Device and method

Abstract

The objective of the present invention is to provide a device and method capable of supporting decision-making based on a user's own determination criteria.　In a prompt generating device 100 according to the present disclosure, a request accepting unit 108a acquires request information from a user. A prompt generating unit 108 functions as a request information acquiring unit, a tendency information acquiring unit, and a generating unit. The prompt generating unit 108 acquires information (user-related information) relating to the user and the request information from the user, accepted by the request accepting unit 108a. Then, the prompt generating unit 108 acquires preference / thought data (behavioral tendency information) relating to the user from a list database of a preference / thought data generating unit 106. The prompt generating unit 108 generates a prompt for outputting a decision-making result based on the request information, on the basis of the preference / thought data (behavioral tendency information).

Description

Apparatus and method

[0001] The present invention relates to an apparatus and method for assisting human decision making.

[0002] Patent Literature 1 describes an image diagnostic system that uses artificial intelligence to diagnose medical images. This image diagnostic system includes an extraction unit that extracts decision-making basis information that influenced the estimation of selected diagnostic content, and outputs the decision-making basis information.

[0003] JP 2022-146822 A

[0004] However, with the technology described in Patent Document 1, it is difficult to clarify the decision-making and the basis thereof based on the criteria of the user who requests the evaluation judgment.

[0005] Therefore, an object of the present disclosure is to provide an apparatus and method that can support decision-making based on the user's own criteria.

[0006] The device of the present invention comprises a request information acquisition unit that acquires a user's request information, a tendency information acquisition unit that acquires the user's behavioral tendency information, and a generation unit that generates a prompt for outputting a decision-making result for the request information based on the behavioral tendency information.

[0007] According to the present invention, decision-making support can be provided based on the user's own criteria.

[0008] FIG. 1 is a schematic diagram illustrating an overview of the processing of the present disclosure. FIG. 2 is a diagram illustrating a system configuration including a prompt generation device 100 of the present disclosure. FIG. 3 is a diagram illustrating a functional configuration of the prompt generation device 100. FIG. 4 is a diagram illustrating a correspondence between a user ID and an MBTI. FIG. 5 is a diagram illustrating decision-making data. FIG. 6 is a diagram illustrating an example of a user information DB. FIG. 7 is a diagram illustrating a personality estimation process. FIG. 8 is a diagram illustrating a decision-making estimation process. FIG. 9 is a diagram illustrating a specific example of preference and thought data. FIG. 10 is a diagram illustrating generation of preference and thought data using attribute information in the preference and thought data generation unit 106. FIG. 11 is a diagram illustrating generation of preference and thought data using MBTI in the preference and thought data generation unit 106. FIG. 12 is a diagram illustrating generation of preference and thought data using questionnaire results in the preference and thought data generation unit 106. FIG. 13 is a diagram illustrating an example format of a prompt generated by the prompt generation unit 108. FIG. 14 is a diagram illustrating a prompt generated by the prompt generation unit 108 according to the above prompt format. Fig. 15 is a diagram showing "My personality" from the prompts. Fig. 16 is a diagram showing a specific example of a process for updating preference and thought data. Fig. 17 is a flowchart showing the model construction operation of the prompt generation device 100 according to the present disclosure. Fig. 18 is a diagram showing a flowchart of a prompt generation process. Fig. 19 is a diagram showing an example of the hardware configuration of the prompt generation device 100 according to an embodiment of the present disclosure.

[0009] The embodiments of the present disclosure will be described with reference to the accompanying drawings. Where possible, the same parts are designated by the same reference numerals and redundant description will be omitted.

[0010] FIG. 1 is a schematic diagram illustrating an overview of the processing of the present disclosure. In the present disclosure, preferences and thoughts are expressed based on information about a user. In the present disclosure, the information about a user includes personality, decision-making data, and user information. Personality and decision-making data are information that can be estimated. Personality and decision-making data are information set by a questionnaire. Personality is set, for example, based on the MBTI (Myers-Briggs Type Indicator). Decision-making data is information consisting of questionnaire-style questions, their answers, and reasons. User information is so-called attribute information, and is information that is registered in advance.

[0011] A prompt sentence for the decision to be made is then generated, and a generative AI model such as an LLM (Large Language Model) makes a decision based on the prompt sentence. Based on the decision, data (or algorithms) for preferences and thoughts are updated.

[0012] 2 is a diagram showing the configuration of a system including a prompt generation device 100 according to the present disclosure. As shown in the figure, this system includes the prompt generation device 100 and an LLM 200. When a user terminal 300 sends a request to the prompt generation device 100, the prompt generation device 100 generates a prompt in response to the request, sends it to the LLM 200, and receives a response. The prompt generation device 100 transfers the response to the user terminal 300, allowing the user of the user terminal 300 to receive the response. However, this is not limiting, and the LLM 200 may be located within the prompt generation device 100. Alternatively, the prompt generation device 100 or the LLM 200 may be located within the user terminal 300.

[0013] For example, the prompt generation device 100 (including the RAG (Retrieval-Augmented Generation) function) and the LLM 200 may be implemented within the user terminal 300. However, the database searched by the RAG function of the prompt generation device 100 may be located within the user terminal 300 or on a network. Alternatively, the RAG function may be located on the user terminal 300 or on a network, and the LLM 200 may also be implemented on the network. The database searched by the RAG function may be located within the user terminal 300 or on a network.

