Information processing device, information processing method, and program

Abstract

This invention provides an information processing device, an information processing method, and a program that reduce the effort required to utilize review submissions. [Solution] In an information processing system, an information processing device that can exchange information with a generation AI device that executes a generation AI that processes prompts and outputs information comprises at least one processor, the processor requests an emotion analysis engine to obtain an emotion phrase which is a predicate containing an emotion expression and a phrase containing the target word of the emotion expression from a review post to a company or store, the processor outputs a request prompt to the generation AI device requesting it to output an evaluation and / or improvement suggestion for the company or store using the obtained emotion phrase, and the processor causes the output information to be output to a terminal using the response output from the generation AI device in response to the request prompt.

Description

Technical Field

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

Background Art

[0002] Conventionally, in a web service of a map or a web review site, a system for posting reviews (so-called word-of-mouth) for companies or stores has been developed (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] For example, when the number of review posts for a target store is large or when the text of the review post by the poster is long, etc., it is laborious to grasp the evaluation by the poster in the review post and / or to extract points to be improved from the review post. There is a problem that it is laborious to utilize the review posts in this way. <�

[0005] The present invention has been made in view of the above problems, and an object thereof is to provide an information processing apparatus, an information processing method, and a program capable of reducing the labor required to utilize review posts.

Means for Solving the Problems

[0006] An information processing apparatus according to a first aspect of the present invention is an information processing apparatus capable of exchanging information with a generation AI device that executes a generation AI that processes a prompt and outputs information, comprising at least one processor, The processor includes a procedure for requesting an emotion analysis engine to obtain an emotion phrase, which is a predicate containing an emotion expression and a phrase containing the target word of the emotion expression, from a review post about a company or store. A procedure for outputting a request prompt to the generating AI device requesting it to output an evaluation and / or improvement suggestion for the company or store using the acquired sentiment phrase, A procedure for outputting output information using the response output from the generating AI device in response to the aforementioned request prompt, Execute this.

[0007] This configuration makes it easy to obtain objective evaluations of companies or stores and / or objective improvement suggestions based on review submissions using a generation AI device, thereby reducing the effort required to utilize review submissions.

[0008] An information processing device according to a second aspect of the present invention is an information processing device according to a first aspect, The processor further performs a procedure to classify the acquired emotion phrase or the target word of the emotion expression within the emotion phrase into one of several categories established for analysis. The aforementioned request prompt is a prompt that requests the output of evaluations and / or improvement suggestions for the company or store for a target category specified by the user.

[0009] This configuration makes it easy for users to obtain objective evaluations of companies or stores and / or objective improvement suggestions for a target category specified by the user.

[0010] An information processing device according to a third aspect of the present invention is an information processing device according to a second aspect, A dictionary linking emotional expression terms and categories is stored in memory for each industry. In the classification procedure described above, the processor refers to the dictionary corresponding to the user's industry and classifies the acquired sentiment phrase or the target word of the sentiment expression within the sentiment phrase into one of several categories provided for analysis. The aforementioned request prompt is a prompt that requests the output of an evaluation and / or improvement suggestion for a company or store for the target category, using the emotional phrase classified in the target category or the target word of the emotional expression in the emotional phrase.

[0011] This configuration allows for the classification of emotional phrases or target words within emotional phrases into appropriate categories by referencing dictionaries established for each industry.

[0012] An information processing device according to a fourth aspect of the present invention is an information processing device according to a third aspect, The aforementioned memory device stores a second category, which is narrower in scope than the first category, associated with each first category, and stores a dictionary for each industry, which associates the target words for emotional expression with the second category. In the classification procedure described above, the processor refers to the dictionary corresponding to the user's company and classifies the acquired sentiment phrase or the target word in the sentiment phrase into one of the plurality of second categories in order to classify the acquired sentiment phrase or the target word in the sentiment phrase into a first category.

[0013] According to this structure, since dictionaries established for each industry are referenced, emotional phrases or target words within emotional phrases can be classified into the appropriate first category by classifying them into the appropriate second category.

[0014] An information processing device according to a fifth aspect of the present invention is an information processing device according to any one of the first to fourth aspects, The processor further performs a procedure to extract a group of sentiment phrases that fall within the extraction range specified by the user from the sentiment phrases obtained in the acquisition procedure. The request prompt is a prompt that requests to output an evaluation of and / or an improvement proposal for the company or the store using the obtained group of sentiment phrases.

[0015] According to this configuration, an objective evaluation of and / or an objective improvement proposal for a company or a store can be easily obtained based on the group of sentiment phrases corresponding to the extraction range specified by the user.

