System

Abstract

To provide a system.SOLUTION: A system includes a feeling engine for recognizing user's feelings from inquiry contents to his / her own company from the user, planning means for extracting elements lowering customer satisfaction in own company services in accordance with the recognized feelings by inputting a specific prompt sentence to a generative AI model, planning improvement measures for the own company services on the basis of the elements lowering the customer satisfaction, and means for feeding back the planned improvement measures to the user.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 a description and related instruction sentence regarding 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] JP 2022-180282 A Summary of the Invention [Problem to be solved by the invention]

[0004] Companies want to improve their services based on customer inquiry histories and online information, but it is difficult to efficiently analyze this information, identify the factors that are lowering customer satisfaction, and devise specific improvement measures. [Means for solving the problem]

[0005] As a means for solving this problem, the system of the present invention includes an emotion engine that recognizes the emotions of a user from the content of the user's inquiry made to the company, a planning means that extracts factors within the company's services that are lowering customer satisfaction in accordance with the recognized emotion by inputting a specific prompt sentence into a generative AI model, and plans improvement measures for the company's services based on the factors that are lowering customer satisfaction, and a means for feeding back the planned improvement measures to the user. [Brief description of the drawings]

[0006] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Diagram 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. FIG. [Diagram 3] FIG. 11 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Diagram 5] FIG. 13 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. 13 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 13 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. 2 is a sequence diagram showing a process flow of the data processing system in the first embodiment of the first form example. [Figure 12]FIG. 11 is a sequence diagram showing a process flow of the data processing system in the application example 1 of the first embodiment. [Figure 13] FIG. 11 is a sequence diagram showing a process flow of the data processing system in the second embodiment of the second embodiment. [Figure 14] FIG. 13 is a sequence diagram showing a process flow of the data processing system in application example 2 of embodiment 2. [Figure 15] FIG. 13 is a sequence diagram showing a process flow of a data processing system in a third embodiment of the present invention. [Figure 16] FIG. 13 is a sequence diagram showing a process flow of the data processing system in the third application example of the third embodiment. [Figure 17] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in the first embodiment of the first form example when an emotion engine is combined. [Figure 18] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Form Example 1 when combined with an emotion engine. [Figure 19] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system according to the embodiment 2 of the second embodiment in the case where an emotion engine is combined. [Figure 20] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Form Example 2 when combined with an emotion engine. [Figure 21] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in the embodiment 3 of the form example 3 in the case where an emotion engine is combined. [Figure 22] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Form Example 3 when combined with an emotion engine. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0007] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0008] First, the terms used in the following description will be explained.

[0009] In the following embodiments, a signed processor (hereinafter simply referred to as a "processor") may be one arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be one 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), an APU (Accelerated Processing Unit), or a TPU (TENSOR PROCESSING UNIT (registered trademark)).

[0010] In the following embodiments, a signed RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by the processor.

[0011] 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.

[0012] In the following embodiments, a communication I / F (Interface) with a code is an interface including a communication processor and an antenna. The communication I / F controls communication between multiple computers. An example of a communication standard applied to the communication I / F is a wireless communication standard including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0013] 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. In addition, in this specification, the same idea as "A and / or B" is also applied when three or more things are expressed by connecting them with "and / or."

[0014] [First embodiment]

[0015] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0016] 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.

[0017] 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 wide area network (WAN) and / or a local area network (LAN).

[0018] 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.

[0019] The reception device 38 includes a touch panel 38A and a microphone 38B, and receives user input. The touch panel 38A detects contact with a pointer (e.g., a pen or a finger) to receive user input by the touch of the pointer. 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.

[0020] 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 (e.g., voice and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs voice according to instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, an aperture, and a shutter, and an imaging element such as a Complementary Metal-Oxide-Semiconductor (CMOS) image sensor or a Charge Coupled Device (CCD) image sensor.

[0021] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54.

[0022] FIG. 2 shows an example of main functions of the data processing device 12 and the smart device 14.

[0023] As shown in Fig. 2, 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. The specific process program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific process program 56 from the storage 32, and executes the read specific process 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 process program 56 executed on the RAM 30.

[0024] 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.

[0025] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores a reception output program 60. The reception output program 60 is used together with the specific processing program 56 by the data processing system 10. The processor 46 reads out the reception output program 60 from the storage 50, and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0026] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0027] "Example 1"

[0028] As one embodiment of the present invention, a web scraping tool or API is used to collect information such as inquiry history and online information. The collected information is saved as text data and becomes input data for analysis by AI.

[0029] "Example 2"

[0030] Next, we use a generative AI (Generative Pretrained Transformer), a type of natural language processing (NLP), as an AI method to analyze the collected information. Generative AI has the ability to learn from large amounts of text data and generate new text. Using this ability, we extract factors that are lowering customer satisfaction from the collected information.

[0031] "Example 3"

[0032] Furthermore, dashboards and graphs are used to visualize the analysis results. This allows companies to understand problems at a glance and devise improvement measures. Specifically, the content and frequency of customer inquiries, online reviews, etc. are visualized, and from this information, factors that are lowering customer satisfaction are identified.

[0033] "Example 4"

[0034] Finally, the improvement measures proposed by AI are considered as a means of planning improvement methods based on the visualized results. Specifically, based on the improvement measures proposed by AI, companies can determine the direction of service improvement and create specific action plans. This allows companies to improve customer satisfaction and reduce unnecessary calls to their call centers.

[0035] The process flow of each embodiment will be described below.

[0036] "Example 1"

[0037] Step 1: Use a web scraping tool or API to collect inquiry history and online information for your company.

[0038] Step 2: The collected information is saved as text data and used as input data for AI analysis.

[0039] "Example 2"

[0040] Step 1: Generative AI, a type of natural language processing (NLP)

[0041] The collected information is analyzed using a trained Transformer.

[0042] Step 2: Using the results of the generative AI's learning, generate new text and extract the factors that are reducing customer satisfaction.

[0043] "Example 3"

[0044] Step 1: Use dashboards and graphs to visualize the analysis results.

[0045] Step 2: From the visualized results, identify problems and develop improvement measures.

[0046] "Example 4"

[0047] Step 1: Based on the improvement measures suggested by AI, the company decides on the direction of service improvement.

[0048] Step 2: Create a concrete action plan to improve customer satisfaction and reduce unnecessary calls to the call center.

[0049] Example 1

[0050] Next, a description will be given of Example 1 of the embodiment 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".

[0051] With the conventional system, it was difficult to efficiently collect and analyze the history of inquiries to the company and information on the Internet. It was also difficult to properly analyze the collected information and devise specific improvement methods to increase customer satisfaction. Furthermore, there was a lack of effective means to reduce unnecessary calls to the call center.

[0052] 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.

[0053] In this invention, the server includes a means for collecting inquiry history and online information to the company using a web scraping tool or API, a means for saving the collected information as text data, a means for preprocessing the saved text data using a natural language processing library, a means for inputting the preprocessed data into a generative AI model and analyzing it, a means for visualizing factors that lower customer satisfaction based on the analysis results, and a means for planning improvement methods based on the visualization results. This enables efficient collection and analysis of information, and allows specific improvement methods to be planned to improve customer satisfaction. It also reduces unnecessary calls to the call center, making it possible to contribute to companies in terms of both quality and cost.

[0054] A "web scraping tool" is software for automatically extracting data from websites.

[0055] "API" stands for Application Programming Interface, an interface for exchanging data and functions between different software programs.

[0056] "Inquiry history" is a record of inquiries made by customers to your company, and is data including the inquiry content, date and time, customer information, etc.

[0057] "Online information" refers to publicly available data or content that exists on the Internet, including social media posts, blog posts, news articles, etc.

[0058] "Text data" is data expressed as a character string, and is a collection of sentences or words.

[0059] A "natural language processing library" is a software library for analyzing text data and understanding its meaning, providing functions such as tokenization, sentiment analysis, and tagging.

[0060] "Preprocessing" refers to the process of converting data into a format that can be analyzed, and includes data cleaning and tokenization.

[0061] A "generative AI model" is an artificial intelligence model that generates new information or analytical results based on input data, and is used for natural language generation and classification tasks.

[0062] "Analysis" is the process of examining data in detail to find meaning and patterns.

[0063] "Visualization" refers to displaying data or analytical results in a visual format such as a graph or chart, and is a means of making information easier to understand intuitively.

[0064] "Improvement methods" are specific measures or action plans for solving current problems and improving the performance and quality of a system or service.

[0065] This invention is a system in which a server uses a web scraping tool or API to collect inquiry history and online information for the company, and saves the collected information as text data. The saved text data is preprocessed using a natural language processing library and input into a generative AI model. The generative AI model analyzes the input data, visualizes factors that are reducing customer satisfaction based on the analysis results, and proposes methods of improvement.

[0066] Hardware and Software Used

[0067] Hardware

[0068] Server: A computer system for data collection, storage, pre-processing, and analysis.

[0069] software

[0070] Web scraping tools: Python's BeautifulSoup and Scrapy, etc.

[0071] API: Twitter API, Google(R) Custom Search API, etc.

[0072] Database: MySQL (registered trademark), PostgreSQL, etc.

[0073] Cloud storage: Amazon S3, Google Cloud Storage, etc.

[0074] Natural language processing libraries: such as NLTK or spaCy.

[0075] Generative AI models: OpenAI's (registered trademark) GPT-3 (registered trademark) and Google's BERT, etc.

[0076] Examples

[0077] Example 1: Collection and analysis of inquiry history

[0078] 1. The server uses BeautifulSoup to collect past inquiry history from its company's inquiry form.

[0079] 2. The server stores the collected query history in a MySQL database.

[0080] 3. The server pre-processes and tokenizes the stored query history using spaCy.

[0081] 4. The server inputs the preprocessed data into GPT-3 and analyzes trends in the query content.

[0082] Example 2: Collecting and analyzing information on the Internet

[0083] 1. The server uses the Twitter API to collect tweets related to specific keywords.

[0084] 2. The server stores the collected tweets in Amazon S3.

[0085] 3. The server preprocesses the stored tweets using NLTK and performs sentiment analysis.

[0086] 4. The server inputs the preprocessed data into BERT to classify the content of the tweets.

[0087] Examples of prompt statements

[0088] Prompt 1: Analyze the query history

[0089] "Based on your inquiry history over the past year, please tell us the most common inquiries you have made."

[0090] Prompt 2: Collecting and analyzing information on the Internet

[0091] "Collect tweets related to a specific keyword and tell me the percentage of positive and negative tweets."

[0092] In this way, the server can efficiently collect information, analyze it, and devise specific improvement methods to increase customer satisfaction. It can also reduce unnecessary calls to the call center, contributing to companies in terms of both quality and cost.

[0093] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0094] Step 1:

[0095] The server collects information using web scraping tools and APIs. Specifically, it uses Python's BeautifulSoup to scrape inquiry history from websites and Twitter API to collect tweets related to specific keywords. The input is the website URL or API endpoint, and the output is the collected raw data.

[0096] Step 2:

[0097] The server stores the collected information as text data. Specifically, it stores inquiry history and tweets in a MySQL database. The input is the collected raw data, and the output is the text data stored in the database.

[0098] Step 3:

[0099] The server preprocesses the stored text data using a natural language processing library. Specifically, it uses spaCy to clean the text (e.g., remove HTML tags and special characters) and tokenize it. The input is the text data stored in the database, and the output is the preprocessed text data.

[0100] Step 4:

[0101] The server inputs the preprocessed data into the generative AI model for analysis. Specifically, it uses OpenAI's GPT-3 to analyze trends in inquiry content and Google's BERT to classify tweet content. The input is the preprocessed text data, and the output is the analysis result.

[0102] Step 5:

[0103] The server visualizes the factors that are lowering customer satisfaction based on the analysis results. Specifically, it uses a visualization tool to display the analysis results as graphs and charts. The input is the analysis results, and the output is the visualized data.

[0104] Step 6:

[0105] The server then creates improvement methods based on the visualized results. Specifically, it generates a specific action plan to improve customer satisfaction based on the visualized data. The input is the visualized data, and the output is a proposal for improvement methods.

[0106] (Application example 1)

[0107] Next, a description will be given of Application Example 1 of Form 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."

[0108] Conventional customer support systems have had problems with not being able to respond quickly and appropriately to user inquiries, leading to lower customer satisfaction and an increase in unnecessary calls to call centers. In addition, they were unable to effectively utilize online information, making it difficult to link it to service improvements. This resulted in issues such as a decline in the company's quality and efficiency in terms of costs.

[0109] 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.

[0110] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, a means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for integrating the collected information and generating answers to customer inquiries in real time using generative AI, and a means for presenting the generated answers to customer support staff. This enables prompt and appropriate responses to inquiries from users, improving customer satisfaction and reducing unnecessary calls to the call center. Furthermore, by effectively utilizing information on the Internet, it is possible to improve services, improving the quality and cost efficiency of the company.

[0111] "Inquiry history to the company" refers to a record of questions, requests, complaints, etc. that a company has received from customers.

[0112] "Online information" refers to text data such as reviews, social media posts, and blog articles that exist on the Internet.

[0113] "Means of collection" refers to the method of obtaining information using web scraping tools or APIs.

[0114] "Means of analysis" refers to the method of inputting collected data into an AI model and extracting patterns and trends.

[0115] "Planning improvements to our own services" is the process of devising ways to improve services based on the results of analysis.

[0116] A "visualization means" is a method for displaying the analysis results in the form of graphs, charts, etc.

[0117] "Means for planning improvement methods" is a method for formulating specific improvement measures based on visualized data.

[0118] "Generative AI" is an artificial intelligence technology that generates text data using natural language processing.

[0119] "A means of generating answers to customer inquiries in real time" refers to a method of using generative AI to instantly create appropriate answers.

[0120] "Means for presentation to customer support representative" refers to the manner in which the generated answers are displayed to support staff.

[0121] The following system configuration and processing procedure will be described as an embodiment of the present invention.

[0122] System Configuration

[0123] The system consists of the following main components:

[0124] 1. Server: Collects data, analyzes it, and generates answers using generative AI.

[0125] 2. Terminal: The device used by the customer support representative (e.g. PC, smartphone).

[0126] 3. User: The customer making the enquiry.

[0127] Hardware and Software Used

[0128] Hardware: Servers, PCs, smartphones

[0129] Software: Python, BeautifulSoup, requests, OpenAI API

[0130] Data processing and calculation

[0131] 1. Data collection:

[0132] The server uses a web scraping tool (BeautifulSoup) and an API (requests) to collect inquiry history and online information for the company.

[0133] The collected data is saved in text format.

[0134] 2. Data Analysis:

[0135] The server inputs the collected text data into a generative AI (OpenAI API) for analysis.

[0136] The analysis results identify and visualize the factors that are lowering customer satisfaction.

[0137] 3. Answer generation:

[0138] The server uses generative AI to generate answers to customer inquiries in real time.

[0139] The generated answer is presented to the customer support representative's terminal.

[0140] Examples

[0141] For example, if a user asks "Do you have product A in stock?", the server will input the following prompt to the generative AI:

[0142] Analyze user inquiries and generate the best answers. Here are some of the inquiries we collected:

[0143] 1. Do you have product A in stock?

[0144] 2. Please tell me how to return product B.

[0145] 3. I would like to see reviews of product C.

[0146] The generative AI generates an appropriate response based on this prompt text and displays it on the customer support representative's device, enabling the representative to respond quickly and appropriately.

[0147] This system enables prompt and appropriate responses to user inquiries, improving customer satisfaction and reducing unnecessary calls to the call center. In addition, by effectively utilizing information on the Internet, it is possible to improve services, improving the quality and cost efficiency of companies.

[0148] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0149] Step 1:

[0150] The server uses a web scraping tool (BeautifulSoup) to collect inquiry history and related information from a specified website. The website URL is given as input, and the collected text data is obtained as output. Specifically, the server sends an HTTP request and analyzes the obtained HTML data to extract text.

[0151] Step 2:

[0152] The server uses API requests to collect inquiry history and related information from the specified API endpoint. The API URL is given as input, and the collected data in JSON format is obtained as output. In concrete terms, the server sends an API request, analyzes the obtained JSON data, and extracts the required information.

[0153] Step 3:

[0154] The server integrates the text data and JSON data collected in steps 1 and 2. Text data and JSON data are given as input, and the integrated text data is obtained as output. In concrete terms, the server combines the text data and JSON data and saves them as a single text data.

[0155] Step 4:

[0156] The server inputs the integrated text data into the generative AI (OpenAI API) and performs analysis. The integrated text data is given as input, and the analysis results are obtained as output. In concrete terms, the server sends a prompt to the generative AI and receives a response from the AI.

[0157] Step 5:

[0158] The server generates answers to customer inquiries in real time based on the analysis results obtained from the generative AI. The analysis results are given as input, and the generated answers are obtained as output. In concrete terms, the server analyzes the analysis results and generates appropriate answers.

[0159] Step 6:

[0160] The server presents the generated answer to the customer support representative's terminal. The generated answer is given as an input, and the answer displayed on the representative's terminal is obtained as an output. In concrete terms, the server transmits the generated answer to the representative's terminal and displays it.

[0161] Step 7:

[0162] The customer support representative responds to the user's inquiry based on the presented answer. The input is the answer presented by the server, and the output is a response to the user. In concrete terms, the representative checks the presented answer and gives an appropriate response to the user.

[0163] Example 2

[0164] Next, a second embodiment of the second embodiment will be described. 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."

[0165] With conventional customer satisfaction improvement systems, it was difficult to effectively analyze customer feedback and reviews and extract specific areas for improvement. In addition, the process of visualizing the analysis results and proposing specific improvement methods was time-consuming and laborious, making it difficult to respond quickly. Furthermore, it was not possible to reduce unnecessary calls to the call center, and there was a need to improve efficiency in both quality and cost.

[0166] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0167] In this invention, the server

[0168] A means of collecting inquiry history and online information for our company,

[0169] A means for pre-processing the collected information;

[0170] A means for analyzing the preprocessed information using a generative AI model; and

[0171] A means for extracting factors that are lowering customer satisfaction based on the analysis results;

[0172] A means for visualizing the extracted factors;

[0173] A means to plan improvement methods based on the visualized results;

[0174] This makes it possible to effectively analyze customer feedback and reviews and quickly extract specific areas for improvement. In addition, visualizing the analysis results makes it easier to plan specific improvement methods, which reduces unnecessary calls to the call center, resulting in improved efficiency in both quality and cost.

[0175] "Inquiry history to our company" refers to a record of inquiries made by customers to our company.

[0176] "Online information" refers to text data such as customer feedback, reviews, and comments that exist on the Internet.

[0177] "Means of collection" refers to methods and technologies for automatically obtaining inquiry history and online information.

[0178] "Preprocessing means" refers to methods or techniques for converting collected text data into a format that is easy to analyze.

[0179] A "generative AI model" is an artificial intelligence model that has the ability to learn from large amounts of text data and generate new text.

[0180] "Means of analysis" refers to the methods and techniques used to analyze the pre-processed data using the generative AI model and extract meaningful information.

[0181] "Factors that reduce customer satisfaction" are elements that cause customer dissatisfaction and negative feedback.

[0182] "Means of extraction" refers to methods or techniques for extracting specific factors from the analysis results.

[0183] A "visualization means" is a method or technique for displaying the extracted factors in a visual format, such as a graph or chart.

[0184] "Means for planning improvement methods" are methods and techniques for devising specific improvement measures based on the visualized results.

[0185] The present invention is a system for improving customer satisfaction, in which a server, a terminal, and a user work together. A specific embodiment of this system will be described below.

[0186] 1. System Configuration

[0187] The server has the following main functions:

[0188] A means of collecting inquiry history and online information for your company

[0189] A means of pre-processing the collected information

[0190] A means of analyzing pre-processed information using generative AI models

[0191] A means of extracting factors that are reducing customer satisfaction based on analysis results

[0192] A means to visualize the extracted factors

[0193] A means to plan improvement methods based on visualized results

[0194] 2. Hardware and Software Used

[0195] The server is a computer system equipped with a high-performance processor and a large amount of memory. The following software is used:

[0196] Web scraping tools and APIs for data collection

[0197] Natural language processing libraries for data preprocessing (e.g. NLTK, spaCy)

[0198] Generative AI models (e.g. GPT-3)

[0199] Tools for data visualization (e.g. Matplotlib, D3.js)

[0200] 3. Program Processing

[0201] The server first collects contact history and online information, including website forms, emails, and social media comments, and then preprocesses the collected text data. Specifically, it cleans, tokenizes, and normalizes the text.