[0014] In this disclosure, the LLM 200 is used as the language model, but other models may be used as long as they are generative AI models. In this disclosure, a generative AI model is a model that can generate content in response to a prompt containing input information, according to the instructions, context, question, and output format indicated by the prompt, and return the content as response information. The prompt can also include input information, in which case the generative AI model generates response information targeted at the input information. The generative AI model may be, for example, an interactive AI model that includes a large language model (LLM) and a user interface (UI) for interacting with the user, enabling text or voice chat with the user. Examples of such generative AI models include ChatGPT, GPT (registered trademark)-3.5, GPT-4V, PaLM2, etc. In this disclosure, the prompt generation device 100 is capable of providing content provision functions using multiple types of interactive AI models. These interactive AI models may be stored within the prompt generation device 100, or may be stored in another device connected to the prompt generation device 100 via a network so that information can be exchanged with the user via the prompt generation device 100. Although only one prompt generation device 100 is shown in the figure, multiple prompt generation devices 100 may be included.

[0015] In this disclosure, a prompt is information indicating an instruction or question entered by a user in an interactive system such as an interaction with a generative AI model or a command line interface (CLI).

[0016] 3 is a diagram showing the functional configuration of the prompt generation device 100. As shown in the figure, the prompt generation device 100 includes a personality data acquisition unit 101, a decision-making data acquisition unit 102, a user information acquisition unit 103, a personality data estimation unit 104, a decision-making data estimation unit 105, a preference and thought data generation unit 106, a preference and thought data update unit 107, a prompt generation unit 108, and a request reception unit 108a.

[0017] The personality data acquisition unit 101 is a part that acquires user personality data serving as a judgment criterion for each user based on instructions from the user terminal 300. This personality data acquisition unit 101 stores a personality DB that associates user IDs with MBTIs, and acquires the corresponding MBTI based on the user ID in the judgment request sent from the user terminal 300. FIG. 4 is a diagram showing the association between user IDs and MBTIs. The personality data acquisition unit 101 holds the personality DB shown in FIG. 4 and acquires the MBTI based on this. The personality DB may be located on an external server, and the personality data acquisition unit 101 may access the personality DB.

[0018] In this disclosure, MBTI is used as the personality data, but it is not necessary to use MBTI, and any widely used data may be used. In this disclosure, users are classified into 32 patterns according to MBTI. In this disclosure, the 32 patterns are INTJ-A, INTP-A, ENTJ-A, ENTP-A, INFJ-A, INFP-A, ENFJ-A, ENFP-A, ISTJ-A, ISFJ-A, ESTJ-A, ESFJ-A, ISTP-A, ISFP-A, ESTP-A, ESFP-A, INTJ-T, INTP-T, ENTJ-T, ENTP-T, INFJ-T, INFP-T, ENFJ-T, ENFP-T, ISTJ-T, ISFJ-T, ESTJ-T, ESFJ-T, ISTP-T, ISFP-T, ESTP-T, and ESFP-T.

[0019] The decision-making data acquisition unit 102 acquires decision-making data. The decision-making data includes the basis for making a decision and the result of that decision. The decision-making data is acquired through a questionnaire.

[0020] Here, the decision-making data will be explained. FIG. 5(a) is a diagram showing an example of the contents of a questionnaire for determining the decision-making data. As shown in the figure, the questionnaire is composed of a question number, a question statement, an answer, a reason, and an answer number. The question statement is composed of a question that can be answered with YES / NO and a sentence asking for the reason. The answer is an item that requires the subject to select YES / NO, and the reason indicates the reason. The answer number is a number that indicates the combination of the answer and the reason.

[0021] For example, in FIG. 5( a), the following questionnaire content is prepared: Question: Your favorite movie, which you own on Blu-ray Disc, is being shown in a rerun. Would you go to the cinema at this time? Please select the most appropriate reason. Answer: I will go to the cinema. I will not go to the cinema. Reason: I want to watch a movie with better sound and a better screen. I like the atmosphere of the cinema. I want to watch a movie with someone. I want to contribute to the sales of my favorite movie. FIG. 5( b) is a diagram illustrating an example of a decision-making DB accessed by the decision-making data acquisition unit 102. As shown in the diagram, the decision-making DB stores survey results, in which answer numbers are assigned to question numbers, for each user (each user ID). This decision-making DB may be maintained by the decision-making data acquisition unit 102 or may be located on an external server. The decision-making data acquisition unit 102 can acquire decision-making data by accessing the decision-making DB.

[0022] The user information acquisition unit 103 is a part that acquires user information. In the present disclosure, the user information refers to user attribute information, but may include or be other information. The user information acquisition unit 103 may hold a user information DB and acquire user information from the user information DB, or may acquire user information from an external user information DB.

[0023] 6 is a diagram showing an example of a user information DB. As shown in the figure, attribute information such as gender, age, prefecture of origin, prefecture of residence, marital status, number of children, occupation, and annual income is included for each user. User information may include at least one of these pieces of information, or may be composed of other information as well. Attribute information is information that indicates the characteristics or properties of a user, and the above is one example.

[0024] The personality data estimation unit 104 is a part that estimates personality information by machine learning for users for whom personality data cannot be obtained. Data from users for whom personality data can be obtained is used as training data, and a personality estimation model that estimates one of 32 types of personality is constructed. The personality estimation model is a machine learning model that inputs an unknown user and estimates a plausible personality.

[0025] Specifically, the personality data estimation unit 104 estimates the user's personality that is not stored in the personality DB based on the information stored in the personality DB and the user information DB. The personality data estimation unit 104 has a personality estimation model, and by inputting user information using this personality estimation model, outputs the corresponding MBTI.

[0026] FIG. 7 is a diagram illustrating the personality estimation process. FIG. 7( a) is an explanatory diagram illustrating the construction of a personality estimation model based on a personality DB and a user information DB. As shown in the figure, data in which attribute information and MBTI are associated with each other is generated for each user based on data in which the attribute information and MBTI are known and stored in the personality DB and the user information DB. A personality estimation model is then constructed based on this data. The personality estimation model is constructed by learning using machine learning with the attribute information as an explanatory variable and the MBTI as a target variable. This learning process is performed by a learning unit (not shown in the figure). The learning unit may be located within the prompt generation device 100 or externally.

