Information processing device, information processing method, and program

Abstract

To provide an information processing apparatus, an information processing method, and a program for realizing targeting in which social connections are considered.SOLUTION: In an information processing apparatus, a target expansion section 104 applies user features of seed users 1201 (target user group) set by a user feature setting section 103 to a relationship graph 1202 generated by a graph generation section 102, adds a user group with social connections to the seed users 1201 to the seed users 1201, and determines the result as expanded user group 1203.SELECTED DRAWING: Figure 12A

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program, and particularly to a technique for determining a user group to be an advertising delivery target.

Background Art

[0002] In recent years, in advertising delivery via the Internet, selection (targeting) of a user group to be an advertising delivery target has been performed. As a targeting technique, for example, the technique disclosed in Patent Document 1 is known. In this document, even when there is no purchase history of a product corresponding to an advertisement, a consumer group estimated to have purchased the product corresponding to the advertisement based on the purchase history of other products is described as being determined as an advertising delivery target.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the technique according to Patent Document 1, an advertising delivery target is determined based on a database representing characteristics related to the consumption behavior of each consumer belonging to a plurality of preset consumer groups. However, in this technique, there is a problem that the social connection between consumers that can affect the consumption behavior of consumers (users) is not considered, and effective targeting has not been realized.

[0005] The present invention has been made in view of the above problems, and an object thereof is to provide a technique for realizing targeting considering social connection.

Means for Solving the Problems

[0006] To solve the above problems, one aspect of the information processing apparatus according to the present invention includes: acquisition means for acquiring factual characteristics of each of a plurality of users as user characteristics; creation means for creating a relationship graph showing the social relationships between the plurality of users based on the user characteristics; target user setting means for setting one or more users from the plurality of users as a target user group; and determination means for determining, based on the relationship graph and the user characteristics of the target user group, one or more users from the plurality of users who have a social relationship with the users included in the target user group, to be added to the target user group as an extended user group.

[0007] The system further includes a prediction means that predicts, based on the user characteristics, one or more users among the plurality of users that have similar user characteristics to the target user as a similar user group, and the determination means can determine, based on the relationship graph and the user characteristics of the similar user group, one or more users among the plurality of users that have a social relationship with a user included in the similar user group, in addition to the target user group and the similar user group, as an extended user group.

[0008] The prediction means can predict the group of similar users using a machine learning model.

[0009] In the relationship graph, each user is represented by a user node, and the creation means can connect each node with a link indicating that it has the social relationship, based on the factual information.

[0010] The creation means can connect pairs of user nodes having the same factual characteristics with explicit links, and, based on the multiple pairs of user nodes connected by such explicit links, can connect pairs of nodes that are not connected by such explicit links with implicit links.

[0011] The creation means may determine the intimacy level of the connected pair based on one or more factual pieces of information shared by the pair.

[0012] The system may further include means for delivering advertisements to the aforementioned expanded user group.

[0013] To solve the above problems, one aspect of the information processing method according to the present invention includes: an acquisition step of acquiring factual characteristics of each of a plurality of users as user characteristics; a creation step of creating a relationship graph showing the social relationships between the plurality of users based on the user characteristics; a target user setting step of setting one or more users from the plurality of users as a target user group; and a determination step of adding one or more users from the plurality of users who have a social relationship with the users included in the target user group to the target user group, and determining them as an extended user group, based on the relationship graph and the user characteristics of the target user group.

[0014] 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 perform information processing, the program for causing the computer to perform the following processes: an acquisition process for acquiring factual characteristics of each of a plurality of users as user characteristics; a creation process for creating a relationship graph showing the social relationships between the plurality of users based on the user characteristics; a target user setting process for setting one or more of the plurality of users as a target user group; and a decision process for determining, based on the relationship graph and the user characteristics of the target user group, that one or more users from the plurality of users who have a social relationship with the users included in the target user group are added to the target user group and designated as an extended user group. [Effects of the Invention]

[0015] According to the present invention, targeting that takes social connections into consideration becomes possible. The objects, embodiments, and effects of the present invention described above, as well as other objects, embodiments, and effects of the present invention not described above, will be understood by those skilled in the art from the following embodiments for carrying out the invention by referring to the accompanying drawings and the claims. [Brief explanation of the drawing]

