System

Abstract

A system is provided.SOLUTION: A system comprising: means for automatically acquiring comment data from a questionnaire system; means for pre-processing the comment data; means for summarizing the pre-processed comment data; means for analyzing an emotion of the summarized comment data; means for accumulating the analysis result in a database; means for analyzing a transition with past data based on the accumulated data; means for analyzing a tendency of an individual respondent; and means for presenting a next action from the analysis result.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] Describe the "problem that the invention aims to solve" and the "means for solving the problem."

[0005] While conventional survey systems automate the compilation and visualization of quantitative questions, the compilation and analysis of qualitative comments is still done manually, which requires a lot of time and effort. Furthermore, there is no way to effectively analyze large amounts of comments and suggest next steps, making continuous data analysis and rapid responses to organizational issues difficult. Therefore, there is a need for a system that can efficiently automatically compile qualitative data and analyze sentiment, and quickly suggest next steps based on the results. [Means for solving the problem]

[0006] The present invention solves the above-mentioned problems by providing a system including: means for automatically acquiring comment data from a survey system; means for preprocessing the comment data; means for summarizing the preprocessed comment data; means for analyzing the sentiment of the summarized comment data; means for storing the analysis results in a database; means for analyzing trends between the accumulated data and past data; means for analyzing trends for individual respondents; and means for suggesting next actions based on the analysis results. Specifically, comment data is automatically acquired from the survey system via an API, and preprocessed by removing unnecessary spaces and special characters. Then, a generative AI model is used to summarize and analyze the sentiment of the comments, and the analysis results are stored in a database. Furthermore, the accumulated data is compared with past data to analyze trends, and trends are analyzed based on the data of individual respondents, thereby identifying management candidates and identifying members who require follow-up. Finally, next actions are automatically suggested based on these analysis results, enabling quick and effective problem solving.

[0007] Understood. Below are definitions of important terms contained in the claims.

[0008] "Survey System" means a platform for collecting comment data, quantitative data, and other survey information.

[0009] "Comment data" is text information written in a free-form field in a questionnaire, and represents the respondent's specific opinions and impressions.

[0010] "Preprocessing" refers to the process of removing unnecessary spaces, special characters, and emojis from the acquired comment data and standardizing the text format.

[0011] "Summarization" is the process of using a generative AI model to concisely summarize the key content of comment data.

[0012] "Sentiment analysis" is the process of analyzing the emotional nuances of comment data and classifying them into categories such as positive, negative, and neutral.

[0013] A "database" is a system for centrally managing and storing acquired and analyzed data.

[0014] "Transition analysis" is the process of comparing accumulated data with past data to analyze changes and trends over time.

[0015] "Individual respondent trend analysis" is the process of analyzing data from specific respondents to understand their characteristics and patterns.

[0016] "Next action suggestion" is a process that automatically suggests specific actions or measures to be taken next based on the analysis results.

[0017] An "API" is an interface that enables data exchange between different systems, and is a means of obtaining data through HTTP requests, etc.

[0018] A "generative AI model" is an artificial intelligence algorithm that uses natural language processing technology to analyze text data and perform summarization and sentiment analysis. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5]FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0020] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0021] First, the terms used in the following description will be explained.

[0022] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0023] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0024] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0025] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0026] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0027] [First embodiment]

[0028] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0029] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0030] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0031] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0032] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0033] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0034] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0035] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0036] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0037] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0038] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0039] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0040] Understood. Below is the "Mode for Carrying Out the Invention" of the specification.

[0041] This invention is a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program of this system operates based on the operations of the server, terminal, and user.

[0042] First, the server automatically retrieves comment data from the survey system. For example, to retrieve survey data for the department monthly meeting in August 2023, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data.

[0043] Next, the server preprocesses the acquired comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the accuracy of the analysis.

[0044] The server then uses a generative AI model to summarize the preprocessed comment data. For example, it converts a comment such as "The meeting was progressing slowly" into a summary statement such as "The delay in the progress was pointed out." The server then uses a sentiment analysis tool to classify the sentiment of the comment as "negative."

[0045] The server then stores the generated summary and the sentiment analysis results in a database. The stored data includes the original comment, summary, sentiment analysis results, and timestamp.

[0046] Based on the accumulated data, the server compares it with past data and analyzes trends. For example, it compares it with data from the past few months to see if there has been an increase in negative comments. This trend analysis allows you to understand trends and take necessary measures.

[0047] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, the server extracts this tendency from the database and identifies potential management positions or employees who need follow-up.

[0048] Finally, the server will suggest the next action based on the results of these analyses, such as "proposing holding a workshop to improve progress" or "recommending feedback follow-up for specific employees," and notify managers of specific actions.

[0049] The terminal provides an interface for users to view survey results and analytical data. Through the terminal, users can view summaries of quantitative and qualitative data, visualize the results in graphs and dashboards, and receive real-time notifications of suggested next steps.

[0050] Based on the information provided on the device, the user can plan and execute specific follow-up actions and measures. For example, the user can decide to hold a workshop and send a notification to employees.

[0051] As a result, the present invention enables rapid collection and analysis of qualitative data, leading to efficient business operations and effective human resource management.

[0052] The processing flow will be explained below.

[0053] Understood. Below are the specific steps of the process.

[0054] Step 1:

[0055] The server automatically retrieves comment data from the survey system by sending an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[0056] Step 2:

[0057] The server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[0058] Step 3:

[0059] The server then uses a generative AI model to summarize the preprocessed comment data. For example, it converts long comments into short summary sentences. This AI model uses natural language processing technology to concisely express the main content of the comments.

[0060] Step 4:

[0061] The server uses a sentiment analysis tool to analyze the sentiment of the comments. Specifically, it categorizes them into categories such as positive, negative, and neutral. For example, a comment such as "The meeting went slowly" would be classified as "negative."

[0062] Step 5:

[0063] The server stores the generated summaries and the results of sentiment analysis in a database. The stored data includes the original comments, summaries, sentiment analysis results, and timestamps. This allows the analysis results to be stored centrally in the database.

[0064] Step 6:

[0065] The server analyzes the trends in past data based on the accumulated data, comparing data from multiple months and analyzing trends in sentiment and opinions to determine whether negative comments are increasing at a particular time.

[0066] Step 7:

[0067] The server analyzes the trends of individual respondents. Based on the respondent ID, it extracts data for specific respondents and analyzes their trends. This process identifies the characteristics and patterns of respondents and identifies management candidates and employees who need follow-up.

[0068] Step 8:

[0069] The server then proposes the next action based on the analysis results, for example, generating specific action suggestions such as "proposing to hold a workshop on improving progress" or "recommending feedback follow-up for specific employees," and notifying the manager.

[0070] Step 9:

[0071] The terminal provides a user interface that displays survey results and analytical data to the user, allowing the user to view summaries of quantitative and qualitative data and visualize the results in the form of graphs and dashboards.

[0072] Step 10:

[0073] The user plans and executes specific follow-up actions and measures based on the information provided on the device. For example, the user decides to hold a workshop and sends a notification to employees.

[0074] Based on the above steps, the present invention enables rapid collection and analysis of qualitative data, thereby realizing efficient business operations and human resource management.

[0075] Example 1

[0076] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0077] Conventional survey analysis systems have struggled to efficiently process and analyze large amounts of comment data and propose effective actions based on that data. Furthermore, manually processing data consumes significant human resources and results in inconsistent analytical accuracy. There is a need for a system that can resolve these issues, quickly aggregate and analyze qualitative data, and achieve efficient business operations and effective human resource management.

[0078] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0079] In this invention, the server includes means for automatically acquiring comment data from a survey system, means for preprocessing the comment data, means for using a generative AI model to summarize the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing tendencies of individual respondents, and means for suggesting next actions based on the analysis results. This makes it possible to automatically process large amounts of comment data and propose specific actions based on highly accurate analysis results.

[0080] A "survey system" is an electronic survey device for collecting opinions and feedback from users.

[0081] "Comment data" is text data that includes user opinions and feedback collected through a questionnaire system.

[0082] "Preprocessing" is the procedure of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format.

[0083] A "generative AI model" is an artificial intelligence model that uses natural language processing to summarize, generate, or analyze text data.

[0084] "Sentiment analysis" is a method for automatically classifying or assessing the emotions contained in text data.

[0085] A "database" is an electronic system for efficiently storing, managing, and retrieving organized data.

[0086] "Transition analysis" is a method of analyzing temporal changes and trends based on accumulated data.

[0087] "Trend analysis" is a method of analyzing the behavioral patterns and opinion trends of specific users or groups.

[0088] "Action suggestion" is the process of suggesting the next course of action based on the results of data analysis.

[0089] This system automatically preprocesses comment data obtained from a survey system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. This system operates based on operations from the server, terminals, and users.

[0090] First, the server automatically retrieves comment data from the survey system. Specifically, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, to retrieve survey data for the department monthly meeting in August 2023, the server uses the API to collect the data. The retrieved data is saved in JSON format or similar.

[0091] The server then preprocesses the retrieved comment data. Specifically, it uses Python and natural language processing (NLP) tools (e.g., NLTK and SpaCy) to remove unnecessary spaces, special characters, and emojis from the comments and standardize the text format. This preprocessing improves the analysis accuracy of the generative AI model.

[0092] After preprocessing the comment data, the server uses a generative AI model (e.g., GPT-4) to summarize the comments. An example prompt is shown below.

[0093] Example prompt sentence:

[0094] Original comment:

[0095] "The meeting was slow"

[0096] Preprocessed text:

[0097] "Slow progress pointed out"

[0098] Generate a summary:

[0099] When this prompt is fed into a generative AI model, the model generates the summary statement, "Delays in progress have been noted."

[0100] The server also performs sentiment analysis on the generated summary using tools such as TextBlob and VADER. For example, a comment such as "Delays in progress have been noted" is classified as "negative."

[0101] The server then stores the generated summary and the results of sentiment analysis in a database. The data to be saved includes the original comment, summary, sentiment analysis results, and timestamp. The database can be MySQL or PostgreSQL.

[0102] The server analyzes the accumulated data by comparing it with past data. For example, it analyzes whether there has been an increase in negative comments compared to data from the past few months. This analysis is performed using statistical analysis libraries such as pandas and numpy in Python.

[0103] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, the server can extract this tendency from the database. This analysis can then be used to identify potential management candidates or employees who need follow-up.

[0104] Finally, the server will suggest the next action based on the results of these analyses, such as "Suggest holding a workshop to improve progress" or "Recommend follow-up on feedback from specific employees," and notify managers of specific actions. This notification can be sent via email or via a dashboard within the system.

[0105] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and view the results visualized in graphs and dashboards. For example, graphs can be generated using JavaScript D3.js and Chart.js, allowing users to interactively check the analysis results. It can also display suggested next actions with real-time notifications.

[0106] Users can plan and implement specific follow-up actions and measures based on the information provided on their devices. For example, they can schedule workshops to improve progress and send notifications to employees. In this way, efficient business operations and effective human resource management are realized.

[0107] As described above, the present invention is a system that enables rapid collection and analysis of qualitative data and aims to improve and streamline business operations by proposing specific actions.

[0108] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0109] Step 1:

[0110] Data Acquisition

[0111] The server sends an HTTP request to the API endpoint to automatically retrieve comment data from the survey system. For example, the server sends a GET request to "https: / / api.example.com / surveys / latest" and receives JSON-formatted comment data as a response. The input is the API endpoint and authentication information (e.g., OAuth token), and the output is the retrieved JSON-formatted comment data. Specifically, data is retrieved from the API using a Python library (such as requests).

[0112] Step 2:

[0113] Data Preprocessing

[0114] The server preprocesses the comment data it receives. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments, and converts all text to a unified format (for example, lowercase). This process uses Python's regular expression module (re) and natural language processing libraries (NLTK and SpaCy). The input is comment data in JSON format, and the output is preprocessed string data. The specific operation is performed using the following code:

[0115] python

[0116] import re

[0117] import spacy

[0118] def preprocess_comment(comment):

[0119] comment = re.sub(r'[^\w\s]', '', comment) Remove special characters

[0120] comment = comment.lower() Convert all to lowercase

[0121] return comment

[0122] Step 3:

[0123] Summary Generation

[0124] The server inputs the preprocessed comment data into a generative AI model (e.g., GPT-4) to generate a summary of the comments. The prompt used here is:

[0125] input:

[0126] Original comment:

[0127] "The meeting was slow"

[0128] Preprocessed text:

[0129] "Slow progress pointed out"

[0130] Generate a summary:

[0131] The output is a summary sentence returned from the generative AI model. Specifically, the prompt sentence is sent to the API of the generative AI model and the summary sentence is received.

[0132] Step 4:

[0133] sentiment analysis

[0134] The server performs sentiment analysis on the generated summary. For sentiment analysis, it uses natural language processing tools such as TextBlob and VADER. The input is the generated summary, and the output is the sentiment classification result of the summary (e.g., negative, positive, neutral). To see the specific operation, use the following code:

[0135] python

[0136] from textblob import TextBlob

[0137] def analyze_sentiment(text):

[0138] analysis = TextBlob(text)

[0139] if analysis.sentiment.polarity > 0:

[0140] return "affirmative"

[0141] elif analysis.sentiment.polarity == 0:

[0142] return "neutral"

[0143] else:

[0144] return "negative"

[0145] Step 5:

[0146] Saving to a database

[0147] The server stores the original comment, the generated summary, the sentiment analysis result, and the timestamp in a database. The input is the original comment, the summary, the sentiment result, and the timestamp, and the output is an entry stored in the database. Specifically, the data is stored in the database using an SQL insert query:

[0148] sql

[0149] INSERT INTO comments (original_comment, summary, sentiment, timestamp)

[0150] VALUES (%s, %s, %s, %s);

[0151] Step 6:

[0152] Trend analysis

[0153] The server analyzes the trends of past data based on the accumulated data. The input is past comment data retrieved from the database, and the output is the results of the trend analysis. Specifically, it performs statistical analysis using the Python pandas library:

[0154] python

[0155] import pandas as pd

[0156] def trend_analysis(comments):

[0157] df = pd.DataFrame(comments)

[0158] trend = df['sentiment'].value_counts()

[0159] return trend

[0160] Step 7:

[0161] Analysis of user trends

[0162] The server analyzes the trends of individual respondents. The input is comment data related to a specific respondent, and the output is the specific trend analysis results. Specifically, it extracts data for a specific respondent by filtering and performs analysis:

[0163] python

[0164] def user_trend_analysis(comments, user_id):

[0165] user_comments = [comment for comment in comments if comment['user_id'] == user_id]

[0166] trend = trend_analysis(user_comments)

[0167] return trend

[0168] Step 8:

[0169] Suggested next actions

[0170] The server proposes the next action based on the analysis results. The input is the transition analysis results and trend analysis results, and the output is a specific action proposal (e.g., a proposal to hold a workshop). Specifically, it performs conditional branching based on the analysis results to generate appropriate actions and notify management:

[0171] python

[0172] def suggest_next_action(trend):

[0173] if trend['negative'] > trend['positive']:

[0174] return "Proposal to hold a workshop on improving progress"

[0175] else:

[0176] return "No follow-up required"

[0177] The above is the processing steps of the system and the specific operations that accompany them.

[0178] (Application example 1)

[0179] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0180] In modern brick-and-mortar stores, efficient staff management and improved service quality are key challenges. Traditional management methods make it difficult to properly collect and analyze feedback from staff, and they are not expected to take appropriate action quickly. Furthermore, it is necessary to accurately summarize the content of feedback and perform sentiment analysis to grasp trends and provide accurate notifications to managers.

[0181] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0182] In this invention, the server includes means for automatically acquiring comment data from a survey system, means for preprocessing the comment data, means for using a generative AI model to summarize the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends from past data based on the stored data, means for analyzing trends of individual respondents, means for suggesting next actions based on the analysis results, and means for visualizing feedback trends using the analysis results and notifying management. This enables efficient collection and analysis of staff feedback and prompt suggestion of appropriate next actions.

[0183] A "survey system" is a system for collecting feedback and opinions from users.

[0184] "Comment Data" refers to opinions and feedback in text form obtained from users through the survey system.

[0185] "Preprocessing" refers to the process of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format to make data analysis easier.

[0186] A "generative AI model" is an artificial intelligence algorithm that learns from large amounts of data and generates new sentences.

[0187] "Sentiment analysis" is a technique that analyzes the content of text data and determines whether the sentiment is positive, negative, or neutral.

[0188] A "database" is a digital storage system for systematically organizing and storing information.

[0189] "Feedback trends" are patterns and tendencies that emerge from the analysis of accumulated feedback data.

[0190] A "management position" is a position within an organization that is responsible for managing staff and running business operations.

[0191] An "HTTP request" is a communication protocol for requesting data from a web server.

[0192] An "API endpoint" is an interface for connecting with external systems and is a connection point for exchanging data.

[0193] This invention is a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program of this system operates based on the operations of the server, terminal, and user.

[0194] First, the server automatically retrieves comment data from the survey system. For example, it sends an HTTP request to an API endpoint to retrieve the latest survey data. The retrieved comment data is preprocessed to remove spaces, special characters, and emojis, and standardize the text format.

[0195] Next, a generative AI model is used on the preprocessed comment data to summarize the comments. For example, a comment such as "The customer service today was terrible" is converted into a summary sentence such as "There was a problem with the service." A sentiment analysis tool is also used to classify the sentiment of this comment as "negative."

[0196] The generated summary and the results of the sentiment analysis are stored in a database. This data includes the original comment, summary, sentiment analysis results, and timestamp. Based on the stored data, the server compares it with past data to analyze trends. For example, it analyzes whether there has been an increase in negative comments compared to data from the past few months. This trend analysis allows trends to be identified and necessary measures to be taken.

[0197] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, it extracts that tendency from the database and identifies staff members who need follow-up. Finally, based on these analysis results, the server proposes the next action to take and notifies management. For example, it may notify specific actions such as "proposing holding a workshop to improve poor customer service" or "recommending follow-up with specific staff members."

[0198] The terminal provides an interface for users to check survey results and analytical data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. Suggested next actions are also notified in real time. For example, specific feedback such as "Today's shift was very busy, but enjoyable" or "The handling of customer complaints did not go well" is displayed in a list. An example of a prompt is "Generate a summary of the following feedback comments: 'Today's shift was very busy, but enjoyable.'"

[0199] As a result, the present invention enables rapid collection and analysis of qualitative data, leading to efficient business operations and effective human resource management.

[0200] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0201] Step 1:

[0202] The server automatically retrieves comment data from the survey system. Specifically, it sends an HTTP request to a pre-configured API endpoint to retrieve the survey data. The input is the API endpoint, and the output is the retrieved comment data.

[0203] Step 2:

[0204] The server preprocesses the acquired comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. The input is the acquired comment data, and the output is the preprocessed comment data.

[0205] Step 3:

[0206] The server summarizes the preprocessed comment data using a generative AI model. Specifically, it converts the comment text into a summary sentence. For example, a comment such as "The customer service today was terrible" is converted into the summary sentence "There was a problem with the service." The input is the preprocessed comment data, and the output is the summarized text data.

[0207] Step 4:

[0208] The server performs sentiment analysis on the summarized comment data. Specifically, it uses a sentiment analysis tool to classify the sentiment of the comment as positive, negative, or neutral. For example, a comment that says "there was a problem with the response" is classified as "negative." The input is the summarized comment data, and the output is the result of the sentiment analysis.

[0209] Step 5:

[0210] The server stores the generated summary and the results of sentiment analysis in a database. Specifically, it saves the original text of the comment, the summarized text, the results of sentiment analysis, and a timestamp. The input is the summary and the results of sentiment analysis, and the output is the information stored in the database.

[0211] Step 6:

[0212] The server analyzes the trends in past data based on the accumulated data. Specifically, it extracts data from the past few months and analyzes whether there has been an increase in negative comments. The input is the past data accumulated in the database, and the output is the results of the trend analysis.

[0213] Step 7:

[0214] The server analyzes the tendencies of individual respondents. Specifically, if a particular respondent repeatedly makes negative comments, it extracts that tendency and identifies staff members who need follow-up. The input is the comment data stored in the database and the analysis results, and the output is a list of staff members who need follow-up.

[0215] Step 8:

[0216] The server proposes the next action based on the analysis results and notifies the manager. Specifically, it notifies specific actions such as "proposing holding a workshop to improve the worsening response" or "recommending follow-up with specific staff members." The input is the transition analysis results and individual trend analysis results, and the output is the notification content to the manager.