[0027] 7(b) is a diagram showing how attribute information and MBTI of an unknown user are estimated based on the constructed personality estimation model. The personality estimation model shown in FIG. 7(a) is a machine learning model that can estimate a user's MBTI by inputting the user's attribute information. As shown in the figure, this personality estimation model inputs attribute information for each user whose attribute information is known, and outputs (estimates) the corresponding MBTI.

[0028] The decision-making data estimation unit 105 is a part that uses machine learning to estimate the decision-making data of users for whom the decision-making data cannot be obtained. Attribute information of users for whom the personality data can be obtained is used as learning data, and a decision-making estimation model is constructed that estimates each question related to decision-making and its answers. The decision-making estimation model inputs attribute information of users whose decision-making data is unknown and estimates various values ​​related to decision-making.

[0029] FIG. 8 is a diagram illustrating the process for decision-making estimation. FIG. 8( a) is an explanatory diagram illustrating the construction of a decision-making estimation model based on a user information DB and decision-making data (survey results). As shown in the figure, data in which attribute information and decision-making data are associated with each other is generated for each user based on data in which attribute information and decision-making data are known and stored in the user information DB and decision-making DB. A decision-making estimation model is then constructed based on this data. The decision-making estimation model is constructed by learning using machine learning with attribute information as an explanatory variable and decision-making data as a target variable. This learning process is performed by a learning unit (not shown in the figure). The learning unit may be located within the prompt generation device 100 or externally.

[0030] 8(b) is a diagram showing how attribute information and decision-making data of an unknown user are estimated based on the constructed decision-making estimation model. The decision-making estimation model shown in FIG. 8(a) is a machine learning model that can estimate a user's decision-making data by inputting the user's attribute information. As shown in the figure, this decision-making estimation model inputs attribute information for each user whose attribute information is known, and outputs (estimates) the corresponding decision-making data.

[0031] The preference and thought data generation unit 106 is a part that generates data expressing the user's preferences and thoughts using personality data, decision-making data, and user information. The basic idea of ​​this disclosure is to generate preference and thought data from three major categories: thoughts seen in the user's attributes, thoughts that are typical of the user's personality, and thoughts unique to the user. The preference and thought data acquisition (generation) method utilizes LLM. The table that is finally generated is as shown in Figure 9, and preference and thought data is linked to the user ID.

[0032] 9 is a diagram showing a specific example of preference and thought data. As shown in the figure, preference and thought data is generated for each user, and the preference and thought data generation unit 106 associates and stores this preference and thought data as a table. As described above, this preference and thought data is generated by a generative AI model such as LLM. The preference and thought data is information that lists the user's way of thinking, the user's behavioral tendencies, etc. in itemized form.

[0033] This generation process will be described. Figure 10 illustrates the generation of preference and thought data using attribute information in the preference and thought data generation unit 106. Figure 10(a) illustrates an example of a prompt output to the LLM 200. As shown in the figure, this prompt includes instructions, attribute information, and an output format instruction. The instructions include, "Please infer 10 general personality traits found in the attribute information below in the following output format." This is instruction information for generating an answer in the LLM 200. The personality traits indicate the user's behavioral tendencies and preference and thought tendencies. This content is an example, and the prompt includes the following: based on attribute information, and inferring preference and thought data. While the number of inferences is 10, it may, of course, be less than or greater than that. The output format indicates a sentence.

[0034] 10(b) is a diagram showing the user's preference and thought data generated by the LLM 200 based on this prompt (attribute information). As shown in the figure and below, ten personality traits of the user are listed.

[0035] (Example output) ・You may value a stable life base and be focused on your current career as a company employee. ・Being from Fukushima Prefecture, you may have a strong sense of hometown pride and attachment to your community. ・Being unmarried and childless, you tend to focus on personal freedom and your career. ・Being living in Tokyo, you may be interested in urban environments and opportunities. ・With a current annual income of 3 million yen, you may be seeking financial stability. ・Being unmarried, you may be prioritizing your time and hobbies. ・Even at the young age of 28, you may have plans and goals for the future. ・Being employed as a company employee, you may have a sense of responsibility and a hardworking personality. ・With an annual income within a certain range, you may be satisfied with your current situation or considering future improvements. ・Moving from Fukushima Prefecture to Tokyo may indicate your ability to adapt to new environments.

[0036] FIG. 11 illustrates the generation of preference and thought data using MBTI by the preference and thought data generation unit 106. FIG. 11(a) illustrates an example of a prompt output to the LLM 200. As shown in the figure, this prompt includes an instruction and an output format instruction. The instruction includes, "Please list my personality traits in bullet points. My MBTI diagnosis result is INTP-T, so please list 10 personality traits in the following output format." This is instruction information for generating an answer in the LLM 200. The personality traits are preference and taste data that indicate the user's behavioral tendencies and preferences / thought tendencies. This content is an example, and the prompt includes information based on MBTI and inferring personality. While the number of inferences is 10, it may, of course, be less or more. The output format indicates a sentence.

[0037] Figure 11(b) shows a user personality generated by the LLM 200 based on this prompt. As shown in the figure and as follows, ten personality traits for the user are listed.