[0016] [Figure 1] Figure 1 shows a configuration example of an information processing system. [Figure 2] Figure 2 shows a functional configuration example of the information processing apparatus 10 according to the embodiment. [Figure 3] Figure 3 shows a flowchart of the process for creating a relationship graph. [Figure 4A] Figure 4A is a diagram for explaining an explicit link. [Figure 4B] Figure 4B is a diagram for explaining an explicit link. [Figure 4C] Figure 4C is a diagram for explaining an explicit link. [Figure 4D] Figure 4D is a diagram for explaining an implicit link. [Figure 5A] Figure 5A is a diagram for explaining the inference process of the relationship between links. [Figure 5B] Figure 5B shows a flowchart of an example of the process for grouping pairs into clusters. [Figure 6A] Figure 6A shows a conceptual diagram of a score (intimacy score) based on the intimacy of the relationship for user pairs. [Figure 6B] Figure 6B shows a schematic architecture of the score prediction model 112. [Figure 7] Figure 7 shows a conceptual diagram of a relationship graph. [Figure 8A] Figure 8A is a diagram for explaining the learning stage of the prospective user prediction model 111. [Figure 8B] Figure 8B is a diagram for explaining the prediction stage of the prospective user prediction model 111. [Figure 9] Figure 9 shows a hardware configuration example of the information processing apparatus 10 and the user apparatus 11. [Figure 10] Figure 10 shows a flowchart of the first target expansion process executed by the information processing apparatus 10. [Figure 11] Figure 11 shows a flowchart of the second target expansion process executed by the information processing apparatus 10. [Figure 12A] Figure 12A shows a conceptual diagram of the first target extension process. [Figure 12B] Figure 12B shows a conceptual diagram of the second target extension process. [Modes for carrying out the invention]

[0017] Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the attached drawings. Among the components disclosed below, those having the same function are denoted by the same reference numeral, and their descriptions are omitted. The embodiments disclosed below are merely examples of means for realizing the present invention, and should be modified or changed as appropriate depending on the configuration of the apparatus to which the present invention is applied and various conditions, and the present invention is not limited to the embodiments below. Furthermore, not all combinations of features described in these embodiments are essential for solving the problem of the present invention.

[0018] [Functional Configuration of Information Processing Equipment] Figure 1 shows an example of the configuration of the information processing system according to this embodiment. As an example, this information processing system, as shown in Figure 1, is composed of an information processing device 10 and a plurality of user devices 11-1 to 11-N (N>1) used by any plurality of 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. Also, in the following description, the terms user device and user may be used synonymously.

[0019] The user device 11 is, for example, a smartphone or tablet, and is configured 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 Wi-Fi (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 the GUI (Graphical User Interface) equipped on the liquid crystal display. These operations include various operations on content such as images displayed on the screen, such as tapping, sliding, and scrolling using a finger or stylus. The user device 11 is not limited to the device shown in Figure 1, but may also be a desktop PC (Personal Computer) or a notebook PC. In that case, each user may operate the device using input devices such as a mouse or keyboard. The user device 11 may also have a separate display surface.

[0020] User device 11 can log in to and use web services (Internet-related services) provided via the information processing device 10 from the information processing device 10 or from other devices (not shown). Such web services may include online malls, online supermarkets, or services related to communications, finance, real estate, sports, and travel, provided via the internet. By using such web services, user device 11 can transmit information about the user of user device 11 to the information processing device 10.

[0021] For example, the user device 11 can transmit information about the user device and its characteristics, such as the IP (Internet Protocol) address of the user device 11, the user's address, and the user's name, to the information processing device 10. Furthermore, the user device 11 can perform positioning calculations based on signals received from GPS (Global Positioning System) satellites (not shown), generate the information obtained from these calculations as the location information of the user device 11, and transmit it to the information processing device 10. The information processing device 10 acquires various information from the user device 11 and, based on that information, creates a relationship graph network (hereinafter referred to as the relationship graph) that shows the social relationships between users. Then, using the created relationship graph, the information processing device 10 determines that the expanded user group, which has social connections (social relationships) with one or more arbitrarily selected users (for example, one or more randomly selected users (seed users)), will be the target of advertising delivery. In other words, the information processing device 10 uses the created relationship graph to perform target expansion (expanded targeting) based on the seed users.

[0022] [Functional configuration of the information processing device 10] The information processing device 10 in this embodiment first acquires various information from user devices 11-1 to 11-N and creates a relationship graph showing the social relationships between users 1 to N. Then, the information processing device 10 uses this relationship graph to perform target expansion.

[0023] Figure 2 shows an example of the functional configuration of the information processing device 10 according to this embodiment. The information processing device 10 shown in Figure 2 comprises a user feature acquisition unit 101, a graph creation unit 102, a user feature setting unit 103, a target expansion unit 104, a prospective user prediction unit 105, a learning unit 106, an output unit 107, a learning model storage unit 110, and a user feature storage unit 120. The learning model storage unit 110 stores a prospective user prediction model 111 and a score prediction model 112, which are machine learning models. These various learning models will be described later. The user feature storage unit 120 stores user features 121.

[0024] The user feature acquisition unit 101 acquires factual features (factual information) (hereinafter referred to as "user features") about each of the user devices 11-1 to 11-N. User features are factual features (information) that are actually or objectively obtained from the user device or user. For example, the user feature acquisition unit 101 can acquire user features directly from user device 11. Alternatively, the user feature acquisition unit 101 can acquire user features as information registered by the user of user device 11 to a predetermined web service.

