Intention dataset generation support program, intention dataset generation support device, intention dataset generation support method, and recording medium

Abstract

To provide an intention dataset generation support program for generating an intention dataset available for learning that considers the internal intention accompanying a behavior.SOLUTION: An intention dataset generation support program of the present disclosure includes a behavior information acquisition procedure, an internal intention acquisition procedure, and a dataset generation procedure. The behavior information acquisition procedure acquires a user's behavior information, the internal intention acquisition procedure acquires the user's internal intention present in the behavior information, and the dataset generation procedure records the behavior information and the internal intention in linkage with each other to generate an intention dataset.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to an intention dataset generation support program, an intention dataset generation support device, an intention dataset generation support method, and a recording medium. [Background technology]

[0002] As a mechanism for outputting information in response to a user's behavior, an information processing device is known that includes a behavior history acquisition unit that acquires behavior history information including a behavior type, which is the type of user behavior, the title of the target of the behavior, and the time point at which the behavior was performed, and a model generation unit that uses multiple pieces of behavior history information in a chronological order as training data and generates a model that outputs recommended information indicating the target corresponding to the behavior history information when the behavior history information is input (Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-163557 Summary of the Invention [Problem to be solved by the invention]

[0004] Incidentally, when people decide to act, there are cases where their inherent will (intention), known as experiential knowledge or tacit knowledge, is strongly involved. Even if we try to use machine learning to recommend actions based on information about such behavior, the collected data does not include inherent intention, making it difficult to make highly accurate recommendations.

[0005] Therefore, an object of the present disclosure is to provide an intention dataset generation support program, an intention dataset generation support device, an intention dataset generation support method, and a recording medium that generate an intention dataset that can be used for learning taking into account inherent intentions accompanying actions. [Means for solving the problem]

[0006] In order to achieve the above object, the intention dataset generation support program of the present disclosure includes: The method includes a procedure for acquiring behavioral information, a procedure for acquiring implicit intentions, and a procedure for generating a dataset. The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step includes acquiring an implicit intention of the user that is implicit in the behavior information; the data set generation step includes linking and recording the behavioral information and the implicit intention to generate an intention data set. The intention data set generation support program causes a computer to execute each of the above procedures.

[0007] The intention dataset generation support device of the present disclosure includes: The system includes a behavioral information acquisition unit, an implicit intention acquisition unit, and a data set generation unit, the behavioral information acquisition unit acquires behavioral information of a user; the implicit intention acquisition unit acquires the user's implicit intention that is implicit in the behavior information; The data set generation unit generates an intention data set by linking and recording the behavior information and the implicit intention.

[0008] The intention dataset generation support method of the present disclosure includes: The method includes a behavioral information acquisition step, an implicit intention acquisition step, and a data set generation step, The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step acquires the user's implicit intention that is implicit in the behavior information, the data set generation step generates an intention data set by linking and recording the behavior information and the implicit intention. Each of the steps is a computer-implemented method.

[0009] The recording medium of the present disclosure includes: The method includes a procedure for acquiring behavioral information, a procedure for acquiring implicit intentions, and a procedure for generating a dataset. The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step includes acquiring an implicit intention of the user that is implicit in the behavior information; the data set generation step includes linking and recording the behavioral information and the implicit intention to generate an intention data set. A computer-readable recording medium on which an intention dataset generation support program for causing a computer to execute each of the above procedures is recorded. [Effects of the Invention]

[0010] According to the present disclosure, it is possible to generate an intention dataset that can be used for learning, taking into account the inherent intentions associated with actions. [Brief explanation of the drawings]

[0011] [Figure 1] FIG. 1 is a block diagram showing a configuration of an example of an intention dataset generation support device according to the present disclosure. [Figure 2] FIG. 2 is a block diagram illustrating an example of a hardware configuration of the intention dataset generation support device of the present disclosure. [Figure 3] FIG. 3 is a flowchart showing an example of a procedure performed by the intention dataset generation support program of the present disclosure. [Figure 4] FIG. 4 is a block diagram showing a configuration of an example of an intention dataset generation support device according to the present disclosure. [Figure 5] FIG. 5 is a flowchart showing an example of a procedure performed by the intention dataset generation support program of the present disclosure. [Figure 6] FIG. 6 is an example screen for explaining an example of a behavior information acquisition procedure of the intention dataset generation support program of the present disclosure. [Figure 7] FIG. 7 is an example of a screen for explaining an example of an implicit intention acquisition procedure of the intention dataset generation support program of the present disclosure. [Figure 8] FIG. 8 is an example of a screen for explaining an example of an implicit intention acquisition procedure of the intention dataset generation support program of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. The present disclosure is not limited to the following embodiments. In the following drawings, identical parts are designated by the same reference numerals. Furthermore, the descriptions of the embodiments can be used interchangeably unless otherwise specified, and the configurations of the embodiments can be combined unless otherwise specified. In the present disclosure, each drawing may apply to one or more embodiments.

