Information processing device, information processing method, information processing program, and system

Abstract

To improve the usefulness of the information provided to users. [Solution] The information processing device 1 performs the following: a first acquisition process for acquiring user attribute information; a second acquisition process for acquiring location information of the user terminal 2 and time information corresponding to the location information; a first generation process for generating user behavior pattern information based on the history of location information and time information; a storage process for associating the user attribute information and behavior pattern information and storing them as user information 32; a selection process for selecting candidate users for the first user based on the user information 32; a second generation process for generating information 37 based on the behavior of the candidate users based on the user attribute information and behavior pattern information; and a provision process for providing predetermined provision information based on the behavior-based information 37 to the user terminal 2A of the first user.

Description

Technical Field

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

Background Art

[0002] In recent years, as in so-called personalized advertisements, it has been practiced to select content to be provided to a user based on the user's attributes and the user's action history on the web or the like, and provide different content for each user.

[0003] As such a technique, for example, Patent Document 1 discloses that an action tendency considered to be of high interest to a user (for example, categories such as sports and history) is extracted from the user's access information to an external network and input information on the external network, and an action proposal (for example, presenting an access destination of information considered to be of high interest) is provided to the user based on the action tendency. Further, Patent Document 1 also proposes setting the time or position for providing the action proposal based on the user's access time to the external network and the position information at the time of access.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] By the way, there is a type of knowledge called experiential knowledge and tacit knowledge. Experiential knowledge refers to the knowledge and insights gained from the user's actual actions and experiences. Tacit knowledge, on the other hand, refers to the knowledge and insights gained from the user's actual actions and experiences, such as knowledge that the user has acquired unconsciously or knowledge that is difficult to put into words. In many cases, the users who possess such experiential knowledge and tacit knowledge (hereinafter also referred to as "tacit knowledge, etc.") are not even aware of it.

[0006] When providing information to users, if information can be provided not only based on verbalized information that the user is aware of, but also on tacit knowledge that is not verbalized—that is, knowledge and insights gained from the actual experiences of the user or other users—it is possible to provide information that is more suitable for the user, or to give the user insights into things they themselves were unaware of. In other words, it is thought that the usefulness of the information provided to users can be increased.

[0007] However, Patent Document 1 selects information that is likely to be of high interest to the user based only on user access information to the external network and input information on the external network, i.e., verbalized information, and makes no mention of experiential knowledge or tacit knowledge.

[0008] This disclosure has been made in view of the above-mentioned issues. In other words, the purpose of this disclosure is to improve the usefulness of the information provided to users. [Means for solving the problem]

[0009] An information processing apparatus according to one aspect of the present disclosure comprises at least one processor, a storage unit for storing an information processing program, and a communication unit capable of communicating with user terminals operated by each user, wherein the processor executes the information processing program to perform a first acquisition process for acquiring user attribute information relating to the user's attributes, a second acquisition process for acquiring location information of the user terminal and time information corresponding to the location information, and a first trained model trained to output patterns of the user's behavior based on the history of the user's location information and time information, and The system performs the following steps: a first generation process for generating behavioral pattern information relating to a user; a storage process for associating the user attribute information and the behavioral pattern information for each user and storing them as user information; a selection process for selecting at least one candidate user for the first user based on the user information; a second generation process for generating information based on the behavior of the candidate users using a second trained model that has been trained to output information based on the user's behavior based on the user attribute information and the behavioral pattern information of the user; and a provision process for providing predetermined information based on the behavioral information to the first user's user terminal.

[0010] An information processing method according to one aspect of the present disclosure is an information processing method performed by an information processing device equipped with a communication unit capable of communicating with a user terminal operated by each user, and includes: a first acquisition step of acquiring user attribute information relating to the attributes of the user; a second acquisition step of acquiring location information of the user terminal and time information corresponding to the location information; a first generation step of generating behavior pattern information relating to the behavior patterns of the user using a first trained model that has been trained to output behavior patterns of the user based on the history of the user's location information and time information; a storage step of associating the user attribute information and the behavior pattern information for each user and storing them as user information; a selection step of selecting at least one candidate user for the first user based on the user information; a second generation step of generating behavior-based information of the candidate user using a second trained model that has been trained to output behavior-based information based on the user's user attribute information and behavior pattern information; and a provision step of providing predetermined provision information based on behavior-based information to the first user's user terminal.

[0011] An information processing program according to one aspect of this disclosure is an information processing device that is executed in an information processing device equipped with a communication unit capable of communicating with user terminals operated by each user, wherein when the information processing program is executed by the processor of the information processing device, the information processing device is instructed to perform a first acquisition process to acquire user attribute information relating to the attributes of the user, a second acquisition process to acquire location information of the user terminal and time information corresponding to the location information, and a first trained model that has been trained to output patterns of the user's behavior based on the history of the user's location information and time information, to output patterns of the user's behavior The system is configured to perform the following steps: a first generation process for generating behavioral pattern information relating to a user; a storage process for associating the user attribute information and the behavioral pattern information for each user and storing them as user information; a selection process for selecting at least one candidate user for the first user based on the user information; a second generation process for generating information based on the behavior of the candidate users using a second trained model that has been trained to output information based on the user's behavior based on the user attribute information and the behavioral pattern information of the user; and a provision process for providing predetermined information based on the behavioral information to the first user's user terminal.

[0012] A system according to one aspect of the present disclosure is a system including a user terminal operated by each user and an information processing device, wherein the information processing device comprises at least one processor, a storage unit for storing an information processing program, and a communication unit capable of communicating with the plurality of user terminals, and the processor executes the information processing program, thereby enabling the information processing device to perform a first acquisition process for acquiring user attribute information relating to the user's attributes, a second acquisition process for acquiring location information of the user terminal and time information corresponding to the location information, and a first learned process which is learned to output patterns of the user's behavior based on the history of the user's location information and time information. The system performs the following steps: a first generation process that generates behavioral pattern information relating to the user's behavioral patterns using a model; a storage process that associates the user attribute information and the behavioral pattern information for each user and stores them as user information; a selection process that selects at least one candidate user for the first user based on the user information; a second generation process that generates information based on the candidate user's behavior using a second trained model that has been trained to output information based on the user's behavior based on the user attribute information and the behavioral pattern information of the user; and a provision process that provides predetermined provision information based on the behavioral information to the first user's user terminal. [Effects of the Invention]

[0013] According to this disclosure, the usefulness of the information provided to users can be improved. [Brief explanation of the drawing]

[0014] [Figure 1] Figure 1 is a schematic diagram showing an example of a system according to the present disclosure. [Figure 2] Figure 2 is a block diagram showing an example of the functional configuration of the system according to the embodiment of this disclosure. [Figure 3] Figure 3 is a data table showing an example of user information according to the embodiment of this disclosure. [Figure 4]FIG. 4 is a data table showing an example of user information according to an embodiment of the present disclosure. [Figure 5] FIG. 5 is a schematic diagram showing an example of a service provided by the system according to the embodiment of the present disclosure. [Figure 6] FIG. 6 is a flowchart showing an example of the information processing method according to the embodiment of the present disclosure. [Figure 7] FIG. 7 is a flowchart showing an example of the acquisition process and the first generation process of FIG. 6. [Figure 8] FIG. 8 is a flowchart showing an example of the second generation process of FIG. 6. [Figure 9] FIG. 9 is a flowchart showing an example of the third generation process of FIG. 6. [Figure 10] FIG. 10 is a diagram showing an example of provision information. [Mode for Carrying Out the Invention]

[0015] Hereinafter, examples of embodiments of the present disclosure will be described with reference to the drawings. Note that the same components and the like are denoted by the same reference numerals, and duplicate descriptions are omitted as appropriate. In addition, the contents of configurations and processes that are not related to the characteristic parts of the embodiments may be omitted.

[0016] (System) FIG. 1 is a schematic diagram showing an example of the system 5 according to the embodiment of the present disclosure. As shown in FIG. 1, the system 5 includes an information processing apparatus 1, and user terminals 2A, 2B, etc. (hereinafter also collectively referred to as "user terminals 2"). The information processing apparatus 1 and the user terminal 2 can be communicatively connected to each other via the network 4, and can transmit and receive data to and from each other.