[0025] User characteristics include the IP address of the user's device, the user's address and name, the user's credit card number, and the user's demographic information (such as gender, age, residential area, occupation, family structure, and other demographic user attributes). User characteristics may also include the registration number and registered name when using a specified web service. Furthermore, user characteristics may include call history, delivery addresses other than the user's address when using a specified web service, usage status, usage history, search history, and information about points that can be accumulated through service use. Thus, user characteristics can include any information related to the user's device or the user themselves, as well as information about the use of a specified service via communication. The user feature acquisition unit 101 stores the acquired user features as user features 121 in the user feature storage unit 120.

[0026] The graph creation unit 102 creates a relationship graph using various user characteristics acquired by the user characteristic acquisition unit 101. The relationship graph will be described later.

[0027] The user characteristic setting unit 103 sets an arbitrary group of target users selected from user devices 11-1 to 11-N as seed users. Note that there may be only one seed user. The seed user may be set by an operator through an input operation using the input unit (input unit 95 in Figure 9), may be set in advance in the system, or may be set by an arbitrary program stored in the storage unit (ROM 92 or RAM 93 in Figure 9). Furthermore, the user characteristic setting unit 103 obtains the user characteristics of the set seed user from user characteristics 121 and sets them in the target expansion unit 104 and the prospective user prediction unit 105. In addition, the user characteristic setting unit 103 obtains the user characteristics of other user groups from user devices 11-1 to 11-N other than the seed user from user characteristics 121 and sets them in the prospective user prediction unit 105.

[0028] The target expansion unit 104 performs target expansion (determination of expanded user groups) using social connections. In one embodiment, the target expansion unit 104 uses a relationship graph created by the graph creation unit 102 to perform target expansion based on seed users set by the user characteristic setting unit 103. In another embodiment, the target expansion unit 104 performs target expansion based on a group of prospective users (described later) predicted from the set seed users. This group of prospective users is predicted by the prospective user prediction unit 105. The target expansion process will be described later.

[0029] The prospective user prediction unit 105 predicts a group of users that are predicted to have similar user characteristics to the seed users set by the user characteristic setting unit 103, as a prospective user group (similar user group). In this embodiment, the prospective user is predicted using the prospective user prediction model 111 that has been trained by the learning unit 106. The prediction process for the prospective user will be described later. Note that the prediction process for the prospective user group is not limited to using the prospective user prediction model 111.

[0030] The learning unit 106 trains the prospective user prediction model 111 and the score prediction model 112, and stores the trained prospective 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.

[0031] The output unit 107 outputs the result of the target expansion performed by the target expansion unit 104 (that is, the information of the expanded user group). The output can be any output process, and can be an output to an external device via the communication I / F (communication I / F 97 in FIG. 9), or can be a display on the display unit (display unit 96 in FIG. 9).

[0032] [Procedure for creating the relationship graph] Next, the procedure for creating the relationship graph according to the present embodiment will be described. In the following description, users A to E are users referred to for explanation and can be users of the user device​​​​​​​​​​ Figure 4A shows an example of creating an explicit link using the IP address of the user's device as a common feature. Figure 4A shows an example where the web services available to users A-C are an online mall 41, a golf course reservation service 42, a travel-related reservation service 43, and a card management system 44. Figures 4A-4C show these four web services, but the number of web services is not limited to a specific number.

[0035] Online Mall 41 is a shopping mall that is accessible online (using the internet). Online Mall 41 can offer a wide variety of goods and services, such as fashion, books, food, concert tickets, and real estate. Golf course reservation service 42 is operated by a website that provides online services related to golf courses, such as golf course search and reservation and lesson information. Travel-related booking services 43 are operated by websites that provide various travel services available online. For example, travel-related booking services 43 can provide hotel and travel tour reservations, airline and rental car reservations, tourist information, hotel information, and information about the area around hotels. The card management system 44 operates on a website that provides services related to credit cards issued and managed by a designated card management company. The card management system 44 may also provide services in connection with at least one of the online mall 41, the golf course reservation service 42, and the travel-related reservation service 43.

[0036] In the example shown in Figure 4A, users A through C each use the same IP address (=198.45.66.xx) to access the online mall 41, the golf course reservation service 42, and the travel-related reservation service 43. The IP address information can be obtained by the user feature acquisition unit 101. In such cases, the graph generation unit 102 creates explicit links (for example, link L1 between user A and user C) between users A to C, using the same IP address characteristics as shown in link status 45.

[0037] Figure 4B shows an example of creating explicit links using the characteristics of the user's address as a common feature. Similar to Figure 4A, Figure 4B shows an example where the web services available to users A to C are an online mall 41, a golf course reservation service 42, a travel-related reservation service 43, and a card management system 44. Here, users A to C each register the same address (delivery address) and use the online mall 41, the golf course reservation service 42, and the travel-related reservation service 43. Address information can be obtained by the user feature acquisition unit 101. In such cases, the graph creation unit 102 creates explicit links (for example, link L1 between user A and user C) between users A to C, using the same address characteristics as shown in link status 46.