[0013] [Embodiment 1] The intention dataset generation support program of the present disclosure is a program for causing a computer to execute a behavioral information acquisition procedure, an implicit intention acquisition procedure, and a dataset generation procedure. The intention dataset generation support program of the present disclosure can also be said to be a program for causing a computer to function as the behavioral information acquisition procedure, the implicit intention acquisition procedure, and the dataset generation procedure. Furthermore, the intention dataset generation support program of the present disclosure can also be said to be a program for causing a computer to execute, for example, each step of the intention dataset generation support method described below.

[0014] The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step includes acquiring an implicit intention of the user that is implicit in the behavior information; The data set generation step generates an intention data set by linking and recording the behavior information and the implicit intention.

[0015] For example, the "procedure" in each step can be read as a "process." The intention dataset generation support program of the present disclosure may be recorded on a computer-readable recording medium. The recording medium is, for example, a non-transitory computer-readable storage medium. The recording medium is not particularly limited, and examples thereof include random access memory (RAM), read-only memory (ROM), hard disk (HD), flash memory (e.g., solid state drive (SSD), USB flash memory, SD / SDHC card, etc.), optical disk (e.g., CD-R / CD-RW, DVD-R / DVD-RW, BD-R / BD-RE, etc.), magneto-optical disk (MO), and floppy disk (FD). The intention dataset generation support program of the present disclosure (also referred to as a programming product or program product) may be distributed from an external computer. The "distribution" may be, for example, distribution via a communication network or a device connected via a wire. The intention dataset generation support program of the present disclosure may be installed and executed on a device to which it is distributed, or may be executed without being installed. An information processing device capable of executing the intention dataset generation support program of the present disclosure can be referred to as, for example, an intention dataset generation support device of the present disclosure.

[0016] Next, the configuration of an example of an intention dataset generation support device according to the present disclosure will be described with reference to FIG. 1. FIG. 1 is a block diagram showing an example of the configuration of an intention dataset generation support device 10 according to the present disclosure (hereinafter also referred to as the present device 10). As shown in FIG. 1, the present device 10 includes a behavioral information acquisition unit 11, an implicit intention acquisition unit 12, and a dataset generation unit 13. Although not shown, the present device 10 may also include, for example, an input unit, an output unit, a display unit, and / or a storage unit. The behavioral information acquisition unit 11, the implicit intention acquisition unit 12, and the dataset generation unit 13 are capable of executing, for example, a behavioral information acquisition procedure, an implicit intention acquisition procedure, and a dataset generation procedure in the intention dataset generation support program according to the present disclosure, respectively.

[0017] The device 10 may be, for example, a single device including the above-described units, or a device in which the units can be connected via a communication network. The device 10 can also be connected to an external device (described later) via the communication network. The communication network is not particularly limited and any known network can be used, for example, a wired or wireless network. Examples of the communication network include the Internet, the World Wide Web (WWW), a telephone line, a Local Area Network (LAN), a Storage Area Network (SAN), a Delay Tolerant Networking (DTN), a Low Power Wide Area Network (LPWA), and a Local 5G (L5G). Examples of wireless communication include Wi-Fi (registered trademark), Bluetooth (registered trademark), Local 5G, and LPWA. Examples of the wireless communication include direct communication between devices (Ad Hoc communication), infrastructure communication, and indirect communication via an access point. The device 10 may be incorporated into a server as a system. Furthermore, the present device 10 may be, for example, a personal computer (PC, for example, desktop or notebook type) on which the program of the present disclosure is installed, a smartphone, a tablet terminal, etc. The present device 10 may be in the form of cloud computing or edge computing, for example, in which at least one of the above-mentioned units is located on a server and the other units are located on a terminal.

[0018] 2 shows a block diagram of the hardware configuration of the device 10. The device 10 includes, for example, a central processing unit (CPU, GPU, etc.) 101, a memory 102, a bus 103, a storage device 104, an input device 105, an output device 106, and a communication device 107. The components of the device 10 are connected to each other via the bus 103 and their respective interfaces (I / F).