[0016] The information processing apparatus according to the sixth aspect of the present invention is the information processing apparatus according to any one of the first to fifth aspects, and the request prompt is a prompt that requests to output an evaluation of and / or an improvement proposal for the company or the store along at least one or more frameworks.

[0017] According to this configuration, an evaluation and / or an improvement proposal along the framework can be obtained.

[0018] The information processing apparatus according to the seventh aspect of the present invention is the information processing apparatus according to any one of the first to sixth aspects, and the request prompt is a prompt that requests the results of evaluating the company or the store along at least one or more frameworks for each of the first period and the second period specified by the user, and the processor further executes a procedure of outputting, to the generation AI apparatus, a prompt that requests to output a transition of evaluation using the prompt processing result for the first period and the prompt processing result for the second period.

[0019] According to this configuration, a transition of evaluation from the first period to the second period along at least one or more frameworks can be obtained, and a transition of the user's evaluation of the company or the store can be easily obtained.

[0020] The information processing apparatus according to the eighth aspect of the present invention is the information processing apparatus according to the second aspect, and <000009A dictionary in which the target word and the category are associated is stored in the storage device for each company. In the classification procedure, the processor refers to the dictionary corresponding to the company of the user, and classifies the acquired emotional phrase or the target word in the emotional phrase into any of a plurality of categories provided for analysis.

[0021] According to this configuration, since the dictionary provided for each company is referred to, the emotional phrase or the target word in the emotional phrase can be classified into an appropriate category.

[0022] The information processing method according to the ninth aspect of the present invention is A procedure in which an acquisition unit requests an emotion analysis engine to acquire an emotional phrase, which is a phrase including a predicate having an emotional expression and a target word of the emotional expression, from a review post for a company or a store. A procedure in which a first output unit outputs a request prompt for requesting an AI generation device to generate an evaluation and / or an improvement proposal for the company or the store by using the acquired emotional phrase to the AI generation device. A procedure in which a second output unit outputs output information by using the answer output from the AI generation device in response to the request prompt. having

[0023] According to this configuration, since an objective evaluation and / or an objective improvement proposal for a company or a store can be easily obtained based on a review post by using an AI generation device, the labor involved in utilizing the review post can be reduced.

[0024] The program according to the tenth aspect of the present invention is for a computer capable of exchanging information with an AI generation device that executes an AI generation that processes a prompt and outputs information. A procedure in which an emotion analysis engine is requested to acquire an emotional phrase, which is a phrase including a predicate having an emotional expression and a target word of the emotional expression, from a review post for a company or a store. A procedure for outputting a request prompt to the generating AI device requesting it to output an evaluation and / or improvement suggestion for the company or store using the acquired sentiment phrase, A procedure for outputting output information using the response output from the generating AI device in response to the aforementioned request prompt, This is a program to execute [the command / action].

[0025] This configuration makes it easy to obtain objective evaluations of companies or stores and / or objective improvement suggestions based on review submissions using a generation AI device, thereby reducing the effort required to utilize review submissions. [Effects of the Invention]

[0026] According to one aspect of the present invention, an AI generation device can be used to easily obtain objective evaluations of companies or stores and / or objective improvement suggestions based on review submissions, thereby reducing the effort required to utilize review submissions. [Brief explanation of the drawing]

[0027] [Figure 1] This is a schematic diagram of the information processing system according to this embodiment. [Figure 2] This is an example of a schematic configuration diagram of the information processing device in this embodiment. [Figure 3] This is an example of a classification master table stored in storage device 23. [Figure 4] This is an example of an industry-specific dictionary stored in memory device 23. [Figure 5] This is a sequence diagram illustrating an example of the process of saving emotional phrases. [Figure 6] This is an example of a table for storing emotional phrases. [Figure 7] This is an example of the segment input screen displayed on terminal 1 in Example 1. [Figure 8] This is a sequence diagram showing an example of the processing flow according to Example 1. [Figure 9] This is an example of a prompt sent in step S223 of Figure 8. [Figure 10] This is an example of a response sent in step S312 of Figure 8. [Figure 11] This is an example of a prompt sent in step S224 of Figure 8. [Figure 12] This is an example of a response sent in step S314 of Figure 8. [Figure 13] Figure 13 shows an example of the segment input screen displayed on terminal 1 in Example 2. [Figure 14] This is a sequence diagram showing an example of the processing flow according to Example 2. [Figure 15] This is an example of a prompt sent in step S233 of Figure 14. [Figure 16] This is an example of a response sent in step S322 of Figure 14. [Figure 17] This is an example of a prompt sent in step S234 of Figure 14. [Figure 18] This is an example of a response sent in step S324 of Figure 14. [Figure 19] This is an example of a prompt sent in step S235 of Figure 14. [Figure 20] This is an example of a response sent in step S326 of Figure 14. [Modes for carrying out the invention]

[0028] The following descriptions of each embodiment will be made with reference to the drawings. However, unnecessarily detailed explanations may be omitted. For example, detailed explanations of already well-known matters and redundant explanations of substantially identical configurations may be omitted. This is to avoid the following explanation becoming unnecessarily verbose and to facilitate understanding for those skilled in the art.