[0038] Figure 4C shows an example of creating an explicit link using the characteristics of the user's credit card number as a common feature. Figure 4B, similar to Figure 4A, shows an example where the web services available to users A to C are an online mall 41, a golf course reservation service 42, a travel-related reservation service 43, and a card management system 44. Here, users A to C each register the same credit card and use the online mall 41, the golf course reservation service 42, and the travel-related reservation service 43. Information including the credit card number can be obtained by the user feature acquisition unit 101. In such cases, the graph creation unit 102 creates explicit links between users A to C (for example, link L1 between user A and user C) based on the same card characteristics, as shown in link status 47.

[0039] FIG. 4D shows an example of creating an implicit link between users. In the example of FIG. 4D, 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 it is inferred that an implicit relationship is constructed between each user (each node), and the inferred link is created (established) as an implicit link. In the example of FIG. 4D, 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 L 2 is created. Note that the graph creation unit 102 predicts and creates an implicit link between users by performing learning (representation learning, relationship learning, embedding learning, knowledge graph embedding) of a user relationship graph composed of nodes (users) connected by explicit links. At this time, the graph creation unit 102 may perform the learning based on a known embedding model or an extension thereof as appropriate.

[0040] <S32: Inference of Relationship between Links> In S32, the graph creation unit 102 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 FIGS. 5A and 5B. FIG. 5A 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.

[0041] The graph creation unit 102 treats user pairs connected by links created in S31 as data points and groups these pairs (data points) into clusters representing a common type using various information acquired by the user feature acquisition unit 101. This information may include IP addresses, addresses, credit card information, age, gender, and friends. Each cluster may represent relationships such as spouse, parent, neighbor, same household, colleague, friend, same-sex sibling, or opposite-sex sibling. In the example in Figure 5A, user pairs are shown with an "X," and the clusters to which these pairs can be grouped are shown as parent-child cluster 51, spouse cluster 52, same-sex sibling cluster 53, friend cluster 54, and colleague cluster. Although Figure 5A shows five clusters, the number of clusters is not limited to a specific number.

[0042] For example, if user A and user B have the same last name, an age difference of less than 10 years, opposite genders, and the same address, the graph creation unit 102 can group the pair of user A and user B into a cluster (spouse cluster 52) that represents the relationship between husband and wife (spouse).

[0043] Figure 5B shows a flowchart of an example of the process performed by the graph creation unit 102 to group pairs into clusters. At the start of S51, for pairs to be grouped, assume they have the characteristics of the same address and the same family name. In S52, the graph creation unit 102 determines whether the pair in question has the characteristic of the same gender. If the pair in question has the characteristic of the same gender (Yes in S52), in S53, the graph creation unit 102 determines whether the age difference of the pair in question is less than or equal to a predetermined threshold value (= X value). If the age difference of the pair in question is greater than the X value (No in S53), the graph creation unit 102 groups the pair in question into the parent-child cluster 51. If the age difference is less than or equal to the X value (Yes in S53), the graph creation unit 102 groups the pair in question into the same-gender sibling cluster 53. Also, if the pair in question does not have the characteristic of the same gender (No in S52), in S54, the graph creation unit 102 determines whether the age difference of the pair in question is less than or equal to a predetermined threshold value (= Y value). If the age difference is greater than the Y value (No in S54), the graph creation unit 102 groups the pair in question into the parent-child cluster 51. If the age difference is less than or equal to the Y value (Yes in S54), the graph creation unit 102 groups the pair in question into the spouse cluster 52.

[0044] <S33: Score Assignment Based on Relationship Closeness> In S33, the graph creation unit 102 predicts a score based on the relationship closeness for the pair inferred in S32 and assigns the score to the pair. In this 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 the score based on the relationship closeness for user pairs (hereinafter referred to as the closeness score).

[0045] In the example in Figure 6A, the degree of intimacy in the relationship between user A and user B, who are connected by an explicit link, changes depending on the characteristics they possess (share). In the upper part of Figure 6A, if user A and user B have characteristics 60 such as being same-sex siblings, having the same address, 1200 call logs, and 50 gift exchanges, the degree of intimacy in the relationship between the user pair (i.e., the intimacy score) will be high. On the other hand, in the lower part of Figure 6A, if user A and user B have characteristics 61 such as being same-sex siblings, having different addresses, 30 call logs, and 2 gift exchanges, the degree of intimacy in the relationship between the user pair (i.e., the intimacy score) will be low. Thus, as in the example in Figure 6A, even if user A and user B are same-sex siblings, the degree of intimacy in the relationship between the user pair will differ depending on other characteristics shared by the user pair. It is observed that pairs with a high degree of relationship intimacy have a closer social distance from each other and exert a greater influence. On the other hand, pairs with a low level of intimacy in their relationship are observed to have a greater social distance from each other and not be in a close relationship.

[0046] In this embodiment, the intimacy score for a user pair is predicted using a score prediction model 112. Figure 6B shows a schematic architecture of the score prediction model 112. The score prediction model 112 is a learning model that takes the features 63 of a user pair as input and predicts the intimacy score 64 for those features 63.

