Note that nodes such as users, items, and genres correspond to entities (head entities or tail entities) in the knowledge graph, and pairs and links correspond to relations.

[0067] The graph construction unit 103 may construct the entire knowledge graph by linking the fact-based user-to-user relationship graph, the item-to-item relationship graph, and the user-to-item relationship graph, each of which is composed of nodes connected by explicit links and explicit links. The graph construction unit 103 may also construct the entire knowledge graph by linking some of the fact-based user relationship graph, the item-to-item relationship graph, and the user-to-item relationship graph, each of which is composed of nodes connected by explicit links and explicit links, and the remaining of the user-to-user relationship graph, the item-to-item relationship graph, and the user-to-item relationship graph, each of which is composed of nodes connected by explicit links and explicit links and nodes connected by implicit links and implicit links. In other words, the entire knowledge graph may not include at least one implicit link of the user-to-user relationship graph, the item-to-item relationship graph, and the user-to-item relationship graph. The graph construction unit 103 may also predict and create an implicit link between nodes by embedding the created knowledge graph into a feature space (vector space). That is, the graph construction unit 103 may predict and create implicit links from a knowledge graph that is composed of only explicit links. Note that extraction (acquisition) of embedded expressions (vector expressions) of entities such as user expressions may be performed based on the entire knowledge graph that does not include at least one implicit link in the user-to-user relationship graph, the item-to-item relationship graph, and the user-to-item relationship graph.

[0068] [User expression extraction procedure] Next, a procedure for extracting user expressions according to this embodiment will be described. The expression extraction unit 104 extracts user expressions for any user from the knowledge graph constructed by the graph construction unit 103. Specifically, the expression extraction unit 104 embeds the knowledge graph in a feature space (vector space) and learns the embedded expressions (vector expressions) of each node (entity) and each link (relation) in the feature space. The expression extraction unit 104 performs knowledge graph learning (representation learning, relation learning, embedding learning) and extracts (acquires) the embedded expressions (low-dimensional vector expressions) of any user as user expressions (user feature vectors).

[0069] The expression extraction unit 104 may employ a Translation-based model that performs learning based on the distance between vector expressions of entities such as TransE, TransD, and RotatE as an embedding model used for learning (embedding) of a knowledge graph. The expression extraction unit 104 may also employ an Embedding-projection model that performs learning by mapping vector expressions of entities such as TransH, TransR, and STransE to a vector space that differs for each relation. The expression extraction unit 104 may also employ a model that performs learning by using a conversion of vector expressions to a complex space such as ComplEx. The expression extraction unit 104 may also employ a model that performs learning by using a convolutional neural network (NN) including NN such as ConvE, ConvR, and R-GCN. The expression extraction unit 104 may also employ a model that uses an attention mechanism such as Knowledge Graph Attention Network (KGAT), and may appropriately employ a known model such as TorusE or an extension thereof. The expression extraction unit 104 may use regularization (such as L1 regularization or L2 regularization) as appropriate when learning the knowledge graph.

[0070] The user expression may reflect the characteristics of each node and the closeness score assigned between nodes (pairs, links). At this time, the expression extraction unit 104 may learn the vector expression of each node and each link (embedding in a knowledge graph) while weighting the vector expression of the link (relation) between the nodes based on the closeness score between the nodes. Also, the expression extraction unit 104 may focus on or limit the learning of the vector expression of each link whose closeness score exceeds or falls below a predetermined threshold and the node connected via the link, and extract the vector expression of an entity such as an arbitrary user. At this time, the node includes at least a node corresponding to an entity such as an arbitrary user. Also, the expression extraction unit 104 may focus on or limit the learning of the vector expression of each node and each link in the N-th order neighborhood (N>1) of the node corresponding to the entity such as an arbitrary user, and extract the vector expression of the entity such as an arbitrary user. In this way, it is possible to reduce the calculation load by screening the learning target according to the closeness score.