[0017] The information processing device 1 is a computer that performs various types of information processing. In the present embodiment, the information processing device 1 functions as a server that outputs predetermined provision information (hereinafter simply referred to as "provision information") to the user terminal 2. The information processing device 1 includes, for example, a processor 11, a memory 12, a storage 13, and a communication I / F 14, which are mutually connected via a communication bus. The information processing device 1 may further include other conventionally known constituent elements.

[0018] The processor 11 performs various types of computation and control. For the processor 11, for example, a general-purpose processor such as a CPU (Central Processing Unit) may be used, or a dedicated processor such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array) may be used. The memory 12 includes, for example, storage devices such as a ROM (Read Only Memory) storing various programs and the like, and a RAM (Random Access Memory) that provides a work area when executing various programs.

[0019] The storage 13 includes, for example, a storage device such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive). The storage 13 only needs to have a storage area accessible by the information processing device 1, and for example, may be configured to have a dedicated storage area outside the information processing device 1. The information processing device 1 executes various types of information processing by the processor 11 using data read into the memory 12, for example, and stores the obtained processing results in the storage 13 as necessary. The communication I / F 14 includes various processing circuits, connection terminals and the like for communicating with other devices via the network 4, and is configured to conform to various communication standards.

[0020] User terminal 2 is an information terminal operated by the user. User terminal 2 may be a portable information terminal such as a smartphone, tablet, PDA (Personal Digital Assistant), or notebook computer, or it may be a wearable device such as smart glasses, VR goggles, AR glasses, or AR contact lenses. Alternatively, user terminal 2 may be a stationary information terminal such as a desktop computer. User terminal 2 may include only one type of the above-mentioned information terminals, or it may include multiple types.

[0021] External device 3 is a computer owned by a provider of an external service different from the services provided by system 5 (for example, car maintenance services, accommodation services, financial services, social networking services, internet shopping, etc.). Information processing device 1 and external device 3 can communicate with each other via network 4 and can send and receive data from each other. In the example in Figure 1, there is only one external device 3, but there may be multiple external devices 3. That is, system 5 may be able to communicate with multiple external devices 3, each corresponding to a different external service.

[0022] Figure 2 is a block diagram showing an example of the functional configuration of system 5 according to the embodiment of this disclosure. As shown in Figure 2, the information processing device 1 comprises a storage unit 30 and a control unit 50. The storage unit 30 is composed of, for example, a storage device such as the ROM, HDD, or SSD described above. The storage unit 30 stores, for example, a program 31, user information 32, and a trained model 38.

[0023] Program 31 is, for example, an information processing program for implementing various functions described later in the information processing device 1. In this embodiment, Program 31 is stored in a storage device such as ROM, which is a non-temporary computer-readable medium. Program 31 may also be stored in any non-temporary computer-readable medium such as an optical disc (CD, DVD, Blu-ray disc, etc.) or semiconductor memory (EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), flash memory, etc.) and provided to the information processing device 1.

[0024] User information 32 is a database that stores various types of information related to a user. User information 32 stores, for example, user attribute information 33, location and time information 34, external service information 35, behavior pattern information 36, and behavior-based information 37, all associated with information that uniquely identifies the user, such as a user ID. In other words, each user has the above types of information associated and stored in user information 32.

[0025] User attribute information 33 is information about the user's attributes. Here, "user attributes" refers to information that characterizes the user. There are no particular restrictions on what constitutes "user attributes," but examples include gender, age, occupation, family structure, address, car model, areas of interest, hobbies, etc.

[0026] Location and time information 34 is information that includes location information of user terminal 2 obtained from user terminal 2 and time information corresponding to said location information. Here, "location information" is information including, for example, the latitude, longitude, and height of user terminal 2, and "time information" is information indicating the time when the location information was measured. Location and time information 34 includes, for example, the history of location and time information of user terminal 2 and the current location information. Location and time information 34 may also include information indicating the movement speed of user terminal 2.

[0027] External service information 35 is information regarding the user's use of external services. Specifically, external service information 35 may include information such as the name and type of external service, the date and time of use, messages sent by the user using the external service, and purchased goods and services.

[0028] Behavioral pattern information 36 is information about the user's behavioral patterns. Behavioral pattern information 36 is generated, for example, by the first trained model 39 (described later) inferring based on the history of location and time information of the user terminal 2. Behavioral pattern information 36 includes, for example, information such as the places the user has visited, the time of day and frequency of such visits, and the routes the user frequently uses.

[0029] Behavior-based information 37 is information obtained based on the user's actions. Behavior-based information 37 can also be expressed as, for example, behavior-derived information, meaning information obtained from the user's actions; behavior-derived knowledge, meaning knowledge derived from the user's actions; or the user's personal know-how information. Behavior-based information 37 may include the user's experiential knowledge and tacit knowledge that the user themselves is not aware of.

[0030] Behavior-based information 37 is generated, for example, by the second trained model 40 described later, which makes inferences based on the user's user attribute information 33 and behavior pattern information 36. Behavior-based information 37 may also be generated based on information that the user has posted on social media, etc. Behavior-based information 37 may include, for example, information such as places the user frequently visits and routes the user often uses (for example, information such as favorite stores and favorite products / services, information on time-limited events such as time sales and limited-time sales, information on the time required for a route and information on stores and facilities along the route).

[0031] Furthermore, the behavior-based information 37 may include, for example, information obtained by inferring the purpose of the user's actions, the benefits that the actions bring, and / or what the user is doing when visiting a particular place. For example, if a user goes to a particular supermarket at a particular time, the second trained model 40 may infer that the supermarket is having a time sale at that time. This inferred information can then become the behavior-based information 37.

[0032] Here, we will describe the user information 32 in detail using Figures 3 and 4. Figures 3 and 4 are data tables showing an example of user information according to the embodiment of this disclosure. Note that the user information 32 is shown separately in Figures 3 and 4 due to space limitations in the figures. The data shown in Figures 3 and 4 may be managed in the same data table.

[0033] In the example in Figure 3, user information 32 stores user attribute information 33, location and time information 34, and external service information 35, each associated with a user ID, which is information that uniquely identifies the user. In the following description, users A, B, and C are the owners of user terminals 2A, 2B, and 2C, respectively. User attribute information 33 stores gender, age, address, make and model year of the car owned by the user, registration genre (genre of information the user is interested in), etc. The make and model year of the car owned may include past history (for example, the name of the car model previously owned). Similarly, other user attribute information 33 such as address may also include past history.

[0034] Location and time information 34 stores the current location information and location history, which is a history of location and time information. In the example in Figure 3, the current time is 20:00:10 on March 13, 2025, and the location information at this time is also stored in the location history. The location history is stored at predetermined intervals, but is not limited to these intervals. For example, in the example in Figure 3, the predetermined interval is 10 seconds, but it may be shorter (e.g., 1 second) or longer (e.g., 30 seconds). This predetermined interval may be the same as, for example, the interval at which location information is acquired from the user terminal 2. Also, the location history may be a subset of data where no change in location information is observed when comparing the preceding and succeeding data in a time series.

[0035] Figure 3 shows examples of external service information 35, including activity history on social media, mobile payment history, vehicle maintenance history, and used car purchase history. Social media activity history includes information on posting dates and content on social media platforms such as Facebook® and X (formerly Twitter®). Mobile payment history is information on the use of financial services, including, for example, the use of electronic money payments using user terminal 2. Vehicle maintenance history is information on the use of vehicle maintenance services such as vehicle inspections and repairs. Used car purchase history is information on used cars purchased by the user.

[0036] In the example in Figure 4, user information 32 stores behavior pattern information 36 and behavior-based information 37, associated with the user ID. Of the behavior pattern information 36, information A1, B1, and C1 are, for example, information indicating frequently used routes. Information A2, A3, B2, B3, and C2 are, for example, information indicating places the user regularly visits and the frequency of those visits. Information C3 is information indicating places the user visited in the most recent predetermined period and the number of visits. This information, exemplified as behavior pattern information 36, may be generated based solely on location and time information 34, or it may be generated based on inferences from other information such as external service information 35. For example, if user A made a mobile payment at restaurant G1, information indicating the dishes user A ate can be added to information A2 by using the mobile payment history as an inference.