[0038] (Example output) - Prefers logical and objective thinking. - Interested in exploring ideas and considering new concepts and theories. - Active in acquiring knowledge, and may immerse themselves deeply in specialized fields. - Has a calm and unemotional side, and tends to prioritize logic over emotion. - Strongly independent, and values ​​their own identity. - Flexible and open to new information. - Creative, and able to take an original approach to problem-solving. - May prefer quiet and calm environments to social situations. - Has a sense of humor, and may have an abstract and sarcastic side. - Able to accept unpredictable situations and uncertainty, and deal with them flexibly.

[0039] FIG. 12 illustrates the generation of preference and thought data using survey results in the preference and thought data generation unit 106. FIG. 12(a) illustrates an example of a prompt output to the LLM 200. As shown in the figure, this prompt includes instructions, survey results, and output format instructions. The instructions include, "Please tell me 10 of my personality traits that can be determined from the following in the following output format." This is instruction information for generating a response in the LLM 200. The personality traits indicate the user's behavioral tendencies and preferences / thought tendencies. This content is an example, and the prompt includes information based on the survey results and inferring personality. While the number of inferences is 10, it is naturally possible to include fewer or more. The output format is shown as text. While the figure shows two survey results, it is preferable to include all survey results.

[0040] Figure 12(b) shows a user personality generated by the LLM 200 based on this prompt. As shown in the figure and below, ten personality traits for the user are listed.

[0041] (Example output) - When pursuing entertainment, you seem to value good sound and a good screen, and enjoy going to the cinema. - Owning an inspiring film allows you to enjoy it whenever you want, and you appear to be willing to make a purchase. - It seems that you place importance on pursuing high-quality experiences and collections in your hobbies and entertainment. - You will find joy in building rich experiences through movies and incorporating them into your daily life. - You may prioritize your sensitivity and emotions when it comes to finding excitement and joy. - Your pursuit of the immersive experience and excitement of a movie theater suggests that you are open to new discoveries and experiences. - You will find fulfillment by following your interests and curiosity and pursuing the good stuff. - You value the passion and excitement you feel through movies, and this has a positive impact on your life. - Inspiring moments and owning good films likely bring joy and excitement to your daily life. - You seek high quality and satisfaction in your hobbies and entertainment, which appears to improve your quality of life.

[0042] The preference and thought data generating unit 106 merges the estimated personality traits (preference and thought data) shown in FIGS. 10 to 12 and stores them as one list (preference and thought data) in the list DB.

[0043] The prompt generation unit 108 is a part that generates prompts to make decisions based on and combine the above content and output the basis for that decision. That is, the prompt generation unit 108 receives input of the content of the decision to be made from the user terminal 300, and acquires decision-making data and preference / thought data linked to the user ID. The prompt generation unit 108 then generates a prompt by formatting the input in accordance with a predetermined format.

[0044] Figure 13 is a diagram showing an example format of a prompt generated by the prompt generation unit 108. The prompt generated by the prompt generation unit 108 includes a request sentence (fixed format), constraints (fixed format), input / output examples, the user's (my) personality (user's preferences and thoughts), output format (fixed format), and output (matters on which a decision is desired).

[0045] The request text and constraints are in a predetermined fixed format. The request text is the content of a user's request for a decision, and is information indicating that the user wants the decision to be made on behalf of the user's avatar.

[0046] The input / output example is information based on decision-making data, and is information based on decision-making data in the decision-making DB that corresponds to the user ID transmitted from the user terminal 300. In the present disclosure, the decision-making time data is based on the results of a questionnaire.

[0047] "My personality (user's preferences and thoughts)" is information based on preference and thought data corresponding to the user ID.

[0048] The output format is a predetermined fixed format that indicates the format of the output. This output format includes the decision-making matter, the decision-making results, the decision-making rationale, and "my personality" that contributed to the decision-making.

[0049] The output (the matter to be decided) is information based on the matter to be decided that is input from the user terminal 300, and is in accordance with the output format. By operating the user terminal 300, the user can input the matter to be decided in text.

[0050] 14 and 15 are diagrams illustrating prompts generated by the prompt generation unit 108 according to the prompt format. The request text is, "You are my alter ego. Please tell me the reasons for making a decision on my behalf and the basis for reaching that decision. Please output the output format based on the input / output examples and my personality, taking into account the constraints." The constraints are, "The decision-making basis must be 100 characters or more," and "For 'What of my personality contributed to the decision?', select one of the listed items." In this disclosure, the constraints are listed in bullet points. This column contains "# constraint," and the LLM 200 determines the constraints based on this entry. The same applies to the input / output examples and my personality. The "My Personality" column is omitted from FIG. 14 for ease of illustration, but an example is shown in FIG. 15.

[0051] The output includes a request sentence, "Things I want to decide on: I originally planned to eat lunch when I got home, but I started to feel hungry while walking around the shopping district. What should I do?" This sentence is input from the user terminal 300. Of the decision-making results, the decision-making basis, and my personality, those that contributed to the decision-making are in a fixed format and are predetermined items.

[0052] 14(b) shows the answers generated by the LLM 200 in response to the prompts. The following answers are generated: the matter I want to decide, the decision result, the reason for the decision, and my personality traits that contributed to the decision.

[0053] (Example output) ・Decision to be made: I had originally planned to have lunch when I got home, but I started to feel hungry while walking through the shopping district. What should I do? ・Decision result: Have lunch in the shopping district ・Rationale for decision: I thought that eating in a new place would help me break the monotony of my daily routine and refresh myself. I also think that enjoying the opportunity to eat out can help relieve stress. ・Part of my personality that contributed to the decision: I am flexible and open to new information.

[0054] The preference and thought data updating unit 107 is a part that provides feedback on whether the user is satisfied with the output result of the LLM 200, and updates the list DB stored in the preference and thought data generating unit 106. Figure 16 is a diagram showing a specific example of the update process. Figure 16(a) is a diagram showing the process when the user is not satisfied with the output result of the LLM 200. Figure 16(b) is a diagram showing the process when the user is satisfied with the output result of the LLM 200.