[0047] The score prediction model 112 is, for example, a learning model that performs weak supervised learning, such as a learning model using a convolutional neural network (CNN). In this embodiment, the score prediction model 112 is a learning model that has been trained using intimacy scores (0 to 1) assigned to multiple features for a user pair as training data, as shown in Figure 6A. For example, in the training phase, the training data used is a combination of data such as an intimacy score close to 1 set for feature 60 in Figure 6A and an intimacy score close to 0 set for feature 61. This training process is performed by the learning unit 108. The score prediction model 112 may differ for each type of relationship between user pairs, and may be a learning model trained according to a single type of relationship.

[0048] In this embodiment, the intimacy score for a user pair was predicted using the score prediction model 112, but the graph creation unit 102 may be configured to predict the score using other methods.

[0049] Through the above process, explicit or implicit links are formed between multiple users, a closeness score is assigned to each link, and a relationship graph is created. A conceptual diagram of the relationship graph is shown in Figure 7. Each user 71-73 has multiple characteristics, and the predicted closeness score is assigned to each pair of users as described above.

[0050] [Predictive processing of potential users] Next, the prospective user prediction process according to this embodiment will be described. In this embodiment, the prospective user prediction process is performed using the prospective user prediction model 111. The learning phase and the prediction phase of the prospective user prediction model 111 will be described separately.

[0051] Figure 8A shows a diagram illustrating the learning stages of the prospective user prediction model 111. First, the user feature setting unit 103 sets active users 81, which consist of multiple users. For example, the user feature setting unit 103 sets multiple users from the users of user devices 11-1 to 11-N shown in Figure 1 who have used a predetermined web service during a certain period in the past (for example, 6 months) as active users 81. Alternatively, the user feature setting unit 103 may set multiple users from the users of user devices 11-1 to 11-N that have been pre-configured as active users 81.

[0052] After setting up the active users 81, the user characteristic setting unit 103 extracts a number of positive users 82 and a number of negative users 83 from the active users 81. Positive users 82 are users who have purchased and / or used a given product or service through the web service and / or have given a positive evaluation of the product or service through the web service. On the other hand, negative users 83 are users other than positive users 82 who are randomly selected from the active users 81.

[0053] Next, the user feature setting unit 103 retrieves user features 84 for positive users 82 and negative users 83 from user features 121 stored in the user feature storage unit 120. For example, user features 84 for positive users 82 and negative users 83 include demographic information and purchase history on the web service (information on product genres and types, etc.). The demographic information and purchase history each include multiple subdivided features. Note that user features 84 may include other features besides demographic information and purchase history, such as point status (available points, etc.) and point features (information on point transactions, such as points earned / used from online or offline shops, etc.). The learning unit 106 uses the user features 84 and the categories (labels (ground truth data)) of positive users or negative users corresponding to those user features to train the prospective user prediction model 111.

[0054] The prospective user prediction model 111 is, for example, a learning model based on XGBoost. During the learning phase, the learning unit 106 performs verification and tuning of hyperparameters (parameters that control the behavior of the prospective user prediction model 111) through grid search and cross-validation. Because XGBoost is a tree (decision tree) based model, the prospective user prediction model 111 can generate results (feature evaluation 85) that show how the input data (user features) affect the model's output. This makes it possible, for example, to verify which user features (combinations of subdivided features) have a greater influence on positive users.

[0055] The trained prospective user prediction model 111 is configured to represent the probability that each user in any given input user group has similar user characteristics to a positive user. This probability is represented by a numerical value between 0 and 1, for example, with the maximum probability being 1. Here, if the threshold is set to 0.5, for example, the prospective user prediction unit 105 can predict (determine) users with a probability greater than 0.5 as prospective users.

[0056] In the prediction (estimation) stage, a group of users who are expected to have similar user characteristics to seed users (corresponding to positive users in the learning stage) are predicted as the prospective user group. Figure 8B shows a diagram illustrating the prediction stage of the prospective user group.

[0057] The prospective user prediction unit 105 inputs the user characteristics 87 of the user group 86 other than the seed user set by the user characteristic setting unit 103 from user devices 11-1 to 11-N into the trained prospective user prediction model 111 to predict the prospective user group 88. The user characteristics 87 include, but are not limited to, demographic information and purchase history on the web service, as explained with reference to Figure 8A. As mentioned above, the prospective user prediction model 111 outputs the probability (for example, a numerical value from 0 to 1) that each user in the user group 86 has similar user characteristics to the seed user. If the threshold is set to 0.5, the prospective user prediction unit 105 can predict one or more users in the user group 86 with a probability higher than 0.5 as the prospective user group 88.

[0058] Next, we will describe the target expansion process using the aforementioned relationship graph that shows social connections, according to this embodiment. First, we will describe a form that uses only the relationship graph as the first target expansion process. Next, we will describe a form that uses the prospective user prediction model 111 and the relationship graph as the second target expansion process.