[0037] In the example in Figure 4, the behavior-based information 37 is exemplified by the information generated from the travel route and each point included in the behavior pattern information 36 and the information stored in the user attribute information 33. For example, "Detailed information on Route E1" could include the time required for Route E1, the advantages of Route E1 when comparing it with an alternative route to Route E1, and detailed information about stores and facilities along Route E1. For example, "Detailed information on Restaurant G1" could include detailed information about the menu offered by Restaurant G1, lunch services, and menu items presumed to be User A's favorites. For example, "Detailed information on Yokohama City ◆◆" could include information related to User A's attribute information such as cars, family, leisure, and vehicle inspections from the information about Yokohama City ◆◆. For example, "Detailed information on Grocery Store H1" could include detailed information about the products offered by Grocery Store H1, sale times, sale days, and products presumed to be User B's favorites.

[0038] "Detailed information on repair shop J1" could include, for example, detailed information about the services and prices offered at repair shop J1, and the repair services that user B is presumed to frequently use. "Detailed information on used car dealership K1" could include, for example, detailed information about the types of cars sold at used car dealership K1, purchase benefits, and the type of car that user C is presumed to have purchased. In addition to the examples given, each of the above detailed information items may also include information about the purpose of visiting each location, the benefits of visiting, and the actions taken during the visit.

[0039] Returning to the explanation of Figure 2, the trained model 38 is created using machine learning techniques such as deep learning and clustering. The trained model 38 may be created using any method, such as supervised learning, unsupervised learning, semi-supervised learning, a combination thereof, or other conventionally known methods such as reinforcement learning. The trained model 38 may also be created by fine-tuning a general-purpose trained model (for example, a natural language processing model such as a large-scale language model).

[0040] In this embodiment, the trained model 38 is stored in the information processing device 1, but it may also be stored in another computer device. That is, the system 5 may be configured to allow the information processing device 1 to acquire data generated by the trained model 38 by linking the information processing device 1 with the other computer device storing the trained model 38 via an API (Application Programming Interface).

[0041] In this embodiment, the trained model 38 includes a first trained model 39, a second trained model 40, and a third trained model 41. The first trained model 39 may be composed of, for example, a single model or a combination of multiple models. The same applies to the second trained model 40 and the third trained model 41.

[0042] The first pre-trained model 39 is a pre-trained model that has been trained to output user behavior pattern information 36 based on the user's location and time history. The first pre-trained model 39 can be created, for example, by machine learning with the user's location and time history as explanatory variables and the user's behavior pattern as the target variable. Furthermore, it is preferable that the first pre-trained model 39 is trained to output user behavior pattern information 36 based on the user's location and time history and external service information 35, by further using information on the usage status of external services as an explanatory variable.

[0043] The second pre-trained model 40 is a pre-trained model that has been trained to output information 37 based on the user's user attribute information 33 and the user's behavior pattern information 36. The second pre-trained model 40 can be created, for example, by machine learning with information about the user's attributes and information about the user's behavior as explanatory variables and information based on the user's behavior as the dependent variable. Alternatively, the second pre-trained model 40 may also be trained to output information 37 based on the user's user attribute information 33, the user's behavior pattern information 36, and external service information 35, by further using information about the usage status of external services (for example, information posted by the user on social media, etc.) as explanatory variables.

[0044] The third pre-trained model 41 is a pre-trained model that, as information provided to the user, is trained to output information that is not identical to the information based on the actions of other users 37, but is related to the information based on the actions of other users 37 and the user's user information 32, based on information based on the actions of other users 37 and the user's user information 32. The third pre-trained model 41 can be created, for example, by machine learning with at least one of the following as explanatory variables: information based on the actions of other users, information about the user's attributes, history of the user's location and time information, the user's current location, history of use of external services, and information about behavioral patterns, and the above-mentioned provided information as the dependent variable. The provided information will be described in detail in a later paragraph.

[0045] The third pre-trained model 41 is preferably capable of predictive analysis processing on user information 32, for example. Specifically, the third pre-trained model 41 is preferably composed of a needs prediction model that predicts future user needs, a behavior prediction model that predicts the user's recent behavior patterns, a change point detection model that detects changes in the user's behavior patterns, and a trend analysis model that predicts changes in the user's behavior over a long period of time. Each of these models will be described in detail in a later paragraph.

[0046] The control unit 50 is composed of, for example, a processor 11 and RAM. When the program 31 is executed by the processor 11, the control unit 50 functions as, for example, a communication unit 51, a receiving unit 52, an acquisition unit 53, a selection unit 57, a generation unit 58, a detection unit 62, a specific unit 63, and a providing unit 64.

[0047] The communication unit 51 transmits and receives data, for example, between the user terminal 2 and the external device 3. The reception unit 52 receives various information inputs and selections from the user, for example, via the user terminal 2. The reception unit 52 can also receive the user's selection of a preferred candidate user from among at least one candidate user selected by the selection unit 57 described later, via the user terminal 2.

[0048] The acquisition unit 53 acquires various types of user information from the user terminal 2 and external devices 3. The various types of information acquired by the acquisition unit 53 are stored in the user information 32. In this embodiment, the acquisition unit 53 includes a first acquisition unit 54, a second acquisition unit 55, and a third acquisition unit 56.

[0049] The first acquisition unit 54 acquires information about the user's attributes. The first acquisition unit 54 receives input of information about the user's attributes from the user, for example, via the user terminal 2. Alternatively, the first acquisition unit 54 may acquire information about the user's attributes from an external device 3, for example.

[0050] The second acquisition unit 55 acquires location information of the user terminal 2 and time information corresponding to said location information. This information is acquired, for example, from the user terminal 2. The third acquisition unit 56 acquires information regarding the user's use of external services. This information may be acquired, for example, from the external device 3 or from the user terminal 2.

[0051] The selection unit 57 selects at least one candidate user for a given user from among multiple users based on the user information 32. Candidate users are, for example, matching candidates when matching users based on the degree of commonality in the user information 32. The selection unit 57 refers to at least one of the following: user attribute information 33, location and time information 34, external service information 35, behavior pattern information 36, and other user information 32, calculates the degree of commonality among each user, and selects users with a high degree of commonality as candidate users.

[0052] The selection unit 57 preferably refers to, for example, the current location information of at least the user terminal 2 when selecting matching candidates. That is, it is preferable that at least one candidate user selected as a matching candidate for the user changes in accordance with changes in the current location information of the user terminal 2. As an example, the selection unit 57 is configured such that the closer the current location information of users is, the more likely those users are to be selected as matching candidates for each other.

[0053] Here, the information provided to the user by the provisioning unit 64 is information based on the candidate user's behavior information 37. That is, the information provided is generated based on the candidate user's behavior information 37. The candidate users selected by the selection unit 57 may be presented to the user via the user terminal 2, and the reception unit 52 may accept the user's selection of a candidate user. In this case, the information provided is information based on the user's behavior information information 37 of the candidate user selected by the user.

[0054] Furthermore, it is preferable that the selection unit 57, for example, when a user selects a candidate user, selects other matching candidates different from the selected user, and that the candidate user selected by the user is more likely to be selected than before the user selected them. In other words, it is preferable that the selection unit 57 is configured such that the more a user has been selected as a candidate user (number of selections), the more likely that user will be selected as a candidate user. It is also preferable that the selection unit 57 is configured such that the higher the confidence level (described later) attached to the information 37 based on the user's actions, the more likely that the user will be selected as a candidate user.

[0055] The generation unit 58 generates various types of information. In this embodiment, the generation unit 58 includes a first generation unit 59, a second generation unit 60, and a third generation unit 61.

[0056] The first generation unit 59 generates behavioral pattern information 36 using the first trained model 39. The first generation unit 59 inputs, for example, the user's location information and time information history, and preferably external service information 35, to the first trained model 39, causing the first trained model 39 to generate the user's behavioral pattern information 36.

[0057] The second generation unit 60 generates information 37 based on the user's behavior using the second trained model 40. For example, the second generation unit 60 inputs the user's user attribute information 33 and behavior pattern information 36 to the second trained model 40, causing the second trained model 40 to generate information 37 based on the user's behavior. The second generation unit 60 may also use external service information 35 as input data to the second trained model 40. Specifically, the second generation unit 60 may generate or modify information 37 based on the user's behavior based on information posted by the user on social media, etc.

[0058] Here, information disseminated on social media, etc., may have low reliability. Therefore, if behavior-based information 37 is generated based on information disseminated by users on social media, etc., the reliability of the generated information may be low. For this reason, it is preferable that the second generation unit 60 has a function to verify the reliability (credibility) of the information generated by the second trained model 40. By verifying the reliability, the second generation unit 60 can, for example, exclude information that may have low reliability without storing it as behavior-based information 37, thereby ensuring the reliability of the behavior-based information 37.