[0202] The preprocessed data is input into a generative AI model (GPT-3). This model has the ability to learn from large amounts of text data and generate new text. The generative AI model extracts factors that are lowering customer satisfaction from the input data. Specifically, it identifies negative feedback and dissatisfaction and lists them as factors.

[0203] 4. Specific Examples

[0204] As a concrete example, the following prompt sentence is input into the generative AI model.

[0205] Example prompt:

[0206] "Please extract the factors that are causing low customer satisfaction from the following text data. Text data: 'Poor product quality. Slow delivery. Unfriendly customer service.'"

[0207] The output of the generative AI model is, "The factors that are lowering customer satisfaction are product quality, delivery delays, and unfriendly customer service."

[0208] 5. Visualize and plan ways to improve

[0209] The server visualizes the extracted factors in visual formats such as graphs and charts. Based on the visualized results, specific improvement measures are proposed. For example, improving product quality, improving the delivery system, or customer service training could be considered.

[0210] In this way, the server can extract factors that are lowering customer satisfaction from the information collected using the generative AI model and quickly devise specific improvement measures.

[0211] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0212] Step 1:

[0213] The server collects the company's inquiry history and other online information. As input, it obtains text data such as website forms, emails, and SNS comments. Specifically, the server automatically collects data using web scraping tools and APIs. The output is the collected raw text data.

[0214] Step 2:

[0215] The server preprocesses the collected text data. As input, it uses the raw text data collected in step 1. Specifically, the server uses natural language processing libraries (e.g., NLTK, spaCy) to clean the text (remove unnecessary characters and symbols), tokenize (split into words and phrases), and normalize (convert into a unified format). The output is the preprocessed, clean text data.

[0216] Step 3:

[0217] The server inputs the preprocessed data into a generative AI model (e.g. GPT-3). The clean text data preprocessed in step 2 is used as input. Specifically, the server divides the preprocessed text data into batches and inputs them into the generative AI model. The generative AI model has the ability to learn from large amounts of text data and generate new text. The output is the result of analysis by the generative AI model.

[0218] Step 4:

[0219] The server extracts factors that are reducing customer satisfaction from the output of the generative AI model. As input, it uses the analysis results obtained from the generative AI model in step 3. Specifically, the server analyzes the output of the generative AI model and extracts negative keywords and phrases. The output is a list of factors that are reducing customer satisfaction.

[0220] Step 5:

[0221] The server visualizes the extracted factors. As input, it uses the list of factors extracted in step 4. Specifically, the server uses a data visualization tool (e.g., Matplotlib, D3.js) to display the factors in a visual form such as a graph or chart. The output is a graph or chart of the visualized factors.

[0222] Step 6:

[0223] The server plans improvement methods based on the visualization results. As input, it uses graphs and charts of the factors visualized in step 5. In terms of specific operations, the server devises specific improvement measures based on the visualized data. For example, improving product quality, improving the delivery system, customer service training, etc. are possible. The output is a list of specific improvement methods.

[0224] (Application example 2)

[0225] Next, a description will be given of Application Example 2 of Form 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."

[0226] With conventional customer satisfaction analysis systems, it was difficult to efficiently collect and analyze customer reviews and feedback and identify specific areas for improvement. In addition, there was a lack of a way to visualize the analysis results and quickly notify managers, which meant it took a long time to improve customer satisfaction. Furthermore, there was a lack of specific measures to reduce unnecessary calls to the call center.

[0227] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0228] In this invention, the server includes a means for collecting inquiry history and online information, an AI means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that lower customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for collecting customer reviews and feedback from mail order sites, a means for extracting factors that lower customer satisfaction using generative AI, and a means for notifying the manager of the extracted factors. This makes it possible to quickly identify specific factors that lower customer satisfaction and notify the manager. In addition, it is possible to take specific measures to reduce unnecessary calls to the call center.

[0229] An "inquiry history" is a record of inquiries made by a customer to a company.

[0230] "Online information" refers to text data such as reviews and feedback that exist on the Internet.

[0231] "Means of collection" refers to the methods and technologies used to collect inquiry history and online information.

[0232] "Means of analysis" are the methods and techniques used to analyze collected information and convert it into meaningful data.

[0233] "AI means for proposing improvements" refers to artificial intelligence technology that proposes improvement measures for services based on the results of analysis.

[0234] A "visualization means" is a method or technique for displaying the results of an analysis in a visual format, such as a graph or chart.

[0235] "Means for planning improvement methods" refers to methods and techniques for devising specific improvement measures based on the visualized results.

[0236] An "online shopping site" is a website that sells products and services over the Internet.

[0237] A "customer review" is an evaluation or comment made by a customer about a product or service.

[0238] "Feedback" refers to opinions and impressions provided by customers.

[0239] "Generative AI" is an artificial intelligence technology that has the ability to learn from large amounts of text data and generate new text.

[0240] "Means of extracting factors" refers to methods and techniques that use generative AI to identify the causes of declining customer satisfaction.

[0241] "Means of notifying the administrator" refers to methods or techniques for notifying the administrator of the extracted causes.

[0242] A system for carrying out this invention has the following configuration: The server includes means for collecting the history of inquiries made to the company and information on the internet, AI means for analyzing the collected information and proposing improvements to the company's services, means for visualizing factors that are lowering customer satisfaction based on the analysis results, means for proposing improvement methods based on the visualization results, means for collecting customer reviews and feedback from mail order sites, means for extracting factors that are lowering customer satisfaction using generative AI, and means for notifying an administrator of the extracted factors.

[0243] Hardware and Software Used

[0244] Hardware: Servers, smartphones

[0245] Software: Python, TENSORFLOW (registered trademark), Transformers library (Hugging Face), Flask (backend), React Native (frontend)

[0246] Data processing and calculation

[0247] Data collection

[0248] The server collects customer reviews and feedback from online shopping sites through the API, allowing for efficient collection of customer opinions and feedback.

[0249] Data Preprocessing

[0250] The collected text data is cleaned and tokenized, which converts the data into a form suitable for analysis.

[0251] Application of generative AI models

[0252] The server uses generative AI (such as GPT-3) to analyze the text data and extract factors that are reducing customer satisfaction. Generative AI has the ability to learn from large amounts of text data and generate new text.

[0253] Viewing the results

[0254] The extracted factors are visualized on the dashboard of a smartphone app, allowing managers to quickly grasp the specific points of dissatisfaction.

[0255] Administrator Notification

[0256] The server notifies the administrator of the extracted cause, allowing the administrator to take prompt action.

[0257] Examples

[0258] For example, by inputting a prompt sentence like the one below into a generative AI model, the AI ​​can analyze the review and extract specific points of dissatisfaction.

[0259] Example prompt:

[0260] "Analyze the following customer reviews to identify factors that are causing low customer satisfaction. Review: 'The product was delivered late and the quality was below expectations.'"

[0261] By inputting this prompt into a generative AI model, the AI ​​can analyze the reviews and extract specific complaints such as "The product was delivered late" or "The quality was below expectations." This allows managers to take swift action to address the complaints, helping to improve customer satisfaction.

[0262] The flow of the specific process in the application example 2 will be described with reference to FIG.

[0263] Step 1:

[0264] The server collects customer reviews and feedback from the online shopping site through an API. The input is the API endpoint of the online shopping site, and the output is the text data of the collected customer reviews and feedback. Specifically, the server periodically calls the API to obtain new reviews and feedback.

[0265] Step 2:

[0266] The server preprocesses the collected text data. The input is the text data collected in step 1, and the output is the cleaned and tokenized data. Specifically, the server removes unnecessary characters and spaces and splits the text into tokens.

[0267] Step 3:

[0268] The server uses a generative AI model (such as GPT-3) to analyze the preprocessed data. The input is the data preprocessed in step 2, and the output is the analysis result including the factors that are reducing customer satisfaction. Specifically, the server inputs a prompt sentence into the generative AI model and obtains the analysis result.

[0269] Step 4:

[0270] The server visualizes the analysis results. The input is the analysis results obtained in step 3, and the output is the data displayed in a visual format such as a graph or chart. Specifically, the server converts the analysis results into graphs and charts for display on the dashboard.

[0271] Step 5:

[0272] The server then proposes an improvement method based on the visualized results. The input is the data visualized in step 4, and the output is a specific improvement measure. Specifically, the server analyzes the visualized data and proposes an improvement measure.

[0273] Step 6:

[0274] The server notifies the administrator of the extracted causes and remedial measures. The input is the remedial measures proposed in step 5, and the output is the notification sent to the administrator. Specifically, the server sends information to the administrator via email or push notification.

[0275] Step 7:

[0276] The administrator receives the notification and takes specific corrective action. The input is the notification received in step 6, and the output is the corrective action taken. Specifically, the administrator makes the necessary changes or actions based on the content of the notification.

[0277] Example 3

[0278] Next, a third embodiment of the third embodiment will be described. 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."

[0279] With conventional customer satisfaction improvement systems, it was difficult to effectively collect and analyze customer inquiries and online evaluation data and quickly propose specific improvement measures. In addition, the process of creating specific action plans based on visualized data was complicated, making it difficult for companies to respond quickly. This led to issues such as insufficient improvement of customer satisfaction and reduction of unnecessary calls to call centers.

[0280] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0281] In this invention, the server includes a means for collecting inquiry history and online information, an artificial intelligence means for analyzing the collected information and proposing improvements to the company's services, a means for visualizing factors lowering customer satisfaction based on the analysis results, a means for proposing improvement methods based on the visualization results, a means for inputting a prompt sentence into the generative AI model to have it propose improvement measures, and a means for creating a specific action plan based on the proposed improvement measures. This makes it possible to effectively collect and analyze customer inquiry contents and online evaluation data, quickly propose specific improvement measures, and create a specific action plan based on the visualized data.

[0282] "Inquiry history to our company" refers to a record of the content, date and time, frequency, etc. of inquiries made by customers to a company.

[0283] "Online information" refers to data such as customer ratings, reviews, and comments that exist on the Internet.

[0284] "Means of collection" refers to the methods and technologies used to obtain the necessary information using databases or APIs.

[0285] "Analytical means" are the methods and techniques used to analyze collected data and extract meaningful information.

[0286] An "artificial intelligence means" is a system that uses techniques such as machine learning and natural language processing to analyze data and gain insights.

[0287] "Visualization means" refers to methods and technologies for displaying analysis results in the form of graphs, dashboards, etc., making them intuitively understandable.

[0288] A "generative AI model" is an artificial intelligence model that is trained to generate appropriate outputs for specific inputs.

[0289] A "prompt" is an instruction entered into a generative AI model to cause it to perform a specific task.

[0290] "Means for proposing improvement measures" refers to methods or technologies that use generative AI models to propose specific improvement measures based on the analysis results.

[0291] "Means for creating an action plan" refers to methods or techniques for formulating a specific action plan based on proposed improvement measures.

[0292] The present invention is directed to a system that aims to improve customer satisfaction and reduce unnecessary calls to a call center. A specific embodiment of this system will be described below.

[0293] First, the server collects the history of inquiries to the company and information on the internet. To do this, a database service or API is used. For example, a general cloud database is used as the database service, and data is obtained from customer inquiry systems and SNS platforms via API.

[0294] Next, the server analyzes the collected data. This analysis is performed using programming languages ​​such as Python and R. Specifically, natural language processing (NLP) libraries (e.g., NLTK and spaCy) are used to analyze the text data and perform frequent keyword and sentiment analysis. Machine learning models are also used to perform data clustering and anomaly detection.

[0295] The analysis results are displayed as dashboards and graphs using visualization tools such as Tableau and Microsoft (registered trademark) Power BI. This allows users to view the dashboard on a browser and intuitively understand the analysis results.

[0296] Based on the visualization results, the user inputs a prompt sentence into the generative AI model to suggest improvement measures. For example, the following prompt sentence can be input into the generative AI model:

[0297] "Please suggest improvement measures to increase customer satisfaction based on the content and frequency of inquiries from customers and online evaluation data."

[0298] The server sends this prompt to the generative AI model and receives improvement measures proposed by the AI. The user creates a specific action plan based on the proposed improvement measures. For example, if the AI ​​suggests an improvement measure such as "There are many complaints about a specific product, so we suggest that you conduct additional tests to improve the quality of the product," the user creates and executes a specific test plan based on the proposal.

[0299] In this way, companies can improve customer satisfaction and reduce unnecessary calls to call centers. This system can consistently perform data collection, analysis, visualization, improvement proposals, and action plan creation, enabling quick and effective response. The flow of the identification process in the third embodiment will be described with reference to FIG. 15.

[0300] Step 1: Data collection

[0301] The server collects the company's inquiry history and online information. As input, it obtains data from customer inquiry systems and social media platforms. Specifically, the server obtains data from these platforms via APIs and stores it in a database. As output, the content of inquiries, their frequency, and online evaluation data are stored in the database.

[0302] Step 2: Data analysis

[0303] The server analyzes the collected data. As input, it uses inquiry content stored in a database and online evaluation data. Specifically, the server uses a natural language processing (NLP) library (e.g., NLTK or spaCy) in Python or R to analyze the text data. This provides the results of frequent keywords and sentiment analysis. It also uses machine learning models to perform data clustering and anomaly detection. The analysis results are obtained as output.

[0304] Step 3: Data visualization

[0305] The server visualizes the analysis results. It uses the results of the data analysis as input. Specifically, the server exports the analysis results in JSON format and imports them into a visualization tool (e.g. Tableau or Microsoft Power BI), which displays the analysis results as dashboards and graphs. As output, it provides the visualized data.

[0306] Step 4: Propose improvements

[0307] Based on the visualization results, the user inputs a prompt sentence into the generative AI model to have it suggest improvements. The visualized data and the prompt sentence are used as input. In concrete terms, the user inputs the following prompt sentence into the generative AI model: "Please suggest improvements to improve customer satisfaction based on the content and frequency of customer inquiries, as well as online evaluation data." The server sends this prompt sentence to the generative AI model and receives the improvements suggested by the AI. The output is the improvements suggested by the AI.

[0308] Step 5: Create an action plan

[0309] The user creates a specific action plan based on the proposed improvement measures. The improvement measures proposed by the AI ​​are used as input. In concrete terms, the user consults with relevant departments within the company based on the proposed improvement measures (e.g., "There are many complaints about a particular product, so we propose that additional testing be carried out to improve the quality of the product") and formulates a specific action plan. A specific action plan is obtained as output.

[0310] In this way, companies can improve customer satisfaction and reduce unnecessary calls to their call centers.

[0311] (Application example 3)

[0312] Next, a description will be given of Application Example 3 of Form Example 3. 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."

[0313] With conventional customer support systems, it was difficult to effectively analyze and visualize customer inquiries and online reviews. In addition, there was a lack of means to plan specific improvement measures based on the analysis results, and it was not possible to fully improve customer satisfaction or reduce unnecessary calls to the call center. This caused companies to face challenges in both quality and cost.

[0314] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0315] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, an AI means for analyzing the collected information and proposing improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for proposing improvement methods based on the visualization results, a means for proposing improvement measures based on specific inquiry contents using generative AI, and a means for displaying the proposed improvement measures. This makes it possible to effectively analyze and visualize the contents of inquiries from customers and online evaluations. In addition, specific improvement measures can be proposed based on the analysis results, which can improve customer satisfaction and reduce unnecessary calls to the call center.

[0316] "Inquiry history" refers to a record of customer inquiries, including their content, frequency, and date and time.

[0317] "Online information" refers to data such as customer ratings, reviews, and comments that exist on the Internet.

[0318] "Collection means" refers to methods or devices for acquiring inquiry history and online information.

[0319] An "analysis means" is a method or device for analyzing collected information and converting it into meaningful data.

[0320] A "visualization means" is a method or device that displays analysis results in the form of graphs, dashboards, etc., to enable intuitive understanding.

[0321] An "improvement planning means" is a method or device for devising improvement measures for a service based on the visualization results.

[0322] "Generative AI" is a type of artificial intelligence, a model that has the ability to generate appropriate outputs for specific inputs.

[0323] A "proposal means" is a method or device for using generative AI to propose improvement measures based on specific inquiry content.

[0324] A "display means" is a method or device for showing the suggested improvement to the user.

[0325] A system for implementing this invention includes a server, a user terminal, and a generative AI model. The server collects inquiry history and online information and analyzes this data. The analysis results are visualized and displayed in a format that the user can intuitively understand. Furthermore, the generative AI model is used to propose improvement measures based on the specific inquiry content and display them on the user terminal.

[0326] The server uses Python and Dash to build a dashboard to visualize the types of inquiries and their frequency. When a user selects a specific inquiry from a drop-down menu, the data related to that inquiry is displayed as a bar graph, allowing the user to see at a glance which inquiries are occurring frequently.

[0327] In addition, the server uses the OpenAI API to generate improvement measures suggested by the generative AI model based on the selected inquiry content. For example, if the user selects "Product defect," the server inputs the following prompt sentence to the generative AI model:

[0328] Example prompt:

[0329] If the customer inquiry is about a "defective product," please provide suggestions for improving the service.

[0330] Based on this prompt, the generative AI model generates "suggestions for reducing product defects," proposing specific improvement measures such as "strengthening quality control" and "simplifying return procedures." These suggestions are displayed on the user's device, allowing the user to use them as a reference when creating a specific action plan.

[0331] This system makes it possible to effectively analyze and visualize customer inquiries and online reviews, and also to devise specific improvement measures based on the analysis results, thereby improving customer satisfaction and reducing unnecessary calls to the call center.

[0332] The flow of the specific process in the application example 3 will be described with reference to FIG.

[0333] Step 1:

[0334] The server collects inquiry history and online information.

[0335] Input: Customer inquiries, online ratings and reviews

[0336] Data processing: Save inquiry details and evaluations in a database

[0337] Output: The collected dataset

[0338] Step 2:

[0339] The server analyzes the collected data.

[0340] Input: Collected dataset

[0341] Data calculations: data cleansing, classification, frequency analysis

[0342] Output: Analysis results (frequency of inquiries, evaluation trends, etc.)

[0343] Step 3:

[0344] The server visualizes the analysis results.

[0345] Input: Analysis results

[0346] Data processing: generating graphs and dashboards

[0347] Output: Visualized data (bar graphs, dashboards)

[0348] Step 4:

[0349] The user reviews the visualized data and selects a particular query.

[0350] Input: visualized data, user selection

[0351] Data processing: Extraction of selected inquiry details

[0352] Output: Selected inquiry details

[0353] Step 5:

[0354] The server uses a generative AI model to suggest improvement measures based on the selected inquiry content.

[0355] Input: Selected inquiry content, prompt text

[0356] Data computation: Proposal generation using generative AI models

[0357] Output: Suggested remedy

[0358] Step 6:

[0359] The server displays the proposed remedy on the user terminal.

[0360] Input: Proposed remedy

[0361] Data processing: Formatting the proposal content

[0362] Output: Remediation measures displayed on the user's terminal

[0363] Step 7:

[0364] The user reviews the proposed improvements and creates a specific action plan.

[0365] Input: Proposed remedy

[0366] Data processing: Creating an action plan

[0367] Output: A concrete action plan

[0368] Furthermore, an emotion engine that estimates the emotion of the user may be combined. That is, the identification processing unit 290 may estimate the emotion of the user using the emotion identification model 59, and perform identification processing using the emotion of the user.

[0369] "Example 1"

[0370] One embodiment of the present invention is a system that incorporates an emotion engine. This system collects inquiry history and online information for the company, and analyzes that information. The analysis is performed by both the AI ​​means and the emotion engine. The AI ​​means understands the content of the information and plans improvements to the service. Meanwhile, the emotion engine analyzes the emotions of the user. For example, the emotional state of the user is estimated from the wording and tone used when the user makes an inquiry. This makes it possible to grasp not only the user's satisfaction level, but also the emotions behind it.

