Information Processing Apparatus, Information Processing Method, and Program

Abstract

To provide a mechanism that makes it easier for a second user to utilize the experience content of a first user. [Solution] The information processing device 2 has an acquisition processing unit 231 that acquires first user behavior data and a model management unit 232 that accumulates first user vector data corresponding to the first user behavior data, the acquisition processing unit 231 acquires second user behavior data, the information processing device 2 further has a recording processing unit 233 that records second user vector data corresponding to the second user behavior data when the second user vector data is similar to the first user vector data, and a detection unit 236 that detects a predetermined behavior by the second user, the model management unit 232 generates a prompt when the detection unit 236 detects the predetermined behavior by the second user, and the information processing device 2 further has a presentation unit 237 that presents output information output by the large-scale language model to the second user by inputting the prompt into the large-scale language model.

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program.

Background Art

[0002] In recent years, Social Networking Services (SNS) have become widespread, and various contents have been posted by SNS users (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] When content indicating the content experienced by a first user is posted on an SNS, a second user who has viewed the content can use it as a reference for his or her own actions. However, even when there is a second user in an environment where the content of the first user's experience can be utilized, the second user may not be aware that the content of the first user's experience can be utilized, or the second user may forget the content of the first user's experience, in which case the content of the first user's experience may not be utilized.

[0005] Therefore, the present invention has been made in view of these points, and an object thereof is to provide a mechanism that makes it easier for a second user to utilize the content of the first user's experience.

Means for Solving the Problems

[0006] The information processing apparatus according to the first aspect of the present invention includes a first user information acquisition unit that acquires first user action data indicating a plurality of actions of a first user, and a first user vector data that is data obtained by converting the first user action data into a vector. The large language model used to present a second action recommended as the next action of the first action to the user who took the first action accumulates the data used by the large language model to infer the second action. A second user information acquisition unit that acquires second user action data indicating the actions of a second user who is a follower of the first user, and when the second user vector data, which is data obtained by converting the second user action data into a vector, is similar to the first user vector data, a recording unit that records the second user vector data, a detection unit that detects a predetermined action by the second user, and the detection unit detects the predetermined action by the second user. A generation unit that generates a prompt for requesting the presentation of the second action when the second user who takes the action specified by the recorded second user vector data takes the predetermined action as the first action, and the large language model outputs the output information by inputting the prompt generated by the generation unit into the large language model. An output information presentation unit that presents the output information to the second user.

[0007] The first user may be a user having a specific preference, and the storage unit may store the first user vector data corresponding to the actions related to the specific preference among the plurality of actions of the first user.

[0008] The information processing apparatus may further include an issuance processing unit that executes a process of issuing a non-fungible token for the second user, and the output information presentation unit may present the output information to the second user on the condition that the second user is permitted to use the non-fungible token.

[0009] When the second user pays a predetermined fee, the issuance processing unit may execute a process of issuing the non-fungible token, or the information processing apparatus may further include a privilege granting unit that grants a privilege to the first user using the predetermined fee paid by the second user as a source.

[0010] When the second user pays a predetermined fee, the issuance processing unit may execute a process of issuing the non-fungible token, or the information processing apparatus may further include a calculation unit that calculates the predetermined fee based on the presentation frequency at which the output information is presented to the second user.

[0011] The issuance processing unit may execute a process of issuing the non-fungible token that has the second user as its owner, or the recording unit may record the second user vector data corresponding to the second user in association with the non-fungible token that the second user owns. The information processing apparatus may include a token list presentation unit that presents a list of commitment tokens, which are non-fungible tokens whose owners have consented to use by other users, to a third user, and a reception unit that receives a designation of a commitment token from the list of commitment tokens from the third user. The detection unit may detect the predetermined action by the third user, and the generation unit may generate the prompt corresponding to the second user vector data associated with the commitment token designated by the third user when the detection unit detects the predetermined action by the third user. The output information presentation unit may present the output information output by the large language model by inputting the prompt generated by the generation unit into the large language model to the third user.

[0012] When the reception unit receives a designation of a commitment token from the third user, the information processing apparatus may further include a privilege granting unit that grants a privilege to the first user whose first user vector data is similar to the second user vector data associated with the commitment token.

[0013] The second user information acquisition unit may acquire post-presentation second user behavior data indicating the behavior of the second user after the output information presentation unit presents the output information to the second user. When the post-presentation second user vector data, which is data obtained by converting the post-presentation second user behavior data into a vector, is similar to the first user vector data, the recording unit may record the post-presentation second user vector data as the second user vector data.