[0029] Figure 1 is a schematic diagram of the information processing system according to this embodiment. As shown in Figure 1, the information processing system S, as an example, comprises terminals 1-1, ..., 1-N (where N is a natural number) used by the user, an information processing device 2, a generation AI device 3, and an emotion analysis device 4. Each of the terminals 1-1, ..., 1-N is connected to the information processing device 2 via a communication network CN. Terminals 1-1, ..., 1-N are, for example, smartphones, tablet devices, laptop computers, or personal computers. Hereinafter, terminals 1-1, ..., 1-N will be collectively referred to as terminal 1.

[0030] For example, the information processing device 2 is connected to the generating AI device 3 via a communication network CN, and is, for example, a server. While this example describes an instance where the information processing device 2 is connected to the generating AI device 3 via a communication network CN, the generating AI device 3 may be built within the information processing device 2, or it may be connected externally; the information processing device 2 only needs to be able to exchange information with the generating AI device 3.

[0031] The generation AI device 3 executes generation AI (Artificial Intelligence) that processes prompts and outputs information (e.g., text). This allows the generation AI device 3 to process prompts, generate information (e.g., text) corresponding to those prompts, and output it. Specifically, for example, when the generation AI device 3 receives a prompt from the information processing device 2, it processes this prompt to output information and then transmits this output information back to the information processing device 2.

[0032] The emotion analysis device 4 is connected to the information processing device 2 via a communication network CN. The emotion analysis device 4 includes an emotion analysis engine 41. This emotion analysis engine 41 extracts emotion phrases, which are predicates containing emotional expressions and phrases containing the target words of those emotional expressions, from text (e.g., sentences). Specifically, for example, the emotion analysis engine 41 extracts emotion phrases, which are phrases containing predicates containing emotional expressions and phrases containing the target words of those emotional expressions, from review posts to companies or stores in response to a request from the information processing device 2, and transmits the extracted emotion phrases to the information processing device 2. Here, an emotion phrase may be, for example, the smallest unit phrase containing a predicate containing an emotional expression and phrases containing the target words of those emotional expressions. This extraction of emotion phrases can be performed using a known emotion analysis engine, or it may be a machine learning model trained using training data that takes text data as input data and emotion phrases as output data. Furthermore, the sentiment analysis engine may be, for example, a machine learning model trained using training data that takes text data as input data and predicates containing sentiment expressions as output data. In this case, the machine learning model may use the predicates containing sentiment expressions output to extract the target words of the sentiment expression through dependency parsing in natural language processing, and output the smallest unit phrase containing the predicate containing the sentiment expression and the target words of the sentiment expression.

[0033] Figure 2 is an example of a schematic configuration diagram of the information processing device in this embodiment. As shown in Figure 2, the information processing device 2 includes, as an example, an input interface 21, a communication module 22, a storage device 23, a memory 24, an output interface 25, and a processor 26. In this embodiment, the information processing device 2 is described as having one processor 26 as an example, but it is not limited to this, and may have multiple processors, or at least one processor.

[0034] The input interface 21 receives input from, for example, an administrator (for example, an employee of a management organization) of the information processing device 2, and outputs an input signal corresponding to the received input to the processor 26. The communication module 22 is connected to the communication network CN and communicates with terminals 1-1 to 1-N, the generation AI device 3, and the emotion analysis device 4. This communication may be wired or wireless, but this explanation will assume it is wired.

[0035] The storage device 23 stores programs and various data for the processor 26 to read and execute. Memory 24 temporarily holds data and programs. Memory 24 is volatile memory, such as RAM (Random Access Memory). The output interface 25 can be connected to an external device and can output signals to that external device.

[0036] The processor 26 loads a program from the storage device 23 into memory 24 and executes a series of instructions contained in the program. An overview of one aspect of the processor 26's processing is described below. The processor 26 performs a procedure to request the sentiment analysis engine 41 to obtain sentiment phrases, which are predicates containing sentiment expressions and phrases containing the target words of those sentiment expressions, from review posts about a company or store. The processor 26 then performs a procedure to output a request prompt to the generating AI device 3, requesting it to output an evaluation and / or improvement suggestion for the company or store using the obtained sentiment phrase. The processor 26 then performs a procedure to output output information using the response output from the generating AI device in response to the request prompt. With this configuration, objective evaluations and / or objective suggestions for a company or store can be easily obtained based on review posts using the generating AI device 3, thus reducing the effort required to utilize review posts.

