Program, method, information processing device, and system

Abstract

This program is to be executed by a computer provided with a processor and a storage unit, wherein the processor executes: a behavior acquisition step for acquiring behavior data pertaining to behavior by a plurality of users; a first extraction step for extracting a first user group comprising one or more users associated with a predetermined theme, among the plurality of users, on the basis of the behavior data acquired in the behavior acquisition step; a representative acquisition step for acquiring representative behavior data representing the first user group extracted in the first extraction step; and a second extraction step for extracting a second user group comprising one or more users from among the plurality of users on the basis of the representative behavior data of the first user group.

Description

Program, method, information processing device, and system

[0001] The present disclosure relates to a program, a method, an information processing device, and a system.

[0002] Technologies for delivering online targeted advertisements are known. Patent Literature 1 proposes a method for analyzing people's hobbies, preferences, behavioral patterns, etc. by performing clustering using user movement history and preference information.

[0003] Japanese Patent Application Laid-Open No. 2019-197460

[0004] Due to the declining domestic population and diversifying lifestyles, hobbies, and preferences, appropriate marketing activities that capture customer personas are more important than ever before. However, data-based persona analysis and marketing activities based on the analysis results each require a high level of expertise. Furthermore, growing momentum for personal information protection, including the abolition of third-party cookies, has made it difficult to conduct individual tracking marketing that relies on the collection of large amounts of personal data across businesses. As a result, there is a need for marketing methods that match big data, statistically processed in various formats, that can capture customer personas with personal data available within first-party websites. However, there is a problem in that it is not possible to identify other user groups (potential customers) that are similar to the typical behavioral patterns (representative behavioral data) of a first user group related to a specific theme or event. The present disclosure is intended to address this problem. Its purpose is to provide a technology that identifies other user groups (potential customers) that are similar to the typical behavioral patterns (representative behavioral data) of a first user group related to a specific theme or event, thereby enabling a more accurate capture of customer profiles.

[0005] A program to be executed by a computer having a processor and a memory unit, the program executing the following steps: a behavior acquisition step in which the processor acquires behavioral data regarding the behavior of multiple users; a first extraction step in which, based on the behavioral data acquired in the behavior acquisition step, a first user group consisting of one or more users from the multiple users who are related to a specified theme; a representative acquisition step in which representative behavior data representing the first user group extracted in the first extraction step; and a second extraction step in which, based on the representative behavior data of the first user group, a second user group consisting of one or more users from the multiple users.

[0006] According to the present disclosure, it is possible to identify other user groups (potential customers) that are similar to the typical behavioral patterns (representative behavioral data) of a first user group related to a specific theme or event, and to identify user groups from user behavioral data.

[0007] 1 is a block diagram showing the functional configuration of the system 1. FIG. 1 is a block diagram showing the functional configuration of the server 10. FIG. 1 is a block diagram showing the functional configuration of the user terminal 20. FIG. 1 is a block diagram showing the functional configuration of the administrator terminal 30. FIG. 2 is a diagram showing the data structure of a user table 1012. FIG. 3 is a diagram showing the data structure of a behavior table 1013. FIG. 4 is a diagram showing the data structure of a theme table 1014. FIG. 5 is a diagram showing the data structure of a cluster table 1015. FIG. 6 is a diagram showing the data structure of an advertisement table 1021. FIG. 7 is a flowchart showing the operation of a cluster processing. FIG. 8 is a flowchart showing the operation of a user extraction processing. FIG. 9 is a flowchart showing the operation of an advertisement distribution processing. FIG. 10 is an example screen showing the operation of an advertisement distribution processing. FIG. 11 is a block diagram showing the basic hardware configuration of a computer 90. FIG. 12 is a diagram showing the relationship between spots before and after a certain spot as a starting point. FIG. 13 is a diagram showing the relationship between each spot taking into account the whole.

[0008] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In all drawings describing the embodiments, common components are designated by the same reference numerals, and repeated description will be omitted. Note that the following embodiments do not unduly limit the content of the present disclosure described in the claims. Furthermore, not all components shown in the embodiments are necessarily essential components of the present disclosure. Furthermore, each drawing is a schematic diagram and is not necessarily a precise illustration.

[0009] <Configuration of System 1> System 1 in the present disclosure is an information processing system that extracts a predetermined user group based on user behavior data. System 1 includes information processing devices, namely, a server 10, a user terminal 20, and an administrator terminal 30, which are connected via a network N. The administrator terminal 30 is not essential. Fig. 1 is a block diagram showing the functional configuration of system 1. Fig. 2 is a block diagram showing the functional configuration of the server 10. Fig. 3 is a block diagram showing the functional configuration of the user terminal 20. Fig. 4 is a block diagram showing the functional configuration of the administrator terminal 30.

[0010] Each information processing device is configured by a computer equipped with an arithmetic unit and a storage device. The basic hardware configuration of the computer and the basic functional configuration of the computer realized by the hardware configuration will be described later. For each of the server 10, the user terminal 20, and the administrator terminal 30, descriptions that overlap with the basic hardware configuration and basic functional configuration of the computer will be omitted.

[0011] The system 1 of the present disclosure can be used as a marketing system / advertising delivery system that analyzes users. Conventional marketing tools that perform user analysis generally do not include ad delivery processing. Marketing tools that perform typical user analysis generally indicate what your customers are like. Conventional ad delivery systems do not include a processing unit that performs user analysis on the advertiser side. Conventional ad delivery systems generally only display advertisements to users corresponding to keywords entered by the advertiser. These data cannot be combined because they use different data. On the other hand, the system 1 of the present disclosure can function as both a persona analysis tool and an ad delivery system by using location information that is common to users of both the marketing system and the ad delivery system and is easy to obtain.

[0012] Until now, when an ad distributor wanted to send information to a certain group of users, it was common to target users only based on specific behaviors. However, this approach had problems such as a small target population, biased attributes and distribution toward the user extraction location, and specific behaviors that were not strictly the user group the ad distributor envisioned. For example, if a distributor wanted to send information to the wealthy by targeting their location information, there was no data indicating a user group called the wealthy. Therefore, the target population was narrowed to users who had checked in at golf courses or luxury hotels. However, if the condition required the user to have visited both a golf course and a luxury hotel, the target population would be reduced, and if the condition required the user to have visited either one of the golf courses or a luxury hotel, the targeting accuracy would be reduced. Furthermore, it was unclear from the data whether the wealthy targeted by the ad distributor commonly visited golf courses and luxury hotels. The system 1 disclosed herein can increase the target population, reduce the geographical influence of bias toward the extraction location, and refine the user group to be considered wealthy.

[0013] The server 10 is an information processing device that provides an information processing service for extracting a predetermined user group based on user behavior data. The server 10 includes a storage unit 101 and a control unit 104.

[0014] <Configuration of Storage Unit 101 of Server 10 > The storage unit 101 of the server 10 includes an application program 1011 , a user table 1012 , a behavior table 1013 , a theme table 1014 , a cluster table 1015 , and an advertisement table 1021 .

[0015] The application programs 1011 are programs for causing the control unit 104 of the server 10 to function as each functional unit. The application programs 1011 include applications such as a web browser application that runs on a user terminal.

[0016] The user table 1012 is a table that stores and manages information about member users (hereinafter referred to as users) who use the service. When a user registers to use the service, the user's information is stored in a new record in the user table 1012. This allows the user to use the service according to the present disclosure. The user table 1012 is a table that uses the user ID as the primary key and has columns for the user ID and user attribute data. FIG. 5 is a diagram showing the data structure of the user table 1012.

[0017] The user ID is an item that stores user identification information for identifying a user. The user identification information is an item that is set to a unique value for each user. The user attribute data is an item that stores a character string that indicates the user's gender, age, occupation, and other arbitrary user attributes. The user attribute data includes the user's unique data (unique information), such as the user's gender and age.

[0018] The behavior table 1013 is a table for storing and managing information related to behavior (behavior information). The behavior table 1013 is a table having columns for user ID, behavior data, and date and time. FIG. 6 is a diagram showing the data structure of the behavior table 1013.

[0019] The user ID is an item that stores user identification information for identifying a user. The behavioral data is an item that stores information (behavioral information) about a user's behavior at a specific date and time. The behavioral information includes information about the user's behavior, such as location information, at each of one or more specific dates and times. Specifically, the behavioral information includes information about spots that indicate the user's location at a specific date and time. For example, the behavioral information includes the following information: The behavioral data includes location information of the user at a specific date and time. The location information includes geographical information such as latitude, longitude, altitude, and polygon data. The location information includes any identification code assigned to each mesh-shaped geographical area divided by specified latitude and longitude. The location information includes an address and a national local government code assigned to each administrative division such as a city, ward, town, or village. The behavioral data includes facility information about facilities visited by the user at a specific date and time. Specifically, the behavioral data includes the names of the facilities visited by the user at a specific date and time and the location information of the facilities. For example, facility information about facilities may include Tokyo Tower, Tokyo Station, Haneda Airport, Tokyo International Exhibition Center, Ise Shrine, and the name and location information of specific stores such as ●● Ramen Funabashi Ekimae Branch. The behavioral data includes information about the type (genre) of facility visited by the user on a specific date and time. For example, the type of facility may include train stations, airports, shrines, art galleries, museums, and other information indicating the type of store, such as ramen shops or chain store names. The behavioral data includes information about events visited by the user on a specific date and time. For example, event information may include the name of an event such as a live performance by a specific artist A or Comic Market, and the location information of the event's location. The behavioral data includes the user's facility genre usage score and feature values. In addition to information about the user's behavior on a specific date and time, the behavioral data also includes information about the user's behavior before and after the specific date and time (hereinafter referred to as "before and after data"). The before and after data includes information about the user's behavior before and after a specific event. The date and time is a field that stores the date and time when new or updated behavioral data is stored in the behavior table. Instead of the date and time, character string information indicating a time such as morning, afternoon, or evening, and information such as a time period (1 o'clock, 2 o'clock) may be stored.Behavioral information can be stored at any time interval or granularity. Information about user behavior before and after a specific event, etc., also includes information about which facility genres the user visited before and after visiting a specific facility genre (check-in or check-out). Feature values ​​are values ​​compared to all users, and are high when usage is not high for all users but is high for a specific user group. On the other hand, general advertising fees are set according to overall usage. By using the features of a user group, advertising can be delivered relatively inexpensively by selecting keywords, locations, websites, and web pages that are effective for that user group. The population of all users can be all users acquired by the system, or each prefecture or a sampled population can be used as the population. A spot can refer to an information mesh unit expressed by latitude and longitude, polygon information, or an area within a radius of Xm.

[0020] The following statistical information can be created based on behavioral data: - Statistics on the population distribution of the spot (location information, facility information, facility genre, etc.) and the population distribution for each record before and after visiting the spot. - By using location information for a specific date and time and the facility genre information associated with that location information, statistics on facility genre features and statistics on facility genre features for each time period including data before and after visiting the spot. - Statistics on relationship values ​​for each spot (values ​​such as a combination of facility genre feature value x population share x population). - Statistics on relationship values ​​for each schedule spot (values ​​such as a combination of facility genre feature value x population share x population). - Statistics on population distribution based on behavioral data other than data before and after visiting the spot. - Statistics on facility genre feature value based on behavioral data other than data before and after visiting the spot. - Statistics on relationship values ​​for each spot based on behavioral data other than data before and after visiting the spot (values ​​such as a combination of facility genre feature value x population share x population). - Statistics on relationship values ​​for each time period other than data before and after visiting the spot. Multiple statistical information can be created from behavioral data, such as the above. These can have statistics for each of the first behavior data, representative behavior data, second behavior data, and cluster.

[0021] It is also possible to create statistics on behavior before and after a visit to a facility category, starting from checking in or checking out of that facility category. For example, when checking in to a hospital, the specific hospital is not taken into consideration. Regarding which facility categories are visited before and after that, the specific facilities are not taken into consideration.

[0022] While search and purchase information can only be obtained from search and shopping sites, location information can be obtained from any website. In other words, while there are only a limited number of websites that can match statistical data on search and purchase information, statistical data on location information can be matched with any website. Furthermore, general advertising distribution systems are based on the user's activity history on the website that displays the advertisement and cannot take into account user activity in the real world, making them unsuitable for analyzing and attracting users to tourist spots and stores. In other words, statistical data on location information, which is the user's activity history in the real world, can be used to analyze and attract users to tourist spots and stores in addition to traditional advertising activities.

[0023] <Example of calculating the relationship value between a cluster and a spot> When the values ​​for Bandya are: Feature quantity of the facility genre to which the spot belongs: 10 Population share of the spot: 30% Population of the spot: 120 The relationship value between Bandya and the spot is 360. When the values ​​for Otaku are: Feature quantity of the facility genre to which the spot belongs: 1 Population share of the spot: 40% Population of the spot: 160 The relationship value between Otaku and the spot is 64. If a user's location information obtained from a website is the spot in question, the cluster to which the user belongs is more likely to be Bandya than Otaku. Furthermore, the accuracy of cluster determination increases by obtaining multiple location information and accumulating the spot relationship values.

[0024] The theme table 1014 is a table for storing and managing information (theme information) related to themes related to a user group consisting of multiple users. Themes include information related to the user group's interests, concerns, and hobbies. For example, themes include character string information indicating that the user group has interests, concerns, and hobbies related to railways, anime, cosplay, and a specific artist A (e.g., anime otaku, cosplay otaku). Themes may also include character string information indicating the user's gender, age, occupation, or other user attributes. The theme table 1014 is a table with a theme ID as a primary key and columns for theme ID, theme name, theme condition, first user group, first behavioral data, and second cluster ID. FIG. 7 illustrates the data structure of the theme table 1014.