[0059] The second generation unit 60 verifies the reliability of the user behavior-based information 37, for example, by comparing it with web search results or user information 32 from other users. The reliability verification can be performed using conventionally known methods, for example, by evaluating the degree of agreement between the information generated by the second trained model 40 and the information obtained through web search, etc.

[0060] Furthermore, the second generation unit 60 may perform reliability verification using a trained model whose reliability can be verified. Conventional known methods can be used to create such a trained model; for example, it can be created by supervised learning using training data with reliability labels. Such a trained model may be configured to be included in the second trained model 40, or it may be configured independently of the second trained model 40.

[0061] The second generation unit 60 preferably assigns a confidence score to the information generated by the second trained model 40, for example, according to the degree of agreement. The behavior-based information 37 is stored in association with the information generated by the second trained model 40 and its confidence score. The confidence score can be used, for example, in the selection of candidate users by the selection unit 57 or in determining the priority of the information provided, as described later.

[0062] Furthermore, reliability verification may be performed not on the information generated by the second trained model 40, but on the information that serves as input data to the second trained model 40. In particular, it is preferable to perform reliability verification on external service information 35, such as information posted on social networking services (SNS). On the other hand, if the input data to the second trained model 40 consists only of user attribute information 33 and behavioral pattern information 36, the system may be configured not to perform reliability verification on this input data and the generated data. This is because user attribute information 33 and behavioral pattern information 36 concern objective facts such as the user's attributes and location history, and are therefore considered to be highly reliable information.

[0063] Here, the detection unit 62 and the identification unit 63 will be described before the third generation unit 61. The detection unit 62 detects changes in the user's behavior patterns based on the history of the user's location information and time information. The detection unit 62 detects changes in the behavior pattern, for example, using a change point detection model included in the third trained model 41. Alternatively, the detection unit 62 may detect changes without using a change point detection model, for example, by analyzing the history of the user's location information and time information based on statistical methods or periodicity analysis methods.

[0064] The identification unit 63 identifies the validity period of the action-based information 37 based on the content of the action-based information 37. Here, "validity period of information" refers to the period during which the information is useful, from the time when the information switches from an invalid state where it is not useful as information to an effective state where it is useful, to the time when the information switches back to an invalid state where it loses its usefulness. For this reason, it can be said that the information switches between an effective state and an invalid state depending on the period (validity period). Provided information generated based on information that is not in an effective state may not be considered highly useful. For this reason, it is preferable to identify the validity period of the action-based information 37 and provide information based on the action-based information 37 within its validity period.

[0065] The identification unit 63, for example, identifies the validity period based on information that can identify the validity period within the behavior-based information 37. Specifically, if the behavior-based information 37 is information about a time sale, the period of that time sale can be identified as the validity period. In other words, if the behavior-based information 37 contains specific information about the date and time or period that can identify the validity period of that information, the validity period can be identified based on that specific information.

[0066] If there is no information in the behavior-based information 37 that can identify the validity period, the identification unit 63 performs a process to infer the validity period by further analyzing the behavior-based information 37 in detail. Specifically, if the behavior-based information 37 is information about a product, the validity period of the behavior-based information 37 may be identified based on information about the product's lifespan, such as the manufacturing date and expiration date, or information about the product's life cycle, such as the release date and release period. The identification unit 63 may identify the validity period using, for example, a trained model capable of inferring the validity period of information. Conventional known methods can be used to create such a trained model, for example, by supervised learning using training data labeled with validity periods. Such a trained model may be configured to be included in the second trained model 40, or it may be configured independently of the second trained model 40.

[0067] The third generation unit 61 generates the provided information using the third trained model 41. For example, the third generation unit 61 inputs information 37 based on the actions of a candidate user and user information 32 of the user to the third trained model 41, and causes the third trained model 41 to generate provided information that is not identical to the information 37 based on the actions of the candidate user, but is related to the information 37 based on the actions of the candidate user and user information 32 of the user. Preferably, the third generation unit 61 generates the provided information using information 37 based on the actions of a candidate user and user information 32 of the user as input data.

[0068] Preferably, the third generation unit 61 can perform predictive analysis of the user's future trends based on user information 32, for example, using the third trained model 41. The third trained model 41 generates information to be provided, for example, based on information 37 based on the behavior of other users and the results of predictive analysis of the user's user information 32. By generating information to be provided based on the results of predictive analysis of future trends, for example, it becomes possible to provide information that matches the user's potential needs.

[0069] Preferably, the third generation unit 61 has a function to set priorities for the information provided by the third trained model 41. The third generation unit 61 may set priorities based on, for example, the validity period, reliability, prediction results such as the user's recent needs and behavior predictions, and the user's current location of the behavior-based information 37 on which the information is based. For example, the priority may be set to be higher the shorter the time remaining until the validity period of the behavior-based information 37 ends, the higher the reliability of the behavior-based information 37, the more it relates to the user A's recent prediction results, and the closer the behavior-based information 37 is to a specific location and the user's current location is to that specific location.

[0070] The third generation unit 61 may set priorities using, for example, a trained model capable of inferring the priority of information. Conventional known methods can be used to create such a trained model, for example, by supervised learning using training data that has been labeled with priority. Furthermore, the explanatory variables and input data of such a trained model may include, for example, environmental data such as season and weather. Moreover, such a trained model may be configured to be included in the third trained model 41, or it may be configured independently of the third trained model 41.

[0071] Preferably, the third generation unit 61 has a function to determine the timing for providing the information generated by the third trained model 41 to the user. The third generation unit 61 may determine the timing for providing the information based on at least one of the following: the validity period of the information 37 based on the actions on which the information is based, its reliability, prediction results such as predictions of the user's future needs and actions, the priority set for the information, and the user's current location.

[0072] The third generation unit 61 may determine the timing of delivery using, for example, a trained model capable of inferring the timing of delivery. Conventional known methods can be used to create such a trained model, for example, by supervised learning using training data labeled with delivery timing. Furthermore, the explanatory variables and input data of such a trained model may include, for example, environmental data such as season and weather, and information priorities. Moreover, such a trained model may be configured to be included in the third trained model 41, or it may be configured independently of the third trained model 41.

[0073] The provisioning unit 64 provides the provisioning information generated by the third generation unit 61 based on the user's user attribute information 33 and information based on the actions of other users 37 to the user terminal 2. The provisioning unit 64 transmits the provisioning information to the user terminal 2 at a provisioning timing determined by the third generation unit 61, for example.

[0074] Furthermore, the providing unit 64 may, for example, provide information 37 based on the actions of other users or information 37 based on the user's actions, generated by the second generation unit 60, as information to be provided to the user. Alternatively, the providing unit 64 may provide information generated by the third generation unit 61 based on the user's user attribute information 33 and the information 37 based on the user's actions, which is not identical to the information 37 based on the user's actions, but is related to the information 37 based on the user's actions and the user's user information 32, as information to be provided to the user.

[0075] The user terminal 2 comprises a storage unit 70, a sensor 72, and a control unit 73. The storage unit 70 is composed of, for example, a storage device such as a ROM, HDD, or SSD included in the user terminal 2. The storage unit 70 stores a program 71. The program 71 is, for example, an information processing program for implementing various functions described later in the user terminal 2.

[0076] Sensor 72 includes a sensor that measures the location information of the user terminal 2. A sensor that measures location information is, for example, a sensor that receives positioning signals transmitted from GPS (Global Positioning System) or other GNSS (Global Navigation Satellite System) positioning satellites.

[0077] The control unit 73 is composed of, for example, a processor and RAM included in the user terminal 2. When program 71 is executed by the processor, the control unit 73 functions as, for example, a communication unit 74, a reception unit 75, a location information acquisition unit 76, and an output unit 77.

[0078] The communication unit 74 transmits and receives data to and from the information processing device 1, for example. The reception unit 75 accepts information input and various selections from the user via a predetermined input interface such as a keyboard or touchscreen, for example. The location information acquisition unit 76 acquires the location information of the user terminal 2 based on the positioning signal received by the sensor 72. The location information acquisition unit 76 also stores the location information in the storage unit 70, associating it with time information indicating the time corresponding to the location information. The output unit 77 outputs various information received from the information processing device 1 to an output device such as the display (not shown) of the user terminal 2, for example.

