Survey management server, survey analysis method, and program

Abstract

This provides a survey management server that can easily acquire profile data showing the tendencies of targets that meet specific criteria. [Solution] The survey management server includes an extraction unit that compares attribute information of a target identified by a first condition with attribute information of a comparison target identified by a second condition from multiple monitors to extract the characteristics of the target; a prompt generation unit that generates prompts for input into a generation AI model based on the extracted characteristics of the target; and a profile data acquisition unit that acquires profile data representing the target generated by the generation AI model by inputting the generated prompts into the generation AI model.

Description

Technical Field

[0001] The present invention relates to a questionnaire management server, a questionnaire analysis method, and a program.

Background Art

[0002] As a method for grasping the tendencies of marketing targets, there is a method of conducting a questionnaire on a plurality of respondents (monitors) and analyzing the response results. Patent Document 1 describes a marketing support system that aggregates each respondent by brand usage frequency and by attribute based on the questionnaire response results and the attributes of the respondents.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In order to effectively support marketing for clients, it is desirable to analyze the tendencies of targets (respondents) that meet certain conditions from the questionnaire response results or attribute information, and create and provide a text showing the profile of the targets, rather than just aggregate values. However, in the prior art, such work has been carried out manually, which is very time-consuming.

[0005] Therefore, one of the objectives is to provide a questionnaire management server, a questionnaire analysis method, and a program that can easily obtain profile data showing the tendencies of targets that meet specific conditions.

Means for Solving the Problems

[0006] According to one aspect of this disclosure, the survey management server includes: an extraction unit that extracts characteristics of a target by comparing attribute information of a target identified by a first condition with attribute information of a comparison target identified by a second condition from a plurality of monitors; a prompt generation unit that generates prompts for input to a generation AI model based on the characteristics of the target extracted by the extraction unit; and a profile data acquisition unit that acquires profile data representing the target generated by the generation AI model by inputting the prompts generated by the prompt generation unit into the generation AI model.

[0007] According to one aspect of the present disclosure, the questionnaire analysis method includes the steps of: extracting characteristics of a target by comparing attribute information of a target identified by a first condition with attribute information of a comparison target identified by a second condition from a plurality of monitors; generating prompts for input into a generative AI model based on the extracted characteristics of the target; and obtaining profile data representing the target generated by the generative AI model by inputting the generated prompts into the generative AI model.

[0008] According to one aspect of this disclosure, the program causes the survey management server to perform the following steps: compare attribute information of a target identified by a first condition with attribute information of a comparison target identified by a second condition from a plurality of monitors to extract the characteristics of the target; generate a prompt for input to a generating AI model based on the extracted characteristics of the target; and obtain profile data representing the target generated by the generating AI model by inputting the generated prompt to the generating AI model. [Effects of the Invention]

[0009] According to the above embodiment, profile data showing the tendencies of targets that meet specific conditions can be easily obtained. [Brief explanation of the drawing]

[0010] [Figure 1] This figure shows an example of a questionnaire system according to the present invention. [Figure 2] This flowchart shows an example of processing by the survey management server according to the embodiment. [Figure 3] This diagram illustrates an example of the processing of the survey management server and the generated AI model according to the embodiment. [Figure 4] The first figure shows an example of the profile analysis results according to the embodiment. [Figure 5] The second figure shows an example of the profile analysis results according to the embodiment. [Figure 6] This flowchart shows an example of processing by a questionnaire management server according to a modified embodiment. [Figure 7] This is a schematic block diagram showing the configuration of a computer according to the embodiment. [Modes for carrying out the invention]

[0011] <Embodiment> (System Configuration) Figure 1 is a diagram showing an example of a survey system according to an embodiment. As shown in the figure, the survey system 1 includes respondent terminals 10a, 10b, ..., a client terminal 20, a survey management server 30, a big data DB (database) 40, and an external server 50. The respondent terminals 10a, 10b, etc. are communicated with the survey management server 30. The client terminal 20 is communicated with the survey management server 30. The big data DB 40 is communicated with the survey management server 30. The external server 50 is communicated with the survey management server 30.

[0012] Respondent terminals 10a, 10b, etc. are terminal devices such as smartphones and PCs (personal computers) used by survey respondents (hereinafter, respondents and potential respondents may be referred to as monitors) when answering the survey. Since survey respondents are selected for each survey and the number of respondents differs for each survey, respondent terminals 10a, 10b, etc. may change from survey to survey, but here they will be referred to as respondent terminals 10a, 10b, etc. for convenience. Figure 1 shows two respondent terminals 10a, etc., but there may be three or more respondent terminals 10a, etc.

[0013] The client terminal 20 is a terminal device such as a smartphone or PC (personal computer) used by companies that request the implementation and analysis of surveys. For example, a company developing a product targeting a specific user group might specify the target user's age, gender, place of residence, hobbies, and preferences, and create questions that can be used for product development and sales. The client terminal 20 sends survey request information, including the survey target conditions and question content, to the survey management server 30, requesting the survey to be conducted. The client terminal 20 also receives the survey results from the survey management server 30. Figure 1 shows one client terminal 20, but since various companies may request the implementation of surveys, there may be multiple client terminals 20. Companies may also request only the analysis of surveys they have conducted themselves from the survey management server 30.