[0019] The central processing unit 101 cooperates with other components via a controller (such as a system controller or an I / O controller) and controls the entire device 10. In the device 10, the central processing unit 101 executes, for example, the program of the present disclosure (the intention dataset generation support program) and other programs, and also reads and writes various types of information. Specifically, for example, the central processing unit 101 functions as a behavioral information acquisition unit 11, an implicit intention acquisition unit 12, and a dataset generation unit 13. The device 10 may include, as a computing device, other computing devices such as a CPU, a GPU (Graphics Processing Unit), an APU (Accelerated Processing Unit), or a combination of these.

[0020] The bus 103 can also be connected to, for example, external devices. Examples of the external devices include an external storage device (such as an external database), a printer, an external input device, an external display device, and an external imaging device. The device 10 can be connected to an external network (the communication line network) by, for example, a communication device 107 connected to the bus 103, and can also be connected to other devices via the external network.

[0021] The memory 102 may be, for example, a main memory (primary storage device). When the central processing unit 101 performs processing, the memory 102 reads various operating programs, such as the program of the present disclosure, stored in the storage device 104 (described later), and the central processing unit 101 receives data from the memory 102 and executes the programs. The main memory may be, for example, a RAM (random access memory). The memory 102 may also be, for example, a ROM (read only memory).

[0022] The storage device 104 is also referred to as an auxiliary storage device, for example, in contrast to the main memory (primary storage device). As described above, the storage device 104 stores an operating program including the program of the present disclosure. The storage device 104 may be, for example, a combination of a recording medium and a drive for reading and writing data from and to the recording medium. The recording medium is not particularly limited and may be, for example, an internal or external type, such as a hard disk (HD), CD-ROM, CD-R, CD-RW, MO, DVD, flash memory, or memory card. The storage device 104 may be, for example, a hard disk drive (HDD) or a solid-state drive (SSD) in which the recording medium and drive are integrated. When the device 10 includes the storage unit, for example, the storage device 104 functions as the storage unit. The storage unit can record, for example, behavioral information, implicit intentions, recommended actions, and hypothetical reason information, which will be described later.

[0023] In the present device 10, the memory 102 and the storage device 104 can also store various information such as log information, information acquired from an external database (not shown) or an external device, information generated by the present device 10, and information used when the present device 10 executes processing. In this case, the memory 102 and the storage device 104 may store, for example, the above-mentioned information on the user of the present device. Note that at least a portion of the information may be stored, for example, in an external server other than the memory 102 and the storage device 104, or may be stored in a distributed manner across multiple terminals using blockchain technology or the like.

[0024] The device 10 further includes, for example, an input device 105 and an output device 106. Examples of the input device 105 include pointing devices such as a touch panel, track pad, and mouse; a keyboard; imaging means such as a camera and scanner; card readers such as an IC card reader and a magnetic card reader; and audio input means such as a microphone. Examples of the output device 106 include display devices such as an LED display and a liquid crystal display; audio output devices such as a speaker; a printer; and the like. In the first embodiment, the input device 105 and the output device 106 are configured separately, but the input device 105 and the output device 106 may be configured as an integrated device, such as a touch panel display.

[0025] An example of processing by the intention dataset generation support program of the present disclosure will be described more specifically with reference to Fig. 3. Fig. 3 is a flowchart showing an example of each procedure of the intention dataset generation support program of the present disclosure.

[0026] The behavioral information acquisition unit 11 acquires behavioral information of the user (S1, behavioral information acquisition step). The behavioral information is not particularly limited and may be, for example, information indicating the type of user behavior, and the content thereof is not particularly limited. The behavioral information may be, for example, information acquired directly or indirectly about the user's behavior. Examples of the directly acquired information include operation information of a specific device and information about the user's behavior input by the user. Examples of the indirectly acquired information include information about the user's movement based on the user's location information. The behavioral information may be, for example, selection information selected by the user from options related to the user's behavior presented by the device 10. Note that the behavioral information acquisition unit 11 may, for example, acquire the user's behavioral information in real time, or may acquire behavioral information (behavioral history) recorded in the memory unit of the device 10 or an external storage device.

[0027] The implicit intention acquisition unit 12 acquires the user's implicit intention implicit in the behavioral information (S2, implicit intention acquisition step). The implicit intention is information about the user's intention (will) when the user performs the behavioral information. The implicit intention acquisition unit 12 can acquire the implicit intention, for example, by outputting a query to the user inquiring about the implicit intention implicit in the behavioral information and having the user input the user's implicit intention as a response to the query. The implicit intention acquisition unit 12 may output the query using, for example, a rule-based algorithm or a large-scale language model. The rule-based algorithm may, for example, associate and record query information corresponding to the behavioral information in advance, and read and output the query information corresponding to the user's behavioral information. In this case, the behavioral information and query information may be recorded in a storage unit of the device 10 or an external recording medium. The large-scale language model is not particularly limited, and examples thereof include ChatGPT by Open AI, Bing by Microsoft, Bard by Google (Alphabet Inc.), and LLM by NEC Corporation. In this case, the implicit intention acquisition unit 12 can generate and output the query by providing the large-scale language model with, for example, the behavioral information and a prompt with constraints imposed so as to output the query.

