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

Abstract

Obtain responses that reproduce the variability in responses from multiple respondents. [Solution] The information processing device includes: an acquisition unit that acquires multiple attribute values ​​for each of multiple attributes of a person's profile, and a probability distribution of the multiple attribute values ​​for each of the multiple attributes; a determination unit that determines the attribute values ​​for each of the multiple attributes for each person's profile of a plurality of respondents based on the probability distribution; a generation unit that generates a prompt instructing each of the plurality of respondents to answer a question as a person's profile having the determined attribute values ​​for each of the multiple attributes; and a processing unit that inputs the generated prompt to a generating AI model for each of the plurality of respondents and acquires the respondent's answer.

Description

Technical Field

[0001] The technology of the present disclosure relates to an information processing apparatus, an information processing method, and an information processing program.

Background Art

[0002] Non-Patent Document 1 discloses the effectiveness when a generative AI replaces respondents in a marketing survey.

[0003] Non-Patent Document 2 discloses an experiment based on subjects simulated by a generative AI.

Prior Art Documents

Non-Patent Documents

[0004]

Non-Patent Document 1

Non-Patent Document 2

Non-Patent Document 3

[0005] Non-patent document 1, mentioned above, discloses that synthetic data using LLM (Large Language Models) closely matches real data in terms of mean values. However, the standard deviation is excessively small, and it cannot reproduce actual variability. Furthermore, GPT (Generative Pretrained Transformer) has low variability, fails to adequately reflect differences in responses such as gender, and tends to generate responses that are overly confident in certain attributes (e.g., high income).

[0006] Non-patent document 2, mentioned above, discloses that LLMs can simulate realistic human behavior based on economic scenarios. Furthermore, LLMs can model classical experiments in economics and behavioral science, and reproduce similar results. While thus capable of replicating human preferences, they tend to favor short-term rewards excessively. They also tend to exhibit lower entropy (diversity) in responses compared to human response data. The model's ability to fully reproduce complex economic behavior in the real world is insufficient.

[0007] The technology disclosed herein aims to provide an information processing device, an information processing method, and an information processing program that can obtain responses that reproduce the variability of responses from multiple respondents. [Means for solving the problem]

[0008] A first aspect of the technology of this disclosure is an information processing device that includes: an acquisition unit that acquires a plurality of attribute values ​​for each of a plurality of attributes of a person's profile, and a probability distribution of the plurality of attribute values ​​for each of the plurality of attributes; a determination unit that determines the attribute values ​​for each of the plurality of attributes for each person's profile of a plurality of respondents based on the probability distribution; a generation unit that generates a prompt instructing each of the plurality of respondents to answer a question as a person's profile having the determined attribute values ​​for each of the plurality of attributes; and a processing unit that inputs the generated prompt to a generating AI model for each of the plurality of respondents and acquires the respondent's answer.

[0009] A second aspect of the technology of this disclosure is an information processing method in which a computer performs the following actions: obtains a plurality of attribute values ​​for each of a plurality of attributes of a person's profile, and a probability distribution of the plurality of attribute values ​​for each of the plurality of attributes; determines the attribute values ​​for each of the plurality of attributes for each person's profile of a plurality of respondents based on the probability distribution; generates a prompt instructing each of the plurality of respondents to answer a question as a person's profile having the determined attribute values ​​for each of the plurality of attributes; and inputs the generated prompt into a generating AI model for each of the plurality of respondents to obtain the respondent's answer.

[0010] A third aspect of the technology of the present disclosure is an information processing program that obtains a plurality of attribute values ​​for each of a plurality of attributes of a person's profile, and a probability distribution of the plurality of attribute values ​​for each of the plurality of attributes, determines the attribute values ​​for each of the plurality of attributes for each of the person's profiles of a plurality of respondents based on the probability distribution, generates a prompt instructing each of the plurality of respondents to answer a question as a person's profile having the determined attribute values ​​for each of the plurality of attributes, and causes a computer to input the generated prompt into a generating AI model for each of the plurality of respondents to obtain the respondent's answer. [Effects of the Invention]

[0011] According to this disclosure, it is possible to obtain answers that reproduce the variability of responses from multiple respondents. [Brief explanation of the drawing]

[0012] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a user terminal. [Figure 3] The functional configuration of a specific processing unit of a data processing device is shown in general terms. [Figure 4] This diagram outlines an example of the operation flow of a specific process performed by a data processing device. [Modes for carrying out the invention]

[0013] Hereinafter, an example of an embodiment of the information processing device, information processing method, and information processing program relating to the technology of this disclosure will be described with reference to the attached drawings.