[0217] Step 9:

[0218] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. The input is the survey results and analysis data stored in the database, and the output is the visualized data displayed on the interface.

[0219] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0220] Understood. Below is the "Mode for Carrying Out the Invention" of the specification based on the invention combining the emotion engine.

[0221] This invention combines an emotion engine with a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program for this system operates based on operations by the server, terminal, and user.

[0222] First, the server automatically retrieves comment data from the survey system. To do this, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[0223] Next, the server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[0224] The server then uses a generative AI model to summarize the preprocessed comment data, for example, converting long comments into short summary sentences. This AI model uses natural language processing techniques to concisely express the main content of the comment.

[0225] The server then analyzes the sentiment of the comments using a sentiment analysis tool and a sentiment engine. Specifically, the sentiment engine categorizes the sentiment into categories such as positive, negative, and neutral. For example, a comment such as "The meeting went slowly" would be classified as "negative."

[0226] The server then stores the generated summary and the results of the sentiment analysis in a database. The data stored includes the original comment, summary, sentiment analysis results, and timestamp. This allows the analysis results to be stored centrally in the database.

[0227] Based on the accumulated data, the server compares it with past data and analyzes trends. By comparing data from multiple months and analyzing trends in sentiment and opinions, it can determine whether there has been an increase in negative comments at a particular time.

[0228] The server then analyzes the trends of individual respondents. Based on the respondent ID, data on specific respondents is extracted and its trends are analyzed. This process identifies the characteristics and patterns of respondents and identifies management candidates and employees who need follow-up.

[0229] Finally, the server generates specific action suggestions based on the results of these analyses, such as "suggest holding a workshop to improve progress" or "recommend following up on feedback from specific employees," and notifies managers.

[0230] The terminal provides an interface for users to view survey results and analytical data. Through the terminal, users can view summaries of quantitative and qualitative data, visualize the results in graphs and dashboards, and receive real-time notifications of suggested next steps.

[0231] The user can plan and implement specific follow-up actions and measures based on the information provided on the device. For example, the user can decide to hold a workshop and send a notification to employees.

[0232] This allows for the rapid collection and analysis of qualitative data, resulting in efficient business operations and human resource management. Furthermore, by combining it with an emotion engine, it is possible to recognize users' emotions and provide specific feedback and follow-up actions based on those emotions.

[0233] The processing flow will be explained below.

[0234] Understood. Below are the specific operations for each processing step.

[0235] Step 1:

[0236] The server automatically retrieves comment data from the survey system by sending an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[0237] Step 2:

[0238] The server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[0239] Step 3:

[0240] The server uses a generative AI model to summarize the preprocessed comment data. For example, a comment such as "The meeting was progressing slowly. There were probably too many agenda items" is converted into a summary such as "The slow progress and the large number of agenda items were pointed out." This AI model uses natural language processing technology to concisely express the main content of the comment.

[0241] Step 4:

[0242] The server analyzes the sentiment of the comments using a sentiment analysis tool and a sentiment engine. For example, a comment such as "The meeting went slowly" may be classified as "negative." The sentiment engine categorizes the sentiment into categories such as positive, negative, and neutral.

[0243] Step 5:

[0244] The server stores the generated summary and the results of sentiment analysis in a database. The stored data includes the original comment, summary, sentiment analysis results, and timestamp. This allows the analysis results to be stored centrally in the database.

[0245] Step 6:

[0246] The server analyzes the accumulated data and compares it with past data, analyzing trends in sentiment and opinions by comparing data from multiple months, for example, to determine whether negative comments are increasing at certain times of the day or during certain events.

[0247] Step 7:

[0248] The server analyzes the trends of individual respondents. Based on the respondent ID, it extracts the data of a specific respondent and analyzes their trends. For example, if a specific employee frequently makes negative comments, it analyzes the employee's trends and generates a detailed report.

[0249] Step 8:

[0250] The server then proposes the next action based on the analysis results. For example, it generates specific action suggestions such as "Suggest holding a workshop to improve progress" or "Recommend follow-up feedback for specific employees." These suggestions are notified to managers in real time via a notification system.

[0251] Step 9:

[0252] The device provides a user interface that displays survey results and analysis data. Through the device, users can view summaries of quantitative and qualitative data and visualize the results in the form of graphs and dashboards. For example, sentiment analysis results can be displayed using color codes for positive, negative, and neutral.

[0253] Step 10:

[0254] The user plans and implements specific follow-up actions and measures based on the information provided on the device. For example, the user decides to hold a proposed workshop and sends an email notification to employees. This allows the user to take the necessary action quickly.

[0255] Based on the above steps, this invention enables the rapid collection and analysis of qualitative data, realizing efficient business operations and human resource management. Furthermore, by combining it with an emotion engine, it is possible to recognize users' emotions and provide specific feedback and follow-up actions based on those emotions.

[0256] Example 2

[0257] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0258] In conventional survey systems, comment data is often collected and analyzed manually, which is not only inefficient but also makes it difficult to ensure data consistency and quality. Furthermore, because sentiment analysis and summarization are not automated, the vast amount of comment data must be reviewed one by one, placing a heavy burden on the person in charge. Furthermore, the process for linking analysis results to subsequent actions is unclear, making it difficult to respond quickly and appropriately. This creates challenges that make it difficult to optimize organizational operations and human resource management.

[0259] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for automatically acquiring comment data from a questionnaire system, means for preprocessing the comment data, means for summarizing the preprocessed comment data using a generative AI model, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing tendencies of individual respondents, means for suggesting next actions based on the analysis results, and means for displaying the analysis results and next actions through a user interface. This consistently automates the process from comment data collection to summarization, sentiment analysis, data storage, trend analysis, trend analysis, and suggestion of next actions, enabling efficient and accurate data analysis and rapid response.

[0260] A "survey system" is a system for collecting comments and opinions from survey subjects, and has the function of acquiring data in digital form.

[0261] "Comment data" refers to data that indicates opinions and impressions in text format collected through a survey system.

[0262] "Preprocessing" refers to the processing step of removing unnecessary spaces, special symbols, and emojis from the acquired comment data and standardizing the text format.

[0263] A "generative AI model" is a model that uses artificial intelligence techniques to analyze a given text and perform a specific purpose, such as summarization or sentiment analysis.

[0264] "Summarization" refers to the process of summarizing the original comment data in a short and concise form and extracting only the main content.

[0265] "Sentiment analysis" is the process of identifying and classifying sentiments, such as positive, negative, or neutral, from text in comment data.

[0266] A "database" is a collection of information that systematically stores collected and processed data so that it can be easily searched, referenced, and analyzed later.

[0267] "Trend analysis" refers to a statistical process that compares past data with current data to reveal fluctuations and trends.

[0268] "Trend analysis" is an analytical technique used to identify recurring patterns or characteristics associated with a particular respondent or data set.

[0269] "Next action suggestion" refers to the process of proposing specific actions or measures to be taken based on the analysis results.

[0270] "User interface" refers to the visual or operational means or screen configuration that allows a user to operate a system and check the results.

[0271] This system automatically preprocesses comment data obtained from a survey system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, and compares and analyzes them with past data. This system operates based on the operations of the server, terminals, and users.

[0272] Hardware and software used

[0273] The main hardware and software used in the present invention are as follows:

[0274] Server: Performs data acquisition, preprocessing, summary generation, sentiment analysis, data accumulation, analysis, and action suggestion.

[0275] Terminal: Provides an interface for users to view results and take next actions.

[0276] Generative AI models: Use natural language processing techniques to perform text summarization and sentiment analysis.

[0277] System operation explanation

[0278] Retrieving comment data

[0279] The server sends an HTTP request to the API endpoint of the survey system to retrieve the comment data. For example, to retrieve feedback on the monthly department meeting, the server sends the following request:

[0280] plaintext

[0281] GET / api / v1 / feedbacks

[0282] Host: survey.example.com

[0283] Authorization: Bearer<API_TOKEN>

[0284] The server extracts the comment field from the response data and converts it into a format that can be processed within the system.

[0285] Preprocessing comment data

[0286] The server pre-processes the retrieved comment data, mainly performing the following tasks:

[0287] Remove unnecessary spaces and special characters.

[0288] Remove emojis or convert them to text.

[0289] Use the same capitalization for text.

[0290] For example, if the comment is "The meeting was 👍," it will be converted to "The meeting was good." This will improve the data quality of the comment text.

[0291] Summary Generation

[0292] The server uses the generative AI model on the preprocessed comment data to summarize the comments. It gives the generative AI model the following prompts:

[0293] plaintext

[0294] Prompt: "Summarize the following comment: 'The meeting went very smoothly and the whole team was united. The results exceeded our expectations.'"

[0295] The model generates a summary in response to this prompt, returning the summary "The meeting went smoothly and produced good results."

[0296] sentiment analysis

[0297] The server performs sentiment analysis on the summary using the sentiment engine, and then provides the following prompts for the generated summary:

[0298] plaintext

[0299] Prompt: "Analyze the sentiment of the following sentence: 'The meeting went smoothly and produced good results.'"

[0300] The emotion engine classifies emotions into categories such as positive, negative, and neutral, and returns the result as "positive."

[0301] Accumulation in the database

[0302] The server stores the analysis results, including the original comment, summary, sentiment analysis results, and timestamps, in a database, allowing the analysis results to be stored centrally.

[0303] Data transition and trend analysis

[0304] The server compares the accumulated data with past data to analyze trends. At the same time, it analyzes the trends of specific respondents based on their respondent IDs to identify patterns and changes. For example, it uses the following SQL query:

[0305] sql

[0306] SELECT COUNT(), emotion

[0307] FROM feedback_analysis

[0308] WHERE timestamp >= '2023-01-01' AND timestamp <= '2023-12-31'

[0309] GROUP BY emotion;

[0310] Next action suggestion

[0311] Based on the analysis results, the server proposes the next action to take, such as "proposing holding a workshop to improve progress" or "recommending feedback follow-up for specific employees," and notifies the manager.

[0312] Terminal and user behavior

[0313] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. Proposed next actions are also notified in real time. This allows users to plan and implement specific follow-ups and measures. For example, a user can decide to hold a workshop and send a notification to employees.

[0314] This allows the present invention to rapidly collect and analyze qualitative data, contributing to the realization of efficient business operations and human resource management.

[0315] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0316] Step 1: Retrieving Comment Data

[0317] The server sends an HTTP request to the API endpoint of the survey system to retrieve the comment data. For example, the survey system might send the following request:

[0318] plaintext

[0319] GET / api / v1 / feedbacks

[0320] Host: survey.example.com

[0321] Authorization: Bearer<API_TOKEN>

[0322] The input is the API endpoint and authentication information. The output is the comment data retrieved in JSON format. The server extracts the comment field from the retrieved JSON data and converts it into a format that can be processed within the system.

[0323] Step 2: Preprocessing the comment data

[0324] The server pre-processes the retrieved comment data, mainly performing the following tasks:

[0325] Remove unnecessary spaces and special characters.

[0326] Remove emojis or convert them to text.

[0327] Use the same capitalization for text.

[0328] For example, if the input is comment data such as "The meeting was a good one," the output will be "The meeting was good."

[0329] Step 3: Summary generation

[0330] The server uses the generative AI model on the preprocessed comment data to summarize the comments. It gives the generative AI model the following prompts:

[0331] plaintext

[0332] Prompt: "Summarize the following comment: 'The meeting went very smoothly and the whole team was united. The results exceeded our expectations.'"

[0333] The input is the preprocessed comment data and a prompt for a summary, and the output is a summary obtained from the generative AI model, such as "The meeting went smoothly and produced good results."

[0334] Step 4: Sentiment analysis

[0335] The server performs sentiment analysis on the summary using the sentiment engine, and then provides the following prompts for the generated summary:

[0336] plaintext

[0337] Prompt: "Analyze the sentiment of the following sentence: 'The meeting went smoothly and produced good results.'"

[0338] The input is the generated summary and a prompt for sentiment analysis. The output is the sentiment analysis result of "positive" obtained from the sentiment engine.

[0339] Step 5: Database storage

[0340] The server stores the analysis results, including the original comment, summary, sentiment analysis, and timestamp, in a database. The inputs are the summary, sentiment analysis, original comment, and timestamp. The output is a database record containing these data. For example, use the following SQL query:

[0341] sql

[0342] INSERT INTO feedback_analysis (original_comment, summary, emotion, timestamp)

[0343] VALUES ('The meeting went very smoothly and the whole team was united. The results exceeded our expectations.', 'The meeting went smoothly and produced good results.', 'Positive', '2023-10-10 10:00:00');

[0344] Step 6: Data transition and trend analysis

[0345] The server compares the accumulated data with past data and analyzes trends. Next, it extracts data for specific respondents based on the respondent ID and analyzes their trends. The input is past data and question data for a specific period. The output is the results of trend analysis of emotions and opinions, as well as the results of trend analysis of specific respondents. For example, the following SQL query is used:

[0346] sql

[0347] SELECT COUNT(), emotion

[0348] FROM feedback_analysis

[0349] WHERE timestamp >= '2023-01-01' AND timestamp <= '2023-12-31'

[0350] GROUP BY emotion;

[0351] Step 7: Present next actions

[0352] The server suggests the next action based on the analysis results. Based on the analyzed data, it suggests necessary actions and follow-ups to staff. The input is the results of the transition and trend analysis. The output is a specific action suggestion. For example, it suggests actions such as "Suggest holding a workshop on improving progress" or "Recommend follow-up on feedback from specific employees." Managers are notified via email, etc.

[0353] email

[0354] To: manager@example.com

[0355] Subject: Next action suggestions

[0356] Body: I ​​suggest holding a workshop on improving progression.

[0357] Step 8: User confirmation and execution

[0358] The terminal provides an interface for users to check survey results and analysis data, and notifies them of proposed next actions in real time. The input is the analysis results and action suggestions sent from the server. The output is visual information and notifications displayed to the user. The user can plan and execute follow-ups and measures through the terminal. For example, a user decides to hold a workshop and sends a notification to employees:

[0359] email

[0360] To: all_employees@example.com

[0361] Subject: Workshop Announcement

[0362] Body: We're hosting a workshop on improving your progression. We'd love for you to join us.

[0363] (Application example 2)

[0364] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0365] Modern brick-and-mortar stores are required to quickly and accurately collect and analyze customer feedback and quickly take appropriate action based on it. However, conventional feedback collection systems have issues with time loss due to manual input and data inaccuracy. Furthermore, sentiment analysis and summarization of feedback are often done manually, making it difficult to respond quickly. In addition, the lack of real-time action suggestions hinders quick action to improve customer satisfaction.

[0366] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for automatically acquiring comment data from a questionnaire system, means for preprocessing the comment data, means for summarizing the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing trends of individual respondents, means for collecting customer feedback using smart glasses, and means for analyzing the collected customer feedback in real time and proposing the next action. This enables rapid collection and analysis of customer feedback and makes it possible to propose appropriate actions in real time.

[0367] A "survey system" is a mechanism for collecting comments and opinions from users.

[0368] "Comment data" is text data containing user opinions and impressions obtained from a questionnaire system.

[0369] "Preprocessing" refers to the process of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format.

[0370] "Summarization" refers to shortening long comment data and expressing the main content concisely.

[0371] "Sentiment analysis" refers to analyzing the sentiment of summarized comment data and classifying it into categories such as positive, negative, or neutral.

[0372] A "database" is a system for storing and managing information such as preprocessed comment data, summaries, and sentiment analysis results.

[0373] "Trend analysis" involves comparing accumulated feedback data with past data to analyze trends in emotions and opinions.

[0374] "Individual respondent trend analysis" refers to analyzing the data of specific respondents to understand their characteristics and patterns.

[0375] "Smart glasses" are glasses-type devices that have a display visible to the wearer and the ability to exchange audio, video, and data in real time.

[0376] "Real-time analysis" means analyzing collected data immediately and obtaining analysis results instantly.

[0377] "Next action proposals" are proposals for specific next steps or improvements based on the analysis results.

[0378] The present invention is a system for automatically collecting and analyzing customer feedback in a physical store and proposing next actions to staff in real time. An embodiment of this system will be described in detail below.

[0379] Program Generation and Processing Description

[0380] This system is constructed using hardware such as a server, smart glasses, and user terminals, as well as software such as generative AI models and emotion analysis tools.

[0381] Hardware and Software Configuration

[0382] Hardware:

[0383] Smart glasses (e.g., interactive eyeglasses)

[0384] Server (high-performance computer)

[0385] User device (e.g., tablet, smartphone)

[0386] software:

[0387] Generative AI models (e.g., models using natural language processing technology)

[0388] Sentiment analysis tools (e.g., software that analyzes the sentiment of text)

[0389] Database management systems (e.g., database software)

[0390] API management systems (e.g., software that manages API endpoints)

[0391] Server Procedure

[0392] 1. Feedback Collection:

[0393] Smart glasses are used to collect customer feedback in-store via voice input, where the voice input is based on requested prompts.

[0394] For example: "I like the coffee at this cafe, but they're slow to serve orders."

[0395] 2. Data Preprocessing:

[0396] The smart glasses convert the voice data into text data and send it to the server, where it removes unnecessary spaces, special characters, and emojis from the comment data and standardizes the text format.

[0397] 3. Summarization and Sentiment Analysis:

[0398] The preprocessed comment data is summarized using a generative AI model, and the summarized comment data is analyzed for sentiment using a sentiment analysis tool, categorizing it into categories such as positive, negative, and neutral.

[0399] Example: "The coffee is good, but the service is slow." (Emotion: Negative)

[0400] 4. Database accumulation and comparison with past data:

[0401] The analysis results are stored in a database that includes the original comments, summaries, sentiment analysis results, and timestamps. This data is used to compare with past data and analyze trends in sentiment and opinion.

[0402] 5. Individual respondent trend analysis and next action suggestions:

[0403] Based on the data of each individual respondent, trends are analyzed to identify specific characteristics and patterns, which then automatically suggest next steps and notify staff via the smart glasses and user devices.

[0404] For example: "Propose staff training to improve service speed."

[0405] Specific use cases

[0406] Below is an example of usage.

[0407] Example prompt sentence:

[0408] "I like the coffee at this cafe, but they're slow to serve orders."

[0409] Example of analysis results:

[0410] Summary: "Good coffee, but slow service."

[0411] Emotion: Negative

[0412] Suggestion: "Propose staff training to improve service speed."

[0413] In this way, we provide a system that automates the entire process from feedback collection to analysis and suggests next actions in real time, thereby improving customer satisfaction and streamlining store operations.

[0414] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0415] Step 1: Gather feedback

[0416] A user wears the smart glasses and provides voice feedback in the store. This feedback is captured by the smart glasses as voice input. The smart glasses use voice recognition technology to convert the voice data into text data and send it to the server.

[0417] Input: User's spoken feedback (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[0418] Output: Text data sent to the server

[0419] Step 2: Data Preprocessing

[0420] The server receives the text data sent from the smart glasses and performs preprocessing, which removes unnecessary spaces, special characters, and emojis from the comment data and standardizes the text format.

[0421] Input: Text data sent from smart glasses (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[0422] Output: Preprocessed text data (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[0423] Step 3: Summary

[0424] The server summarizes the preprocessed text data using a generative AI model, which analyzes the text data and generates a summary that succinctly expresses the main content.

[0425] Input: Preprocessed text data (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[0426] Output: Summarized text data (e.g., "The coffee is good, but the service is slow.")

[0427] Step 4: Sentiment analysis

[0428] The server analyzes the generated summary using a sentiment analysis tool, which classifies the sentiment of the summary into categories such as positive, negative, and neutral.

[0429] Input: Summarized text data (e.g., "The coffee is good, but the service is slow.")

[0430] Output: Sentiment analysis result (e.g. "negative")

[0431] Step 5: Database accumulation

[0432] The server stores the results of the sentiment analysis, the summary, and the original text data in a database, which also stores the analysis results and timestamps.

[0433] Input: Original text data, summary, sentiment analysis results, timestamp

[0434] Output: Data accumulation in database

[0435] Step 6: Analyze the progress

[0436] The server analyzes trends based on the feedback data stored in the database, comparing it with past data, thereby analyzing fluctuations in sentiment and opinion over a specific period of time.