[0028] The implicit intention acquisition unit 12 may also output inferred reason information that infers the user's implicit intention implicit in the behavioral information. Examples of the inferred reason information include, but are not limited to, a sentence such as, "The reason you selected XX (behavior) is because XXX (reason), right?" The implicit intention acquisition unit 12 may output the inferred reason information using, for example, a rule-based algorithm, or may output the inferred reason information using a large-scale language model. The rule-based algorithm may, for example, associate and record in advance inferred reason information corresponding to the behavioral information, and read and output inferred reason information corresponding to the user's behavioral information. In this case, the behavioral information and inferred reason information may be recorded in a storage unit of the device 10 or may be recorded on an external recording medium. The large-scale language model is not particularly limited, and examples thereof include ChatGPT by Open AI, Bing by Microsoft, Bard by Google (Alphabet Inc.), and LLM by NEC Corporation. In this case, the implicit intention acquisition unit 12 can generate and output the query by, for example, providing the large-scale language model with the behavioral information and a prompt on which a constraint is imposed so as to output the inferred reason information. Furthermore, the implicit intention acquisition unit 12 may, for example, inquire about the accuracy of the inferred reason information. If the inferred reason information matches the user's implicit intention, the implicit intention acquisition unit 12 may, for example, acquire the inferred reason information again as the user's implicit intention. If the inferred reason information does not match the user's implicit intention, the implicit intention acquisition unit 12 can, for example, acquire the implicit intention by outputting a query to the user inquiring about the implicit intention implicit in the behavioral information and having the user input the user's implicit intention as a response to the query.

[0029] The dataset generation unit 13 generates an intention dataset by linking the behavioral information and the implicit intention and recording them (S3, dataset generation procedure). The dataset generation unit 13 may, for example, link the behavioral information and the implicit intention and record them in a storage unit of the device 10 to generate the intention dataset, or may link the behavioral information and the implicit intention and record them in a database external to the device 10. Furthermore, for example, if the user does not input an implicit intention in step S2, the dataset generation unit 13 may record information such as "no implicit intention input", "no response", or "no intention" in association with the behavioral information.

[0030] The intention dataset generation support method of the present disclosure is a method implemented by, for example, replacing the "procedures" in the intention dataset generation support program of the present disclosure with "processes." Specifically, the intention dataset generation support method of the present disclosure includes a behavioral information acquisition process, an implicit intention acquisition process, and a dataset generation process. The behavioral information acquisition process acquires user behavioral information. The implicit intention acquisition process acquires the user's implicit intention implicit in the behavioral information. The dataset generation process associates and records the behavioral information and the implicit intention to generate an intention dataset. The intention dataset generation support method of the present disclosure can be implemented, for example, using the intention dataset generation support device 10 of the present disclosure shown in FIG. 1 or FIG. 2. Note that the intention dataset generation support method of the present disclosure is not limited to a method using the intention dataset generation support device 10, for example. For the intention dataset generation support method of the present disclosure, the descriptions of the intention dataset generation support program and the intention dataset generation support device of the present disclosure can be cited, for example.

[0031] According to the intention dataset generation support program of the present disclosure, a behavioral information acquisition step acquires user behavioral information, a implicit intention acquisition step acquires the user's implicit intention implicit in the behavioral information, and a dataset generation step associates and records the behavioral information and the implicit intention to generate an intention dataset. Therefore, for example, according to the intention dataset generation support program of the present disclosure, when acquiring behavioral information in a scene involving complex decision-making, the behavioral information can be associated with the implicit intention and acquired. Therefore, by using the intention dataset generated by the intention dataset generation support program of the present disclosure, it is possible to learn behavioral information with higher accuracy.

[0032] [Embodiment 2] Another example of the intention dataset generation support program of the present disclosure will be described.