[0071] A conceptual diagram (user representation 91) of a user representation (node ​​representation) extracted for user A is shown at the bottom of FIG. 9. In the user representation 91, information on one or more users, items, genres, shops, etc. connected to user A by implicit or explicit links is reflected in one user representation. The user representation may also represent a closeness score of each node (item, genre, etc.) to user A. In other words, the user representation of user A may correspond to an adjacent representation that reflects the closeness of the relationship between user A and each node. The expression extraction unit 104 may extract an item expression (node ​​expression) for any item from the knowledge graph constructed by the graph construction unit 103.

[0072] [Prediction of potential users] Next, a process for predicting a potential user according to the present embodiment will be described. The potential user prediction model 111 is a learning model that predicts a user having similar characteristics to a seed user as a potential user. A seed user is a user who has purchased and / or used a given product or service through a web service and / or has positively evaluated the product or service through the web service.

[0073] The prospective user prediction model 111 is, for example, a learning model for machine learning based on XGBoost. In the learning stage, the learning unit 106 trains the prospective user prediction model 111 using the user features of a seed user (positive user), the user features of users other than the seed user (negative users), and the user expressions of these users. The user features are base user features and include a purchase history (information on the genre and type of products, etc.) in the web service. The demographic information and the purchase history may each include a plurality of subdivided features. Note that the user features are not limited to demographic information and purchase history, and may include other features such as a point status (available points, etc.), point features (information on point transactions such as points acquired / used from online or offline shops, etc.), etc.

[0074] The learning unit 106 verifies and tunes (adjusts) hyperparameters (parameters that control the behavior of the potential user prediction model 111) by grid search and cross validation. Since XGBoost is a tree (decision tree) based model, the potential user prediction model 111 can generate results that show how input data (user features) affect the model output. This makes it possible to verify, for example, which user features (combinations of subdivided features) are more influential for a seed user.

[0075] The trained prospective user prediction model 111 is configured to output, for any user, the possibility that the user has user characteristics similar to those of the seed user (prospective user possibility). The possibility is expressed, for example, as a numerical value from 0 to 1, with the maximum possibility being 1. Here, for example, if the threshold is set to 0.5, the prospective user prediction unit 105 can predict (determine) a user having a possibility greater than 0.5 as a prospective user (i.e., a potential user having characteristics similar to those of the seed user). Note that there may be a plurality of seed users, and in this case, the prospective user prediction unit 105 can predict, as a prospective user, a user having user characteristics similar to those of a user included in the plurality of seed users (seed user group (model user group)).

[0076] In this embodiment, the prospective user prediction unit 105 inputs the user expressions extracted by the expression extraction unit 105 and the base user features for an arbitrary user into the prospective user prediction model 111 to predict whether the arbitrary user is a prospective user. FIG. 10 shows a diagram for explaining the prospective user prediction process according to this embodiment. When the user A is set as an arbitrary user (target user), the prospective user prediction unit 105 inputs the base user features 1001 of the user A and the user expressions 91 extracted by the expression extraction unit 104 into the prospective user prediction model 111 to predict whether the user A is a prospective user. Specifically, the prospective user prediction unit 105 predicts and outputs the possibility that the user A has the same user features as the seed user from the user features 1001 and the user expressions 91 of the user A (the possibility 1002 of the prospective user).

[0077] [Hardware configuration of information processing device 10] FIG. 11 is a block diagram showing an example of a hardware configuration of an information processing device 10 according to the present embodiment. The information processing apparatus 10 according to the present embodiment can be implemented on any single or multiple computers, mobile devices, or any other processing platform. 11, an example in which the information processing device 10 is implemented in a single computer is shown, but the information processing device 10 according to the present embodiment may be implemented in a computer system including multiple computers. The multiple computers may be connected to each other so as to be able to communicate with each other via a wired or wireless network.

[0078] 11, the information processing device 10 may include a CPU 201, a ROM 202, a RAM 203, a HDD 204, an input unit 205, a display unit 206, a communication I / F 207, and a system bus 208. The information processing device 10 may also include an external memory. A CPU (Central Processing Unit) 201 generally controls the operations of the information processing device 10, and controls each of the components (202 to 207) via a system bus 208, which is a data transmission path.