[0079] Furthermore, various types of information received by the reception unit 75, as well as location and time information acquired by the location information acquisition unit 76, are transmitted to the information processing device 1 via the communication unit 74. In addition, data related to behavioral logs, such as the usage history of external service applications used on the user terminal 2, may also be transmitted to the information processing device 1 via the communication unit 74.

[0080] System 5 may be implemented by including a native application downloaded to the user terminal 2, by including a web application, or by being implemented as a cloud-based service such as SaaS.

[0081] The external device 3 comprises a storage unit 80 and a control unit 83. The storage unit 80 is composed of, for example, a storage device such as a ROM, HDD, or SSD included in the external device 3. The storage unit 80 stores external service information 82, which is information regarding the user's use of external services. The control unit 83 is composed of, for example, a processor or RAM included in the external device 3. The control unit 83 transmits the information contained in the external service information 82 to the information processing device 1 in response to receiving a request from the information processing device 1 to refer to the external service information 82. The information processing device 1 stores the information contained in the external service information 82 received from the external device 3 as external service information 35. That is, the external service information 35 may be at least a part of the external service information 82.

[0082] Figure 5 is a conceptual diagram showing an example of a service provided by System 5 according to the embodiment of this disclosure. An example of providing information to User A will be described below using Figure 5. First, the information processing device 1 appropriately acquires the various types of information described above as user information 32 from user terminals 2A, 2B, and 2C. Note that the information processing device 1 generates the behavioral pattern information within the user information 32 based on the history of location information and other data acquired from user terminal 2.

[0083] Furthermore, if users B and C are using external services provided by repair shops, travel agencies, etc., the information processing device 1 appropriately obtains information regarding the usage status of these users' external services from the external device 3 and stores the obtained information as user information 32. The same applies when user A is using external services. In addition, using the method described above, the information processing device 1 generates behavior-based information 37 (personal know-how information in Figure 5) for each user based on at least some of the information in the user information 32.

[0084] When selecting candidate users to match with User A, for example, one or more candidate users are selected based on user information 32 as shown in Figure 5. Let's assume that Users B and C are selected as candidate users for User A. In this case, information provided to User A is generated based on the behavioral information 37 of Users B and C, and User A's user information 32. This generated information can be described as personalized information for User A, such as tacit knowledge for Users B and C. Such information is considered useful information that may provide User A with new insights.

[0085] Information provided to User A could include, for example, information about User A's tasks and information about solutions to those tasks. As an example, consider a case where, as shown in Figures 3 and 4, information 37 based on User B's actions includes information that User B frequently uses a specific repair shop J1, and User A's user information 32 identifies that User B's vehicle inspection deadline is approaching, or that vehicle inspection is included in the registered categories of User A's user information 32. In this case, information about User A's tasks could include, for example, the implementation of a vehicle inspection or the vehicle inspection deadline. Information about solutions to the tasks could include information about the vehicle inspection fees and services of repair shop J1, and the URL of the vehicle inspection reservation site.

[0086] Furthermore, in this example, the information provided to User A may not be information about repair shop J1 itself, but rather information related to repair shop J1. For example, information related to repair shop J1 in terms of information about cars (e.g., used car appraisal services, new car information, special offers when purchasing a new car, etc.) may be provided to User A. Also, if the information 37 based on User C's actions includes information that User C used used car dealer K1 and purchased a specific car model, then the information provided to User A may include information about the car models and prices sold at used car dealer K1, and special offers when purchasing that specific car model.

[0087] Furthermore, consider the case where, as shown in the examples in Figures 3 and 4, the information 37 based on user B's behavior includes information that user B frequently uses a specific grocery store H1. Also, assume that the address shown in user A's user information 32 is near grocery store H1, and that the current time is 7 PM on a weekday. In this case, for example, information regarding user A's task could be shopping. Information regarding the solution to the task could include the products offered by grocery store H1, or the time of any sales. If the current time has passed the time of any sales, information from other stores could become the information regarding the solution.

[0088] Furthermore, in this example, the information provided to user A may not be information about grocery store H1 itself, but rather information related to grocery store H1. For example, if grocery store H1 is having a limited-time sale on specialty products from Gunma Prefecture, the information provided to user A may include information about a limited-time sale on travel to Gunma Prefecture, as information related to grocery store H1. In this example, user A may be unaware that they want to go to Gunma Prefecture because of a sale on specialty products from Gunma Prefecture in their neighborhood. Therefore, generating information as in this example can make it possible to bring to light needs that user A may not be aware of. Also, from the perspective of broadening the variety of information provided, it is preferable that the information provided is not the information 37 based on the actions of other users such as user B itself, but rather information related to the information 37 based on the actions of other users, as in this example.

[0089] After the information is provided to user A, if the information processing device 1 detects that user A is considering using an external service included in the information, the information processing device 1 notifies the external device 3, which corresponds to that external service, that there is a potential customer. With this configuration, the external device 3 can prepare to provide services or products to user A, for example, by registering virtual inventory. The method for detecting that user A is considering the service is not particularly limited, and conventionally known methods such as analyzing the frequency of access to the information, the user's behavior on the information display page, and the time spent on the page can be used.

[0090] (Information processing methods) Next, we will describe the information processing method performed by System 5. Figure 6 is a flowchart showing an example of the information processing method according to the embodiment of this disclosure. The order of the various processes that constitute the various flows described below may be in any order and may be executed in parallel, as long as no inconsistencies occur in the processing content.

[0091] In step S1, the information processing device 1 acquires various information to be stored as user information 32. In step S2, the information processing device 1 performs a first generation process. In step S3, the information processing device 1 performs a second generation process.

[0092] Here, the processes of steps S1 and S2 will be described in detail using Figure 7. Figure 7 is a flowchart of an example of the acquisition process and the first generation process in Figure 6. In step S21A, user terminal 2A, for example, executes program 71 to log in to system 5, and then receives input from user A regarding information about user A's attributes. The information received in step S21A is transmitted from user terminal 2 to information processing device 1. In step S22, information processing device 1 acquires the information transmitted from user terminal 2A in step S21A and stores it as user attribute information 33.

[0093] In step S23A, the user terminal 2A acquires current location and time information and transmits this information to the information processing device 1. Step S23A is executed, for example, at predetermined intervals (e.g., every second, every 10 seconds, etc.) while logged into system 5. The processing in step S23A may be executed by background processing on the user terminal 2A. In step S24, the information processing device 1 acquires the information transmitted from the user terminal 2A in step S23A and stores it as location and time information 34.

[0094] In step S25A, the user terminal 2A transmits information regarding the usage status of external services provided by the external device 3 to the information processing device 1. Step S25A is executed, for example, in response to the use of external services by the user terminal 2A. In step S26, the information processing device 1 acquires the information transmitted from the user terminal 2A in step S25A and stores it as external service information 35. Note that the external service information 35 may also be acquired from the external device 3 through API linkage with the external device 3, etc.

[0095] In step S27A, user terminal 2A transmits activity log data to information processing device 1. Step S27A may be executed at predetermined intervals, for example, while logged into system 5, or at predetermined times such as when logging into system 5 or when user A performs a predetermined operation. In step S28, information processing device 1 acquires the activity log data transmitted from user terminal 2A in step S27A and stores it in user information 32.

[0096] Steps S21B, S23B, S25B, and S27B are the same processes as steps S21A, S23A, S25A, and S27A, respectively, except that they are executed on the user terminal 2B and the timing of their execution. Therefore, the explanation of these processes with respect to the user terminal 2B is omitted.

[0097] In step S29, the information processing device 1 takes the history of user A's location information and time information as input data to the first trained model 39. In step S29, it is preferable to also take user A's external service information 35 as input data to the first trained model 39. Based on this input data, the first trained model 39 generates information about user A's behavior patterns. In step S30, the information processing device 1 acquires the information generated by the first trained model 39 and stores it as behavior pattern information 36.

[0098] For user B, information to be stored as behavioral pattern information 36 is generated in the same manner as for user A. Although not shown in Figure 7, for other users such as user C, information to be stored as behavioral pattern information 36 is also generated in the same manner as for user A.