[0025] The theme ID is a field that stores theme identification information for identifying a theme. The theme identification information is a field that is assigned a unique value for each piece of theme information. The theme name is a field that stores the name of the theme. Any character string can be set for the theme name. The theme condition is a field that stores conditions for identifying a user group associated with a theme. Specifically, in the present disclosure, information that specifies conditions related to the behavioral information stored in the behavior table is stored. The theme condition stores information that specifies conditions related to behavioral data and date and time. For example, information that specifies demographic data by gender and age, administrative districts such as cities, wards, towns, and villages, weekdays, holidays, time periods, and the Rokuyo calendar, or information that specifies the behavior of being in a specific geographic space (Tokyo Big Sight) on a specific date and time (August XX), etc., may be stored. Note that the theme condition may include a condition related to a combination of multiple pieces of behavioral data. For example, the theme condition may include multiple pieces of chronologically consecutive behavioral data, such as attending a concert and then going to a ramen restaurant. The first user group is a field that stores user IDs included in a user group associated with a theme. The first behavioral data is an item that stores representative (typical) behavioral data of a user group related to a theme. For example, moving from a convenience store to Tokyo Big Sight, a ramen shop, and an izakaya in that order. Note that the first behavioral data does not necessarily need to store a single predetermined behavioral data, but may be configured to store multiple representative (typical) behavioral data related to the user group. Furthermore, each of the multiple behavioral data may be associated with a predetermined score according to the degree of association with the first user group. The second cluster ID is an item that stores cluster identification information for identifying a cluster.

[0026] The cluster table 1015 is a table for storing and managing information about clusters (cluster information). The cluster table 1015 has columns for cluster ID, user group, and representative behavior data, with the cluster ID as the primary key. FIG. 8 is a diagram showing the data structure of the cluster table 1015.

[0027] The cluster ID is an item that stores second cluster identification information for identifying the second cluster. The second cluster identification information is an item in which a unique value is set for each piece of second cluster information. The user group is an item that stores user IDs included in a user group associated with the second cluster identification information. The representative behavior data is an item that stores representative (typical) behavior data of a user group associated with the second cluster identification information. Note that the representative (typical) behavior data is similar to the first behavior data, and therefore a description thereof will be omitted.

[0028] Advertisement table 1021 is a table for storing and managing information related to advertisements (advertisement information). Advertisement table 1021 is a table having columns for advertisement ID, advertisement data, and advertisement theme ID, with advertisement ID as the primary key. FIG. 9 is a diagram showing the data structure of advertisement table 1021.

[0029] The advertisement ID is an item that stores advertisement identification information for identifying advertisements. The advertisement identification information is an item in which a unique value is set for each advertisement information. The advertisement data is an item that stores information about advertisements. Specifically, the advertisement data may include information about the interests of a user or cluster, attribute information of the user or cluster, and unique information. The advertisement data also includes information such as the text of the advertisement to be delivered and an avatar image. For tourist spots and brick-and-mortar stores, the advertisement data also includes location information such as the latitude and longitude of the spot. The advertisement theme ID is an item that stores theme identification information associated with users to whom the advertisement is to be delivered. Specifically, the advertisement theme ID is an item that stores theme identification information for specifying a theme stored in the theme table 1014. This item enables targeted advertising that narrows down the users to whom the advertisement is to be delivered.

[0030] <Configuration of the control unit 104 of the server 10> The control unit 104 of the server 10 includes a user registration control unit 1041. The control unit 104 executes the application program 1011 stored in the storage unit 101, thereby realizing each functional unit.

[0031] The user registration control unit 1041 performs a process of storing information about users who wish to use the service disclosed herein in the user table 1012. The information stored in the user table 1012 is generated when a user opens a web page or application operated by the service provider from an information processing terminal, enters information into a predetermined input form, and transmits the information to the server 10. The user registration control unit 1041 stores the received information in a new record in the user table 1012, completing user registration. This allows the user stored in the user table 1012 to use the service. Prior to the user registration control unit 1041 registering the user information in the user table 1012, the service provider may conduct a predetermined screening process to restrict whether or not the user can use the service. The user ID may be any string or number that can identify the user. It may be any string or number desired by the user, or the user registration control unit 1041 may automatically set any string or number. User registration is not required in this system.

[0032] <Configuration of User Terminal 20> The user terminal 20 is an information processing device operated by a user who uses a service. The user terminal 20 may be, for example, a mobile terminal such as a smartphone or tablet, a stationary personal computer (PC), or a laptop PC. It may also be a wearable terminal such as a head-mounted display (HMD) or a wristwatch-type terminal, or a device capable of acquiring location information such as a car navigation system. The user terminal 20 may also be a terminal used exclusively for video distribution consisting only of a display such as digital signage. The user terminal 20 includes a storage unit 201, a control unit 204, an input device 206, and an output device 208. If the user terminal 20 only has the function of displaying received information, it may not include the storage unit 201 or the input device 206. For example, this may be the case for digital signage.

[0033] To obtain a user's location information, information such as a Global Positioning System (GPS) installed on a user's smartphone, smartwatch, or the like may be used. Alternatively, location information may be obtained by installing a terminal such as a magnetic reader or IC reader and having the user have the terminal read the user's own device. Furthermore, location information obtained from a base station of a communication device such as a mobile phone or from Wi-Fi (Wireless Fidelity, registered trademark) may also be obtained. Location information estimated from an IP address may also be used.

[0034] <Configuration of Storage Unit 201 of User Terminal 20 > The storage unit 201 of the user terminal 20 includes a user ID 2011 and an application program 2012 .

[0035] The user ID 2011 is the user's account ID. The user transmits the user ID 2011 from the user terminal 20 to the server 10. The server 10 identifies the user based on the user ID 2011 and provides the user with the services according to the present disclosure. The user ID 2011 includes information such as a session ID temporarily assigned by the server 10 to identify the user using the user terminal 20.

[0036] The application program 2012 may be stored in advance in the storage unit 201, or may be downloaded from a web server operated by a service provider via a communication IF. The application program 2012 includes an application such as a web browser application. The application program 2012 includes a programming language such as JavaScript (registered trademark) that is executed on the web browser application stored in the user terminal 20.

[0037] <Configuration of control unit 204 of user terminal 20> The control unit 204 of the user terminal 20 includes an input control unit 2041 and an output control unit 2042. The control unit 204 executes an application program 2012 stored in the storage unit 201, thereby realizing each functional unit.

[0038] <Configuration of Input Device 206 of User Terminal 20> The input device 206 of the user terminal 20 includes a camera 2061 , a microphone 2062 , a position information sensor 2063 , a motion sensor 2064 , and a touch device 2065 .

[0039] <Configuration of Output Device 208 of User Terminal 20 > The output device 208 of the user terminal 20 includes a display 2081 and a speaker 2082 .

[0040] <Configuration of Administrator Terminal 30> The administrator terminal 30 is an information processing device operated by an administrator who manages services. The administrator terminal 30 may be, for example, a mobile terminal such as a smartphone or tablet, or may be a stationary personal computer (PC) or laptop PC. It may also be a wearable terminal such as an HMD (Head Mount Display) or a wristwatch terminal. The administrator terminal 30 includes a storage unit 301, a control unit 304, an input device 306, and an output device 308.

[0041] <Configuration of Storage Unit 301 of Administrator Terminal 30 > The storage unit 301 of the administrator terminal 30 includes a user ID 3011 and an application program 3012 .

[0042] The user ID 3011 is the user's account ID. The user transmits the user ID 3011 from the administrator terminal 30 to the server 10. The server 10 identifies the user based on the user ID 3011 and provides the user with the services disclosed herein. The user ID 3011 includes information such as a session ID temporarily assigned by the server 10 to identify the user using the administrator terminal 30.

[0043] The application program 3012 may be stored in advance in the storage unit 301, or may be downloaded from a web server operated by a service provider via a communication IF. The application program 3012 includes a programming language such as JavaScript (registered trademark) that is executed on a web browser application stored in the administrator terminal 30.

[0044] <Configuration of control unit 304 of administrator terminal 30> The control unit 304 of the administrator terminal 30 includes an input control unit 3041 and an output control unit 3042. The control unit 304 executes an application program 3012 stored in the storage unit 301, thereby realizing each functional unit.

[0045] <Configuration of Input Device 306 of Administrator Terminal 30 > The input device 306 of the administrator terminal 30 includes a camera 3061 , a microphone 3062 , a position information sensor 3063 , a motion sensor 3064 , and a keyboard 3065 .

[0046] <Configuration of Output Device 308 of Administrator Terminal 30 > The output device 308 of the administrator terminal 30 includes a display 3081 and a speaker 3082 .

[0047] <Operation of System 1> Each process of System 1 will be described below. Fig. 10 is a flowchart showing the operation of cluster processing. Fig. 11 is a flowchart showing the operation of user extraction processing. Fig. 12 is a flowchart showing the operation of advertisement delivery processing. Fig. 13 is an example screen showing the operation of advertisement delivery processing. Note that some or all of the information processing of the cluster processing, user extraction processing, and advertisement delivery processing according to the present disclosure may be realized using an artificial intelligence system such as generation AI.

[0048] <Clustering> Clustering is a process of dividing a plurality of users into groups (clustering) according to the similarity of their behavioral information.

[0049] <Overview of Cluster Processing> Cluster processing is a series of processes that store behavioral information of multiple users, perform clustering processing based on the similarity of the stored behavioral information of the users, and assign and store a cluster ID to each user to group them. Alternatively, it may be possible to only assign a similarity.

[0050] <Details of Cluster Processing> Details of cluster processing will be described below.

[0051] In step S101, the control unit 104 of the server 10 executes a behavior acquisition step of acquiring behavior data related to the time-series behavior of multiple users. Specifically, the control unit 104 of the server 10 acquires information (behavior data) related to user behavior based on location information and the like periodically received from the user terminal 20. Details of the behavior data are the same as those described for the behavior data items in the behavior table 1013. The behavior data can be identified based on information (date and time and location information) transmitted by the user terminal 20, by referring to a table (not shown) in which location information is associated with information such as addresses, facility information, information about facility types (genres), and event information. Furthermore, the behavior data may be identified based on information from a wireless base station or the like (including a mobile base station or the like that provides the user terminal 20 with mobile wireless communication functions such as LTE, 4G, and 5G) that is located at a predetermined location where the user terminal 20 is located and that can wirelessly connect to the user terminal 20, rather than based on information transmitted by the user terminal 20. For example, location information estimated by a wireless base station or the like may be acquired as the location information of the user terminal 20. The control unit 104 of the server 10 may collect user behavior data in step S101 periodically, constantly, or all at once, regardless of the processing from step S102 onwards in the cluster processing, and the timing of collection is not important. The control unit 104 of the server 10 may also be configured to obtain user behavior data from another information processing service without directly collecting it. Specifically, steps S101 to S103 of the cluster processing may be configured to be executed independently and asynchronously (for example, at different frequencies).

[0052] In step S102, the control unit 104 of the server 10 calculates feature values ​​for each facility genre for all users using all of the behavior data acquired in the behavior acquisition step. Specifically, the control unit 104 of the server 10 calculates facility genre importance and feature values ​​based on the behavior data of multiple users acquired in step S101. Specifically, the user's behavior data can be viewed as array data spanning a predetermined period, such as every hour, one day, one week, one month, or one year. For example, assuming that information about the user's behavior (location information, facility name, facility type, numerical values ​​such as event name, text information, and category information) can be acquired for each hour of a 24-hour day (today's behavior data), the user's behavior data for a specific day can be viewed as array elements [P1, P2, ..., P24] having 24 elements (24 hours). The control unit 104 of the server 10 may determine, as the daily activity data for all users, representative values ​​such as the mode, average, median, sum, difference, weighted sum, weighted difference, etc. (mode, average, median, sum, difference, weighted sum, weighted difference of array elements) of the daily activity data for all users over a predetermined period, such as one week, one month, or one year, values ​​obtained by detecting and excluding outliers, or values ​​obtained by excluding facility genres with a high co-occurrence rate between facility genres. For example, the daily activity data for all users may be a typical behavior pattern (e.g., frequently moving to facility A, facility B, facility C, etc.) typical of all users. The control unit 104 of the server 10 may also vectorize the daily activity data for all users using any machine learning, deep learning, artificial intelligence model, etc. In each process of the present disclosure, behavior data related to daily activity data for 24 hours of one day will be mainly described as an example, but the present disclosure can also be applied to array data for any predetermined period, such as one week, one month, or one year.