[0033] The intention dataset generation support program of this embodiment is similar to the intention dataset generation support program of the first embodiment, except that it includes, for example, a recommended action output procedure and a match determination procedure in addition to the configuration of the intention dataset generation support program of the first embodiment, and the description thereof can be cited. The intention dataset generation support program of this embodiment includes, for example, a recommended action output procedure and a match determination procedure, where the recommended action output procedure outputs recommended actions for the user, the match determination procedure determines whether the recommended actions match the behavioral information, and the implicit intention acquisition procedure acquires the user's implicit intention implicit in the behavioral information when the recommended actions do not match the behavioral information.

[0034] Next, the intention dataset generation support device of this embodiment will be described with reference to Fig. 6. The intention dataset generation support device 10A of the present disclosure is similar to the intention dataset generation support device 10 of the first embodiment, except that it includes a recommended behavior output unit 14 and a match determination unit 15 in addition to the configuration of the intention dataset generation support device 10 of the first embodiment, and the description thereof can be cited. The intention dataset generation support device 10A of this embodiment includes, for example, the recommended behavior output unit 14 and the match determination unit 15. The recommended behavior output unit 14 outputs recommended behaviors for the user. The match determination unit 15 determines whether the recommended behaviors match the behavioral information. If the recommended behaviors do not match the behavioral information, the implicit intention acquisition unit 12 acquires the user's implicit intention implicit in the behavioral information.

[0035] 4, the intention dataset generation support device 10A includes a recommended action output unit 14 and a match determination unit 15 in addition to the configuration of the intention dataset generation support device 10 of embodiment 1. The hardware configuration of the intention dataset generation support device 10A is the same as that of the intention dataset generation support device 10 of FIG. 2, except that the central processing unit 101 includes the configuration of the intention dataset generation support device 10A of FIG. 4 instead of the configuration of the intention dataset generation support device 10 of FIG. 1.

[0036] An example of processing by the intention dataset generation support program of the present disclosure will be described more specifically with reference to Fig. 7. Fig. 7 is a flowchart showing an example of each procedure of the intention dataset generation support program of the present disclosure.

[0037] First, the recommended action output unit 14 outputs, for example, a recommended action for the user (S11, recommended action output step). The recommended action is not particularly limited, and is, for example, information indicating the type of action recommended to the user, and its content is not particularly limited. The recommended action may include, for example, the same information as the type of the action information, or may include different information. Specific examples of the recommended action include, for example, information recommending an operation of a specific device, text information to be input by the user, etc. The recommended action may be, for example, an option related to the user's action presented by the device 10. Furthermore, the recommended action output unit 14 may output the recommended action based on, for example, premise information. The premise information is, for example, information that is a premise for the user's action, and is not particularly limited. Examples of the premise information include information such as a user's dialogue log and user attributes.

[0038] The recommended action output unit 14 may output, for example, hypothetical reason information explaining the reason for recommending the recommended action together with the output of the recommended action. The hypothetical reason information is, for example, information in which the reason for recommendation is associated with each recommended action. The hypothetical reason information may, for example, be recorded in a storage unit of the device 10A or an external storage medium. The hypothetical reason information may, for example, be generated by using a machine learning model that performs machine learning on the intention dataset. The recommended action output unit 14 may, for example, output the hypothetical reason information using a rule-based algorithm or may output the hypothetical reason information using a large-scale language model. The rule-based algorithm may, for example, associate and record hypothetical reason information corresponding to the recommended action in advance, and read and output the hypothetical reason information corresponding to the recommended action. In this case, the recommended action and the hypothetical reason information may be recorded in a storage unit of the device 10A or an external storage medium. The large-scale language model is not particularly limited, and examples thereof include ChatGPT by Open AI, Bing by Microsoft, Bard by Google (Alphabet Inc.), and LLM by NEC Corporation. In this case, the recommended action output unit 14 can generate and output the hypothetical reason information by, for example, providing the large-scale language model with the recommended action and a prompt that imposes constraints so as to output the hypothetical reason information.

[0039] Next, S1 is performed in the same manner as S1 in the embodiment 1. Note that, when the recommended behavior is a behavior option of the user, the behavior information acquisition unit 11 can acquire, for example, the selection result of the option as behavior information of the user.

[0040] The match determination unit 15 determines, for example, whether the recommended action matches the action information (S12, match determination step). When the recommended action is an action option of the user and the action is a selection result of the option, the match determination unit 15 can determine whether the recommended action matches the action information by determining, for example, whether the option recommended as the recommended action matches the option selected by the user. When the recommended action does not match the action information, for example, the implicit intention acquisition unit 12 can acquire the user's implicit intention implicit in the action information.