[0437] Input: Feedback data stored in the database

[0438] Output: Trend analysis report

[0439] Step 7: Individual respondent trend analysis

[0440] The server extracts data from each respondent and analyzes trends and patterns, which allows for the identification of specific respondent characteristics and behavioral patterns.

[0441] Input: Individual respondent data

[0442] Output: Trend analysis report

[0443] Step 8: Proposed next actions

[0444] The server then proposes specific next steps based on the results of trend analysis, which are then sent to the smart glasses or the user's device, where staff can check them in real time.

[0445] Input: Trend analysis results, Trend analysis results

[0446] Output: Notification of next action suggestion (e.g. "Propose staff training to improve service speed.")

[0447] The above steps enable fast and accurate collection, analysis, and action proposal of customer feedback.

[0448] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0449] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0450] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0451] [Second embodiment]

[0452] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0453] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0454] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0455] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0456] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0457] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0458] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0459] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0460] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0461] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0462] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0463] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0464] Understood. Below is the "Mode for Carrying Out the Invention" of the specification.

[0465] This invention is a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program of this system operates based on the operations of the server, terminal, and user.

[0466] First, the server automatically retrieves comment data from the survey system. For example, to retrieve survey data for the department monthly meeting in August 2023, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data.

[0467] Next, the server preprocesses the acquired comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the accuracy of the analysis.

[0468] The server then uses a generative AI model to summarize the preprocessed comment data. For example, it converts a comment such as "The meeting was progressing slowly" into a summary statement such as "The delay in the progress was pointed out." The server then uses a sentiment analysis tool to classify the sentiment of the comment as "negative."

[0469] The server then stores the generated summary and the sentiment analysis results in a database. The stored data includes the original comment, summary, sentiment analysis results, and timestamp.

[0470] Based on the accumulated data, the server compares it with past data and analyzes trends. For example, it compares it with data from the past few months to see if there has been an increase in negative comments. This trend analysis allows you to understand trends and take necessary measures.

[0471] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, the server extracts this tendency from the database and identifies potential management positions or employees who need follow-up.

[0472] Finally, the server will suggest the next action based on the results of these analyses, such as "proposing holding a workshop to improve progress" or "recommending feedback follow-up for specific employees," and notify managers of specific actions.

[0473] The terminal provides an interface for users to view survey results and analytical data. Through the terminal, users can view summaries of quantitative and qualitative data, visualize the results in graphs and dashboards, and receive real-time notifications of suggested next steps.

[0474] Based on the information provided on the device, the user can plan and execute specific follow-up actions and measures. For example, the user can decide to hold a workshop and send a notification to employees.

[0475] As a result, the present invention enables rapid collection and analysis of qualitative data, leading to efficient business operations and effective human resource management.

[0476] The processing flow will be explained below.

[0477] Understood. Below are the specific steps of the process.

[0478] Step 1:

[0479] The server automatically retrieves comment data from the survey system by sending an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[0480] Step 2:

[0481] The server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[0482] Step 3:

[0483] The server then uses a generative AI model to summarize the preprocessed comment data. For example, it converts long comments into short summary sentences. This AI model uses natural language processing technology to concisely express the main content of the comments.

[0484] Step 4:

[0485] The server uses a sentiment analysis tool to analyze the sentiment of the comments. Specifically, it categorizes them into categories such as positive, negative, and neutral. For example, a comment such as "The meeting went slowly" would be classified as "negative."

[0486] Step 5:

[0487] The server stores the generated summaries and the results of sentiment analysis in a database. The stored data includes the original comments, summaries, sentiment analysis results, and timestamps. This allows the analysis results to be stored centrally in the database.

[0488] Step 6:

[0489] The server analyzes the trends in past data based on the accumulated data, comparing data from multiple months and analyzing trends in sentiment and opinions to determine whether negative comments are increasing at a particular time.

[0490] Step 7:

[0491] The server analyzes the trends of individual respondents. Based on the respondent ID, it extracts data for specific respondents and analyzes their trends. This process identifies the characteristics and patterns of respondents and identifies management candidates and employees who need follow-up.

[0492] Step 8:

[0493] The server then proposes the next action based on the analysis results, for example, generating specific action suggestions such as "proposing to hold a workshop on improving progress" or "recommending feedback follow-up for specific employees," and notifying the manager.

[0494] Step 9:

[0495] The terminal provides a user interface that displays survey results and analytical data to the user, allowing the user to view summaries of quantitative and qualitative data and visualize the results in the form of graphs and dashboards.

[0496] Step 10:

[0497] The user plans and executes specific follow-up actions and measures based on the information provided on the device. For example, the user decides to hold a workshop and sends a notification to employees.

[0498] Based on the above steps, the present invention enables rapid collection and analysis of qualitative data, thereby realizing efficient business operations and human resource management.

[0499] Example 1

[0500] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0501] Conventional survey analysis systems have struggled to efficiently process and analyze large amounts of comment data and propose effective actions based on that data. Furthermore, manually processing data consumes significant human resources and results in inconsistent analytical accuracy. There is a need for a system that can resolve these issues, quickly aggregate and analyze qualitative data, and achieve efficient business operations and effective human resource management.

[0502] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0503] In this invention, the server includes means for automatically acquiring comment data from a survey system, means for preprocessing the comment data, means for using a generative AI model to summarize the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing tendencies of individual respondents, and means for suggesting next actions based on the analysis results. This makes it possible to automatically process large amounts of comment data and propose specific actions based on highly accurate analysis results.

[0504] A "survey system" is an electronic survey device for collecting opinions and feedback from users.

[0505] "Comment data" is text data that includes user opinions and feedback collected through a questionnaire system.

[0506] "Preprocessing" is the procedure of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format.

[0507] A "generative AI model" is an artificial intelligence model that uses natural language processing to summarize, generate, or analyze text data.

[0508] "Sentiment analysis" is a method for automatically classifying or assessing the emotions contained in text data.

[0509] A "database" is an electronic system for efficiently storing, managing, and retrieving organized data.

[0510] "Transition analysis" is a method of analyzing temporal changes and trends based on accumulated data.

[0511] "Trend analysis" is a method of analyzing the behavioral patterns and opinion trends of specific users or groups.

[0512] "Action suggestion" is the process of suggesting the next course of action based on the results of data analysis.

[0513] This system automatically preprocesses comment data obtained from a survey system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. This system operates based on operations from the server, terminals, and users.

[0514] First, the server automatically retrieves comment data from the survey system. Specifically, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, to retrieve survey data for the department monthly meeting in August 2023, the server uses the API to collect the data. The retrieved data is saved in JSON format or similar.

[0515] The server then preprocesses the retrieved comment data. Specifically, it uses Python and natural language processing (NLP) tools (e.g., NLTK and SpaCy) to remove unnecessary spaces, special characters, and emojis from the comments and standardize the text format. This preprocessing improves the analysis accuracy of the generative AI model.

[0516] After preprocessing the comment data, the server uses a generative AI model (e.g., GPT-4) to summarize the comments. An example prompt is shown below.

[0517] Example prompt sentence:

[0518] Original comment:

[0519] "The meeting was slow"

[0520] Preprocessed text:

[0521] "Slow progress pointed out"

[0522] Generate a summary:

[0523] When this prompt is fed into a generative AI model, the model generates the summary statement, "Delays in progress have been noted."

[0524] The server also performs sentiment analysis on the generated summary using tools such as TextBlob and VADER. For example, a comment such as "Delays in progress have been noted" is classified as "negative."

[0525] The server then stores the generated summary and the results of sentiment analysis in a database. The data to be saved includes the original comment, summary, sentiment analysis results, and timestamp. The database can be MySQL or PostgreSQL.

[0526] The server analyzes the accumulated data by comparing it with past data. For example, it analyzes whether there has been an increase in negative comments compared to data from the past few months. This analysis is performed using statistical analysis libraries such as pandas and numpy in Python.

[0527] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, the server can extract this tendency from the database. This analysis can then be used to identify potential management candidates or employees who need follow-up.

[0528] Finally, the server will suggest the next action based on the results of these analyses, such as "Suggest holding a workshop to improve progress" or "Recommend follow-up on feedback from specific employees," and notify managers of specific actions. This notification can be sent via email or via a dashboard within the system.

[0529] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and view the results visualized in graphs and dashboards. For example, graphs can be generated using JavaScript D3.js and Chart.js, allowing users to interactively check the analysis results. It can also display suggested next actions with real-time notifications.

[0530] Users can plan and implement specific follow-up actions and measures based on the information provided on their devices. For example, they can schedule workshops to improve progress and send notifications to employees. In this way, efficient business operations and effective human resource management are realized.

[0531] As described above, the present invention is a system that enables rapid collection and analysis of qualitative data and aims to improve and streamline business operations by proposing specific actions.

[0532] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0533] Step 1:

[0534] Data Acquisition

[0535] The server sends an HTTP request to the API endpoint to automatically retrieve comment data from the survey system. For example, the server sends a GET request to "https: / / api.example.com / surveys / latest" and receives JSON-formatted comment data as a response. The input is the API endpoint and authentication information (e.g., OAuth token), and the output is the retrieved JSON-formatted comment data. Specifically, data is retrieved from the API using a Python library (such as requests).

[0536] Step 2:

[0537] Data Preprocessing

[0538] The server preprocesses the comment data it receives. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments, and converts all text to a unified format (for example, lowercase). This process uses Python's regular expression module (re) and natural language processing libraries (NLTK and SpaCy). The input is comment data in JSON format, and the output is preprocessed string data. The specific operation is performed using the following code:

[0539] python

[0540] import re

[0541] import spacy

[0542] def preprocess_comment(comment):

[0543] comment = re.sub(r'[^\w\s]', '', comment) Remove special characters

[0544] comment = comment.lower() Convert all to lowercase

[0545] return comment

[0546] Step 3:

[0547] Summary Generation

[0548] The server inputs the preprocessed comment data into a generative AI model (e.g., GPT-4) to generate a summary of the comments. The prompt used here is:

[0549] input:

[0550] Original comment:

[0551] "The meeting was slow"

[0552] Preprocessed text:

[0553] "Slow progress pointed out"

[0554] Generate a summary:

[0555] The output is a summary sentence returned from the generative AI model. Specifically, the prompt sentence is sent to the API of the generative AI model and the summary sentence is received.

[0556] Step 4:

[0557] sentiment analysis

[0558] The server performs sentiment analysis on the generated summary. For sentiment analysis, it uses natural language processing tools such as TextBlob and VADER. The input is the generated summary, and the output is the sentiment classification result of the summary (e.g., negative, positive, neutral). To see the specific operation, use the following code:

[0559] python

[0560] from textblob import TextBlob

[0561] def analyze_sentiment(text):

[0562] analysis = TextBlob(text)

[0563] if analysis.sentiment.polarity > 0:

[0564] return "affirmative"

[0565] elif analysis.sentiment.polarity == 0:

[0566] return "neutral"

[0567] else:

[0568] return "negative"

[0569] Step 5:

[0570] Saving to a database

[0571] The server stores the original comment, the generated summary, the sentiment analysis result, and the timestamp in a database. The input is the original comment, the summary, the sentiment result, and the timestamp, and the output is an entry stored in the database. Specifically, the data is stored in the database using an SQL insert query:

[0572] sql

[0573] INSERT INTO comments (original_comment, summary, sentiment, timestamp)

[0574] VALUES (%s, %s, %s, %s);

[0575] Step 6:

[0576] Trend analysis

[0577] The server analyzes the trends of past data based on the accumulated data. The input is past comment data retrieved from the database, and the output is the results of the trend analysis. Specifically, it performs statistical analysis using the Python pandas library:

[0578] python

[0579] import pandas as pd

[0580] def trend_analysis(comments):

[0581] df = pd.DataFrame(comments)

[0582] trend = df['sentiment'].value_counts()

[0583] return trend

[0584] Step 7:

[0585] Analysis of user trends

[0586] The server analyzes the trends of individual respondents. The input is comment data related to a specific respondent, and the output is the specific trend analysis results. Specifically, it extracts data for a specific respondent by filtering and performs analysis:

[0587] python

[0588] def user_trend_analysis(comments, user_id):

[0589] user_comments = [comment for comment in comments if comment['user_id'] == user_id]

[0590] trend = trend_analysis(user_comments)

[0591] return trend

[0592] Step 8:

[0593] Suggested next actions

[0594] The server proposes the next action based on the analysis results. The input is the transition analysis results and trend analysis results, and the output is a specific action proposal (e.g., a proposal to hold a workshop). Specifically, it performs conditional branching based on the analysis results to generate appropriate actions and notify management:

[0595] python

[0596] def suggest_next_action(trend):

[0597] if trend['negative'] > trend['positive']:

[0598] return "Proposal to hold a workshop on improving progress"

[0599] else:

[0600] return "No follow-up required"

[0601] The above is the processing steps of the system and the specific operations that accompany them.

[0602] (Application example 1)

[0603] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0604] In modern brick-and-mortar stores, efficient staff management and improved service quality are key challenges. Traditional management methods make it difficult to properly collect and analyze feedback from staff, and they are not expected to take appropriate action quickly. Furthermore, it is necessary to accurately summarize the content of feedback and perform sentiment analysis to grasp trends and provide accurate notifications to managers.

[0605] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0606] In this invention, the server includes means for automatically acquiring comment data from a survey system, means for preprocessing the comment data, means for using a generative AI model to summarize the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends from past data based on the stored data, means for analyzing trends of individual respondents, means for suggesting next actions based on the analysis results, and means for visualizing feedback trends using the analysis results and notifying management. This enables efficient collection and analysis of staff feedback and prompt suggestion of appropriate next actions.

[0607] A "survey system" is a system for collecting feedback and opinions from users.

[0608] "Comment Data" refers to opinions and feedback in text form obtained from users through the survey system.

[0609] "Preprocessing" refers to the process of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format to make data analysis easier.

[0610] A "generative AI model" is an artificial intelligence algorithm that learns from large amounts of data and generates new sentences.

[0611] "Sentiment analysis" is a technique that analyzes the content of text data and determines whether the sentiment is positive, negative, or neutral.

[0612] A "database" is a digital storage system for systematically organizing and storing information.

[0613] "Feedback trends" are patterns and tendencies that emerge from the analysis of accumulated feedback data.

[0614] A "management position" is a position within an organization that is responsible for managing staff and running business operations.

[0615] An "HTTP request" is a communication protocol for requesting data from a web server.

[0616] An "API endpoint" is an interface for connecting with external systems and is a connection point for exchanging data.

[0617] This invention is a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program of this system operates based on the operations of the server, terminal, and user.

[0618] First, the server automatically retrieves comment data from the survey system. For example, it sends an HTTP request to an API endpoint to retrieve the latest survey data. The retrieved comment data is preprocessed to remove spaces, special characters, and emojis, and standardize the text format.

[0619] Next, a generative AI model is used on the preprocessed comment data to summarize the comments. For example, a comment such as "The customer service today was terrible" is converted into a summary sentence such as "There was a problem with the service." A sentiment analysis tool is also used to classify the sentiment of this comment as "negative."

[0620] The generated summary and the results of the sentiment analysis are stored in a database. This data includes the original comment, summary, sentiment analysis results, and timestamp. Based on the stored data, the server compares it with past data to analyze trends. For example, it analyzes whether there has been an increase in negative comments compared to data from the past few months. This trend analysis allows trends to be identified and necessary measures to be taken.

[0621] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, it extracts that tendency from the database and identifies staff members who need follow-up. Finally, based on these analysis results, the server proposes the next action to take and notifies management. For example, it may notify specific actions such as "proposing holding a workshop to improve poor customer service" or "recommending follow-up with specific staff members."

[0622] The terminal provides an interface for users to check survey results and analytical data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. Suggested next actions are also notified in real time. For example, specific feedback such as "Today's shift was very busy, but enjoyable" or "The handling of customer complaints did not go well" is displayed in a list. An example of a prompt is "Generate a summary of the following feedback comments: 'Today's shift was very busy, but enjoyable.'"

[0623] As a result, the present invention enables rapid collection and analysis of qualitative data, leading to efficient business operations and effective human resource management.

[0624] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0625] Step 1:

[0626] The server automatically retrieves comment data from the survey system. Specifically, it sends an HTTP request to a pre-configured API endpoint to retrieve the survey data. The input is the API endpoint, and the output is the retrieved comment data.

[0627] Step 2:

[0628] The server preprocesses the acquired comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. The input is the acquired comment data, and the output is the preprocessed comment data.

[0629] Step 3:

[0630] The server summarizes the preprocessed comment data using a generative AI model. Specifically, it converts the comment text into a summary sentence. For example, a comment such as "The customer service today was terrible" is converted into the summary sentence "There was a problem with the service." The input is the preprocessed comment data, and the output is the summarized text data.

[0631] Step 4:

[0632] The server performs sentiment analysis on the summarized comment data. Specifically, it uses a sentiment analysis tool to classify the sentiment of the comment as positive, negative, or neutral. For example, a comment that says "there was a problem with the response" is classified as "negative." The input is the summarized comment data, and the output is the result of the sentiment analysis.

[0633] Step 5:

[0634] The server stores the generated summary and the results of sentiment analysis in a database. Specifically, it saves the original text of the comment, the summarized text, the results of sentiment analysis, and a timestamp. The input is the summary and the results of sentiment analysis, and the output is the information stored in the database.

[0635] Step 6:

[0636] The server analyzes the trends in past data based on the accumulated data. Specifically, it extracts data from the past few months and analyzes whether there has been an increase in negative comments. The input is the past data accumulated in the database, and the output is the results of the trend analysis.

[0637] Step 7:

[0638] The server analyzes the tendencies of individual respondents. Specifically, if a particular respondent repeatedly makes negative comments, it extracts that tendency and identifies staff members who need follow-up. The input is the comment data stored in the database and the analysis results, and the output is a list of staff members who need follow-up.

[0639] Step 8:

[0640] The server proposes the next action based on the analysis results and notifies the manager. Specifically, it notifies specific actions such as "proposing holding a workshop to improve the worsening response" or "recommending follow-up with specific staff members." The input is the transition analysis results and individual trend analysis results, and the output is the notification content to the manager.

[0641] Step 9:

[0642] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. The input is the survey results and analysis data stored in the database, and the output is the visualized data displayed on the interface.

[0643] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0644] Understood. Below is the "Mode for Carrying Out the Invention" of the specification based on the invention combining the emotion engine.

[0645] This invention combines an emotion engine with a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program for this system operates based on operations by the server, terminal, and user.

[0646] First, the server automatically retrieves comment data from the survey system. To do this, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[0647] Next, the server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[0648] The server then uses a generative AI model to summarize the preprocessed comment data, for example, converting long comments into short summary sentences. This AI model uses natural language processing techniques to concisely express the main content of the comment.

[0649] The server then analyzes the sentiment of the comments using a sentiment analysis tool and a sentiment engine. Specifically, the sentiment engine categorizes the sentiment into categories such as positive, negative, and neutral. For example, a comment such as "The meeting went slowly" would be classified as "negative."

[0650] The server then stores the generated summary and the results of the sentiment analysis in a database. The data stored includes the original comment, summary, sentiment analysis results, and timestamp. This allows the analysis results to be stored centrally in the database.

[0651] Based on the accumulated data, the server compares it with past data and analyzes trends. By comparing data from multiple months and analyzing trends in sentiment and opinions, it can determine whether there has been an increase in negative comments at a particular time.

[0652] The server then analyzes the trends of individual respondents. Based on the respondent ID, data on specific respondents is extracted and its trends are analyzed. This process identifies the characteristics and patterns of respondents and identifies management candidates and employees who need follow-up.

[0653] Finally, the server generates specific action suggestions based on the results of these analyses, such as "suggest holding a workshop to improve progress" or "recommend following up on feedback from specific employees," and notifies managers.

[0654] The terminal provides an interface for users to view survey results and analytical data. Through the terminal, users can view summaries of quantitative and qualitative data, visualize the results in graphs and dashboards, and receive real-time notifications of suggested next steps.

[0655] The user can plan and implement specific follow-up actions and measures based on the information provided on the device. For example, the user can decide to hold a workshop and send a notification to employees.

[0656] This allows for the rapid collection and analysis of qualitative data, resulting in efficient business operations and human resource management. Furthermore, by combining it with an emotion engine, it is possible to recognize users' emotions and provide specific feedback and follow-up actions based on those emotions.

[0657] The processing flow will be explained below.

[0658] Understood. Below are the specific operations for each processing step.