[0371] "Example 2"

[0372] The emotion analysis results obtained from the emotion engine are provided to the AI ​​means. The AI ​​means uses these results to plan improvements to the service. For example, if anger is detected among many users,

[0373] In this case, the system identifies the cause and proposes a remedy. This makes it possible to improve services based on the user's feelings, not just the content of the inquiry.

[0374] "Example 3"

[0375] The system of the present invention also includes a means for visualizing the analysis results. This makes it possible to grasp at a glance the factors that are lowering customer satisfaction and the emotional state of users. For example, the dashboard displays the most frequently inquired items and the distribution of emotional states. This allows companies to quickly grasp problems and devise appropriate improvement measures.

[0376] "Example 4"

[0377] The system of the present invention also includes a means for proposing improvement methods based on the visualization results. This makes it possible to create specific action plans and improve customer satisfaction. For example, based on information obtained from the emotion engine, improvement measures such as providing new services that resolve user dissatisfaction can be proposed. This not only improves user satisfaction, but also contributes to the company's quality and cost aspects, such as reducing unnecessary calls to call centers.

[0378] The process flow of each embodiment will be described below.

[0379] "Example 1"

[0380] Step 1: The system collects inquiry history and online information for the company.

[0381] Step 2: The collected information is analyzed by both AI means and the emotion engine.

[0382] Step 3: The emotion engine analyzes the user's emotions and provides the results to the AI ​​means.

[0383] "Example 2"

[0384] Step 1: The AI ​​means plans service improvements based on the emotion analysis results provided by the emotion engine.

[0385] Step 2: If anger is detected among many users, identify the cause and develop measures to improve it.

[0386] "Example 3"

[0387] Step 1: The system visualizes the analysis results.

[0388] Step 2: Display the most frequently inquired items and distribution of emotional states on the dashboard.

[0389] "Example 4"

[0390] Step 1: The system plans improvement measures based on the visualized results.

[0391] Step 2: Create a concrete action plan to improve customer satisfaction.

[0392] Step 3: Based on the information obtained from the emotion engine, develop improvement measures, such as providing new services to resolve user dissatisfaction.

[0393] Example 1

[0394] Next, a description will be given of Example 1 of the embodiment 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".

[0395] With the conventional system, it was difficult to not only collect the history of inquiries to the company and information on the Internet, but also to effectively analyze that information and understand the emotional state of the customer. It was also difficult to devise specific service improvement measures based on the analysis results. Furthermore, there was a lack of a process to visualize the factors that were lowering customer satisfaction and to devise improvement methods based on that. As a result, the system was unable to fully contribute to the company's quality improvement and cost reduction.

[0396] 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.

[0397] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, a means for saving the collected information as text data, an artificial intelligence means for analyzing the saved text data, an emotion engine means for analyzing the emotional state of the customer based on the analysis result, a means for planning improvements to the company's services based on the analysis result, a means for visualizing factors that lower customer satisfaction based on the analysis result, and a means for planning improvement methods based on the visualization result. This makes it possible to effectively analyze the collected information, grasp the emotional state of the customer, and plan specific service improvement measures. In addition, by visualizing factors that lower customer satisfaction and planning improvement methods based on the visualization, it becomes possible to contribute to improving the quality of the company and reducing costs.

[0398] "Inquiry history to our company" refers to the content and history of inquiries that a company has received from customers.

[0399] "Online information" refers to information that exists on the Internet, such as reviews, comments, blog posts, and social media posts.

[0400] "Means of collection" refers to methods of obtaining information using web scraping tools, APIs, etc.

[0401] "Text data" refers to information stored in the form of strings of characters.

[0402] "Means of storage" refers to the way information is stored, such as in a database or file system.

[0403] "Artificial intelligence means" refers to methods of analyzing data using machine learning models or natural language processing models.

[0404] "Emotion engine means" refers to algorithms or tools for analyzing emotions from text data.

[0405] "Means for proposing improvements to services" refers to methods for proposing specific improvement measures based on the results of analysis.

[0406] "Means of visualizing factors that are reducing customer satisfaction" refers to methods of displaying the analysis results in a visual format such as graphs or charts.

[0407] "Means for planning improvement methods" refers to a method for formulating specific improvement measures based on visualized data.

[0408] The present invention is a system for collecting inquiry histories and information on the Internet and analyzing the collected information to improve services. A specific embodiment of this system will be described below.

[0409] Information gathering

[0410] The server collects information using web scraping tools and APIs. Specifically, it uses web scraping tools such as BeautifulSoup and Scrapy, or the Twitter API and Google Custom Search API. For example, the server performs a task such as "collect the latest tweets about the service."

[0411] Data Storage

[0412] The server stores the collected information as text data. Specifically, it stores it in a database such as MySQL or MongoDB. For example, the server performs the operation "save the collected tweets in MongoDB."

[0413] 4. Data Analysis

[0414] The server inputs the stored text data into an AI model and performs analysis. Specifically, it uses natural language processing models such as BERT and GPT-3. For example, the server executes the process of "inputting the stored tweets into the BERT model and extracting problems related to the service."

[0415] Emotion analysis

[0416] The server analyzes the user's emotions using an emotion engine. Specifically, it uses IBM Watson (registered trademark) Tone Analyzer or Microsoft Azure (registered trademark) Text Analytics. For example, the server executes a task of "analyzing the user's emotions from the contents of tweets and detecting dissatisfaction or anger."

[0417] Proposal for service improvement

[0418] The server will propose service improvements based on the analysis results of the AI ​​model. Specifically, it will analyze user inquiries and online feedback to extract areas for improvement. For example, the server will perform a process such as "proposing improvement measures for service delays based on the analysis results."

[0419] Understanding emotional states

[0420] The server grasps the user's emotional state based on the analysis results of the emotion engine. Specifically, it estimates the user's emotional state from the language and tone used when the user makes an inquiry. For example, the server executes a task such as "evaluating the user's satisfaction or dissatisfaction level based on the content of the inquiry."

[0421] Examples

[0422] For example, if a user inquires that "the service is slow," the server collects the content of the inquiry and saves it as text data. The saved data is then input into an AI model to analyze the specific issues related to the service delay. At the same time, an emotion engine is used to analyze the user's emotions, such as dissatisfaction or anger. This allows the server to understand not only the user's satisfaction level, but also the emotions behind it.

[0423] Examples of prompt statements

[0424] "Generate prompts to infer a user's emotional state from the language and tone used when making a query."

[0425] In this way, it is possible to explain a specific embodiment of the system while clarifying the roles of the server, the terminal, and the user.

[0426] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0427] Step 1: Gather information

[0428] The server collects information using web scraping tools and APIs. Specifically, it uses BeautifulSoup or Scrapy to extract data from websites. It also uses the Twitter API and Google Custom Search API to obtain information on the Internet. As input, a specific website URL or API endpoint is given. As output, the collected text data is obtained. For example, the server collects the latest reviews from "https: / / example.com / reviews".

[0429] Step 2: Save Data

[0430] The server saves the collected information as text data. Specifically, it saves it in a database such as MySQL or MongoDB. The collected text data is given as input. The data saved in the database is obtained as output. For example, the server performs the operation "save the collected reviews in the 'reviews' collection in MongoDB."

[0431] Step 3: Data analysis

[0432] The server inputs the stored text data into an AI model and performs analysis. Specifically, it uses natural language processing models such as BERT and GPT-3. The text data stored in the database is given as input. The analysis results are obtained as output. For example, the server executes the process of "inputting the stored reviews into the BERT model and extracting problems with the service."

[0433] Step 4: Sentiment Analysis

[0434] The server uses an emotion engine to analyze the user's emotions. Specifically, it uses IBM Watson Tone Analyzer or Microsoft Azure Text Analytics. Analyzed text data is provided as input. The emotion analysis results are obtained as output. For example, the server executes a task such as "analyzing the user's emotions from the content of the review and detecting dissatisfaction or anger."

[0435] Step 5: Plan service improvements

[0436] The server proposes service improvements based on the analysis results of the AI ​​model. Specifically, it analyzes user inquiries and online feedback to extract areas for improvement. The analysis results and sentiment analysis results are given as input. Specific improvement measures are obtained as output. For example, the server might carry out the process of "proposing improvement measures for service delays based on the analysis results."

[0437] Step 6: Understand your emotional state

[0438] The server grasps the user's emotional state based on the analysis results from the emotion engine. Specifically, the server estimates the user's emotional state from the language and tone used when the user makes an inquiry. The inquiry content and the emotion analysis results are given as input. The user's emotional state is obtained as output. For example, the server executes a task such as "evaluating the user's satisfaction or dissatisfaction level based on the inquiry content."

[0439] (Application example 1)

[0440] Next, a description will be given of Application Example 1 of Form 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."

[0441] Conventional customer support systems can collect and analyze user inquiries and reviews, but they have the problem that it is difficult to grasp the user's emotional state in real time. Even if the analysis results are visualized, it is difficult to quickly come up with specific improvement methods. This leads to a decrease in customer satisfaction and an increase in unnecessary calls to call centers, which has a negative impact on the quality and costs of companies.

[0442] 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.

[0443] In this invention, the server includes a means for collecting inquiry history and online information, an AI means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for analyzing the emotional state of the user using an emotion engine, and a means for displaying the analysis results in real time. This makes it possible to grasp the emotional state of the user in real time and quickly plan specific improvement methods.

[0444] "Inquiry history to our company" refers to data that records the inquiries received by the company from customers.

[0445] "Online information" refers to text data such as reviews, comments, blog posts, forum posts, etc. that exist on the Internet.

[0446] "Means of collection" refers to the method of obtaining information using web scraping tools or APIs.

[0447] "Analyzing" is the process of analyzing collected data to extract useful information.

[0448] "AI means for proposing improvements to our own services" is a system that uses artificial intelligence to propose ways to improve services.

[0449] A "means for visualizing factors that are reducing customer satisfaction" is a method for displaying the results of data analysis in a visual format such as a graph or chart.

[0450] "Means for planning improvement methods" is the process of formulating specific improvement measures based on visualized data.

[0451] An "emotion engine" is software for analyzing a user's emotional state from text data.

[0452] The "means for analyzing the user's emotional state" is a method for analyzing the user's emotions using an emotion engine.

[0453] "Means for displaying the analysis results in real time" refers to a system for instantly displaying the analyzed data to the user.

[0454] A system for carrying out this invention has the following configuration: The server includes means for collecting the history of inquiries made to the company and information on the internet, AI means for analyzing the collected information and proposing improvements to the company's services, means for visualizing factors that are lowering customer satisfaction based on the analysis results, means for proposing improvement methods based on the visualization results, means for analyzing the emotional state of the user using an emotion engine, and means for displaying the analysis results in real time.

[0455] Hardware and Software Used

[0456] Hardware: Servers, smartphones

[0457] Software: Python, BeautifulSoup, requests, OpenAI API, TextBlob

[0458] Data processing and calculation

[0459] 1. Data collection: The server uses web scraping tools (BeautifulSoup, requests) and APIs to collect inquiry history and online information for the company. This information is saved as text data.

[0460] 2. Information analysis: The server analyzes the collected text data using AI means (OpenAI API). A generative AI model is used to understand the content of the information and to plan service improvements.

[0461] 3. Emotion Analysis: The server uses an emotion engine (TextBlob) to analyze the user's emotional state, for example by inferring the user's emotional state from the wording and tone of the query.

[0462] 4. Visualization: The server displays the analysis results in a visual format such as graphs and charts, allowing users to intuitively understand the factors that are causing low customer satisfaction.

[0463] 5. Real-time Display: The server provides immediate feedback to the user by providing a means for displaying the analysis results in real-time.

[0464] Examples

[0465] For example, if a user reviews "This product was very good, but the delivery was slow," the server collects this text data and analyzes it using AI means. The analysis results provide insights such as "Satisfaction with the product is high, but dissatisfaction with the delivery" The sentiment analysis reveals a mix of positive and negative emotions. These analysis results are visualized in graphs and charts and displayed in real time.

[0466] Examples of prompt statements

[0467] Analyze the following text and provide insights: "This product was very good, but the delivery was slow."

[0468] In this way, it is possible to grasp the user's emotional state and level of satisfaction in real time and devise quick and specific methods of improvement.

[0469] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0470] Step 1:

[0471] The server uses web scraping tools (BeautifulSoup, requests) and APIs to collect the company's inquiry history and other online information. The input is the website URL and API endpoint, and the output is text data. Specifically, the server accesses the specified URL, retrieves the HTML content, and analyzes it. As a result of the analysis, text data is extracted and stored in a database.

[0472] Step 2:

[0473] The server analyzes the collected text data using AI means (OpenAI API). The input is the text data collected in step 1, and the output is the analysis result. Specifically, the server sends the text data to a generative AI model and performs analysis using a prompt sentence. As an analysis result, it obtains insight into the content of the text data.

[0474] Step 3:

[0475] The server uses an emotion engine (TextBlob) to analyze the user's emotional state. The input is the text data collected in step 1, and the output is the emotion analysis result. Specifically, the server inputs the text data into the emotion engine and analyzes the emotional state into positive, negative, and neutral. As the analysis result, the user's emotional state is quantified and output.

[0476] Step 4:

[0477] The server displays the analysis results in a visual format such as a graph or chart. The input is the analysis results obtained in step 2 and step 3, and the output is visualized data. Specifically, the server inputs the analysis results into a data visualization tool to generate graphs and charts. The generated visualized data is displayed in a format that the user can intuitively understand.

[0478] Step 5:

[0479] The server provides immediate feedback to the user using a means for displaying the analysis results in real time. The input is the data visualized in step 4, and the output is the feedback displayed in real time. Specifically, the server transmits the visualized data to the user's terminal and displays it in real time. This allows the user to immediately check the analysis results and take any necessary action.

[0480] Example 2

[0481] Next, a second embodiment of the second embodiment will be described. 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."

[0482] In the conventional system, it was difficult to effectively analyze the information and plan specific service improvement measures by simply collecting inquiries and feedback from users. In addition, since service improvements were not made based on the user's feelings, there was a problem that customer satisfaction was not sufficiently improved. Furthermore, there were many unnecessary calls to the call center, which had a negative impact on the company's quality and costs.

[0483] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0484] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, a means for sentiment analysis of the collected information, a means for providing the sentiment analysis result to a generative AI, a means for analyzing the information using the generative AI and extracting factors that are lowering customer satisfaction, a means for planning service improvement measures based on the extracted factors, and a means for feeding back the planned improvement measures to the user. This makes it possible to plan specific service improvement measures based on the user's sentiment, thereby improving customer satisfaction and reducing unnecessary calls to the call center.

[0485] "Inquiry history" is a record of inquiries made by a user to a company.

[0486] "Online information" refers to information such as user feedback, reviews, and comments that exist on the Internet.

[0487] "Sentiment analysis" is the process of analyzing user emotions (anger, joy, sadness, etc.) from text data.

[0488] "Generative AI" is artificial intelligence that has the ability to learn from large amounts of text data and generate new text.

[0489] "Information analysis" is the process of analyzing collected data and extracting useful information.

[0490] "Customer satisfaction" refers to the level of satisfaction that users feel toward a company's services or products.

[0491] "Cause extraction" is the process of identifying the causes of lower customer satisfaction.

[0492] "Service improvement measures" are specific methods of improvement implemented to enhance customer satisfaction.

[0493] "Feedback" refers to information or notification of improvements provided to a user.

[0494] A "call center" is a telephone response center for accepting inquiries and support from users.

[0495] "Unnecessary calls" are unnecessary phone inquiries due to resolved issues or lack of information.

[0496] "Quality" is a characteristic of a product or service that relates to its performance, reliability, and user satisfaction.

[0497] "Cost" refers to the cost incurred in providing a product or service.

[0498] The present invention is a system for collecting inquiries from users and information on the Internet, analyzing the collected information, and formulating measures to improve services. A specific embodiment of this system will be described below.

[0499] The server first collects the user's inquiry history and online information. This information is stored in a database in text format. For example, if a user sends an inquiry such as "I haven't received my product," the content of that inquiry is stored in the database.

[0500] Next, the server sends the collected text data to an emotion analysis engine to analyze the user's emotions. The emotion analysis engine detects emotions such as "anger," "joy," and "sadness" from the text data. For example, it detects the emotion of "anger" from the text "The product has not arrived."

[0501] The server provides the results of the emotion analysis to the generative AI. The generative AI uses a natural language processing model, such as OpenAI's GPT-3. The generative AI performs information analysis based on the emotion analysis results and extracts factors that are lowering customer satisfaction. For example, it identifies that "delayed delivery" is the cause of the emotion "anger."

[0502] Next, the generative AI proposes specific service improvement measures for the extracted factors, such as "changing the delivery company" or "reviewing the delivery process."

[0503] Finally, the server feeds back the improvements obtained from the generative AI to the user, for example, sending feedback to the user such as "The delivery company has been changed."

[0504] As a concrete example, consider a situation where many users of an online shopping site are dissatisfied with "slow delivery." If the emotion engine detects a lot of "anger" emotions, the server uses generative AI to identify the "cause of the delivery delay" and proposes improvement measures such as "changing the delivery company" or "reviewing the delivery process."

[0505] An example of a prompt sentence might be:

[0506] “If many users are angry about delivery delays, what remedial measures can you suggest?”

[0507] In this way, the server can utilize generative AI to improve services based on user emotions.

[0508] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0509] Step 1: Gather information

[0510] The server collects user inquiry history and online information.

[0511] Input: User inquiry emails, chat logs, online reviews and comments.

[0512] Data processing: The collected information is stored in a database in text format.

[0513] Output: Text-formatted inquiry history and online information are saved in a database.

[0514] Specific operation: The server automatically collects inquiry emails and chat logs sent by users and stores them in a database. For example, if a user sends an inquiry such as "I haven't received my product," the content of the inquiry is stored in the database.

[0515] Step 2: Sentiment Analysis

[0516] The server sends the collected text data to an emotion analysis engine to analyze the user's emotions.

[0517] Input: Text-based inquiry history stored in the database and online information.

[0518] Data calculation: The sentiment analysis engine analyzes the user's emotions (anger, joy, sadness, etc.) from the text data.

[0519] Output: Sentiment analysis results (e.g. emotion tags such as "anger", "joy", "sadness" etc.).

[0520] Specific operation: The server inputs the text data into a sentiment analysis engine and analyzes the user's sentiment. For example, it detects the sentiment of "anger" from the text "The product has not arrived."

[0521] Step 3: Provide information

[0522] The server provides the emotion analysis results to the generative AI.

[0523] Input: Sentiment analysis results.

[0524] Data processing: Converting the emotion analysis results into a format for sending to the generative AI.

[0525] Output: Sentiment analysis results provided to the generative AI.

[0526] Specific operation: The server sends the analysis results obtained from the emotion engine to the generative AI. For example, it provides text data in which the emotion "anger" has been detected to the generative AI.

[0527] Step 4: Factor extraction

[0528] Based on the emotion analysis results provided, the generative AI extracts the factors that are lowering customer satisfaction.

[0529] Input: Sentiment analysis results.

[0530] Data Computation: Generative AI analyzes sentiment analysis results to identify the root cause of problems.

[0531] Output: Factors that are causing low customer satisfaction (e.g. "delayed delivery").

[0532] How it works: The generative AI analyzes the results of emotion analysis and identifies the root cause of the problem. For example, it identifies that "delayed delivery" is the cause of the emotion "anger."

[0533] Step 5: Plan improvements

[0534] The generative AI then devise improvement measures for the extracted factors.

[0535] Input: Factors that are reducing customer satisfaction.

[0536] Data Computation: Generative AI generates specific solutions to identified problems.

[0537] Output: Service improvement measures (e.g., "change delivery company" or "review delivery process").

[0538] Specific behavior: Generative AI generates specific improvement measures for the identified problems. For example, it proposes improvement measures such as "changing the delivery company" or "reviewing the delivery process."

[0539] Step 6: Feedback on results

[0540] The server provides feedback to the user on improvements obtained from the generative AI.

[0541] Input: Service improvement measures from generative AI.

[0542] Data processing: Convert the improvement measures into a format that can be communicated to users.