[0037] The processor 26 may also further include a procedure for classifying the acquired emotion phrase or the target word of the emotion expression within the emotion phrase into one of several categories provided for analysis. In this case, the request prompt output to the generating AI device 3 may be a prompt requesting that the device output an evaluation and / or improvement suggestion for the company or store for the target category specified by the user. This makes it easy to obtain an objective evaluation and / or objective improvement suggestion for the company or store for the target category specified by the user.

[0038] In this embodiment, as an example, the memory device 23 stores dictionaries for each industry, in which target words and categories of emotional expressions are associated. In this case, the processor 26, in the procedure of classifying into one of the plurality of categories, may refer to the dictionary corresponding to the user's industry and classify the acquired emotional phrase or the target word of the emotional expression in the emotional phrase into one of the plurality of categories provided for analysis. In this case, the request prompt output to the generating AI device 3 is a prompt requesting that the AI ​​output an evaluation and / or improvement suggestion for the company or store for the target category using the emotional phrase or the target word of the emotional expression in the emotional phrase classified into the target category. According to this, since dictionaries provided for each industry are referred to, the emotional phrase or the target word in the emotional phrase can be classified into an appropriate category.

[0039] Figure 3 shows an example of a classification master table stored in the memory device 23. As shown in Figure 3, the classification master table T1 stores records of pairs of major categories (also called first categories) and intermediate categories (also called second categories). In this way, the memory device 23 stores intermediate categories that are narrower in scope than the major category associated with each major category. As a concrete example of dictionaries that associate target words for emotional expressions with categories being stored for each industry, dictionaries that associate target words for emotional expressions with intermediate categories are stored for each industry. Figure 4 shows an example of industry-specific dictionaries stored in the memory device 23. In the food industry dictionary T2-1, records of pairs of target words for emotional expressions with intermediate categories are stored. In the beauty industry dictionary T2-2, records of pairs of target words for emotional expressions with intermediate categories are stored. In this way, as an example, dictionaries that associate target words for emotional expressions with intermediate categories are stored for each industry. In this case, for example, in the procedure of classifying into one of the multiple categories, the processor 26 refers to the dictionary corresponding to the user's company and classifies the acquired sentiment phrase or the target word in the sentiment phrase into one of the multiple subcategories in order to classify the acquired sentiment phrase or the target word in the sentiment phrase into a major category. As a result, since dictionaries provided for each industry are referred to, the sentiment phrase or the target word in the sentiment phrase can be classified into an appropriate major category by classifying it into an appropriate subcategory.

[0040] Next, we will explain the flow of preprocessing that is performed periodically (for example, once a week) as an example, using Figure 5. Figure 5 is a sequence diagram showing an example of the process of saving emotional phrases. The following explanation will describe the processing of each device's processor in accordance with Figure 5, but to avoid redundancy, the processors will be omitted.

[0041] (Step S211) The information processing device 2, for example, acquires the differences in review posts on the posting site in bulk and stores them in the storage device 23. The review posts acquired here may be review posts for a target company or store, review posts for all companies or stores that have been registered in advance, or review posts for all companies or stores in a specific region.

[0042] (Step S212) The information processing device 2 requests the sentiment analysis device 4 to perform sentiment analysis on the review posts acquired in batch differentially, for example.

[0043] (Step S411) When the emotion analysis device 4 receives a request, it performs emotion analysis and extracts from the review post a group of emotion phrases that include the predicate of the emotion expression and the target word of that emotion expression.

[0044] (Step S412) The emotion analysis device 4 then transmits the extracted emotion phrases to the information processing device 2.

[0045] (Step S213) When the information processing device 2 receives a group of emotion phrases, it classifies each emotion phrase or the target word within the emotion phrase into one of several subcategories, and reads out the main category corresponding to the classified subcategories.

[0046] (Step S214) The information processing device 2 then associates each emotional phrase with a major category and a medium category and stores them in the storage device 23. As a result, for example, the data in table T3 shown in Figure 6 is stored in the storage device 23.

[0047] Figure 6 shows an example of a table for storing sentiment phrases. As shown in Figure 6, table T3 stores records of pairs of the following: the year and month of the review post, the store name, the sentiment phrase extracted from the review post, the major category to which the sentiment phrase belongs, and the subcategory to which the sentiment phrase belongs.