[0014] The information processing method according to the second aspect of the present invention includes steps of: a computer acquiring first user behavior data indicating a plurality of behaviors of a first user; accumulating, as data used by a large language model for inferring a second action for presenting the second action recommended as the next action of the first action to the user who took the first action, first user vector data which is data obtained by converting the first user behavior data into a vector; a computer acquiring second user behavior data indicating the behavior of a second user who is a follower of the first user; recording the second user vector data when the second user vector data, which is data obtained by converting the second user behavior data into a vector, is similar to the first user vector data; detecting a predetermined action by the second user; generating a prompt for requesting the presentation of the second action when the second user who takes the action specified by the recorded second user vector data takes the predetermined action as the first action; and presenting, to the second user, output information output by the large language model by inputting the generated prompt into the large language model.

[0015] The program according to the third aspect of the present invention causes a computer to function as a first user information acquisition unit that acquires first user action data indicating a plurality of actions of a first user, a first user vector data that is data obtained by converting the first user action data into a vector, and stores the data as data used by a large language model to infer a second action to be presented as a next action of the first action to a user who has taken the first action; a second user information acquisition unit that acquires second user action data indicating the actions of a second user who is a follower of the first user; a recording unit that records the second user vector data when the second user vector data, which is data obtained by converting the second user action data into a vector, is similar to the first user vector data; a detection unit that detects a predetermined action by the second user; a generation unit that generates a prompt for requesting the presentation of the second action when the second user who takes the action specified by the recorded second user vector data takes the predetermined action as the first action; and an output information presentation unit that presents the output information output by the large language model by inputting the prompt generated by the generation unit into the large language model, to the second user.

Effect of the Invention

[0016] According to the present invention, there is an effect that a mechanism can be provided that makes it easier for a second user to utilize the experience content of a first user.

Brief Description of the Drawings

[0017]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Embodiments for Carrying Out the Invention

[0018] [Overview of Information Processing System S] FIG. 1 is a diagram for explaining the overview of the information processing system S. The information processing system S is a system for providing an information processing service. The information processing service is a service for presenting the experience content of a first user to a second user. Specifically, it is a service for presenting, to a second user who has taken a first action, a second action that the first user took after the first action in the past, as a recommended action. The first user is a user who is proficient in a specific field or has a specific hobby, such as food, fashion, tourism, music, etc. Note that the first user may be a user who is active in a specific area such as an SNS (Social Networking Service), a video distribution service, etc. The second user is another user different from the first user and follows the first user.

[0019] The actions (the first action and the second action) in the present embodiment are direct actions taken by the user, such as movement, stay, and operation of the terminal by the user who holds the terminal, but may further include geographical elements, time elements, etc. The second action is an action indicating the user's experience content. For example, when the first action is getting off at Shibuya Station, the action of eating or drinking at a store existing around Shibuya Station; when the first action is eating or drinking at a first store, the action of eating or drinking at a second store; when the first action is staying at the user's home, the action of playing a predetermined music; when the first action is playing a first music, the action of playing a second music, etc. The information processing system S includes a user terminal 1 and an information processing apparatus 2.

[0020] The user terminal 1 is a terminal used by a user (the first user or the second user) who uses an information processing service, and is, for example, a smartphone, a tablet terminal, or the like. A dedicated application program (hereinafter referred to as the "dedicated app") for providing the information processing service to the user is installed on the user terminal 1. The dedicated app has a function of collecting the user's behavior data and a function of presenting recommended behaviors. The behavior data is data indicating the user's behavior, and is, for example, the position history of the user terminal 1 and the operation history of the user terminal 1, or information specified by the position history of the user terminal 1 and the operation history of the user terminal 1 (for example, the frequency of the user visiting a store, the time and hour when the user visits a store, the location of the store where the user visits, the services and products purchased by the user at the store, etc.).

[0021] The information processing device 2 is a device that manages the information processing service, and is, for example, a server. Information about the first user and information about the second user are stored in the information processing device 2. For example, the information about the first user includes information indicating the target behavior, which is the first behavior that triggers the second behavior. Also, for example, the information about the second user includes information indicating the first user followed by the second user.

[0022] In addition, a large language model used to infer the second behavior to be taken after the first behavior is stored in the information processing device 2. The large language model is, for example, ChatGPT. Time-series data indicating the behavior of the first user, that is, data indicating the experience content of the first user, is accumulated in the large language model.

[0023] In addition, data indicating the behavior of the second user similar to the behavior of the first user is stored in the information processing device 2. The data indicating the behavior of the second user similar to the behavior of the first user is data used to identify, for example, the behavior of the second user itself, the characteristics of the behavior of the second user, the tendency of the behavior of the second user, etc.