[0099] Here, we will describe the process of step S3 in detail using Figure 8. Figure 8 is a flowchart of an example of the second generation process in Figure 6. In step S41, the information processing device 1 uses user attribute information 33 of user A and user A's behavior pattern information 36 as input data to the second trained model 40. In step S41, external service information 35, such as information transmitted by user A on social networking services, etc., may also be used as input data to the second trained model 40. Based on this input data, the second trained model 40 generates information 37 based on user A's behavior. In step S42, the information processing device 1 acquires the information generated by the second trained model 40.

[0100] In step S42, the information processing device 1 verifies the reliability of the information 37 based on user A's actions, for example, by performing a web search or comparing it with user information 32 of other users. In step S42, it is preferable to add a reliability score to the information 37 based on actions.

[0101] Furthermore, the reliability verification may be performed on the external service information 35 input to the second trained model 40, rather than on the information generated by the second trained model 40. Also, if the input data to the second trained model 40 consists only of user attribute information 33 and behavior pattern information 36 of user A, step S42 may be omitted.

[0102] In step S44, the information processing device 1 determines the validity period of the action-based information 37. In step S44, for example, if there is information in the action-based information 37 that can specify the validity period, the validity period is specified based on that information. If there is no information in the action-based information 37 that can specify the validity period, for example, the action-based information 37 is further analyzed to infer the validity period. This process may use the second trained model 40 or other trained models.

[0103] Similarly, for User B, the generation, verification, and determination of the validity period of information 37 based on User B's actions are carried out in the same manner as for User A. The same applies to other users such as User C.

[0104] Returning to the explanation of Figure 6, in step S4, the information processing device 1 updates the user information 32 each time it acquires new user information 32 in step S1. Since the user information 32 includes the current location information of each user, the user information 32 may be updated, for example, when each user moves.

[0105] In step S5, the information processing device 1 constructs a correlation between users based on the degree of commonality of the information contained in the user information 32. This correlation may be updated, for example, when the user information 32 is updated. Furthermore, this correlation may be updated at predetermined timings, for example, when it is detected that user A has performed a search for a predetermined external service based on the activity log data of user terminal 2A, or when it is detected that appointment information has been registered in user A's calendar.

[0106] In step S6, the information processing device 1 selects one or more candidate users based on the correlation between users. In step S6, for example, it is preferable that the higher the reliability of the behavior-based information 37, the more likely that user is to be selected as a candidate user.

[0107] Information on selected candidate users is transmitted from the information processing device 1 to the user terminal 2 and displayed as a list on the user terminal 2A. Preferably, at least a portion of the candidate users' user information 32 is also displayed on the user terminal 2A as reference information for user A to select a preferred candidate user from one or more candidates. However, for the protection of personal information, names and addresses are not displayed.

[0108] In step S7, user terminal 2A accepts the selection of a candidate user from user A. The selection result by user A is transmitted from user terminal 2 to information processing device 1. Note that a candidate user selected by user A will have an increased number of selections as a candidate user, making them more likely to be selected as a candidate user by other users other than user A.

[0109] Furthermore, since the candidate users selected by User A will have their own behavior-based information 37 used to generate information provided to User A, it is likely that some users will be reluctant to have their information used by others from a privacy perspective. Therefore, candidate users selected by User A may be given the option to choose whether or not to allow the use of their own information. Alternatively, each user may be presented with their own behavior-based information 37 and be able to pre-set whether or not other users can use each piece of information included in their behavior-based information 37.

[0110] In the following explanation, we will assume that user B was selected in step S7. Note that in step S7, it is also possible to select other users, such as user C. Alternatively, step S7 may be omitted, and the information 37 based on the actions of all candidate users selected in step S6 may be used in the third generation process described later.

[0111] In step S8, the information processing device 1 performs a third generation process. The process in step S8 will be described in detail using Figure 9. Figure 9 is a flowchart showing an example of the third generation process in Figure 6.

[0112] Steps S51 to S56 are preprocessing steps for the data to be input to the third trained model 41. Note that such preprocessing may also be performed in steps S2 and S3. In step S51, the information processing device 1 performs, for example, normalization of the location information acquired in step S24. Also in step S51, the location information is analyzed to obtain information such as user A's travel path, place of stay, and travel speed.

[0113] In step S52, the information processing device 1 analyzes the behavior log data acquired in step S28, for example, to obtain information about behavior log patterns. Specifically, it obtains information such as the usage status of system 5 and other applications, user A's search history, and website browsing patterns.

[0114] In step S53, the information processing device 1 analyzes the external service information 35 acquired in step S26, for example, to obtain information regarding the usage patterns of external services. Specifically, it obtains information such as the type and timing of the external service used, the frequency of use, and the evaluation of the external service by user A.

[0115] In step S54, the information processing device 1 performs preprocessing on the user attribute information 33 acquired in step S22, such as data cleansing (detection and imputation of missing values, removal of noise data, etc.). Data cleansing may also be performed as appropriate in steps S51 to S53.

[0116] In step S55, the information processing device 1 integrates the data obtained in steps S51 to S54. In step S56, the information processing device 1 performs feature engineering. In step S56, for example, time-series features are generated (e.g., the average number of actions over the last N days, the frequency of visits to a specific location, the number of times external services are used, the variability of visit times and usage time periods, etc.). These time-series features are used in the predictive analysis in steps S61 to S65, which will be described later.

[0117] Steps S61 to S65 are predictive analysis processes of user information 32 using the third trained model 41. The needs prediction model, behavior prediction model, change point detection model, and trend analysis model described below may be, for example, independent trained models included in the third trained model 41, or they may be submodels that constitute the third trained model 41. Furthermore, each of these models can be created by machine learning, for example, using information included in the information obtained in step S56 as explanatory variables and information output from each model as the target variable. The explanatory variables and input data of each model may include, for example, environmental data such as season and weather.

[0118] In step S61, the information processing device 1 uses a needs prediction model to predict the future needs of user A. Specifically, the information processing device 1 inputs the information obtained in step S56 into the needs prediction model and obtains information from the needs prediction model regarding products and services that user A will be in demand in the future, and information regarding the estimation of user A's potential interests.

[0119] In step S62, the information processing device 1 uses a behavior prediction model to analyze the periodicity of the user's behavior and predict the user's recent behavior. Specifically, the information processing device 1 inputs the information obtained in step S56 into the behavior prediction model and obtains a prediction of user A's recent behavior and / or information about user A's specific behavior from the behavior prediction model.

[0120] In step S63, the information processing device 1 uses a change point detection model to detect changes in the user's behavior patterns and estimate the points of change to a new lifestyle. Specifically, the information processing device 1 inputs the information obtained in step S56 into the change point detection model and obtains information from the change point detection model regarding the points of change in user A's behavior and the lifestyle after the change.

[0121] In step S64, the information processing device 1 uses a trend analysis model to evaluate the influence of seasonality on user behavior and predict changes in user behavior over long-term periods such as years. Specifically, the information processing device 1 inputs the information obtained in step S56 into a change point detection model and obtains information regarding changes in user A's behavior over long-term periods from the trend analysis model.

[0122] In step S65, the information processing device 1 integrates the various types of information (prediction results) acquired in steps S61 to S64. In step S65, for example, the various types of information acquired in steps S61 to S64 are weighted based on the results of the analysis of user A's search history and location information, and an integration process is performed.

[0123] Steps S66 to S68 involve the generation of information provided by the third trained model 41. In step S66, the information processing device 1 uses the prediction results integrated in step S65 and the information 37 based on user B's actions as input data to the third trained model 41. Based on this input data, the third trained model 40 generates information to be provided to user A. The information processing device 1 then retrieves the information generated by the third trained model 40.

[0124] The information provided in step S66 is preferably information related to information within the validity period or information with a high degree of reliability from among the information 37 based on user B's actions. Furthermore, the information provided in step S66 is preferably information related to information extracted from among the information 37 based on user B's actions based on various prediction results of user A integrated in step S65. In step S66, for example, multiple pieces of information are provided.

[0125] In step S67, the information processing device 1 sets a priority for the information provided in step S66. In step S67, the information processing device 1 sets a priority based on, for example, the validity period and reliability of the information 37 based on the actions on which the information provided is based, prediction results such as the recent needs and actions of user A, and the user's current location.