[0055] 16(a), when the preference and thought data updating unit 107 receives information from the user terminal 300 that the output result of the LLM 200 is not satisfactory, the preference and thought data generating unit 106 deletes the preference and thought data picked up in "What is in my personality that contributed to decision-making" stored in the list DB. This is because it can be interpreted that the accuracy of the preference and thought data (personality characteristics) is low. On the other hand, when the preference and thought data updating unit 107 receives information from the user terminal 300 that the output result of the LLM 200 is satisfactory, the preference and thought data generating unit 106 does not perform the deletion process.

[0056] Furthermore, the preference and thought data updating unit 107 adds preference and thought data (personality traits) inferred from the current decision-making to the list DB as new preference and thought data. For example, the preference and thought data updating unit 107 may cause the LLM 200 to output new preference and thought data based on the decision-making data shown in Fig. 12 and add the new preference and thought data to the list DB.

[0057] For example, in the example in Figure 14, the following information was obtained. These are the same items as in the questionnaire results. - Matter to decide: I had originally planned to have lunch when I got home, but I started to feel hungry while in the shopping district. What should I do? - Decision result: Have lunch in the shopping district - Decision rationale: I thought that eating in a new place would help me break the monotony of everyday life and refresh myself. I also think that enjoying the opportunity to eat out can help relieve stress.

[0058] The preference and thought data updating unit 107 instructs the preference and thought data generating unit 106 to add new preference and thought data to the list DB, and generates a prompt using the output result. The preference and thought data generating unit 106 causes the LLM 200 to output new preference and thought data, and performs processing to add the new preference and thought data to the list DB of the preference and thought data generating unit 106.

[0059] With this configuration, the prompt generation device 100 can make decisions based on the user's preference and thought data and provide the reasons for those decisions.

[0060] Next, the operation of the prompt generation device 100 configured as described above will be described. Fig. 17 is a flowchart showing each model operation of the prompt generation device 100 according to the present disclosure. A pre-operation (model data construction) is initiated according to the pre-operation conditions (for example, time or instructions from the administrator). First, the personality data acquisition unit 101, decision-making data acquisition unit 102, and user information acquisition unit 103 acquire their respective data (S101 to S103).

[0061] Then, the personality data acquiring unit 101 and the user information acquiring unit 103 output the acquired data to the personality data estimating unit 104 (S104 and S105).

[0062] The personality data estimation unit 104 constructs a personality estimation model based on this information and estimates the personality of the unknown user using the constructed personality estimation model (S106). The personality data estimation unit 104 stores the estimated personality of the unknown user in a personality DB and updates it (S107).

[0063] Meanwhile, the decision-making data estimation unit 105 acquires data from the decision-making data acquisition unit 102 and the user information acquisition unit 103 (S108 and S109). The decision-making data estimation unit 105 then constructs a decision-making estimation model and estimates the results of a survey regarding the decision-making of the unknown user using the constructed decision-making estimation model (S110). The decision-making data estimation unit 105 updates the decision-making DB of the decision-making data acquisition unit 102 by adding the estimated survey results (S111).

[0064] Furthermore, the preference and thought data generation unit 106 acquires data from the personality data acquisition unit 101, the decision-making data acquisition unit 102, and the user information acquisition unit 103, and constructs preference and thought data (S112 and S113).

[0065] In this way, unknown personality data, unknown decision-making data, and preference / thinking data can be generated. Even for users whose personality data and decision-making data are unknown, the personality data and decision-making data can be estimated by constructing various prediction models.

[0066] Next, a prompt generation process for causing the LLM 200 to perform a decision process using the estimated personality data, decision-making data, and preference / thought data will be described. Figure 18 is a flowchart showing this process.

[0067] In the present disclosure, the prompt generation unit 108 starts operating when a request sentence is input and sent by the user of the user terminal 300. When the request receiving unit 108a receives a request from the user terminal 300, the prompt generation unit 108 acquires preference and thought data (user behavioral tendencies) and decision-making data (survey results) from the decision-making data acquisition unit 102 and the preference and thought data generation unit 106 (S114). The prompt generation unit 108 then generates a prompt based on the decision-making data and preference and thought data, and sends it to the LLM 200 (S115).

[0068] The prompt generation unit 108 obtains the answer from the LLM 200 and transmits it to the user terminal 300 (S116). The prompt generation unit 108 obtains from the user terminal 300 the result of the judgment on the validity of the answer notified to the user terminal 300 and transmits the result of the judgment to the preference and thought data update unit 107 (S117). The preference and thought data update unit 107 transmits the result of the user's judgment to the preference and thought data generation unit 106 (S118) and updates the list DB of the preference and thought data generation unit 106 (S119).

[0069] Next, the effects of the prompt generation device 100 of the present disclosure will be described. In the prompt generation device 100 of the present disclosure, the request receiving unit 108a receives request information from a user. The prompt generation unit 108 functions as a request information acquisition unit, a tendency information acquisition unit, and a generation unit. The prompt generation unit 108 receives information about the user (user-related information) and the request information from the user received by the request receiving unit 108a. The prompt generation unit 108 then receives the user's preference and thought data (behavioral tendency information) from the list DB of the preference and thought data generation unit 106. Based on the preference and thought data (behavioral tendency information), the prompt generation unit 108 generates a prompt for outputting a decision-making result based on the request information.

[0070] This makes it possible to generate prompts that allow users to make decisions based on their preferences and thought data, thereby enabling appropriate support for users in making decisions.