[0014] The survey management server 30 is a server device operated by a research company that conducts and analyzes surveys at the request of companies and other organizations. For example, a research company has members who are willing to answer various types of surveys. The survey management server 30 extracts monitors from the members for each survey and conducts the survey targeting the extracted monitors. For example, members who answer the survey may be awarded points that can be used for online shopping, etc. The survey management server 30 receives survey requests, conducts and compiles surveys, outputs survey results, and performs profile analysis of the target audience based on the survey results. At this time, the survey management server 30 analyzes the characteristics of the target monitors and has an AI (artificial intelligence) model generate a summary of the analysis results. The survey management server 30 comprises an input reception unit 31, a survey processing unit 32, an extraction unit 33, a prompt generation unit 34, a profile data acquisition unit 35, an output unit 36, and a storage unit 37.

[0015] The input reception unit 31 receives requests for the implementation of surveys and for the analysis of target profiles. The input reception unit 31 receives survey implementation request information from the client terminal 20, including the conditions for the survey participants, the content of the questions, the survey implementation period, the deadline, etc. The input reception unit 31 may also receive analysis request information from the client terminal 20, including the conditions for the monitors to be analyzed (target) (first condition), the conditions for the monitors to be compared (second condition), etc. If the input reception unit 31 only receives requests from companies for the analysis of survey results, the input reception unit 31 may also receive the survey results from the client terminal 20, including the questions and answers of the survey conducted by the company.

[0016] The survey processing unit 32 performs tasks such as creating survey information based on the questions included in the request information, extracting monitors based on the target conditions for the survey included in the request information, distributing the survey information to the extracted monitors, collecting survey responses, and registering them in the survey database 372.

[0017] The extraction unit 33 compares the attribute information of the target specified by the first condition and the attribute information of the comparison target specified by the second condition from a plurality of monitors, and extracts the characteristics of the target.

[0018] The prompt generation unit 34 generates a prompt for input to the generation AI model 53 based on the characteristics of the target extracted by the extraction unit 33. As shown in FIG. 1, the generation AI model 53 includes a natural language model 51 that generates text based on the input prompt and an image generation AI model 52 that generates an image based on the input prompt. The prompt generation unit 34 generates a prompt for input to the natural language model 51 and a prompt for input to the image generation AI model, respectively. In the present embodiment, an example in which the generation AI model 53 has the natural language model 51 and the image generation AI model will be described, but the present invention is not limited thereto. In other embodiments, the generation AI model 53 may be one model that generates both text and images based on the input prompt. In this case, the prompt generation unit 34 generates a prompt for generating both a profile sentence and a profile image.

[0019] The profile data acquisition unit 35 acquires profile data representing the target generated by the generation AI model by inputting the prompt generated by the prompt generation unit 34 to the generation AI model 53. The profile data includes a profile sentence that represents the characteristics of the target in text and a profile image that represents the characteristics of the target in an image. In the example of FIG. 1, the profile data acquisition unit 35 acquires the profile sentence generated by the natural language model 51 based on the prompt. Further, when the prompt generation unit 34 generates a prompt for the profile image and inputs it to the image generation AI model 52, the profile data acquisition unit 35 acquires the profile image generated by the image generation AI model 52 based on this prompt.

[0020] The output unit 36 ​​outputs the survey results and the target profile analysis results to the client terminal 20.

[0021] The memory unit 37 stores various data. For example, the memory unit 37 stores the monitor DB 371 and the survey DB 372. The monitor DB 371 stores basic attribute information of members. For example, the monitor DB 371 stores demographic information such as age and gender, marketing information such as hobbies, preferences and tendencies, points earned by members, and member rank, all linked to the monitor ID. The survey DB 372 stores the response results for each monitor for each survey. For example, the survey DB 372 stores the monitor ID and the answers to each question of the survey, linked to the survey ID. In the configuration example in Figure 1, the monitor DB 371 and the survey DB 372 are stored in the memory unit 37 of the survey management server 30, but these DBs may also be stored in an external storage device (for example, a big data DB 40) of the survey management server 30.

[0022] Big Data DB 40 stores survey results, such as large-scale surveys conducted independently by research companies on their members, linked to monitor IDs. For example, Big Data DB 40 registers lifestyle awareness data, which includes survey results on attitudes and values ​​regarding various categories such as daily life, shopping, and work, as well as hobbies and how members spend their time and money. It should be noted that large-scale surveys are not necessarily conducted on all members; some surveys are conducted only on certain members (for example, those with points above a certain value). Also, some members who are selected as targets for large-scale surveys may not respond. Therefore, there is variation among members in the response results registered in Big Data DB 40. It is desirable to register as much data as possible regarding members' thoughts, attitudes, and behaviors in Big Data DB 40, so the survey management server 30 may register the results of surveys conducted at the request of companies to Big Data DB 40 with the approval of those companies. In addition, various basic attribute information of monitors (members) may be registered in Big Data DB 40 instead of Monitor DB 371. It should be noted that Big Data DB 40 may be registered in the storage unit 37 of the survey management server 30. In this embodiment, the basic attribute information (demographic information, marketing information, etc.) included in the monitor DB371 and the lifestyle awareness data included in the big data DB40 are collectively referred to simply as "attribute information."