[0053] In step S102, the control unit 104 of the server 10 executes a grouping step of classifying multiple users into multiple groups based on the similarity of the behavioral data acquired in the behavior acquisition step. Specifically, the control unit 104 of the server 10 executes clustering based on the similarity of the time-series behavioral data of multiple users acquired in step S101. Specifically, the user's behavioral data can be viewed as array data spanning a predetermined period, such as every hour, one day, one week, one month, or one year. For example, assuming that information about the user's behavior (location information, facility name, facility type, numerical values ​​such as event name, text information, and category information) can be acquired for each hour of a 24-hour day (daily behavior data), the user's behavioral data for a particular day can be viewed as array elements [P1, P2, ..., P24] having 24 (24-hour) elements. Alternatively, the period of time spent traveling from home to home can be considered as the current day. The control unit 104 of the server 10 may determine the user's daily activity data (referred to as "user daily activity data") as a representative value such as the mode, average, median, sum, difference, weighted sum, or weighted difference (the mode, average, median, sum, difference, weighted sum, or weighted difference of the array elements) of the user's daily activity data over a predetermined period such as one week, one month, or one year, a value obtained by detecting and excluding outliers, or a value obtained by excluding facility genres with a high co-occurrence rate between facility genres. For example, the user's daily activity data may be a typical user behavior pattern (e.g., frequently moving to facility A, facility B, facility C, etc.). The control unit 104 of the server 10 may also vectorize the user's daily activity data using any machine learning, deep learning, artificial intelligence model, or the like. Specifically, based on the behavior data of multiple users acquired in step S101, appropriate processing such as vectorization may be performed to calculate the similarity between the behaviors of each user. In addition, the control unit 104 of the server 10 can apply any normalization method to the obtained vectors to execute a process of extracting the similarity between the vectors of multiple users, a vector (representative vector) that characterizes each of the vectors of the users, etc. Also, the facility genre importance of each user may be calculated and a feature value compared with the overall importance may be provided.In each process of the present disclosure, behavioral data relating to daily behavioral data for 24 hours per day will be mainly described as an example, but the present disclosure can also be applied to sequence data spanning any predetermined period such as one week, one month, one year, etc. Furthermore, it may be determined whether the user works, lives, or visits a location based on the length of stay, frequency of visits, usual behavior, etc., and the occupation type may be determined and stored by combining the determination of work status with the facility genre.

[0054] In step S102, the control unit 104 of the server 10 calculates the similarity of vectors related to the behavioral data calculated for each of multiple users. Specifically, the control unit 104 of the server 10 calculates the similarity value between vectors related to the behavioral data using the inverse of Euclidean distance, Manhattan distance, or other methods such as cosine similarity. As a result, the control unit 104 of the server 10 calculates a similarity matrix (the inverse of the distance matrix) between multiple users. Based on the similarity matrix, the control unit 104 of the server 10 clusters multiple users according to the similarity of their behavioral data. For example, clustering algorithms include k-means (kmeans). By performing the clustering process, cluster labels such as cluster IDs can be obtained for each user who is similar to the other users. Users with the same cluster ID are considered to have similar behavioral data (high similarity in behavioral data, close distance). Facility genre values ​​that indicate unreliable values ​​may also be excluded. For example, if the average number of spots that served as the source of the feature quantity of facility genre A is close to 1, it may be a value indicating that the user was at that location by chance, and is therefore less reliable than a feature quantity consisting of multiple spots. If the average number of spots that served as the source of the feature quantity of facility genre A for a certain user group is close to 1, and the average number of spots that served as the source of the feature quantity of facility genre A for another user group with high feature quantities of facility genre A is farther than 1, this means that a facility genre that normally involves the use of many stores is being used only by this user group, and the reliability is low. If the average number of spots that served as the source of the feature quantity of facility genre A is close to 1, and another facility genre B with high feature quantities is located in the same location and the average number of spots that served as the source of the feature quantity of facility genre B is farther than 1, the feature quantity of facility genre A is likely to be influenced by facility genre B, and the reliability is low. By comparing the co-occurrence rates of the locations of spots of facility genre A and facility genre B, if the rate at which facility genre B occurs when facility genre A is present is higher than the rate at which facility genre A occurs when facility genre B is present, the feature quantity of facility genre A is likely to be influenced by facility genre B, and the reliability is low.If facility genre A is a toy store (one store within the aggregation range) and facility genre B is a department store (three stores within the aggregation range), and the rate at which facility genre B is also present when facility genre A is present is 100%, and the rate at which facility genre B is also present when facility genre A is present is 33%, then facility genre A is likely to be subordinate to facility genre B, and the feature value of facility genre A is unreliable. If the co-occurrence rate of facility genre A and facility genre B is low when viewed from the overall or other user group usage behavior, and the co-occurrence rate is high only for the relevant user group, then it is likely to be influenced by facility genre B, which is subordinate, and therefore unreliable. If the co-occurrence rate of the locations of spots viewed from both facility genre A and facility genre B is 100%, it is best to consider them to be effectively the same facility. In this way, facility genres that show unreliable values ​​can be identified using the "average number of spots that are the source of feature values" and the "co-occurrence rate between facility genres," and facility genres that meet certain conditions can be excluded from the representative behavior data. Furthermore, when multiple clusters with high similarity are determined, a user may belong to the cluster with the highest similarity, or to multiple clusters with high similarity. The control unit 104 of the server 10 calculates representative values ​​such as the mode, average, median, sum, difference, weighted sum, weighted difference, etc. (mode, average, median, sum, difference, weighted sum, weighted difference of array elements) of the behavior data of multiple users with the same cluster ID (users belonging to the same group) as representative behavior data of the group. When calculating the representative behavior data, outliers may be detected and excluded, or values ​​of facility genres with high co-occurrence rates may be excluded. Additionally, any processing may be performed so that the representative behavior data represents typical behavior of a group of users. Furthermore, the representative behavior data may be calculated from clusters that prioritize rare facility genres and carefully select users with many commonalities. For example, representative behavioral data calculated based on the behavioral data of other clusters that have a high similarity between the first behavioral data and behavioral data related to facility genres, or other clusters that have been determined to have a correlation of a predetermined value or more with the results of pre-clustering (such as anime otaku or train otaku), may be used as the representative behavioral data.Alternatively, representative behavior data may be calculated using facility genre statistical feature values ​​that are lower than the standard. In this case, the cluster will be one in which users have not visited that facility genre, so there is a possibility of errors if users have visited that facility genre, but there is no error if users have not visited that facility genre, so the reliability is high. Also, feature values ​​may be weighted. Machine learning, deep learning, or any artificial intelligence model may be used to calculate representative behavior data. Furthermore, the representative behavior data may be one with a high total value of each user's feature values, or one with a large number of users may be used as the representative value. Even if the similarity in facility genre importance is high, if the variance within the cluster is high, it may not be selected as representative behavior data.

[0055] The control unit 104 of the server 10 stores the cluster ID, the user ID of the user having the cluster ID, and representative behavior data calculated from the behavior data of the user having the cluster ID in the cluster ID, user group, and representative behavior data fields of a new record in the cluster table 1015. This makes it possible to divide multiple users into multiple groups (user groups) based on the similarity of their behavior data. <User Extraction Process> The user extraction process is a process of identifying a first user group related to the input theme information and a second cluster whose behavior information is similar to that of the first user group.

[0056] <Overview of user extraction process> The user extraction process is a series of processes that accepts input of theme information, identifies a first user group consisting of multiple users related to the theme information, identifies representative first behavioral data of the first user group, identifies one or more clusters that have behavioral data similar to the first behavioral data, and stores the clusters in association with the first user group.

[0057] <Details of User Extraction Processing> Details of the user extraction processing will be described below.

[0058] In step S301, the control unit 104 of the server 10 executes a theme reception step of receiving theme information related to a theme. An administrator, such as an employee of a business providing the information processing service according to the present disclosure, operates the input device 306 of the administrator terminal 30 to input the URL of a page (theme reception processing page) for executing the theme reception processing into a web browser or the like, and open the theme reception processing page. The control unit 304 of the administrator terminal 30 sends a request to open the theme reception processing page to the server 10. Based on the received request, the control unit 104 of the server 10 generates a theme reception processing page and sends it to the administrator terminal 30. The control unit 304 of the administrator terminal 30 displays the received theme reception processing page on the display 3081 of the administrator terminal 30. The theme reception processing page includes an input field for receiving a theme name that identifies a theme related to the user's interests, concerns, and hobbies and preferences, and theme conditions for identifying the theme. The theme conditions are conditions for identifying a user group related to the theme. The theme conditions are similar to the description of the theme condition items in the theme table 1014.

[0059] The administrator operates the input device 306 of the administrator terminal 30 to input the theme name and theme conditions into the input fields of the theme reception processing page. The administrator operates the input device 306 of the administrator terminal 30 to send the theme name and theme conditions entered into the input fields to the server 10. The control unit 104 of the server 10 stores the received theme name and theme conditions in the theme name and theme condition fields of a new record in the theme table 1014.

[0060] In the user extraction process, steps S301 to S305 do not necessarily need to be executed consecutively. For example, the administrator may execute the process of step S301 in advance and store multiple themes in the theme table 1014. Furthermore, the control unit 104 of the server 10 may be configured to execute steps S302 to S305 asynchronously with the completion of step S301.

[0061] In step S302, the control unit 104 of the server 10 executes a first extraction step of extracting a first user group consisting of one or more users associated with a predetermined theme from among the multiple users based on the behavioral data acquired in the behavior acquisition step. Specifically, the control unit 104 of the server 10 references the theme table 1014 to acquire the theme condition items. For example, the control unit 104 of the server 10 may be configured to acquire a record of theme information for which the processing from step S302 onward has not been executed and for which information is not stored in at least one of the first user group, first behavioral data, and second cluster ID items. Based on the theme condition, the control unit 104 of the server 10 searches the behavioral data items in the behavior table 1013 and identifies multiple user IDs (users) whose behavioral data meets the theme condition as the first user group.

[0062] The first extraction step extracts a first user group related to an event held during a predetermined period. For example, assume that an event is held at a predetermined facility (facility A) on a predetermined date and time (date and time A). In this case, the control unit 104 of the server 10 searches the behavior data items in the behavior table 1013, searches for one or more users visiting facility A on date and time A, and identifies them as the first user group.

[0063] The first extraction step extracts a first user group based on the behavior of multiple users before and after a predetermined period of time, excluding one or more users not related to an event held during the predetermined period. Specifically, the theme conditions may include not only conditions that must be satisfied by the behavior data, but also conditions that the behavior data do not satisfy. For example, the theme conditions may include a condition that a user does not visit facilities B, C, etc., during a time period before or after visiting (checking in or checking out of) an event held on date and time A. In this case, the control unit 104 of the server 10 searches the behavior data items in the behavior table 1013 and, from one or more users who visited facility A on date and time A, excludes users who visited facilities B, C, etc. during a time period before or after the time period, thereby identifying the first user group.

[0064] For example, if an event is an artist event for minors, facility A, which is the venue for the event, may include multiple facilities such as a live music venue and a hospital. In this case, it is possible that multiple users visiting facility A have completely different interests when visiting the live music venue and the hospital. For example, let's say that a user visiting the live music venue is user A and a user visiting the hospital is user B, and there is behavioral data for when they visit the hospital. In this case, it is possible that user A and user B have in common the fact that they both visited facility A on a certain date and time A, and that they cannot be distinguished from the user behavioral data. Even in such a case, it is possible that the behavioral patterns of user A and user B, such as visiting facilities, will be significantly different between the time periods before and after their visit to facility A. For example, it is possible that user B visits facility B and facility C, such as a nearby elderly care facility, before and after their visit. In this case, if behavioral data showing a visit to a nursing home is available before and after the behavioral data showing a visit to a hospital, it is assumed that user B visited facility A on date and time A, but also visited a nursing home or other facility before and after the date and time. Therefore, it is assumed that the user did not attend the live venue, but visited facility A for other reasons, such as a hospital. Therefore, when identifying a first user group based on the theme of the artist event, user A is included in the first user group, and user B, who is assumed to have low relevance, is excluded from the first user group. This allows for more accurate identification of a first user group consisting of multiple users with uniform interests and concerns from abstract behavioral data. Furthermore, the clusters that visit hospitals or the behavior of the target clusters when visiting hospitals may be used as targets for user groups to be excluded. The exclusion process changes the population distribution and schedule of the user group, allowing for more effective delivery of advertisements to locations.

[0065] In step S303, the control unit 104 of the server 10 executes a representative acquisition step of acquiring representative behavior data representing the first user group extracted in the first extraction step. The representative acquisition step executes a step of acquiring representative behavior data representing the first user group for a period other than the predetermined period. Specifically, the control unit 104 of the server 10 calculates representative values ​​such as the mode, average, median, sum, difference, weighted sum, weighted difference, etc. (mode, average, median, sum, difference, weighted sum, weighted difference of array elements) of the behavior data of multiple users included in the first user group as the representative behavior data of the first user group. Note that, when calculating the representative behavior data of the first user group, outliers may be detected and excluded, or facility genres with a high co-occurrence rate among facility genres may be excluded. Alternatively, any processing may be performed so that the representative behavior data represents typical behavior of the user group. Furthermore, the representative behavior data of the first user group may be calculated from a cluster that prioritizes facility genres with high rarity and carefully selects users with many commonalities. For example, the representative behavior data may be calculated based on the behavior data of another cluster that has a high similarity between the first behavior data and behavior data related to facility genres, or based on the behavior data of another cluster that has been determined to have a correlation of a predetermined value or greater with the results of previous clustering (e.g., anime otaku or train otaku). Specifically, the representative behavior data may be a typical behavior pattern based on the behavior data of many users included in the first user group on that day. Furthermore, multiple representative behavior data may be calculated for the first user group. For example, the representative behavior data may be based on typical behavior of otaku event participants (first user group). Alternatively, multiple representative behavior data may be generated from otaku event participants based on typical behavior of anime otaku, train otaku, etc., and the combined representative behavior data may be used as the representative behavior data of the otaku event participants.

[0066] In step S304, the control unit 104 of the server 10 executes a second extraction step of extracting a second user group consisting of one or more users from the plurality of users based on the representative behavior data of the first user group. Specifically, the control unit 104 of the server 10 searches the cluster table 1015 for representative behavior data items based on the representative behavior data of the first user group calculated in step S303. For example, the control unit 104 of the server 10 retrieves a predetermined number of cluster information items having representative behavior data similar to the representative behavior data of the first user group (high similarity in behavior data, close distance) from the cluster table 1015. The control unit 104 of the server 10 identifies, as second cluster IDs, the cluster IDs of the predetermined number of cluster information items in the order of cluster information whose representative behavior similarity is equal to or greater than a predetermined value, and similarity between the representative behavior data items. This makes it possible to identify a second user group consisting of one or more users having behavior data similar to the representative behavior data of the first user group (one or more users included in the clusters corresponding to the second cluster IDs).