[0024] Hereinafter, the processing of the information processing system S will be described. As a prerequisite for FIG. 1, the information processing apparatus 2 stores information about the first user including information indicating an action of staying in Shibuya Ward as a target action which is the first action that triggers the second action. Also, in the large language model, time-series data indicating the actions of the first user including data indicating the action of eating and drinking at Store A, which is a Japanese restaurant where the first user who is proficient in food exists in Shibuya Ward, is accumulated. Further, the information processing apparatus 2 stores data indicating the actions of the second user similar to the actions of the first user including data indicating the action of eating and drinking at Store B, which is a Japanese restaurant.

[0025] In this case, first, when the second user who is on the train gets off at Shibuya Station, the user terminal 1 of the second user collects second user action data indicating the action of the second user getting off at Shibuya Station, and transmits the collected second user action data to the information processing apparatus 2 ((1) in FIG. 1).

[0026] Based on the second user action data acquired from the user terminal 1 of the second user, the information processing apparatus 2 detects a predetermined action by the second user ((2) in FIG. 1). The predetermined action is the target action of the first user followed by the second user. For example, it is an action of the second user staying in Shibuya Ward (for example, an action of the second user getting off at Shibuya Station).

[0027] When the information processing apparatus 2 detects a predetermined action by the second user, it uses the large language model to infer the second action to be taken next after the first action when the second user who takes the action specified by the data indicating the actions of the second user similar to the actions of the first user stored takes the predetermined action as the first action ((3) in FIG. 1). For example, the information processing apparatus 2 infers that the second action to be taken next after the action of staying in Shibuya Ward which is the first action is to eat and drink at Store A.

[0028] Then, the information processing device 2 transmits the inference result to the user terminal 1 of the second user. The information processing device 2 notifies the user terminal 1 of the second user with a message recommending dining at Store A as the inference result. After that, the user terminal 1 of the second user causes the inference result to be displayed on the display.

[0029] By doing so, the information processing system S can present the second action taken by the first user after the first action to the second user who has taken the first action in the past. Thereby, the information processing system S can cause the second user to recognize the experience content of the first user in an environment where the experience content of the first user can be utilized. As a result, the information processing system S can provide a mechanism that makes it easier for the second user to utilize the experience content of the first user. Hereinafter, the configuration of the information processing device 2 will be described.

[0030] [Functional Configuration of Information Processing Device 2] FIG. 2 is a diagram schematically showing the functional configuration of the information processing device 2. The information processing device 2 includes a communication unit 21, a storage unit 22, and a control unit 23. In FIG. 2, the arrows indicate the main data flow, and there may be a data flow not shown in FIG. 2. In FIG. 2, each functional block shows a configuration in terms of functional units, not in terms of hardware (devices). Therefore, the functional blocks shown in FIG. 2 may be implemented in a single device, or may be divided and implemented in a plurality of devices. The exchange of data between the functional blocks may be performed via any means such as a data bus, a network, a portable storage medium, or the like.

[0031] The communication unit 21 is a communication interface for connecting to a network and has a communication controller for receiving data from an external device.

[0032] The storage unit 22 is a large-capacity storage device such as a ROM (Read Only Memory) that stores the BIOS (Basic Input Output System) of a computer that realizes the information processing apparatus 2, a RAM (Random Access Memory) that serves as a working area of the information processing apparatus 2, an OS (Operating System), application programs, and an HDD (Hard Disk Drive) or SSD (Solid State Drive) that stores various information referred to during the execution of the application programs. The storage unit 22 stores a first user management database that manages information about the first user and a second user management database that manages information about the second user.

[0033] FIG. 3 is a diagram showing an example of the configuration of the first user management database. In the example shown in FIG. 3, the first user management database stores by associating a user ID, an attribute, and a target action. The attribute is information indicating, for example, the type of the first user (a user proficient in a specific field, a user active in a predetermined area, etc.) and the field in which the first user is proficient. The target action is, for example, an action designated by the first user, an action specified based on the tendency of the actions of the first user, etc. Note that the configuration of the first user management database is not limited to the example shown in FIG. 3, and may further include information about members of the information processing service, such as the address of the first user.

[0034] FIG. 4 is a diagram showing an example of the configuration of the second user management database. In the example shown in FIG. 4, the second user management database stores by associating a user ID, a followed user ID, and payment status. The followed user ID is the ID of the first user followed by the second user. The payment status is information indicating whether a predetermined fee has been paid. The predetermined fee is, for example, a fee associated with the use of an information processing service, a fee associated with the issuance of a non-fungible token, etc. Although details will be described later, the non-fungible token is information issued by a blockchain with the second user as the owner. Note that the configuration of the second user management database is not limited to the example shown in FIG. 4, and may further include information related to members of the information processing service, such as the address of the second user, etc.

[0035] Returning to FIG. 2, the control unit 23 is a processor such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), or NPU (Neural network Processing Unit) of the information processing apparatus 2, and by executing the program stored in the storage unit 22, functions as an acquisition processing unit 231, a model management unit 232, a recording processing unit 233, an issuance processing unit 234, a calculation unit 235, a detection unit 236, a presentation unit 237, a reception processing unit 238, and a privilege granting unit 239.