[0023] The external server 50 has a generation AI model 53. In this embodiment, as illustrated in Figure 1, the external server 50 has a natural language model 51 and an image generation AI model 52 as the generation AI model 53. The natural language model 51 generates and outputs text corresponding to the input prompt. The image generation AI model 52 generates and outputs an image corresponding to the input prompt. The natural language model 51 and the image generation AI model 52 may each be general-purpose models used for various purposes, or they may be models that have been uniquely retrained (fine-tuned) from a general-purpose model. In this embodiment, an example in which the generation AI model 53 has a natural language model 51 and an image generation AI model 52 is described, but it is not limited to this. In other embodiments, the generation AI model 53 may have only one model that generates both text and images based on the input prompt. Even the generation AI model 53 having only one model may be a general-purpose model or a model that has been retrained from a general-purpose model. Furthermore, the generation AI model 53 (natural language model 51 and image generation AI model 52) may be built on the survey management server 30.

[0024] (Example of processing) Figure 2 is a flowchart showing an example of processing by the survey management server according to the embodiment. The example of processing in the survey management server 30 will be explained in detail with reference to Figure 2.

[0025] First, the input reception unit 31 receives a request to conduct and analyze a survey (step S101). For example, a client terminal 20 transmits survey request information, including the conditions of the monitors to be surveyed, the content of the questions, the deadline, etc., and the input reception unit 31 receives the survey request information. In addition to receiving survey request information from the client terminal 20, an operator who has received a request from a company may input the survey request information and analysis request information into the survey management server 30. When the input reception unit 31 obtains the survey request information and analysis request information transmitted from the client terminal 20 or directly entered, the acceptance of the survey request is completed.

[0026] Next, the survey processing unit 32 conducts the survey based on the survey request information (step S102). The survey processing unit 32 creates survey information, including a unique survey ID to be registered in the survey DB 372 and question information to be distributed to respondents. The question information may be automatically generated by an application, or it may be generated by an operator referring to the survey request information and inputting the question content as text. The question information is, for example, an HTML web page that displays the question as text and has areas where respondents can select answers from options using radio buttons or checkboxes, and areas where they can input answers as text. This web page is configured so that when a respondent enters their answer and presses a submit button, the respondent's answer is sent to the survey management server 30. The survey processing unit 32 refers to the monitor's attribute information (basic attribute information included in the monitor DB 371, lifestyle awareness data included in the big data DB 40, etc.) and extracts monitors that match the survey target conditions of the survey request information. The survey processing unit 32 distributes the created question information to the extracted monitors.

[0027] The monitor uses the respondent terminal 10 to check the distributed questions, enter their answers, and operate the respondent terminal 10 to send the entered answers to the survey management server 30. The survey management server 30's survey processing unit 32 retrieves the answers to the distributed question information. Upon retrieving the answers, the survey processing unit 32 registers the survey answers (survey results) in the survey database 372 (step S103). The survey processing unit 32 also registers the answers of monitors associated with the monitor ID in the survey database 372, linked to the monitor ID.

[0028] Next, based on the registered survey results, a profile analysis is performed on the target respondents among the survey respondents. First, the extraction unit 33 sets conditions (first condition and second condition) to identify the target and comparison subjects (step S104). The first and second conditions are, for example, those specified in advance by the company when the survey was requested to be conducted and analyzed in step S101. For example, the first condition is information indicating that the respondent is a first-time purchaser of the requesting company's brand product, and the second condition is information indicating that the respondent is a repeat customer who has purchased the brand's product multiple times. One way to recognize whether the information indicates a new purchaser or a repeat customer is to use the answers to the purchase-related questions in the survey. For example, the first condition is a respondent who selected answer a (e.g., first purchase, purchase count is less than ○ times) to the purchase-related question, and the second condition is a respondent who selected answer b (e.g., not first purchase, purchase count is ○ times or more) to the same question. In other embodiments, the first and second conditions may be information other than that indicating a new purchaser or a repeat customer. For example, the first condition may be a respondent who selected answer a (e.g., an affirmative answer) to a question 1 in the questionnaire, and the second condition may be a respondent who selected a different answer b (e.g., a negative answer) to the same question 1. Furthermore, for example, the first condition may be all respondents to the questionnaire, and the second condition may be monitors who are not included in the questionnaire (have not responded) but whose lifestyle awareness data is registered in the big data DB40. The first and second conditions are not limited to these. In addition, each of the first and second conditions may be a combination of multiple conditions (for example, a respondent who selected answer a to question 1 and answer b to question 2).

[0029] Furthermore, the extraction unit 33 determines whether profile analysis is possible based on the number of targets and comparison subjects identified based on the set first and second conditions (step S105). For example, if both the number of targets n1 identified from the first condition and the number of comparison subjects n2 from the second condition are less than or equal to the minimum value (e.g., n1 ≤ 2 and n2 ≤ 2), the extraction unit 33 determines that analysis is not possible (step S105; NO). In this case, the extraction unit 33 transmits the determination that analysis is not possible to the output unit 36, and the output unit 36 ​​outputs an analysis result to the client terminal 20 prompting a change in conditions because sufficient profile analysis cannot be performed under the current conditions (step S110). The requesting company operates the client terminal 20 to change the first and second conditions and sends them to the questionnaire management server 30. Then, returning to step S104, the extraction unit 33 resets the first and second conditions newly acquired through the input reception unit 31. The output unit 36 ​​may also output a result indicating that the analysis is not possible to the display device (not shown) of the questionnaire management server 30, and the operator may refer to this result and request a change in the conditions via telephone or client terminal 20.

[0030] On the other hand, if the number of targets n1 and the number of comparison targets n2 are greater than the minimum value, it is determined that profile analysis is possible (Step S105; YES). In this case, the extraction unit 33 compares the differences in attribute information and extracts the characteristics of the targets (Step S106). The attribute information includes basic attribute information contained in the monitor DB 371 and lifestyle awareness data contained in the big data DB 40.