[0041] For example, when the recommended action matches the action information, the dataset generation unit 13 may link at least one of the action information and the recommended action with the hypothesis reason information and record them to generate an intention dataset. Furthermore, the dataset generation unit 13 may further link and record the premise information, for example. Furthermore, the dataset generation unit 13 may further link and record, for example, recommendation success / failure information in the intention dataset. Examples of the success / failure information include information indicating that the recommended action matches the action information (recommendation successful) and information indicating that the recommended action does not match the action information (recommendation failed). Furthermore, when the recommendation fails, the dataset generation unit 13 may further record whether or not an implicit intention was input.

[0042] In this embodiment, the recommended action output unit 14 may update the hypothetical reason information based on, for example, the intention dataset. The recommended action output unit 14 may, for example, analyze each piece of information recorded in the intention dataset and update the hypothetical reason information based on the analysis results. The analysis may, for example, be an analysis using natural language processing. Specifically, the recommended action output unit 14 may convert the information recorded in the intention dataset into vectors using W2V (Word2Vec) and perform analysis of variance based on the converted vectors. This allows the information recorded in the intention dataset to be classified into successful recommendations (recommended actions adopted by the user), corrective recommendations (recommended actions not adopted by the user, but for which an implicit intention was input), ignored recommendations (recommended actions not adopted by the user, and for which no implicit intention was input), etc. Furthermore, the recommended action output unit 14 may, for example, pass the intention dataset to an LLM to analyze the linguistic information (behavioral information, recommended actions, hypothetical reason information, and implicit intention) included in the intention dataset. In this case, the recommended behavior output unit 14 can, for example, pass a prompt with constraints imposed on it to classify the information in the intention dataset (e.g., the success recommendation, the correction recommendation, and the ignore recommendation) and the information in the intention dataset to the LLM for analysis.

[0043] Next, a specific example of the processing of the device 10A will be described with reference to Figures 6 to 10. In the following description, the device 10A will be described as being incorporated into a chat system used for consultation between a mentor and a client, but the present disclosure is not limited to the following example.

[0044] First, a mentor, which is the device 10A, is consulting with a client in the chat system shown in FIG. 6. The recommended action output unit 14 of the device 10A monitors the interaction between the mentor (the user of the device 10A) and the client in the chat system shown in FIG. 6, receives the client's consultation ("What I want to achieve: Implementing the technology I'm researching in a useful way in society." To achieve this, the first action is to "clarify who it will be useful for.") and determines a recommended action (selection of the "WILL" button and "fundamental reason") using the consultation content as premise information. Then, the recommended action output unit 14 displays candidate buttons for replying to the client to the mentor, as shown in FIG. 6. At this time, the recommended action output unit 14 also outputs hypothetical reason information ("It is recommended that you ask WILL for the fundamental reason for the reason of XXX") as the basis for the recommended action.

[0045] Next, the behavioral information acquisition unit 11 monitors the behavior (input to the chat system) of the mentor (user of the device 10) in the chat system shown in Figure 7. The mentor checks the consultation's request ("What I want to achieve: Implement the technology I'm researching in a useful way in society." To achieve this, the first action is to "clarify who it will be useful for"), the recommended action ("WILL" button and "fundamental reason") presented by the device 10A, and the hypothetical reason information (for some reason, it is recommended that you ask the fundamental reason for WILL), and then inputs a question ("What does 'research' mean?"). The behavioral information acquisition unit 11 of the device 10 acquires the question from the mentor ("What does 'research' mean?") as the user's behavioral information.

[0046] Because the recommended action (the "WILL" button and the "fundamental reason") does not match the user's action (the input of a question ("What does 'research' mean?"), the match determination unit 15 of the device 10A executes processing by the implicit intention acquisition unit 12, as shown in FIG. 7. Specifically, the implicit intention acquisition unit 12 presents the user, or mentor, with a question asking about the intention of the behavioral information (if you don't mind, please tell us why you selected the button: "Teach" or "Don't teach"). When the mentor presses the "Teach" button, the implicit intention acquisition unit 12 displays a pop-up window on the chat system for inputting an implicit intention, as shown in FIG. 8, and prompts the mentor to input the intention of the behavior. Then, the implicit intention acquisition unit 12 can acquire, for example, the input intention as the user's implicit intention implicit in the behavioral information. Then, the dataset generation unit 13 of the device 10A associates and records the behavioral information and the implicit intention, thereby generating an intention dataset.