[0067] This allows for the extraction of a second user group (potential customers) that differs from the first user group but is expected to potentially exhibit similar behavioral patterns to the first user group based on the typical behavioral patterns (representative behavioral data) of the first user group related to a specific theme or event. For example, if the first user group is a group of users who attended a specific artist event, a second user group with similar daily behavioral data on days other than the event date is identified. In this case, the second user group is not necessarily a group of users who attended the specific artist event, but because their behavior is similar, it is expected that their interests, concerns, hobbies, and preferences are similar to those of the first user group. For example, it is expected that the purchasing behavior (interests, concerns, hobbies, and preferences in goods and services purchased) of the second user group and the first user group are similar.

[0068] In addition, the identification of the second user group does not necessarily require the use of cluster information identified by cluster processing; it is also possible to identify the second user group that is expected to exhibit behavior patterns similar to those of the first user group from multiple users who have behavior data similar to the representative behavior data of the first user group.

[0069] In addition, if the second user group includes the first user group, the second user group may be identified by excluding some or all of the first user group from the second user group. For example, after identifying potential customers (second user group) who visit a luxury car dealership and have behavioral data trends similar to those of luxury car buyers (first user group), a more suitable second user group can be identified by excluding the first user group who have already purchased luxury cars from the second user group. Based on the distribution of the second user group, advertisements aimed at attracting customers can be displayed to users who are likely to visit within a few hours, while advertisements aimed at promoting awareness can be displayed to users who are likely to visit within a period of, for example, six months. Furthermore, when matching using external context, it is possible to choose not to deliver advertisements to users who have visited a Lexus store because they are likely to be potential buyers.

[0070] The number of users (number of people) included in the second user group identified by the first user group may be greater or less than the number of users (number of people) included in the first user group. The second user group may be a part of the first user group. For example, the second user group may be identified from one or more user groups included in the first user group. Specifically, a second user group consisting of a part of the users included in the first user group may be identified by applying cluster processing to the users included in the first user group. Alternatively, the distance (similarity) between the behavioral data of each of the multiple users included in the first user group and the representative behavioral data may be calculated, and a user group whose distance is less than a predetermined value (whose similarity is greater than or equal to a predetermined value) may be identified as the second user group. In this case, the second user group, which is a part of the first user group and has a behavior pattern similar to the representative behavior data of the first user group, can be said to be a user group consisting of a part of users in the first user group who exhibit behavior data that is particularly typical of the first user group (if the first user group is otaku, then the second user group is a genuine otaku who exhibits behavior that is particularly typical of otaku among the otaku). In this way, by performing the user extraction process, it is possible to identify the second user group consisting of a part of users in the first user group who exhibit behavior data that is particularly similar to the representative behavior data of the first user group, from among the multiple user groups included in the first user group.

[0071] In cases where a certain facility type or spot is visited by the majority of users, a second user group (potential customers) can be generated by extracting users who have not visited that facility type or spot. The second user group created from the first user group who visited a certain facility type or spot is a user group that has the potential to visit that facility type or spot, but the user group that has not visited a certain facility type or spot may be a customer persona that has factors that prevent them from visiting that facility type or spot, and is effective as marketing data.

[0072] Each spot is assigned a relationship value with each user group, and user groups can be determined by obtaining location information from a website that identifies the spots where users have checked in and adding up the relationship values. For example, if a user has checked in at spots B and C and has a high relationship value with otaku, the user is determined to be part of the otaku cluster.

[0073] In step S305, the control unit 104 of the server 10 executes a second storage step of storing the second user group extracted in the second extraction step. Specifically, the control unit 104 of the server 10 stores the cluster ID of the identified second user group in the second cluster ID field of the theme table 1014. This identifies the first user group and the second cluster IDs of second clusters whose behavioral data is similar to that of the first user group (the second cluster ID identifies a user group by referring to the cluster IDs in the cluster table 1015, and the multiple user groups become the second user group).

[0074] <Advertisement Distribution Processing> The advertisement distribution processing is processing for distributing advertisements to user terminals of users.

[0075] <Overview of Advertisement Delivery Process> The advertisement delivery process is a series of processes that identify a second user group according to a set advertisement, acquire second user information related to the second user group, generate an advertisement corresponding to the second user group, and deliver the advertisement. Note that, although advertisement delivery is disclosed as an example in this disclosure, the information delivered to a user terminal is not limited to advertisements, and any information, such as recommendation information (text data, image data, audio data, video data, etc. describing recommended places, products, services, etc.), may be provided to the user. Furthermore, the user terminal may not have a function for transmitting user information including location information, and may be configured to only receive recommendation information, such as digital signage.

[0076] In the advertisement distribution process, steps S501 to S505 do not necessarily need to be executed consecutively. For example, it is preferable to execute the advertisement generation process in steps S501 to S504 collectively in advance for multiple pieces of advertisement information. In this case, in step S505, one or more advertisements that match the behavior of the user are distributed to the user who accessed the specified website.

[0077] <Details of Advertisement Distribution Processing> Details of the advertisement distribution processing will be described below.

[0078] In step S501, the control unit 104 of the server 10 executes an information input step of accepting input of information related to a user's interests from an information publisher. An administrator, such as an employee of a business providing the information processing service according to the present disclosure, operates the input device 306 of the administrator terminal 30 to input the URL of a page (advertisement delivery processing page) for executing ad delivery processing into a web browser or the like, and open the ad delivery processing page. The control unit 304 of the administrator terminal 30 sends a request to open the ad delivery processing page to the server 10. Based on the received request, the control unit 104 of the server 10 generates an ad delivery processing page and sends it to the administrator terminal 30. The control unit 304 of the administrator terminal 30 displays the received ad delivery processing page on the display 3081 of the administrator terminal 30. The ad delivery processing page includes theme identification information that identifies a theme related to the user's interests, concerns, hobbies, and preferences, and input fields that can accept input of text, images, etc. to be used in advertisements to be delivered to users who are interested in the theme. The theme identification information is a theme ID in the theme table 1014. The control unit 104 of the server 10 may search the theme table 1014 and display the acquired theme ID and theme name in a selectable manner on the display 3081 of the administrator terminal 30. For example, theme identification information and theme names related to themes such as "Otaku," "Visitors to the Hakata Ramen Higashi Station Store," and "Customers Eating Ramen for Dinner" may be displayed in a list on the display 3081 of the administrator terminal 30 in a selectable manner by the administrator.

[0079] The administrator can select one or more theme IDs from the displayed list by operating the input device 306 of the administrator terminal 30. This makes it possible to provide information such as advertisements according to themes such as the user's interests, concerns, hobbies, and preferences, as well as user attributes such as gender and age.

[0080] The control unit 304 of the administrator terminal 30 transmits one or more pieces of theme identification information and advertisement data input by the administrator to the server 10. The control unit 104 of the server 10 stores the received one or more pieces of theme identification information and advertisement data in the advertisement theme ID and advertisement data fields of a new record in the advertisement table 1021.

[0081] 13 illustrates an example of the advertisement delivery processing page D1. The advertisement delivery processing page D1 includes avatars D11 visually depicting the personalities of a user group who have interests, concerns, hobbies, and preferences in a selected theme, visited places D12 indicating a breakdown of facilities visited by the user group, facility genres, and the like, attributes D13 indicating a breakdown of the user attribute data of the user group, similarities D14 indicating the similarities between the user group and user groups related to other themes, and interests D15 indicating the interests, concerns, hobbies, and preferences of the user group. In the present disclosure, an example of an advertisement delivery processing page D1 for a user group consisting of multiple users identified by one or more theme identification information stored in the advertisement theme ID field of the advertisement table 1021 with reference to the theme table 1014 will be described as an example, but is not limited thereto. For example, the present invention may be applied to presenting information on a user group consisting of any number of users, such as an arbitrary user group in the cluster table 1015 stored in the cluster processing, a first user group in the theme table 1014 stored in the user extraction processing, or a second user group identified by referencing the user group in the cluster table 1015 using a second cluster ID expected to exhibit a behavior pattern similar to that of the first user group. Furthermore, advertisements may be delivered based on keywords associated with themes. For example, when keywords such as anime, trains, and karaoke are associated with the theme of "otaku," an advertisement may be delivered by selecting only karaoke, or conversely, an advertisement may be delivered excluding karaoke. Furthermore, the determination of the theme does not need to be made in real time; the theme may be determined using location information of previously visited locations.

[0082] <Configuration of Avatar D11> The control unit 104 of the server 10 executes a persona creation step of creating a person image that characterizes a predetermined user group based on the behavioral data of multiple users included in the predetermined user group. The control unit 104 of the server 10 executes a persona presentation step of presenting the person image created in the persona creation step. Specifically, the control unit 104 of the server 10 searches an avatar database (not shown) based on the behavioral data of multiple users included in the predetermined user group, user attribute data, representative behavioral data of the predetermined user group, and representative values ​​(such as mode, average, median, maximum, and minimum) of the user attribute data of multiple users. Note that the search may be performed from behavioral data that detects outliers and removes them, values ​​that remove facility genres with a high co-occurrence rate between facility genres, or behavioral data that prioritizes rare facility genres and selects users with many commonalities. The search may be based on other clusters with high similarity between the first behavior data and behavior data related to facility genres, behavior data of clusters with high similarity when excluding facility genres with high co-occurrence rates, or behavior data of other clusters determined to have a predetermined or higher correlation with the results of pre-clustering (e.g., anime otaku, train otaku). The avatar database stores each feature of a character in a digital space (the character does not necessarily have to resemble a human), such as a human face, eyes, nose, hairstyle, body shape, and clothing, in association with behavior data, user attribute data, etc. The control unit 104 of the server 10 refers to the avatar database based on the behavior data, user attribute data, etc. of a predetermined user group to identify the features (features) that make up the avatar, such as a human face, eyes, nose, hairstyle, body shape, and clothing. The control unit 104 of the server 10 creates a human image that characterizes the predetermined user group by combining the identified features that make up the avatar. The human image may be created using a learning model, such as machine learning, deep learning, or artificial intelligence model, using the behavior data, user attribute data, etc. of the predetermined user group as input data. For example, this can be realized by training a learning model in advance so that a predetermined person image is output as output data for input data such as behavior data of a user group and user attribute data.Alternatively, behavioral data, user attribute data, etc. of a predetermined user group may be used as input data and applied to the generation AI to create a person image that characterizes the predetermined user group. The control unit 104 of the server 10 outputs the created person image to the avatar D11. For example, if the predetermined user group is a group of users with a theme related to train enthusiasts, an image of a person wearing glasses, a polo shirt with cargo pants, and sneakers may be registered as a train enthusiast, and the avatar D11 may be output when the person has a high similarity to a train enthusiast.

[0083] A person image, i.e., an avatar, that defines the person's image is assigned to a portion of the second user group. In other words, the number of users assigned avatars is equal to or less than the number of users assigned to the second user group. Naturally, the same avatar will be displayed for a group of users with similar representative behavior data. Creating a small number of avatars as person images for a large group of users with similarity reduces creation costs. The second user group to assign avatars to is preferably one with a common image in which the public's interest and appearance correlate, such as participants in an otaku event. User groups that can be extracted using location information are groups of users who gather at specific real-world locations, such as events or stores, and it is easy to form an image of their appearance by actually seeing them. On the other hand, in the case of search information, for example, it is difficult to evoke a common image of appearance from a group of users who searched for a specific keyword. If an event is considered a facility genre that appears only a few days a year, the rarity of this facility genre becomes extremely high, and its frequency is concentrated among a certain number of users. In other words, specific images that correlate with interests and appearance are more likely to be recognized by the general public in users of facility genres where the rarity x frequency value is high and users are extremely concentrated rather than dispersed, as well as in users who have such facility genres as co-occurring genres. This facilitates implicit understanding of the interests and appearances of users other than the user group who sets the avatar, without the need to interpret the numerical values ​​of each data. Images of people that define a person's image can be illustrations, photographs, or 3D, and can even have each part separated. For example, photos of people taken on the date and time of a specific event or illustrations based on such photos are desirable, and statistically more common are even more desirable. The base of the avatar can be changed based on gender and age, or separate images can be prepared for each gender and age. Images of people that define a person's image can be set by reconstructing statistical combinations from data from services that allow users to set a general avatar with each part separated and post it linked to location information, event information names, etc., or data obtained by morphological analysis of each part of images from fixed-point cameras, etc.It is desirable that the statistical combination of each feature be common in that area or event compared to the overall population. Similar to the concept of facility genre features, for example, an Akihabara-style appearance is not the majority appearance in Akihabara, but rather an appearance / clothing that is relatively common or only seen in Akihabara. When displaying the avatar of a user group setting an avatar as the avatar of another user group based on similarity, each feature may be changed based on the feature of the facility genre of the other user group. For example, some facility genres related to appearance, such as tanning salons, nail salons, tattoo shops, and contact lenses, may be associated with each feature of the avatar. In other words, even when displaying an otaku avatar to a user group with a high similarity to otaku, changes may be made, such as darkening the skin color if the feature of tanning salons is high and removing glasses if the feature of contact lenses is high. When displaying the avatar of a user group setting an avatar as the avatar of another user group based on similarity, each feature may be changed based on the similarity with the user group setting the other avatar. For example, if the user group for which avatars are assigned is set such that otaku tend to wear clothes in dull colors and dull designs, and delinquents tend to wear clothes in bright colors and dull designs, when displaying an otaku avatar to a user group with a high similarity to otaku, if the next most similar user group is delinquents, the avatar may be changed to reflect the characteristics of each group, such as dull colors and dull designs, or bright colors and dull designs. This allows images that better match the characteristics of each user group to be displayed without preparing many images. It is not necessary to assign one avatar to each user group; multiple avatars may be assigned. Furthermore, each user group is distributed to various locations, and even if they are the same otaku, clothing and backgrounds may be displayed that are appropriate for the location, such as clothing for attending an event or outdoor camping.