[0031] For example, the extraction unit 33 compares the basic attribute information of the target and the comparison target (demographic information such as age and gender, and marketing information such as hobbies and preferences) and calculates the difference between the target and the comparison target. For example, the extraction unit 33 calculates the difference between the average age of the target and the average age of the comparison target. The extraction unit 33 calculates the difference between the gender composition ratio of the target (female 0%, male 0%) and the gender composition ratio of the comparison target (female ±×%, male ±×). Similarly, the extraction unit 33 calculates the difference between the geographical composition ratio of the target (Kanto 0%, Kansai 0%, ...) and the geographical composition ratio of the comparison target (Kanto ±×%, Kansai ±×%, ...). In addition, the extraction unit 33 calculates the difference between the geographical composition ratio of the target (hobby A 0%, hobby B 0%, ...) and the geographical composition ratio of the comparison target (hobby A ±×%, hobby B ±×%, ...). Differences are calculated similarly for other basic attribute information. Furthermore, the extraction unit 33 compares the proportion of lifestyle awareness data (responses to each question) between the target and the comparison group and calculates the difference. For example, the extraction unit 33 calculates the difference (±×%) between the proportion of respondents in the target group who selected answer a for question 1 and the proportion of respondents in the comparison group who selected answer a for question 1. Note that this question 1 is not a question from a questionnaire conducted at the request of a client, but a question from a questionnaire conducted to acquire lifestyle awareness data. The extraction unit 33 similarly calculates the difference between the proportion of target respondents and the proportion of comparison group respondents for all responses to each question. At this time, the extraction unit 33 extracts items from among the multiple items of attribute information that show a significant difference between the target and the comparison group, as items that represent the characteristics of the target (age group of monitors that make up the target, lifestyle awareness trends, etc.). Specifically, the extraction unit 33 extracts items from the attribute information where the difference between the target and the comparison group is above a threshold (for example, 5 points or more).

[0032] The extraction unit 33 extracts a predetermined number (for example, 20 items) from among multiple items of basic attribute information (demographic information, etc.) and lifestyle attitude data, in order of the largest differences, as items that represent the overall characteristics of the target. In addition, the lifestyle attitude data may be divided into multiple categories A, B, C, ... In this case, the extraction unit 33 may extract a predetermined number (for example, 3 items) from each category A, B, C, ... in order of the largest differences, as features that represent the characteristics (attitude tendencies) of the target for each category. At this time, if there are no items in a certain category whose difference exceeds a threshold, the extraction unit 33 may classify that category as "no characteristic attitudes."

[0033] Next, the prompt generation unit 34 generates prompts based on the target's characteristics and inputs them to the generation AI model 53 (natural language model 51 and image generation AI model 52) (step S107). Here, the prompts generated by the prompt generation unit 34 are prompts for the generation AI model 53 to generate profile data representing the target. Specifically, the prompt generation unit 34 generates prompts that include the target's characteristics extracted by the extraction unit 33. More specifically, the prompt generation unit 34 generates prompts for the natural language model 51 to generate profile sentences that summarize items representing the target's overall characteristics, and for category-specific profile sentences that summarize features representing the target's characteristics (tendencies in awareness) for each category of lifestyle awareness data. The prompt generation unit 34 also generates prompts for the image generation AI model 52 to generate images representing a person's image that summarizes items representing the target's overall characteristics.

[0034] In order to obtain accurate profile sentences and profile images from the natural language model 51 and the image generation AI model 52, it is necessary to appropriately generate prompts to be input to each model. For this reason, the prompt generation unit 34 generates prompts to be input to each model, for example, by the method described below.

[0035] Figure 3 is a diagram illustrating an example of processing by the questionnaire management server and the generation AI model according to the embodiment. First, an example of generating a prompt to be input to the natural language model 51 will be explained with reference to Figure 3. The prompt generation unit 34 has in advance text data P1 which will be the content of the instruction to the natural language model 51 in the prompt. The text data P1 is a standard phrase such as, for example, "Please summarize the following three features in Japanese within 50 characters." The prompt generation unit 34 also extracts data representing the target features (feature data P2) extracted by the extraction unit 33 from the attribute information as information necessary for the natural language model 51 to generate a profile sentence. The prompt generation unit 34 may also prepare structured data P3 which has structured attribute information in advance, and extract the feature data P2 from this structured data P3. For example, suppose the extraction unit 33 has extracted items (answers) a1, a3, and a6 as target features for category A of the lifestyle awareness data. In this case, the prompt generation unit 34 combines the text data P1 and the feature data P2, which is information (text) describing items (answers) a1, a3, and a6, to form a prompt sentence that the natural language model 51 can accept. In other words, the prompt generation unit 34 generates a prompt that includes pre-stored text data and extracted target features. For example, the prompt generation unit 34 generates a prompt for generating a profile sentence for category A, such as, "Summarize the following three features in Japanese of 50 characters or less: 'Lifestyle Awareness Data Item a1', 'Lifestyle Awareness Data Item a3', 'Lifestyle Awareness Data Item a6'." The prompt generation unit 34 inputs the generated category-specific prompts, along with the attribute information of the target and comparison target, into the natural language model 51. The natural language model 51 then generates a profile sentence representing the target's features in each category (e.g., category A of lifestyle awareness) based on the instructions included in the prompt and the attribute information, which is the target's feature data P2. Furthermore, the prompt generation unit 34 generates a prompt to summarize the items that represent the overall characteristics of the target.For example, suppose the extraction unit 33 extracts "average age" and "gender" from the basic attribute information, and "item a1 of category A," "item b2 of category B," "item c1 of category C," etc. from the lifestyle awareness data as items representing the characteristics of the target. In this case, the prompt generation unit 34 combines the text data P1 and the feature data P2, which is information that explains the extracted items, to form a prompt sentence that the natural language model 51 can accept. For example, the prompt generation unit 34 generates a prompt to generate a profile sentence that represents the overall characteristics of this target, such as, "Please summarize the following characteristics in Japanese of 50 characters or less: 'average age,' 'gender,' 'item a1 of category A of the lifestyle awareness data,' 'item b2 of category B of the lifestyle awareness data,' 'item c1 of category C of the lifestyle awareness data,' etc." The prompt generation unit 34 does not generate a prompt for the category "no distinctive characteristics."