[0047] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0048] <Additional Notes> Some or all of the above embodiments can be described as, but not limited to, the following supplementary notes. (Appendix 1) The method includes a procedure for acquiring behavioral information, a procedure for acquiring implicit intentions, and a procedure for generating a dataset. The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step includes acquiring an implicit intention of the user that is implicit in the behavior information; the data set generation step includes linking and recording the behavioral information and the implicit intention to generate an intention data set. An intention dataset generation support program for causing a computer to execute each of the above procedures. (Appendix 2) 2. The intention dataset generation support program according to claim 1, wherein the implicit intention acquisition step outputs inferred reason information for inferring the user's implicit intention underlying the behavior information. (Appendix 3) A recommended action output procedure and a matching determination procedure are included; The recommended action output step outputs a recommended action for the user; the matching determination step determines whether the recommended action matches the action information; 3. The intention dataset generation support program according to claim 1, wherein the implicit intention acquisition step acquires the user's implicit intention implicit in the behavior information when the recommended behavior and the behavior information are inconsistent. (Appendix 4) the recommended action output step includes outputting hypothetical reason information indicating a hypothetical reason why the recommended action was recommended; The intention dataset generation support program according to claim 3, wherein the dataset generation procedure generates an intention dataset by linking and recording at least one of the behavioral information and the recommended behavior with the hypothesis reason information when the recommended behavior and the behavioral information match. (Appendix 5) 5. The intention dataset generation support program according to claim 4, wherein the recommended action output step updates the hypothesis reason information based on the intention dataset. (Appendix 6) The system includes a behavioral information acquisition unit, an implicit intention acquisition unit, and a data set generation unit, the behavioral information acquisition unit acquires behavioral information of a user; the implicit intention acquisition unit acquires the user's implicit intention that is implicit in the behavior information; The data set generation unit generates an intention data set by linking and recording the behavior information and the implicit intention. (Appendix 7) 7. The intention dataset generation assistance device according to claim 6, wherein the implicit intention acquisition unit outputs inference reason information for inferring the user's implicit intention that is implicit in the behavior information. (Appendix 8) A recommended action output unit and a match determination unit are included, the recommended action output unit outputs a recommended action for the user; the matching determination unit determines whether the recommended behavior matches the behavior information; The intention dataset generation assistance device according to claim 6 or 7, wherein the implicit intention acquisition unit acquires the user's implicit intention implicit in the behavior information when the recommended behavior and the behavior information are inconsistent. (Appendix 9) the recommended action output unit outputs hypothetical reason information indicating a hypothetical reason why the recommended action was recommended; The intention dataset generation support device according to claim 8, wherein, when the recommended action matches the action information, the dataset generation unit links at least one of the action information and the recommended action with the hypothesis reason information and records them to generate an intention dataset. (Appendix 10) 10. The intention dataset generation support device according to claim 9, wherein the recommended action output unit updates the hypothesis reason information based on the intention dataset. (Appendix 11) The method includes a behavioral information acquisition step, an implicit intention acquisition step, and a data set generation step, The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step acquires the user's implicit intention that is implicit in the behavior information, the data set generation step generates an intention data set by linking and recording the behavior information and the implicit intention. The intention dataset generation support method, in which each of the steps is executed by a computer. (Appendix 12) 12. The intention dataset generation support method according to claim 11, wherein the implicit intention acquisition step outputs inferred reason information for inferring the user's implicit intention underlying the behavior information. (Appendix 13) A recommended action output step and a match determination step are included, The recommended action output step outputs a recommended action for the user, the matching determination step determines whether the recommended behavior matches the behavior information; 13. The intention dataset generation support method according to claim 11, wherein the implicit intention acquisition step acquires the user's implicit intention implicit in the behavior information when the recommended behavior and the behavior information are inconsistent. (Appendix 14) The recommended action output step outputs hypothetical reason information indicating a hypothetical reason why the recommended action was recommended, The intention dataset generation support method according to claim 13, wherein, when the recommended action matches the action information, the dataset generation step links and records at least one of the action information and the recommended action with the hypothesis reason information to generate an intention dataset. (Appendix 15) The intention dataset generation support method according to claim 14, wherein the recommended action output step updates the hypothesis reason information based on the intention dataset. (Appendix 16) The method includes a procedure for acquiring behavioral information, a procedure for acquiring implicit intentions, and a procedure for generating a dataset. The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step includes acquiring an implicit intention of the user that is implicit in the behavior information; the data set generation step includes linking and recording the behavioral information and the implicit intention to generate an intention data set. A computer-readable recording medium on which an intention dataset generation support program for causing a computer to execute each of the above procedures is recorded. (Appendix 17) 17. The recording medium according to claim 16, wherein the step of acquiring the implicit intention outputs inferred reason information for inferring the user's implicit intention that is implicit in the behavior information. (Appendix 18) A recommended action output procedure and a matching determination procedure are included; The recommended action output step outputs a recommended action for the user; the matching determination step determines whether the recommended action matches the action information; 18. The recording medium according to claim 16, wherein the implicit intention acquisition step acquires the user's implicit intention that is implicit in the behavior information when the recommended behavior and the behavior information are inconsistent. (Appendix 19) the recommended action output step includes outputting hypothetical reason information indicating a hypothetical reason why the recommended action was recommended; The recording medium of claim 18, wherein the dataset generation procedure generates an intention dataset by linking and recording at least one of the behavioral information and the recommended behavior with the hypothesis reason information when the recommended behavior and the behavioral information match. (Appendix 20) 20. The recording medium of claim 19, wherein the recommended action output step updates the hypothesis reason information based on the intention dataset. [Industrial Applicability]