[0036] [Accumulation Processing] First, the accumulation processing executed by the information processing apparatus 2 will be described. The accumulation processing is a process of accumulating data indicating the experience content of the first user, specifically, a process of accumulating data related to the actions of the first user used for inference by a large language model.

[0037] The acquisition processing unit 231 functions as a first user information acquisition unit and acquires first user action data from the user terminal 1 of the first user. The first user action data is time-series action data indicating a plurality of actions of the first user.

[0038] The acquisition processing unit 231 acquires the first user behavior data from the user terminal 1 of each first user at a predetermined interval (for example, every hour, every day, every week, etc.). Note that the acquisition processing unit 231 may acquire the data collected by the dedicated application as the first user behavior data each time the dedicated application collects data (each time the first user acts).

[0039] When the acquisition processing unit 231 acquires the first user behavior data, it generates first user vector data, which is data obtained by converting the first user behavior data into a vector. The acquisition processing unit 231 can convert the first user behavior data into a vector using, for example, known techniques.

[0040] The model management unit 232 functions as a storage unit and stores the first user vector data generated by the acquisition processing unit 231 as data used by the large language model, which presents the second action recommended as the next action after the first action for the user who took the first action, for inferring the second action. The model management unit 232, for example, uses each first user vector data generated by the acquisition processing unit 231 to let the large language model learn the actions of each first user. The model management unit 232 may store each first user vector data in the storage unit 22 as external data to be referred to when the large language model infers the second action.

[0041] Here, the first user behavior data of the first user who is proficient in a specific field or has a specific preference may include behaviors not related to the specific field or specific preference. For example, when the first user is proficient in food, in addition to a series of behaviors for eating and drinking in a restaurant (for example, the behavior of going to the restaurant, the behavior of eating and drinking in the restaurant, etc.), the first user behavior data may include behaviors such as staying at the first user's home and behaviors for using facilities other than restaurants that are not related to food. If such data is accumulated, the inference accuracy of the large language model may decrease. Therefore, the model management unit 232 may accumulate first user vector data from which behaviors not related to the specific field or specific preference that the first user is proficient in are excluded.

[0042] For example, first, the acquisition processing unit 231 generates first user vector data filtered by behaviors related to the specific field that the first user is proficient in or the specific preference that the first user has. The acquisition processing unit 231 generates first user vector data, for example, after filtering the first user behavior data. The acquisition processing unit 231 may filter the first user vector data after generating the first user vector data. The acquisition processing unit 231 can generate, for example, using known techniques, first user vector data filtered by behaviors related to the specific field that the first user is proficient in or the specific preference that the first user has. Then, the model management unit 232 accumulates the first user vector data filtered by behaviors related to the specific field that the first user is proficient in or the specific preference that the first user has.

[0043] Note that the model management unit 232 may accumulate the first user vector data without filtering. For example, a dedicated application is provided with a function of setting whether to permit the collection of behavior data according to a user's operation. When the first user uses the user terminal 1 to set permission for collecting behavior data and setting rejection of collecting behavior data, the acquisition processing unit 231 acquires the first user behavior data collected during a period from when the collection of behavior data is permitted until the collection of behavior data is rejected (for example, a period of taking actions related to a specific field or a period of activities in a specific area). Then, the model management unit 232 accumulates the first user vector data generated based on the first user behavior data acquired by the acquisition processing unit 231.

[0044] [Recording process] Subsequently, the recording process executed by the information processing apparatus 2 will be described. The recording process is a process of recording data for specifying a tendency of the second user's behavior similar to the first user's behavior among tendencies of a plurality of behaviors of the second user.

[0045] The acquisition processing unit 231 further functions as a second user information acquisition unit, and acquires second user behavior data from the user terminal 1 of the second user. The second user behavior data is behavior data indicating the behavior of the second user who is a follower of the first user. The acquisition processing unit 231 acquires the second user behavior data from the user terminal 1 of the second user at a predetermined interval (for example, every hour, every day, every week, etc.). Note that the acquisition processing unit 231 may acquire the data collected by the dedicated application as the second user behavior data each time the dedicated application collects data (each time the second user takes an action).

[0046] When the acquisition processing unit 231 acquires the second user behavior data, the acquisition processing unit 231 generates second user vector data which is data obtained by converting the second user behavior data into a vector. The acquisition processing unit 231 can convert the second user behavior data into a vector using, for example, a known technique.

[0047] When the second user vector data generated by the acquisition processing unit 231 is similar to the first user vector data, the recording processing unit 233 records the second user vector data. Specifically, when the second user vector data generated by the acquisition processing unit 231 is similar to the first user vector data corresponding to the first user followed by the second user corresponding to the second user vector data, the recording processing unit 233 records the second user vector data.