[0126] In step S68, the information processing device 1 determines the timing for providing each piece of information generated in step S66 to user A. In step S68, the information processing device 1 determines the timing for providing information based on at least one of the following: the validity period and reliability of the information 37 based on the actions on which the information is based; prediction results such as predictions of user A's future needs and actions; the priority set in step S67; and the user's current location.

[0127] Returning to the explanation of Figure 6, in step S9, the information processing device 1 transmits the information to the user terminal 2A. Step S9 is performed, for example, at the timing determined in step S68.

[0128] In step S10, user terminal 2A receives the information transmitted from the information processing device in step S9, outputs the information to an output device such as a display, and presents it to user A. With this, the process of providing information to user A is completed.

[0129] Figure 10 shows an example of the information provided. In Figure 10, "Vehicle Inspection" in column 32A, "Company Z" in column 32B, "Driving" in column 32C, "Family of 4" in column 32D, and "Three-Day Weekend Next Week" in column 32E are, for example, pieces of information included in user information 32, such as the user's registered genre and family structure, which are selected or entered by the user. The columns directly below these columns 32A to 32E are related words that have been identified as terms related to the respective information described in columns 32A to 32E. This identification is performed, for example, by the third trained model 41.

[0130] Next, the third pre-trained model 41 extracts common elements shared by related words, for example. In the example in Figure 10, the common elements are "leisure" (driving, three-day weekend next week), "travel" (driving, three-day weekend next week), and "maintenance" (vehicle inspection, Company Z). The third pre-trained model 41 extracts multiple pieces of information related to these common elements from the multiple pieces of information contained in the candidate user's behavior-based information 37, for example, and combines the extracted pieces of information to generate new information. In Figure 10, five pieces of information extracted from the behavior-based information 37—"Conduct a vehicle safety check before the long weekend," "Provide information on tourist spots for families," "Plan for using nearby facilities while waiting for maintenance," "Meal coupon for a family of four," and "Joint project between Company Z dealer and tourist facilities"—are combined to generate new information: "Day trip drive package with vehicle maintenance before Golden Week." This new information is an example of information provided to the user. In other words, it shows that derived information, which is new information derived from the related information extracted from the behavior-based information 37, is generated through a combination process.

[0131] The generation process of derived information, which is new information in the third trained model 41, may be performed by detecting a change in the behavior pattern using the change point detection model included in the third trained model 41, or after detection. As a result, derived information based on the behavior pattern is generated before and after the change point indicating a change in the user's behavior pattern, resulting in the generation of more effective and useful information.

[0132] Although one embodiment of the present invention has been described in detail above, the present invention is not limited to the above-described embodiment and can be modified, improved, etc. as appropriate. The present invention includes all modifications within the meaning and scope of the claims and equivalents thereof.

[0133] [Note] The embodiments of the present disclosure described above include at least the following:

[0134] (Item 1) An information processing apparatus comprising at least one processor, a storage unit for storing information processing programs, and a communication unit capable of communicating with user terminals operated by each user, The processor executes the information processing program, A first acquisition process for acquiring user attribute information relating to the attributes of the aforementioned user, A second acquisition process that acquires the location information of the user terminal and the time information corresponding to the location information, A first generation process generates behavioral pattern information relating to the user's behavioral patterns using a first trained model that has been trained to output patterns of the user's behavior based on the user's location information and time information history. A memory process that associates the user attribute information and the behavioral pattern information for each user and stores them as user information, A selection process for selecting at least one candidate user for the first user based on the user information, A second generation process generates information based on the actions of a candidate user using a second trained model that has been trained to output information based on the user's actions based on the user attribute information and behavior pattern information of the user, A provision process that provides predetermined information based on the information derived from the aforementioned actions to the user terminal of the first user, An information processing device that performs this task. This improves the usefulness of the information provided to users. Specifically, it allows for the provision of information to users based on information that may include the tacit knowledge of other users, making it possible to provide information that gives users some kind of insight.

[0135] (Item 2) The predetermined information provided is information based on the candidate user's actions and the first user's user information. The information processing device described in item 1. This allows, for example, to make the information provided to users more tailored to their needs. Furthermore, when predetermined information is generated based on the behavior of multiple candidate users, the variety of information provided to users can be broadened, and as a result, the usefulness of the information provided to users can be further improved.

[0136] (Item 3) Furthermore, a third generation process is executed to generate the predetermined provided information using a third trained model that has been trained to output information that is not identical to the information based on the actions of the candidate user and the user information of the first user, but is related to the information based on the actions and the user information of the first user, based on the information based on the actions of the candidate user and the user information of the first user. An information processing device as described in item 1 or 2. This allows us to provide users with information that is more relevant to them, even if it may include tacit knowledge from other users. Specifically, by providing users with information that is not based on the actions of other users themselves, but is related to that action and also related to the user's own user information, we can provide users with information that is more likely to be of interest to them.

[0137] (Item 4) Furthermore, based on the user's location information and time information history, a detection process is performed to detect changes in the user's behavior patterns. The third generation process is a process that generates the predetermined information to be provided based on information based on the actions of the candidate user, the user information of the first user, and information regarding changes in the patterns of the actions. The information processing device described in item 3. When user behavior patterns change, for example, it may indicate that user needs have changed, either implicitly or explicitly. According to the above configuration, it is possible to predict such implicit or explicit user needs, and as a result, further improve the usefulness of the information provided to users.

[0138] (Item 5) Furthermore, based on the content of the information based on the aforementioned action, a specific process is performed to determine the validity period of the information based on the aforementioned action. The aforementioned predetermined information is information based on the aforementioned actions within the validity period. An information processing device as described in any one of items 1 through 4. This reduces the risk of providing users with unnecessary information, such as information about services that have already ended. As a result, the usefulness of the information provided to users can be further improved. In addition, it becomes possible to provide information to users at an appropriate time.

[0139] (Item 6) Furthermore, a selection acceptance process is executed to accept the selection of a candidate user desired by the first user from among the at least one candidate user selected by the selection process. The aforementioned predetermined information provided is information based on the actions of candidate users selected by the first user. An information processing device as described in any one of items 1 through 5. This allows users to select potential users who are likely to have the information they want, thereby further improving the usefulness of the information provided to them.

[0140] (Item 7) In the selection process described above, when selecting candidate users other than the first user, the candidate users selected by the first user are more likely to be selected than before they were selected by the first user. The information processing device described in item 6. Candidate users who are supported by other users are likely to possess information of high utility or reliability. By making such candidates more likely to be selected as matching candidates, it is possible to further improve the utility of the information provided to multiple users. On the other hand, candidate users who are not supported by other users may possess information of low utility or reliability. According to the above structure, for example, such candidate users will be relatively less likely to be selected as matching candidates, and as a result, it is possible to further improve the utility of the information provided to multiple users.

[0141] (Item 8) The user information includes the current location information of the user terminal, Depending on the current location information of the user terminal, at least one candidate user selected by the selection process may be changed. An information processing device as described in any one of items 1 through 7. For example, the usefulness of the information may vary significantly depending on the user's current location. With the above configuration, the candidate users to be matched can be changed according to the user's current location, making it easier to provide the user with more useful information, such as information about places close to the user's current location that can be easily reached.

[0142] (Item 9) Furthermore, a third acquisition process is executed to obtain external service information regarding the user's usage of external services. The first generation process is a process that generates the behavior pattern information based on the history of the location information and the time information, and the external service information. An information processing device as described in any one of items 1 through 8. This allows for, for example, more detailed predictions of user behavior patterns. As a result, it becomes possible to further improve the usefulness of the information provided to users.

[0143] (Item 10) An information processing method performed by an information processing device equipped with a communication unit capable of communicating with user terminals operated by each user, A first acquisition step involves obtaining user attribute information relating to the attributes of the aforementioned user, A second acquisition step involves acquiring the location information of the user terminal and the time information corresponding to the location information. A first generation step of generating behavioral pattern information relating to the user's behavioral patterns using a first trained model that has been trained to output patterns of the user's behavior based on the user's location information and time information history, A storage step in which, for each user, the user attribute information and the behavior pattern information are associated and stored as user information, A selection step in which at least one candidate user is selected for the first user based on the user information, A second generation step of generating information based on the user's behavior using a second trained model that has been trained to output information based on the user's behavior based on the user attribute information and behavior pattern information of the user, A provision step of providing predetermined information based on the information obtained from the aforementioned actions to the user terminal of the first user, Information processing methods, including those mentioned above. This improves the usefulness of the information provided to users. Specifically, it allows for the provision of information to users based on information that may include the tacit knowledge of other users, making it possible to provide information that gives users some kind of insight.