[0659] Step 1:

[0660] The server automatically retrieves comment data from the survey system by sending an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[0661] Step 2:

[0662] The server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[0663] Step 3:

[0664] The server uses a generative AI model to summarize the preprocessed comment data. For example, a comment such as "The meeting was progressing slowly. There were probably too many agenda items" is converted into a summary such as "The slow progress and the large number of agenda items were pointed out." This AI model uses natural language processing technology to concisely express the main content of the comment.

[0665] Step 4:

[0666] The server analyzes the sentiment of the comments using a sentiment analysis tool and a sentiment engine. For example, a comment such as "The meeting went slowly" may be classified as "negative." The sentiment engine categorizes the sentiment into categories such as positive, negative, and neutral.

[0667] Step 5:

[0668] The server stores the generated summary and the results of sentiment analysis in a database. The stored data includes the original comment, summary, sentiment analysis results, and timestamp. This allows the analysis results to be stored centrally in the database.

[0669] Step 6:

[0670] The server analyzes the accumulated data and compares it with past data, analyzing trends in sentiment and opinions by comparing data from multiple months, for example, to determine whether negative comments are increasing at certain times of the day or during certain events.

[0671] Step 7:

[0672] The server analyzes the trends of individual respondents. Based on the respondent ID, it extracts the data of a specific respondent and analyzes their trends. For example, if a specific employee frequently makes negative comments, it analyzes the employee's trends and generates a detailed report.

[0673] Step 8:

[0674] The server then proposes the next action based on the analysis results. For example, it generates specific action suggestions such as "Suggest holding a workshop to improve progress" or "Recommend follow-up feedback for specific employees." These suggestions are notified to managers in real time via a notification system.

[0675] Step 9:

[0676] The device provides a user interface that displays survey results and analysis data. Through the device, users can view summaries of quantitative and qualitative data and visualize the results in the form of graphs and dashboards. For example, sentiment analysis results can be displayed using color codes for positive, negative, and neutral.

[0677] Step 10:

[0678] The user plans and implements specific follow-up actions and measures based on the information provided on the device. For example, the user decides to hold a proposed workshop and sends an email notification to employees. This allows the user to take the necessary action quickly.

[0679] Based on the above steps, this invention enables the rapid collection and analysis of qualitative data, realizing efficient business operations and human resource management. Furthermore, by combining it with an emotion engine, it is possible to recognize users' emotions and provide specific feedback and follow-up actions based on those emotions.

[0680] Example 2

[0681] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0682] In conventional survey systems, comment data is often collected and analyzed manually, which is not only inefficient but also makes it difficult to ensure data consistency and quality. Furthermore, because sentiment analysis and summarization are not automated, the vast amount of comment data must be reviewed one by one, placing a heavy burden on the person in charge. Furthermore, the process for linking analysis results to subsequent actions is unclear, making it difficult to respond quickly and appropriately. This creates challenges that make it difficult to optimize organizational operations and human resource management.

[0683] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for automatically acquiring comment data from a questionnaire system, means for preprocessing the comment data, means for summarizing the preprocessed comment data using a generative AI model, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing tendencies of individual respondents, means for suggesting next actions based on the analysis results, and means for displaying the analysis results and next actions through a user interface. This consistently automates the process from comment data collection to summarization, sentiment analysis, data storage, trend analysis, trend analysis, and suggestion of next actions, enabling efficient and accurate data analysis and rapid response.

[0684] A "survey system" is a system for collecting comments and opinions from survey subjects, and has the function of acquiring data in digital form.

[0685] "Comment data" refers to data that indicates opinions and impressions in text format collected through a survey system.

[0686] "Preprocessing" refers to the processing step of removing unnecessary spaces, special symbols, and emojis from the acquired comment data and standardizing the text format.

[0687] A "generative AI model" is a model that uses artificial intelligence techniques to analyze a given text and perform a specific purpose, such as summarization or sentiment analysis.

[0688] "Summarization" refers to the process of summarizing the original comment data in a short and concise form and extracting only the main content.

[0689] "Sentiment analysis" is the process of identifying and classifying sentiments, such as positive, negative, or neutral, from text in comment data.

[0690] A "database" is a collection of information that systematically stores collected and processed data so that it can be easily searched, referenced, and analyzed later.

[0691] "Trend analysis" refers to a statistical process that compares past data with current data to reveal fluctuations and trends.

[0692] "Trend analysis" is an analytical technique used to identify recurring patterns or characteristics associated with a particular respondent or data set.

[0693] "Next action suggestion" refers to the process of proposing specific actions or measures to be taken based on the analysis results.

[0694] "User interface" refers to the visual or operational means or screen configuration that allows a user to operate a system and check the results.

[0695] This system automatically preprocesses comment data obtained from a survey system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, and compares and analyzes them with past data. This system operates based on the operations of the server, terminals, and users.

[0696] Hardware and software used

[0697] The main hardware and software used in the present invention are as follows:

[0698] Server: Performs data acquisition, preprocessing, summary generation, sentiment analysis, data accumulation, analysis, and action suggestion.

[0699] Terminal: Provides an interface for users to view results and take next actions.

[0700] Generative AI models: Use natural language processing techniques to perform text summarization and sentiment analysis.

[0701] System operation explanation

[0702] Retrieving comment data

[0703] The server sends an HTTP request to the API endpoint of the survey system to retrieve the comment data. For example, to retrieve feedback on the monthly department meeting, the server sends the following request:

[0704] plaintext

[0705] GET / api / v1 / feedbacks

[0706] Host: survey.example.com

[0707] Authorization: Bearer<API_TOKEN>

[0708] The server extracts the comment field from the response data and converts it into a format that can be processed within the system.

[0709] Preprocessing comment data

[0710] The server pre-processes the retrieved comment data, mainly performing the following tasks:

[0711] Remove unnecessary spaces and special characters.

[0712] Remove emojis or convert them to text.

[0713] Use the same capitalization for text.

[0714] For example, if the comment is "The meeting was 👍," it will be converted to "The meeting was good." This will improve the data quality of the comment text.

[0715] Summary Generation

[0716] The server uses the generative AI model on the preprocessed comment data to summarize the comments. It gives the generative AI model the following prompts:

[0717] plaintext

[0718] Prompt: "Summarize the following comment: 'The meeting went very smoothly and the whole team was united. The results exceeded our expectations.'"

[0719] The model generates a summary in response to this prompt, returning the summary "The meeting went smoothly and produced good results."

[0720] sentiment analysis

[0721] The server performs sentiment analysis on the summary using the sentiment engine, and then provides the following prompts for the generated summary:

[0722] plaintext

[0723] Prompt: "Analyze the sentiment of the following sentence: 'The meeting went smoothly and produced good results.'"

[0724] The emotion engine classifies emotions into categories such as positive, negative, and neutral, and returns the result as "positive."

[0725] Accumulation in the database

[0726] The server stores the analysis results, including the original comment, summary, sentiment analysis results, and timestamps, in a database, allowing the analysis results to be stored centrally.

[0727] Data transition and trend analysis

[0728] The server compares the accumulated data with past data to analyze trends. At the same time, it analyzes the trends of specific respondents based on their respondent IDs to identify patterns and changes. For example, it uses the following SQL query:

[0729] sql

[0730] SELECT COUNT(), emotion

[0731] FROM feedback_analysis

[0732] WHERE timestamp >= '2023-01-01' AND timestamp <= '2023-12-31'

[0733] GROUP BY emotion;

[0734] Next action suggestion

[0735] Based on the analysis results, the server proposes the next action to take, such as "proposing holding a workshop to improve progress" or "recommending feedback follow-up for specific employees," and notifies the manager.

[0736] Terminal and user behavior

[0737] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. Proposed next actions are also notified in real time. This allows users to plan and implement specific follow-ups and measures. For example, a user can decide to hold a workshop and send a notification to employees.

[0738] This allows the present invention to rapidly collect and analyze qualitative data, contributing to the realization of efficient business operations and human resource management.

[0739] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0740] Step 1: Retrieving Comment Data

[0741] The server sends an HTTP request to the API endpoint of the survey system to retrieve the comment data. For example, the survey system might send the following request:

[0742] plaintext

[0743] GET / api / v1 / feedbacks

[0744] Host: survey.example.com

[0745] Authorization: Bearer<API_TOKEN>

[0746] The input is the API endpoint and authentication information. The output is the comment data retrieved in JSON format. The server extracts the comment field from the retrieved JSON data and converts it into a format that can be processed within the system.

[0747] Step 2: Preprocessing the comment data

[0748] The server pre-processes the retrieved comment data, mainly performing the following tasks:

[0749] Remove unnecessary spaces and special characters.

[0750] Remove emojis or convert them to text.

[0751] Use the same capitalization for text.

[0752] For example, if the input is comment data such as "The meeting was a good one," the output will be "The meeting was good."

[0753] Step 3: Summary generation

[0754] The server uses the generative AI model on the preprocessed comment data to summarize the comments. It gives the generative AI model the following prompts:

[0755] plaintext

[0756] Prompt: "Summarize the following comment: 'The meeting went very smoothly and the whole team was united. The results exceeded our expectations.'"

[0757] The input is the preprocessed comment data and a prompt for a summary, and the output is a summary obtained from the generative AI model, such as "The meeting went smoothly and produced good results."

[0758] Step 4: Sentiment analysis

[0759] The server performs sentiment analysis on the summary using the sentiment engine, and then provides the following prompts for the generated summary:

[0760] plaintext

[0761] Prompt: "Analyze the sentiment of the following sentence: 'The meeting went smoothly and produced good results.'"

[0762] The input is the generated summary and a prompt for sentiment analysis. The output is the sentiment analysis result of "positive" obtained from the sentiment engine.

[0763] Step 5: Database storage

[0764] The server stores the analysis results, including the original comment, summary, sentiment analysis, and timestamp, in a database. The inputs are the summary, sentiment analysis, original comment, and timestamp. The output is a database record containing these data. For example, use the following SQL query:

[0765] sql

[0766] INSERT INTO feedback_analysis (original_comment, summary, emotion, timestamp)

[0767] VALUES ('The meeting went very smoothly and the whole team was united. The results exceeded our expectations.', 'The meeting went smoothly and produced good results.', 'Positive', '2023-10-10 10:00:00');

[0768] Step 6: Data transition and trend analysis

[0769] The server compares the accumulated data with past data and analyzes trends. Next, it extracts data for specific respondents based on the respondent ID and analyzes their trends. The input is past data and question data for a specific period. The output is the results of trend analysis of emotions and opinions, as well as the results of trend analysis of specific respondents. For example, the following SQL query is used:

[0770] sql

[0771] SELECT COUNT(), emotion

[0772] FROM feedback_analysis

[0773] WHERE timestamp >= '2023-01-01' AND timestamp <= '2023-12-31'

[0774] GROUP BY emotion;

[0775] Step 7: Present next actions

[0776] The server suggests the next action based on the analysis results. Based on the analyzed data, it suggests necessary actions and follow-ups to staff. The input is the results of the transition and trend analysis. The output is a specific action suggestion. For example, it suggests actions such as "Suggest holding a workshop on improving progress" or "Recommend follow-up on feedback from specific employees." Managers are notified via email, etc.

[0777] email

[0778] To: manager@example.com

[0779] Subject: Next action suggestions

[0780] Body: I ​​suggest holding a workshop on improving progression.

[0781] Step 8: User confirmation and execution

[0782] The terminal provides an interface for users to check survey results and analysis data, and notifies them of proposed next actions in real time. The input is the analysis results and action suggestions sent from the server. The output is visual information and notifications displayed to the user. The user can plan and execute follow-ups and measures through the terminal. For example, a user decides to hold a workshop and sends a notification to employees:

[0783] email

[0784] To: all_employees@example.com

[0785] Subject: Workshop Announcement

[0786] Body: We're hosting a workshop on improving your progression. We'd love for you to join us.

[0787] (Application example 2)

[0788] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0789] Modern brick-and-mortar stores are required to quickly and accurately collect and analyze customer feedback and quickly take appropriate action based on it. However, conventional feedback collection systems have issues with time loss due to manual input and data inaccuracy. Furthermore, sentiment analysis and summarization of feedback are often done manually, making it difficult to respond quickly. In addition, the lack of real-time action suggestions hinders quick action to improve customer satisfaction.

[0790] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for automatically acquiring comment data from a questionnaire system, means for preprocessing the comment data, means for summarizing the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing trends of individual respondents, means for collecting customer feedback using smart glasses, and means for analyzing the collected customer feedback in real time and proposing the next action. This enables rapid collection and analysis of customer feedback and makes it possible to propose appropriate actions in real time.

[0791] A "survey system" is a mechanism for collecting comments and opinions from users.

[0792] "Comment data" is text data containing user opinions and impressions obtained from a questionnaire system.

[0793] "Preprocessing" refers to the process of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format.

[0794] "Summarization" refers to shortening long comment data and expressing the main content concisely.

[0795] "Sentiment analysis" refers to analyzing the sentiment of summarized comment data and classifying it into categories such as positive, negative, or neutral.

[0796] A "database" is a system for storing and managing information such as preprocessed comment data, summaries, and sentiment analysis results.

[0797] "Trend analysis" involves comparing accumulated feedback data with past data to analyze trends in emotions and opinions.

[0798] "Individual respondent trend analysis" refers to analyzing the data of specific respondents to understand their characteristics and patterns.

[0799] "Smart glasses" are glasses-type devices that have a display visible to the wearer and the ability to exchange audio, video, and data in real time.

[0800] "Real-time analysis" means analyzing collected data immediately and obtaining analysis results instantly.

[0801] "Next action proposals" are proposals for specific next steps or improvements based on the analysis results.

[0802] The present invention is a system for automatically collecting and analyzing customer feedback in a physical store and proposing next actions to staff in real time. An embodiment of this system will be described in detail below.

[0803] Program Generation and Processing Description

[0804] This system is constructed using hardware such as a server, smart glasses, and user terminals, as well as software such as generative AI models and emotion analysis tools.

[0805] Hardware and Software Configuration

[0806] Hardware:

[0807] Smart glasses (e.g., interactive eyeglasses)

[0808] Server (high-performance computer)

[0809] User device (e.g., tablet, smartphone)

[0810] software:

[0811] Generative AI models (e.g., models using natural language processing technology)

[0812] Sentiment analysis tools (e.g., software that analyzes the sentiment of text)

[0813] Database management systems (e.g., database software)

[0814] API management systems (e.g., software that manages API endpoints)

[0815] Server Procedure

[0816] 1. Feedback Collection:

[0817] Smart glasses are used to collect customer feedback in-store via voice input, where the voice input is based on requested prompts.

[0818] For example: "I like the coffee at this cafe, but they're slow to serve orders."

[0819] 2. Data Preprocessing:

[0820] The smart glasses convert the voice data into text data and send it to the server, where it removes unnecessary spaces, special characters, and emojis from the comment data and standardizes the text format.

[0821] 3. Summarization and Sentiment Analysis:

[0822] The preprocessed comment data is summarized using a generative AI model, and the summarized comment data is analyzed for sentiment using a sentiment analysis tool, categorizing it into categories such as positive, negative, and neutral.

[0823] Example: "The coffee is good, but the service is slow." (Emotion: Negative)

[0824] 4. Database accumulation and comparison with past data:

[0825] The analysis results are stored in a database that includes the original comments, summaries, sentiment analysis results, and timestamps. This data is used to compare with past data and analyze trends in sentiment and opinion.

[0826] 5. Individual respondent trend analysis and next action suggestions:

[0827] Based on the data of each individual respondent, trends are analyzed to identify specific characteristics and patterns, which then automatically suggest next steps and notify staff via the smart glasses and user devices.

[0828] For example: "Propose staff training to improve service speed."

[0829] Specific use cases

[0830] Below is an example of usage.

[0831] Example prompt sentence:

[0832] "I like the coffee at this cafe, but they're slow to serve orders."

[0833] Example of analysis results:

[0834] Summary: "Good coffee, but slow service."

[0835] Emotion: Negative

[0836] Suggestion: "Propose staff training to improve service speed."

[0837] In this way, we provide a system that automates the entire process from feedback collection to analysis and suggests next actions in real time, thereby improving customer satisfaction and streamlining store operations.

[0838] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0839] Step 1: Gather feedback

[0840] A user wears the smart glasses and provides voice feedback in the store. This feedback is captured by the smart glasses as voice input. The smart glasses use voice recognition technology to convert the voice data into text data and send it to the server.

[0841] Input: User's spoken feedback (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[0842] Output: Text data sent to the server

[0843] Step 2: Data Preprocessing

[0844] The server receives the text data sent from the smart glasses and performs preprocessing, which removes unnecessary spaces, special characters, and emojis from the comment data and standardizes the text format.

[0845] Input: Text data sent from smart glasses (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[0846] Output: Preprocessed text data (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[0847] Step 3: Summary

[0848] The server summarizes the preprocessed text data using a generative AI model, which analyzes the text data and generates a summary that succinctly expresses the main content.

[0849] Input: Preprocessed text data (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[0850] Output: Summarized text data (e.g., "The coffee is good, but the service is slow.")

[0851] Step 4: Sentiment analysis

[0852] The server analyzes the generated summary using a sentiment analysis tool, which classifies the sentiment of the summary into categories such as positive, negative, and neutral.

[0853] Input: Summarized text data (e.g., "The coffee is good, but the service is slow.")

[0854] Output: Sentiment analysis result (e.g. "negative")

[0855] Step 5: Database accumulation

[0856] The server stores the results of the sentiment analysis, the summary, and the original text data in a database, which also stores the analysis results and timestamps.

[0857] Input: Original text data, summary, sentiment analysis results, timestamp

[0858] Output: Data accumulation in database

[0859] Step 6: Analyze the progress

[0860] The server analyzes trends based on the feedback data stored in the database, comparing it with past data, thereby analyzing fluctuations in sentiment and opinion over a specific period of time.

[0861] Input: Feedback data stored in the database

[0862] Output: Trend analysis report

[0863] Step 7: Individual respondent trend analysis

[0864] The server extracts data from each respondent and analyzes trends and patterns, which allows for the identification of specific respondent characteristics and behavioral patterns.

[0865] Input: Individual respondent data

[0866] Output: Trend analysis report

[0867] Step 8: Proposed next actions

[0868] The server then proposes specific next steps based on the results of trend analysis, which are then sent to the smart glasses or the user's device, where staff can check them in real time.

[0869] Input: Trend analysis results, Trend analysis results

[0870] Output: Notification of next action suggestion (e.g. "Propose staff training to improve service speed.")

[0871] The above steps enable fast and accurate collection, analysis, and action proposal of customer feedback.

[0872] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0873] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0874] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0875] [Third embodiment]

[0876] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0877] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0878] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0879] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0880] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0881] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0882] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0883] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0884] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0885] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0886] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0887] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0888] Understood. Below is the "Mode for Carrying Out the Invention" of the specification.

[0889] This invention is a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program of this system operates based on the operations of the server, terminal, and user.

[0890] First, the server automatically retrieves comment data from the survey system. For example, to retrieve survey data for the department monthly meeting in August 2023, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data.

[0891] Next, the server preprocesses the acquired comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the accuracy of the analysis.

[0892] The server then uses a generative AI model to summarize the preprocessed comment data. For example, it converts a comment such as "The meeting was progressing slowly" into a summary statement such as "The delay in the progress was pointed out." The server then uses a sentiment analysis tool to classify the sentiment of the comment as "negative."

[0893] The server then stores the generated summary and the sentiment analysis results in a database. The stored data includes the original comment, summary, sentiment analysis results, and timestamp.

[0894] Based on the accumulated data, the server compares it with past data and analyzes trends. For example, it compares it with data from the past few months to see if there has been an increase in negative comments. This trend analysis allows you to understand trends and take necessary measures.

[0895] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, the server extracts this tendency from the database and identifies potential management positions or employees who need follow-up.

[0896] Finally, the server will suggest the next action based on the results of these analyses, such as "proposing holding a workshop to improve progress" or "recommending feedback follow-up for specific employees," and notify managers of specific actions.

[0897] The terminal provides an interface for users to view survey results and analytical data. Through the terminal, users can view summaries of quantitative and qualitative data, visualize the results in graphs and dashboards, and receive real-time notifications of suggested next steps.

[0898] Based on the information provided on the device, the user can plan and execute specific follow-up actions and measures. For example, the user can decide to hold a workshop and send a notification to employees.