[0048] For example, the recording processing unit 233 calculates the similarity between the first user vector data and the second user vector data using a known method such as cosine similarity or Euclidean distance, and records the second user vector data when the calculated similarity is less than a predetermined threshold. The predetermined threshold is a numerical value predetermined for use when the recording processing unit 233 determines whether the actions of the first user and the second user are similar.

[0049] For example, as the recording of the second user vector data, the recording processing unit 233 associates the ID of the second user with the second user vector data and stores it in the storage unit 22. The recording processing unit 233 may store the second user vector data in association with a non-fungible token owned by the second user. The information processing apparatus 2 stores the second user vector data in association with a non-fungible token owned by the second user, for example, by executing the following two steps.

[0050] As a first step, the issuance processing unit 234 executes a process of issuing a non-fungible token of the second user. Specifically, the issuance processing unit 234 executes a process of issuing a non-fungible token owned by the second user.

[0051] The issuance processing unit 234, for example, requests the blockchain to issue a non-fungible token with the second user as the owner, thereby causing the blockchain to issue the non-fungible token. The issuance processing unit 234 executes a process of issuing a non-fungible token of the second user, for example, when the second user registers as a member of an information processing service.

[0052] The issuance processing unit 234 may execute a process of issuing a non-fungible token when the second user pays a predetermined fee. The predetermined fee is, for example, a fee associated with the use of an information processing service, a fee associated with the issuance of a non-fungible token, etc. The predetermined fee may be a fee paid only once, or a fee paid at predetermined intervals (for example, monthly, annually, etc.). The issuance processing unit 234, for example, refers to the presence or absence of a payment associated with the ID of the second user in the user management database, and executes a process of issuing a non-fungible token when the second user pays the predetermined fee.

[0053] Here, if the predetermined fee is a fee paid at a predetermined interval, the information processing device 2 may calculate the predetermined fee to be charged to the second user according to the presentation frequency of the output information presented to the second user by the presentation unit 237 described later. Specifically, the calculation unit 235 calculates the predetermined fee based on the presentation frequency of the output information presented to the second user.

[0054] For example, the calculation unit 235 calculates the predetermined cost to be higher the higher the presentation frequency of the output information, and calculates the predetermined cost to be lower the lower the presentation frequency of the output information. Note that the calculation unit 235 may calculate the predetermined cost to be lower the higher the presentation frequency of the output information, and calculate the predetermined cost to be higher the lower the presentation frequency of the output information. In this way, the information processing device 2 can charge the second user a cost according to the presentation frequency of the output information.

[0055] As a second step, the recording processing unit 233 records the second user vector data corresponding to the second user in association with the non-fungible token of which the second user is the owner. For example, the recording processing unit 233 requests the blockchain to store the second user vector data, thereby storing the second user vector data in association with the non-fungible token of which the second user is the owner. The recording processing unit 233 may store in the storage unit 22 the ID of the non-fungible token of which the second user is the owner and the second user vector data in association with each other.

[0056] Also, the predetermined fee may be a fee associated with the recording of the second user vector data. In this case, when the second user pays the predetermined fee, the recording processing unit 233 records the second user vector data corresponding to the second user in association with the non-fungible token of which the second user is the owner. By doing so, the information processing apparatus 2 can record data for identifying the tendency of the behavior of the second user similar to the behavior of the first user in association with the non-fungible token.

[0057] [Presentation Processing] Subsequently, the presentation processing executed by the information processing apparatus 2 will be described. The presentation processing is processing for presenting the experience content of the first user to the second user, specifically, processing for presenting, to the second user who has taken a predetermined action, an action recommended as the next action to be taken after the predetermined action.

[0058] The detection unit 236 detects a predetermined action by the second user. Specifically, first, the detection unit 236 detects, based on the second user behavior data acquired by the acquisition processing unit 231, the target action of the first user followed by the second user as the predetermined action. Note that the detection unit 236 may detect a predetermined time as an element other than the behavior of the second user. For example, in the first user management database, the target time is stored in association with the user ID, and the detection unit 236 detects the predetermined time when the current time reaches the target time of the first user followed by the second user.

[0059] The model management unit 232 further functions as a generation unit, and generates a prompt to be input to the large language model when the detection unit 236 detects a predetermined action by the second user. Specifically, when the second user who tends to perform the action specified by the recorded second user vector data performs a predetermined action as the first action, the model management unit 232 generates a prompt for requesting the presentation of the second action.

[0060] For example, when the second user who tends to perform the action specified by one or more second user vector data associated with the non-fungible token of which the second user is the owner performs a predetermined action as the first action, the model management unit 232 generates a prompt for requesting the presentation of the second action.