[0079] The ROM (Read Only Memory) 202 is a non-volatile memory that stores control programs and the like necessary for the CPU 201 to execute processing. Note that the programs may be stored in a non-volatile memory such as a HDD (Hard Disk Drive) 204 or an SSD (Solid State Drive) or an external memory such as a removable storage medium (not shown). The RAM (Random Access Memory) 203 is a volatile memory and functions as a main memory, a work area, etc. of the CPU 201. That is, the CPU 201 loads necessary programs, etc. from the ROM 202 to the RAM 203 when executing processing, and realizes various functional operations by executing the programs, etc. The learning model storage unit 110 and the feature storage unit 102 shown in FIG. 2 can be configured by the RAM 203.

[0080] The HDD 204 stores, for example, various data and various information required when the CPU 201 performs processing using a program. The HDD 204 also stores, for example, various data and various information obtained when the CPU 201 performs processing using a program. The input unit 205 is composed of a keyboard and a pointing device such as a mouse. The display unit 206 is configured with a monitor such as a liquid crystal display (LCD), etc. The display unit 206 may be configured in combination with the input unit 205 to function as a GUI (Graphical User Interface).

[0081] The communication I / F 207 is an interface that controls communication between the information processing device 10 and an external device. The communication I / F 207 provides an interface with a network and executes communication with an external device via the network. Various data, various parameters, etc. are transmitted and received between the external device and the communication I / F 207. In this embodiment, the communication I / F 207 may execute communication via a wired LAN (Local Area Network) or a dedicated line conforming to a communication standard such as Ethernet (registered trademark). However, the network available in this embodiment is not limited to this, and may be configured as a wireless network. This wireless network includes wireless PANs (Personal Area Networks) such as Bluetooth (registered trademark), ZigBee (registered trademark), and UWB (Ultra Wide Band). It also includes wireless LANs (Local Area Networks) such as Wi-Fi (Wireless Fidelity) (registered trademark), and wireless MANs (Metropolitan Area Networks) such as WiMAX (registered trademark). It also includes wireless WANs (Wide Area Networks) such as LTE / 3G, 4G, and 5G. Note that the network only needs to connect each device so that they can communicate with each other, and the communication standard, scale, and configuration are not limited to those described above.

[0082] At least some of the functions of each element of the information processing device 10 shown in Fig. 2 can be realized by the CPU 201 executing a program. However, at least some of the functions of each element of the information processing device 10 shown in Fig. 2 may be operated as dedicated hardware. In this case, the dedicated hardware operates under the control of the CPU 201.

[0083] [Hardware configuration of user device 11] The hardware configuration of the user device 11 shown in Fig. 1 may be the same as that of Fig. 11. That is, the user device 10 may include a CPU 201, a ROM 202, a RAM 203, a HDD 204, an input unit 205, a display unit 206, a communication I / F 207, and a system bus 208. The user device 11 can display various information provided by the information processing device 10 on the display unit 206, and perform processing corresponding to an input operation received from a user via a GUI (configured by the input unit 205 and the display unit 206).

[0084] [Processing flow] Fig. 12 shows a flowchart of processing executed by the information processing device 10 according to this embodiment. The processing shown in Fig. 12 can be realized by the CPU 201 of the information processing device 10 loading a program stored in the ROM 202 or the like into the RAM 203 and executing it. For the explanation of Fig. 12, the information processing system shown in Fig. 1 will be referred to. It is assumed that the prospective user prediction model 111 and the score prediction model 112, which have been learned by the learning unit 105, are stored in the learning model storage unit 110.

[0085] In S1201, the user characteristic acquisition unit 101 acquires the user characteristics of each user from the user devices 11-1 to 11-N, and stores them in the characteristic storage unit 120 as user characteristics 121. In addition, the item characteristic acquisition unit 102 acquires item characteristics (attributes) based on registration information and transaction history in various web services from a predetermined database, and stores them in the characteristic storage unit 120 as item characteristics 122. The process of S1201 may be a process of acquiring (collecting) user characteristics and item characteristics for a certain period of time in the past.

[0086] In S1202, the graph constructing unit 103 constructs a knowledge graph. The procedure for constructing the knowledge graph is as described above with reference to FIG.

[0087] In S1203, the expression extraction unit 104 extracts user expressions for an arbitrary user from the knowledge graph created in S1202. The procedure for extracting the user expressions is as described above with reference to FIG.