[0059] [First target expansion process] Figure 12A shows a conceptual diagram of the first target expansion process. The target expansion unit 104 applies the user characteristics of the seed user 1201 (target user group) set by the user characteristic setting unit 103 to the relationship graph 1202 created by the graph creation unit 102, and determines the group of users who have a social connection with the seed user 1201 as the expanded user group 1203, in addition to the seed user 1201. The user characteristics of the seed user 1201 can be any information that can identify the seed user 1201 in the relationship graph, such as the seed user 1201's address, name, or demographic information. The target expansion unit 104 may also determine users whose intimacy score with the seed user 1201 is higher than a predetermined threshold in the relationship graph as the expanded user group 1203, in addition to the seed user 1201.

[0060] [Second target expansion process] Next, the target expansion process using the aforementioned relationship graph that shows social connections, according to this embodiment, will be explained. Figure 12B shows a conceptual diagram of the second target expansion process. The prospective user prediction unit 105 predicts the prospective user group 1204 from the seed users 1201 (target user group) set by the user characteristic setting unit 103, using the prospective user prediction model 111 described above. Specifically, the prospective user prediction unit 105 predicts the prospective user group 1204 as a group of users (not shown) other than the seed users that are predicted to have similar user characteristics to the seed users 1201.

[0061] Next, the target expansion unit 104 applies the user characteristics of the prospective user group 1204 to the relationship graph 1202 created by the graph creation unit 102, and determines the group of users who have social connections with the prospective user group 1204 as the expanded user group 1205, in addition to the seed users 1201 and the prospective user group 1204. The user characteristics of the prospective user group 1204 can be any information that can identify the prospective user group 1204 in the relationship graph, such as the address, name, or demographic information of the prospective user group 1204. The target expansion unit 104 may also determine, in the relationship graph, users whose intimacy score with the prospective user group 1204 is higher than a predetermined threshold, as the expanded user group 1205, in addition to the seed users 1201 and the prospective user group 1204.

[0062] [Hardware configuration of the information processing device 10] Figure 9 is a block diagram showing an example of the hardware configuration of the information processing device 10 according to this embodiment. The information processing device 10 according to this embodiment can be implemented on one or more computers, mobile devices, or any other processing platform. Referring to Figure 9, an example is shown in which the information processing device 10 is implemented in a single computer; however, the information processing device 10 according to this embodiment may be implemented in a computer system including multiple computers. The multiple computers may be connected to each other via a wired or wireless network.

[0063] As shown in Figure 9, the information processing device 10 may include a CPU 91, a ROM 92, a RAM 93, an HDD 94, an input unit 95, a display unit 96, a communication interface 97, and a system bus 98. The information processing device 10 may also include external memory. The CPU (Central Processing Unit) 91 comprehensively controls the operation of the information processing device 10 and controls each component (92-97) via the system bus 98, which is a data transmission path.

[0064] ROM (Read Only Memory) 92 is a non-volatile memory that stores control programs and other data necessary for the CPU 91 to execute processing. These programs may also be stored in non-volatile memory such as an HDD (Hard Disk Drive) 94 or SSD (Solid State Drive), or in external memory such as a removable storage medium (not shown). The RAM (Random Access Memory) 93 is a volatile memory and functions as the main memory, work area, etc., of the CPU 91. In other words, when executing processing, the CPU 91 loads necessary programs, etc., from the ROM 92 into the RAM 93 and executes these programs, etc., to realize various functional operations. The learning model storage unit 110 and the user feature storage unit 120 shown in Figure 2 can be composed of RAM 93.

[0065] HDD94 stores various data and information necessary for CPU91 to perform processing using programs, for example. Furthermore, HDD94 also stores various data and information obtained through processing performed by CPU91 using programs, for example. The input unit 95 consists of a pointing device such as a keyboard or mouse. The display unit 96 is comprised of a monitor such as a liquid crystal display (LCD). The display unit 96 may function as a GUI (Graphical User Interface) when combined with the input unit 95.

[0066] Communication I / F97 is an interface that controls communication between the information processing device 10 and an external device. The communication interface 97 provides an interface to the network and performs communication with external devices via the network. Various data and parameters are sent and received between the external device and the communication interface 97. In this embodiment, the communication interface 97 may perform 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 usable in this embodiment is not limited to this and may consist of 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). Furthermore, it 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 it can communicate with each other, and the communication standard, scale, and configuration are not limited to those described above.

[0067] At least some of the functions of the information processing device 10 shown in Figure 2 can be realized by the CPU 91 executing a program. However, at least some of the functions of the information processing device 10 shown in Figure 2 may be operated as dedicated hardware. In this case, the dedicated hardware operates based on the control of the CPU 91.

[0068] [Hardware configuration of user device 11] The hardware configuration of the user device 11 shown in Figure 1 may be the same as that in Figure 9. That is, the user device 11 may include a CPU 91, a ROM 92, a RAM 93, an HDD 94, an input unit 95, a display unit 96, a communication interface 97, and a system bus 98. The user device 11 can display various information provided by the information processing device 10 on the display unit 96 and perform processing corresponding to input operations received from the user via a GUI (configured with an input unit 95 and a display unit 96).