[0899] As a result, the present invention enables rapid collection and analysis of qualitative data, leading to efficient business operations and effective human resource management.

[0900] The processing flow will be explained below.

[0901] Understood. Below are the specific steps of the process.

[0902] Step 1:

[0903] The server automatically retrieves comment data from the survey system by sending an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[0904] Step 2:

[0905] The server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[0906] Step 3:

[0907] The server then uses a generative AI model to summarize the preprocessed comment data. For example, it converts long comments into short summary sentences. This AI model uses natural language processing technology to concisely express the main content of the comments.

[0908] Step 4:

[0909] The server uses a sentiment analysis tool to analyze the sentiment of the comments. Specifically, it categorizes them into categories such as positive, negative, and neutral. For example, a comment such as "The meeting went slowly" would be classified as "negative."

[0910] Step 5:

[0911] The server stores the generated summaries and the results of sentiment analysis in a database. The stored data includes the original comments, summaries, sentiment analysis results, and timestamps. This allows the analysis results to be stored centrally in the database.

[0912] Step 6:

[0913] The server analyzes the trends in past data based on the accumulated data, comparing data from multiple months and analyzing trends in sentiment and opinions to determine whether negative comments are increasing at a particular time.

[0914] Step 7:

[0915] The server analyzes the trends of individual respondents. Based on the respondent ID, it extracts data for specific respondents and analyzes their trends. This process identifies the characteristics and patterns of respondents and identifies management candidates and employees who need follow-up.

[0916] Step 8:

[0917] The server then proposes the next action based on the analysis results, for example, generating specific action suggestions such as "proposing to hold a workshop on improving progress" or "recommending feedback follow-up for specific employees," and notifying the manager.

[0918] Step 9:

[0919] The terminal provides a user interface that displays survey results and analytical data to the user, allowing the user to view summaries of quantitative and qualitative data and visualize the results in the form of graphs and dashboards.

[0920] Step 10:

[0921] The user plans and executes specific follow-up actions and measures based on the information provided on the device. For example, the user decides to hold a workshop and sends a notification to employees.

[0922] Based on the above steps, the present invention enables rapid collection and analysis of qualitative data, thereby realizing efficient business operations and human resource management.

[0923] Example 1

[0924] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0925] Conventional survey analysis systems have struggled to efficiently process and analyze large amounts of comment data and propose effective actions based on that data. Furthermore, manually processing data consumes significant human resources and results in inconsistent analytical accuracy. There is a need for a system that can resolve these issues, quickly aggregate and analyze qualitative data, and achieve efficient business operations and effective human resource management.

[0926] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0927] In this invention, the server includes means for automatically acquiring comment data from a survey system, means for preprocessing the comment data, means for using a generative AI model to summarize the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing tendencies of individual respondents, and means for suggesting next actions based on the analysis results. This makes it possible to automatically process large amounts of comment data and propose specific actions based on highly accurate analysis results.

[0928] A "survey system" is an electronic survey device for collecting opinions and feedback from users.

[0929] "Comment data" is text data that includes user opinions and feedback collected through a questionnaire system.

[0930] "Preprocessing" is the procedure of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format.

[0931] A "generative AI model" is an artificial intelligence model that uses natural language processing to summarize, generate, or analyze text data.

[0932] "Sentiment analysis" is a method for automatically classifying or assessing the emotions contained in text data.

[0933] A "database" is an electronic system for efficiently storing, managing, and retrieving organized data.

[0934] "Transition analysis" is a method of analyzing temporal changes and trends based on accumulated data.

[0935] "Trend analysis" is a method of analyzing the behavioral patterns and opinion trends of specific users or groups.

[0936] "Action suggestion" is the process of suggesting the next course of action based on the results of data analysis.

[0937] This system automatically preprocesses comment data obtained from a survey system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. This system operates based on operations from the server, terminals, and users.

[0938] First, the server automatically retrieves comment data from the survey system. Specifically, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, to retrieve survey data for the department monthly meeting in August 2023, the server uses the API to collect the data. The retrieved data is saved in JSON format or similar.

[0939] The server then preprocesses the retrieved comment data. Specifically, it uses Python and natural language processing (NLP) tools (e.g., NLTK and SpaCy) to remove unnecessary spaces, special characters, and emojis from the comments and standardize the text format. This preprocessing improves the analysis accuracy of the generative AI model.

[0940] After preprocessing the comment data, the server uses a generative AI model (e.g., GPT-4) to summarize the comments. An example prompt is shown below.

[0941] Example prompt sentence:

[0942] Original comment:

[0943] "The meeting was slow"

[0944] Preprocessed text:

[0945] "Slow progress pointed out"

[0946] Generate a summary:

[0947] When this prompt is fed into a generative AI model, the model generates the summary statement, "Delays in progress have been noted."

[0948] The server also performs sentiment analysis on the generated summary using tools such as TextBlob and VADER. For example, a comment such as "Delays in progress have been noted" is classified as "negative."

[0949] The server then stores the generated summary and the results of sentiment analysis in a database. The data to be saved includes the original comment, summary, sentiment analysis results, and timestamp. The database can be MySQL or PostgreSQL.

[0950] The server analyzes the accumulated data by comparing it with past data. For example, it analyzes whether there has been an increase in negative comments compared to data from the past few months. This analysis is performed using statistical analysis libraries such as pandas and numpy in Python.

[0951] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, the server can extract this tendency from the database. This analysis can then be used to identify potential management candidates or employees who need follow-up.

[0952] Finally, the server will suggest the next action based on the results of these analyses, such as "Suggest holding a workshop to improve progress" or "Recommend follow-up on feedback from specific employees," and notify managers of specific actions. This notification can be sent via email or via a dashboard within the system.

[0953] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and view the results visualized in graphs and dashboards. For example, graphs can be generated using JavaScript D3.js and Chart.js, allowing users to interactively check the analysis results. It can also display suggested next actions with real-time notifications.

[0954] Users can plan and implement specific follow-up actions and measures based on the information provided on their devices. For example, they can schedule workshops to improve progress and send notifications to employees. In this way, efficient business operations and effective human resource management are realized.

[0955] As described above, the present invention is a system that enables rapid collection and analysis of qualitative data and aims to improve and streamline business operations by proposing specific actions.

[0956] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0957] Step 1:

[0958] Data Acquisition

[0959] The server sends an HTTP request to the API endpoint to automatically retrieve comment data from the survey system. For example, the server sends a GET request to "https: / / api.example.com / surveys / latest" and receives JSON-formatted comment data as a response. The input is the API endpoint and authentication information (e.g., OAuth token), and the output is the retrieved JSON-formatted comment data. Specifically, data is retrieved from the API using a Python library (such as requests).

[0960] Step 2:

[0961] Data Preprocessing

[0962] The server preprocesses the comment data it receives. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments, and converts all text to a unified format (for example, lowercase). This process uses Python's regular expression module (re) and natural language processing libraries (NLTK and SpaCy). The input is comment data in JSON format, and the output is preprocessed string data. The specific operation is performed using the following code:

[0963] python

[0964] import re

[0965] import spacy

[0966] def preprocess_comment(comment):

[0967] comment = re.sub(r'[^\w\s]', '', comment) Remove special characters

[0968] comment = comment.lower() Convert all to lowercase

[0969] return comment

[0970] Step 3:

[0971] Summary Generation

[0972] The server inputs the preprocessed comment data into a generative AI model (e.g., GPT-4) to generate a summary of the comments. The prompt used here is:

[0973] input:

[0974] Original comment:

[0975] "The meeting was slow"

[0976] Preprocessed text:

[0977] "Slow progress pointed out"

[0978] Generate a summary:

[0979] The output is a summary sentence returned from the generative AI model. Specifically, the prompt sentence is sent to the API of the generative AI model and the summary sentence is received.

[0980] Step 4:

[0981] sentiment analysis

[0982] The server performs sentiment analysis on the generated summary. For sentiment analysis, it uses natural language processing tools such as TextBlob and VADER. The input is the generated summary, and the output is the sentiment classification result of the summary (e.g., negative, positive, neutral). To see the specific operation, use the following code:

[0983] python

[0984] from textblob import TextBlob

[0985] def analyze_sentiment(text):

[0986] analysis = TextBlob(text)

[0987] if analysis.sentiment.polarity > 0:

[0988] return "affirmative"

[0989] elif analysis.sentiment.polarity == 0:

[0990] return "neutral"

[0991] else:

[0992] return "negative"

[0993] Step 5:

[0994] Saving to a database

[0995] The server stores the original comment, the generated summary, the sentiment analysis result, and the timestamp in a database. The input is the original comment, the summary, the sentiment result, and the timestamp, and the output is an entry stored in the database. Specifically, the data is stored in the database using an SQL insert query:

[0996] sql

[0997] INSERT INTO comments (original_comment, summary, sentiment, timestamp)

[0998] VALUES (%s, %s, %s, %s);

[0999] Step 6:

[1000] Trend analysis

[1001] The server analyzes the trends of past data based on the accumulated data. The input is past comment data retrieved from the database, and the output is the results of the trend analysis. Specifically, it performs statistical analysis using the Python pandas library:

[1002] python

[1003] import pandas as pd

[1004] def trend_analysis(comments):

[1005] df = pd.DataFrame(comments)

[1006] trend = df['sentiment'].value_counts()

[1007] return trend

[1008] Step 7:

[1009] Analysis of user trends

[1010] The server analyzes the trends of individual respondents. The input is comment data related to a specific respondent, and the output is the specific trend analysis results. Specifically, it extracts data for a specific respondent by filtering and performs analysis:

[1011] python

[1012] def user_trend_analysis(comments, user_id):

[1013] user_comments = [comment for comment in comments if comment['user_id'] == user_id]

[1014] trend = trend_analysis(user_comments)

[1015] return trend

[1016] Step 8:

[1017] Suggested next actions

[1018] The server proposes the next action based on the analysis results. The input is the transition analysis results and trend analysis results, and the output is a specific action proposal (e.g., a proposal to hold a workshop). Specifically, it performs conditional branching based on the analysis results to generate appropriate actions and notify management:

[1019] python

[1020] def suggest_next_action(trend):

[1021] if trend['negative'] > trend['positive']:

[1022] return "Proposal to hold a workshop on improving progress"

[1023] else:

[1024] return "No follow-up required"

[1025] The above is the processing steps of the system and the specific operations that accompany them.

[1026] (Application example 1)

[1027] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1028] In modern brick-and-mortar stores, efficient staff management and improved service quality are key challenges. Traditional management methods make it difficult to properly collect and analyze feedback from staff, and they are not expected to take appropriate action quickly. Furthermore, it is necessary to accurately summarize the content of feedback and perform sentiment analysis to grasp trends and provide accurate notifications to managers.

[1029] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1030] In this invention, the server includes means for automatically acquiring comment data from a survey system, means for preprocessing the comment data, means for using a generative AI model to summarize the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends from past data based on the stored data, means for analyzing trends of individual respondents, means for suggesting next actions based on the analysis results, and means for visualizing feedback trends using the analysis results and notifying management. This enables efficient collection and analysis of staff feedback and prompt suggestion of appropriate next actions.

[1031] A "survey system" is a system for collecting feedback and opinions from users.

[1032] "Comment Data" refers to opinions and feedback in text form obtained from users through the survey system.

[1033] "Preprocessing" refers to the process of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format to make data analysis easier.

[1034] A "generative AI model" is an artificial intelligence algorithm that learns from large amounts of data and generates new sentences.

[1035] "Sentiment analysis" is a technique that analyzes the content of text data and determines whether the sentiment is positive, negative, or neutral.

[1036] A "database" is a digital storage system for systematically organizing and storing information.

[1037] "Feedback trends" are patterns and tendencies that emerge from the analysis of accumulated feedback data.

[1038] A "management position" is a position within an organization that is responsible for managing staff and running business operations.

[1039] An "HTTP request" is a communication protocol for requesting data from a web server.

[1040] An "API endpoint" is an interface for connecting with external systems and is a connection point for exchanging data.

[1041] This invention is a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program of this system operates based on the operations of the server, terminal, and user.

[1042] First, the server automatically retrieves comment data from the survey system. For example, it sends an HTTP request to an API endpoint to retrieve the latest survey data. The retrieved comment data is preprocessed to remove spaces, special characters, and emojis, and standardize the text format.

[1043] Next, a generative AI model is used on the preprocessed comment data to summarize the comments. For example, a comment such as "The customer service today was terrible" is converted into a summary sentence such as "There was a problem with the service." A sentiment analysis tool is also used to classify the sentiment of this comment as "negative."

[1044] The generated summary and the results of the sentiment analysis are stored in a database. This data includes the original comment, summary, sentiment analysis results, and timestamp. Based on the stored data, the server compares it with past data to analyze trends. For example, it analyzes whether there has been an increase in negative comments compared to data from the past few months. This trend analysis allows trends to be identified and necessary measures to be taken.

[1045] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, it extracts that tendency from the database and identifies staff members who need follow-up. Finally, based on these analysis results, the server proposes the next action to take and notifies management. For example, it may notify specific actions such as "proposing holding a workshop to improve poor customer service" or "recommending follow-up with specific staff members."

[1046] The terminal provides an interface for users to check survey results and analytical data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. Suggested next actions are also notified in real time. For example, specific feedback such as "Today's shift was very busy, but enjoyable" or "The handling of customer complaints did not go well" is displayed in a list. An example of a prompt is "Generate a summary of the following feedback comments: 'Today's shift was very busy, but enjoyable.'"

[1047] As a result, the present invention enables rapid collection and analysis of qualitative data, leading to efficient business operations and effective human resource management.

[1048] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1049] Step 1:

[1050] The server automatically retrieves comment data from the survey system. Specifically, it sends an HTTP request to a pre-configured API endpoint to retrieve the survey data. The input is the API endpoint, and the output is the retrieved comment data.

[1051] Step 2:

[1052] The server preprocesses the acquired comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. The input is the acquired comment data, and the output is the preprocessed comment data.

[1053] Step 3:

[1054] The server summarizes the preprocessed comment data using a generative AI model. Specifically, it converts the comment text into a summary sentence. For example, a comment such as "The customer service today was terrible" is converted into the summary sentence "There was a problem with the service." The input is the preprocessed comment data, and the output is the summarized text data.

[1055] Step 4:

[1056] The server performs sentiment analysis on the summarized comment data. Specifically, it uses a sentiment analysis tool to classify the sentiment of the comment as positive, negative, or neutral. For example, a comment that says "there was a problem with the response" is classified as "negative." The input is the summarized comment data, and the output is the result of the sentiment analysis.

[1057] Step 5:

[1058] The server stores the generated summary and the results of sentiment analysis in a database. Specifically, it saves the original text of the comment, the summarized text, the results of sentiment analysis, and a timestamp. The input is the summary and the results of sentiment analysis, and the output is the information stored in the database.

[1059] Step 6:

[1060] The server analyzes the trends in past data based on the accumulated data. Specifically, it extracts data from the past few months and analyzes whether there has been an increase in negative comments. The input is the past data accumulated in the database, and the output is the results of the trend analysis.

[1061] Step 7:

[1062] The server analyzes the tendencies of individual respondents. Specifically, if a particular respondent repeatedly makes negative comments, it extracts that tendency and identifies staff members who need follow-up. The input is the comment data stored in the database and the analysis results, and the output is a list of staff members who need follow-up.

[1063] Step 8:

[1064] The server proposes the next action based on the analysis results and notifies the manager. Specifically, it notifies specific actions such as "proposing holding a workshop to improve the worsening response" or "recommending follow-up with specific staff members." The input is the transition analysis results and individual trend analysis results, and the output is the notification content to the manager.

[1065] Step 9:

[1066] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. The input is the survey results and analysis data stored in the database, and the output is the visualized data displayed on the interface.

[1067] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1068] Understood. Below is the "Mode for Carrying Out the Invention" of the specification based on the invention combining the emotion engine.

[1069] This invention combines an emotion engine with a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program for this system operates based on operations by the server, terminal, and user.

[1070] First, the server automatically retrieves comment data from the survey system. To do this, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[1071] Next, the server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[1072] The server then uses a generative AI model to summarize the preprocessed comment data, for example, converting long comments into short summary sentences. This AI model uses natural language processing techniques to concisely express the main content of the comment.

[1073] The server then analyzes the sentiment of the comments using a sentiment analysis tool and a sentiment engine. Specifically, the sentiment engine categorizes the sentiment into categories such as positive, negative, and neutral. For example, a comment such as "The meeting went slowly" would be classified as "negative."

[1074] The server then stores the generated summary and the results of the sentiment analysis in a database. The data stored includes the original comment, summary, sentiment analysis results, and timestamp. This allows the analysis results to be stored centrally in the database.

[1075] Based on the accumulated data, the server compares it with past data and analyzes trends. By comparing data from multiple months and analyzing trends in sentiment and opinions, it can determine whether there has been an increase in negative comments at a particular time.

[1076] The server then analyzes the trends of individual respondents. Based on the respondent ID, data on specific respondents is extracted and its trends are analyzed. This process identifies the characteristics and patterns of respondents and identifies management candidates and employees who need follow-up.

[1077] Finally, the server generates specific action suggestions based on the results of these analyses, such as "suggest holding a workshop to improve progress" or "recommend following up on feedback from specific employees," and notifies managers.

[1078] The terminal provides an interface for users to view survey results and analytical data. Through the terminal, users can view summaries of quantitative and qualitative data, visualize the results in graphs and dashboards, and receive real-time notifications of suggested next steps.

[1079] The user can plan and implement specific follow-up actions and measures based on the information provided on the device. For example, the user can decide to hold a workshop and send a notification to employees.

[1080] This allows for the rapid collection and analysis of qualitative data, resulting in efficient business operations and human resource management. Furthermore, by combining it with an emotion engine, it is possible to recognize users' emotions and provide specific feedback and follow-up actions based on those emotions.

[1081] The processing flow will be explained below.

[1082] Understood. Below are the specific operations for each processing step.

[1083] Step 1:

[1084] The server automatically retrieves comment data from the survey system by sending an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[1085] Step 2:

[1086] The server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[1087] Step 3:

[1088] The server uses a generative AI model to summarize the preprocessed comment data. For example, a comment such as "The meeting was progressing slowly. There were probably too many agenda items" is converted into a summary such as "The slow progress and the large number of agenda items were pointed out." This AI model uses natural language processing technology to concisely express the main content of the comment.

[1089] Step 4:

[1090] The server analyzes the sentiment of the comments using a sentiment analysis tool and a sentiment engine. For example, a comment such as "The meeting went slowly" may be classified as "negative." The sentiment engine categorizes the sentiment into categories such as positive, negative, and neutral.

[1091] Step 5:

[1092] The server stores the generated summary and the results of sentiment analysis in a database. The stored data includes the original comment, summary, sentiment analysis results, and timestamp. This allows the analysis results to be stored centrally in the database.

[1093] Step 6:

[1094] The server analyzes the accumulated data and compares it with past data, analyzing trends in sentiment and opinions by comparing data from multiple months, for example, to determine whether negative comments are increasing at certain times of the day or during certain events.

[1095] Step 7:

[1096] The server analyzes the trends of individual respondents. Based on the respondent ID, it extracts the data of a specific respondent and analyzes their trends. For example, if a specific employee frequently makes negative comments, it analyzes the employee's trends and generates a detailed report.

[1097] Step 8:

[1098] The server then proposes the next action based on the analysis results. For example, it generates specific action suggestions such as "Suggest holding a workshop to improve progress" or "Recommend follow-up feedback for specific employees." These suggestions are notified to managers in real time via a notification system.

[1099] Step 9:

[1100] The device provides a user interface that displays survey results and analysis data. Through the device, users can view summaries of quantitative and qualitative data and visualize the results in the form of graphs and dashboards. For example, sentiment analysis results can be displayed using color codes for positive, negative, and neutral.

[1101] Step 10:

[1102] The user plans and implements specific follow-up actions and measures based on the information provided on the device. For example, the user decides to hold a proposed workshop and sends an email notification to employees. This allows the user to take the necessary action quickly.

[1103] Based on the above steps, this invention enables the rapid collection and analysis of qualitative data, realizing efficient business operations and human resource management. Furthermore, by combining it with an emotion engine, it is possible to recognize users' emotions and provide specific feedback and follow-up actions based on those emotions.