[0088] In S1204, the prospective user prediction unit 105 predicts the possibility that a given user is a prospective user having similar characteristics to a given seed user. In this embodiment, as shown in Fig. 10, for the given user, the base user characteristics of the given user and the user expression of the given user are input to the prospective user prediction model 111, and the possibility that the given user has similar characteristics to the seed user (the possibility of the given user being a prospective user) is predicted and output.

[0089] The prospective user prediction unit 105 may also select and set a plurality of users (target user group) other than the seed user (or seed user group) from among the user devices 11-1 to 11-N. Then, the prospective user prediction unit 105 can predict prospective users having similar characteristics to the seed user as a prospective user group (similar user group) from the target user group. In this case, the prospective user prediction unit 105 inputs the user characteristics that are the base of each user in the target user group and the user expression of the user to the prospective user prediction model 111 to predict the prospective user group. In addition, when there are a plurality of seed users (seed user group), the prospective user prediction unit 105 can predict users having user characteristics similar to those of users included in the seed user group as a prospective user group from the target user group.

[0090] In S1205, the output unit 107 outputs the result of the possibility of the prospective user predicted in S1204. For example, when the processes of S1203 and S1204 are performed for a plurality of users, the output unit 107 can generate information on the information of a prospective user group predicted from the plurality of users and output the information to an external device (not shown).

[0091] In this way, the information processing device according to this embodiment creates a user-to-user relationship graph, an item-to-item relationship graph, and a user-item relationship graph from the user characteristics of multiple users and the characteristics of multiple items, and creates a knowledge graph using each of these relationship graphs. The knowledge graph links all other users and items (including genres, shops, etc.) related to any given user, and is expected to be utilized in platform businesses.

[0092] Furthermore, the information processing device according to this embodiment creates a user representation for an arbitrary user, which represents all other users and items related to the arbitrary user, from the knowledge graph. The user representation is a representation in which the characteristics (attributes) of all other users and items are linked to one user. In other words, the user representation is not an individual representation for each other user or item, but is a representation that includes composite relationships with other users and items for an arbitrary user. This allows the connections of an arbitrary user with other users and items to be treated as one representation, and when the user representation is used for arbitrary prediction processing for a platform business, the amount of computation can be advantageously reduced.

[0093] Furthermore, the information processing device according to this embodiment uses the user expression and user features of an arbitrary user to predict whether the user has similar features to a predetermined seed user (possibility of the user being a prospective user). By using the user expression as well as the user features, prediction can be made with more consideration given to the user's preferences, and the accuracy of the prediction can be improved. Furthermore, the information on prospective users obtained by the prediction enables more targeted marketing.

[0094] Second embodiment Next, a second embodiment will be described. The information processing system according to this embodiment can be configured as shown in Fig. 1, similarly to the first embodiment. Below, differences from the first embodiment will be described, and descriptions of similar configurations and features will be omitted.

[0095] [Functional configuration of information processing device 10] The information processing device 10 according to this embodiment first acquires various user features from the user devices 11-1 to 11-N, and also acquires features related to items from a predetermined database. The information processing device 10 then creates a graph neural network (GNN) as a graph network (graph) from the acquired features, and extracts a user expression for an arbitrary user from the GNN. Furthermore, the information processing device 10 uses the user expression to predict a prospective user who has similar user features to a predetermined seed user (e.g., who is likely to purchase a predetermined item with the seed user).

[0096] FIG. 13 shows an example of the functional configuration of the information processing device 10 according to the present embodiment. 13 includes a user feature acquisition unit 101, an item feature acquisition unit 102, a GNN construction unit 1301, an expression extraction unit 1302, a potential user prediction unit 1303, a learning unit 1304, an output unit 107, a learning model storage unit 110, and a feature storage unit 120. The learning model storage unit 110 stores a potential user prediction model 113 and a score prediction model 112. The feature storage unit 120 is configured to store a user feature 121 and an item feature 122.

[0097] The GNN construction unit 1301 creates a GNN based on user features, item features, and tasks. The construction procedure of the GNN will be described later. The expression extraction unit 1302 extracts a user expression for an arbitrary user from the GNN constructed by the GNN construction unit 1301. The expression extraction unit 1302 may also extract an item expression for an arbitrary item from the GNN. The extraction process of the user expression (or item expression) will be described later.