[0061] Also, for example, when the detection unit 236 detects a predetermined time, the model management unit 232 may generate a prompt for requesting the presentation of the action to be taken when the second user who tends to perform the action specified by one or more second user vector data associated with the non-fungible token of which the second user is the owner has passed the predetermined time.

[0062] When the model management unit 232 generates a prompt, it obtains output information indicating the inference result output by the large language model by inputting the prompt to the large language model.

[0063] The presentation unit 237 functions as an output information presentation unit, and presents the output information output by the large language model by inputting the prompt generated by the model management unit 232 to the large language model to the second user. Specifically, the presentation unit 237 transmits the output information to the user terminal 1 of the second user, thereby causing the output information to be displayed on the user terminal 1 of the second user.

[0064] The prompting unit 237 may prompt the second user with the output information when the second user meets a predetermined condition. Specifically, the prompting unit 237 prompts the second user with the output information on the condition that the use of non-fungible tokens is permitted for the second user. The condition for permitting the use of non-fungible tokens for the second user is, for example, that the second user has paid a predetermined fee such as a fee associated with the use of the information processing service or a fee for purchasing the right to use non-fungible tokens.

[0065] The prompting unit 237 refers to, for example, the payment status associated with the ID of the second user in the user management database, and when it determines that the use of non-fungible tokens is permitted for the second user because the second user has paid a predetermined fee, the prompting unit 237 prompts the second user with the output information. By doing so, the information processing apparatus 2 can prompt the second user who wishes to be prompted with the recommended action after paying a predetermined fee with the output information.

[0066] The information processing apparatus 2 may execute a recording process after prompting the output information. Specifically, first, the acquisition processing unit 231 acquires post-prompt second user action data indicating the actions of the second user after the prompting unit 237 has prompted the second user with the output information. Then, when the post-prompt second user vector data, which is data obtained by converting the post-prompt second user action data into a vector, is similar to the first user vector data, the recording processing unit 233 records the post-prompt second user vector data as the second user vector data. By doing so, the information processing apparatus 2 can record the actions of the second user after the output information has been prompted as the experience content of the second user.

[0067] Here, for a non-fungible token whose second user is the owner, since the experience content of the second user (the experience content that the second user has acted on alone or the experience content that the second user has acted on based on the experience of the first user) is accumulated, it is considered that the accumulated experience content of the second user has a certain value. Therefore, the information processing apparatus 2 may present the accumulated experience content of the second user to a third user who is another user different from the second user. Specifically, the accumulated experience content of the second user is presented to the third user by executing the following four steps.

[0068] As a first step, the presentation unit 237 further functions as a token list presentation unit and presents a list of the issued non-fungible tokens to the third user. The list of non-fungible tokens includes, for example, information indicating the owner of the non-fungible token, information indicating the second user vector data associated with the non-fungible token (such as a title like the experience of eating and drinking at a restaurant in Shibuya Ward), and the like.

[0069] The presentation unit 237 may present a list of non-fungible tokens whose owners have consented to use by other users to the third user. For example, the storage unit 22 stores information regarding non-fungible tokens. The information regarding non-fungible tokens includes the ID of the owner of the non-fungible token and information indicating the presence or absence of consent for use of the non-fungible token by other users. Note that the information indicating the presence or absence of consent for use of the non-fungible token by other users may be included in the non-fungible token issued by the blockchain. In this case, the presentation unit 237 refers to the information regarding the non-fungible token and causes a list of non-fungible tokens for which the owners of the non-fungible tokens have consented to use by other users among the plurality of non-fungible tokens to be displayed on the user terminal 1 of the third user.

[0070] As a second step, the reception processing unit 238 receives from the third user a designation of a non-fungible token from the list of non-fungible tokens. When the reception processing unit 238 receives the designation of the non-fungible token, it associates the ID of the third user with the ID of the non-fungible token and stores them in the storage unit 22.

[0071] As a third step, the detection unit 236 detects a predetermined action by the third user. The predetermined action is the target action of the first user followed by the owner of the non-fungible token designated by the third user.

[0072] As a fourth step, upon the detection unit 236 detecting a predetermined action by the third user, the model management unit 232 infers an action to be recommended as the next action to be taken by the third user after the predetermined action. Specifically, first, the model management unit 232 generates a prompt corresponding to the second user vector data associated with the non-fungible token designated by the third user. For example, when the third user takes an action specified by the second user vector data associated with the designated non-fungible token, the model management unit 232 generates a prompt for requesting the presentation of a second action when the predetermined action is taken as the first action. Then, the model management unit 232 obtains output information indicating the inference result output by the large language model by inputting the generated prompt into the large language model.