[0071] In the present disclosure, the prompt generation unit 108 may generate a prompt by further adding decision-making data including the user's decision result and the basis for that decision. This decision-making data is based on questionnaire questions and their answers and represents an example of a user's decision-making process. In the present disclosure, the prompt includes the decision-making data as an example of input and output.

[0072] This allows prompts to be generated that prompt the LLM 200 to respond appropriately.

[0073] In the present disclosure, the prompt includes requesting the rationale for the decision-making result, which allows the user to know the reason for the result and feel satisfied with the result.

[0074] In the present disclosure, the prompt includes an instruction to request "what part of my personality contributed to the decision-making" from the behavioral tendency information as the behavioral tendency information that contributed to the decision-making result. This allows the user to know the preference and thought data (behavioral tendency) of the user who obtained the result, and to feel satisfied with the result.

[0075] In the present disclosure, user preference and thought data (behavioral tendency information) is generated based on user-related information, which is at least one of the user's personality, decision-making data (survey results) including the user's judgment results and the reasons for them, and the user's attribute information.

[0076] In the present disclosure, the prompt generation unit 108 sends a prompt to a generative AI model and obtains an answer corresponding to the prompt from the generative AI model.

[0077] In the present disclosure, the prompt generation unit 108 accesses a list DB that stores preference and thought data (behavioral tendency information) to acquire the preference and thought data (behavioral tendency information). The preference and thought data update unit 107 updates the preference and thought data (behavioral tendency information) in the list DB based on the answer. Here, the user views the answer from the LLM 200 and judges its validity. The preference and thought data update unit 107 updates the list DB based on the user's judgment of the validity of the answer.

[0078] This allows the information in the list DB based on an inappropriate answer to be changed, so that from the next time onwards, the user will no longer receive an answer that does not satisfy him or her.

[0079] For example, if the answer is not determined to be appropriate for the user, the preference and thought data updating unit 107 removes the preference and thought data (behavioral tendency information) included in the answer from the list DB, so that the preference and thought data selected here will not be selected from the next time onwards.

[0080] If the answer is judged to be appropriate for the user, the preference and thought data updating unit 107 uses the decision-making data included in the prompt to acquire new preference and thought data (behavioral tendency information), which makes it possible to select new preference and thought data from the next time onwards, thereby obtaining a more convincing answer.

[0081] The device and method of the present disclosure have the following configuration.

[0082] [1] A device comprising: a request information acquisition unit that acquires request information of a user; a tendency information acquisition unit that acquires behavioral tendency information (preference / thought data: my personality) of the user; and a generation unit that generates a prompt for outputting a decision-making result for the request information based on the behavioral tendency information.

[0083] [2] The device according to [1], wherein the generation unit generates the prompt by further adding decision-making data including a judgment result and a reason for the judgment result by the user.

[0084] [3] The device according to [1] or [2], wherein the prompt includes a request for reasons that led to the decision-making result.

[0085] [4] The device according to [3], wherein the prompt includes an instruction to request, from the behavioral tendency information, behavioral tendency information that contributed to the decision-making result.

[0086] [5] The device described in any one of [1] to [4], wherein the user's behavioral tendency information is generated based on the user's personality, decision-making data (survey results) including the user's judgment results and the reasons for them, and user-related information which is at least one of the user's attribute information.

[0087] [6] The device according to any one of [1] to [5], wherein the generation unit transmits the prompt to a generative AI model and obtains an answer corresponding to the prompt from the generative AI model.

[0088] [7] The device described in [6], wherein the tendency information acquisition unit accesses a list DB that stores the behavioral tendency information to acquire the behavioral tendency information, and further includes an update unit that updates the behavioral tendency information in the list DB based on the response.

[0089] [8] The device according to [7], wherein the update unit excludes behavioral tendency information included in the answer from the list DB when the answer is not determined to be appropriate for the user.

[0090] [9] The device according to [7] or [8], wherein the update unit, if the answer is determined to be appropriate for the user, uses decision-making data included in the prompt to obtain new behavioral tendency information.

[0091]

[10] A method comprising: a request information acquisition step of acquiring request information of a user; a tendency information acquisition step of acquiring behavioral tendency information of the user; and a generation step of generating a prompt for outputting a decision-making result for the request information based on the behavioral tendency information.

[0092] The block diagrams used to explain the above embodiments show functional blocks. These functional blocks (components) are realized by any combination of hardware and / or software. Furthermore, the method for realizing each functional block is not particularly limited. That is, each functional block may be realized using a single device that is physically or logically coupled, or may be realized using two or more physically or logically separated devices that are connected directly or indirectly (e.g., via wire, wirelessly, etc.) and these multiple devices. The functional block may also be realized by combining the single device or multiple devices with software.

[0093] Functions include, but are not limited to, judgment, determination, assessment, calculation, computation, processing, derivation, investigation, search, confirmation, reception, transmission, output, access, resolution, selection, selection, establishment, comparison, assumption, expectation, consideration, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating, mapping, and assignment. For example, a functional block (component) that performs transmission is called a transmitting unit or transmitter. As mentioned above, there are no particular limitations on how these functions are implemented.

[0094] For example, the prompt generation device 100 according to an embodiment of the present disclosure may function as a computer that performs processing of the prompt generation method of the present disclosure. Fig. 19 is a diagram illustrating an example of the hardware configuration of the prompt generation device 100 according to an embodiment of the present disclosure. The prompt generation device 100 described above may be physically configured as a computer device including a processor 1001, a memory 1002, a storage device 1003, a communication device 1004, an input device 1005, an output device 1006, a bus 1007, and the like.

[0095] In the following description, the term "device" may be interpreted as a circuit, a device, a unit, etc. The hardware configuration of the prompt generation device 100 may be configured to include one or more of the devices shown in the figures, or may be configured to exclude some of the devices.