[0048] The following describes each of the embodiments.

[0049] <Example 1> First, let's describe Example 1. In Example 1, the generating AI device 3 is requested to output evaluations and / or improvement suggestions for a company or store according to at least one framework, and these are obtained. Figure 7 is an example of the segment input screen displayed on terminal 1 in Example 1. As shown in Figure 7, the segment input screen G1 is provided with a select box R11 for specifying the extraction range from review submissions. This extraction range is specified by at least one combination of, for example, company, store, analysis period (e.g., year and month), category (e.g., major category or medium category), and language. Here, as an example, we will describe an example of specifying the extraction range by company, store, and analysis period.

[0050] Furthermore, the segment input screen G1 is provided with a select box R12 for specifying the category to be analyzed (hereinafter referred to as the target category, in this example being a major category). The segment input screen G1 is also provided with a send button R13, and when this send button R13 is pressed, the extraction range (for example, companies, stores, and analysis period) and the target category are transmitted from terminal 1 to information processing device 2.

[0051] Next, an example of the processing related to Example 1 will be explained with reference to Figure 8. Figure 8 is a sequence diagram showing an example of the processing flow related to Example 1. Figure 9 is an example of a prompt sent in step S223 of Figure 8. Figure 10 is an example of a response sent in step S312 of Figure 8. Figure 11 is an example of a prompt sent in step S224 of Figure 8. Figure 12 is an example of a response sent in step S314 of Figure 8. The processing of the processor of each device will be explained below with reference to Figure 8, but the processors will be omitted to avoid redundancy.

[0052] (Step S111) Terminal 1-i (where i is an integer from 1 to N) accepts the specification of the extraction range, such as companies, stores, and analysis period, and the target category.

[0053] (Step S112) Terminal 1-i then transmits the received extraction range and target category to the information processing device 2.

[0054] (Step S221) When the information processing device 2 receives the extraction range and target category, it reads the sentiment phrase group corresponding to the specified extraction range from the storage device 23. In this way, the processor 26 extracts the sentiment phrase group corresponding to the extraction range specified by the user from the sentiment phrases acquired in the acquisition procedure described above. This makes it easy to obtain an objective evaluation of the company or store and / or objective improvement suggestions based on the sentiment phrase group corresponding to the extraction range specified by the user.

[0055] (Step S222) The information processing device 2 then refers to the industry dictionary of the company of the user using terminal 1-i and classifies each extracted sentiment phrase into a category. Specifically, for example, the information processing device 2 classifies each sentiment phrase into a subcategory and then into a major category.

[0056] (Step S223) The information processing device 2 then sends a request prompt to the generating AI device 3, for example, requesting that it output an evaluation and / or improvement suggestion for a company or store in accordance with the QSCVHA evaluation using a group of emotional phrases belonging to the target category. The QSCVHA evaluation is an evaluation of Quality, Service, Cleanliness, Value, Hospitality, and Atmosphere. Figure 9 shows an example prompt P1 for step S223. Thus, the request prompt is a prompt that requests that it output an evaluation and / or improvement suggestion for a company or store using the group of emotional phrases extracted in step S221.

[0057] (Step S311) When the generating AI device 3 receives the request prompt sent in step S223, it processes the request prompt and generates a QSCVHA evaluation and / or improvement suggestion as a response.

[0058] (Step S312) The generating AI device 3 then sends the generated QSCVHA evaluation and / or improvement suggestion as a response to the information processing device 2. Figure 10 shows an example response A1 sent in step S312.

[0059] (Step S224) When the information processing device 2 receives a QSCVHA evaluation and / or improvement suggestion, it sends a request prompt requesting that the device output an evaluation and / or improvement suggestion for the company or store in accordance with a SWOT analysis, for example, using a group of sentiment phrases belonging to the target category. SWOT analysis is a factor analysis of a company's external and internal environment using four elements: Strength, Weakness, Opportunity, and Threat. Figure 11 shows an example prompt P2 for step S224.

[0060] (Step S313) When the generating AI device 3 receives the request prompt sent in step S224, it processes the request prompt and generates SWOT analysis results and / or improvement suggestions as a response.

[0061] (Step S314) The generating AI device 3 then sends the generated SWOT analysis results and / or improvement suggestions as a response to the information processing device 2. Figure 12 shows an example response A2 that is sent in step S314.

[0062] (Step S225) When the information processing device 2 receives SWOT analysis results and / or improvement suggestions, it generates output information including, for example, QSCVHA evaluation and / or improvement suggestions and SWOT analysis results and / or improvement suggestions.