[0036] Similarly, the prompt generation unit 34 generates a prompt for generating an image representing a person that summarizes items representing the overall characteristics of the target. This prompt for generating an image representing a person may be generated using the same items as the prompt for summarizing the items representing the overall characteristics of the target. The prompt generation unit 34 combines the text data P1 and the feature data P2 to form a prompt sentence that the image generation AI model 52 can accept. For example, the prompt generation unit 34 generates a prompt for generating a profile image representing the overall characteristics of this target, such as, "Please generate a person image that summarizes the following features: 'Average age', 'Gender', 'Item a1 from category A of lifestyle awareness data', 'Item b2 from category B of lifestyle awareness data', 'Item c1 from category C of lifestyle awareness data', ..."

[0037] Furthermore, the prompt generation unit 34 may generate prompts that conform to a predetermined "type" so that the natural language model 51 and the image generation AI model 52 can generate more accurate profile sentences and profile images. Specifically, the prompt generation unit 34 generates prompts that, for example, have labels consisting of symbols and text indicating the meaning of the information, such as "#Role:", "#Task:", and "#Example:", placed at the beginning of instructions (text data P1) or feature data P2. This enables the natural language model 51 and the image generation AI model 52 to interpret the roles and instructions (tasks) required of each model more correctly, thereby improving the accuracy of the output profile sentences and profile images. The prompt generation unit 34 may also include a prompt history P4 as "#Example:". The prompt generation unit 34 stores and accumulates a prompt history P4, which includes previously generated prompts and the profile data generated by these prompts, in a memory unit 37, for example. The prompt generation unit 34 retrieves the prompt history P4 from the memory unit 37 and appends it as "#Example:" to the prompt. In this way, by providing at least one of the previously given prompts and the generated profile data as examples, each model can be made to generate profile data that closely matches the desired content.

[0038] Next, the profile data acquisition unit 35 inputs each prompt generated by the prompt generation unit 34 to the natural language model 51 and the image generation AI model 52, respectively, thereby acquiring the target profile sentence generated by the natural language model 51 and the profile image generated by the image generation AI model 52 (step S108). As shown in the example in Figure 3, when the prompt generation unit 34 inputs prompts into the input fields, the natural language model 51 and the image generation AI model 52 generate profile sentences (profile sentences for each category and profile sentences representing overall characteristics) and profile images based on the input prompts and output them to the output fields. The profile data acquisition unit 35 acquires the profile sentences and profile images output to the output fields. Profile sentences and profile images are collectively referred to as profile data.

[0039] Next, the output unit 36 ​​creates a target profile summary D1 and profile sheet D2 based on the profile data generated by the generation AI model 53. As shown in the example in Figure 3, the output unit 36 ​​inputs the profile sentence generated by the natural language model 51 and the profile image generated by the image generation AI 52 into the designated locations in the profile summary D1 and profile sheet D2, respectively. Once these are created, the output unit 36 ​​outputs the questionnaire results registered in the questionnaire DB 372 and the created profile analysis results (profile summary and profile sheet) to the client terminal 20 (step S109).

[0040] (Example of profile analysis results) Figure 4 is the first figure showing an example of profile analysis results according to the embodiment. Figure 5 is the second figure showing an example of profile analysis results according to the embodiment. Figure 4 shows an example of profile summary D1, and Figure 5 shows an example of profile sheet D2.

[0041] As shown in Figure 4, the profile summary D1 is data that shows an overview of the target's profile and includes target information D11, comparison information D12, profile summary text D13, profile image D14, basic attribute information D15, lifestyle awareness information D16, etc. Target information D11 includes the first condition used to identify the target and the number of targets n1 extracted by the first condition. Comparison information D12 includes the second condition used to identify the comparison targets and the number of comparison targets n2 extracted by the second condition. The profile summary text D13 is a profile text (summary text) that represents the overall characteristics of the target, generated by the natural language model 51. The profile image D14 is an image representing the person of the target, generated by the image generation AI model 52. Basic attribute information D15 is information that aggregates basic attribute information such as the target's demographic information by item and includes the group to which the largest number of targets belong out of the total number of targets n1, the average age of the targets, etc. In the example in Figure 4, the "gender" item shows that the largest number of targets belong to the "female" group. Lifestyle awareness information D16 is a list of profile sentences (summary sentences) that describe the characteristics of each of the target categories A, B, C, ... generated by the natural language model 51.