[0069] [Process flow] Figure 10 shows a flowchart of the first target extension process described above, which is executed by the information processing device 10 according to this embodiment. The process shown in Figure 10 can be realized by the CPU 91 of the information processing device 10 loading a program stored in ROM 92 or the like into RAM 93 and executing it. For the explanation of Figure 10, refer to the information processing system shown in Figure 1. Assume that the score prediction model 112, which has been trained by the learning unit 106, is stored in the learning model storage unit 110.

[0070] In S101, the user feature acquisition unit 101 acquires user features for each user from user devices 11-1 to 11-N and stores them in the user feature storage unit 120 as user features 121. The process in S101 may also be a process for acquiring (collecting) user features over a certain period in the past.

[0071] In S102, the graph creation unit 102 creates a relationship graph for users 1 to N using the various user features acquired by the user feature acquisition unit 101. The procedure for creating the relationship graph is as described above.

[0072] In S103, the user characteristic setting unit 103 sets an arbitrary group of target users from among users 1 to N as seed users. As mentioned above, the seed users may be set by the operator through input operations using the input unit 95, may be set in advance by the system, or may be set by any program stored in ROM 92 or RAM 93. Furthermore, in S103, the user characteristic setting unit 103 obtains the user characteristics of the seed users from the user characteristics 121 and sets them in the target extension unit 104.

[0073] In S104, the target extension unit 104 applies the user characteristics of the seed user, which were set in S103, to the relationship graph created in S102, and determines that the group of users who have social connections with the seed user, in addition to the seed user, will be the extended user group for the seed user. The process in S104 corresponds to the process described with reference to Figure 12A.

[0074] In S105, the output unit 107 outputs information about the extended user group determined in S104. The output unit 107 may generate various information about the extended user group and output it to an external device (not shown). For example, the output unit 107 can distribute advertisements created for the seed users to the extended user group.

[0075] Next, Figure 11 shows a flowchart of the second target extension process described above, which is executed by the information processing device 10 according to this embodiment. The process shown in Figure 11 can be realized by the CPU 91 of the information processing device 10 loading a program stored in ROM 92 or the like into RAM 93 and executing it. Similar to Figure 10, the information processing system shown in Figure 1 will be referred to for the explanation of Figure 11. The score prediction model 112, which has been trained by the learning unit 106, is assumed to be stored in the learning model storage unit 110.

[0076] The processes in S111 and S112 are the same as those in S101 and S102 in Figure 10, so their explanation is omitted. In S113, the user feature setting unit 103 sets an arbitrary group of target users from users 1 to N as seed users. As mentioned above, these seed users may be set by the operator through input operations using the input unit 95, may be set in advance by the system, or may be set by any program stored in ROM 92 or RAM 93. Furthermore, in S113, the learning unit 106 sets these seed users as positive users and trains the prospective user prediction model 111. This learning process corresponds to the process described with reference to Figure 8A.

[0077] In S114, the prospective user prediction unit 105 predicts a group of prospective users from the seed user. Specifically, the prospective user prediction unit 105 inputs the user characteristics of the other user groups 86 besides the seed user into the prospective user prediction model 111 that was trained in S113, and predicts a group of prospective users. The prospective user prediction unit 105 may also determine the group of users that is the predicted group of prospective users plus the seed user as the final group of prospective users. In S115, the user characteristic setting unit 103 obtains the user characteristics of the prospective user group from the user characteristics 121 and sets them in the target extension unit 104.

[0078] In S116, the target expansion unit 104 applies the user characteristics of the prospective user to the relationship graph created in S112, and determines a group of users who have social connections with the prospective user group, in addition to the seed user and the prospective user group, as an expanded user group for the prospective user group. The process in S116 corresponds to the process described with reference to Figure 12B.

[0079] In S117, the output unit 107 outputs information about the extended user group determined in S116. The output unit 107 may generate various information about the extended user group and output it to an external device (not shown). For example, the output unit 107 can distribute advertisements created for the seed users to the extended user group.

[0080] Thus, the information processing device 10 is configured to determine an extended user group as a group of users that have a social connection with a seed user (one or more target users) that has been arbitrarily set. Since this extended user group has a social connection with the seed user, such as parent and child, it is highly likely that the information of the advertisement delivered to the seed user will be shared among this extended user group. Therefore, the same advertising effect as that of the advertisement delivered to the seed user can be expected for this extended user group.

[0081] The information processing device 10 is also configured to estimate a group of potential users from the seed user and to determine a group of users who have social connections with the group of potential users as an extended user group. In other words, for the seed user, users who are expected to perform similar actions (such as purchasing products) on the web service are predicted as a group of potential users, and an extended user group who have social connections with the group of potential users is determined. This makes it possible to identify a wider range of extended user groups as advertising destinations, and further improvements in advertising effectiveness can be expected.

[0082] According to the embodiment described above, it becomes possible to perform extended targeting based on social connections from a small number of seed users, resulting in efficient and effective targeting.