[0543] Output: Remediation measures communicated to the user.

[0544] Specific operation: The server notifies the user of the improvement measures obtained from the generative AI. For example, it sends feedback to the user such as "The delivery company has been changed."

[0545] (Application example 2)

[0546] Next, a description will be given of Application Example 2 of Form 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."

[0547] Conventional customer satisfaction improvement systems only collect and analyze inquiry history and online information, and are unable to provide specific improvement suggestions based on user emotions. This makes it difficult to effectively improve customer satisfaction, and there is a lack of specific measures to alleviate the dissatisfaction of users who have negative emotions. In addition, it is not possible to reduce unnecessary calls to the call center, and companies are required to improve their quality and cost efficiency.

[0548] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0549] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, an AI means for analyzing the collected information and proposing improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for proposing improvement methods based on the visualization results, a means for collecting feedback and reviews from users and performing sentiment analysis, and a means for extracting feedback with negative sentiments and generating improvement proposals. This makes it possible to make specific improvement proposals based on the user's sentiments, and to effectively improve customer satisfaction. In addition, it is possible to reduce unnecessary calls to the call center, and to realize efficiency in terms of quality and cost for the company.

[0550] An "inquiry history" is a record of inquiries made by a customer to a company.

[0551] "Online information" refers to data such as reviews, comments, blog articles, and social media posts that exist on the Internet.

[0552] "Means of collection" refers to software or hardware mechanisms for automatically collecting inquiry history and online information.

[0553] "Means of analysis" are the algorithms and tools used to analyze collected information and convert it into meaningful data.

[0554] "AI means" is a system that uses artificial intelligence technology to analyze information and make suggestions for improving services.

[0555] A "visualization tool" is any tool or software that allows the analysis results to be displayed in a visual format, such as a graph or chart.

[0556] "Means for planning improvement methods" refers to processes and tools for devising specific improvement measures based on the visualized analysis results.

[0557] "Feedback" means opinions or thoughts provided by a user.

[0558] A "review" is a piece of writing or comment in which a user evaluates a product or service.

[0559] "Sentiment analysis" is a technology that automatically determines a user's emotions from text data.

[0560] "Negative emotions" refers to negative feelings such as anger, frustration, and sadness.

[0561] A "means for generating improvement suggestions" is an algorithm or tool for automatically generating specific improvement measures based on feedback with negative sentiment.

[0562] A system for carrying out the present invention has the following configuration: The server includes means for collecting inquiry history and online information, AI means for analyzing the collected information and proposing improvements to the company's services, means for visualizing factors that are lowering customer satisfaction based on the analysis results, means for proposing improvement methods based on the visualization results, means for collecting feedback and reviews from users and performing sentiment analysis, and means for extracting feedback with negative sentiment and generating improvement proposals.

[0563] Hardware and Software Used

[0564] Hardware: Servers, user devices (smartphones, PCs, etc.)

[0565] Software: Python, OpenAI API, TextBlob

[0566] Data processing and calculation

[0567] 1. Collecting inquiry history and online information:

[0568] The server automatically collects inquiry history and online information using web scraping tools and APIs.

[0569] 2. Analysis of information:

[0570] The server uses AI means to analyze the collected information. Specifically, it uses a generative AI (generative pretrained transformer) to analyze the collected text data and extract factors that are lowering customer satisfaction.

[0571] 3. Emotion analysis:

[0572] The server uses TextBlob to parse the user feedback / reviews for sentiment and extracts the feedback if a negative sentiment (e.g. anger, dissatisfaction, sadness, etc.) is detected.

[0573] 4. Generate improvement suggestions:

[0574] The server uses the OpenAI API to generate specific improvement suggestions based on the negative sentiment feedback, which are then displayed in the form of graphs and charts using visualization tools.

[0575] Examples

[0576] For example, suppose a user gives the following feedback:

[0577] "The delivery of the product was late."

[0578] "The customer service was poor."

[0579] "The item was damaged."

[0580] The server collects these feedbacks, performs sentiment analysis using TextBlob, extracts feedback where negative sentiment is detected, and generates improvement suggestions using the OpenAI API, such as:

[0581] Examples of prompt statements

[0582] Please provide your suggestions for improving our service based on your feedback below:

[0583] Delivery of the product was late.

[0584] Customer service was poor.

[0585] The product was damaged.

[0586] By inputting this prompt text into the OpenAI API, specific improvement suggestions can be obtained. The generated improvement suggestions are displayed in the form of graphs and charts using a visualization tool, allowing companies to plan specific improvement measures based on this.

[0587] The flow of the identification process in the application example 2 will be described with reference to FIG.

[0588] Step 1:

[0589] The server collects the company's inquiry history and online information. Specifically, it uses web scraping tools and APIs to collect data such as inquiry history, reviews, comments, and SNS posts. The input is the inquiry history and online information, and the output is the collected data.

[0590] Step 2:

[0591] The server analyzes the collected information. Specifically, it uses a generative AI (Generative Pretrained Transformer) to analyze the collected text data and extract the factors that are reducing customer satisfaction. The input is the collected data, and the output is the analysis results.

[0592] Step 3:

[0593] The server visualizes the factors that are lowering customer satisfaction based on the analysis results. Specifically, it uses a visualization tool to display the analysis results in the form of graphs and charts. The input is the analysis results, and the output is the visualized data.

[0594] Step 4:

[0595] The server plans improvement methods based on the visualized results. Specifically, it uses processes and tools to devise improvement measures based on the visualized data. The input is the visualized data, and the output is the improvement methods.

[0596] Step 5:

[0597] The server collects feedback and reviews from users and performs sentiment analysis. Specifically, it uses TextBlob to analyze the sentiment of feedback and reviews. The input is the feedback and reviews from users, and the output is the sentiment analysis result.

[0598] Step 6:

[0599] The server extracts feedback with negative emotions. Specifically, it extracts feedback in which negative emotions (e.g., anger, dissatisfaction, sadness, etc.) are detected from the emotion analysis results. The input is the emotion analysis results, and the output is negative feedback.

[0600] Step 7:

[0601] The server generates improvement suggestions based on feedback with negative sentiment. Specifically, it uses the OpenAI API to generate specific improvement suggestions based on negative feedback. The input is the negative feedback, and the output is the improvement suggestions.

[0602] Step 8:

[0603] The server visualizes the generated improvement proposals. Specifically, it uses a visualization tool to display the improvement proposals in the form of graphs and charts. The input is the improvement proposals, and the output is the visualized improvement proposals.

[0604] Example 3

[0605] Next, a third embodiment of the third embodiment will be described. 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."

[0606] With the conventional system, it was difficult to effectively collect and analyze customer inquiries and online evaluations, visualize them, and propose improvement measures.In addition, there was a lack of means to create specific action plans, so it was not possible to fully improve customer satisfaction or reduce unnecessary calls to the call center.

[0607] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0608] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, a means for analyzing the collected information and identifying the emotional state of customers and items of frequent inquiries using natural language processing technology, a means for visualizing the analysis results using a dashboard and graphs, a means for proposing improvement methods using a generative AI model based on the visualized results, and a means for creating a specific action plan based on the proposed improvement methods. This enables a company to quickly plan and implement specific improvement measures to improve customer satisfaction.

[0609] "Inquiry history" refers to a record of the content, date and time, frequency, etc. of inquiries made by a customer to a company.

[0610] "Online information" refers to data such as customer ratings, comments, and reviews that exist on the Internet.

[0611] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[0612] "Emotional state" refers to the emotional state of a customer when making an inquiry or evaluation, and includes emotions such as "satisfied" and "dissatisfied."

[0613] A "dashboard" is an interface for visually displaying data, using graphs and charts to allow information to be grasped at a glance.

[0614] A "generative AI model" is an artificial intelligence model that generates new information or suggestions based on input data.

[0615] A "prompt" is an instruction or question input to a generative AI model that causes the AI ​​to generate a response or suggestion based on that instruction.

[0616] An "action plan" is a specific plan or procedure to be followed to achieve a particular goal.

[0617] This invention is a system for enabling a company to improve customer satisfaction, and is realized by a server, a terminal, and a user fulfilling their respective roles.

[0618] Data collection

[0619] The server collects customer inquiry history and online information from the company's CRM system and SNS API. For example, the content and frequency of customer inquiries are obtained from the CRM system, and evaluation comments from Twitter and Facebook are collected from the SNS API. This data is stored in a database.

[0620] 4. Data Analysis

[0621] The server analyzes the collected data. Specifically, it uses a Python NLP library (e.g., spaCy) to analyze the text data and identify the emotional state of customers and the most frequently inquired items. For example, it extracts emotions such as "dissatisfied" or "satisfied" from the content of inquiries and counts their frequency.

[0622] Visualization

[0623] The server visualizes the analysis results. Tableau or Power BI is used to generate dashboards and graphs. For example, the dashboard could display a "ranking of frequently inquired items" or a "distribution of emotional states." Company personnel can access the dashboard via a browser and check the results.

[0624] Proposal for improvement

[0625] Based on the visualized results, the server uses a generative AI model (e.g., GPT-3) to suggest ways to improve the service. For example, it generates specific suggestions such as "improving FAQs for frequently asked questions" or "improving services that cause complaints." The generated suggestions are displayed on the dashboard.

[0626] Create an action plan

[0627] The user (company representative) creates a specific action plan based on the improvement measures proposed by the AI. For example, the user decides on specific actions such as "creating a new FAQ page" or "assembling a project team to improve a specific service."

[0628] Examples

[0629] For example, let's say a company wants to improve customer satisfaction. The system is used in the following steps:

[0630] 1. Data Collection

[0631] The server calls the CRM system API at a fixed time every day to obtain the latest inquiry data. It also uses the SNS API to collect evaluation comments from the past 24 hours. The collected data is stored in a database.

[0632] 2. Data Analysis

[0633] The server runs Python scripts to analyze the collected data. The scripts tokenize the text data and perform sentiment analysis, for example extracting the sentiment "unhappy" from the comment "This service is terrible."

[0634] 3. Visualization

[0635] The server sends the analysis results to Tableau and updates the dashboard. The dashboard displays the latest ranking of items with the most inquiries and the distribution of emotional states. Company personnel can access the dashboard through a browser to check the results.

[0636] 4. Proposal for improvement

[0637] The server sends a prompt to a generative AI model (e.g., GPT-3) to generate improvement measures. For example, a suggestion such as "improving FAQs for frequently asked questions" is generated. The generated suggestions are displayed on the dashboard.

[0638] 5. Create an action plan

[0639] The user (company staff) checks the suggestions on the dashboard and creates a specific action plan. For example, the user may create an action plan such as "select a person to create a new FAQ page and publish the page within a week."

[0640] Examples of prompt statements

[0641] Below is an example of a prompt sentence to be input to the generative AI model.

[0642] "Please suggest improvements to improve customer satisfaction based on the content and frequency of customer inquiries and online reviews."

[0643] In this way, the company can improve customer satisfaction and reduce unnecessary calls to the call center. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0644] Step 1: Data collection

[0645] The server collects customer inquiry history and online information from the company's CRM system and SNS API. As input, it uses the API key of the CRM system and authentication information of the SNS API. Specifically, the server calls the API at a fixed time every day to obtain the latest inquiry data and evaluation comments. As output, the collected data is stored in a database.

[0646] Step 2: Data analysis

[0647] The server analyzes the collected data. As input, it uses the inquiry content and evaluation comments stored in the database. Specifically, it uses a Python NLP library (e.g., spaCy) to tokenize the text data and perform sentiment analysis. For example, it extracts the sentiment "dissatisfied" from the comment "This service is terrible." As output, the analysis results are stored in the database.

[0648] Step 3: Visualization

[0649] The server visualizes the analysis results. As input, it uses the analysis results stored in the database. Specifically, it generates dashboards and graphs using Tableau or Power BI. For example, it displays "ranking of frequently inquired items" or "distribution of emotional states" on the dashboard. As output, the visualized data is displayed on the dashboard.

[0650] Step 4: Propose improvements

[0651] Based on the visualization results, the server uses a generative AI model (e.g. GPT-3) to suggest improvement methods. As input, it uses the visualized data and a prompt text. In concrete terms, it sends the prompt text to the generative AI model to generate improvement measures. For example, a suggestion such as "improving FAQs for items with many inquiries" is generated. As output, the generated suggestions are displayed on a dashboard.

[0652] Step 5: Create an action plan

[0653] The user (company staff) creates a specific action plan based on the improvement measures proposed by the AI. As input, the suggestions displayed on the dashboard are used. As specific operations, the user reviews the suggestions, selects a person to be responsible for creating a new FAQ page, and creates a specific action plan, such as publishing the page within a week. As output, the created action plan is shared within the company and put into action.

[0654] (Application example 3)

[0655] Next, a description will be given of Application Example 3 of Form Example 3. 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."

[0656] With conventional customer satisfaction improvement systems, it was difficult to effectively collect and analyze customer inquiries and online evaluations, and quickly develop specific improvement measures. In addition, the proposals for improvement measures based on visualized data were insufficient, and it took a lot of effort for companies to create specific action plans. Furthermore, there was a lack of concrete measures to reduce unnecessary calls to the call center, so there was a need to improve efficiency in both quality and cost.

[0657] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0658] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, an AI means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that lower customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for proposing improvement measures using a generative AI model, a means for creating a specific action plan based on the proposed improvement measures, and a means for collecting feedback from customers in real time and visualizing it in a dashboard or graph. This allows specific improvement measures to improve customer satisfaction to be quickly proposed, and companies to efficiently create action plans. It also reduces unnecessary calls to the call center, contributing to companies in terms of both quality and cost.

[0659] "Inquiry history to our company" refers to the content and history of inquiries that a company has received from customers.

[0660] "Online information" refers to information such as customer ratings, reviews, and comments that are published on the Internet.

[0661] "Means of collection" refers to technologies and methods for automatically collecting inquiry history and online information.

[0662] "Means of analysis" refers to the techniques and methods used to analyze collected information and convert it into meaningful data.

[0663] "AI means" refers to technologies and methods that use artificial intelligence to analyze data and develop improvement measures.

[0664] "Visualization means" refers to techniques and methods for displaying analytical results in a visual format, such as graphs or dashboards.

[0665] "Means for planning improvement methods" refers to techniques and methods for devising specific improvement measures based on visualized data.

[0666] A "generative AI model" refers to a generative artificial intelligence model, and is a technology for generating data and information for specific purposes.

[0667] "Proposed means" refers to techniques and methods for proposing improvement measures using generative AI models.

[0668] "Means of creating an action plan" refers to techniques and methods for creating a specific action plan based on proposed improvement measures.

[0669] "Means of collecting feedback in real time" refers to technologies and methods for collecting feedback from customers immediately.

[0670] "Dashboard and / or graphical visualization means" refers to techniques and methods for displaying collected feedback in the form of dashboards and / or graphical displays.

[0671] The system for implementing this invention is composed of three main elements: a server, a terminal, and a user. The server is responsible for collecting customer inquiry history and information on the Internet and analyzing this data. Specifically, the server includes the following means:

[0672] 1. Collection method: The server automatically collects inquiry history and online evaluations. This is done using web scraping technology or APIs. For example, Python libraries such as BeautifulSoup and Scrapy can be used.

[0673] 2. Means of analysis: To analyze the collected data, the server uses Natural Language Processing (NLP) techniques, specifically vectorizing the text data and classifying the data using a clustering algorithm (e.g. KMeans), which allows the customer feedback to be divided into meaningful categories.

[0674] 3. AI method: The server uses a generative AI model to propose improvement measures based on the analysis results. Here, a generative AI model such as OpenAI's GPT-3 is used. A specific improvement measure is generated using the prompt sentence as input.

[0675] 4. Visualization: To visualize the analysis results, the server generates dashboards and graphs using data visualization libraries such as Matplotlib and Plotly, allowing users to understand the analysis results at a glance.

[0676] 5. Means of planning improvement methods: The server creates a specific action plan based on the visualized data. Based on the improvement measures proposed by the generative AI model, the server plans an action plan that the user can execute.

[0677] 6. Means of collecting feedback in real time: The server collects feedback from customers in real time. For this, a real-time database (e.g. Firebase) can be used.

[0678] 7. Visualization in the form of dashboards and graphs: The server displays the collected feedback in the form of dashboards and graphs, allowing users to understand the feedback trends in real time.

[0679] As a concrete example, the following prompt sentence can be input to a generative AI model:

[0680] Please suggest improvements to improve customer satisfaction. Below is some feedback from customers.

[0681] Product delivery is slow

[0682] Poor customer support

[0683] The quality of the product was not as expected

[0684] Based on this prompt, the generative AI model will suggest specific improvements, such as "reviewing the delivery process," "strengthening customer support training," and "stricter product quality control."

[0685] In this way, the server can quickly propose specific improvement measures to improve customer satisfaction, enabling companies to create action plans efficiently. It can also reduce unnecessary calls to the call center, contributing to companies in terms of both quality and cost.

[0686] The flow of the specific process in the application example 3 will be described with reference to FIG.

[0687] Step 1:

[0688] The server collects customer inquiry history and online information. Specifically, it uses web scraping technology and APIs to automatically collect customer feedback and ratings. Website URLs and API endpoints are required as input, and collected text data is obtained as output.

[0689] Step 2:

[0690] The server analyzes the collected data. Specifically, it uses natural language processing (NLP) techniques to vectorize the text data and classifies the data using a clustering algorithm (e.g., KMeans). As input, it requires the collected text data, and as output, it obtains the clustered data.

[0691] Step 3:

[0692] The server uses a generative AI model to propose improvement measures based on the analysis results. Specifically, it uses a generative AI model such as OpenAI's GPT-3 to generate improvement measures using the prompt sentence as input. Clustered data and the prompt sentence are required as input, and the generated improvement measures are obtained as output.

[0693] Step 4:

[0694] The server visualizes the analysis results, specifically by generating dashboards and graphs using data visualization libraries such as Matplotlib and Plotly. As input, it requires the clustered data, and as output, it gets the visualized graphs and dashboards.

[0695] Step 5:

[0696] The server creates a concrete action plan based on the visualized data. Based on the improvement measures proposed by the generative AI model, the server creates an action plan that the user can execute. As input, the generated improvement measures are required, and as output, a concrete action plan is obtained.

[0697] Step 6:

[0698] The server collects customer feedback in real time. Specifically, it uses a real-time database (e.g. Firebase) to collect feedback instantly. As input, it requires customer feedback and as output, it gets feedback data that is updated in real time.

[0699] Step 7:

[0700] The server displays the collected feedback in the form of dashboards and graphs. Specifically, it uses a data visualization library to display feedback trends in real time. As input, it requires feedback data updated in real time, and as output, it gets visualized dashboards and graphs.

[0701] 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 a voice indicating a user input for the result of the specific processing. The control unit 46A transmits the voice 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 voice data.

[0702] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by making a neural network perform deep learning. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image is input. The data generation model 58 performs inference on the input inference data according to 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.

[0703] Another example of generative AI is Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) are mentioned.

[0704] In the above embodiment, an example was given in which the specific process was performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0705] [Second embodiment]

[0706] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0707] 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.

[0708] 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 wide area network (WAN) and / or a local area network (LAN).

[0709] 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.

[0710] 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 the voice according to instructions from the processor 46.

[0711] Camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a 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 (e.g., an imaging range defined by an angle of view equivalent to the width of the field of vision of an average healthy person).

[0712] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54. The transmission and reception of various types of information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is performed in a secure state.

[0713] Fig. 4 shows an example of 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.

[0714] 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.

[0715] 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.

[0716] 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.

[0717] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0718] "Example 1"

[0719] As one embodiment of the present invention, a web scraping tool or API is used to collect information such as inquiry history and online information. The collected information is saved as text data and becomes input data for analysis by AI.

[0720] "Example 2"

[0721] Next, we use a generative AI (Generative Pretrained Transformer), a type of natural language processing (NLP), as an AI method to analyze the collected information. Generative AI has the ability to learn from large amounts of text data and generate new text. Using this ability, we extract factors that are lowering customer satisfaction from the collected information.