[0073] As a fifth step, the presentation unit 237 presents the output information output by the large language model, which is obtained by inputting the prompt generated by the model management unit 232 into the large language model, to the third user. Specifically, the presentation unit 237 transmits the output information to the user terminal 1 of the third user, causing the output information to be displayed on the user terminal 1 of the third user. By doing so, the information processing apparatus 2 can provide the accumulated experience content of the second user to the third user.

[0074] The information processing apparatus 2 may give a privilege to the first user who provided the experience content to the second user. The privilege may be, for example, cash, virtual currency, points available for information processing services, or the like.

[0075] Specifically, the privilege granting unit 239 gives a privilege to the first user using the predetermined fee paid by the second user as a source. For example, in the information processing apparatus 2, a ratio to be applied to the privilege fee is predetermined from the predetermined fee paid by the second user, and the privilege granting unit 239 calculates the privilege fee based on the predetermined fee paid by the second user and the predetermined ratio, and gives the privilege corresponding to the calculated fee to the first user. By doing so, the information processing apparatus 2 can provide an incentive for the first user to provide experience content to other users.

[0076] When the reception processing unit 238 receives the designation of the non-fungible token from the third user, the privilege granting unit 239 may give a privilege to the first user followed by the owner of the non-fungible token. Specifically, when the reception processing unit 238 receives the designation of the non-fungible token from the third user, the privilege granting unit 239 gives a privilege to the first user of the first user vector data similar to the second user vector data associated with the non-fungible token.

[0077] By doing so, the information processing apparatus 2 can provide an incentive for the first user to provide experience content to other users. Note that the privilege granting unit 239 may further give a privilege to the second user who is the owner of the non-fungible token from the third user to the reception processing unit 238.

[0078] [Processing of Information Processing Apparatus 2] Subsequently, the flow of processing executed by the information processing apparatus 2 will be described. FIG. 5 is a flowchart showing the flow of processing executed by the information processing apparatus 2. This flowchart starts when the acquisition processing unit 231 acquires the second user behavior data (S1).

[0079] When the detection unit 236 fails to detect a predetermined action by the second user based on the second user behavior data acquired by the acquisition processing unit 231 (when the answer is NO in S2), the process returns to S1. On the other hand, when the detection unit 236 detects a predetermined action by the second user based on the second user behavior data (when the answer is YES in S2), the model management unit 232 generates a prompt for requesting the presentation of a second action when the second user who takes the action specified by the recorded second user vector data takes the predetermined action as the first action (S3).

[0080] The model management unit 232 acquires output information indicating the inference result output by the large language model by inputting the generated prompt into the large language model (S4). Then, the presentation unit 237 presents the output information acquired by the model management unit 232 to the second user (S5).

[0081] [Effects of the present embodiment] As described above, when the information processing apparatus 2 detects a predetermined action by the second user, the information processing apparatus 2 generates a prompt for requesting the presentation of a second action when the second user who takes the action specified by the second user vector data similar to the first user vector data takes the predetermined action as the first action, and inputs the prompt into the large language model in which the first user vector data is accumulated, and presents the output information output by the large language model to the second user. By doing so, the information processing apparatus 2 can present the second action that the first user took after the first action to the second user who took the first action. As a result, the information processing apparatus 2 can make the second user recognize the experience content of the first user in an environment where the experience content of the first user can be utilized. As a result, the information processing apparatus 2 can provide a mechanism that makes it easier for the second user to utilize the experience content of the first user.

[0082] Note that the present invention can contribute to Goal 9, "Build the infrastructure for industry and innovation," of the Sustainable Development Goals (SDGs) led by the United Nations.

[0083] As described above, the present invention has been described using embodiments. However, the technical scope of the present invention is not limited to the scope described in the above embodiments, and various modifications and changes are possible within the scope of the gist. For example, all or part of the device can be configured by functionally or physically dispersing and integrating it in any unit. Also, new embodiments resulting from any combination of a plurality of embodiments are included in the embodiments of the present invention. The effects of the new embodiments resulting from the combination have the effects of the original embodiments combined.

Explanation of Reference Numerals

[0084] 1 User terminal 2 Information processing apparatus 21 Communication unit 22 Storage unit 23 Control unit 231 Acquisition processing unit 232 Model management unit 233 Recording processing unit 234 Issuance processing unit 235 Calculation unit 236 Detection unit 237 Presentation unit 238 Reception processing unit 239 Privilege granting unit S Information processing system

Claims

1. a first user information acquisition unit that acquires first user behavior data indicating a plurality of behaviors of a first user; a storage unit that stores first user vector data, which is data obtained by converting the first user action data into a vector, as data used by a large-scale language model that is used to present a second action recommended as a next action following the first action to a user who has taken a first action, to infer the second action; a second user information acquisition unit that acquires second user behavior data indicating behavior of a second user who is a follower of the first user; a recording unit that records second user vector data, which is data obtained by converting the second user behavior data into a vector, when the second user vector data is similar to the first user vector data; A detection unit that detects a predetermined action by the second user; a generation unit that generates, in response to the detection unit detecting the predetermined behavior of the second user, a prompt for requesting presentation of the second behavior in a case where the second user, who takes a behavior specified by the recorded second user vector data, takes the predetermined behavior as the first behavior; and an output information presenting unit that presents to the second user output information output by the large-scale language model by inputting the prompt generated by the generation unit into the large-scale language model; An information processing device having the above configuration.