[0014] First, let's explain the terminology used in the following explanation.

[0015] In the following embodiments, the numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), or an APU (Accelerated Processing Unit), etc.

[0016] In the following embodiments, the numbered RAM (Random Access Memory) is a memory where information is temporarily stored and is used as a work memory by the processor.

[0017] In the following embodiments, the numbered storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, etc.

[0018] In the following embodiments, the numbered communication I / F (Interface) is an interface that includes a communication processor and an antenna, etc. The communication I / F manages communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark), etc.

[0019] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B". That is, "A and / or B" means that it may be only A, only B, or a combination of A and B. Also, in this specification, when expressing three or more matters connected by "and / or", the same concept as "A and / or B" is applied.

[0020] <Overview of the Embodiment>

[0021] Surveys and experiments targeting a plurality of agents (participants, subjects, respondents) generated using a generative AI model have been known to have an approximately similar average value, but the entropy of the responses is extremely small, and as a result, the standard deviation is excessively small, and it is almost impossible to reproduce the actual variation.

[0022] There is a technique of creating an approximate agent for each respondent based on an existing database and prompting the respondent group to answer a questionnaire or the like. However, in addition to the decreasing entropy, if the distribution of the population originally targeted by the respondents in the existing database and the important characteristics of the respondents and the population are different, there can be a major issue with external validity in the first place. From this, even if responses similar to the respondent group are obtained, generalization in other scenarios becomes extremely difficult.

[0023] On the other hand, while WEB questionnaire surveys have become easier to conduct using Google (registered trademark) forms and the like, when conducting a questionnaire survey, it cannot be redone, and significant time and costs are required, including the costs borne by the subjects and respondents.

[0024] Therefore, in this embodiment, important attributes for the respondent's profile are identified according to the purpose of the survey, and a probability distribution of attribute values ​​for each important attribute that approximates the desired respondent group is obtained. Then, the profile of each respondent is determined according to the probability distribution of attribute values ​​for each important attribute, a prompt is generated for each respondent instructing them to answer the question based on the determined profile, and the answer is obtained from the generating AI model.

[0025] <System Configuration>

[0026] Figure 1 shows an example of the configuration of the data processing system 10 according to the embodiment.

[0027] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a user terminal 14. An example of the data processing device 12 is a server. An example of the user terminal 14 is a personal computer, tablet, or smartphone.

[0028] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0029] The user terminal 14 comprises a computer 36, a receiving device 38, an output device 40, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The receiving device 38 and output device 40 are also connected to the bus 52.

[0030] The reception device 38 is equipped with a keyboard, mouse, etc., and accepts user input. The reception device 38 may also accept user input by touching an object (e.g., a pen or finger) via a touch panel, or by detecting the user's voice via a microphone. The control unit 46A transmits data indicating the received user input to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0031] The output device 40 includes a display 40A and a speaker 40B, and presents data to a person by outputting it in a form perceptible to the person (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46.

[0032] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0033] Figure 2 shows an example of the main functions of the data processing device 12 and the user terminal 14.

[0034] As shown in Figure 2, in the data processing device 12, specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "data processing program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0035] The storage 32 stores the generated AI model 58. The generated AI model 58 is used by the specific processing unit 290.

[0036] Generative AI Model 58 is a type of so-called generative AI (Artificial Intelligence). An example of Generative AI Model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The generative AI model 58 is obtained by performing deep learning on a neural network. The generative AI model 58 is input with prompts containing instructions, and inference data such as audio data representing speech, text data representing text, and image data representing images. The generative AI model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0037] At the user terminal 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0038] Next, we will describe the processing of the specific processing unit 290 when the data processing device 12 performs specific processing to obtain answers to the question from each of the virtual respondents.

[0039] As shown in Figure 3, the specific processing unit 290 includes an acquisition unit 291, a determination unit 292, a generation unit 293, a processing unit 294, and an analysis unit 295.

[0040] The acquisition unit 291 acquires multiple attribute values ​​for each of the multiple attributes of the hypothetical respondent's profile, and the probability distribution of the multiple attribute values ​​for each of the multiple attributes.

[0041] Specifically, attributes are important attributes in relation to the question text for a hypothetical respondent, and include information about age, gender, occupation, income, address, or family structure. The type of attribute and the attribute value for each attribute are determined by a human or computer. Alternatively, the type of attribute and the attribute value for each attribute may be determined using the generative AI model 58.