[0722] "Example 3"

[0723] Furthermore, dashboards and graphs are used to visualize the analysis results. This allows companies to understand problems at a glance and devise improvement measures. Specifically, the content and frequency of customer inquiries, online reviews, etc. are visualized, and from this information, factors that are lowering customer satisfaction are identified.

[0724] "Example 4"

[0725] Finally, the improvement measures proposed by AI are considered as a means of planning improvement methods based on the visualized results. Specifically, based on the improvement measures proposed by AI, companies can determine the direction of service improvement and create specific action plans. This allows companies to improve customer satisfaction and reduce unnecessary calls to their call centers.

[0726] The process flow of each embodiment will be described below.

[0727] "Example 1"

[0728] Step 1: Use a web scraping tool or API to collect inquiry history and online information for your company.

[0729] Step 2: The collected information is saved as text data and used as input data for AI analysis.

[0730] "Example 2"

[0731] Step 1: Generative AI, a type of natural language processing (NLP)

[0732] The collected information is analyzed using a trained Transformer.

[0733] Step 2: Using the results of the generative AI's learning, generate new text and extract the factors that are reducing customer satisfaction.

[0734] "Example 3"

[0735] Step 1: Use dashboards and graphs to visualize the analysis results.

[0736] Step 2: From the visualized results, identify problems and develop improvement measures.

[0737] "Example 4"

[0738] Step 1: Based on the improvement measures suggested by AI, the company decides on the direction of service improvement.

[0739] Step 2: Create a concrete action plan to improve customer satisfaction and reduce unnecessary calls to the call center.

[0740] Example 1

[0741] Next, a description will be given of Example 1 of Form 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".

[0742] With the conventional system, it was difficult to efficiently collect and analyze the history of inquiries to the company and information on the Internet. It was also difficult to properly analyze the collected information and devise specific improvement methods to increase customer satisfaction. Furthermore, there was a lack of effective means to reduce unnecessary calls to the call center.

[0743] 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.

[0744] In this invention, the server includes a means for collecting inquiry history and online information to the company using a web scraping tool or API, a means for saving the collected information as text data, a means for preprocessing the saved text data using a natural language processing library, a means for inputting the preprocessed data into a generative AI model and analyzing it, a means for visualizing factors that lower customer satisfaction based on the analysis results, and a means for planning improvement methods based on the visualization results. This enables efficient collection and analysis of information, and allows specific improvement methods to be planned to improve customer satisfaction. It also reduces unnecessary calls to the call center, making it possible to contribute to companies in terms of both quality and cost.

[0745] A "web scraping tool" is software for automatically extracting data from websites.

[0746] "API" stands for Application Programming Interface, an interface for exchanging data and functions between different software programs.

[0747] "Inquiry history" is a record of inquiries made by customers to your company, and is data including the inquiry content, date and time, customer information, etc.

[0748] "Online information" refers to publicly available data or content that exists on the Internet, including social media posts, blog posts, news articles, etc.

[0749] "Text data" is data expressed as a character string, and is a collection of sentences or words.

[0750] A "natural language processing library" is a software library for analyzing text data and understanding its meaning, providing functions such as tokenization, sentiment analysis, and tagging.

[0751] "Preprocessing" refers to the process of converting data into a format that can be analyzed, and includes data cleaning and tokenization.

[0752] A "generative AI model" is an artificial intelligence model that generates new information or analytical results based on input data, and is used for natural language generation and classification tasks.

[0753] "Analysis" is the process of examining data in detail to find meaning and patterns.

[0754] "Visualization" refers to displaying data or analytical results in a visual format such as a graph or chart, and is a means of making information easier to understand intuitively.

[0755] "Improvement methods" are specific measures or action plans for solving current problems and improving the performance and quality of a system or service.

[0756] This invention is a system in which a server uses a web scraping tool or API to collect inquiry history and online information for the company, and saves the collected information as text data. The saved text data is preprocessed using a natural language processing library and input into a generative AI model. The generative AI model analyzes the input data, visualizes factors that are reducing customer satisfaction based on the analysis results, and proposes methods of improvement.

[0757] Hardware and Software Used

[0758] Hardware

[0759] Server: A computer system for data collection, storage, pre-processing, and analysis.

[0760] software

[0761] Web scraping tools: Python's BeautifulSoup and Scrapy, etc.

[0762] API: Twitter API, Google Custom Search API, etc.

[0763] Database: MySQL, PostgreSQL, etc.

[0764] Cloud storage: Amazon S3, Google Cloud Storage, etc.

[0765] Natural language processing libraries: such as NLTK or spaCy.

[0766] Generative AI models: such as OpenAI's GPT-3 and Google's BERT.

[0767] Examples

[0768] Example 1: Collection and analysis of inquiry history

[0769] 1. The server uses BeautifulSoup to collect past inquiry history from its company's inquiry form.

[0770] 2. The server stores the collected query history in a MySQL database.

[0771] 3. The server pre-processes and tokenizes the stored query history using spaCy.

[0772] 4. The server inputs the preprocessed data into GPT-3 and analyzes trends in the query content.

[0773] Example 2: Collecting and analyzing information on the Internet

[0774] 1. The server uses the Twitter API to collect tweets related to specific keywords.

[0775] 2. The server stores the collected tweets in Amazon S3.

[0776] 3. The server preprocesses the stored tweets using NLTK and performs sentiment analysis.

[0777] 4. The server inputs the preprocessed data into BERT to classify the content of the tweets.

[0778] Examples of prompt statements

[0779] Prompt 1: Analyze the query history

[0780] "Based on your inquiry history over the past year, please tell us the most common inquiries you have made."

[0781] Prompt 2: Collecting and analyzing information on the Internet

[0782] "Collect tweets related to a specific keyword and tell me the percentage of positive and negative tweets."

[0783] In this way, the server can efficiently collect information, analyze it, and devise specific improvement methods to increase customer satisfaction. It can also reduce unnecessary calls to the call center, contributing to companies in terms of both quality and cost.

[0784] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0785] Step 1:

[0786] The server collects information using web scraping tools and APIs. Specifically, it uses Python's BeautifulSoup to scrape inquiry history from websites and Twitter API to collect tweets related to specific keywords. The input is the website URL or API endpoint, and the output is the collected raw data.

[0787] Step 2:

[0788] The server stores the collected information as text data. Specifically, it stores inquiry history and tweets in a MySQL database. The input is the collected raw data, and the output is the text data stored in the database.

[0789] Step 3:

[0790] The server preprocesses the stored text data using a natural language processing library. Specifically, it uses spaCy to clean the text (e.g., remove HTML tags and special characters) and tokenize it. The input is the text data stored in the database, and the output is the preprocessed text data.

[0791] Step 4:

[0792] The server inputs the preprocessed data into the generative AI model for analysis. Specifically, it uses OpenAI's GPT-3 to analyze trends in inquiry content and Google's BERT to classify tweet content. The input is the preprocessed text data, and the output is the analysis result.

[0793] Step 5:

[0794] The server visualizes the factors that are lowering customer satisfaction based on the analysis results. Specifically, it uses a visualization tool to display the analysis results as graphs and charts. The input is the analysis results, and the output is the visualized data.

[0795] Step 6:

[0796] The server then creates improvement methods based on the visualized results. Specifically, it generates a specific action plan to improve customer satisfaction based on the visualized data. The input is the visualized data, and the output is a proposal for improvement methods.

[0797] (Application example 1)

[0798] Next, a description will be given of Application Example 1 of Form 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".

[0799] Conventional customer support systems have had problems with not being able to respond quickly and appropriately to user inquiries, leading to lower customer satisfaction and an increase in unnecessary calls to call centers. In addition, they were unable to effectively utilize online information, making it difficult to link it to service improvements. This resulted in issues such as a decline in the company's quality and efficiency in terms of costs.

[0800] 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.

[0801] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, a means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for integrating the collected information and generating answers to customer inquiries in real time using generative AI, and a means for presenting the generated answers to customer support staff. This enables prompt and appropriate responses to inquiries from users, improving customer satisfaction and reducing unnecessary calls to the call center. Furthermore, by effectively utilizing information on the Internet, it is possible to improve services, improving the quality and cost efficiency of the company.

[0802] "Inquiry history to the company" refers to a record of questions, requests, complaints, etc. that a company has received from customers.

[0803] "Online information" refers to text data such as reviews, social media posts, and blog articles that exist on the Internet.

[0804] "Means of collection" refers to the method of obtaining information using web scraping tools or APIs.

[0805] "Means of analysis" refers to the method of inputting collected data into an AI model and extracting patterns and trends.

[0806] "Planning improvements to our own services" is the process of devising ways to improve services based on the results of analysis.

[0807] A "visualization means" is a method for displaying the analysis results in the form of graphs, charts, etc.

[0808] "Means for planning improvement methods" is a method for formulating specific improvement measures based on visualized data.

[0809] "Generative AI" is an artificial intelligence technology that generates text data using natural language processing.

[0810] "A means of generating answers to customer inquiries in real time" refers to a method of using generative AI to instantly create appropriate answers.

[0811] "Means for presentation to customer support representative" refers to the manner in which the generated answers are displayed to support staff.

[0812] The following system configuration and processing procedure will be described as an embodiment of the present invention.

[0813] System Configuration

[0814] The system consists of the following main components:

[0815] 1. Server: Collects data, analyzes it, and generates answers using generative AI.

[0816] 2. Terminal: The device used by the customer support representative (e.g. PC, smartphone).

[0817] 3. User: The customer making the enquiry.

[0818] Hardware and Software Used

[0819] Hardware: Servers, PCs, smartphones

[0820] Software: Python, BeautifulSoup, requests, OpenAI API

[0821] Data processing and calculation

[0822] 1. Data collection:

[0823] The server uses a web scraping tool (BeautifulSoup) and an API (requests) to collect inquiry history and online information for the company.

[0824] The collected data is saved in text format.

[0825] 2. Data Analysis:

[0826] The server inputs the collected text data into a generative AI (OpenAI API) for analysis.

[0827] The analysis results identify and visualize the factors that are lowering customer satisfaction.

[0828] 3. Answer generation:

[0829] The server uses generative AI to generate answers to customer inquiries in real time.

[0830] The generated answer is presented to the customer support representative's terminal.

[0831] Examples

[0832] For example, if a user asks "Do you have product A in stock?", the server will input the following prompt to the generative AI:

[0833] Analyze user inquiries and generate the best answers. Here are some of the inquiries we collected:

[0834] 1. Do you have product A in stock?

[0835] 2. Please tell me how to return product B.

[0836] 3. I would like to see reviews of product C.

[0837] The generative AI generates an appropriate response based on this prompt text and displays it on the customer support representative's device, enabling the representative to respond quickly and appropriately.

[0838] This system enables prompt and appropriate responses to user inquiries, improving customer satisfaction and reducing unnecessary calls to the call center. In addition, by effectively utilizing information on the Internet, it is possible to improve services, improving the quality and cost efficiency of companies.

[0839] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0840] Step 1:

[0841] The server uses a web scraping tool (BeautifulSoup) to collect inquiry history and related information from a specified website. The website URL is given as input, and the collected text data is obtained as output. Specifically, the server sends an HTTP request and analyzes the obtained HTML data to extract text.

[0842] Step 2:

[0843] The server uses API requests to collect inquiry history and related information from the specified API endpoint. The API URL is given as input, and the collected data in JSON format is obtained as output. In concrete terms, the server sends an API request, analyzes the obtained JSON data, and extracts the required information.

[0844] Step 3:

[0845] The server integrates the text data and JSON data collected in steps 1 and 2. Text data and JSON data are given as input, and the integrated text data is obtained as output. In concrete terms, the server combines the text data and JSON data and saves them as a single text data.

[0846] Step 4:

[0847] The server inputs the integrated text data into the generative AI (OpenAI API) and performs analysis. The integrated text data is given as input, and the analysis results are obtained as output. In concrete terms, the server sends a prompt to the generative AI and receives a response from the AI.

[0848] Step 5:

[0849] The server generates answers to customer inquiries in real time based on the analysis results obtained from the generative AI. The analysis results are given as input, and the generated answers are obtained as output. In concrete terms, the server analyzes the analysis results and generates appropriate answers.

[0850] Step 6:

[0851] The server presents the generated answer to the customer support representative's terminal. The generated answer is given as an input, and the answer displayed on the representative's terminal is obtained as an output. In concrete terms, the server transmits the generated answer to the representative's terminal and displays it.

[0852] Step 7:

[0853] The customer support representative responds to the user's inquiry based on the presented answer. The input is the answer presented by the server, and the output is a response to the user. In concrete terms, the representative checks the presented answer and gives an appropriate response to the user.

[0854] Example 2

[0855] Next, a description will be given of Example 2 of the embodiment 2. In the following description, the data processing device 12 is referred to as a "server" and the smart glasses 214 are referred to as a "terminal".

[0856] With conventional customer satisfaction improvement systems, it was difficult to effectively analyze customer feedback and reviews and extract specific areas for improvement. In addition, the process of visualizing the analysis results and proposing specific improvement methods was time-consuming and laborious, making it difficult to respond quickly. Furthermore, it was not possible to reduce unnecessary calls to the call center, and there was a need to improve efficiency in both quality and cost.

[0857] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0858] In this invention, the server

[0859] A means of collecting inquiry history and online information for our company,

[0860] A means for pre-processing the collected information;

[0861] A means for analyzing the preprocessed information using a generative AI model; and

[0862] A means for extracting factors that are lowering customer satisfaction based on the analysis results;

[0863] A means for visualizing the extracted factors;

[0864] A means to plan improvement methods based on the visualized results;

[0865] This makes it possible to effectively analyze customer feedback and reviews and quickly extract specific areas for improvement. In addition, visualizing the analysis results makes it easier to plan specific improvement methods, which reduces unnecessary calls to the call center, resulting in improved efficiency in both quality and cost.

[0866] "Inquiry history to our company" refers to a record of inquiries made by customers to our company.

[0867] "Online information" refers to text data such as customer feedback, reviews, and comments that exist on the Internet.

[0868] "Means of collection" refers to methods and technologies for automatically obtaining inquiry history and online information.

[0869] "Preprocessing means" refers to methods or techniques for converting collected text data into a format that is easy to analyze.

[0870] A "generative AI model" is an artificial intelligence model that has the ability to learn from large amounts of text data and generate new text.

[0871] "Means of analysis" refers to the methods and techniques used to analyze the pre-processed data using the generative AI model and extract meaningful information.

[0872] "Factors that reduce customer satisfaction" are elements that cause customer dissatisfaction and negative feedback.

[0873] "Means of extraction" refers to methods or techniques for extracting specific factors from the analysis results.

[0874] A "visualization means" is a method or technique for displaying the extracted factors in a visual format, such as a graph or chart.

[0875] "Means for planning improvement methods" are methods and techniques for devising specific improvement measures based on the visualized results.

[0876] The present invention is a system for improving customer satisfaction, in which a server, a terminal, and a user work together. A specific embodiment of this system will be described below.

[0877] 1. System Configuration

[0878] The server has the following main functions:

[0879] A means of collecting inquiry history and online information for your company

[0880] A means of pre-processing the collected information

[0881] A means of analyzing pre-processed information using generative AI models

[0882] A means of extracting factors that are reducing customer satisfaction based on analysis results

[0883] A means to visualize the extracted factors

[0884] A means to plan improvement methods based on visualized results

[0885] 2. Hardware and Software Used

[0886] The server is a computer system equipped with a high-performance processor and a large amount of memory. The following software is used:

[0887] Web scraping tools and APIs for data collection

[0888] Natural language processing libraries for data preprocessing (e.g. NLTK, spaCy)

[0889] Generative AI models (e.g. GPT-3)

[0890] Tools for data visualization (e.g. Matplotlib, D3.js)

[0891] 3. Program Processing

[0892] The server first collects contact history and online information, including website forms, emails, and social media comments, and then preprocesses the collected text data. Specifically, it cleans, tokenizes, and normalizes the text.

[0893] The preprocessed data is input into a generative AI model (GPT-3). This model has the ability to learn from large amounts of text data and generate new text. The generative AI model extracts factors that are lowering customer satisfaction from the input data. Specifically, it identifies negative feedback and dissatisfaction and lists them as factors.

[0894] 4. Specific Examples

[0895] As a concrete example, the following prompt sentence is input into the generative AI model.

[0896] Example prompt:

[0897] "Please extract the factors that are causing low customer satisfaction from the following text data. Text data: 'Poor product quality. Slow delivery. Unfriendly customer service.'"

[0898] The output of the generative AI model is, "The factors that are lowering customer satisfaction are product quality, delivery delays, and unfriendly customer service."

[0899] 5. Visualize and plan ways to improve

[0900] The server visualizes the extracted factors in visual formats such as graphs and charts. Based on the visualized results, specific improvement measures are proposed. For example, improving product quality, improving the delivery system, or customer service training could be considered.

[0901] In this way, the server can extract factors that are lowering customer satisfaction from the information collected using the generative AI model and quickly devise specific improvement measures.

[0902] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0903] Step 1:

[0904] The server collects the company's inquiry history and other online information. As input, it obtains text data such as website forms, emails, and SNS comments. Specifically, the server automatically collects data using web scraping tools and APIs. The output is the collected raw text data.

[0905] Step 2:

[0906] The server preprocesses the collected text data. As input, it uses the raw text data collected in step 1. Specifically, the server uses natural language processing libraries (e.g., NLTK, spaCy) to clean the text (remove unnecessary characters and symbols), tokenize (split into words and phrases), and normalize (convert into a unified format). The output is the preprocessed, clean text data.

[0907] Step 3:

[0908] The server inputs the preprocessed data into a generative AI model (e.g. GPT-3). The clean text data preprocessed in step 2 is used as input. Specifically, the server divides the preprocessed text data into batches and inputs them into the generative AI model. The generative AI model has the ability to learn from large amounts of text data and generate new text. The output is the result of analysis by the generative AI model.

[0909] Step 4:

[0910] The server extracts factors that are reducing customer satisfaction from the output of the generative AI model. As input, it uses the analysis results obtained from the generative AI model in step 3. Specifically, the server analyzes the output of the generative AI model and extracts negative keywords and phrases. The output is a list of factors that are reducing customer satisfaction.

[0911] Step 5:

[0912] The server visualizes the extracted factors. As input, it uses the list of factors extracted in step 4. Specifically, the server uses a data visualization tool (e.g., Matplotlib, D3.js) to display the factors in a visual form such as a graph or chart. The output is a graph or chart of the visualized factors.

[0913] Step 6:

[0914] The server plans improvement methods based on the visualization results. As input, it uses graphs and charts of the factors visualized in step 5. In terms of specific operations, the server devises specific improvement measures based on the visualized data. For example, improving product quality, improving the delivery system, customer service training, etc. are possible. The output is a list of specific improvement methods.

[0915] (Application example 2)

[0916] Next, a description will be given of Application Example 2 of Form 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".

[0917] With conventional customer satisfaction analysis systems, it was difficult to efficiently collect and analyze customer reviews and feedback and identify specific areas for improvement. In addition, there was a lack of a way to visualize the analysis results and quickly notify managers, which meant it took a long time to improve customer satisfaction. Furthermore, there was a lack of specific measures to reduce unnecessary calls to the call center.

[0918] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0919] In this invention, the server includes a means for collecting inquiry history and online information, an AI means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that lower customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for collecting customer reviews and feedback from mail order sites, a means for extracting factors that lower customer satisfaction using generative AI, and a means for notifying the manager of the extracted factors. This makes it possible to quickly identify specific factors that lower customer satisfaction and notify the manager. In addition, it is possible to take specific measures to reduce unnecessary calls to the call center.

[0920] An "inquiry history" is a record of inquiries made by a customer to a company.

[0921] "Online information" refers to text data such as reviews and feedback that exist on the Internet.

[0922] "Means of collection" refers to the methods and technologies used to collect inquiry history and online information.

[0923] "Means of analysis" are the methods and techniques used to analyze collected information and convert it into meaningful data.

[0924] "AI means for proposing improvements" refers to artificial intelligence technology that proposes improvement measures for services based on the results of analysis.

[0925] A "visualization means" is a method or technique for displaying the results of an analysis in a visual format, such as a graph or chart.

[0926] "Means for planning improvement methods" refers to methods and techniques for devising specific improvement measures based on the visualized results.