[0098] The prospective user prediction unit 1303 predicts a group of users who are predicted to have similar characteristics to a predetermined seed user as prospective users. In this embodiment, the prospective users are predicted using the prospective user prediction model 113 that has been trained by the training unit 106. The prediction process of the prospective users will be described later.

[0099] [Procedure for constructing GNN] Next, the procedure for constructing a GNN according to this embodiment will be described. Fig. 14 shows a flowchart of the procedure for constructing a knowledge graph executed by the GNN construction unit 1301 according to this embodiment. First, in S1401, the GNN construction unit 1301 creates an inter-user relationship graph according to the procedure for creating an inter-user relationship graph described in the first embodiment. The GNN construction unit 1301 also creates an inter-item relationship graph according to the procedure for creating an inter-item relationship graph described in the first embodiment.

[0100] In S1402, the GNN construction unit 1301 creates a relationship graph (user-item relationship graph) between multiple users (user nodes) and multiple items (item nodes) based on the user features 121 stored in the feature storage unit 120. Specifically, the GNN construction unit 1301 first acquires user features related to items for each user, such as the purchase history, search history, or browsing history (including click history) of each user, from the user features 121 stored in the feature storage unit 120. Next, the GNN construction unit 1301 creates a user-item relationship graph using the user features related to the items for each user. For example, the GNN construction unit 1301 forms edges (relationships between nodes) for one or more items for which a purchase history, search history, or browsing history exists for any user.

[0101] Next, in S1403, the GNN construction unit 1301 sets (defines) a plurality of tasks (task nodes) representing classifications of a plurality of items included in the user-item relationship graph, and adds them to the graph. The task is one classification targeted in the market, and may be information indicating, for example, the brand name of the item, the name of the region where the item is sold, or the name of the agency that handles the item. In S1404, the GNN construction unit 1301 trains the GNN, organizes the edges between the nodes, and constructs the GNN.

[0102] A conceptual diagram of a GNN (GNN150) is shown at the top of Fig. 15. In the GNN150, in addition to edges between a plurality of user nodes and a plurality of item nodes, edges between a plurality of task nodes are added. The GNN construction unit 1301 acquires relationships between the plurality of user nodes and the plurality of task nodes and relationships between the plurality of item nodes and the plurality of task nodes from user features and item features, and causes the GNN to learn using the acquired relationships.

[0103] [User expression extraction procedure] Next, the procedure for extracting user expressions according to this embodiment will be described. The expression extraction unit 1302 extracts a user expression for any user from the GNN constructed by the GNN construction unit 1301. Specifically, the expression extraction unit 1302 embeds the GNN in a common feature space (vector space) and learns the embedded expressions (vector expressions) of each node (entity) and each edge (relation) in the feature space. It performs learning (representation learning, relation learning, embedding learning) of the graph network indicated by the GNN, and extracts (obtains) the embedded expression (low-dimensional vector expression) of any user as the user expression (user feature vector).

[0104] 15 shows a conceptual diagram (user representation 151) of a user representation extracted for user A. In user representation 151, in addition to information on one or more users, items, genres, shops, etc. having a relationship with user A, information on a task (e.g., a brand name) is reflected in one user representation. The expression extraction unit 104 may extract an item expression for an arbitrary item from the GNN constructed by the GNN construction unit 1301.

[0105] [Prediction of potential users] Next, a process for predicting a potential user according to the present embodiment will be described. The potential user prediction model 113 is a learning model that predicts a user having similar characteristics to a seed user as a potential user. A seed user is a user who has purchased and / or used a given product or service through a web service and / or has positively evaluated the product or service through the web service.

[0106] The prospective user prediction model 113 is, for example, a learning model based on XGBoost. In the learning stage, the learning unit 106 trains the prospective user prediction model 113 using the user features of the seed user, the user features of users other than the seed user (negative users), and the user expressions of these users. The user features may be base user features (demographic information and purchase history) as in the first embodiment.