[1104] Example 2

[1105] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1106] In conventional survey systems, comment data is often collected and analyzed manually, which is not only inefficient but also makes it difficult to ensure data consistency and quality. Furthermore, because sentiment analysis and summarization are not automated, the vast amount of comment data must be reviewed one by one, placing a heavy burden on the person in charge. Furthermore, the process for linking analysis results to subsequent actions is unclear, making it difficult to respond quickly and appropriately. This creates challenges that make it difficult to optimize organizational operations and human resource management.

[1107] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for automatically acquiring comment data from a questionnaire system, means for preprocessing the comment data, means for summarizing the preprocessed comment data using a generative AI model, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing tendencies of individual respondents, means for suggesting next actions based on the analysis results, and means for displaying the analysis results and next actions through a user interface. This consistently automates the process from comment data collection to summarization, sentiment analysis, data storage, trend analysis, trend analysis, and suggestion of next actions, enabling efficient and accurate data analysis and rapid response.

[1108] A "survey system" is a system for collecting comments and opinions from survey subjects, and has the function of acquiring data in digital form.

[1109] "Comment data" refers to data that indicates opinions and impressions in text format collected through a survey system.

[1110] "Preprocessing" refers to the processing step of removing unnecessary spaces, special symbols, and emojis from the acquired comment data and standardizing the text format.

[1111] A "generative AI model" is a model that uses artificial intelligence techniques to analyze a given text and perform a specific purpose, such as summarization or sentiment analysis.

[1112] "Summarization" refers to the process of summarizing the original comment data in a short and concise form and extracting only the main content.

[1113] "Sentiment analysis" is the process of identifying and classifying sentiments, such as positive, negative, or neutral, from text in comment data.

[1114] A "database" is a collection of information that systematically stores collected and processed data so that it can be easily searched, referenced, and analyzed later.

[1115] "Trend analysis" refers to a statistical process that compares past data with current data to reveal fluctuations and trends.

[1116] "Trend analysis" is an analytical technique used to identify recurring patterns or characteristics associated with a particular respondent or data set.

[1117] "Next action suggestion" refers to the process of proposing specific actions or measures to be taken based on the analysis results.

[1118] "User interface" refers to the visual or operational means or screen configuration that allows a user to operate a system and check the results.

[1119] This system automatically preprocesses comment data obtained from a survey system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, and compares and analyzes them with past data. This system operates based on the operations of the server, terminals, and users.

[1120] Hardware and software used

[1121] The main hardware and software used in the present invention are as follows:

[1122] Server: Performs data acquisition, preprocessing, summary generation, sentiment analysis, data accumulation, analysis, and action suggestion.

[1123] Terminal: Provides an interface for users to view results and take next actions.

[1124] Generative AI models: Use natural language processing techniques to perform text summarization and sentiment analysis.

[1125] System operation explanation

[1126] Retrieving comment data

[1127] The server sends an HTTP request to the API endpoint of the survey system to retrieve the comment data. For example, to retrieve feedback on the monthly department meeting, the server sends the following request:

[1128] plaintext

[1129] GET / api / v1 / feedbacks

[1130] Host: survey.example.com

[1131] Authorization: Bearer<API_TOKEN>

[1132] The server extracts the comment field from the response data and converts it into a format that can be processed within the system.

[1133] Preprocessing comment data

[1134] The server pre-processes the retrieved comment data, mainly performing the following tasks:

[1135] Remove unnecessary spaces and special characters.

[1136] Remove emojis or convert them to text.

[1137] Use the same capitalization for text.

[1138] For example, if the comment is "The meeting was 👍," it will be converted to "The meeting was good." This will improve the data quality of the comment text.

[1139] Summary Generation

[1140] The server uses the generative AI model on the preprocessed comment data to summarize the comments. It gives the generative AI model the following prompts:

[1141] plaintext

[1142] Prompt: "Summarize the following comment: 'The meeting went very smoothly and the whole team was united. The results exceeded our expectations.'"

[1143] The model generates a summary in response to this prompt, returning the summary "The meeting went smoothly and produced good results."

[1144] sentiment analysis

[1145] The server performs sentiment analysis on the summary using the sentiment engine, and then provides the following prompts for the generated summary:

[1146] plaintext

[1147] Prompt: "Analyze the sentiment of the following sentence: 'The meeting went smoothly and produced good results.'"

[1148] The emotion engine classifies emotions into categories such as positive, negative, and neutral, and returns the result as "positive."

[1149] Accumulation in the database

[1150] The server stores the analysis results, including the original comment, summary, sentiment analysis results, and timestamps, in a database, allowing the analysis results to be stored centrally.

[1151] Data transition and trend analysis

[1152] The server compares the accumulated data with past data to analyze trends. At the same time, it analyzes the trends of specific respondents based on their respondent IDs to identify patterns and changes. For example, it uses the following SQL query:

[1153] sql

[1154] SELECT COUNT(), emotion

[1155] FROM feedback_analysis

[1156] WHERE timestamp >= '2023-01-01' AND timestamp <= '2023-12-31'

[1157] GROUP BY emotion;

[1158] Next action suggestion

[1159] Based on the analysis results, the server proposes the next action to take, such as "proposing holding a workshop to improve progress" or "recommending feedback follow-up for specific employees," and notifies the manager.

[1160] Terminal and user behavior

[1161] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. Proposed next actions are also notified in real time. This allows users to plan and implement specific follow-ups and measures. For example, a user can decide to hold a workshop and send a notification to employees.

[1162] This allows the present invention to rapidly collect and analyze qualitative data, contributing to the realization of efficient business operations and human resource management.

[1163] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1164] Step 1: Retrieving Comment Data

[1165] The server sends an HTTP request to the API endpoint of the survey system to retrieve the comment data. For example, the survey system might send the following request:

[1166] plaintext

[1167] GET / api / v1 / feedbacks

[1168] Host: survey.example.com

[1169] Authorization: Bearer<API_TOKEN>

[1170] The input is the API endpoint and authentication information. The output is the comment data retrieved in JSON format. The server extracts the comment field from the retrieved JSON data and converts it into a format that can be processed within the system.

[1171] Step 2: Preprocessing the comment data

[1172] The server pre-processes the retrieved comment data, mainly performing the following tasks:

[1173] Remove unnecessary spaces and special characters.

[1174] Remove emojis or convert them to text.

[1175] Use the same capitalization for text.

[1176] For example, if the input is comment data such as "The meeting was a good one," the output will be "The meeting was good."

[1177] Step 3: Summary generation

[1178] The server uses the generative AI model on the preprocessed comment data to summarize the comments. It gives the generative AI model the following prompts:

[1179] plaintext

[1180] Prompt: "Summarize the following comment: 'The meeting went very smoothly and the whole team was united. The results exceeded our expectations.'"

[1181] The input is the preprocessed comment data and a prompt for a summary, and the output is a summary obtained from the generative AI model, such as "The meeting went smoothly and produced good results."

[1182] Step 4: Sentiment analysis

[1183] The server performs sentiment analysis on the summary using the sentiment engine, and then provides the following prompts for the generated summary:

[1184] plaintext

[1185] Prompt: "Analyze the sentiment of the following sentence: 'The meeting went smoothly and produced good results.'"

[1186] The input is the generated summary and a prompt for sentiment analysis. The output is the sentiment analysis result of "positive" obtained from the sentiment engine.

[1187] Step 5: Database storage

[1188] The server stores the analysis results, including the original comment, summary, sentiment analysis, and timestamp, in a database. The inputs are the summary, sentiment analysis, original comment, and timestamp. The output is a database record containing these data. For example, use the following SQL query:

[1189] sql

[1190] INSERT INTO feedback_analysis (original_comment, summary, emotion, timestamp)

[1191] VALUES ('The meeting went very smoothly and the whole team was united. The results exceeded our expectations.', 'The meeting went smoothly and produced good results.', 'Positive', '2023-10-10 10:00:00');

[1192] Step 6: Data transition and trend analysis

[1193] The server compares the accumulated data with past data and analyzes trends. Next, it extracts data for specific respondents based on the respondent ID and analyzes their trends. The input is past data and question data for a specific period. The output is the results of trend analysis of emotions and opinions, as well as the results of trend analysis of specific respondents. For example, the following SQL query is used:

[1194] sql

[1195] SELECT COUNT(), emotion

[1196] FROM feedback_analysis

[1197] WHERE timestamp >= '2023-01-01' AND timestamp <= '2023-12-31'

[1198] GROUP BY emotion;

[1199] Step 7: Present next actions

[1200] The server suggests the next action based on the analysis results. Based on the analyzed data, it suggests necessary actions and follow-ups to staff. The input is the results of the transition and trend analysis. The output is a specific action suggestion. For example, it suggests actions such as "Suggest holding a workshop on improving progress" or "Recommend follow-up on feedback from specific employees." Managers are notified via email, etc.

[1201] email

[1202] To: manager@example.com

[1203] Subject: Next action suggestions

[1204] Body: I ​​suggest holding a workshop on improving progression.

[1205] Step 8: User confirmation and execution

[1206] The terminal provides an interface for users to check survey results and analysis data, and notifies them of proposed next actions in real time. The input is the analysis results and action suggestions sent from the server. The output is visual information and notifications displayed to the user. The user can plan and execute follow-ups and measures through the terminal. For example, a user decides to hold a workshop and sends a notification to employees:

[1207] email

[1208] To: all_employees@example.com

[1209] Subject: Workshop Announcement

[1210] Body: We're hosting a workshop on improving your progression. We'd love for you to join us.

[1211] (Application example 2)

[1212] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1213] Modern brick-and-mortar stores are required to quickly and accurately collect and analyze customer feedback and quickly take appropriate action based on it. However, conventional feedback collection systems have issues with time loss due to manual input and data inaccuracy. Furthermore, sentiment analysis and summarization of feedback are often done manually, making it difficult to respond quickly. In addition, the lack of real-time action suggestions hinders quick action to improve customer satisfaction.

[1214] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for automatically acquiring comment data from a questionnaire system, means for preprocessing the comment data, means for summarizing the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing trends of individual respondents, means for collecting customer feedback using smart glasses, and means for analyzing the collected customer feedback in real time and proposing the next action. This enables rapid collection and analysis of customer feedback and makes it possible to propose appropriate actions in real time.

[1215] A "survey system" is a mechanism for collecting comments and opinions from users.

[1216] "Comment data" is text data containing user opinions and impressions obtained from a questionnaire system.

[1217] "Preprocessing" refers to the process of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format.

[1218] "Summarization" refers to shortening long comment data and expressing the main content concisely.

[1219] "Sentiment analysis" refers to analyzing the sentiment of summarized comment data and classifying it into categories such as positive, negative, or neutral.

[1220] A "database" is a system for storing and managing information such as preprocessed comment data, summaries, and sentiment analysis results.

[1221] "Trend analysis" involves comparing accumulated feedback data with past data to analyze trends in emotions and opinions.

[1222] "Individual respondent trend analysis" refers to analyzing the data of specific respondents to understand their characteristics and patterns.

[1223] "Smart glasses" are glasses-type devices that have a display visible to the wearer and the ability to exchange audio, video, and data in real time.

[1224] "Real-time analysis" means analyzing collected data immediately and obtaining analysis results instantly.

[1225] "Next action proposals" are proposals for specific next steps or improvements based on the analysis results.

[1226] The present invention is a system for automatically collecting and analyzing customer feedback in a physical store and proposing next actions to staff in real time. An embodiment of this system will be described in detail below.

[1227] Program Generation and Processing Description

[1228] This system is constructed using hardware such as a server, smart glasses, and user terminals, as well as software such as generative AI models and emotion analysis tools.

[1229] Hardware and Software Configuration

[1230] Hardware:

[1231] Smart glasses (e.g., interactive eyeglasses)

[1232] Server (high-performance computer)

[1233] User device (e.g., tablet, smartphone)

[1234] software:

[1235] Generative AI models (e.g., models using natural language processing technology)

[1236] Sentiment analysis tools (e.g., software that analyzes the sentiment of text)

[1237] Database management systems (e.g., database software)

[1238] API management systems (e.g., software that manages API endpoints)

[1239] Server Procedure

[1240] 1. Feedback Collection:

[1241] Smart glasses are used to collect customer feedback in-store via voice input, where the voice input is based on requested prompts.

[1242] For example: "I like the coffee at this cafe, but they're slow to serve orders."

[1243] 2. Data Preprocessing:

[1244] The smart glasses convert the voice data into text data and send it to the server, where it removes unnecessary spaces, special characters, and emojis from the comment data and standardizes the text format.

[1245] 3. Summarization and Sentiment Analysis:

[1246] The preprocessed comment data is summarized using a generative AI model, and the summarized comment data is analyzed for sentiment using a sentiment analysis tool, categorizing it into categories such as positive, negative, and neutral.

[1247] Example: "The coffee is good, but the service is slow." (Emotion: Negative)

[1248] 4. Database accumulation and comparison with past data:

[1249] The analysis results are stored in a database that includes the original comments, summaries, sentiment analysis results, and timestamps. This data is used to compare with past data and analyze trends in sentiment and opinion.

[1250] 5. Individual respondent trend analysis and next action suggestions:

[1251] Based on the data of each individual respondent, trends are analyzed to identify specific characteristics and patterns, which then automatically suggest next steps and notify staff via the smart glasses and user devices.

[1252] For example: "Propose staff training to improve service speed."

[1253] Specific use cases

[1254] Below is an example of usage.

[1255] Example prompt sentence:

[1256] "I like the coffee at this cafe, but they're slow to serve orders."

[1257] Example of analysis results:

[1258] Summary: "Good coffee, but slow service."

[1259] Emotion: Negative

[1260] Suggestion: "Propose staff training to improve service speed."

[1261] In this way, we provide a system that automates the entire process from feedback collection to analysis and suggests next actions in real time, thereby improving customer satisfaction and streamlining store operations.

[1262] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1263] Step 1: Gather feedback

[1264] A user wears the smart glasses and provides voice feedback in the store. This feedback is captured by the smart glasses as voice input. The smart glasses use voice recognition technology to convert the voice data into text data and send it to the server.

[1265] Input: User's spoken feedback (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[1266] Output: Text data sent to the server

[1267] Step 2: Data Preprocessing

[1268] The server receives the text data sent from the smart glasses and performs preprocessing, which removes unnecessary spaces, special characters, and emojis from the comment data and standardizes the text format.

[1269] Input: Text data sent from smart glasses (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[1270] Output: Preprocessed text data (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[1271] Step 3: Summary

[1272] The server summarizes the preprocessed text data using a generative AI model, which analyzes the text data and generates a summary that succinctly expresses the main content.

[1273] Input: Preprocessed text data (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[1274] Output: Summarized text data (e.g., "The coffee is good, but the service is slow.")

[1275] Step 4: Sentiment analysis

[1276] The server analyzes the generated summary using a sentiment analysis tool, which classifies the sentiment of the summary into categories such as positive, negative, and neutral.

[1277] Input: Summarized text data (e.g., "The coffee is good, but the service is slow.")

[1278] Output: Sentiment analysis result (e.g. "negative")

[1279] Step 5: Database accumulation

[1280] The server stores the results of the sentiment analysis, the summary, and the original text data in a database, which also stores the analysis results and timestamps.

[1281] Input: Original text data, summary, sentiment analysis results, timestamp

[1282] Output: Data accumulation in database

[1283] Step 6: Analyze the progress

[1284] The server analyzes trends based on the feedback data stored in the database, comparing it with past data, thereby analyzing fluctuations in sentiment and opinion over a specific period of time.

[1285] Input: Feedback data stored in the database

[1286] Output: Trend analysis report

[1287] Step 7: Individual respondent trend analysis

[1288] The server extracts data from each respondent and analyzes trends and patterns, which allows for the identification of specific respondent characteristics and behavioral patterns.

[1289] Input: Individual respondent data

[1290] Output: Trend analysis report

[1291] Step 8: Proposed next actions

[1292] The server then proposes specific next steps based on the results of trend analysis, which are then sent to the smart glasses or the user's device, where staff can check them in real time.

[1293] Input: Trend analysis results, Trend analysis results

[1294] Output: Notification of next action suggestion (e.g. "Propose staff training to improve service speed.")

[1295] The above steps enable fast and accurate collection, analysis, and action proposal of customer feedback.

[1296] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1297] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1298] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1299] [Fourth embodiment]

[1300] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1301] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1302] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1303] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1304] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1305] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1306] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1307] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1308] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1309] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1310] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1311] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1312] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1313] Understood. Below is the "Mode for Carrying Out the Invention" of the specification.

[1314] This invention is a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program of this system operates based on the operations of the server, terminal, and user.

[1315] First, the server automatically retrieves comment data from the survey system. For example, to retrieve survey data for the department monthly meeting in August 2023, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data.

[1316] Next, the server preprocesses the acquired comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the accuracy of the analysis.

[1317] The server then uses a generative AI model to summarize the preprocessed comment data. For example, it converts a comment such as "The meeting was progressing slowly" into a summary statement such as "The delay in the progress was pointed out." The server then uses a sentiment analysis tool to classify the sentiment of the comment as "negative."

[1318] The server then stores the generated summary and the sentiment analysis results in a database. The stored data includes the original comment, summary, sentiment analysis results, and timestamp.

[1319] Based on the accumulated data, the server compares it with past data and analyzes trends. For example, it compares it with data from the past few months to see if there has been an increase in negative comments. This trend analysis allows you to understand trends and take necessary measures.

[1320] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, the server extracts this tendency from the database and identifies potential management positions or employees who need follow-up.

[1321] Finally, the server will suggest the next action based on the results of these analyses, such as "proposing holding a workshop to improve progress" or "recommending feedback follow-up for specific employees," and notify managers of specific actions.

[1322] The terminal provides an interface for users to view survey results and analytical data. Through the terminal, users can view summaries of quantitative and qualitative data, visualize the results in graphs and dashboards, and receive real-time notifications of suggested next steps.

[1323] Based on the information provided on the device, the user can plan and execute specific follow-up actions and measures. For example, the user can decide to hold a workshop and send a notification to employees.

[1324] As a result, the present invention enables rapid collection and analysis of qualitative data, leading to efficient business operations and effective human resource management.

[1325] The processing flow will be explained below.

[1326] Understood. Below are the specific steps of the process.

[1327] Step 1:

[1328] The server automatically retrieves comment data from the survey system by sending an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[1329] Step 2:

[1330] The server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[1331] Step 3:

[1332] The server then uses a generative AI model to summarize the preprocessed comment data. For example, it converts long comments into short summary sentences. This AI model uses natural language processing technology to concisely express the main content of the comments.

[1333] Step 4:

[1334] The server uses a sentiment analysis tool to analyze the sentiment of the comments. Specifically, it categorizes them into categories such as positive, negative, and neutral. For example, a comment such as "The meeting went slowly" would be classified as "negative."

[1335] Step 5:

[1336] The server stores the generated summaries and the results of sentiment analysis in a database. The stored data includes the original comments, summaries, sentiment analysis results, and timestamps. This allows the analysis results to be stored centrally in the database.

[1337] Step 6:

[1338] The server analyzes the trends in past data based on the accumulated data, comparing data from multiple months and analyzing trends in sentiment and opinions to determine whether negative comments are increasing at a particular time.

[1339] Step 7:

[1340] The server analyzes the trends of individual respondents. Based on the respondent ID, it extracts data for specific respondents and analyzes their trends. This process identifies the characteristics and patterns of respondents and identifies management candidates and employees who need follow-up.

[1341] Step 8:

[1342] The server then proposes the next action based on the analysis results, for example, generating specific action suggestions such as "proposing to hold a workshop on improving progress" or "recommending feedback follow-up for specific employees," and notifying the manager.

[1343] Step 9:

[1344] The terminal provides a user interface that displays survey results and analytical data to the user, allowing the user to view summaries of quantitative and qualitative data and visualize the results in the form of graphs and dashboards.

[1345] Step 10:

[1346] The user plans and executes specific follow-up actions and measures based on the information provided on the device. For example, the user decides to hold a workshop and sends a notification to employees.

[1347] Based on the above steps, the present invention enables rapid collection and analysis of qualitative data, thereby realizing efficient business operations and human resource management.

[1348] Example 1

[1349] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1350] Conventional survey analysis systems have struggled to efficiently process and analyze large amounts of comment data and propose effective actions based on that data. Furthermore, manually processing data consumes significant human resources and results in inconsistent analytical accuracy. There is a need for a system that can resolve these issues, quickly aggregate and analyze qualitative data, and achieve efficient business operations and effective human resource management.