[0927] An "online shopping site" is a website that sells products and services over the Internet.

[0928] A "customer review" is an evaluation or comment made by a customer about a product or service.

[0929] "Feedback" refers to opinions and impressions provided by customers.

[0930] "Generative AI" is an artificial intelligence technology that has the ability to learn from large amounts of text data and generate new text.

[0931] "Means of extracting factors" refers to methods and techniques that use generative AI to identify the causes of declining customer satisfaction.

[0932] "Means of notifying the administrator" refers to methods or techniques for notifying the administrator of the extracted causes.

[0933] A system for carrying out this invention has the following configuration: The server includes means for collecting the history of inquiries made to the company and information on the internet, AI means for analyzing the collected information and proposing improvements to the company's services, means for visualizing factors that are lowering customer satisfaction based on the analysis results, means for proposing improvement methods based on the visualization results, means for collecting customer reviews and feedback from mail order sites, means for extracting factors that are lowering customer satisfaction using generative AI, and means for notifying an administrator of the extracted factors.

[0934] Hardware and Software Used

[0935] Hardware: Servers, smartphones

[0936] Software: Python, TensorFlow, Transformers library (Hugging Face), Flask (backend), React Native (frontend)

[0937] Data processing and calculation

[0938] Data collection

[0939] The server collects customer reviews and feedback from online shopping sites through the API, allowing for efficient collection of customer opinions and feedback.

[0940] Data Preprocessing

[0941] The collected text data is cleaned and tokenized, which converts the data into a form suitable for analysis.

[0942] Application of generative AI models

[0943] The server uses generative AI (such as GPT-3) to analyze the text data and extract factors that are reducing customer satisfaction. Generative AI has the ability to learn from large amounts of text data and generate new text.

[0944] Viewing the results

[0945] The extracted factors are visualized on the dashboard of a smartphone app, allowing managers to quickly grasp the specific points of dissatisfaction.

[0946] Administrator Notification

[0947] The server notifies the administrator of the extracted cause, allowing the administrator to take prompt action.

[0948] Examples

[0949] For example, by inputting a prompt sentence like the one below into a generative AI model, the AI ​​can analyze the review and extract specific points of dissatisfaction.

[0950] Example prompt:

[0951] "Analyze the following customer reviews to identify factors that are causing low customer satisfaction. Review: 'The product was delivered late and the quality was below expectations.'"

[0952] By inputting this prompt into a generative AI model, the AI ​​can analyze the reviews and extract specific complaints such as "The product was delivered late" or "The quality was below expectations." This allows managers to take swift action to address the complaints, helping to improve customer satisfaction.

[0953] The flow of the specific process in the application example 2 will be described with reference to FIG.

[0954] Step 1:

[0955] The server collects customer reviews and feedback from the online shopping site through an API. The input is the API endpoint of the online shopping site, and the output is the text data of the collected customer reviews and feedback. Specifically, the server periodically calls the API to obtain new reviews and feedback.

[0956] Step 2:

[0957] The server preprocesses the collected text data. The input is the text data collected in step 1, and the output is the cleaned and tokenized data. Specifically, the server removes unnecessary characters and spaces and splits the text into tokens.

[0958] Step 3:

[0959] The server uses a generative AI model (such as GPT-3) to analyze the preprocessed data. The input is the data preprocessed in step 2, and the output is the analysis result including the factors that are reducing customer satisfaction. Specifically, the server inputs a prompt sentence into the generative AI model and obtains the analysis result.

[0960] Step 4:

[0961] The server visualizes the analysis results. The input is the analysis results obtained in step 3, and the output is the data displayed in a visual format such as a graph or chart. Specifically, the server converts the analysis results into graphs and charts for display on the dashboard.

[0962] Step 5:

[0963] The server then proposes an improvement method based on the visualized results. The input is the data visualized in step 4, and the output is a specific improvement measure. Specifically, the server analyzes the visualized data and proposes an improvement measure.

[0964] Step 6:

[0965] The server notifies the administrator of the extracted causes and remedial measures. The input is the remedial measures proposed in step 5, and the output is the notification sent to the administrator. Specifically, the server sends information to the administrator via email or push notification.

[0966] Step 7:

[0967] The administrator receives the notification and takes specific corrective action. The input is the notification received in step 6, and the output is the corrective action taken. Specifically, the administrator makes the necessary changes or actions based on the content of the notification.

[0968] Example 3

[0969] Next, a third embodiment of the third embodiment will be described. 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".

[0970] With conventional customer satisfaction improvement systems, it was difficult to effectively collect and analyze customer inquiries and online evaluation data and quickly propose specific improvement measures. In addition, the process of creating specific action plans based on visualized data was complicated, making it difficult for companies to respond quickly. This led to issues such as insufficient improvement of customer satisfaction and reduction of unnecessary calls to call centers.

[0971] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0972] In this invention, the server includes a means for collecting inquiry history and online information, an artificial intelligence means for analyzing the collected information and proposing improvements to the company's services, a means for visualizing factors lowering customer satisfaction based on the analysis results, a means for proposing improvement methods based on the visualization results, a means for inputting a prompt sentence into the generative AI model to have it propose improvement measures, and a means for creating a specific action plan based on the proposed improvement measures. This makes it possible to effectively collect and analyze customer inquiry contents and online evaluation data, quickly propose specific improvement measures, and create a specific action plan based on the visualized data.

[0973] "Inquiry history to our company" refers to a record of the content, date and time, frequency, etc. of inquiries made by customers to a company.

[0974] "Online information" refers to data such as customer ratings, reviews, and comments that exist on the Internet.

[0975] "Means of collection" refers to the methods and technologies used to obtain the necessary information using databases or APIs.

[0976] "Analytical means" are the methods and techniques used to analyze collected data and extract meaningful information.

[0977] An "artificial intelligence means" is a system that uses techniques such as machine learning and natural language processing to analyze data and gain insights.

[0978] "Visualization means" refers to methods and technologies for displaying analysis results in the form of graphs, dashboards, etc., making them intuitively understandable.

[0979] A "generative AI model" is an artificial intelligence model that is trained to generate appropriate outputs for specific inputs.

[0980] A "prompt" is an instruction entered into a generative AI model to cause it to perform a specific task.

[0981] "Means for proposing improvement measures" refers to methods or technologies that use generative AI models to propose specific improvement measures based on the analysis results.

[0982] "Means for creating an action plan" refers to methods or techniques for formulating a specific action plan based on proposed improvement measures.

[0983] The present invention is directed to a system that aims to improve customer satisfaction and reduce unnecessary calls to a call center. A specific embodiment of this system will be described below.

[0984] First, the server collects the history of inquiries to the company and information on the internet. To do this, a database service or API is used. For example, a general cloud database is used as the database service, and data is obtained from customer inquiry systems and SNS platforms via API.

[0985] Next, the server analyzes the collected data. This analysis is performed using programming languages ​​such as Python and R. Specifically, natural language processing (NLP) libraries (e.g., NLTK and spaCy) are used to analyze the text data and perform frequent keyword and sentiment analysis. Machine learning models are also used to perform data clustering and anomaly detection.

[0986] The analysis results are displayed as dashboards and graphs using visualization tools such as Tableau and Microsoft Power BI. This allows users to view the dashboards on a browser and intuitively understand the analysis results.

[0987] Based on the visualization results, the user inputs a prompt sentence into the generative AI model to suggest improvement measures. For example, the following prompt sentence can be input into the generative AI model:

[0988] "Please suggest improvement measures to increase customer satisfaction based on the content and frequency of inquiries from customers and online evaluation data."

[0989] The server sends this prompt to the generative AI model and receives improvement measures proposed by the AI. The user creates a specific action plan based on the proposed improvement measures. For example, if the AI ​​suggests an improvement measure such as "There are many complaints about a specific product, so we suggest that you conduct additional tests to improve the quality of the product," the user creates and executes a specific test plan based on the proposal.

[0990] In this way, companies can improve customer satisfaction and reduce unnecessary calls to call centers. This system can consistently perform data collection, analysis, visualization, improvement proposals, and action plan creation, enabling quick and effective response. The flow of the identification process in the third embodiment will be described with reference to FIG. 15.

[0991] Step 1: Data collection

[0992] The server collects the company's inquiry history and online information. As input, it obtains data from customer inquiry systems and social media platforms. Specifically, the server obtains data from these platforms via APIs and stores it in a database. As output, the content of inquiries, their frequency, and online evaluation data are stored in the database.

[0993] Step 2: Data analysis

[0994] The server analyzes the collected data. As input, it uses inquiry content stored in a database and online evaluation data. Specifically, the server uses a natural language processing (NLP) library (e.g., NLTK or spaCy) in Python or R to analyze the text data. This provides the results of frequent keywords and sentiment analysis. It also uses machine learning models to perform data clustering and anomaly detection. The analysis results are obtained as output.

[0995] Step 3: Data visualization

[0996] The server visualizes the analysis results. It uses the results of the data analysis as input. Specifically, the server exports the analysis results in JSON format and imports them into a visualization tool (e.g. Tableau or Microsoft Power BI), which displays the analysis results as dashboards and graphs. As output, it provides the visualized data.

[0997] Step 4: Propose improvements

[0998] Based on the visualization results, the user inputs a prompt sentence into the generative AI model to have it suggest improvements. The visualized data and the prompt sentence are used as input. In concrete terms, the user inputs the following prompt sentence into the generative AI model: "Please suggest improvements to improve customer satisfaction based on the content and frequency of customer inquiries, as well as online evaluation data." The server sends this prompt sentence to the generative AI model and receives the improvements suggested by the AI. The output is the improvements suggested by the AI.

[0999] Step 5: Create an action plan

[1000] The user creates a specific action plan based on the proposed improvement measures. The improvement measures proposed by the AI ​​are used as input. In concrete terms, the user consults with relevant departments within the company based on the proposed improvement measures (e.g., "There are many complaints about a particular product, so we propose that additional testing be carried out to improve the quality of the product") and formulates a specific action plan. A specific action plan is obtained as output.

[1001] In this way, companies can improve customer satisfaction and reduce unnecessary calls to their call centers.

[1002] (Application example 3)

[1003] Next, a description will be given of Application Example 3 of Form Example 3. 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".

[1004] With conventional customer support systems, it was difficult to effectively analyze and visualize customer inquiries and online reviews. In addition, there was a lack of means to plan specific improvement measures based on the analysis results, and it was not possible to fully improve customer satisfaction or reduce unnecessary calls to the call center. This caused companies to face challenges in both quality and cost.

[1005] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1006] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, an AI means for analyzing the collected information and proposing improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for proposing improvement methods based on the visualization results, a means for proposing improvement measures based on specific inquiry contents using generative AI, and a means for displaying the proposed improvement measures. This makes it possible to effectively analyze and visualize the contents of inquiries from customers and online evaluations. In addition, specific improvement measures can be proposed based on the analysis results, which can improve customer satisfaction and reduce unnecessary calls to the call center.

[1007] "Inquiry history" refers to a record of customer inquiries, including their content, frequency, and date and time.

[1008] "Online information" refers to data such as customer ratings, reviews, and comments that exist on the Internet.

[1009] "Collection means" refers to methods or devices for acquiring inquiry history and online information.

[1010] An "analysis means" is a method or device for analyzing collected information and converting it into meaningful data.

[1011] A "visualization means" is a method or device that displays analysis results in the form of graphs, dashboards, etc., to enable intuitive understanding.

[1012] An "improvement planning means" is a method or device for devising improvement measures for a service based on the visualization results.

[1013] "Generative AI" is a type of artificial intelligence, a model that has the ability to generate appropriate outputs for specific inputs.

[1014] A "proposal means" is a method or device for using generative AI to propose improvement measures based on specific inquiry content.

[1015] A "display means" is a method or device for showing the suggested improvement to the user.

[1016] A system for implementing this invention includes a server, a user terminal, and a generative AI model. The server collects inquiry history and online information and analyzes this data. The analysis results are visualized and displayed in a format that the user can intuitively understand. Furthermore, the generative AI model is used to propose improvement measures based on the specific inquiry content and display them on the user terminal.

[1017] The server uses Python and Dash to build a dashboard to visualize the types of inquiries and their frequency. When a user selects a specific inquiry from a drop-down menu, the data related to that inquiry is displayed as a bar graph, allowing the user to see at a glance which inquiries are occurring frequently.

[1018] In addition, the server uses the OpenAI API to generate improvement measures suggested by the generative AI model based on the selected inquiry content. For example, if the user selects "Product defect," the server inputs the following prompt sentence to the generative AI model:

[1019] Example prompt:

[1020] If the customer inquiry is about a "defective product," please provide suggestions for improving the service.

[1021] Based on this prompt, the generative AI model generates "suggestions for reducing product defects," proposing specific improvement measures such as "strengthening quality control" and "simplifying return procedures." These suggestions are displayed on the user's device, allowing the user to use them as a reference when creating a specific action plan.

[1022] This system makes it possible to effectively analyze and visualize customer inquiries and online reviews, and also to devise specific improvement measures based on the analysis results, thereby improving customer satisfaction and reducing unnecessary calls to the call center.

[1023] The flow of the specific process in the application example 3 will be described with reference to FIG.

[1024] Step 1:

[1025] The server collects inquiry history and online information.

[1026] Input: Customer inquiries, online ratings and reviews

[1027] Data processing: Save inquiry details and evaluations in a database

[1028] Output: Collected dataset

[1029] Step 2:

[1030] The server analyzes the collected data.

[1031] Input: Collected dataset

[1032] Data calculations: data cleansing, classification, frequency analysis

[1033] Output: Analysis results (frequency of inquiries, evaluation trends, etc.)

[1034] Step 3:

[1035] The server visualizes the analysis results.

[1036] Input: Analysis results

[1037] Data processing: generating graphs and dashboards

[1038] Output: Visualized data (bar graphs, dashboards)

[1039] Step 4:

[1040] The user reviews the visualized data and selects a particular query.

[1041] Input: visualized data, user selection

[1042] Data processing: Extraction of selected inquiry details

[1043] Output: Selected inquiry details

[1044] Step 5:

[1045] The server uses a generative AI model to suggest improvement measures based on the selected inquiry content.

[1046] Input: Selected inquiry content, prompt text

[1047] Data computation: Proposal generation using generative AI models

[1048] Output: Suggested remedy

[1049] Step 6:

[1050] The server displays the proposed remedy on the user terminal.

[1051] Input: Proposed remedy

[1052] Data processing: Formatting the proposal content

[1053] Output: Remediation measures displayed on the user's terminal

[1054] Step 7:

[1055] The user checks the proposed improvements and creates a specific action plan.

[1056] Input: Proposed remedy

[1057] Data processing: Creating an action plan

[1058] Output: A concrete action plan

[1059] In addition, 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.

[1060] "Example 1"

[1061] One embodiment of the present invention is a system that incorporates an emotion engine. This system collects inquiry history and online information for the company, and analyzes that information. The analysis is performed by both the AI ​​means and the emotion engine. The AI ​​means understands the content of the information and plans improvements to the service. Meanwhile, the emotion engine analyzes the emotions of the user. For example, the emotional state of the user is estimated from the wording and tone used when the user makes an inquiry. This makes it possible to grasp not only the user's satisfaction level, but also the emotions behind it.

[1062] "Example 2"

[1063] The emotion analysis results obtained from the emotion engine are provided to the AI ​​means. The AI ​​means uses these results to plan improvements to the service. For example, if anger is detected from many users, the cause is identified and a remedial measure is proposed. This makes it possible to improve the service in response to the user's emotions, not just the content of the inquiry.

[1064] "Example 3"

[1065] The system of the present invention also includes a means for visualizing the analysis results. This makes it possible to grasp at a glance the factors that are lowering customer satisfaction and the emotional state of users. For example, the dashboard displays the most frequently inquired items and the distribution of emotional states. This allows companies to quickly grasp problems and devise appropriate improvement measures.

[1066] "Example 4"

[1067] The system of the present invention also includes a means for proposing improvement methods based on the visualization results. This makes it possible to create specific action plans and improve customer satisfaction. For example, based on information obtained from the emotion engine, improvement measures such as providing new services that resolve user dissatisfaction can be proposed. This not only improves user satisfaction, but also contributes to the company's quality and cost aspects, such as reducing unnecessary calls to call centers.

[1068] The process flow of each embodiment will be described below.

[1069] "Example 1"

[1070] Step 1: The system collects inquiry history and online information for the company.

[1071] Step 2: The collected information is analyzed by both AI means and the emotion engine.

[1072] Step 3: The emotion engine analyzes the user's emotions and provides the results to the AI ​​means.

[1073] "Example 2"

[1074] Step 1: The AI ​​means plans service improvements based on the emotion analysis results provided by the emotion engine.

[1075] Step 2: If anger is detected among many users, identify the cause and develop measures to improve it.

[1076] "Example 3"

[1077] Step 1: The system visualizes the analysis results.

[1078] Step 2: Display the most frequently inquired items and distribution of emotional states on the dashboard.

[1079] "Example 4"

[1080] Step 1: The system plans improvement measures based on the visualized results.

[1081] Step 2: Create a concrete action plan to improve customer satisfaction.

[1082] Step 3: Based on the information obtained from the emotion engine, develop improvement measures, such as providing new services to resolve user dissatisfaction.

[1083] Example 1

[1084] Next, a description will be given of Example 1 of Form 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".

[1085] With the conventional system, it was difficult to not only collect the history of inquiries to the company and information on the Internet, but also to effectively analyze that information and understand the emotional state of the customer. It was also difficult to devise specific service improvement measures based on the analysis results. Furthermore, there was a lack of a process to visualize the factors that were lowering customer satisfaction and to devise improvement methods based on that. As a result, the system was unable to fully contribute to the company's quality improvement and cost reduction.

[1086] 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.

[1087] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, a means for saving the collected information as text data, an artificial intelligence means for analyzing the saved text data, an emotion engine means for analyzing the emotional state of the customer based on the analysis result, a means for planning improvements to the company's services based on the analysis result, a means for visualizing factors that lower customer satisfaction based on the analysis result, and a means for planning improvement methods based on the visualization result. This makes it possible to effectively analyze the collected information, grasp the emotional state of the customer, and plan specific service improvement measures. In addition, by visualizing factors that lower customer satisfaction and planning improvement methods based on the visualization, it becomes possible to contribute to improving the quality of the company and reducing costs.

[1088] "Inquiry history to our company" refers to the content and history of inquiries that a company has received from customers.

[1089] "Online information" refers to information that exists on the Internet, such as reviews, comments, blog posts, and social media posts.

[1090] "Means of collection" refers to methods of obtaining information using web scraping tools, APIs, etc.

[1091] "Text data" refers to information stored in the form of strings of characters.

[1092] "Means of storage" refers to the way information is stored, such as in a database or file system.

[1093] "Artificial intelligence means" refers to methods of analyzing data using machine learning models or natural language processing models.

[1094] "Emotion engine means" refers to algorithms or tools for analyzing emotions from text data.

[1095] "Means for proposing improvements to services" refers to methods for proposing specific improvement measures based on the results of analysis.

[1096] "Means of visualizing factors that are reducing customer satisfaction" refers to methods of displaying the analysis results in a visual format such as graphs or charts.

[1097] "Means for planning improvement methods" refers to a method for formulating specific improvement measures based on visualized data.

[1098] The present invention is a system for collecting inquiry histories and information on the Internet and analyzing the collected information to improve services. A specific embodiment of this system will be described below.

[1099] Information gathering

[1100] The server collects information using web scraping tools and APIs. Specifically, it uses web scraping tools such as BeautifulSoup and Scrapy, or the Twitter API and Google Custom Search API. For example, the server performs a task such as "collect the latest tweets about the service."

[1101] Data Storage

[1102] The server stores the collected information as text data. Specifically, it stores it in a database such as MySQL or MongoDB. For example, the server performs the operation "save the collected tweets in MongoDB."

[1103] 4. Data Analysis

[1104] The server inputs the stored text data into an AI model and performs analysis. Specifically, it uses natural language processing models such as BERT and GPT-3. For example, the server executes the process of "inputting the stored tweets into the BERT model and extracting problems related to the service."