[0084] <Configuration of visited place D12> The control unit 104 of the server 10 executes a behavior aggregation step of aggregating the number of predetermined one or more behavioral data included in the behavioral data of a plurality of users included in a predetermined user group. The control unit 104 of the server 10 executes a behavior presentation step of presenting an index value based on the number of predetermined one or more behavioral data aggregated in the behavior aggregation step, in association with the predetermined one or more behavioral data.

[0085] The behavior aggregation step executes a step of aggregating the number of facility genres related to one or more predetermined facility types included in the behavior data of multiple users included in a predetermined user group. The behavior presentation step executes a step of presenting an index value based on the number of the predetermined one or more facility genres aggregated in the behavior aggregation step in association with the predetermined one or more facility genres.

[0086] The behavior aggregation step executes a step of aggregating the number of predetermined one or more facilities included in the behavior data of a plurality of users included in a predetermined user group, and the behavior presentation step executes a step of presenting an index value based on the number of predetermined one or more facilities aggregated in the behavior aggregation step in association with the predetermined one or more facilities.

[0087] Specifically, the control unit 104 of the server 10 tabulates a breakdown of the behavioral data possessed by multiple users included in the user group. In this case, the behavioral data to be tabulated (predetermined behavioral data) may be a single behavioral data from a specific time period, or a pattern of continuous behavioral data spanning multiple time periods (e.g., a case where users travel to facilities A, B, and C in that order over time periods A, B, and C). Tabulation may also be based on information regarding any behavior defined in the behavioral data item of the behavior table 1013 (location information, facility, facility genre, store, store name, event, etc.). For example, a tabulation result may be obtained in which A% of users visit facility A, B% of users visit facility B, C% of users visit facility genre C, and D% of users visit facility genre D. The control unit 104 of the server 10 may also be configured to calculate a predetermined index value (score) based on the tabulation result. In this case, the index value indicates the degree of association between a specific user group and the predetermined behavioral data, the importance of the behavioral data, etc. The control unit 104 of the server 10 associates the counting result with each piece of behavior data and outputs it to the visited place D12.

[0088] <Output example 1 of visited place D12 (facility genre)> Karaoke box 42 points Computer sales and repair shop 32 points CD and DVD shop 48 points Anime shop 64 points Card shop 42 points Bar and club 12 points Family restaurant 22 points Convenience store 43 points Amusement facility 82 points

[0089] The output of visited places D12 may be configured to be hierarchically subdivided by facility genre and facility, and output hierarchically. <Output example 2 of visited places D12> Convenience stores Seven-Eleven 44 points Family Mart 28 points Lawson 32 points Amusement facilities Aquariums 42 points Art museums 22 points Movie theaters 12 points

[0090] The control unit 104 of the server 10 executes an attribute aggregation step of counting the number of one or more pieces of predetermined attribute information included in attribute information related to the attributes of multiple users included in a predetermined user group. The control unit 104 of the server 10 executes an attribute presentation step of presenting an index value based on the number of the predetermined one or more pieces of attribute information aggregated in the attribute aggregation step, in association with the predetermined one or more pieces of attribute information. Specifically, the control unit 104 of the server 10 aggregates a breakdown of the user attribute data (e.g., gender, age, place of residence) possessed by multiple users included in the user group. For example, an aggregation result may be obtained such as A% of users being male, B% of users being female, C% of users being in their 20s, and D% of users being in their 30s. The control unit 104 of the server 10 may also be configured to calculate a predetermined index value (score) based on the aggregation result. In this case, the index value indicates the degree of association between a specific user group and the predetermined user attribute data, the importance of the user attribute data, etc. The control unit 104 of the server 10 outputs the breakdown of the aggregated user attributes to attribute D13.

[0091] <Attribute D13 output example 1 (gender)> ・Male 72 points ・Female 28 points <Attribute D13 output example 2 (age group)> ・10-19 years old 14 points ・20-29 years old 18 points ・30-39 years old 22 points ・40-49 years old 21 points <Attribute D13 output example 3 (place of residence)> ・Kanto 42 points ・Tohoku 18 points ・Chubu 10 points ・Kansai 5 points

[0092] <Configuration of Similarity D14> The control unit 104 of the server 10 executes a similarity calculation step of calculating similarities between a predetermined user group and multiple user groups. The similarity calculation step executes a step of calculating similarities between behavioral data of multiple users included in the predetermined user group and behavioral data of multiple users included in each of the multiple user groups. The control unit 104 of the server 10 executes a similarity presentation step of presenting the similarities for each of the multiple user groups calculated in the similarity calculation step in association with a theme related to the user group. Specifically, the control unit 104 of the server 10 calculates similarities between the representative behavior data of the predetermined user group and first behavior data of theme information stored in the theme table 1014, the representative behavior data stored in the cluster table 1015, or representative behavior data of users related to a theme identified by referring to the theme table 1014 based on the advertising theme ID of another theme stored in the advertisement table 1021. The method for calculating the similarity of behavioral data is similar to the method for calculating the similarity of behavioral data described in the cluster processing and user extraction processing, and therefore will not be described here. Alternatively, the similarity of an arbitrary user group consisting of multiple users to a predetermined user group may be calculated. The control unit 104 of the server 10 may also be configured to calculate a predetermined index value (score) based on the similarity. In this case, the index value indicates the similarity between the predetermined user group and other user groups. This allows the similarity between the representative behavior data of the predetermined user group and the representative behavior data of user groups included in other theme information and advertising information, i.e., the similarity between the behavior patterns of multiple users included in each user group, to be calculated. The control unit 104 of the server 10 associates the calculated similarity with the theme name, theme ID, advertising ID, etc., and outputs it to the similarity D14 of the advertisement delivery processing page D1.

[0093] <Example 1 of output of similarity D14> Overseas travel 64 points Outdoor festival 28 points Leisure 12 points Car 82 points Otaku 24 points

[0094] <Configuration of Interest D15> The control unit 104 of the server 10 executes an interest presentation step of presenting interest information, which is an index value indicating the interests of multiple users included in a predetermined user group, based on the behavioral data of multiple users included in the predetermined user group. Specifically, the control unit 104 of the server 10 compiles a breakdown of the behavioral data of multiple users included in the user group. In this case, the behavioral data to be compiled (predetermined behavioral data) may be a single behavioral data from a specific time period, or a pattern of continuous behavioral data spanning multiple time periods (e.g., moving to facilities A, B, and C in that order over time periods A, B, and C). Alternatively, the data may be compiled based on information related to any behavior defined in the behavioral data item of the behavior table 1013 (e.g., location information, facility, facility genre, store, store name, event, etc.). For example, a compilation result may be obtained in which, for facility genre A, A1% of users visit facility genre A1, A2% of users visit facility genre A2, and A3% of users visit facility genre A3. The control unit 104 of the server 10 may also be configured to calculate a predetermined index value (score) based on the aggregation results. In this case, the index value indicates the degree of association between a specific user group and predetermined behavioral data, the importance of the behavioral data, etc. In the present disclosure, unlike the visited places D12, the interests D15 present information indicating the user's interests and preferences in terms of a predetermined genre type, etc. For example, in facility genre A, it is possible to visually confirm the proportion of A1, A2, and A3 groups in the interests and preferences of users included in a predetermined user group. <Output example 2 of interests D15> Drinks Coffee 44 points Black tea 12 points Barley tea 22 points Water 32 points

[0095] By checking the display contents of the advertisement distribution processing page D1, the information publisher can learn in detail about themes such as the interests, concerns, hobbies, and preferences of a user group consisting of multiple users identified by one or more selected theme IDs.By checking the display contents of the advertisement distribution processing page D1, the information publisher can create advertising data suitable for the user group.In addition, in accordance with the request of the information publisher's customer (for example, a store such as the Hakata Ramen Higashi Station store), the information publisher can confirm whether the user group is likely to visit the store as a potential customer.It is desirable to group the themes into uniform themes such as by baseball team, as this makes it easier to understand when distributing advertisements.

[0096] In step S502, the control unit 104 of the server 10 executes a second user group identification step of identifying one or more second user groups related to the user interests input in step S501. Specifically, the control unit 104 of the server 10 searches the theme ID item in the theme table 1014 based on the one or more theme identification information stored in the advertising theme ID of the advertisement table 1021 in step S501, and obtains information on the second cluster ID. This makes it possible to identify a second user group (potential customers) that is expected to exhibit a behavior pattern similar to that of the first user group linked to the one or more theme identification information input by the administrator in step S501.

[0097] In step S503, the control unit 104 of the server 10 executes a second user information acquisition step of acquiring second user information of users included in one or more second user groups identified in the second user group identification step. Specifically, the control unit 104 of the server 10 searches the cluster ID field in the cluster table 1015 based on the second cluster ID acquired in step S502, and acquires the user group field. The control unit 104 of the server 10 searches the user ID field in the user table 1012 based on one or more user identification information included in the acquired user group, and acquires user information (second user information of the second user) including one or more user attribute data. Note that the control unit 104 of the server 10 may acquire information such as representative behavior data of the second user group in addition to the user attribute data.

[0098] In step S504, the control unit 104 of the server 10 executes a data creation step to generate information. Specifically, the control unit 104 of the server 10 creates advertising materials (information and data related to advertisements) using the advertising data stored in the advertisement table 1021 in step S501. For example, the advertising materials include text advertisements (catchphrases), banner advertisements (images), and advertisements that combine images and text. The advertising materials may be text data, image data, audio data, or video data. The control unit 104 of the server 10 may create advertisements suited to the purchasing behavior of the second user, taking into account the user attribute data and representative behavior data of the user acquired in step S503. The data created by the server 10 is not limited to advertising data, and may be any information, such as recommendation information (text data, image data, audio data, and video data describing recommended places, products, services, etc.).

[0099] It is preferable to configure the system so that the processing of steps S501 to S504 is executed in advance to create multiple second user groups corresponding to each of the multiple pieces of advertising information, and advertising data corresponding to the multiple second user groups.

[0100] In step S505, the control unit 104 of the server 10 executes an information providing step of providing information according to a predetermined theme to the second user group extracted in the second extraction step. Specifically, the control unit 104 of the server 10 delivers the advertisement created in step S504 to the second user group identified in step S502. This allows an appropriate advertisement according to the advertisement theme ID set in step S501 to be delivered to the second user group.

[0101] In the present disclosure, an example will be described in which, when a second user included in a second user group accesses a predetermined website using a user terminal 20, advertising material provided by the server 10 according to the present disclosure is posted on the website. The second user can access the predetermined website by operating the input device 206 of the user terminal 20 to execute a browser application or the like. The predetermined website is assumed to have a script (e.g., JavaScript) such as a predetermined tag for linking with the information processing service according to the present disclosure embedded in advance. Upon accessing the predetermined website, the control unit 204 of the user terminal 20 executes the embedded script and transmits the user ID 2011 of the user terminal 20, the URL of the predetermined website, and other user information to the server 10 according to the present disclosure. As a result, the control unit 104 of the server 10 obtains information indicating that the second user identified by the user ID 2011 (second user ID) has accessed the predetermined website.

[0102] If the control unit 104 of the server 10 determines that the received second user ID is included in the user IDs of the second user group identified in step S502, it transmits the advertisement data generated in step S504 to the user terminal 20 of the second user. The control unit 204 of the user terminal 20 of the second user displays the received advertisement data on the display 2081 of the user terminal 20, and presents the advertisement data to the second user.

[0103] In step S505, the information provision step executes a step of providing information to a predetermined user in accordance with the information on the interests received in the information input step. The control unit 104 of the server 10 may execute an information provision step of providing information according to a predetermined theme to a predetermined user identified based on the second user group extracted in the second extraction step. Specifically, the target of advertisement delivery is not limited to the second user group. For example, advertisements may be delivered to users (target users) other than the second user group whose representative behavior data is similar to that of the second user group. Furthermore, the target users do not necessarily have to be users of the information processing service according to the present disclosure (including the first user and the second user). For example, the user terminal 20 may be configured not to transmit user identification information such as the user ID 2011. In this case, the user can be identified based on information from a wireless base station or the like that is located at a predetermined location where the user terminal 20 is located and can wirelessly connect to the user terminal 20. For example, location information estimated by a wireless base station or the like may be used as the location information of the user terminal 20. In this case, as the user moves, the wireless base station or the like used by the user terminal 20 changes, and time-series changes in the user's location information can be acquired as behavioral data. The acquisition of behavioral data is similar to step S101 of the cluster processing, and therefore a description thereof will be omitted. This allows the control unit 104 of the server 10 to acquire behavioral data of users (target users) other than the second user group. The control unit 104 of the server 10 compares the acquired behavioral data of the target user (target behavioral data) with the behavioral data of the second user group acquired in step S503, and identifies the second user group similar to the target behavioral data. The method for calculating the similarity of the behavioral data is similar to the method for calculating the behavioral data described in the cluster processing and user extraction processing, and therefore a description thereof will be omitted. The control unit 104 of the server 10 delivers the advertising data created by the processing of steps S501 to S504 to the identified second user group to the target users. The delivery of advertising data is similar to the delivery of advertisements to second users, and therefore a description thereof will be omitted.