[1351] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1352] In this invention, the server includes means for automatically acquiring comment data from a survey system, means for preprocessing the comment data, means for using a generative AI model to summarize the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing tendencies of individual respondents, and means for suggesting next actions based on the analysis results. This makes it possible to automatically process large amounts of comment data and propose specific actions based on highly accurate analysis results.

[1353] A "survey system" is an electronic survey device for collecting opinions and feedback from users.

[1354] "Comment data" is text data that includes user opinions and feedback collected through a questionnaire system.

[1355] "Preprocessing" is the procedure of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format.

[1356] A "generative AI model" is an artificial intelligence model that uses natural language processing to summarize, generate, or analyze text data.

[1357] "Sentiment analysis" is a method for automatically classifying or assessing the emotions contained in text data.

[1358] A "database" is an electronic system for efficiently storing, managing, and retrieving organized data.

[1359] "Transition analysis" is a method of analyzing temporal changes and trends based on accumulated data.

[1360] "Trend analysis" is a method of analyzing the behavioral patterns and opinion trends of specific users or groups.

[1361] "Action suggestion" is the process of suggesting the next course of action based on the results of data analysis.

[1362] This system automatically preprocesses comment data obtained from a survey system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. This system operates based on operations from the server, terminals, and users.

[1363] First, the server automatically retrieves comment data from the survey system. Specifically, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, to retrieve survey data for the department monthly meeting in August 2023, the server uses the API to collect the data. The retrieved data is saved in JSON format or similar.

[1364] The server then preprocesses the retrieved comment data. Specifically, it uses Python and natural language processing (NLP) tools (e.g., NLTK and SpaCy) to remove unnecessary spaces, special characters, and emojis from the comments and standardize the text format. This preprocessing improves the analysis accuracy of the generative AI model.

[1365] After preprocessing the comment data, the server uses a generative AI model (e.g., GPT-4) to summarize the comments. An example prompt is shown below.

[1366] Example prompt sentence:

[1367] Original comment:

[1368] "The meeting was slow"

[1369] Preprocessed text:

[1370] "Slow progress pointed out"

[1371] Generate a summary:

[1372] When this prompt is fed into a generative AI model, the model generates the summary statement, "Delays in progress have been noted."

[1373] The server also performs sentiment analysis on the generated summary using tools such as TextBlob and VADER. For example, a comment such as "Delays in progress have been noted" is classified as "negative."

[1374] The server then stores the generated summary and the results of sentiment analysis in a database. The data to be saved includes the original comment, summary, sentiment analysis results, and timestamp. The database can be MySQL or PostgreSQL.

[1375] The server analyzes the accumulated data by comparing it with past data. For example, it analyzes whether there has been an increase in negative comments compared to data from the past few months. This analysis is performed using statistical analysis libraries such as pandas and numpy in Python.

[1376] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, the server can extract this tendency from the database. This analysis can then be used to identify potential management candidates or employees who need follow-up.

[1377] Finally, the server will suggest the next action based on the results of these analyses, such as "Suggest holding a workshop to improve progress" or "Recommend follow-up on feedback from specific employees," and notify managers of specific actions. This notification can be sent via email or via a dashboard within the system.

[1378] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and view the results visualized in graphs and dashboards. For example, graphs can be generated using JavaScript D3.js and Chart.js, allowing users to interactively check the analysis results. It can also display suggested next actions with real-time notifications.

[1379] Users can plan and implement specific follow-up actions and measures based on the information provided on their devices. For example, they can schedule workshops to improve progress and send notifications to employees. In this way, efficient business operations and effective human resource management are realized.

[1380] As described above, the present invention is a system that enables rapid collection and analysis of qualitative data and aims to improve and streamline business operations by proposing specific actions.

[1381] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1382] Step 1:

[1383] Data Acquisition

[1384] The server sends an HTTP request to the API endpoint to automatically retrieve comment data from the survey system. For example, the server sends a GET request to "https: / / api.example.com / surveys / latest" and receives JSON-formatted comment data as a response. The input is the API endpoint and authentication information (e.g., OAuth token), and the output is the retrieved JSON-formatted comment data. Specifically, data is retrieved from the API using a Python library (such as requests).

[1385] Step 2:

[1386] Data Preprocessing

[1387] The server preprocesses the comment data it receives. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments, and converts all text to a unified format (for example, lowercase). This process uses Python's regular expression module (re) and natural language processing libraries (NLTK and SpaCy). The input is comment data in JSON format, and the output is preprocessed string data. The specific operation is performed using the following code:

[1388] python

[1389] import re

[1390] import spacy

[1391] def preprocess_comment(comment):

[1392] comment = re.sub(r'[^\w\s]', '', comment) Remove special characters

[1393] comment = comment.lower() Convert all to lowercase

[1394] return comment

[1395] Step 3:

[1396] Summary Generation

[1397] The server inputs the preprocessed comment data into a generative AI model (e.g., GPT-4) to generate a summary of the comments. The prompt used here is:

[1398] input:

[1399] Original comment:

[1400] "The meeting was slow"

[1401] Preprocessed text:

[1402] "Slow progress pointed out"

[1403] Generate a summary:

[1404] The output is a summary sentence returned from the generative AI model. Specifically, the prompt sentence is sent to the API of the generative AI model and the summary sentence is received.

[1405] Step 4:

[1406] sentiment analysis

[1407] The server performs sentiment analysis on the generated summary. For sentiment analysis, it uses natural language processing tools such as TextBlob and VADER. The input is the generated summary, and the output is the sentiment classification result of the summary (e.g., negative, positive, neutral). To see the specific operation, use the following code:

[1408] python

[1409] from textblob import TextBlob

[1410] def analyze_sentiment(text):

[1411] analysis = TextBlob(text)

[1412] if analysis.sentiment.polarity > 0:

[1413] return "affirmative"

[1414] elif analysis.sentiment.polarity == 0:

[1415] return "neutral"

[1416] else:

[1417] return "negative"

[1418] Step 5:

[1419] Saving to a database

[1420] The server stores the original comment, the generated summary, the sentiment analysis result, and the timestamp in a database. The input is the original comment, the summary, the sentiment result, and the timestamp, and the output is an entry stored in the database. Specifically, the data is stored in the database using an SQL insert query:

[1421] sql

[1422] INSERT INTO comments (original_comment, summary, sentiment, timestamp)

[1423] VALUES (%s, %s, %s, %s);

[1424] Step 6:

[1425] Trend analysis

[1426] The server analyzes the trends of past data based on the accumulated data. The input is past comment data retrieved from the database, and the output is the results of the trend analysis. Specifically, it performs statistical analysis using the Python pandas library:

[1427] python

[1428] import pandas as pd

[1429] def trend_analysis(comments):

[1430] df = pd.DataFrame(comments)

[1431] trend = df['sentiment'].value_counts()

[1432] return trend

[1433] Step 7:

[1434] Analysis of user trends

[1435] The server analyzes the trends of individual respondents. The input is comment data related to a specific respondent, and the output is the specific trend analysis results. Specifically, it extracts data for a specific respondent by filtering and performs analysis:

[1436] python

[1437] def user_trend_analysis(comments, user_id):

[1438] user_comments = [comment for comment in comments if comment['user_id'] == user_id]

[1439] trend = trend_analysis(user_comments)

[1440] return trend

[1441] Step 8:

[1442] Suggested next actions

[1443] The server proposes the next action based on the analysis results. The input is the transition analysis results and trend analysis results, and the output is a specific action proposal (e.g., a proposal to hold a workshop). Specifically, it performs conditional branching based on the analysis results to generate appropriate actions and notify management:

[1444] python

[1445] def suggest_next_action(trend):

[1446] if trend['negative'] > trend['positive']:

[1447] return "Proposal to hold a workshop on improving progress"

[1448] else:

[1449] return "No follow-up required"

[1450] The above is the processing steps of the system and the specific operations that accompany them.

[1451] (Application example 1)

[1452] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1453] In modern brick-and-mortar stores, efficient staff management and improved service quality are key challenges. Traditional management methods make it difficult to properly collect and analyze feedback from staff, and they are not expected to take appropriate action quickly. Furthermore, it is necessary to accurately summarize the content of feedback and perform sentiment analysis to grasp trends and provide accurate notifications to managers.

[1454] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1455] In this invention, the server includes means for automatically acquiring comment data from a survey system, means for preprocessing the comment data, means for using a generative AI model to summarize the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends from past data based on the stored data, means for analyzing trends of individual respondents, means for suggesting next actions based on the analysis results, and means for visualizing feedback trends using the analysis results and notifying management. This enables efficient collection and analysis of staff feedback and prompt suggestion of appropriate next actions.

[1456] A "survey system" is a system for collecting feedback and opinions from users.

[1457] "Comment Data" refers to opinions and feedback in text form obtained from users through the survey system.

[1458] "Preprocessing" refers to the process of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format to make data analysis easier.

[1459] A "generative AI model" is an artificial intelligence algorithm that learns from large amounts of data and generates new sentences.

[1460] "Sentiment analysis" is a technique that analyzes the content of text data and determines whether the sentiment is positive, negative, or neutral.

[1461] A "database" is a digital storage system for systematically organizing and storing information.

[1462] "Feedback trends" are patterns and tendencies that emerge from the analysis of accumulated feedback data.

[1463] A "management position" is a position within an organization that is responsible for managing staff and running business operations.

[1464] An "HTTP request" is a communication protocol for requesting data from a web server.

[1465] An "API endpoint" is an interface for connecting with external systems and is a connection point for exchanging data.

[1466] This invention is a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program of this system operates based on the operations of the server, terminal, and user.

[1467] First, the server automatically retrieves comment data from the survey system. For example, it sends an HTTP request to an API endpoint to retrieve the latest survey data. The retrieved comment data is preprocessed to remove spaces, special characters, and emojis, and standardize the text format.

[1468] Next, a generative AI model is used on the preprocessed comment data to summarize the comments. For example, a comment such as "The customer service today was terrible" is converted into a summary sentence such as "There was a problem with the service." A sentiment analysis tool is also used to classify the sentiment of this comment as "negative."

[1469] The generated summary and the results of the sentiment analysis are stored in a database. This data includes the original comment, summary, sentiment analysis results, and timestamp. Based on the stored data, the server compares it with past data to analyze trends. For example, it analyzes whether there has been an increase in negative comments compared to data from the past few months. This trend analysis allows trends to be identified and necessary measures to be taken.

[1470] The server also analyzes the tendencies of individual respondents. For example, if a particular respondent repeatedly makes negative comments, it extracts that tendency from the database and identifies staff members who need follow-up. Finally, based on these analysis results, the server proposes the next action to take and notifies management. For example, it may notify specific actions such as "proposing holding a workshop to improve poor customer service" or "recommending follow-up with specific staff members."

[1471] The terminal provides an interface for users to check survey results and analytical data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. Suggested next actions are also notified in real time. For example, specific feedback such as "Today's shift was very busy, but enjoyable" or "The handling of customer complaints did not go well" is displayed in a list. An example of a prompt is "Generate a summary of the following feedback comments: 'Today's shift was very busy, but enjoyable.'"

[1472] As a result, the present invention enables rapid collection and analysis of qualitative data, leading to efficient business operations and effective human resource management.

[1473] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1474] Step 1:

[1475] The server automatically retrieves comment data from the survey system. Specifically, it sends an HTTP request to a pre-configured API endpoint to retrieve the survey data. The input is the API endpoint, and the output is the retrieved comment data.

[1476] Step 2:

[1477] The server preprocesses the acquired comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. The input is the acquired comment data, and the output is the preprocessed comment data.

[1478] Step 3:

[1479] The server summarizes the preprocessed comment data using a generative AI model. Specifically, it converts the comment text into a summary sentence. For example, a comment such as "The customer service today was terrible" is converted into the summary sentence "There was a problem with the service." The input is the preprocessed comment data, and the output is the summarized text data.

[1480] Step 4:

[1481] The server performs sentiment analysis on the summarized comment data. Specifically, it uses a sentiment analysis tool to classify the sentiment of the comment as positive, negative, or neutral. For example, a comment that says "there was a problem with the response" is classified as "negative." The input is the summarized comment data, and the output is the result of the sentiment analysis.

[1482] Step 5:

[1483] The server stores the generated summary and the results of sentiment analysis in a database. Specifically, it saves the original text of the comment, the summarized text, the results of sentiment analysis, and a timestamp. The input is the summary and the results of sentiment analysis, and the output is the information stored in the database.

[1484] Step 6:

[1485] The server analyzes the trends in past data based on the accumulated data. Specifically, it extracts data from the past few months and analyzes whether there has been an increase in negative comments. The input is the past data accumulated in the database, and the output is the results of the trend analysis.

[1486] Step 7:

[1487] The server analyzes the tendencies of individual respondents. Specifically, if a particular respondent repeatedly makes negative comments, it extracts that tendency and identifies staff members who need follow-up. The input is the comment data stored in the database and the analysis results, and the output is a list of staff members who need follow-up.

[1488] Step 8:

[1489] The server proposes the next action based on the analysis results and notifies the manager. Specifically, it notifies specific actions such as "proposing holding a workshop to improve the worsening response" or "recommending follow-up with specific staff members." The input is the transition analysis results and individual trend analysis results, and the output is the notification content to the manager.

[1490] Step 9:

[1491] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. The input is the survey results and analysis data stored in the database, and the output is the visualized data displayed on the interface.

[1492] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1493] Understood. Below is the "Mode for Carrying Out the Invention" of the specification based on the invention combining the emotion engine.

[1494] This invention combines an emotion engine with a system that automatically preprocesses comment data obtained from a questionnaire system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, compares them with past data to analyze trends, and analyzes the tendencies of individual respondents, automatically suggesting the next course of action. The program for this system operates based on operations by the server, terminal, and user.

[1495] First, the server automatically retrieves comment data from the survey system. To do this, it sends an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[1496] Next, the server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[1497] The server then uses a generative AI model to summarize the preprocessed comment data, for example, converting long comments into short summary sentences. This AI model uses natural language processing techniques to concisely express the main content of the comment.

[1498] The server then analyzes the sentiment of the comments using a sentiment analysis tool and a sentiment engine. Specifically, the sentiment engine categorizes the sentiment into categories such as positive, negative, and neutral. For example, a comment such as "The meeting went slowly" would be classified as "negative."

[1499] The server then stores the generated summary and the results of the sentiment analysis in a database. The data stored includes the original comment, summary, sentiment analysis results, and timestamp. This allows the analysis results to be stored centrally in the database.

[1500] Based on the accumulated data, the server compares it with past data and analyzes trends. By comparing data from multiple months and analyzing trends in sentiment and opinions, it can determine whether there has been an increase in negative comments at a particular time.

[1501] The server then analyzes the trends of individual respondents. Based on the respondent ID, data on specific respondents is extracted and its trends are analyzed. This process identifies the characteristics and patterns of respondents and identifies management candidates and employees who need follow-up.

[1502] Finally, the server generates specific action suggestions based on the results of these analyses, such as "suggest holding a workshop to improve progress" or "recommend following up on feedback from specific employees," and notifies managers.

[1503] The terminal provides an interface for users to view survey results and analytical data. Through the terminal, users can view summaries of quantitative and qualitative data, visualize the results in graphs and dashboards, and receive real-time notifications of suggested next steps.

[1504] The user can plan and implement specific follow-up actions and measures based on the information provided on the device. For example, the user can decide to hold a workshop and send a notification to employees.

[1505] This allows for the rapid collection and analysis of qualitative data, resulting in efficient business operations and human resource management. Furthermore, by combining it with an emotion engine, it is possible to recognize users' emotions and provide specific feedback and follow-up actions based on those emotions.

[1506] The processing flow will be explained below.

[1507] Understood. Below are the specific operations for each processing step.

[1508] Step 1:

[1509] The server automatically retrieves comment data from the survey system by sending an HTTP request to a pre-configured API endpoint to retrieve the latest survey data. For example, it retrieves survey data from monthly department meetings and feedback from general meetings.

[1510] Step 2:

[1511] The server preprocesses the retrieved comment data. Specifically, it removes unnecessary spaces, special characters, and emojis from the comments and standardizes the text format. This preprocessing improves the data quality of the comment text.

[1512] Step 3:

[1513] The server uses a generative AI model to summarize the preprocessed comment data. For example, a comment such as "The meeting was progressing slowly. There were probably too many agenda items" is converted into a summary such as "The slow progress and the large number of agenda items were pointed out." This AI model uses natural language processing technology to concisely express the main content of the comment.

[1514] Step 4:

[1515] The server analyzes the sentiment of the comments using a sentiment analysis tool and a sentiment engine. For example, a comment such as "The meeting went slowly" may be classified as "negative." The sentiment engine categorizes the sentiment into categories such as positive, negative, and neutral.

[1516] Step 5:

[1517] The server stores the generated summary and the results of sentiment analysis in a database. The stored data includes the original comment, summary, sentiment analysis results, and timestamp. This allows the analysis results to be stored centrally in the database.

[1518] Step 6:

[1519] The server analyzes the accumulated data and compares it with past data, analyzing trends in sentiment and opinions by comparing data from multiple months, for example, to determine whether negative comments are increasing at certain times of the day or during certain events.

[1520] Step 7:

[1521] The server analyzes the trends of individual respondents. Based on the respondent ID, it extracts the data of a specific respondent and analyzes their trends. For example, if a specific employee frequently makes negative comments, it analyzes the employee's trends and generates a detailed report.

[1522] Step 8:

[1523] The server then proposes the next action based on the analysis results. For example, it generates specific action suggestions such as "Suggest holding a workshop to improve progress" or "Recommend follow-up feedback for specific employees." These suggestions are notified to managers in real time via a notification system.

[1524] Step 9:

[1525] The device provides a user interface that displays survey results and analysis data. Through the device, users can view summaries of quantitative and qualitative data and visualize the results in the form of graphs and dashboards. For example, sentiment analysis results can be displayed using color codes for positive, negative, and neutral.

[1526] Step 10:

[1527] The user plans and implements specific follow-up actions and measures based on the information provided on the device. For example, the user decides to hold a proposed workshop and sends an email notification to employees. This allows the user to take the necessary action quickly.

[1528] Based on the above steps, this invention enables the rapid collection and analysis of qualitative data, realizing efficient business operations and human resource management. Furthermore, by combining it with an emotion engine, it is possible to recognize users' emotions and provide specific feedback and follow-up actions based on those emotions.

[1529] Example 2

[1530] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1531] In conventional survey systems, comment data is often collected and analyzed manually, which is not only inefficient but also makes it difficult to ensure data consistency and quality. Furthermore, because sentiment analysis and summarization are not automated, the vast amount of comment data must be reviewed one by one, placing a heavy burden on the person in charge. Furthermore, the process for linking analysis results to subsequent actions is unclear, making it difficult to respond quickly and appropriately. This creates challenges that make it difficult to optimize organizational operations and human resource management.

[1532] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for automatically acquiring comment data from a questionnaire system, means for preprocessing the comment data, means for summarizing the preprocessed comment data using a generative AI model, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing tendencies of individual respondents, means for suggesting next actions based on the analysis results, and means for displaying the analysis results and next actions through a user interface. This consistently automates the process from comment data collection to summarization, sentiment analysis, data storage, trend analysis, trend analysis, and suggestion of next actions, enabling efficient and accurate data analysis and rapid response.

[1533] A "survey system" is a system for collecting comments and opinions from survey subjects, and has the function of acquiring data in digital form.

[1534] "Comment data" refers to data that indicates opinions and impressions in text format collected through a survey system.

[1535] "Preprocessing" refers to the processing step of removing unnecessary spaces, special symbols, and emojis from the acquired comment data and standardizing the text format.

[1536] A "generative AI model" is a model that uses artificial intelligence techniques to analyze a given text and perform a specific purpose, such as summarization or sentiment analysis.

[1537] "Summarization" refers to the process of summarizing the original comment data in a short and concise form and extracting only the main content.

[1538] "Sentiment analysis" is the process of identifying and classifying sentiments, such as positive, negative, or neutral, from text in comment data.

[1539] A "database" is a collection of information that systematically stores collected and processed data so that it can be easily searched, referenced, and analyzed later.

[1540] "Trend analysis" refers to a statistical process that compares past data with current data to reveal fluctuations and trends.

[1541] "Trend analysis" is an analytical technique used to identify recurring patterns or characteristics associated with a particular respondent or data set.