[0096] Each function in the prompt generating device 100 is realized by loading specified software (programs) onto hardware such as the processor 1001 and memory 1002, causing the processor 1001 to perform calculations, control communication via the communication device 1004, and control at least one of reading and writing data in the memory 1002 and storage 1003.

[0097] The processor 1001, for example, runs an operating system to control the entire computer. The processor 1001 may be configured as a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic unit, a register, etc. For example, the decision-making data estimation unit 105, the thought / thought data generation unit 106, the preference / thought data update unit 107, the prompt generation unit 108, etc. may be realized by the processor 1001.

[0098] The processor 1001 also loads programs (program code), software modules, data, etc. from at least one of the storage 1003 and the communication device 1004 into the memory 1002 and executes various processes in accordance with these programs. The programs used may be programs that cause a computer to execute at least some of the operations described in the above-described embodiments. For example, the decision-making data estimation unit 105, the thought / thought data generation unit 106, the preference / thought data update unit 107, and the prompt generation unit 108 may be implemented by a control program stored in the memory 1002 and running on the processor 1001, and similar implementations may be used for other functional blocks. While the above-described various processes have been described as being executed by a single processor 1001, they may also be executed simultaneously or sequentially by two or more processors 1001. The processor 1001 may be implemented on one or more chips. The programs may also be transmitted over a network via a telecommunications line.

[0099] The memory 1002 is a computer-readable recording medium and may be configured, for example, by at least one of a read-only memory (ROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), a random access memory (RAM), etc. The memory 1002 may also be referred to as a register, a cache, a main memory (primary storage device), etc. The memory 1002 may store executable programs (program codes), software modules, etc. for implementing a prompt generation method according to one embodiment of the present disclosure.

[0100] Storage 1003 is a computer-readable recording medium, and may be composed of at least one of, for example, an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (e.g., a compact disk, a digital versatile disk, a Blu-ray (registered trademark) disk), a smart card, a flash memory (e.g., a card, a stick, a key drive), a floppy (registered trademark) disk, a magnetic strip, etc. Storage 1003 may also be referred to as an auxiliary storage device. The above-mentioned storage medium may be, for example, a database, a server, or other appropriate medium including at least one of memory 1002 and storage 1003.

[0101] The communication device 1004 is hardware (transmission / reception device) for communicating between computers via at least one of a wired network and a wireless network, and is also referred to as, for example, a network device, a network controller, a network card, or a communication module. The communication device 1004 may be configured to include a high-frequency switch, a duplexer, a filter, a frequency synthesizer, etc. to realize at least one of frequency division duplex (FDD) and time division duplex (TDD). For example, the request reception unit 108a described above may be realized by the communication device 1004. The communication device 1004 may be implemented with a transmitter and a receiver that are physically or logically separated.

[0102] The input device 1005 is an input device (e.g., a keyboard, a mouse, a microphone, a switch, a button, a sensor, etc.) that receives input from the outside. The output device 1006 is an output device (e.g., a display, a speaker, an LED lamp, etc.) that outputs to the outside. Note that the input device 1005 and the output device 1006 may be integrated into one device (e.g., a touch panel).

[0103] Furthermore, each device, such as the processor 1001 and the memory 1002, is connected by a bus 1007 for communicating information. The bus 1007 may be configured using a single bus, or may be configured using different buses between each device.

[0104] Furthermore, prompt generation device 100 may be configured to include hardware such as a microprocessor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a programmable logic device (PLD), or a field programmable gate array (FPGA), and some or all of the functional blocks may be realized by such hardware. For example, processor 1001 may be implemented using at least one of these pieces of hardware.

[0105] The notification of information is not limited to the aspects / embodiments described in the present disclosure and may be performed using other methods. For example, the notification of information may be performed by physical layer signaling (e.g., Downlink Control Information (DCI) and Uplink Control Information (UCI)), higher layer signaling (e.g., Radio Resource Control (RRC) signaling, Medium Access Control (MAC) signaling, broadcast information (Master Information Block (MIB) and System Information Block (SIB))), other signals, or a combination thereof. Furthermore, the RRC signaling may be referred to as an RRC message, and may be, for example, an RRC Connection Setup message, an RRC Connection Reconfiguration message, or the like.

[0106] The order of the procedures, sequences, flowcharts, etc. of each aspect / embodiment described in this disclosure may be changed unless it is consistent. For example, the methods described in this disclosure present elements of various steps using an example order, and are not limited to the particular order presented.

[0107] Input and output information may be stored in a specific location (for example, memory) or may be managed using a management table. Input and output information may be overwritten, updated, or added to. Output information may be deleted. Input information may be sent to another device.

[0108] The determination may be made based on a value represented by one bit (0 or 1), a Boolean value (true or false), or a numerical comparison (e.g., comparison with a predetermined value).

[0109] The aspects / embodiments described in this disclosure may be used alone, in combination, or switched depending on the implementation. Notification of predetermined information (e.g., notification that "X is true") is not limited to explicit notification, but may be implicit (e.g., not notifying the predetermined information).

[0110] Although the present disclosure has been described in detail above, it is clear to those skilled in the art that the present disclosure is not limited to the embodiments described herein. The present disclosure can be implemented in modified and altered forms without departing from the spirit and scope of the present disclosure as defined by the claims. Therefore, the description of the present disclosure is intended to be illustrative and does not have any limiting meaning on the present disclosure.

[0111] Software shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, etc., whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise.

[0112] Software, instructions, information, etc. may also be transmitted or received over a transmission medium. For example, if software is transmitted from a website, server, or other remote source using wired technologies (such as coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL)), and / or wireless technologies (such as infrared, microwave), then these wired and / or wireless technologies are included within the definition of transmission media.