[1105] Emotion analysis

[1106] The server uses an emotion engine to analyze the user's emotions. Specifically, it uses IBM Watson Tone Analyzer and Microsoft Azure Text Analytics. For example, the server performs a task such as "analyzing the user's emotions from the content of tweets and detecting dissatisfaction or anger."

[1107] Proposal for service improvement

[1108] The server will propose service improvements based on the analysis results of the AI ​​model. Specifically, it will analyze user inquiries and online feedback to extract areas for improvement. For example, the server will perform a process such as "proposing improvement measures for service delays based on the analysis results."

[1109] Understanding emotional states

[1110] The server grasps the user's emotional state based on the analysis results of the emotion engine. Specifically, it estimates the user's emotional state from the language and tone used when the user makes an inquiry. For example, the server executes a task such as "evaluating the user's satisfaction or dissatisfaction level based on the content of the inquiry."

[1111] Examples

[1112] For example, if a user inquires that "the service is slow," the server collects the content of the inquiry and saves it as text data. The saved data is then input into an AI model to analyze the specific issues related to the service delay. At the same time, an emotion engine is used to analyze the user's emotions, such as dissatisfaction or anger. This allows the server to understand not only the user's satisfaction level, but also the emotions behind it.

[1113] Examples of prompt statements

[1114] "Generate prompts to infer a user's emotional state from the language and tone used when making a query."

[1115] In this way, it is possible to explain a specific embodiment of the system while clarifying the roles of the server, the terminal, and the user.

[1116] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1117] Step 1: Gather information

[1118] The server collects information using web scraping tools and APIs. Specifically, it uses BeautifulSoup or Scrapy to extract data from websites. It also uses the Twitter API and Google Custom Search API to obtain information on the Internet. As input, a specific website URL or API endpoint is given. As output, the collected text data is obtained. For example, the server collects the latest reviews from "https: / / example.com / reviews".

[1119] Step 2: Save Data

[1120] The server saves the collected information as text data. Specifically, it saves it in a database such as MySQL or MongoDB. The collected text data is given as input. The data saved in the database is obtained as output. For example, the server performs the operation "save the collected reviews in the 'reviews' collection in MongoDB."

[1121] Step 3: Data analysis

[1122] The server inputs the stored text data into an AI model and performs analysis. Specifically, it uses natural language processing models such as BERT and GPT-3. The text data stored in the database is given as input. The analysis results are obtained as output. For example, the server executes the process of "inputting the stored reviews into the BERT model and extracting problems with the service."

[1123] Step 4: Sentiment Analysis

[1124] The server uses an emotion engine to analyze the user's emotions. Specifically, it uses IBM Watson Tone Analyzer or Microsoft Azure Text Analytics. Analyzed text data is provided as input. The emotion analysis results are obtained as output. For example, the server executes a task such as "analyzing the user's emotions from the content of the review and detecting dissatisfaction or anger."

[1125] Step 5: Plan service improvements

[1126] The server proposes service improvements based on the analysis results of the AI ​​model. Specifically, it analyzes user inquiries and online feedback to extract areas for improvement. The analysis results and sentiment analysis results are given as input. Specific improvement measures are obtained as output. For example, the server might carry out the process of "proposing improvement measures for service delays based on the analysis results."

[1127] Step 6: Understand your emotional state

[1128] The server grasps the user's emotional state based on the analysis results from the emotion engine. Specifically, the server estimates the user's emotional state from the language and tone used when the user makes an inquiry. The inquiry content and the emotion analysis results are given as input. The user's emotional state is obtained as output. For example, the server executes a task such as "evaluating the user's satisfaction or dissatisfaction level based on the inquiry content."

[1129] (Application example 1)

[1130] Next, a description will be given of Application Example 1 of Form 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".

[1131] Conventional customer support systems can collect and analyze user inquiries and reviews, but they have the problem that it is difficult to grasp the user's emotional state in real time. Even if the analysis results are visualized, it is difficult to quickly come up with specific improvement methods. This leads to a decrease in customer satisfaction and an increase in unnecessary calls to call centers, which has a negative impact on the quality and costs of companies.

[1132] 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.

[1133] In this invention, the server includes a means for collecting inquiry history and online information, an AI means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for analyzing the emotional state of the user using an emotion engine, and a means for displaying the analysis results in real time. This makes it possible to grasp the emotional state of the user in real time and quickly plan specific improvement methods.

[1134] "Inquiry history to our company" refers to data that records the inquiries received by the company from customers.

[1135] "Online information" refers to text data such as reviews, comments, blog posts, forum posts, etc. that exist on the Internet.

[1136] "Means of collection" refers to the method of obtaining information using web scraping tools or APIs.

[1137] "Analyzing" is the process of analyzing collected data to extract useful information.

[1138] "AI means for proposing improvements to our own services" is a system that uses artificial intelligence to propose ways to improve services.

[1139] A "means for visualizing factors that are reducing customer satisfaction" is a method for displaying the results of data analysis in a visual format such as a graph or chart.

[1140] "Means for planning improvement methods" is the process of formulating specific improvement measures based on visualized data.

[1141] An "emotion engine" is software for analyzing a user's emotional state from text data.

[1142] The "means for analyzing the user's emotional state" is a method for analyzing the user's emotions using an emotion engine.

[1143] "Means for displaying the analysis results in real time" refers to a system for instantly displaying the analyzed data to the user.

[1144] A system for carrying out this invention has the following configuration: The server includes means for collecting the history of inquiries made to the company and information on the internet, AI means for analyzing the collected information and proposing improvements to the company's services, means for visualizing factors that are lowering customer satisfaction based on the analysis results, means for proposing improvement methods based on the visualization results, means for analyzing the emotional state of the user using an emotion engine, and means for displaying the analysis results in real time.

[1145] Hardware and Software Used

[1146] Hardware: Servers, smartphones

[1147] Software: Python, BeautifulSoup, requests, OpenAI API, TextBlob

[1148] Data processing and calculation

[1149] 1. Data collection: The server uses web scraping tools (BeautifulSoup, requests) and APIs to collect inquiry history and online information for the company. This information is saved as text data.

[1150] 2. Information analysis: The server analyzes the collected text data using AI means (OpenAI API). A generative AI model is used to understand the content of the information and to plan service improvements.

[1151] 3. Emotion Analysis: The server uses an emotion engine (TextBlob) to analyze the user's emotional state, for example by inferring the user's emotional state from the wording and tone of the query.

[1152] 4. Visualization: The server displays the analysis results in a visual format such as graphs and charts, allowing users to intuitively understand the factors that are causing low customer satisfaction.

[1153] 5. Real-time Display: The server provides immediate feedback to the user by providing a means for displaying the analysis results in real-time.

[1154] Examples

[1155] For example, if a user reviews "This product was very good, but the delivery was slow," the server collects this text data and analyzes it using AI means. The analysis results provide insights such as "Satisfaction with the product is high, but dissatisfaction with the delivery" The sentiment analysis reveals a mix of positive and negative emotions. These analysis results are visualized in graphs and charts and displayed in real time.

[1156] Examples of prompt statements

[1157] Analyze the following text and provide insights: "This product was very good, but the delivery was slow."

[1158] In this way, it is possible to grasp the user's emotional state and level of satisfaction in real time and devise quick and specific methods of improvement.

[1159] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1160] Step 1:

[1161] The server uses web scraping tools (BeautifulSoup, requests) and APIs to collect the company's inquiry history and other online information. The input is the website URL and API endpoint, and the output is text data. Specifically, the server accesses the specified URL, retrieves the HTML content, and analyzes it. As a result of the analysis, text data is extracted and stored in a database.

[1162] Step 2:

[1163] The server analyzes the collected text data using AI means (OpenAI API). The input is the text data collected in step 1, and the output is the analysis result. Specifically, the server sends the text data to a generative AI model and performs analysis using a prompt sentence. As an analysis result, it obtains insight into the content of the text data.

[1164] Step 3:

[1165] The server uses an emotion engine (TextBlob) to analyze the user's emotional state. The input is the text data collected in step 1, and the output is the emotion analysis result. Specifically, the server inputs the text data into the emotion engine and analyzes the emotional state into positive, negative, and neutral. As the analysis result, the user's emotional state is quantified and output.

[1166] Step 4:

[1167] The server displays the analysis results in a visual format such as a graph or chart. The input is the analysis results obtained in step 2 and step 3, and the output is visualized data. Specifically, the server inputs the analysis results into a data visualization tool to generate graphs and charts. The generated visualized data is displayed in a format that the user can intuitively understand.

[1168] Step 5:

[1169] The server provides immediate feedback to the user using a means for displaying the analysis results in real time. The input is the data visualized in step 4, and the output is the feedback displayed in real time. Specifically, the server transmits the visualized data to the user's terminal and displays it in real time. This allows the user to immediately check the analysis results and take any necessary action.

[1170] Example 2

[1171] Next, a description will be given of Example 2 of the embodiment 2. In the following description, the data processing device 12 is referred to as a "server" and the smart glasses 214 are referred to as a "terminal".

[1172] In the conventional system, it was difficult to effectively analyze the information and plan specific service improvement measures by simply collecting inquiries and feedback from users. In addition, since service improvements were not made based on the user's feelings, there was a problem that customer satisfaction was not sufficiently improved. Furthermore, there were many unnecessary calls to the call center, which had a negative impact on the company's quality and costs.

[1173] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1174] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, a means for sentiment analysis of the collected information, a means for providing the sentiment analysis result to a generative AI, a means for analyzing the information using the generative AI and extracting factors that are lowering customer satisfaction, a means for planning service improvement measures based on the extracted factors, and a means for feeding back the planned improvement measures to the user. This makes it possible to plan specific service improvement measures based on the user's sentiment, thereby improving customer satisfaction and reducing unnecessary calls to the call center.

[1175] "Inquiry history" is a record of inquiries made by a user to a company.

[1176] "Online information" refers to information such as user feedback, reviews, and comments that exist on the Internet.

[1177] "Sentiment analysis" is the process of analyzing user emotions (anger, joy, sadness, etc.) from text data.

[1178] "Generative AI" is artificial intelligence that has the ability to learn from large amounts of text data and generate new text.

[1179] "Information analysis" is the process of analyzing collected data and extracting useful information.

[1180] "Customer satisfaction" refers to the level of satisfaction that users feel toward a company's services or products.

[1181] "Cause extraction" is the process of identifying the causes of lower customer satisfaction.

[1182] "Service improvement measures" are specific methods of improvement implemented to enhance customer satisfaction.

[1183] "Feedback" refers to information or notification of improvements provided to a user.

[1184] A "call center" is a telephone response center for accepting inquiries and support from users.

[1185] "Unnecessary calls" are unnecessary phone inquiries due to resolved issues or lack of information.

[1186] "Quality" is a characteristic of a product or service that relates to its performance, reliability, and user satisfaction.

[1187] "Cost" refers to the cost incurred in providing a product or service.

[1188] The present invention is a system for collecting inquiries from users and information on the Internet, analyzing the collected information, and formulating measures to improve services. A specific embodiment of this system will be described below.

[1189] The server first collects the user's inquiry history and online information. This information is stored in a database in text format. For example, if a user sends an inquiry such as "I haven't received my product," the content of that inquiry is stored in the database.

[1190] Next, the server sends the collected text data to an emotion analysis engine to analyze the user's emotions. The emotion analysis engine detects emotions such as "anger," "joy," and "sadness" from the text data. For example, it detects the emotion of "anger" from the text "The product has not arrived."

[1191] The server provides the results of the emotion analysis to the generative AI. The generative AI uses a natural language processing model, such as OpenAI's GPT-3. The generative AI performs information analysis based on the emotion analysis results and extracts factors that are lowering customer satisfaction. For example, it identifies that "delayed delivery" is the cause of the emotion "anger."

[1192] Next, the generative AI proposes specific service improvement measures for the extracted factors, such as "changing the delivery company" or "reviewing the delivery process."

[1193] Finally, the server feeds back the improvements obtained from the generative AI to the user, for example, sending feedback to the user such as "The delivery company has been changed."

[1194] As a concrete example, consider a situation where many users of an online shopping site are dissatisfied with "slow delivery." If the emotion engine detects a lot of "anger" emotions, the server uses generative AI to identify the "cause of the delivery delay" and proposes improvement measures such as "changing the delivery company" or "reviewing the delivery process."

[1195] An example of a prompt sentence might be:

[1196] “If many users are angry about delivery delays, what remedial measures can you suggest?”

[1197] In this way, the server can utilize generative AI to improve services based on user emotions.

[1198] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1199] Step 1: Gather information

[1200] The server collects user inquiry history and online information.

[1201] Input: User inquiry emails, chat logs, online reviews and comments.

[1202] Data processing: The collected information is stored in a database in text format.

[1203] Output: Text-formatted inquiry history and online information are saved in a database.

[1204] Specific operation: The server automatically collects inquiry emails and chat logs sent by users and stores them in a database. For example, if a user sends an inquiry such as "I haven't received my product," the content of the inquiry is stored in the database.

[1205] Step 2: Sentiment Analysis

[1206] The server sends the collected text data to an emotion analysis engine to analyze the user's emotions.

[1207] Input: Text-based inquiry history stored in the database and online information.

[1208] Data calculation: The sentiment analysis engine analyzes the user's emotions (anger, joy, sadness, etc.) from the text data.

[1209] Output: Sentiment analysis results (e.g. emotion tags such as "anger", "joy", "sadness" etc.).

[1210] Specific operation: The server inputs the text data into a sentiment analysis engine and analyzes the user's sentiment. For example, it detects the sentiment of "anger" from the text "The product has not arrived."

[1211] Step 3: Provide information

[1212] The server provides the emotion analysis results to the generative AI.

[1213] Input: Sentiment analysis results.

[1214] Data processing: Converting the emotion analysis results into a format for sending to the generative AI.

[1215] Output: Sentiment analysis results provided to the generative AI.

[1216] Specific operation: The server sends the analysis results obtained from the emotion engine to the generative AI. For example, it provides text data in which the emotion "anger" has been detected to the generative AI.

[1217] Step 4: Factor extraction

[1218] Based on the emotion analysis results provided, the generative AI extracts the factors that are lowering customer satisfaction.

[1219] Input: Sentiment analysis results.

[1220] Data Computation: Generative AI analyzes sentiment analysis results to identify the root cause of problems.

[1221] Output: Factors that are causing low customer satisfaction (e.g. "delayed delivery").

[1222] How it works: The generative AI analyzes the results of emotion analysis and identifies the root cause of the problem. For example, it identifies that "delayed delivery" is the cause of the emotion "anger."

[1223] Step 5: Plan improvements

[1224] The generative AI then devise improvement measures for the extracted factors.

[1225] Input: Factors that are reducing customer satisfaction.

[1226] Data Computation: Generative AI generates specific solutions to identified problems.

[1227] Output: Service improvement measures (e.g., "change delivery company" or "review delivery process").

[1228] Specific behavior: Generative AI generates specific improvement measures for the identified problems. For example, it proposes improvement measures such as "changing the delivery company" or "reviewing the delivery process."

[1229] Step 6: Feedback on results

[1230] The server provides feedback to the user on improvements obtained from the generative AI.

[1231] Input: Service improvement measures from generative AI.

[1232] Data processing: Convert the improvement measures into a format that can be communicated to users.

[1233] Output: Remediation measures communicated to the user.

[1234] Specific operation: The server notifies the user of the improvement measures obtained from the generative AI. For example, it sends feedback to the user such as "The delivery company has been changed."

[1235] (Application example 2)

[1236] Next, a description will be given of Application Example 2 of Form 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".

[1237] Conventional customer satisfaction improvement systems only collect and analyze inquiry history and online information, and are unable to provide specific improvement suggestions based on user emotions. This makes it difficult to effectively improve customer satisfaction, and there is a lack of specific measures to alleviate the dissatisfaction of users who have negative emotions. In addition, it is not possible to reduce unnecessary calls to the call center, and companies are required to improve their quality and cost efficiency.

[1238] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1239] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, an AI means for analyzing the collected information and proposing improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for proposing improvement methods based on the visualization results, a means for collecting feedback and reviews from users and performing sentiment analysis, and a means for extracting feedback with negative sentiments and generating improvement proposals. This makes it possible to make specific improvement proposals based on the user's sentiments, and to effectively improve customer satisfaction. In addition, it is possible to reduce unnecessary calls to the call center, and to realize efficiency in terms of quality and cost for the company.

[1240] An "inquiry history" is a record of inquiries made by a customer to a company.

[1241] "Online information" refers to data such as reviews, comments, blog articles, and social media posts that exist on the Internet.

[1242] "Means of collection" refers to software or hardware mechanisms for automatically collecting inquiry history and online information.

[1243] "Means of analysis" are the algorithms and tools used to analyze collected information and convert it into meaningful data.

[1244] "AI means" is a system that uses artificial intelligence technology to analyze information and make suggestions for improving services.

[1245] A "visualization tool" is any tool or software that allows the analysis results to be displayed in a visual format, such as a graph or chart.

[1246] "Means for planning improvement methods" refers to processes and tools for devising specific improvement measures based on the visualized analysis results.

[1247] "Feedback" means opinions or thoughts provided by a user.

[1248] A "review" is a piece of writing or comment in which a user evaluates a product or service.

[1249] "Sentiment analysis" is a technology that automatically determines a user's emotions from text data.

[1250] "Negative emotions" refers to negative feelings such as anger, frustration, and sadness.

[1251] A "means for generating improvement suggestions" is an algorithm or tool for automatically generating specific improvement measures based on feedback with negative sentiment.

[1252] A system for carrying out the present invention has the following configuration: The server includes means for collecting inquiry history and online information, AI means for analyzing the collected information and proposing improvements to the company's services, means for visualizing factors that are lowering customer satisfaction based on the analysis results, means for proposing improvement methods based on the visualization results, means for collecting feedback and reviews from users and performing sentiment analysis, and means for extracting feedback with negative sentiment and generating improvement proposals.

[1253] Hardware and Software Used

[1254] Hardware: Servers, user devices (smartphones, PCs, etc.)

[1255] Software: Python, OpenAI API, TextBlob

[1256] Data processing and calculation

[1257] 1. Collecting inquiry history and online information:

[1258] The server automatically collects inquiry history and online information using web scraping tools and APIs.

[1259] 2. Analysis of information:

[1260] The server uses AI means to analyze the collected information. Specifically, it uses a generative AI (generative pretrained transformer) to analyze the collected text data and extract factors that are lowering customer satisfaction.

[1261] 3. Emotion analysis:

[1262] The server uses TextBlob to parse the user feedback / reviews for sentiment and extracts the feedback if a negative sentiment (e.g. anger, dissatisfaction, sadness, etc.) is detected.

[1263] 4. Generate improvement suggestions:

[1264] The server uses the OpenAI API to generate specific improvement suggestions based on the negative sentiment feedback, which are then displayed in the form of graphs and charts using visualization tools.

[1265] Examples

[1266] For example, suppose a user gives the following feedback:

[1267] "The delivery of the product was late."

[1268] "The customer service was poor."

[1269] "The item was damaged."

[1270] The server collects these feedbacks, performs sentiment analysis using TextBlob, extracts feedback where negative sentiment is detected, and generates improvement suggestions using the OpenAI API, such as:

[1271] Examples of prompt statements

[1272] Please provide your suggestions for improving our service based on your feedback below:

[1273] Delivery of the product was late.

[1274] Customer service was poor.

[1275] The product was damaged.

[1276] By inputting this prompt text into the OpenAI API, specific improvement suggestions can be obtained. The generated improvement suggestions are displayed in the form of graphs and charts using a visualization tool, allowing companies to plan specific improvement measures based on this.

[1277] The flow of the identification process in the application example 2 will be described with reference to FIG.

[1278] Step 1:

[1279] The server collects the company's inquiry history and online information. Specifically, it uses web scraping tools and APIs to collect data such as inquiry history, reviews, comments, and SNS posts. The input is the inquiry history and online information, and the output is the collected data.

[1280] Step 2:

[1281] The server analyzes the collected information. Specifically, it uses a generative AI (Generative Pretrained Transformer) to analyze the collected text data and extract the factors that are reducing customer satisfaction. The input is the collected data, and the output is the analysis results.