2. The first user is a user having a particular preference, The storage unit stores the first user vector data corresponding to an action related to the specific preference among a plurality of actions of the first user. The information processing device according to claim 1 .

3. The information processing device further includes an issuance processing unit that executes a process of issuing a non-fungible token of the second user, The output information presenting unit presents the output information to the second user on the condition that the second user is permitted to use the non-fungible token. The information processing device according to claim 1 .

4. The issuance processing unit executes a process of issuing the non-fungible token when the second user pays a predetermined fee, The information processing device further includes a reward providing unit that provides a reward to the first user using the predetermined fee paid by the second user as a resource. The information processing device according to claim 3 .

5. The issuance processing unit executes a process of issuing the non-fungible token when the second user pays a predetermined fee, The information processing device further includes a calculation unit that calculates the predetermined cost based on a presentation frequency at which the output information is presented to the second user. The information processing device according to claim 3 .

6. The issuance processing unit executes a process of issuing the non-fungible token whose owner is the second user, The recording unit records the second user vector data corresponding to the second user in association with the non-fungible token owned by the second user; The information processing device includes: a token list presentation unit that presents to a third user a list of consent tokens, which are non-fungible tokens that the owner of the non-fungible token has consented to be used by other users; a reception unit that receives, from the third user, a designation of the agreement token from the list of agreement tokens; having The detection unit detects the predetermined action by the third user, the generation unit generates the prompt corresponding to the second user vector data associated with the consent token designated by the third user, when the detection unit detects the predetermined behavior by the third user; the output information presenting unit presents to the third user the output information output by the large-scale language model by inputting the prompt generated by the generation unit into the large-scale language model. The information processing device according to claim 3 .

7. The information processing device further includes a privilege granting unit that grants a privilege to the first user of the first user vector data similar to the second user vector data associated with the acceptance token when the acceptance token is designated by the third user. The information processing device according to claim 6.

8. the second user information acquisition unit acquires post-presentation second user behavior data indicating a behavior of the second user after the output information presentation unit presents the output information to the second user; The recording unit records the post-presentation second user vector data as the second user vector data when the post-presentation second user vector data, which is data obtained by converting the post-presentation second user behavior data into a vector, is similar to the first user vector data. The information processing device according to claim 1 .

9. The computer executes acquiring first user behavior data indicative of a plurality of behaviors of a first user; a step of storing first user vector data, which is data obtained by converting the first user action data into a vector, as data used by a large-scale language model used to present a second action recommended as a next action of the first action to a user who has taken the first action, for inferring the second action; acquiring second user behavior data indicating behavior of a second user who is a follower of the first user; a step of recording second user vector data, which is data obtained by converting the second user behavior data into a vector, when the second user vector data is similar to the first user vector data; detecting a predetermined action by the second user; generating a prompt for requesting presentation of the second behavior when the second user, who takes a behavior specified by the recorded second user vector data, takes the predetermined behavior as the first behavior, in response to detection of the predetermined behavior by the second user; presenting output information output by the large-scale language model by inputting the generated prompt into the large-scale language model to the second user; An information processing method comprising the steps of:

10. Computer, a first user information acquisition unit that acquires first user behavior data indicating a plurality of behaviors of a first user; a storage unit that stores first user vector data, which is data obtained by converting the first user action data into vectors, as data used by a large-scale language model that is used to present a second action recommended as a next action following the first action to a user who has taken a first action, to infer the second action; a second user information acquisition unit that acquires second user behavior data indicating behavior of a second user who is a follower of the first user; a recording unit that records second user vector data, which is data obtained by converting the second user behavior data into a vector, when the second user vector data is similar to the first user vector data; A detection unit that detects a predetermined action by the second user; a generation unit that generates, in response to the detection unit detecting the predetermined behavior of the second user, a prompt for requesting presentation of the second behavior in a case where the second user, who takes a behavior specified by the recorded second user vector data, takes the predetermined behavior as the first behavior, when the second user takes the predetermined behavior as the first behavior; and an output information presenting unit that presents to the second user output information output by the large-scale language model by inputting the prompt generated by the generation unit into the large-scale language model; A program to function as a

Images