[0049] According to the present disclosure, when acquiring behavioral information in a scene involving complex decision-making, the behavioral information can be linked to the intentions inherent in the behavioral information and acquired. Therefore, by using an intention dataset generated by the intention dataset generation support program of the present disclosure, it becomes possible to learn behavioral information with higher accuracy. Therefore, the present disclosure can be widely and usefully used in various fields that utilize machine learning. [Explanation of symbols]

[0050] 10, 10A Intention dataset generation support device 11 Behavioral information acquisition unit 12. Intention Acquisition Unit 13 Dataset Generation Unit 14 Recommended Action Output Section 15 Match determination section 101 Central Processing Unit 102 memory 103 Bus 104 Storage device 105 Input Device 106 Output Device 107 Communication Devices

Claims

1. The method includes a procedure for acquiring behavioral information, a procedure for acquiring implicit intentions, and a procedure for generating a dataset. The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step includes acquiring an implicit intention of the user that is implicit in the behavior information; the data set generation step includes linking and recording the behavioral information and the implicit intention to generate an intention data set. An intention dataset generation support program for causing a computer to execute each of the above procedures.

2. 2. The intention dataset generation support program according to claim 1, wherein the implicit intention acquisition step outputs inferred reason information for inferring the user's implicit intention that is implicit in the behavior information.

3. A procedure for outputting recommended actions and a procedure for determining whether a match has been made; The recommended action output step outputs a recommended action for the user; the matching determination step determines whether the recommended behavior matches the behavior information; 3. The intention dataset generation support program according to claim 1, wherein the implicit intention acquisition step acquires the user's implicit intention implicit in the behavior information when the recommended behavior and the behavior information are inconsistent.

4. the recommended action output step includes outputting hypothetical reason information indicating a hypothetical reason why the recommended action was recommended; 4. The intention dataset generation support program according to claim 3, wherein the dataset generation procedure generates an intention dataset by linking and recording at least one of the behavioral information and the recommended behavior with the hypothesis reason information when the recommended behavior and the behavioral information match.

5. 5. The intention dataset generation support program according to claim 4, wherein the recommended action output step updates the hypothesis reason information based on the intention dataset.

6. The system includes a behavioral information acquisition unit, an implicit intention acquisition unit, and a data set generation unit, the behavioral information acquisition unit acquires behavioral information of a user; the implicit intention acquisition unit acquires the user's implicit intention that is implicit in the behavior information; The data set generation unit generates an intention data set by linking and recording the behavior information and the implicit intention.

7. The intention dataset generation support device according to claim 6 , wherein the implicit intention acquisition unit outputs inferred reason information for inferring the user's implicit intention that is implicit in the behavior information.

8. a recommended action output unit and a match determination unit; the recommended action output unit outputs a recommended action for the user; the matching determination unit determines whether the recommended behavior matches the behavior information; The intention dataset generation assistance device according to claim 6 , wherein the implicit intention acquisition unit acquires the user's implicit intention implicit in the behavior information when the recommended behavior and the behavior information do not match.

9. The method includes a behavioral information acquisition step, an implicit intention acquisition step, and a data set generation step, The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step acquires the user's implicit intention that is implicit in the behavior information, the data set generation step generates an intention data set by linking and recording the behavior information and the implicit intention. The intention dataset generation support method, in which each of the steps is executed by a computer.

10. The method includes a procedure for acquiring behavioral information, a procedure for acquiring implicit intentions, and a procedure for generating a dataset. The behavioral information acquisition step acquires behavioral information of a user, the implicit intention acquisition step includes acquiring an implicit intention of the user that is implicit in the behavior information; the data set generation step includes linking and recording the behavioral information and the implicit intention to generate an intention data set. A computer-readable recording medium on which an intention dataset generation support program for causing a computer to execute each of the above procedures is recorded.

Images