[0113] The information, signals, etc. described in this disclosure may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.

[0114] Note that terms described in this disclosure and terms necessary for understanding this disclosure may be replaced with terms having the same or similar meanings. For example, at least one of a channel and a symbol may be a signal (signaling). Furthermore, a signal may be a message. Furthermore, a component carrier (CC) may be called a carrier frequency, a cell, a frequency carrier, etc.

[0115] Furthermore, the information, parameters, etc. described in the present disclosure may be expressed using absolute values, relative values ​​from a predetermined value, or other corresponding information. For example, a radio resource may be indicated by an index.

[0116] The names used for the above-described parameters are not intended to be limiting in any way. Furthermore, the mathematical expressions using these parameters may differ from those explicitly disclosed in this disclosure. The various channels (e.g., PUCCH, PDCCH, etc.) and information elements may be identified by any suitable names, and therefore the various names assigned to these various channels and information elements are not intended to be limiting in any way.

[0117] In this disclosure, the terms "Mobile Station (MS)," "user terminal," "User Equipment (UE)," "terminal," and the like may be used interchangeably.

[0118] A mobile station may also be referred to by those skilled in the art as a subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, or some other suitable terminology.

[0119] As used in this disclosure, the terms "determining" and "determining" may encompass a wide variety of actions. "Determining" and "determining" may include, for example, judging, calculating, computing, processing, deriving, investigating, looking up, searching, inquiring (e.g., searching in a table, database, or other data structure), ascertaining, and the like. "Determining" and "determining" may also include receiving (e.g., receiving information), transmitting (e.g., sending information), input, output, accessing (e.g., accessing data in memory), and the like. Furthermore, "judgment" and "decision" can include regarding resolving, selecting, choosing, establishing, comparing, etc. as having been "judged" or "decided." In other words, "judgment" and "decision" can include regarding some action as having been "judged" or "decided." Furthermore, "judgment (decision)" can be interpreted as "assuming," "expecting," "considering," etc.

[0120] The terms "connected," "coupled," or any variation thereof, refer to any direct or indirect connection or coupling between two or more elements, and may include the presence of one or more intermediate elements between two elements that are "connected" or "coupled" to each other. The coupling or connection between elements may be physical, logical, or a combination thereof. For example, "connected" may be read as "access." As used in this disclosure, two elements may be considered to be "connected" or "coupled" to each other using one or more wires, cables, and / or printed electrical connections, as well as electromagnetic energy having wavelengths in the radio frequency range, microwave range, and optical (both visible and invisible) range, as some non-limiting and non-exhaustive examples.

[0121] As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly stated otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

[0122] As used in this disclosure, any reference to an element using a designation such as "first," "second," etc. does not generally limit the quantity or order of those elements. These designations may be used in this disclosure as a convenient method of distinguishing between two or more elements. Thus, a reference to a first and a second element does not imply that only two elements may be employed or that the first element must in some way precede the second element.

[0123] When the terms "include," "including," and variations thereof are used in this disclosure, these terms are intended to be inclusive, similar to the term "comprising." Furthermore, when the term "or" is used in this disclosure, it is not intended to be an exclusive or.

[0124] In this disclosure, where articles are added by translation, such as a, an, and the in English, the disclosure may include that the nouns following these articles are in the plural form.

[0125] In the present disclosure, the term "A and B are different" may mean "A and B are different from each other." The term may also mean "A and B are each different from C." Terms such as "separate" and "coupled" may also be interpreted in the same way as "different."

[0126] 100...prompt generation device, 200...LLM, 300...user terminal, 101...personality data acquisition unit, 102...decision-making data acquisition unit, 103...user information acquisition unit, 104...personality data estimation unit, 105...decision-making data estimation unit, 106...thought data generation unit, 107...thought data update unit, 108...prompt generation unit, 108a...request reception unit.

Claims

1. An apparatus comprising: a request information acquisition unit that acquires request information of a user; a tendency information acquisition unit that acquires behavioral tendency information of the user; and a generation unit that generates a prompt for outputting a decision-making result for the request information based on the behavioral tendency information.

2. The device according to claim 1, wherein the generator generates the prompt by further adding decision-making data including the result of the user's decision and the basis for that decision.

3. The device of claim 1, wherein the prompt includes a request for reasons that led to the decision-making outcome.

4. The device according to claim 3, wherein the prompt includes an instruction to request, from the behavioral tendency information, behavioral tendency information that contributed to the decision-making result.

5. The device according to claim 1, wherein the user's behavioral tendency information is generated based on user-related information, which is at least one of the user's personality, decision-making data including the user's judgment results and the reasons for them, and the user's attribute information.

6. The device according to claim 1, wherein the generator transmits the prompt to a generative AI model and obtains an answer corresponding to the prompt from the generative AI model.

7. The device described in claim 6, wherein the tendency information acquisition unit accesses a list DB that stores the behavioral tendency information to acquire the behavioral tendency information, and further includes an update unit that updates the behavioral tendency information in the list DB based on the response.

8. The device according to claim 7, wherein the update unit excludes the behavioral tendency information included in the answer from the list DB if the answer is not determined to be appropriate for the user.

9. The device according to claim 7, wherein the update unit, if the answer is determined to be appropriate for the user, uses decision-making data included in the prompt to obtain new behavioral tendency information.

10. A method comprising: a request information acquisition step of acquiring request information of a user; a tendency information acquisition step of acquiring behavioral tendency information of the user; and a generation step of generating a prompt for outputting a decision-making result for the request information based on the behavioral tendency information.

Images