[0104] The information providing step executes a step of providing information to a predetermined user in accordance with information related to the interests received in the information input step. The information providing step executes a step of providing information in accordance with a predetermined theme related to an input operation received from the predetermined user. The information providing step executes a step of providing information in accordance with a predetermined theme related to location information in accordance with location information of the predetermined user. The information providing step executes a step of providing information in accordance with a predetermined theme related to attribute information in accordance with attribute information related to the attributes of the predetermined user. The information providing step executes a step of providing information in accordance with a predetermined theme related to attribute information identified based on the appearance of the predetermined user. The information providing step executes a step of providing information in accordance with a predetermined theme related to the attribute information in accordance with time.

[0105] Although the target user of the device and the user ID of the medium are not the same, matching can be achieved using the following method. For example, the control unit 104 of the server 10 may identify a second user group related to the target user based on the target user's user attribute data, the date and time of access to a specific website, the target user's location information, the specific website accessed by the user terminal 20 in response to the target user's input operations on the user terminal 20, other information about the target user's interests (e.g., information such as the user's facility genre importance determined from the history of websites accessed and the user terminal's location information), etc. Furthermore, the identified second user group does not need to be one; multiple second user groups may be identified. The control unit 104 of the server 10 may deliver advertisements to the target user by combining known targeting advertising techniques using the target user's user attribute data, the date and time of access to a specific website, the target user's location information, the specific website accessed by the target user, the history of websites accessed by the target user, etc. For example, when comparing the acquired behavioral data of the target user (target behavioral data) with the behavioral data of a second user group and calculating the similarity (score) with the target behavioral data, the control unit 104 of the server 10 may identify one or more second user groups by taking into account the score obtained by a known targeting advertising technique. For example, in the same way as identifying a second user group having representative behavioral data similar to the target behavioral data based on the target user's target behavioral data, the control unit 104 of the server 10 may identify a second user group having similar user attribute data, location information, the date and time of access to a specific website, the specific website accessed, the history of websites accessed by the target user, etc. (having similar interests).As a result, the control unit 104 of the server 10 can deliver to the target user advertisements that are expected to interest the target user based not only on the target user's target behavior data, but also on user attribute data, location information, date and time, accessed websites, the IP address of the access source, the user-agent of the user terminal, etc. Furthermore, when a city, ward, town, or village or a company is determined from the IP address, advertisements corresponding to clusters of the city, ward, town, town, or village or company may be delivered. In addition, depending on the time of day, for example, advertisements for restaurants suitable for dinner may be delivered to the user terminal 20 if it is dinnertime.

[0106] Furthermore, user attribute data of a user does not necessarily need to be obtained from user behavior data, etc. For example, consider a case where the user terminal 20 is a terminal consisting only of a display such as a digital signage. In this case, a camera 2061 installed on the front of the digital signage captures an image of the appearance of a user (a predetermined user) standing in front of the display. The control unit 204 of the user terminal 20 transmits the captured appearance image data of the predetermined user to the server 10. The control unit 104 of the server 10 may be configured to identify (infer) the user attributes of the predetermined user by analyzing the appearance image data. Specifically, the user attributes of a user are identified using the following procedure. The control unit 104 of the server 10 analyzes (morphologically analyzes) the acquired appearance image data to capture the predetermined user as an expression vector (appearance vector) consisting of components such as gender, age, hairstyle, clothing, color scheme, and belongings. In this case, the user attribute data of a group of users who have an appearance similar to that of the predetermined user based on the features that make up the predetermined user's appearance may be used as the user attribute data of the predetermined user.

[0107] Note that the specified user does not necessarily have to be one person, and the configuration may be such that the average appearance of multiple users photographed by camera 2061 is considered to be the appearance of the specified user and the processing described below is executed.

[0108] The control unit 104 of the server 10 performs clustering processing on multiple users (an unspecified number of users) in advance, based on the similarity of their expression vectors. The control unit 104 of the server 10 associates typical appearance vectors (representative appearance vectors) with typical user attribute data (representative user attribute data) for a user group consisting of multiple users and stores them in a database (not shown). Note that the user attribute data of the users to be processed in the clustering processing is known in advance. The database stores the user group, the representative appearance vector, and the representative user attribute data in association with each other. Based on the acquired expression vector of a specific user, the control unit 104 of the server 10 searches for similar records from the representative appearance vectors of multiple user groups stored in a database, and acquires the representative user attribute data stored in association with the representative appearance vector. The control unit 104 of the server 10 uses the representative user attribute data as the specific user attribute data.

[0109] Note that, in the present disclosure, an example has been disclosed in which, based on one appearance vector of a specific user, representative user attribute data of a group of users having a representative appearance vector similar to the specific user's appearance vector is used as the user attribute data of the specific user. However, this is not limited to this. For example, a specific user may have multiple independent appearance vectors consisting of at least one combination of gender, age, hairstyle, clothing, color scheme, belongings, etc. Specifically, each feature (gender, age, hairstyle, clothing, color scheme, belongings, etc.) that constitutes the specific user's appearance may have an independent appearance vector. In this case, the control unit 104 of the server 10 may perform a clustering process for multiple users (an unspecified number of users) in advance, based on the similarity of each feature. For a user group consisting of multiple users, the control unit 104 associates a typical appearance vector (representative appearance vector) with typical user attribute data (representative user attribute data) of the user group for each feature, and stores the associated data in a database (not shown). In this case, typical user attribute data (representative user attribute data) of a group of similar users is associated with each feature such as gender, age, hairstyle, clothing, color scheme, and belongings and stored in a database (not shown). The control unit 104 of the server 10 identifies multiple pieces of user attribute data for each feature that makes up the appearance of a specific user by referring to the database (not shown). The control unit 104 of the server 10 identifies the user attribute data of the specific user as, for example, a weighted sum of multiple pieces of user attribute data. The weight of each feature may be set to an arbitrary value (a score, etc. may be assigned to each feature). Furthermore, a score, etc. may be assigned to each combination of features that make up the appearance.

[0110] Alternatively, the control unit 104 of the server 10 may identify a group of users with similar appearances based on the features that make up the appearance of a specific user, using the TF-IDF scores for each feature shown below, and use the user attribute data of the group of users as the user attribute data of the specific user.

[0111] Each part of multiple people's age, gender, belongings, hairstyle, etc. is vectorized, and a score can be calculated for this set in the same way as calculating the TF-IDF score for text data. Specifically, person data is structured by part, and each part is treated as an independent "word." - Age: 25, 30, 35, ... - Gender: Male, Female - Possessions: Sunglasses, bag, hat, ... - Hairstyle: Short hair, long hair, short hair, ... TF (occurrence frequency) is calculated by counting the frequency of occurrence of features (words) for each part: - Number of times age "25" appears - Number of times gender "male" appears - Number of times possession "sunglasses" appears - Number of times hairstyle "short hair" appears DF (document frequency) is calculated by counting the number of people in whom each feature appears. - Number of people with the age "25" - Number of people with the gender "male" - Number of people with the item "sunglasses" - Number of people with the hairstyle "short hair" IDF (inverse document frequency) is calculated based on the calculated DF. The method for calculating IDF is the same as the procedure for calculating TF-IDF for normal text documents. The TF-IDF score is calculated by calculating the product of TF and IDF. The TF-IDF score calculated in this way can be used to evaluate the importance of a specific feature (feature) within a set. For example, if the item "sunglasses" is common to many people, the IDF will be low, and the importance of the item unique to a specific person will be evaluated highly. By calculating TF-IDF scores for a person's appearance features, similar to document search using TF-IDF scores, it is possible to more appropriately identify a group of users similar to a specific person based on features that are highly important within the entire set, while taking into account the frequency of appearance of each feature.

[0112] For information publishers who own stores, etc., it is possible to analyze the area where the store is located and understand the store's customer persona, and then distribute information to a group of users who have a high degree of similarity to the store's customer persona. In addition, at that time, it is also possible to distribute information taking into account the trade area.

[0113] Generally, a trade area is roughly defined by the radius of each facility type as the area within which a store or other facility can attract customers. Furthermore, typical advertising distribution systems do not obtain location information from websites with sufficient resolution to determine whether users are visiting a tourist attraction or store, and therefore do not aggregate visitors to the tourist attraction or store. Therefore, when attracting customers to a tourist attraction or store, they distribute information to users within a certain radius of the tourist attraction or store. In contrast, each user group can obtain population distribution data for each spot before and after check-in (or check-out), which can be considered as the actual trade area of ​​each spot. When users are directly sampled from a particular spot, the distribution can be considered as the trade area of ​​that spot. These trade areas vary in scope depending on the spot and user group, and reflect not only the influence of distance, transportation, and population, but also the characteristics of the spot and user group. These trade areas do not have a simple radius, and can take into account factors such as particularly large populations or a particularly high population ratio compared to other user groups. For example, in the case of tourist destinations, areas near tourist destinations are generally populated by local residents and not by tourists in their daily lives, so they are not considered commercial areas. Instead, areas where tourists are frequently present in their daily lives can be treated as actual commercial areas. Furthermore, the period before and after check-in (or check-out) for each spot can be arbitrarily changed depending on the purpose and server load, allowing for multiple commercial areas with different durations. In this case, a six-hour period can be considered a range that can attract customers within six hours, while a six-month period can be considered a range that can attract customers within six months. In other words, by considering commercial areas, if you want to attract customers within six hours, you can deliver ads to users who can attract customers within six hours. If you want to attract customers within six months, you can deliver ads to users who can attract customers within six months. For example, a store can deliver ads when it has leftover inventory and wants to attract customers who will visit that day. At a regional tourist destination, you can deliver ads when it wants to attract wealthy urban residents who will visit within six months. Furthermore, even if location information and attribute information obtained from a website identify a target user group, those outside the commercial area can be excluded from ad delivery.In addition, for example, when a second user group (potential customers) for a particular tourist destination is generated, its distribution may differ from the actual commercial area, but this can be corrected by using the distribution of the original user group or the population distribution of another user group in the period before and after checking in (or checking out) at the tourist destination.

[0114] When targeting by matching statistical big data with personal data obtainable within a first-party service, it is expected that only a small amount of information can be obtained from the target user, making it difficult to capture a user profile from such a small amount of information. However, by matching the information sent by the target user with a combination of multiple statistics in different formats (multimodal), a user profile can be captured. Specifically, in the disclosed invention, by combining multiple statistical information in different formats calculated based on behavioral data, more appropriate user groups (clusters) can be identified. This method allows ad distributors and advertising media to perform optimal targeting with a smaller amount of data, without having to retain large amounts of data that can identify individuals or that could lead to the identification of individuals. For example, if two or three pieces of location information can be obtained from a website user, and the user can be inferred to be an otaku or a band girl based on the occupancy rate of the statistical data of the user group of the mesh or spot corresponding to that location information, but if further location information cannot be obtained and the occupancy rate cannot be distinguished from the otaku or band girl, by combining schedule statistics, it can be determined that the user is a band girl because there are fewer otaku and more band girls using the facility genre corresponding to the above location information at 10 PM.In addition, if two or three keywords and one piece of location information can be obtained from a website user, and the user can be inferred to be an otaku or band girl based on the statistical data of the user group corresponding to that keyword (feature amount of the facility genre), but if further keywords cannot be obtained and the feature amount cannot be distinguished from the otaku or band girl, it can be determined that the user is an otaku by combining statistical data of the user group of the mesh or spot and the occupancy rate of that spot is low for band girls and high for otaku.

[0115] While typical advertising delivery systems deliver information based on the content of websites that users are viewing or have viewed, by matching this information with location information collected in various formats, it is possible to provide more appropriate information (advertisement delivery) based on the user's real-world environment (external context information and its transitions), which is unrelated to the information the user is viewing or has viewed on the website. For example, if there is time-series representative behavior data for a certain user group, such as a visit to a coffee shop, a visit to a tourist spot, and then a return to a hotel, and current location information is obtained showing that a user presumed to be part of the user group is at a coffee shop, the user can be provided with information about tourist spots that is unrelated to the information the user is viewing or has viewed on the website. Furthermore, if the facility category "tourist spot" in the time-series representative behavior data is concentrated in a specific spot, it is presumed that the user group is traveling to a specific tourist spot. Therefore, it is preferable to present information related to the specific tourist spot to the presumed user group, rather than recommending other tourist spot candidates. This also applies when the time-series representative behavior data for the user group's spots is concentrated in a specific tourist spot. On the other hand, if the facility genre "tourist destination" in the time-series representative behavior data is distributed across multiple spots, it is estimated that the user group is sightseeing but not planning to visit any particular tourist destination, so it is preferable to deliver information about potential tourist destinations to users estimated to be part of that user group. In this way, the information provided can be changed depending on the degree of distribution of spots that make up the time-series representative behavior data of the facility genre for each user group.Note that providing information according to the environment in which the user is placed in the real world (external context information and its changes) is not intended to be unrelated to the information the user is viewing or has viewed on a website. For example, if a user is viewing a website related to learning and it is estimated that the user has just visited a sports gym or other facility in the real world based on the time-series representative behavior data of a user group to which the user is estimated to belong, information combining the content of the website the user is viewing and the environment in which the user is placed in the real world (external context information and its changes) may be provided. Information about real-world locations is not necessarily provided. For example, if the time-series representative behavior data of the user group's spots and facility genres is not characteristic or important data for the user group, information unrelated to real-world locations according to the user group's interests (feature values ​​of the facility genre) may be provided.