[0042] As shown in Figure 5, profile sheet D2 includes target information D21, comparison information D22, profile summary D23, profile image D24, basic attribute information D25, lifestyle awareness information D26, etc. Target information D21, comparison information D22, profile summary D23, and profile image D24 are the same as the corresponding items D11-D24 in profile summary D1. Basic attribute information D25 includes item information D251, which is the same as basic attribute information D15 in profile summary D1, and aggregated / comparative information D252 for each item. Aggregated / comparative information D252 shows the difference between the proportion of the most frequent group relative to the total number of targets n1 and the proportion of the most frequent group in the comparison group. For example, for the item "place of residence," the "Kanto" group has the most targets, accounting for 45% of the total number of targets n1, and the difference from the proportion of the comparison group belonging to the "Kanto" group is +4.0%. Lifestyle awareness information D26 represents the category-specific items of lifestyle awareness data extracted by the extraction unit 33 as characteristics of the target. Lifestyle awareness information D26 includes the extracted response content D261 and the aggregated / comparative information D262 for each response content. The aggregated / comparative information D262 shows the difference between the percentage of targets who selected the extracted response and the percentage of comparison subjects who selected the same response. In addition, the response content D261 and the aggregated / comparative information D262 are displayed in descending order of difference. In the example in Figure 5, for category A, the difference in the percentage of targets who selected response A6 is the largest between the target and the comparison subjects, with 84% of targets selecting response A6, and a difference of +10.6% compared to the comparison subjects. For example, in category E, if there are no items where the difference from the comparison target exceeds a threshold, and the extraction unit 33 sets it to "No distinctive awareness," then, as shown in Figure 5, the lifestyle awareness information D26 displays information D263 indicating "No distinctive awareness" instead of the response content D261 and the aggregated / comparison information D262.

[0043] (modified version) Figure 6 is a flowchart showing an example of processing by a survey management server according to a modified embodiment. Figure 2 shows an example in which the survey management server 30 creates and conducts a survey, but it is not limited to this. For example, the company may create and conduct the survey, and only the analysis of the survey results may be performed by the research company (survey management server 30). An example of processing in this case will be explained with reference to Figure 6. First, in step S201, the input reception unit 31 accepts the upload of the survey results conducted by the company along with a request for survey analysis. Also, step S102 in Figure 2 is omitted. Next, in step S202, the survey processing unit 32 registers the uploaded survey results in the survey DB 372. The processing in step S202 is the same as the processing in step S103 in Figure 2. Also, the subsequent processing S203 to S209 is the same as S104 to S110 in Figure 2, respectively. Furthermore, the requesting company can repeatedly execute steps S203 to S209 on the survey management server 30 via the client terminal 20, changing the first and second conditions, thereby performing target profile analysis according to various conditions.

[0044] Furthermore, Figures 1 and 2 illustrate an example in which the prompt generation unit 34 generates prompts for generating an image representing a person's profile that summarizes the overall characteristics of the target, inputs these prompts along with attribute information to the image generation AI model 52, and the image generation AI model 52 generates an image corresponding to the input prompt. However, the system is not limited to this example. For instance, the memory unit 37 stores multiple candidate images representing the target's profile that have been prepared in advance by the research company, and the profile data acquisition unit 35 may use the candidate images stored in the memory unit 37 instead of the image generated by the image generation AI model 52. For example, each candidate image may be associated with multiple keywords indicating attribute information. Keywords include, for example, keywords indicating basic attribute information such as "female (gender)" and "Kanto (place of residence)," and keywords included in the responses to lifestyle attitude data ("small daily luxuries," "long-standing traditions," etc.). In this case, in step S107, the prompt generation unit 34 does not generate prompts for input to the image generation AI model 52. Also, in step S108, the profile data acquisition unit 35 acquires the candidate image containing the most keywords that match the features extracted by the extraction unit 33 from among multiple candidate images as the target's profile image. The profile data acquisition unit 35 may also make it a requirement that the image contains keywords indicating features in the target's basic attribute information ("female," etc.), and acquire the candidate image containing the most keywords indicating features in the target's lifestyle attitude data from among the candidate images extracted based on this requirement as the target's profile image. This makes it possible to suppress the selection of a profile image that deviates from the target's image.

[0045] (effect) As described above, according to this embodiment, the survey management server 30 includes an extraction unit 33 that extracts the characteristics of a target by comparing the attribute information of a target identified by a first condition with the attribute information of a comparison target identified by a second condition from multiple monitors; a prompt generation unit 34 that inputs the attribute information of the target and a prompt generated based on the characteristics of the target to a natural language model 51 that generates a profile sentence representing the target; and a profile data acquisition unit 35 that acquires the profile sentence generated by the natural language model 51 based on the prompt.

[0046] In this way, the survey management server 30 can easily obtain profile data showing the tendencies of targets that meet specific conditions, using the natural language model 51.

[0047] Furthermore, the prompt generation unit 34 inputs the target's attribute information and the prompt generated based on the target's characteristics to the image generation AI model 52, which generates a profile image representing the target, and the profile data acquisition unit 35 acquires the profile image generated by the image generation AI model 52 based on the prompt.

[0048] In this way, the survey management server 30 can easily obtain an image of the target person using the image generation AI model 52.

[0049] Furthermore, the extraction unit 33 compares the differences between the target and the comparison target for each of the multiple attribute information items, and extracts items whose differences are greater than a predetermined threshold as features of the target.

[0050] In this way, the survey management server 30 can appropriately extract attribute information that shows a significant difference between the target and the comparison target as features that represent the target's attributes.