[0063] (Step S226) The information processing device 2 then transmits the output information to terminal 1-i.

[0064] (Step S113) When terminal 1-i receives output information, it displays this output information. This allows the user of terminal 1-i to obtain, for example, QSCVHA evaluations and / or improvement suggestions and SWOT analysis results and / or improvement suggestions generated based on review submissions, and to easily obtain objective evaluations and / or objective improvement suggestions for the user's company or store.

[0065] Thus, the processor 26 executes a procedure to request the sentiment analysis engine 41 to obtain sentiment phrases, which are predicates containing sentiment expressions and phrases containing the target words of those sentiment expressions, from review posts about a company or store. The processor 26 then executes a procedure to output a request prompt to the generating AI device 3, requesting that it output an evaluation and / or improvement suggestion for the company or store using the obtained sentiment phrases. Here, the request prompt may be a prompt requesting that it output an evaluation and / or improvement suggestion for the company or store according to at least one framework. This makes it possible to obtain an evaluation and / or improvement suggestion according to the framework. The processor 26 then executes a procedure to output output information using the response output from the generating AI device 3 in response to the request prompt. With this configuration, objective evaluations and / or objective suggestions for a company or store can be easily obtained based on review posts using the generating AI device 3, thus reducing the effort required to utilize review posts.

[0066] <Example 2> Next, we will describe Example 2. In Example 2, the generating AI device 3 is requested to output the trend of evaluation by comparing the results of evaluations of the company or store according to at least one framework for each of the first and second periods specified by the user. Figure 13 is an example of the segment input screen displayed on terminal 1 in Example 2. As shown in Figure 13, the segment input screen G2 is provided with a select box R21 for specifying the extraction range from review posts. This extraction range is specified by at least one combination of, for example, company, store, analysis period (e.g., year and month), category (e.g., major category or medium category), and language. Here, as an example, we will explain assuming that company and store are specified as the extraction range. The segment input screen G2 is also provided with a select box R22 for specifying the first and second periods to be compared. Here, as an example, we will explain assuming that 2023 is specified as the first period and 2024 is specified as the second period.

[0067] Furthermore, the segment input screen G2 is provided with a select box R23 for specifying the category to be analyzed (hereinafter referred to as the target category, in this example being a major category). The segment input screen G2 is also provided with a send button R24, and when this send button R24 is pressed, the extraction range (for example, companies and stores), the first period, the second period, and the target category are transmitted from terminal 1 to information processing device 2.

[0068] Next, an example of the processing according to Example 2 will be explained with reference to Figure 14. Figure 14 is a sequence diagram showing an example of the processing flow according to Example 2. Figure 15 is an example of a prompt sent in step S233 of Figure 14. Figure 16 is an example of a response sent in step S322 of Figure 14. Figure 17 is an example of a prompt sent in step S234 of Figure 14. Figure 18 is an example of a response sent in step S324 of Figure 14. Figure 19 is an example of a prompt sent in step S235 of Figure 14. Figure 20 is an example of a response sent in step S326 of Figure 14. The processing of the processor of each device will be explained below with reference to Figure 14, but the processor will be omitted to avoid redundancy.

[0069] (Step S121) Terminal 1-i accepts the specification of the extraction range (e.g., companies and stores), the first period, the second period, and the target category.

[0070] (Step S122) Terminal 1-i then transmits the received extraction range (e.g., companies and stores), first period (e.g., 2023), second period (e.g., 2024), and target category to the information processing device 2.

[0071] (Step S231) When the information processing device 2 receives the extraction range (e.g., companies and stores), the first period, the second period, and the target category, it extracts the sentiment phrase group corresponding to the first period (e.g., 2023) and the sentiment phrase group corresponding to the second period (e.g., 2024).

[0072] (Step S232) The information processing device 2 then refers to the industry dictionary of the user's company and classifies each sentiment phrase for the first period (e.g., 2023) into a category (e.g., a subcategory) and each sentiment phrase for the second period (e.g., 2024) into a category (e.g., a subcategory).

[0073] (Step S233) Next, the information processing device 2 sends a request prompt to the generating AI device 3, for example, requesting that it output an evaluation and / or improvement suggestion for a company or store in accordance with the QSCVHA evaluation using the sentiment phrase group for the first period (e.g., 2023) belonging to the target category (e.g., goods and services) received in step S231. Figure 15 shows an example of the request prompt P3 in step S233.

[0074] (Step S321) When the generating AI device 3 receives the request prompt sent in step S233, it processes the request prompt and outputs a QSCVHA evaluation and / or improvement suggestion as a response.