[0116] For example, if the facility genres before checking in to a ramen restaurant include many izakayas in the time-series representative behavior data for all users, users who check in to izakayas can be grouped together to provide information that is perceived as a "group of users taking actions before eating ramen" or a "group of users who will eat ramen as their next action." If the facility genres before checking in to a ramen restaurant differ for each user group, users who check in to the facility genres before checking in to the ramen restaurant for each user group can be grouped together to provide information that is perceived as a "group of users taking actions before eating ramen" or a "group of users who will eat ramen as their next action." Furthermore, by combining actions before and after the ramen restaurant within six hours of the ramen restaurant, it is possible to create a "group of users who can visit the restaurant within six hours and will eat ramen as their next action." This allows advertisers who want to quickly attract customers to the ramen restaurant to select a "group of users who can visit the restaurant within six hours and will eat ramen as their next action" without having to make complex decisions. When the advertiser is a tourist destination or a store, the user group extracted from the tourist destination or store is the user group that suits the tourist destination or store, and by using the user group as a first user group and identifying the representative behavior data or the second user group, it is possible to provide even more suitable information. Also, when the advertiser is a tourist destination or a store, for example, if the user group extracted from the tourist destination or store has high characteristics related to prep schools and renovations in its daily behavior data and high characteristics related to athletics and natural parks in its daily behavior data for the period before and after checking in at the tourist destination or store, it is possible to use both keywords to match with keywords obtained from website users and perform more suitable targeting that reflects the users' interests in their daily lives and their interests when visiting the tourist destination or store.

[0117] When the context of daily activities is added to complicate the trend, there are roughly nine trends. Reference numeral 1001 in FIG. 15 indicates that spot X is a location where there is no pattern in the pre- and post-traffic activities. Reference numeral 1002 in FIG. 15 indicates that spot X is a location where multiple spots of the same facility category are headed. Reference numeral 1003 in FIG. 15 indicates that spot X is a location where multiple spots of the same facility category are headed to locations with no pattern. Reference numeral 1004 in FIG. 15 indicates that spot X is a location where multiple spots of the same facility category are headed to the same facility category. Reference numeral 1005 in FIG. 15 indicates that spot X is a location where multiple spots of the same facility category are headed to the same spot. Reference numeral 1006 in FIG. 15 indicates that spot X is a location where there is no pattern in the pre- and post-traffic activities where people head to the same spot. Reference numeral 1007 in FIG. 15 indicates that spot X is a location where people head from the same spot to the same spot from the pre- and post-traffic activities. Reference numeral 1008 in FIG. 15 indicates that spot X is a location where people head from the same spot to multiple spots of the same facility category. 1009 in FIG. 15 indicates that spot X is a location from the same spot to a location where there is no regularity in the subsequent behavior.

[0118] The significance of Spot X itself as a spot or facility genre for a cluster can be understood by combining, for example, the behavior of a spot on that day where a large number of people gather, regardless of user group, such as a train station or airport, or the overall behavior of that day starting from check-ins in a certain area. 1101 in FIG. 16 indicates that Spot X is one of the locations within the no-rule. 1102 in FIG. 16 indicates that Spot X is the central location within the no-rule. 1103 in FIG. 16 indicates that Spot X is one of the locations that is the derivative (origin) of the center of the no-rule. 1104 in FIG. 16 indicates that Spot X is one of the locations within a larger flow. 1105 in FIG. 16 indicates that Spot X is the central location of a larger flow. 1106 in FIG. 16 indicates that Spot X is one of the locations that is complementary to the center of the larger flow.

[0119] Spots or facility genres where clusters (including the entire cluster) share common behavior can be further categorized by the length of stay and frequency of visits. The more variance there is between the two locations and the longer the stay compared to the two locations, the more likely it is a destination. - Places where stay times are long and the frequency is high can be interpreted as spots or facility genres such as schools or workplaces. - Places where stay times are long and the frequency is low can be interpreted as spots or facility genres such as events or tourist destinations. - Places where stay times are short and the frequency is high can be interpreted as spots or facility genres that serve as transportation hubs such as stations and airports. - Places where stay times are short and the frequency is low can be interpreted as spots or facility genres that serve as special shopping destinations or sacred places for specific groups of users.

[0120] There is a master-servant relationship between spots, with the destination or purpose being the master, and the spots or actions that complement it being the subordinate. For example, a shrine and a souvenir shop have a master-servant relationship, and the relationship is not "Let's go to the shrine after buying souvenirs," but rather "Let's go to the shrine after buying souvenirs," and it is appropriate to provide information that is tailored to this relationship.

[0121] These facility genre features indicate the degree of connection between each user group and the facility genre, and the commercial area and distribution indicate the degree of connection between each user group and the spot (area), so facility genres / spots (areas) that are highly related to a certain user group may be visualized using a co-occurrence network representation method.

[0122] The information providing step may include providing information according to a predetermined theme related to the combination based on at least two or more of the following: one or more pieces of information related to the user's interests; one or more operations received as input from the predetermined user; one or more pieces of location information indicating the location of the predetermined user; one or more pieces of attribute information related to the attributes of the predetermined user; and one or more pieces of time. Specifically, the second user group may be identified based on a combination of at least two or more pieces of information related to the target user's interests, including the target user's behavioral data (including location information), the target user's user attribute data, the date and time of access to a predetermined website, the predetermined website accessed in response to the target user's input operation on the user terminal 20, the history of websites accessed in response to the target user's input operation on the user terminal 20, the IP address of the access source, information about the user agent and browser of the user terminal, and other information related to the target user's interests. This allows the control unit 104 of the server 10 to deliver to the target user more suitable advertisements that are expected to interest the target user based on a combination of not only the target user's target behavioral data but also the user attribute data, the date and time, the website accessed, and the like.

[0123] 14 is a block diagram showing the basic hardware configuration of a computer 90. The computer 90 includes at least a processor 901, a main storage device 902, an auxiliary storage device 903, and a communication IF 991 (interface), which are electrically connected to one another by a communication bus 921.

[0124] The processor 901 is hardware for executing an instruction set written in a program, and is composed of an arithmetic unit, registers, peripheral circuits, and the like.

[0125] The main storage device 902 is used to temporarily store programs and data to be processed by the programs, etc. For example, it is a volatile memory such as a DRAM (Dynamic Random Access Memory).

[0126] The auxiliary storage device 903 is a storage device for saving data and programs, such as a flash memory, a hard disk drive (HDD), a magneto-optical disk, a CD-ROM, a DVD-ROM, or a semiconductor memory.

[0127] The communication IF 991 is an interface for inputting and outputting signals for communicating with other computers via a network using a wired or wireless communication standard. The network is composed of the Internet, a LAN, various mobile communication systems constructed using wireless base stations, etc. For example, the network includes 3G, 4G, and 5G mobile communication systems, LTE (Long Term Evolution), and wireless networks (e.g., Wi-Fi (registered trademark)) that can connect to the Internet via a predetermined access point. In the case of a wireless connection, communication protocols include, for example, Z-Wave (registered trademark), ZigBee (registered trademark), and Bluetooth (registered trademark). In the case of a wired connection, the network also includes a network that is directly connected using a USB (Universal Serial Bus) cable, etc.

[0128] It should be noted that the computer 90 can be virtually realized by distributing all or part of each hardware configuration across multiple computers 90 and interconnecting them via a network. In this way, the concept of the computer 90 includes not only a computer 90 housed in a single housing or case, but also a virtualized computer system.

[0129] <Basic Functional Configuration of Computer 90> A description will be given of the functional configuration of the computer realized by the basic hardware configuration (FIG. 14) of the computer 90. The computer includes at least the functional units of a control unit, a storage unit, and a communication unit.

[0130] The functional units of the computer 90 can also be realized by distributing all or part of the functional units among multiple computers 90 interconnected via a network. The computer 90 is a concept that includes not only a single computer 90 but also a virtualized computer system.

[0131] The control unit is realized by the processor 901 reading various programs stored in the auxiliary storage device 903, loading them into the main storage device 902, and executing processing in accordance with the programs. The control unit can realize functional units that perform various types of information processing depending on the type of program. In this way, the computer is realized as an information processing device that performs information processing.

[0132] The functions performed by the components described herein may be implemented in circuitry or processing circuitry, including general-purpose processors, application-specific processors, integrated circuits, ASICs (Application Specific Integrated Circuits), CPUs (Central Processing Units), conventional circuits, and / or combinations thereof, programmed to perform the described functions. Processors include transistors and other circuits and are considered circuitry or processing circuitry. A processor may also be a programmed processor that executes a program stored in memory. In this specification, a circuit, unit, or means is hardware that is programmed to perform or executes the described functions. The hardware may be any hardware disclosed herein or any hardware known to be programmed to perform or execute the described functions. When the hardware is a processor, which is considered a type of circuitry, the circuit, means, or unit is a combination of hardware and software used to configure the hardware and / or processor.

[0133] The storage unit is realized by a main storage device 902 and an auxiliary storage device 903. The storage unit stores data, various programs, and various databases. The processor 901 can allocate a storage area corresponding to the storage unit in the main storage device 902 or the auxiliary storage device 903 in accordance with the programs. The control unit can cause the processor 901 to add, update, and delete data stored in the storage unit in accordance with the various programs.

[0134] The term "database" refers to a relational database, which manages data sets called tables and masters in a tabular format structurally defined by rows and columns, by associating them with each other. In a database, a table is called a table or master, a column in a table is called a column, and a row in a table is called a record. In a relational database, relationships between tables and masters can be established and associated. Typically, each table and each master is set with a column that serves as a primary key to uniquely identify a record, but setting a primary key to a column is not required. The control unit can cause the processor 901 to add, delete, or update records in specific tables and masters stored in the storage unit according to various programs. By storing data, various programs, and various databases in the storage unit, the information processing device and information processing system according to the present disclosure can be considered to have been manufactured.

[0135] Note that the databases and masters in this disclosure may include any data structure in which information is structurally defined (such as a list, dictionary, associative array, or object). The data structure also includes data that can be considered as a data structure by combining data with functions, classes, methods, etc. written in any programming language.

[0136] The communication unit is realized by the communication IF 991. The communication unit realizes the function of communicating with other computers 90 via a network. The communication unit can receive information transmitted from other computers 90 and input the information to the control unit. The control unit can cause the processor 901 to execute information processing on the received information in accordance with various programs. Furthermore, the communication unit can transmit information output from the control unit to other computers 90.

[0137] <Additional Notes> The matters described in the above embodiments will be added below.

[0138] (Supplementary Note 1) A program to be executed by a computer including a processor and a storage unit, the program comprising: a behavior acquisition step (S101) of acquiring behavioral data related to the behaviors of a plurality of users; a first extraction step (S302) of extracting a first user group consisting of one or more users associated with a predetermined theme from among the plurality of users based on the behavioral data acquired in the behavior acquisition step; a representative acquisition step (S303) of acquiring representative behavior data representative of the first user group extracted in the first extraction step; and a second extraction step (S304) of extracting a second user group consisting of one or more users from the plurality of users based on the representative behavior data of the first user group. This makes it possible to extract a second user group (potential customers) that is different from the first user group but is expected to potentially exhibit behavioral patterns similar to those of the first user, based on typical behavioral patterns (representative behavioral data) of the first user group associated with a specific theme or event.

[0139] (Supplementary Note 2) The program according to Supplementary Note 1, wherein the first extraction step (S302) is a step of extracting a first user group associated with an event held during a predetermined period, and the representative acquisition step (S303) is a step of acquiring representative behavior data representing the first user group during a period other than the predetermined period. By extracting a first user group associated with an event held at a predetermined date, time, and location, it is possible to extract a first user group that exhibits more homogeneous, high-quality, and similar behavior patterns. Based on the similarity of the behavior patterns of the first user group during periods other than the event, it is possible to extract a second user group (potential customers) that is expected to potentially exhibit behavior patterns similar to those of the first user.

[0140] (Supplementary Note 3) The program according to Supplementary Note 2, wherein the first extraction step (S302) is a step of extracting a first user group by excluding one or more users who are not related to an event held during a predetermined period based on the behavior of a plurality of users before and after the predetermined period. This makes it possible to exclude from the first user group users who are dissimilar from the perspective of similarity in behavior patterns. This makes it possible to extract a first user group that exhibits more homogeneous, high-quality, and similar behavior patterns.

[0141] (Supplementary Note 4) The program according to any one of Supplementary Notes 1 to 3, wherein the processor executes a grouping step (S102) of classifying a plurality of users into a plurality of groups based on the similarity of the behavioral data acquired in the behavior acquisition step, and the second extraction step (S304) is a step of extracting, from the plurality of groups, one or more groups having representative behavioral data similar to the representative behavioral data of the first user group as a second user group. This makes it possible to extract, as the second user group, from groups (clusters) obtained by grouping (clustering) all of the plurality of users according to the similarity of their behavioral patterns, groups having behavioral patterns similar to the behavioral patterns of the first user group.

[0142] (Supplementary Note 5) The program according to any one of Supplementary Notes 1 to 4, wherein the processor executes an information providing step (S505) of providing information according to a predetermined theme to a predetermined user identified based on the second user group extracted in the second extraction step. This makes it possible to provide information related to a predetermined theme that is expected to interest the second user group to a predetermined user (potential customer) identified based on the second user group who is expected to potentially exhibit behavioral patterns similar to those of the first user. This makes it possible to improve the response to the provided information. The predetermined user may be a user included in the second user group.