[0042] The probability distribution of multiple attribute values ​​for a given attribute is obtained based on pre-prepared statistical data, such as labor force statistics, retail price statistics, basic employment structure statistics, census, household expenditure statistics, basic social life statistics, population estimates, vital statistics, basic national life statistics, data from core systems such as the Basic Resident Register, citizen opinion surveys, and planning surveys (such as surveys for formulating child and childcare support plans). These are determined manually or by computer.

[0043] As an example, we obtain multiple attribute values ​​for each of the multiple attributes of a hypothetical respondent's profile, as well as the probability distribution of the multiple attribute values ​​for each of the multiple attributes, as shown below.

[0044] factor_A = { "husband": 0.5, "wife": 0.5} factor_B = { "Both spouses work for companies in Prefecture B": 0.1065, "Husband works for a company in Prefecture B, and wife works for a company in City A": 0.2130, "Both spouses work for companies in City A": 0.2130, "Husband works for a company in Prefecture B, and wife is a full-time housewife": 0.1633, "Husband works for a company in City A, and wife is a full-time housewife": 0.1089, "〇 households": 0.059} factor_C = { "One child is of elementary school age or older": 0.25, "One child is a preschooler": 0.25, "Two or more children are all of elementary school age or older": 0.25, "Two or more children are the youngest, and the youngest is a preschooler": 0.25} factor_D = { "Household income is high and comfortable": 0.2, "Household income is middle income and somewhat comfortable": 0.5, "Household income is middle income and not comfortable": 0.2, "Household income is low income and very difficult": 0.1} factor_E = { "Owned home / apartment": 0.5, "Rented home": 0.5} factor_F = { "Living in a convenient urban area": ​​0.241, "Living in a less convenient suburban area, but with abundant greenery and parks": 0.759}

[0045] In the example above, factor_A to factor_F are multiple attributes, and the probability distributions of the multiple attribute values ​​for those attributes are shown within the {}.

[0046] The acquisition unit 291 may further acquire, for each attribute value of one attribute, the weight of each attribute value of the other attribute, or an adjusted probability distribution of each attribute value of the other attribute, which is predetermined between two correlated attributes. For example, if there is a correlation between factor_B and factor_F, the following adjusted probability distribution may be acquired.

[0047] def adjust_factor_F(factor_B_key): if factor_B_key in ["Both spouses work for companies in Prefecture B", "The husband works for a company in Prefecture B, and the wife works for a company in City A"]: return {"Living in a convenient urban area": ​​0.8, "Living in a less convenient suburban area, but with plenty of greenery and parks": 0.2} elif factor_B_key in ["Both husband and wife work for companies in City A", "Husband works for a company in City A, and wife is a full-time housewife"]: return {"Live in a convenient urban area": ​​0.3, "Live in a less convenient suburban area, but with plenty of greenery and parks": 0.7} else: return factor_F

[0048] In the example above, adjust_factor_F(factor_B_key) is a function that returns the probability distribution of each attribute value of attribute factor_F, adjusted according to the attribute values ​​of attribute factor_B.

[0049] The decision unit 292 determines the attribute values ​​for each of the multiple attributes for each of the multiple hypothetical respondents' profiles based on the acquired probability distribution.

[0050] Specifically, the determination unit 292 repeatedly determines the attribute value of each of the multiple attributes for each hypothetical respondent, according to the probability distribution of each attribute value for that attribute.

[0051] The decision unit 292, when it finds that there are two correlated attributes, determines the attribute values ​​for each of the multiple attributes for each of the multiple respondents' personal profiles based on the probability distribution and the weighted or adjusted probability distribution.

[0052] The generation unit 293 generates prompts that instruct each of the multiple virtual respondents to answer the question as a person having attribute values ​​for each of the determined multiple attributes.

[0053] For example, the following prompt is generated, which includes a person profile named persona_prompts and a set of questions.

[0054] persona_prompts = { f”{a}-{b}-{c}-{d}-{e}-{f}”: f”As {a}, you are a married person living in City A, Japan, and you are {b}, {c}, {e}, {f}, and {d}. Please answer the following questions.” for a in factor_A for b in factor_B for c in factor_C for d in factor_D for e in factor_E for f in factor_F} Questions = [ "Overall, how happy do you usually feel? On a scale of 10 ("very happy") to 0 ("very unhappy"), what score would you rate yourself on? Please answer with only a score, such as "〇 points". Please consider the significant randomness of each answer and carefully consider your response so that the data distribution across attributes is wide enough to include everything from a maximum of 10 to a minimum of 0." ]

[0055] In the example above, a, b, c, d, e, and f represent the attribute values ​​for the determined attributes factor_A to factor_F.