[0107] The learning unit 106 verifies and tunes (adjusts) hyperparameters (parameters that control the behavior of the potential user prediction model 113) by grid search and cross validation. Since XGBoost is a tree (decision tree) based model, the potential user prediction model 113 can generate results that show how input data (user features) affect the model output. This makes it possible to verify, for example, which user features (combinations of subdivided features) are more influenced by the seed user.

[0108] The trained potential user prediction model 113 is configured to output, for any user, the possibility that the user has the same user characteristics as the seed user (potential of the potential user). The possibility is expressed, for example, as a numerical value from 0 to 1, with the maximum possibility being 1. Here, for example, if the threshold is set to 0.5, the potential user prediction unit 1303 can predict (determine) a user having a possibility greater than 0.5 as a potential user (i.e., a potential user having the same characteristics as the seed user).

[0109] In this embodiment, the prospective user prediction unit 1303 inputs the user expressions extracted by the expression extraction unit 1302 and the base user features for an arbitrary user into the prospective user prediction model 113 to predict whether the arbitrary user is a prospective user. FIG. 16 shows a diagram for explaining the prospective user prediction process according to this embodiment. When the user A is set as an arbitrary user (target user), the prospective user prediction unit 1303 inputs the base user features 1601 of the user A and the user expressions 151 extracted by the expression extraction unit 1302 into the prospective user prediction model 113 to predict whether the user A is a prospective user. Specifically, the prospective user prediction unit 1303 predicts and outputs the possibility that the user A has the same user features as the seed user from the user features 1601 and the user expressions 151 of the user A (the possibility of the prospective user 1602).

[0110] [Processing flow] Fig. 17 shows a flowchart of processing executed by the information processing device 10 according to this embodiment. The processing shown in Fig. 17 can be realized by the CPU 201 of the information processing device 10 loading a program stored in the ROM 202 or the like into the RAM 203 and executing it. For the explanation of Fig. 17, the information processing system shown in Fig. 1 will be referred to. It is assumed that the prospective user prediction model 113 and the score prediction model 112, which have been learned by the learning unit 105, are stored in the learning model storage unit 110.

[0111] The process of S1701 is the same as the process of S1201 in Fig. 12. In S1702, the GNN constructing unit 1301 constructs a GNN. The procedure for constructing the GNN is as described above with reference to Fig. 14.

[0112] In S1703, the expression extraction unit 1302 extracts a user expression for an arbitrary user from the GNN created in S1702. The procedure for extracting the user expression is as described above with reference to FIG.

[0113] In S1704, the potential user prediction unit 1303 predicts the possibility that a given user is a potential user having similar characteristics to a given seed user. In this embodiment, as shown in Fig. 16, for the given user, the base user characteristics of the given user and the user expression of the given user are input to the potential user prediction model 113, and the possibility that the given user has similar characteristics to the seed user (the possibility of the given user being a potential user) is predicted and output.

[0114] The prospective user prediction unit 1303 may also select and set a plurality of users (target user group) other than the seed user (or seed user group) from among the user devices 11-1 to 11-N. Then, the prospective user prediction unit 1303 can predict a prospective user having similar characteristics to the seed user from the target user group as a prospective user group (similar user group). In this case, the prospective user prediction unit 1303 inputs the user characteristics that are the base of each user in the target user group and the user expression of the user to the prospective user prediction model 113 to predict the prospective user group. Also, when there are a plurality of seed users (seed user group), the prospective user prediction unit 1303 can predict a user having user characteristics similar to those of a user included in the seed user group from the target user group as a prospective user group.

[0115] In S1705, the output unit 107 outputs the result of the possibility of the prospective user predicted in S1704. For example, when the processes of S1703 and S1704 are performed for a plurality of users, the output unit 107 can generate information on the information of a prospective user group predicted from the plurality of users and output the information to an external device (not shown).

[0116] In this way, the information processing device according to this embodiment creates a GNN from the user characteristics of multiple users, the characteristics of multiple items, and multiple tasks defined as one classification of items. For any given user, the GNN is connected by edges to all other users, items (including genres, shops, etc.), and tasks related to that user, and is expected to be utilized in platform businesses.