[0075] (Step S322) The generating AI device 3 then sends the outputted QSCVHA evaluation and / or improvement suggestions to the information processing device 2 as a response for the first period (e.g., 2023). Figure 16 shows an example response A3 for the first period (an example response generated using the sentiment phrases for 2023) that is sent in step S322.

[0076] (Step S234) When the information processing device 2 receives a sample response for the first period (e.g., 2023), it sends a request prompt to the generating AI device 3 requesting that it output an evaluation and / or improvement suggestion for the company or store in accordance with the QSCVHA evaluation using a group of sentiment phrases for the second period (e.g., 2024) belonging to the target category (e.g., goods and services) received in step S231. Figure 17 shows an example of the request prompt P4 in step S234.

[0077] (Step S323) When the generating AI device 3 receives the request prompt sent in step S234, it processes the request prompt and outputs a QSCVHA evaluation and / or improvement suggestion as a response.

[0078] (Step S324) The generating AI device 3 then sends the outputted QSCVHA evaluation and / or improvement suggestions to the information processing device 2 as the answer for the second period (e.g., 2024). Figure 18 shows an example answer A4 for the second period (an example answer generated using the sentiment phrases for 2024) that is sent in step S324.

[0079] (Step S235) When the information processing device 2 receives the answer example for the second period (e.g., 2024), it sends a request prompt to the generating AI device 3 requesting the evaluation progress. Figure 19 shows an example of the request prompt P5 in step S235.

[0080] (Step S325) When the generating AI device 3 receives the request prompt sent in step S235, it processes the request prompt and outputs the evaluation progress as the answer.

[0081] (Step S326) The generating AI device 3 then sends the outputted evaluation trends as a response to the information processing device 2. Figure 20 shows an example of response A5 sent in step S326 (for example, the evaluation trends for each indicator of QSCVHA and an overall summary of the evaluation trends).

[0082] (Step S236) When the information processing device 2 receives the evaluation progress, it generates output information that includes this evaluation progress.

[0083] (Step S237) The information processing device 2 then transmits the output information to terminal 1-i.

[0084] (Step S123) When terminal 1-i receives output information, it displays this output information. This allows the user of terminal 1-i to obtain, for example, the evaluation trends for each QSCVHA indicator from the first period to the second period and / or an overall evaluation of the evaluation trends, and to easily obtain the evaluation trends of the user's company or store.

[0085] In this embodiment, the QSCVHA evaluation was used as an example of a framework, but it is not limited to this, and other frameworks may be used, as long as there is at least one framework. That is, the request prompt may be a prompt that requests the results of an evaluation of the company or store according to at least one framework for each of the first and second periods specified by the user. In this case, the processor 26 further executes a procedure to output a prompt to the generating AI device 3 that requests the output of the evaluation transition using the prompt processing results for the first period and the prompt processing results for the second period. This makes it possible to obtain the evaluation transition from the first period to the second period according to at least one framework, and to easily obtain the evaluation transition of the user's company or store.

[0086] <Variation> In this embodiment, it has been described that a dictionary associating target words and categories of emotional expressions is stored in a memory device for each industry, but this is not the only example. A dictionary associating target words and categories may be stored in a memory device for each company. In this case, in the procedure for classifying into one of the multiple categories, the processor 26 may refer to the dictionary corresponding to the user's company and classify the acquired emotional phrase or the target word in the emotional phrase into one of the multiple categories provided for analysis. By doing so, since a dictionary provided for each company is referred to, the emotional phrase or the target word in the emotional phrase can be classified into an appropriate category.

[0087] Furthermore, at least a part of the information processing device 2 described in the above-mentioned embodiment may be configured as hardware or as software. In the case of software configuration, a program that realizes at least a part of the functions of the information processing device 2 may be stored on a computer-readable recording medium and loaded and executed by a computer. The recording medium is not limited to removable ones such as magnetic disks or optical disks, but may also be a fixed recording medium such as a hard disk drive or memory.

[0088] Furthermore, a program that implements at least some of the functions of the information processing device 2 may be distributed via communication lines such as the Internet (including wireless communication). In addition, the program may be encrypted, modulated, or compressed and distributed via wired or wireless lines such as the Internet, or stored on a recording medium.

[0089] Furthermore, the information processing device 2 may be made to function using one or more information devices. When multiple information devices are used, at least one of them may be a computer, and the computer may execute a predetermined program to realize the function of at least one means of the information processing device 2.

[0090] Furthermore, in the invention of a method, all steps may be automatically controlled by a computer. Alternatively, each step may be performed by a computer while the progress between steps is controlled manually. Furthermore, at least a portion of all steps may be performed manually.