[0051] Furthermore, the extraction unit 33 extracts a predetermined number of items as target features, starting with those with the largest differences among the items whose differences exceed a predetermined threshold.

[0052] In this way, the survey management server 30 can appropriately narrow down attribute information that has a significant difference between the target and the comparison target. This prevents the prompts generated by the prompt generation unit 34 from including unnecessary features with small differences, and thus prevents important features with larger differences from being omitted from the profile statement.

[0053] Furthermore, the extraction unit 33 extracts a predetermined number of items from among those whose differences exceed a predetermined standard, starting with the items with the largest differences, as features of the target category, for each category of attribute information. Categories in which there are no items whose differences exceed a predetermined standard are set to have no features.

[0054] In this way, the survey management server 30 can exclude the generation of prompts for categories with small differences from the comparison target by classifying them as having no features. This prevents the prompts generated by the prompt generation unit 34 from including features that do not differ significantly from the comparison target, thus preventing the target profile statement from becoming inaccurate.

[0055] Furthermore, the extraction unit 33 outputs a result indicating that analysis is not possible if both the number of targets and comparison targets are below a predetermined number.

[0056] In this way, the survey management server 30 can prevent the output of inaccurate analysis results when the number of targets and comparison subjects is insufficient.

[0057] (Computer configuration) Figure 7 is a schematic block diagram showing the configuration of a computer according to an embodiment. The computer 900 comprises a processor 901, main memory 902, auxiliary memory 903, and interface 904. The above-mentioned survey management server 30 is implemented in the computer 900. The operation of each processing unit described above is stored in the auxiliary memory 903 in the form of a program. The processor 901 reads the program from the auxiliary memory 903, expands it into the main memory 902, and executes the above processing according to the program. The processor 901 also allocates memory area in the main memory 902 to be used for the above processing according to the program.

[0058] The program may be for implementing a part of the functions to be performed by the computer 900. For example, the program may perform functions in combination with other programs already stored in the auxiliary storage device 903, or in combination with other programs implemented in other devices. In other embodiments, the computer may be equipped with a custom LSI (Large Scale Integrated Circuit) such as a PLD (Programmable Logic Device) in addition to, or in place of, the above configuration. Examples of PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), FPGA (Field Programmable Gate Array), etc. In this case, some or all of the functions implemented by the processor may be implemented by the integrated circuit.

[0059] Examples of auxiliary storage devices 903 include HDDs (Hard Disk Drives), SSDs (Solid State Drives), magnetic disks, magneto-optical disks, CD-ROMs (Compact Disc Read Only Memory), DVD-ROMs (Digital Versatile Disc Read Only Memory), and semiconductor memory. The auxiliary storage device 903 may be an internal medium directly connected to the bus of the computer 900, or it may be an external medium (external storage device 910) connected to the computer 900 via an interface 904 or a communication line. Furthermore, if this program is distributed to the computer 900 via a communication line, the computer 900 that receives the distribution may expand the program into the main memory 902 and execute the above processing. In at least one embodiment, the auxiliary storage device 903 is a tangible storage medium that is not temporary.

[0060] (Note 1) An online survey management server comprising: an extraction unit that compares attribute information of a target identified by a first condition with attribute information of a comparison target identified by a second condition from multiple monitors to extract the characteristics of the target; a prompt generation unit that generates prompts for input to a generating AI model based on the characteristics of the target extracted by the extraction unit; and a profile data acquisition unit that acquires profile data representing the target generated by the generating AI model by inputting the prompts generated by the prompt generation unit into the generating AI model.

[0061] (Note 2) The questionnaire management server described in Appendix 1 further comprises an output unit that creates a profile analysis result for the target using the profile data acquired by the profile data acquisition unit.

[0062] (Note 3) The survey management server according to Appendix 1 or 2, wherein the generation AI model includes a natural language model that generates text based on input prompts, the prompt generation unit generates prompts for input to the natural language model based on the characteristics of the target, and the profile data acquisition unit acquires a profile sentence representing the target generated by the natural language model as the profile data.

[0063] (Note 4) The survey management server according to any one of the appendices 1 to 3, wherein the generation AI model includes an image generation AI model that generates an image based on an input prompt, the prompt generation unit generates a prompt for input to the image generation AI model based on the characteristics of the target, and the profile data acquisition unit acquires a profile image representing the target generated by the image generation AI as the profile data.

[0064] (Note 5) The extraction unit compares the differences between the target and the comparison target for each of the attribute information items of multiple items, and extracts items whose differences are greater than a predetermined threshold as features of the target, as described in any one of the appendices 1 to 4.

[0065] (Note 6) The extraction unit is the questionnaire management server described in Appendix 5, which extracts a predetermined number of items as characteristics of the target, starting with the items whose differences are greater than a predetermined threshold.

[0066] (Note 7) The extraction unit extracts a predetermined number of items from among those whose differences are greater than a predetermined threshold, starting with the items with the largest differences, as features for the target category, for each category of attribute information, and sets categories where there are no items whose differences are greater than a predetermined standard to have no features, as described in Appendix 5 or 6.

[0067] (Note 8) The extraction unit outputs a result indicating that analysis is not possible if the number of targets and comparison targets is less than or equal to a predetermined number, as described in any one of the appendices 1 to 7.

[0068] (Note 9) The prompt generation unit generates the prompt by combining the characteristics of the target with predetermined text data, as described in any one of the appendices 1 to 8 of the survey management server.