[0117] In addition, the information processing device according to this embodiment creates a user representation for an arbitrary user from the GNN, which represents all other users, items, and tasks related to the arbitrary user. The user representation is a representation in which the features (attributes) of all other users, items, and tasks are linked to one user. In other words, the user representation is not an individual representation for each other user or item, but is a representation that includes composite relationships with other users, items, and tasks for an arbitrary user. This allows the connections of an arbitrary user with other users, items, and tasks to be treated as one representation, and when the user representation is used for arbitrary prediction processing for a platform business, the advantageous effect of reducing the amount of calculation processing can be obtained.

[0118] Furthermore, the information processing device according to this embodiment uses the user expression and user features of an arbitrary user to predict whether the arbitrary user has similar features to a predetermined seed user (possibility of the arbitrary user being a prospective user). By using the user expression including the task as well as the user features, prediction can be made with more consideration given to the user's preferences, and the accuracy of the prediction can be improved. Furthermore, the information on the prospective user obtained by the prediction enables more targeted marketing. For example, if the task is set to a brand name, the brand preferences of the arbitrary user become clear, and it becomes possible to develop marketing focused on that brand for the arbitrary user.

[0119] Although specific embodiments have been described above, the embodiments are merely illustrative and are not intended to limit the scope of the present invention. The apparatus and method described in this specification can be embodied in forms other than those described above. Furthermore, the above-described embodiments can be appropriately omitted, substituted, and modified without departing from the scope of the present invention. Such omitted, substituted, and modified forms are included in the scope of the claims and their equivalents, and belong to the technical scope of the present invention. [Explanation of symbols]

[0120] 1 to N: user, 10: information processing device, 11-1 to 11-N: user device, 101: user feature acquisition unit, 102: item feature acquisition unit, 103: graph construction unit, 104; 1302: expression extraction unit, 105; 1303: potential user prediction unit, 106: learning unit, 107: output unit, 110: learning model storage unit, 111; 113: potential user prediction model, 112: score prediction model, 120: feature storage unit, 121: user feature, 122: item feature, 1301: GNN construction unit

Claims

1. An acquisition means for acquiring user features of a plurality of users from a plurality of user devices used by the plurality of users, the user features including features related to items, which are purchase histories, search histories, or browsing histories for items including products or services via web services, and acquiring item features of the plurality of items from a predetermined database; a selection means for selecting, from the user characteristics of the plurality of users, user characteristics of a model user group, which is a group of users among the plurality of users who have at least one of purchasing, using, and positively evaluating a given product or service through a web service, and user characteristics of a target user group, which is a user group among the plurality of users different from the model user group; a user-to-user graph constructing means for constructing a user relationship graph representing relationships between the plurality of users by connecting a plurality of user nodes representing the plurality of users with links using commonalities between user characteristics of the plurality of users; an item-to-item graph constructing means for constructing an item-to-item relationship graph representing the relationships between the items by connecting a plurality of item nodes representing the items with links using the relationships between the item features of the items; a user-item graph construction means for constructing a user-item relationship graph representing relationships between each of the users and the items by connecting each of the user nodes and each of the item nodes with a link using item-related features among the user features of the users; a graph construction means for constructing a graph representing the mutual relationships between the plurality of user nodes and the plurality of item nodes by connecting the plurality of user nodes and the plurality of item nodes with links using common user nodes or item nodes as connection points in the inter-user relationship graph, the inter-item relationship graph, and the user-item graph; a generation means for embedding the graph in a vector space, learning a vector representation of each node and each link in the vector space, and generating, as a user representation, a vector representation in the graph for each user node of the target user group; and a prediction means for predicting, as a similar user group, one or more users among the target user group having characteristics similar to those of users included in the model user group based on the user expression; The prediction means uses a learning model for machine learning configured to input user features and the user expressions of the target user group and output the probability of having user features similar to the user features of the model user group, and predicts one or more users whose probability is greater than a predetermined threshold as the similar user group.

2. The method further comprises: setting means for setting a classification of the plurality of items based on the item characteristics; The information processing device according to claim 1, characterized in that the graph construction means constructs, as the graph, a graph network representing the mutual relationships between item-related features among the user features of the plurality of users, item features of the plurality of items, and classifications of the plurality of items.