[0056] When two correlated attributes are included, the persona_prompts included in the prompt will be represented as follows:

[0057] persona_prompts = {} for a, b, c, d, e in itertools.product(factor_A.items(), factor_B.items(), factor_C.items(), factor_D.items(), factor_E.items()): adjusted_factor_F = adjust_factor_F(b[0]) for f, f_prob in adjusted_factor_F.items(): prob = a[1] * b[1] * c[1] * d[1] * e[1] * f_prob if prob > 0: key = f”{a[0]}-{b[0]}-{c[0]}-{d[0]}-{e[0]}-{f}” value = f”As {a[0]}, you are a married {a[0]}, {b[0]}, {c[0]}, {e[0]}, {f}, {d[0]}, {a persona_prompts[key] = value

[0058] The processing unit 294 inputs the generated prompt for each of the multiple virtual respondents to the generating AI model 58, and obtains the virtual respondents' responses from the output of the generating AI model 58.

[0059] The analysis unit 295 aggregates the responses of multiple hypothetical respondents obtained by the processing unit 294. For example, if the response consists of multiple choices, the unit aggregates the percentage of respondents who selected each choice. If the response is a free-form text, a generative AI model may be used when aggregating the responses.

[0060] Next, the operation of the data processing system 10 will be explained.

[0061] An example of the specific processing flow by the data processing device 12 will be explained with reference to Figure 4. Note that the specific processing flow shown in Figure 4 is an example of a "data processing method" related to the technology of this disclosure. When the data processing device 12 receives from the user terminal 14 multiple attribute values ​​for each of the multiple attributes of the virtual respondent's profile, and the probability distribution of the multiple attribute values ​​for each of the multiple attributes, it executes the specific processing shown in Figure 4.

[0062] In step S100, the acquisition unit 291 acquires multiple attribute values ​​for each of the multiple attributes of the virtual respondent's profile, as well as the probability distribution of the multiple attribute values ​​for each of the multiple attributes, which are received from the user terminal 14.

[0063] In step S102, the decision unit 292 determines the attribute values ​​for each of the multiple attributes for each of the multiple hypothetical respondent profiles based on the acquired probability distribution.

[0064] In step S104, the generation unit 293 generates a prompt instructing each of the multiple virtual respondents to answer the question as a person having attribute values ​​for each of the determined multiple attributes.

[0065] In step S106, the processing unit 294 inputs the generated prompt for each of the multiple virtual respondents to the generating AI model 58 and obtains the virtual respondents' responses from the output of the generating AI model 58.

[0066] In step S108, the analysis unit 295 aggregates the responses of multiple virtual respondents obtained by the processing unit 294, displays the aggregated results on the user terminal 14, and terminates the specific processing.

[0067] <Examples>

[0068] The results of an experiment conducted to confirm the effectiveness of the method of this embodiment described above will now be explained.

[0069] An experiment was conducted in which participants answered the following question: "Overall, how happy do you usually feel? On a scale of 10 ("very happy") to 0 ("very unhappy"), the answer would be approximately 100%. For comparison, two methods were used: one involving multiple actual participants, and another using a simple-random distribution.

[0070] The aggregated results of the responses are shown in Table 1. As shown in Table 1, it was found that in this embodiment, aggregated results closer to those obtained by having multiple actual subjects respond, compared to the method using a simple-random distribution.

[0071] [Table 1]

[0072] As described above, according to the data processing system of this embodiment, a probability distribution of multiple attribute values ​​is obtained, and based on the probability distribution, the attribute values ​​for each of the multiple attributes are determined for each of the multiple respondent's person profiles, a prompt is generated instructing each of the multiple respondent to answer the question as a person profile having the determined attribute values ​​for each of the multiple attributes, and the generated prompt is input to a generation AI model for each of the multiple respondent to obtain the respondent's answer. This makes it possible to obtain answers that reproduce the variability of answers from multiple respondents.

[0073] Furthermore, by creating a probability distribution that approximates the population of respondents and obtaining responses from a generated AI model using prompts that include randomly generated person profiles based on that distribution, the accuracy of the responses improves, and the standard deviation and entropy increase.