[0143] (Supplementary Note 6) The program according to Supplementary Note 5, wherein the processor executes an information input step (S501) of accepting input of information relating to a user's interests, and an information provision step (S505) of providing information to a predetermined user in accordance with the information relating to the interests accepted in the information input step. This makes it possible to provide information to a predetermined user who is expected to have the user's interests in accordance with the input of information relating to the user's interests accepted from an information publisher, including an information medium, etc. This improves the effectiveness of providing information.

[0144] (Supplementary Note 7) The program according to Supplementary Note 5 or 6, wherein the information providing step (S505) is a step of providing information according to a predetermined theme related to an input operation received from a predetermined user in response to the input operation. This makes it possible to provide information to a predetermined user that is more in line with the user's interests in response to a predetermined operation performed by the user on the Internet. This improves the effectiveness of providing information. For example, it is possible to provide information that is more effective for the user by taking into account the genre and article content of the website visited by the user.

[0145] (Supplementary Note 8) The program according to any one of Supplementary Notes 5 to 7, wherein the information providing step (S505) is a step of providing information according to a predetermined theme related to the location information of a predetermined user, in accordance with the location information of the predetermined user. This makes it possible to provide information that is more in line with the interests of each user, in accordance with the user's location information. This improves the effectiveness of providing information.

[0146] (Supplementary Note 9) The program according to any one of Supplementary Notes 5 to 8, wherein the information providing step (S505) is a step of providing information according to a predetermined theme related to attribute information of a predetermined user, in accordance with the attribute information of the predetermined user. This makes it possible to provide information that is more in line with the interests of each user, in accordance with the user's attribute information, such as the user's gender, age, and occupation. This improves the effectiveness of providing information.

[0147] (Supplementary Note 10) The program according to Supplementary Note 9, wherein the information providing step (S505) is a step of providing information according to a predetermined theme related to attribute information of a predetermined user, in accordance with the attribute information identified based on the appearance of the predetermined user. This makes it possible to provide information to a predetermined user that is more in line with the interests of each user, in accordance with the attribute information of the user identified based on the user's appearance. This improves the effectiveness of providing information.

[0148] (Supplementary Note 11) The program according to any one of Supplementary Notes 5 to 10, wherein the information providing step (S505) is a step of providing information according to a predetermined theme related to the time, depending on the time. This makes it possible to provide a predetermined user with information that is in line with the user's interests, depending on the time of day. This improves the effectiveness of providing information.

[0149] (Supplementary Note 12) The information providing step (S505) is a step of providing information according to a predetermined theme related to a combination of at least two or more of the following: one or more pieces of information related to the user's interests; one or more operations input from a predetermined user; one or more pieces of location information indicating the location of the predetermined user; one or more pieces of attribute information related to the attributes of the predetermined user; and one or more pieces of time. This makes it possible to provide information that is in line with the user's interests in accordance with multiple conditions to the predetermined user. This improves the effectiveness of providing information.

[0150] (Supplementary Note 13) A program to be executed by a computer including a processor and a storage unit, the program comprising: a behavior acquisition step (S101) of acquiring behavioral data related to the time-series behavior of a plurality of users; a grouping step (S102, S304, S502) of classifying the plurality of users into a plurality of user groups related to a plurality of different themes based on similarities in the behavioral data acquired in the behavior acquisition step; a similarity calculation step (S501) of calculating similarities between a predetermined user group and the plurality of user groups; and a similarity presentation step (S501) of presenting the similarities for each of the plurality of user groups calculated in the similarity calculation step in association with the themes related to the user groups, wherein the similarity calculation step (S501) is a step of calculating similarities between the behavioral data of a plurality of users included in the predetermined user group and the behavioral data of a plurality of users included in each of the plurality of user groups. This makes it possible to present similarities between the behavioral patterns of a user group related to a predetermined theme and a plurality of user groups related to other themes. For example, it is possible to present the degree of similarity between representative behavior data of a user group related to a specific theme and representative behavior data of multiple user groups related to other themes, which can be used as reference information when delivering advertisements to a user group.

[0151] (Supplementary Note 14) The program according to Supplementary Note 13, wherein the processor executes a persona creation step (S501) of creating a person image that characterizes a predetermined user group based on behavioral data of a plurality of users included in the predetermined user group, and a persona presentation step (S501) of presenting the person image created in the persona creation step. This makes it possible to present typical appearances, styles, etc. of users included in the user group. This can be used as reference information when delivering advertisements to the user group.

[0152] (Supplementary Note 15) The program according to Supplementary Note 13, wherein the processor executes a behavior aggregation step (S501) of aggregating the number of predetermined one or more behavioral data included in the behavioral data of a plurality of users included in a predetermined user group, and a behavior presentation step (S501) of presenting an index value based on the number of predetermined one or more behavioral data aggregated in the behavior aggregation step, in association with the predetermined one or more behavioral data. This makes it possible to present the behavioral patterns of users included in the user group. By referring to the behavioral patterns of a plurality of users included in the user group, it is possible to use this as reference information when delivering advertisements to the user group.

[0153] (Supplementary Note 16) The program according to Supplementary Note 15, wherein the behavior aggregation step (S501) is a step of aggregating the number of facility genres related to one or more predetermined facility types included in the behavior data of a plurality of users included in a predetermined user group, and the behavior presentation step (S501) is a step of presenting an index value based on the number of one or more predetermined facility genres aggregated in the behavior aggregation step in association with the one or more predetermined facility genres. This makes it possible to present what facility genres users included in the user group have visited. Checking what facility genres multiple users included in the user group have visited can be used as reference information when delivering advertisements to the user group.

[0154] (Supplementary Note 17) The program according to Supplementary Note 15, wherein the behavior aggregation step (S501) is a step of aggregating the number of one or more predetermined facilities included in the behavior data of multiple users included in a predetermined user group, and the behavior presentation step (S501) is a step of presenting an index value based on the number of one or more predetermined facilities aggregated in the behavior aggregation step in association with the one or more predetermined facilities. This makes it possible to present what facilities users included in the user group are visiting. By checking what facilities multiple users included in the user group have visited, it is possible to use this information as reference information when delivering advertisements to the user group.

[0155] (Supplementary Note 18) The program according to Supplementary Note 13, wherein the processor executes an attribute aggregation step (S501) of aggregating the number of predetermined one or more pieces of attribute information included in attribute information relating to the attributes of a plurality of users included in a predetermined user group, and an attribute presentation step (S501) of presenting an index value based on the number of predetermined one or more pieces of attribute information aggregated in the attribute aggregation step, in association with the predetermined one or more pieces of attribute information. This makes it possible to present the attributes possessed by users included in the user group. Checking the attributes of a plurality of users included in the user group can be used as reference information when delivering advertisements to the user group.

[0156] (Supplementary Note 19) A method executed by a computer having a processor and a memory, wherein the processor executes all of the steps executed in any of the inventions according to Supplementary Note 1 to Supplementary Note 18. This makes it possible to extract a second user group (potential customers) that is different from the first user group but is expected to potentially exhibit behavioral patterns similar to those of the first users, based on typical behavioral patterns (representative behavioral data) of the first user group related to a specific theme or event. Note that this also includes methods executed by generative AI, large-scale language models, etc.

[0157] (Supplementary Note 20) An information processing device comprising a control unit and a storage unit, wherein the control unit executes all of the steps executed in any of the inventions according to Supplementary Note 1 to Supplementary Note 18. This makes it possible to extract a second user group (potential customers) that differs from the first user group but is expected to potentially exhibit behavioral patterns similar to those of the first users, based on typical behavioral patterns (representative behavioral data) of the first user group related to a specific theme or event. This also includes information processing devices comprising generative AI, large-scale language models, etc.

[0158] (Supplementary Note 21) A system comprising means for executing all steps executed in any of the inventions according to Supplementary Note 1 to Supplementary Note 18. This makes it possible to extract a second user group (potential customers) who are different from the first user group but are expected to potentially exhibit behavioral patterns similar to those of the first users, based on typical behavioral patterns (representative behavioral data) of the first user group related to a specific theme or event. This also includes systems consisting of generative AI, large-scale language models, etc.

[0159] 1 System, 10 Server, 101 Memory unit, 104 Control unit, 106 Input device, 108 Output device, 20 User terminal, 201 Memory unit, 204 Control unit, 206 Input device, 208 Output device, 30 Administrator terminal, 301 Memory unit, 304 Control unit, 306 Input device, 308 Output device

Claims

1. A program to be executed by a computer having a processor and a memory unit, wherein the processor executes the following steps: a behavior acquisition step of acquiring behavioral data regarding the behavior of a plurality of users; a first extraction step of extracting a first user group consisting of one or more users from the plurality of users who are related to a predetermined theme based on the behavioral data acquired in the behavior acquisition step; a representative acquisition step of acquiring representative behavior data representative of the first user group extracted in the first extraction step; and a second extraction step of extracting a second user group consisting of one or more users from the plurality of users based on the representative behavior data of the first user group.

2. The program described in claim 1, wherein the first extraction step is a step of extracting the first user group related to an event held during a specified period, and the representative acquisition step is a step of acquiring representative behavior data representing the first user group during a period other than the specified period.

3. The program described in claim 2, wherein the first extraction step is a step of extracting the first group of users based on the behavior of the plurality of users before and after the specified period, excluding one or more users who are not related to an event being held during the specified period.

4. The program of claim 1, wherein the processor executes a grouping step of classifying the plurality of users into a plurality of groups based on the similarity of the behavioral data acquired in the behavior acquisition step, and the second extraction step is a step of extracting one or more groups from the plurality of groups that have representative behavioral data similar to the representative behavioral data of the first user group as the second user group.

5. The program according to claim 1, wherein the processor executes an information providing step of providing information according to the predetermined theme to a predetermined user identified based on the second user group extracted in the second extraction step.

6. The program according to claim 5, wherein the processor executes an information input step of accepting input of information relating to a user's interests, and the information provision step is a step of providing information to the specified user in accordance with the information relating to the interests accepted in the information input step.

7. The program according to claim 5, wherein the information providing step is a step of providing information according to the predetermined theme related to an input operation received from the predetermined user in response to the input operation.

8. The program according to claim 5, wherein the information providing step is a step of providing information according to the predetermined theme related to the location information of the predetermined user in accordance with the location information.

9. The program according to claim 5, wherein the information providing step is a step of providing information according to the predetermined theme related to attribute information in accordance with attribute information relating to the predetermined user attribute.

10. The program according to claim 9, wherein the information providing step is a step of providing information according to the predetermined theme related to the attribute information identified based on the appearance of the specified user.

11. The program according to claim 5, wherein the information providing step is a step of providing information according to the predetermined theme related to the time in accordance with the time.

12. The program of claim 5, wherein the information providing step is a step of providing information according to the specified theme related to a combination of at least two or more of the following: one or more pieces of information related to the user's interests; one or more operations input from the specified user; one or more pieces of location information indicating the location of the specified user; one or more pieces of attribute information related to the attributes of the specified user; and one or more pieces of time.

13. A program to be executed by a computer having a processor and a memory unit, wherein the processor executes the following steps: a behavior acquisition step of acquiring behavioral data related to the time-series behavior of a plurality of users; a grouping step of classifying the plurality of users into a plurality of user groups related to a plurality of different themes based on the similarity of the behavioral data acquired in the behavior acquisition step; a similarity calculation step of calculating the similarity between a predetermined user group and a plurality of user groups; and a similarity presentation step of presenting the similarity for each of the plurality of user groups calculated in the similarity calculation step in association with the theme related to the user group, wherein the similarity calculation step is a step of calculating the similarity between the behavioral data of a plurality of users included in the predetermined user group and the behavioral data of a plurality of users included in each of the plurality of user groups.

14. The program of claim 13, wherein the processor executes a persona creation step of creating a person image that characterizes a specified user group based on behavioral data of multiple users included in the specified user group, and a persona presentation step of presenting the person image created in the persona creation step.

15. The program of claim 13, wherein the processor executes: a behavior aggregation step of aggregating the number of one or more predetermined behavioral data included in the behavioral data of multiple users included in the predetermined user group; and a behavior presentation step of presenting an index value based on the number of the predetermined one or more behavioral data aggregated in the behavior aggregation step in association with the predetermined one or more behavioral data.

16. The program described in claim 15, wherein the behavior aggregation step is a step of aggregating the number of facility genres related to one or more predetermined facility types included in the behavior data of multiple users included in the predetermined user group, and the behavior presentation step is a step of presenting an index value based on the number of the one or more predetermined facility genres aggregated in the behavior aggregation step in association with the one or more predetermined facility genres.

17. The program described in claim 15, wherein the behavior aggregation step is a step of aggregating the number of one or more predetermined facilities included in the behavior data of multiple users included in the predetermined user group, and the behavior presentation step is a step of presenting an index value based on the number of the predetermined one or more facilities aggregated in the behavior aggregation step in association with the predetermined one or more facilities.

18. The program of claim 13, wherein the processor executes an attribute aggregation step of aggregating the number of one or more pieces of predetermined attribute information included in attribute information relating to the attributes of multiple users included in the predetermined user group, and an attribute presentation step of presenting an index value based on the number of the predetermined one or more pieces of attribute information aggregated in the attribute aggregation step in association with the predetermined one or more pieces of attribute information.

19. A method implemented on a computer having a processor and memory, wherein the processor performs all of the steps performed in any one of claims 1 to 18.

20. An information processing device comprising a control unit and a memory unit, wherein the control unit executes all steps executed in any one of the inventions according to claims 1 to 18.

21. A system comprising means for executing all the steps performed in any one of the inventions according to claims 1 to 18.

Images