[1542] "Next action suggestion" refers to the process of proposing specific actions or measures to be taken based on the analysis results.

[1543] "User interface" refers to the visual or operational means or screen configuration that allows a user to operate a system and check the results.

[1544] This system automatically preprocesses comment data obtained from a survey system, performs summarization and sentiment analysis using a generative AI model, stores the results in a database, and compares and analyzes them with past data. This system operates based on the operations of the server, terminals, and users.

[1545] Hardware and software used

[1546] The main hardware and software used in the present invention are as follows:

[1547] Server: Performs data acquisition, preprocessing, summary generation, sentiment analysis, data accumulation, analysis, and action suggestion.

[1548] Terminal: Provides an interface for users to view results and take next actions.

[1549] Generative AI models: Use natural language processing techniques to perform text summarization and sentiment analysis.

[1550] System operation explanation

[1551] Retrieving comment data

[1552] The server sends an HTTP request to the API endpoint of the survey system to retrieve the comment data. For example, to retrieve feedback on the monthly department meeting, the server sends the following request:

[1553] plaintext

[1554] GET / api / v1 / feedbacks

[1555] Host: survey.example.com

[1556] Authorization: Bearer<API_TOKEN>

[1557] The server extracts the comment field from the response data and converts it into a format that can be processed within the system.

[1558] Preprocessing comment data

[1559] The server pre-processes the retrieved comment data, mainly performing the following tasks:

[1560] Remove unnecessary spaces and special characters.

[1561] Remove emojis or convert them to text.

[1562] Use the same capitalization for text.

[1563] For example, if the comment is "The meeting was 👍," it will be converted to "The meeting was good." This will improve the data quality of the comment text.

[1564] Summary Generation

[1565] The server uses the generative AI model on the preprocessed comment data to summarize the comments. It gives the generative AI model the following prompts:

[1566] plaintext

[1567] Prompt: "Summarize the following comment: 'The meeting went very smoothly and the whole team was united. The results exceeded our expectations.'"

[1568] The model generates a summary in response to this prompt, returning the summary "The meeting went smoothly and produced good results."

[1569] sentiment analysis

[1570] The server performs sentiment analysis on the summary using the sentiment engine, and then provides the following prompts for the generated summary:

[1571] plaintext

[1572] Prompt: "Analyze the sentiment of the following sentence: 'The meeting went smoothly and produced good results.'"

[1573] The emotion engine classifies emotions into categories such as positive, negative, and neutral, and returns the result as "positive."

[1574] Accumulation in the database

[1575] The server stores the analysis results, including the original comment, summary, sentiment analysis results, and timestamps, in a database, allowing the analysis results to be stored centrally.

[1576] Data transition and trend analysis

[1577] The server compares the accumulated data with past data to analyze trends. At the same time, it analyzes the trends of specific respondents based on their respondent IDs to identify patterns and changes. For example, it uses the following SQL query:

[1578] sql

[1579] SELECT COUNT(), emotion

[1580] FROM feedback_analysis

[1581] WHERE timestamp >= '2023-01-01' AND timestamp <= '2023-12-31'

[1582] GROUP BY emotion;

[1583] Next action suggestion

[1584] Based on the analysis results, the server proposes the next action to take, such as "proposing holding a workshop to improve progress" or "recommending feedback follow-up for specific employees," and notifies the manager.

[1585] Terminal and user behavior

[1586] The terminal provides an interface for users to check survey results and analysis data. Through the terminal, users can check summaries of quantitative and qualitative data and visualize the results in graphs and dashboards. Proposed next actions are also notified in real time. This allows users to plan and implement specific follow-ups and measures. For example, a user can decide to hold a workshop and send a notification to employees.

[1587] This allows the present invention to rapidly collect and analyze qualitative data, contributing to the realization of efficient business operations and human resource management.

[1588] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1589] Step 1: Retrieving Comment Data

[1590] The server sends an HTTP request to the API endpoint of the survey system to retrieve the comment data. For example, the survey system might send the following request:

[1591] plaintext

[1592] GET / api / v1 / feedbacks

[1593] Host: survey.example.com

[1594] Authorization: Bearer<API_TOKEN>

[1595] The input is the API endpoint and authentication information. The output is the comment data retrieved in JSON format. The server extracts the comment field from the retrieved JSON data and converts it into a format that can be processed within the system.

[1596] Step 2: Preprocessing the comment data

[1597] The server pre-processes the retrieved comment data, mainly performing the following tasks:

[1598] Remove unnecessary spaces and special characters.

[1599] Remove emojis or convert them to text.

[1600] Use the same capitalization for text.

[1601] For example, if the input is comment data such as "The meeting was a good one," the output will be "The meeting was good."

[1602] Step 3: Summary generation

[1603] The server uses the generative AI model on the preprocessed comment data to summarize the comments. It gives the generative AI model the following prompts:

[1604] plaintext

[1605] Prompt: "Summarize the following comment: 'The meeting went very smoothly and the whole team was united. The results exceeded our expectations.'"

[1606] The input is the preprocessed comment data and a prompt for a summary, and the output is a summary obtained from the generative AI model, such as "The meeting went smoothly and produced good results."

[1607] Step 4: Sentiment analysis

[1608] The server performs sentiment analysis on the summary using the sentiment engine, and then provides the following prompts for the generated summary:

[1609] plaintext

[1610] Prompt: "Analyze the sentiment of the following sentence: 'The meeting went smoothly and produced good results.'"

[1611] The input is the generated summary and a prompt for sentiment analysis. The output is the sentiment analysis result of "positive" obtained from the sentiment engine.

[1612] Step 5: Database storage

[1613] The server stores the analysis results, including the original comment, summary, sentiment analysis, and timestamp, in a database. The inputs are the summary, sentiment analysis, original comment, and timestamp. The output is a database record containing these data. For example, use the following SQL query:

[1614] sql

[1615] INSERT INTO feedback_analysis (original_comment, summary, emotion, timestamp)

[1616] VALUES ('The meeting went very smoothly and the whole team was united. The results exceeded our expectations.', 'The meeting went smoothly and produced good results.', 'Positive', '2023-10-10 10:00:00');

[1617] Step 6: Data transition and trend analysis

[1618] The server compares the accumulated data with past data and analyzes trends. Next, it extracts data for specific respondents based on the respondent ID and analyzes their trends. The input is past data and question data for a specific period. The output is the results of trend analysis of emotions and opinions, as well as the results of trend analysis of specific respondents. For example, the following SQL query is used:

[1619] sql

[1620] SELECT COUNT(), emotion

[1621] FROM feedback_analysis

[1622] WHERE timestamp >= '2023-01-01' AND timestamp <= '2023-12-31'

[1623] GROUP BY emotion;

[1624] Step 7: Present next actions

[1625] The server suggests the next action based on the analysis results. Based on the analyzed data, it suggests necessary actions and follow-ups to staff. The input is the results of the transition and trend analysis. The output is a specific action suggestion. For example, it suggests actions such as "Suggest holding a workshop on improving progress" or "Recommend follow-up on feedback from specific employees." Managers are notified via email, etc.

[1626] email

[1627] To: manager@example.com

[1628] Subject: Next action suggestions

[1629] Body: I ​​suggest holding a workshop on improving progression.

[1630] Step 8: User confirmation and execution

[1631] The terminal provides an interface for users to check survey results and analysis data, and notifies them of proposed next actions in real time. The input is the analysis results and action suggestions sent from the server. The output is visual information and notifications displayed to the user. The user can plan and execute follow-ups and measures through the terminal. For example, a user decides to hold a workshop and sends a notification to employees:

[1632] email

[1633] To: all_employees@example.com

[1634] Subject: Workshop Announcement

[1635] Body: We're hosting a workshop on improving your progression. We'd love for you to join us.

[1636] (Application example 2)

[1637] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1638] Modern brick-and-mortar stores are required to quickly and accurately collect and analyze customer feedback and quickly take appropriate action based on it. However, conventional feedback collection systems have issues with time loss due to manual input and data inaccuracy. Furthermore, sentiment analysis and summarization of feedback are often done manually, making it difficult to respond quickly. In addition, the lack of real-time action suggestions hinders quick action to improve customer satisfaction.

[1639] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for automatically acquiring comment data from a questionnaire system, means for preprocessing the comment data, means for summarizing the preprocessed comment data, means for analyzing the sentiment of the summarized comment data, means for storing the analysis results in a database, means for analyzing trends with past data based on the stored data, means for analyzing trends of individual respondents, means for collecting customer feedback using smart glasses, and means for analyzing the collected customer feedback in real time and proposing the next action. This enables rapid collection and analysis of customer feedback and makes it possible to propose appropriate actions in real time.

[1640] A "survey system" is a mechanism for collecting comments and opinions from users.

[1641] "Comment data" is text data containing user opinions and impressions obtained from a questionnaire system.

[1642] "Preprocessing" refers to the process of removing unnecessary spaces, special characters, and emojis from comment data and standardizing the text format.

[1643] "Summarization" refers to shortening long comment data and expressing the main content concisely.

[1644] "Sentiment analysis" refers to analyzing the sentiment of summarized comment data and classifying it into categories such as positive, negative, or neutral.

[1645] A "database" is a system for storing and managing information such as preprocessed comment data, summaries, and sentiment analysis results.

[1646] "Trend analysis" involves comparing accumulated feedback data with past data to analyze trends in emotions and opinions.

[1647] "Individual respondent trend analysis" refers to analyzing the data of specific respondents to understand their characteristics and patterns.

[1648] "Smart glasses" are glasses-type devices that have a display visible to the wearer and the ability to exchange audio, video, and data in real time.

[1649] "Real-time analysis" means analyzing collected data immediately and obtaining analysis results instantly.

[1650] "Next action proposals" are proposals for specific next steps or improvements based on the analysis results.

[1651] The present invention is a system for automatically collecting and analyzing customer feedback in a physical store and proposing next actions to staff in real time. An embodiment of this system will be described in detail below.

[1652] Program Generation and Processing Description

[1653] This system is constructed using hardware such as a server, smart glasses, and user terminals, as well as software such as generative AI models and emotion analysis tools.

[1654] Hardware and Software Configuration

[1655] Hardware:

[1656] Smart glasses (e.g., interactive eyeglasses)

[1657] Server (high-performance computer)

[1658] User device (e.g., tablet, smartphone)

[1659] software:

[1660] Generative AI models (e.g., models using natural language processing technology)

[1661] Sentiment analysis tools (e.g., software that analyzes the sentiment of text)

[1662] Database management systems (e.g., database software)

[1663] API management systems (e.g., software that manages API endpoints)

[1664] Server Procedure

[1665] 1. Feedback Collection:

[1666] Smart glasses are used to collect customer feedback in-store via voice input, where the voice input is based on requested prompts.

[1667] For example: "I like the coffee at this cafe, but they're slow to serve orders."

[1668] 2. Data Preprocessing:

[1669] The smart glasses convert the voice data into text data and send it to the server, where it removes unnecessary spaces, special characters, and emojis from the comment data and standardizes the text format.

[1670] 3. Summarization and Sentiment Analysis:

[1671] The preprocessed comment data is summarized using a generative AI model, and the summarized comment data is analyzed for sentiment using a sentiment analysis tool, categorizing it into categories such as positive, negative, and neutral.

[1672] Example: "The coffee is good, but the service is slow." (Emotion: Negative)

[1673] 4. Database accumulation and comparison with past data:

[1674] The analysis results are stored in a database that includes the original comments, summaries, sentiment analysis results, and timestamps. This data is used to compare with past data and analyze trends in sentiment and opinion.

[1675] 5. Individual respondent trend analysis and next action suggestions:

[1676] Based on the data of each individual respondent, trends are analyzed to identify specific characteristics and patterns, which then automatically suggest next steps and notify staff via the smart glasses and user devices.

[1677] For example: "Propose staff training to improve service speed."

[1678] Specific use cases

[1679] Below is an example of usage.

[1680] Example prompt sentence:

[1681] "I like the coffee at this cafe, but they're slow to serve orders."

[1682] Example of analysis results:

[1683] Summary: "Good coffee, but slow service."

[1684] Emotion: Negative

[1685] Suggestion: "Propose staff training to improve service speed."

[1686] In this way, we provide a system that automates the entire process from feedback collection to analysis and suggests next actions in real time, thereby improving customer satisfaction and streamlining store operations.

[1687] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1688] Step 1: Gather feedback

[1689] A user wears the smart glasses and provides voice feedback in the store. This feedback is captured by the smart glasses as voice input. The smart glasses use voice recognition technology to convert the voice data into text data and send it to the server.

[1690] Input: User's spoken feedback (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[1691] Output: Text data sent to the server

[1692] Step 2: Data Preprocessing

[1693] The server receives the text data sent from the smart glasses and performs preprocessing, which removes unnecessary spaces, special characters, and emojis from the comment data and standardizes the text format.

[1694] Input: Text data sent from smart glasses (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[1695] Output: Preprocessed text data (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[1696] Step 3: Summary

[1697] The server summarizes the preprocessed text data using a generative AI model, which analyzes the text data and generates a summary that succinctly expresses the main content.

[1698] Input: Preprocessed text data (e.g., "I like the coffee at this cafe, but they're slow to serve my order.")

[1699] Output: Summarized text data (e.g., "The coffee is good, but the service is slow.")

[1700] Step 4: Sentiment analysis

[1701] The server analyzes the generated summary using a sentiment analysis tool, which classifies the sentiment of the summary into categories such as positive, negative, and neutral.

[1702] Input: Summarized text data (e.g., "The coffee is good, but the service is slow.")

[1703] Output: Sentiment analysis result (e.g. "negative")

[1704] Step 5: Database accumulation

[1705] The server stores the results of the sentiment analysis, the summary, and the original text data in a database, which also stores the analysis results and timestamps.

[1706] Input: Original text data, summary, sentiment analysis results, timestamp

[1707] Output: Data accumulation in database

[1708] Step 6: Analyze the progress

[1709] The server analyzes trends based on the feedback data stored in the database, comparing it with past data, thereby analyzing fluctuations in sentiment and opinion over a specific period of time.

[1710] Input: Feedback data stored in the database

[1711] Output: Trend analysis report

[1712] Step 7: Individual respondent trend analysis

[1713] The server extracts data from each respondent and analyzes trends and patterns, which allows for the identification of specific respondent characteristics and behavioral patterns.

[1714] Input: Individual respondent data

[1715] Output: Trend analysis report

[1716] Step 8: Proposed next actions

[1717] The server then proposes specific next steps based on the results of trend analysis, which are then sent to the smart glasses or the user's device, where staff can check them in real time.

[1718] Input: Trend analysis results, Trend analysis results

[1719] Output: Notification of next action suggestion (e.g. "Propose staff training to improve service speed.")

[1720] The above steps enable fast and accurate collection, analysis, and action proposal of customer feedback.

[1721] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1722] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1723] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1724] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1725] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1726] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1727] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1728] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1729] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1730] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1731] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1732] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1733] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1734] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1735] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1736] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1737] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1738] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1739] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1740] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1741] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1742] The following is further disclosed regarding the above embodiment.

[1743] Understood. The draft claims are as follows:

[1744] (Claim 1)

[1745] means for automatically obtaining comment data from the survey system;

[1746] means for preprocessing the comment data;

[1747] means for summarizing the preprocessed comment data;

[1748] means for analyzing sentiment of the summarized comment data;

[1749] means for storing the analysis results in a database;

[1750] A means of analyzing the trends of past data based on accumulated data, and

[1751] A means of analyzing individual respondent trends;

[1752] means for suggesting a next action based on the analysis results;

[1753] A system including:

[1754] (Claim 2)

[1755] 2. The system according to claim 1, wherein the means for automatically acquiring comment data from the survey system acquires the comment data by sending an HTTP request to a pre-configured API endpoint.

[1756] (Claim 3)

[1757] 2. The system according to claim 1, wherein the preprocessing means includes a process for deleting unnecessary spaces, special characters, and emojis from comments and standardizing the text format.

[1758] "Example 1"

[1759] (Claim 1)

[1760] means for automatically obtaining comment data from the survey system;

[1761] means for preprocessing the comment data;

[1762] means for using a generative AI model to summarize the preprocessed comment data;

[1763] means for analyzing sentiment of the summarized comment data;

[1764] means for storing the analysis results in a database;

[1765] A means of analyzing the trends of past data based on accumulated data, and

[1766] A means of analyzing individual respondent trends;

[1767] means for suggesting a next action based on the analysis results;

[1768] A system including:

[1769] (Claim 2)

[1770] 2. The system according to claim 1, wherein the means for automatically acquiring comment data from the survey system acquires the comment data by sending an HTTP request to a pre-configured API endpoint.

[1771] (Claim 3)

[1772] 2. The system according to claim 1, wherein the preprocessing means includes a process for deleting unnecessary spaces, special characters, and emojis from comments and standardizing the text format.

[1773] "Application Example 1"

[1774] (Claim 1)

[1775] means for automatically obtaining comment data from the survey system;

[1776] means for preprocessing the comment data;

[1777] means for using a generative AI model to summarize the pre-processed comment data;

[1778] means for analyzing sentiment of the summarized comment data;

[1779] means for storing the analysis results in a database;

[1780] A means of analyzing the trends of past data based on accumulated data, and

[1781] A means of analyzing individual respondent trends;

[1782] means for suggesting a next action based on the analysis results;

[1783] a means for using the analysis results to visualize feedback trends and notify management;

[1784] A system including:

[1785] (Claim 2)

[1786] 2. The system according to claim 1, wherein the means for automatically acquiring comment data from the survey system acquires the comment data by sending an HTTP request to a pre-configured API endpoint.

[1787] (Claim 3)

[1788] 2. The system according to claim 1, wherein the preprocessing means includes a process for deleting unnecessary spaces, special characters, and emojis from comments and standardizing the text format.

[1789] "Example 2: Combining Emotion Engines"

[1790] (Claim 1)

[1791] means for automatically obtaining comment data from the survey system;

[1792] means for preprocessing the comment data;

[1793] means for summarizing the preprocessed comment data using a generative AI model;

[1794] means for analyzing sentiment of the summarized comment data;

[1795] means for storing the analysis results in a database;

[1796] A means of analyzing the trends of past data based on accumulated data, and

[1797] A means of analyzing individual respondent trends;

[1798] means for suggesting a next action based on the analysis results;

[1799] means for displaying the analysis results and next actions through a user interface;

[1800] A system including:

[1801] (Claim 2)

[1802] 2. The system according to claim 1, wherein the means for automatically acquiring comment data from the survey system acquires the comment data by sending an HTTP request to a pre-configured API endpoint.

[1803] (Claim 3)

[1804] 2. The system according to claim 1, wherein the preprocessing means includes a process for deleting unnecessary spaces, special symbols, and emojis from comments and standardizing the text format.

[1805] "Application example 2 when combining emotion engines"

[1806] (Claim 1)

[1807] means for automatically obtaining comment data from the survey system;

[1808] means for preprocessing the comment data;

[1809] means for summarizing the preprocessed comment data;

[1810] means for analyzing sentiment of the summarized comment data;

[1811] means for storing the analysis results in a database;

[1812] A means of analyzing the trends of past data based on accumulated data, and

[1813] A means of analyzing individual respondent trends;

[1814] a means for collecting customer feedback using smart glasses;

[1815] a means for analyzing the collected customer feedback in real time and proposing next actions;

[1816] A system including:

[1817] (Claim 2)

[1818] 2. The system according to claim 1, wherein the means for automatically acquiring comment data from the survey system acquires the comment data by sending an HTTP request to a pre-configured API endpoint.

[1819] (Claim 3)

[1820] 2. The system according to claim 1, wherein the preprocessing means includes a process for deleting unnecessary spaces, special characters, and emojis from comments and standardizing the text format. [Explanation of symbols]

[1821] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for automatically obtaining comment data from the survey system; means for preprocessing the comment data; means for summarizing the preprocessed comment data; means for analyzing sentiment of the summarized comment data; means for storing the analysis results in a database; A means of analyzing the trends of past data based on accumulated data, and A means of analyzing individual respondent trends; means for suggesting a next action based on the analysis results; A system including:

2. The system according to claim 1 , wherein the means for automatically acquiring comment data from the survey system acquires the comment data by sending an HTTP request to a pre-configured API endpoint.

3. 2. The system according to claim 1, wherein the preprocessing means includes a process for deleting unnecessary spaces, special characters, and emojis from comments and standardizing the text format.

Images