[0074] Furthermore, it is possible to experimentally obtain and compile responses before collecting actual answers to questions. For example, the present invention can be applied to experiments on citizens regarding infectious disease prevention behavior, experiments on environmentally conscious behavior (whether or not redelivery is used and whether or not parcel lockers are used), budget decision-making regarding environmentally conscious behavior by local government officials and corporate managers, and experiments on the consolidation of junior high schools by administrative managers and citizens, in which the results of compiling responses from anticipated respondents can be obtained in advance.

[0075] Furthermore, the present invention can be applied to pre-evaluation of decision-making content that is highly irreversible. For example, the present invention can be applied to understanding consumer behavior targeting a specific population, conducting A / B testing of products and services targeting a specific population, and understanding the sensitivity of residents and stakeholders in policy choices in government, thereby allowing for the prior understanding of aggregated results from anticipated respondents.

[0076] Furthermore, the present invention can be applied to preliminary evaluations in situations where experiments are not practically feasible. For example, the present invention can be applied to experiments on the acceptance level of stakeholders in the event of mergers or closures of businesses, or to experiments on the acceptance level of employees regarding personnel-related policies, allowing for the prior assessment of the aggregated results of responses from anticipated respondents.

[0077] Furthermore, the data processing system according to this embodiment can replace parts of questionnaire surveys, monitor surveys, and monitor experiments.

[0078] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0079] In the above embodiment, the example described was obtaining answers to a question about feelings of happiness, but it is not limited to this. Answers to other types of questions may also be obtained.

[0080] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing method for the specific process may be used, which includes computer 22 and multiple other computers.

[0081] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0082] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[0083] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0084] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0085] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0086] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0087] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0088] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0089] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference. [Explanation of Symbols]

[0090] 10 Data Processing Systems 12 Data Processing Devices 14 User terminals 22 Computers 24 Databases 28 processors 32 storage 56 Specific Processing Program 58 Generative AI Models 290 Specific Processing Unit 291 Acquisition Department 292 Decision Section 293 Generation part 294 Processing Unit 295 Analysis Department< / url:>

Claims

1. An acquisition unit that acquires multiple attribute values ​​for each of the multiple attributes of a person's profile, and the probability distribution of the multiple attribute values ​​for each of the multiple attributes, Based on the aforementioned probability distribution, a determination unit determines the attribute value for each of the aforementioned multiple attributes with respect to the individual profile of each of the multiple respondents, A generation unit that generates prompts instructing each of the multiple respondents to answer the question as a person having attribute values ​​in each of the multiple attributes determined above, A processing unit that inputs the generated prompt to a generating AI model for each of the aforementioned multiple respondents and obtains the respondent's answer, Information processing device including

2. The acquisition unit further acquires, for each of the attribute values ​​of one attribute, the weight of each attribute value of the other attribute, or the adjusted probability distribution of each attribute value of the other attribute, which are predetermined between two correlated attributes. The information processing apparatus according to claim 1, wherein the determination unit determines the attribute value for each of the multiple attributes with respect to each of the person profiles of the multiple respondents, based on the probability distribution and the weight or the adjusted probability distribution.

3. The information processing apparatus according to claim 1, further comprising an analysis unit for aggregating the responses of the plurality of respondents obtained by the processing unit.

4. The information processing apparatus according to claim 1, wherein the attributes include age, gender, occupation, income, address, or family structure.

5. Obtain multiple attribute values ​​for each of the multiple attributes of the person's profile, and the probability distribution of the multiple attribute values ​​for each of the multiple attributes. Based on the aforementioned probability distribution, the attribute values ​​for each of the aforementioned attributes are determined for each of the personal profiles of the multiple respondents. A prompt is generated for each of the aforementioned multiple respondents, instructing them to answer the question as a person having the attribute values ​​for each of the determined multiple attributes. For each of the aforementioned respondents, the generated prompt is input into the generating AI model to obtain the respondent's answer. A method of information processing performed by a computer.

6. Obtain multiple attribute values ​​for each of the multiple attributes of the person's profile, and the probability distribution of the multiple attribute values ​​for each of the multiple attributes. Based on the aforementioned probability distribution, the attribute values ​​for each of the aforementioned attributes are determined for each of the personal profiles of the multiple respondents. A prompt is generated for each of the aforementioned multiple respondents, instructing them to answer the question as a person having the attribute values ​​for each of the determined multiple attributes. For each of the aforementioned respondents, the generated prompt is input into the generating AI model to obtain the respondent's answer. An information processing program that causes a computer to perform a task.

Images