[0069] (Note 10) The prompt generation unit generates the prompt by further combining text that explains the prompt history, including prompts generated in the past, as an example, as described in Appendix 9 of the survey management server.

[0070] (Note 11) The extraction unit identifies target respondents from among the respondents of the questionnaire based on a first condition, and identifies comparison respondents based on a second condition, wherein the first condition is a respondent who gave a specific answer to a specific question in the questionnaire, and the second condition is a respondent who gave a different answer to the target in the questionnaire, as described in any one of the appendices 1 to 10.

[0071] (Note 12) The extraction unit is the questionnaire management server described in Appendix 11, which accepts the setting of the first and second conditions.

[0072] (Note 13) A questionnaire analysis method comprising the steps of: extracting the characteristics of a target by comparing attribute information of a target identified by a first condition with attribute information of a comparison target identified by a second condition from multiple monitors; generating prompts for input to a generative AI model based on the extracted characteristics of the target; and obtaining profile data representing the target generated by the generative AI model by inputting the generated prompts to the generative AI model.

[0073] (Note 14) A program that causes a survey management server to execute the following steps: comparing attribute information of a target identified by a first condition with attribute information of a comparison target identified by a second condition from multiple monitors to extract the characteristics of the target; generating prompts for input into a generating AI model based on the extracted characteristics of the target; and obtaining profile data representing the target generated by the generating AI model by inputting the generated prompts into the generating AI model. [Explanation of Symbols]

[0074] 1. Survey System 10, 10a, 10b Respondent terminals 20 client terminals 30 Survey Management Server 31 Input Reception Section 32 Questionnaire Processing Department 33 Extraction part 34 Prompt generation unit 35 Profile data acquisition unit 36 Output section 37 Memory section 371 Monitor DB 372 Questionnaire Database 50 External Servers 51 Natural Language Models 52 AI Models for Image Generation 53 Generative AI Models 40 Big Data Databases

Claims

1. An extraction unit that compares the attribute information of a target identified by a first condition with the attribute information of a comparison target identified by a second condition from multiple monitors to extract the characteristics of the target, A prompt generation unit generates prompts for input to the generated AI model based on the characteristics of the target extracted by the extraction unit, A profile data acquisition unit acquires profile data representing the target generated by the AI ​​model by inputting the prompt generated by the prompt generation unit into the AI ​​model. A survey management server equipped with the following features.

2. The system further includes an output unit that uses the profile data acquired by the profile data acquisition unit to create a profile analysis result for the target. The questionnaire management server according to claim 1.

3. The aforementioned generation AI model includes a natural language model that generates text based on an input prompt. The prompt generation unit generates a prompt for input to the natural language model based on the characteristics of the target, The profile data acquisition unit acquires a profile statement representing the target generated by the natural language model as the profile data. The questionnaire management server according to claim 1.

4. The aforementioned generation AI model includes an image generation AI model that generates an image based on an input prompt. The prompt generation unit generates prompts for input to the image generation AI model based on the characteristics of the target. The profile data acquisition unit acquires a profile image representing the target generated by the image generation AI model as the profile data. The questionnaire management server according to claim 1.

5. The extraction unit compares the differences between the target and the comparison target for each of the multiple attribute information items, and extracts items whose differences are greater than a predetermined threshold as features of the target. The questionnaire management server according to claim 1.

6. The extraction unit extracts a predetermined number of items from among those whose differences are greater than a predetermined threshold, starting with those with larger differences, as features of the target. The questionnaire management server according to claim 5.

7. The extraction unit is For each category of attribute information, a predetermined number of items whose differences are greater than a predetermined threshold are extracted as characteristics of the target category, starting with those with the largest differences. For categories where no item exceeds a predetermined standard in the aforementioned difference, set the category to "no distinguishing features." The questionnaire management server according to claim 5.

8. The extraction unit outputs a result indicating that analysis is not possible if both the number of targets and the number of comparison targets are less than or equal to a predetermined number. The questionnaire management server according to any one of claims 1 to 7.

9. The prompt generation unit generates the prompt by combining the characteristics of the target with predetermined text data. The questionnaire management server according to any one of claims 1 to 7.

10. The prompt generation unit generates a prompt by further combining text that explains the prompt history, including prompts generated in the past, as an example. The questionnaire management server according to claim 9.

11. The extraction unit identifies the target respondents from among the survey respondents based on the first condition, and identifies the comparison target respondents based on the second condition. The first and second conditions are, respectively, respondents who gave specific answers to specific questions in the questionnaire. The questionnaire management server according to any one of claims 1 to 7.

12. The extraction unit accepts the setting of the first and second conditions. The questionnaire management server according to claim 11.

13. The process involves comparing the attribute information of a target identified by a first condition with the attribute information of a comparison target identified by a second condition, from multiple monitors, to extract the characteristics of the target. The steps include generating prompts for input to the generated AI model based on the extracted target characteristics, The steps include: obtaining profile data representing the target generated by the generating AI model by inputting the generated prompt into the generating AI model; A survey analysis method that includes [specific characteristics / features].

14. The process involves comparing the attribute information of a target identified by a first condition with the attribute information of a comparison target identified by a second condition, from multiple monitors, to extract the characteristics of the target. The steps include generating prompts for input to the generated AI model based on the extracted target characteristics, The steps include: obtaining profile data representing the target generated by the generating AI model by inputting the generated prompt into the generating AI model; A program that instructs the survey management server to execute the command.

Images