[0091] As described above, the present invention is not limited to the embodiments described above, and the components can be modified and implemented in practice without departing from the spirit of the invention. Furthermore, various inventions can be formed by appropriately combining the multiple components disclosed in the above embodiments. For example, some components may be deleted from all the components shown in the embodiments. Moreover, components from different embodiments may be appropriately combined. [Explanation of Symbols]

[0092] 1-1, ..., 1-N terminals 2. Information Processing Device 21 Input Interfaces 22 Communication Module 23 Storage device 24 memory 25 Output Interfaces 26 processors 3 Generation AI device S Information Processing System

Claims

1. An information processing device that can exchange information with a generation AI device that executes generation AI that processes prompts and outputs information, Equipped with at least one processor, The processor includes a procedure for requesting an emotion analysis engine to obtain an emotion phrase, which is a predicate containing an emotion expression and a phrase containing the target word of the emotion expression, from a review post about a company or store. A procedure for outputting a request prompt to the generating AI device requesting it to output an evaluation and / or improvement suggestion for the company or store using the acquired emotion phrase, A procedure for outputting output information using the response output from the generating AI device in response to the aforementioned request prompt, An information processing device that performs the following actions.

2. The processor further performs a procedure to classify the acquired emotion phrase or the target word of the emotion expression within the emotion phrase into one of several categories established for analysis. The aforementioned request prompt is a prompt that requests the output of evaluations and / or improvement suggestions for the company or store for a target category specified by the user. The information processing apparatus according to claim 1.

3. A dictionary linking emotional expression terms and categories is stored in memory for each industry. In the classification procedure described above, the processor refers to the dictionary corresponding to the user's industry and classifies the acquired sentiment phrase or the target word of the sentiment expression within the sentiment phrase into one of several categories provided for analysis. The request prompt is a prompt that requests the output of an evaluation and / or improvement suggestion for a company or store for the target category, using the emotional phrase classified in the target category or the target word of the emotional expression in the emotional phrase. The information processing apparatus according to claim 2.

4. The aforementioned memory device stores a second category, which has a narrower scope than the first category, associated with each first category, and stores a dictionary for each industry, which associates the target words for emotional expression with the second category. In the classification procedure described above, the processor refers to the dictionary corresponding to the user's company and classifies the acquired sentiment phrase or the target word in the sentiment phrase into one of the plurality of second categories in order to classify the sentiment phrase or the target word in the sentiment phrase into a first category. The information processing apparatus according to claim 3.

5. The processor further performs a procedure to extract a group of sentiment phrases that fall within the extraction range specified by the user from the sentiment phrases obtained in the acquisition procedure. The request prompt is a prompt that requests the output of an evaluation and / or improvement suggestion for the company or store using the acquired sentiment phrases. The information processing apparatus according to any one of claims 1 to 4.

6. The request prompt is a prompt that requests the output of an evaluation and / or improvement suggestion for the company or store in accordance with at least one framework. The information processing apparatus according to any one of claims 1 to 4.

7. The aforementioned request prompt is a prompt that requests the results of an evaluation of the company or store in accordance with at least one framework for each of the first and second periods specified by the user, The processor further executes a procedure to output a prompt to the generating AI device requesting that the AI ​​output the evaluation progress using the prompt processing results for the first period and the prompt processing results for the second period. The information processing apparatus according to any one of claims 1 to 4.

8. A dictionary linking target words and categories is stored in a memory device for each company. In the classification procedure described above, the processor refers to the dictionary corresponding to the user's company and classifies the acquired sentiment phrase or the target word within the sentiment phrase into one of several categories provided for analysis. The information processing apparatus according to claim 2.

9. The acquisition method is a procedure that requests an emotion analysis engine to acquire an emotion phrase, which is a predicate containing an emotion expression and a phrase containing the target word of that emotion expression, from a review post about a company or store. A procedure in which the first output means outputs a request prompt to a generating AI device requesting it to output an evaluation and / or improvement suggestion for the company or the store using the acquired sentiment phrase, The second output means outputs output information using the response output from the generating AI device in response to the request prompt. An information processing method having

10. A computer capable of exchanging information with a generative AI device that executes generative AI to process prompts and output information, A procedure for obtaining sentiment phrases, which are predicates containing sentiment expressions and phrases containing the target words of those sentiment expressions, from review posts about a company or store by requesting them from a sentiment analysis engine. A procedure for outputting a request prompt to the generating AI device requesting it to output an evaluation and / or improvement suggestion for the company or store using the acquired emotion phrase, A procedure for outputting output information using the response output from the generating AI device in response to the aforementioned request prompt, A program to execute.

Images