[0144] (Item 11) An information processing program executed in an information processing device equipped with a communication unit capable of communicating with user terminals operated by each user, When the information processing program is executed by the processor of the information processing device, the information processing device will: A first acquisition process for acquiring user attribute information relating to the attributes of the aforementioned user, A second acquisition process that acquires the location information of the user terminal and the time information corresponding to the location information, A first generation process generates behavioral pattern information relating to the user's behavioral patterns using a first trained model that has been trained to output patterns of the user's behavior based on the user's location information and time information history. A memory process that associates the user attribute information and the behavioral pattern information for each user and stores them as user information, A selection process for selecting at least one candidate user for the first user based on the user information, A second generation process generates information based on the actions of a candidate user using a second trained model that has been trained to output information based on the user's actions based on the user attribute information and behavior pattern information of the user, A provision process that provides predetermined information based on the information derived from the aforementioned actions to the user terminal of the first user, An information processing program that executes [something]. This improves the usefulness of the information provided to users. Specifically, it allows for the provision of information to users based on information that may include the tacit knowledge of other users, making it possible to provide information that gives users some kind of insight.

[0145] (Item 12) A system including user terminals operated by each user and an information processing device, The information processing device comprises at least one processor, a storage unit for storing information processing programs, and a communication unit capable of communicating with the plurality of user terminals. By executing the information processing program, the information processing device will A first acquisition process for acquiring user attribute information relating to the attributes of the aforementioned user, A second acquisition process that acquires the location information of the user terminal and the time information corresponding to the location information, A first generation process generates behavioral pattern information relating to the user's behavioral patterns using a first trained model that has been trained to output patterns of the user's behavior based on the user's location information and time information history. A memory process that associates the user attribute information and the behavioral pattern information for each user and stores them as user information, A selection process for selecting at least one candidate user for the first user based on the user information, A second generation process generates information based on the actions of a candidate user using a second trained model that has been trained to output information based on the user's actions based on the user attribute information and behavior pattern information of the user, The system performs a provision process that provides predetermined information based on the information obtained from the aforementioned actions to the user terminal of the first user. system. This improves the usefulness of the information provided to users. Specifically, it allows for the provision of information to users based on information that may include the tacit knowledge of other users, making it possible to provide information that gives users some kind of insight.

[0146] (Item 13) A non-temporary computer-readable medium on which the information processing program described in item 12 is stored. This improves the usefulness of the information provided to users. Specifically, it allows for the provision of information to users based on information that may include the tacit knowledge of other users, making it possible to provide information that gives users some kind of insight. [Explanation of Symbols]

[0147] 1: Information Processing Device 2: User terminal 3: External device 5: System

Claims

1. An information processing apparatus comprising at least one processor, a storage unit for storing information processing programs, and a communication unit capable of communicating with user terminals operated by each user, The processor executes the information processing program, A first acquisition process for acquiring user attribute information relating to the attributes of the aforementioned user, A second acquisition process that acquires the location information of the user terminal and the time information corresponding to the location information, A first generation process generates behavioral pattern information relating to the user's behavioral patterns using a first trained model that has been trained to output patterns of the user's behavior based on the user's location information and time information history. A memory process that associates the user attribute information and the behavioral pattern information for each user and stores them as user information, A selection process for selecting at least one candidate user for the first user based on the user information, A second generation process generates information based on the actions of a candidate user using a second trained model that has been trained to output information based on the user's actions based on the user attribute information and behavior pattern information of the user, A provision process that provides predetermined information based on the information derived from the aforementioned actions to the user terminal of the first user, An information processing device that performs this task.

2. The predetermined information provided is information based on the candidate user's actions and the first user's user information. The information processing apparatus according to claim 1.

3. Furthermore, a third generation process is executed to generate the predetermined provided information using a third trained model that has been trained to output information that is not identical to the information based on the actions of the candidate user and the user information of the first user, but is related to the information based on the actions and the user information of the first user, based on the information based on the actions of the candidate user and the user information of the first user. The information processing apparatus according to claim 1 or 2.

4. Furthermore, based on the user's location information and time information history, a detection process is performed to detect changes in the user's behavior patterns. The third generation process is a process that generates the predetermined information to be provided based on information based on the actions of the candidate user, user information of the first user, and information regarding changes in the patterns of the actions. The information processing apparatus according to claim 3.

5. Furthermore, based on the content of the information based on the aforementioned action, a specific process is performed to determine the validity period of the information based on the aforementioned action. The aforementioned predetermined information is information based on the aforementioned actions within the validity period. The information processing apparatus according to claim 1 or 2.

6. Furthermore, a selection acceptance process is executed to accept the selection of a candidate user desired by the first user from among the at least one candidate user selected by the selection process. The aforementioned predetermined information provided is information based on the actions of candidate users selected by the first user. The information processing apparatus according to claim 1 or 2.

7. In the selection process described above, when selecting candidate users other than the first user, the candidate users selected by the first user are more likely to be selected than before they were selected by the first user. The information processing apparatus according to claim 6.

8. The user information includes the current location information of the user terminal, Depending on the current location information of the user terminal, at least one candidate user selected by the selection process may be changed. The information processing apparatus according to claim 1 or 2.

9. Furthermore, a third acquisition process is performed to obtain external service information regarding the user's usage of external services. The first generation process is a process that generates the behavior pattern information based on the history of the location information and the time information, and the external service information. The information processing apparatus according to claim 1 or 2.

10. An information processing method performed by an information processing device equipped with a communication unit capable of communicating with user terminals operated by each user, A first acquisition step involves obtaining user attribute information relating to the attributes of the aforementioned user, A second acquisition step involves acquiring the location information of the user terminal and the time information corresponding to the location information. A first generation step of generating behavioral pattern information relating to the user's behavioral patterns using a first trained model that has been trained to output patterns of the user's behavior based on the user's location information and time information history, A storage step in which, for each user, the user attribute information and the behavior pattern information are associated and stored as user information, A selection step in which at least one candidate user is selected for the first user based on the user information, A second generation step of generating information based on the user's behavior using a second trained model that has been trained to output information based on the user's behavior based on the user attribute information and behavior pattern information of the user, A provision step of providing predetermined information based on the information obtained from the aforementioned actions to the user terminal of the first user, Information processing methods, including those mentioned above.

11. An information processing program executed in an information processing device equipped with a communication unit capable of communicating with user terminals operated by each user, When the information processing program is executed by the processor of the information processing device, the information processing device will: A first acquisition process for acquiring user attribute information relating to the attributes of the aforementioned user, A second acquisition process that acquires the location information of the user terminal and the time information corresponding to the location information, A first generation process generates behavioral pattern information relating to the user's behavioral patterns using a first trained model that has been trained to output patterns of the user's behavior based on the user's location information and time information history. For each of the aforementioned multiple users, a storage process is performed to associate the user attribute information and the behavior pattern information and store them as user information. Based on the user information, a selection process is performed to identify at least one candidate user for the first user, A second generation process generates information based on the actions of a candidate user using a second trained model that has been trained to output information based on the user's actions based on the user attribute information and behavior pattern information of the user, A provision process that provides predetermined information based on the information derived from the aforementioned actions to the user terminal of the first user, An information processing program that executes [something].

12. A system including user terminals operated by each user and an information processing device, The information processing device comprises at least one processor, a storage unit for storing an information processing program, and a communication unit capable of communicating with the user terminal. By executing the information processing program, the information processing device will A first acquisition process for acquiring user attribute information relating to the attributes of the aforementioned user, A second acquisition process that acquires the location information of the user terminal and the time information corresponding to the location information, A first generation process generates behavioral pattern information relating to the user's behavioral patterns using a first trained model that has been trained to output patterns of the user's behavior based on the user's location information and time information history. For each user, a storage process is performed to associate the user attribute information and the behavior pattern information and store them as user information. A selection process for selecting at least one candidate user for the first user based on the user information, A second generation process generates information based on the actions of a candidate user using a second trained model that has been trained to output information based on the user's actions based on the user attribute information and behavior pattern information of the user, The system performs a provision process that provides predetermined information based on the information obtained from the aforementioned actions to the user terminal of the first user. system.

Images