3. 3. The information processing apparatus according to claim 2, wherein each of the categories of the plurality of items is a brand name of each of the plurality of items.

4. 4. The information processing apparatus according to claim 1, further comprising a distribution unit that distributes advertisements to the similar user group.

5. An information processing method executed by an information processing device, an acquisition step of acquiring user features of a plurality of users from a plurality of user devices used by the plurality of users, the user features including features related to items, which are purchase histories, search histories, or browsing histories for items including products and services via a web service, and acquiring item features of the plurality of items from a predetermined database; a selection process for selecting, from the user characteristics of the plurality of users, user characteristics of a model user group, which is a group of users among the plurality of users who have at least one of purchasing, using, and positively evaluating a given product or service through a web service, and user characteristics of a target user group, which is a user group among the plurality of users different from the model user group; a user-to-user graph construction step of constructing a user-to-user relationship graph representing relationships between the plurality of users by connecting a plurality of user nodes representing the plurality of users with links using commonalities between user characteristics of the plurality of users; an item-to-item graph construction step of constructing an item-to-item relationship graph representing the relationships between the plurality of items by connecting a plurality of item nodes representing the plurality of items with links using the relationships between the item features of the plurality of items; a user-item graph construction step of constructing a user-item relationship graph representing relationships between each of the users and the items by connecting each of the user nodes and each of the item nodes with a link using item-related features among the user features of the users; a graph construction step of constructing a graph representing the mutual relationships between the plurality of user nodes and the plurality of item nodes by connecting the plurality of user nodes and the plurality of item nodes with links using common user nodes or item nodes as connection points in the inter-user relationship graph, the inter-item relationship graph, and the user-item graph; a generation step of embedding the graph into a vector space, learning a vector representation of each node and each link in the vector space, and generating, as a user representation, a vector representation in the graph for each user node of the target user group; a prediction step of predicting, as a similar user group, one or more users among the target user group having characteristics similar to characteristics of users included in the model user group based on the user representation; The information processing method is characterized in that, in the prediction process, a learning model for machine learning configured to input user features and the user expressions of the target user group and output the probability of having user features similar to the user features of the model user group is used to predict one or more users whose probability is greater than a predetermined threshold as the similar user group.

6. An information processing program for causing a computer to execute information processing, the program comprising: an acquisition process for acquiring user features of a plurality of users from a plurality of user devices used by the plurality of users, the user features including features related to items, which are purchase histories, search histories, or browsing histories for items including products and services via a web service, and acquiring item features of the plurality of items from a predetermined database; a selection process for selecting, from the user characteristics of the plurality of users, user characteristics of a model user group, which is a group of users among the plurality of users who have at least one of purchasing, using, and positively evaluating a given product or service through a web service, and user characteristics of a target user group, which is a user group among the plurality of users different from the model user group; a user-to-user graph construction process for constructing a user relationship graph representing relationships between the plurality of users by connecting a plurality of user nodes representing the plurality of users with links using commonalities between user characteristics of the plurality of users; an item-to-item graph construction process for constructing an item-to-item relationship graph representing the relationships between the items by connecting a plurality of item nodes representing the plurality of items with links using the relationships between the item features of the plurality of items; a user-item graph construction process for constructing a user-item relationship graph representing relationships between each of the users and the items by connecting each of the user nodes and each of the item nodes with a link using item-related features among the user features of the users; a graph construction process for constructing a graph representing the mutual relationships between the plurality of user nodes and the plurality of item nodes by connecting the plurality of user nodes and the plurality of item nodes with links using common user nodes or item nodes as connection points in the inter-user relationship graph, the inter-item relationship graph, and the user-item graph; a generation process of embedding the graph in a vector space, learning a vector representation of each node and each link in the vector space, and generating, as a user representation, a vector representation in the graph for each user node of the target user group; a prediction process of predicting, as a similar user group, one or more users among the target user group who have characteristics similar to those of users included in the model user group, based on the user representation; The prediction process includes predicting, as the similar user group, one or more users whose probability is greater than a predetermined threshold, using a learning model for machine learning configured to input user features of the target user group and the user expression and output a probability of having user features similar to the user features of the model user group. Information processing program.

Images