[0083] Although specific embodiments are described above, these embodiments are merely illustrative and not intended to limit the scope of the present invention. Apparatuses and methods described herein can be embodied in forms other than those described above. Furthermore, the embodiments described above can be appropriately omitted, substituted, and modified without departing from the scope of the present invention. Such omitted, substituted, and modified forms fall within the scope of the claims and their equivalents and are within the technical scope of the present invention. [Explanation of Symbols]

[0084] 1~N: User, 10: Information Processing Device, 11-1~11-N: User Device, 101: User Feature Acquisition Unit, 102: Graph Creation Unit, 103: User Feature Setting Unit, 104: Target Expansion Unit, 105: Prospective User Prediction Unit, 106: Learning Unit, 107: Output Unit, 110: Learning Model Storage Unit, 111: Prospective User Prediction Model, 112: Score Prediction Model, 120: User Feature Storage Unit, 121: User Features

Claims

1. A means for acquiring factual characteristics of each of multiple users as user characteristics, A target user setting means for setting one or more users from the aforementioned group of users as a target user group, A learning means for training a first machine learning model to output the intimacy of a pair, using a plurality of features shared between each user and one or more other users, which are identified when the information of each user satisfies pre-set conditions, based on the information contained in the user features of each of the plurality of users and one or more other users, wherein the plurality of features include features that indicate the relationship between each user and one or more other users. A determination means that, based on the user characteristics of the target user group, determines one or more users from among the multiple users who have a social relationship with the users included in the target user group, adds the determined one or more users to the target user group, and determines them as an extended user group. It has, The information processing apparatus is characterized in that the determination means inputs a plurality of features shared by pairs of users included in the target user group and users from the plurality of users that are not included in the target user group into the first machine learning model that has been trained, thereby predicting the level of intimacy between pairs of users included in the target user group and users from the plurality of users that are not included in the target user group, and determining one or more users according to the level of the predicted level of intimacy.

2. The information processing apparatus according to Claim 1, wherein the plurality of features shared between each user and one or more other users include features indicating a relationship between each user and one or more other users, and the relationship includes spouse, parent and child, neighbor, same household, colleague, friend, same-sex sibling, opposite-sex sibling.

3. The system further includes a prediction means for predicting, based on the user characteristics, one or more users among the plurality of users that have similar user characteristics to the users included in the target user group, as a similar user group. The information processing apparatus according to claim 1 or 2, wherein the determination means determines, based on the user characteristics of the similar user group, one or more users from among the plurality of users who have a social relationship with the users of the similar user group, to be added to the target user group and the similar user group, and to be designated as the extended user group, using a second machine learning model.

4. The information processing apparatus according to any one of claims 1 to 3, further comprising a distribution means for distributing advertisements to the aforementioned extended user group.

5. An information processing method performed by an information processing device, The acquisition process involves obtaining factual characteristics for each of multiple users as user characteristics, A target user setting step in which one or more users from the aforementioned group of users are set as the target user group, A learning step in which a first machine learning model is trained to output the intimacy of a pair, using a plurality of features shared between each user and one or more other users, which are identified when the information of each user satisfies pre-set conditions, based on the information contained in the user features of each of the plurality of users and one or more other users, wherein the plurality of features include features that indicate the relationship between each user and one or more other users, and the intimacy assigned to the plurality of features, the learning step being to train a first machine learning model to output the intimacy of the pair, the plurality of features including features that indicate the relationship between each user and one or more other users, A determination step in which, based on the user characteristics of the target user group, one or more users from among the multiple users who have a social relationship with the users included in the target user group are determined, and the determined one or more users are added to the target user group and determined to be an extended user group, It has, The information processing method is characterized in that, in the determination step, a plurality of features shared by pairs of users included in the target user group and users from the plurality of users that are not included in the target user group are input into the first machine learning model that has been trained, thereby predicting the level of intimacy between pairs of users included in the target user group and users from the plurality of users that are not included in the target user group, and determining one or more users according to the level of the predicted level of intimacy.

6. An information processing program for causing a computer to perform information processing, wherein the program causes the computer to perform information processing. A process to acquire factual characteristics of each of multiple users as user characteristics, A target user setting process that sets one or more of the aforementioned multiple users as the target user group, A learning process to train a first machine learning model to output the intimacy of a pair, using a plurality of features shared between each user and one or more other users, which are identified when the information of each user satisfies pre-set conditions, based on the information contained in the user features of each of the plurality of users and one or more other users, wherein the plurality of features include features that indicate the relationship between each user and one or more other users. This process includes a determination process that, based on the user characteristics of the target user group, determines one or more users from among the multiple users who have a social relationship with the users included in the target user group, adds the determined one or more users to the target user group, and determines them as an extended user group. The decision process includes inputting a plurality of features shared by pairs of users included in the target user group and users not included in the target user group from among the plurality of users into the first machine learning model that has been trained, predicting the level of intimacy between pairs of users included in the target user group and users not included in the target user group, and determining one or more users according to the level of the predicted level of intimacy. Information processing program.

Images