[1282] Step 3:

[1283] The server visualizes the factors that are lowering customer satisfaction based on the analysis results. Specifically, it uses a visualization tool to display the analysis results in the form of graphs and charts. The input is the analysis results, and the output is the visualized data.

[1284] Step 4:

[1285] The server plans improvement methods based on the visualized results. Specifically, it uses processes and tools to devise improvement measures based on the visualized data. The input is the visualized data, and the output is the improvement methods.

[1286] Step 5:

[1287] The server collects feedback and reviews from users and performs sentiment analysis. Specifically, it uses TextBlob to analyze the sentiment of feedback and reviews. The input is the feedback and reviews from users, and the output is the sentiment analysis result.

[1288] Step 6:

[1289] The server extracts feedback with negative emotions. Specifically, it extracts feedback in which negative emotions (e.g., anger, dissatisfaction, sadness, etc.) are detected from the emotion analysis results. The input is the emotion analysis results, and the output is negative feedback.

[1290] Step 7:

[1291] The server generates improvement suggestions based on feedback with negative sentiment. Specifically, it uses the OpenAI API to generate specific improvement suggestions based on negative feedback. The input is the negative feedback, and the output is the improvement suggestions.

[1292] Step 8:

[1293] The server visualizes the generated improvement proposals. Specifically, it uses a visualization tool to display the improvement proposals in the form of graphs and charts. The input is the improvement proposals, and the output is the visualized improvement proposals.

[1294] Example 3

[1295] Next, a third embodiment of the third embodiment will be described. 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".

[1296] With the conventional system, it was difficult to effectively collect and analyze customer inquiries and online evaluations, visualize them, and propose improvement measures.In addition, there was a lack of means to create specific action plans, so it was not possible to fully improve customer satisfaction or reduce unnecessary calls to the call center.

[1297] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1298] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, a means for analyzing the collected information and identifying the emotional state of customers and items of frequent inquiries using natural language processing technology, a means for visualizing the analysis results using a dashboard and graphs, a means for proposing improvement methods using a generative AI model based on the visualized results, and a means for creating a specific action plan based on the proposed improvement methods. This enables a company to quickly plan and implement specific improvement measures to improve customer satisfaction.

[1299] "Inquiry history" refers to a record of the content, date and time, frequency, etc. of inquiries made by a customer to a company.

[1300] "Online information" refers to data such as customer ratings, comments, and reviews that exist on the Internet.

[1301] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[1302] "Emotional state" refers to the emotional state of a customer when making an inquiry or evaluation, and includes emotions such as "satisfied" and "dissatisfied."

[1303] A "dashboard" is an interface for visually displaying data, using graphs and charts to allow information to be grasped at a glance.

[1304] A "generative AI model" is an artificial intelligence model that generates new information or suggestions based on input data.

[1305] A "prompt" is an instruction or question input to a generative AI model that causes the AI ​​to generate a response or suggestion based on that instruction.

[1306] An "action plan" is a specific plan or procedure to be followed to achieve a particular goal.

[1307] This invention is a system for enabling a company to improve customer satisfaction, and is realized by a server, a terminal, and a user fulfilling their respective roles.

[1308] Data collection

[1309] The server collects customer inquiry history and online information from the company's CRM system and SNS API. For example, the content and frequency of customer inquiries are obtained from the CRM system, and evaluation comments from Twitter and Facebook are collected from the SNS API. This data is stored in a database.

[1310] 4. Data Analysis

[1311] The server analyzes the collected data. Specifically, it uses a Python NLP library (e.g., spaCy) to analyze the text data and identify the emotional state of customers and the most frequently inquired items. For example, it extracts emotions such as "dissatisfied" or "satisfied" from the content of inquiries and counts their frequency.

[1312] Visualization

[1313] The server visualizes the analysis results. Tableau or Power BI is used to generate dashboards and graphs. For example, the dashboard could display a "ranking of frequently inquired items" or a "distribution of emotional states." Company personnel can access the dashboard via a browser and check the results.

[1314] Proposal for improvement

[1315] Based on the visualized results, the server uses a generative AI model (e.g., GPT-3) to suggest ways to improve the service. For example, it generates specific suggestions such as "improving FAQs for frequently asked questions" or "improving services that cause complaints." The generated suggestions are displayed on the dashboard.

[1316] Create an action plan

[1317] The user (company representative) creates a specific action plan based on the improvement measures proposed by the AI. For example, the user decides on specific actions such as "creating a new FAQ page" or "assembling a project team to improve a specific service."

[1318] Examples

[1319] For example, let's say a company wants to improve customer satisfaction. The system is used in the following steps:

[1320] 1. Data Collection

[1321] The server calls the CRM system API at a fixed time every day to obtain the latest inquiry data. It also uses the SNS API to collect evaluation comments from the past 24 hours. The collected data is stored in a database.

[1322] 2. Data Analysis

[1323] The server runs Python scripts to analyze the collected data. The scripts tokenize the text data and perform sentiment analysis, for example extracting the sentiment "unhappy" from the comment "This service is terrible."

[1324] 3. Visualization

[1325] The server sends the analysis results to Tableau and updates the dashboard. The dashboard displays the latest ranking of items with the most inquiries and the distribution of emotional states. Company personnel can access the dashboard through a browser to check the results.

[1326] 4. Proposal for improvement

[1327] The server sends a prompt to a generative AI model (e.g., GPT-3) to generate improvement measures. For example, a suggestion such as "improving FAQs for frequently asked questions" is generated. The generated suggestions are displayed on the dashboard.

[1328] 5. Create an action plan

[1329] The user (company staff) checks the suggestions on the dashboard and creates a specific action plan. For example, the user may create an action plan such as "select a person to create a new FAQ page and publish the page within a week."

[1330] Examples of prompt statements

[1331] Below is an example of a prompt sentence to be input to the generative AI model.

[1332] "Please suggest improvements to improve customer satisfaction based on the content and frequency of customer inquiries and online reviews."

[1333] In this way, the company can improve customer satisfaction and reduce unnecessary calls to the call center. The flow of the identification process in the third embodiment will be described with reference to FIG.

[1334] Step 1: Data collection

[1335] The server collects customer inquiry history and online information from the company's CRM system and SNS API. As input, it uses the API key of the CRM system and authentication information of the SNS API. Specifically, the server calls the API at a fixed time every day to obtain the latest inquiry data and evaluation comments. As output, the collected data is stored in a database.

[1336] Step 2: Data analysis

[1337] The server analyzes the collected data. As input, it uses the inquiry content and evaluation comments stored in the database. Specifically, it uses a Python NLP library (e.g., spaCy) to tokenize the text data and perform sentiment analysis. For example, it extracts the sentiment "dissatisfied" from the comment "This service is terrible." As output, the analysis results are stored in the database.

[1338] Step 3: Visualization

[1339] The server visualizes the analysis results. As input, it uses the analysis results stored in the database. Specifically, it generates dashboards and graphs using Tableau or Power BI. For example, it displays "ranking of frequently inquired items" or "distribution of emotional states" on the dashboard. As output, the visualized data is displayed on the dashboard.

[1340] Step 4: Propose improvements

[1341] Based on the visualization results, the server uses a generative AI model (e.g. GPT-3) to suggest improvement methods. As input, it uses the visualized data and a prompt text. In concrete terms, it sends the prompt text to the generative AI model to generate improvement measures. For example, a suggestion such as "improving FAQs for items with many inquiries" is generated. As output, the generated suggestions are displayed on a dashboard.

[1342] Step 5: Create an action plan

[1343] The user (company staff) creates a specific action plan based on the improvement measures proposed by the AI. As input, the suggestions displayed on the dashboard are used. As specific operations, the user reviews the suggestions, selects a person to be responsible for creating a new FAQ page, and creates a specific action plan, such as publishing the page within a week. As output, the created action plan is shared within the company and put into action.

[1344] (Application example 3)

[1345] Next, a description will be given of Application Example 3 of Form Example 3. 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".

[1346] With conventional customer satisfaction improvement systems, it was difficult to effectively collect and analyze customer inquiries and online evaluations, and quickly develop specific improvement measures. In addition, the proposals for improvement measures based on visualized data were insufficient, and it took a lot of effort for companies to create specific action plans. Furthermore, there was a lack of concrete measures to reduce unnecessary calls to the call center, so there was a need to improve efficiency in both quality and cost.

[1347] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1348] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, an AI means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that lower customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for proposing improvement measures using a generative AI model, a means for creating a specific action plan based on the proposed improvement measures, and a means for collecting feedback from customers in real time and visualizing it in a dashboard or graph. This allows specific improvement measures to improve customer satisfaction to be quickly proposed, and companies to efficiently create action plans. It also reduces unnecessary calls to the call center, contributing to companies in terms of both quality and cost.

[1349] "Inquiry history to our company" refers to the content and history of inquiries that a company has received from customers.

[1350] "Online information" refers to information such as customer ratings, reviews, and comments that are published on the Internet.

[1351] "Means of collection" refers to technologies and methods for automatically collecting inquiry history and online information.

[1352] "Means of analysis" refers to the techniques and methods used to analyze collected information and convert it into meaningful data.

[1353] "AI means" refers to technologies and methods that use artificial intelligence to analyze data and develop improvement measures.

[1354] "Visualization means" refers to techniques and methods for displaying analytical results in a visual format, such as graphs or dashboards.

[1355] "Means for planning improvement methods" refers to techniques and methods for devising specific improvement measures based on visualized data.

[1356] A "generative AI model" refers to a generative artificial intelligence model, and is a technology for generating data and information for specific purposes.

[1357] "Proposed means" refers to techniques and methods for proposing improvement measures using generative AI models.

[1358] "Means of creating an action plan" refers to techniques and methods for creating a specific action plan based on proposed improvement measures.

[1359] "Means of collecting feedback in real time" refers to technologies and methods for collecting feedback from customers immediately.

[1360] "Dashboard and / or graphical visualization means" refers to techniques and methods for displaying collected feedback in the form of dashboards and / or graphical displays.

[1361] The system for implementing this invention is composed of three main elements: a server, a terminal, and a user. The server is responsible for collecting customer inquiry history and information on the Internet and analyzing this data. Specifically, the server includes the following means:

[1362] 1. Collection method: The server automatically collects inquiry history and online evaluations. This is done using web scraping technology or APIs. For example, Python libraries such as BeautifulSoup and Scrapy can be used.

[1363] 2. Means of analysis: To analyze the collected data, the server uses Natural Language Processing (NLP) techniques, specifically vectorizing the text data and classifying the data using a clustering algorithm (e.g. KMeans), which allows the customer feedback to be divided into meaningful categories.

[1364] 3. AI method: The server uses a generative AI model to propose improvement measures based on the analysis results. Here, a generative AI model such as OpenAI's GPT-3 is used. A specific improvement measure is generated using the prompt sentence as input.

[1365] 4. Visualization: To visualize the analysis results, the server generates dashboards and graphs using data visualization libraries such as Matplotlib and Plotly, allowing users to understand the analysis results at a glance.

[1366] 5. Means of planning improvement methods: The server creates a specific action plan based on the visualized data. Based on the improvement measures proposed by the generative AI model, the server plans an action plan that the user can execute.

[1367] 6. Means of collecting feedback in real time: The server collects feedback from customers in real time. For this, a real-time database (e.g. Firebase) can be used.

[1368] 7. Visualization in the form of dashboards and graphs: The server displays the collected feedback in the form of dashboards and graphs, allowing users to understand the feedback trends in real time.

[1369] As a concrete example, the following prompt sentence can be input to a generative AI model:

[1370] Please suggest improvements to improve customer satisfaction. Below is some feedback from customers.

[1371] Product delivery is slow

[1372] Poor customer support

[1373] The quality of the product was not as expected

[1374] Based on this prompt, the generative AI model will suggest specific improvements, such as "reviewing the delivery process," "strengthening customer support training," and "stricter product quality control."

[1375] In this way, the server can quickly propose specific improvement measures to improve customer satisfaction, enabling companies to create action plans efficiently. It can also reduce unnecessary calls to the call center, contributing to companies in terms of both quality and cost.

[1376] The flow of the specific process in the application example 3 will be described with reference to FIG.

[1377] Step 1:

[1378] The server collects customer inquiry history and online information. Specifically, it uses web scraping technology and APIs to automatically collect customer feedback and ratings. Website URLs and API endpoints are required as input, and collected text data is obtained as output.

[1379] Step 2:

[1380] The server analyzes the collected data. Specifically, it uses natural language processing (NLP) techniques to vectorize the text data and classifies the data using a clustering algorithm (e.g., KMeans). As input, it requires the collected text data, and as output, it obtains the clustered data.

[1381] Step 3:

[1382] The server uses a generative AI model to propose improvement measures based on the analysis results. Specifically, it uses a generative AI model such as OpenAI's GPT-3 to generate improvement measures using the prompt sentence as input. Clustered data and the prompt sentence are required as input, and the generated improvement measures are obtained as output.

[1383] Step 4:

[1384] The server visualizes the analysis results, specifically by generating dashboards and graphs using data visualization libraries such as Matplotlib and Plotly. As input, it requires the clustered data, and as output, it gets the visualized graphs and dashboards.

[1385] Step 5:

[1386] The server creates a concrete action plan based on the visualized data. Based on the improvement measures proposed by the generative AI model, the server creates an action plan that the user can execute. As input, the generated improvement measures are required, and as output, a concrete action plan is obtained.

[1387] Step 6:

[1388] The server collects customer feedback in real time. Specifically, it uses a real-time database (e.g. Firebase) to collect feedback instantly. As input, it requires customer feedback and as output, it gets feedback data that is updated in real time.

[1389] Step 7:

[1390] The server displays the collected feedback in the form of dashboards and graphs. Specifically, it uses a data visualization library to display feedback trends in real time. As input, it requires feedback data updated in real time, and as output, it gets visualized dashboards and graphs.

[1391] 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 a voice indicating a user input for the result of the specific processing. The control unit 46A transmits the 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.

[1392] 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> ) and other generation AIs. The data generation model 58 is obtained by making a neural network perform deep learning. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image is input. The data generation model 58 performs inference on the input inference data according to 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.

[1393] Another example of generative AI is Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) are mentioned.

[1394] 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 smart glasses 214.

[1395] [Third embodiment]

[1396] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1397] 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.

[1398] 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 wide area network (WAN) and / or a local area network (LAN).

[1399] 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.

[1400] 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 the voice according to instructions from the processor 46.

[1401] Camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a 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 (e.g., an imaging range defined by an angle of view equivalent to the width of the field of vision of an average healthy person).

[1402] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54. The transmission and reception of various types of information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is performed in a secure state.

[1403] Fig. 6 shows an example of 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.

[1404] 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.

[1405] 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.

[1406] In the headset type terminal 314, 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.

[1407] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1408] "Example 1"

[1409] As one embodiment of the present invention, a web scraping tool or API is used to collect information such as inquiry history and online information. The collected information is saved as text data and becomes input data for analysis by AI.

[1410] "Example 2"

[1411] Next, we use a generative AI (Generative Pretrained Transformer), a type of natural language processing (NLP), as an AI method to analyze the collected information. Generative AI has the ability to learn from large amounts of text data and generate new text. Using this ability, we extract factors that are lowering customer satisfaction from the collected information.

[1412] "Example 3"

[1413] Furthermore, dashboards and graphs are used to visualize the analysis results. This allows companies to understand problems at a glance and devise improvement measures. Specifically, the content and frequency of customer inquiries, online reviews, etc. are visualized, and from this information, factors that are lowering customer satisfaction are identified.

[1414] "Example 4"

[1415] Finally, the improvement measures proposed by AI are considered as a means of planning improvement methods based on the visualized results. Specifically, based on the improvement measures proposed by AI, companies can determine the direction of service improvement and create specific action plans. This allows companies to improve customer satisfaction and reduce unnecessary calls to their call centers.

[1416] The process flow of each embodiment will be described below.

[1417] "Example 1"

[1418] Step 1: Use a web scraping tool or API to collect inquiry history and online information for your company.

[1419] Step 2: The collected information is saved as text data and used as input data for AI analysis.

[1420] "Example 2"

[1421] Step 1: Generative AI, a type of natural language processing (NLP)

[1422] The collected information is analyzed using a trained Transformer.

[1423] Step 2: Using the results of the generative AI's learning, generate new text and extract the factors that are reducing customer satisfaction.

[1424] "Example 3"

[1425] Step 1: Use dashboards and graphs to visualize the analysis results.

[1426] Step 2: From the visualized results, identify problems and develop improvement measures.

[1427] "Example 4"

[1428] Step 1: Based on the improvement measures suggested by AI, the company decides on the direction of service improvement.

[1429] Step 2: Create a concrete action plan to improve customer satisfaction and reduce unnecessary calls to the call center.

[1430] Example 1

[1431] Next, a description will be given of Example 1 of the embodiment 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".

[1432] With the conventional system, it was difficult to efficiently collect and analyze the history of inquiries to the company and information on the Internet. It was also difficult to properly analyze the collected information and devise specific improvement methods to increase customer satisfaction. Furthermore, there was a lack of effective means to reduce unnecessary calls to the call center.

[1433] 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.

[1434] In this invention, the server includes a means for collecting inquiry history and online information to the company using a web scraping tool or API, a means for saving the collected information as text data, a means for preprocessing the saved text data using a natural language processing library, a means for inputting the preprocessed data into a generative AI model and analyzing it, a means for visualizing factors that lower customer satisfaction based on the analysis results, and a means for planning improvement methods based on the visualization results. This enables efficient collection and analysis of information, and allows specific improvement methods to be planned to improve customer satisfaction. It also reduces unnecessary calls to the call center, making it possible to contribute to companies in terms of both quality and cost.

[1435] A "web scraping tool" is software for automatically extracting data from websites.

[1436] "API" stands for Application Programming Interface, an interface for exchanging data and functions between different software programs.

[1437] "Inquiry history" is a record of inquiries made by customers to your company, and is data including the inquiry content, date and time, customer information, etc.

[1438] "Online information" refers to publicly available data or content that exists on the Internet, including social media posts, blog posts, news articles, etc.

[1439] "Text data" is data expressed as a character string, and is a collection of sentences or words.

[1440] A "natural language processing library" is a software library for analyzing text data and understanding its meaning, providing functions such as tokenization, sentiment analysis, and tagging.

[1441] "Preprocessing" refers to the process of converting data into a format that can be analyzed, and includes data cleaning and tokenization.

[1442] A "generative AI model" is an artificial intelligence model that generates new information or analytical results based on input data, and is used for natural language generation and classification tasks.

[1443] "Analysis" is the process of examining data in detail to find meaning and patterns.

[1444] "Visualization" refers to displaying data and analytical results in a visual format such as a graph or chart, and is a means of making information easier to understand intuitively.

[1445] "Improvement methods" are specific measures or action plans for solving current problems and improving the performance and quality of a system or service.

[1446] This invention is a system in which a server uses a web scraping tool or API to collect inquiry history and online information for the company, and saves the collected information as text data. The saved text data is preprocessed using a natural language processing library and input into a generative AI model. The generative AI model analyzes the input data, visualizes factors that are reducing customer satisfaction based on the analysis results, and proposes methods of improvement.

[1447] Hardware and Software Used

[1448] Hardware

[1449] Server: A computer system for data collection, storage, pre-processing, and analysis.

[1450] software

[1451] Web scraping tools: Python's BeautifulSoup and Scrapy, etc.

[1452] API: Twitter API, Google Custom Search API, etc.

[1453] Database: MySQL, PostgreSQL, etc.

[1454] Cloud storage: Amazon S3, Google Cloud Storage, etc.

[1455] Natural language processing libraries: such as NLTK or spaCy.

[1456] Generative AI models: such as OpenAI's GPT-3 and Google's BERT.

[1457] Examples

[1458] Example 1: Collection and analysis of inquiry history

[1459] 1. The server uses BeautifulSoup to collect past inquiry history from its company's inquiry form.

[1460] 2. The server stores the collected query history in a MySQL database.

[1461] 3. The server pre-processes and tokenizes the stored query history using spaCy.

[1462] 4. The server inputs the preprocessed data into GPT-3 and analyzes trends in the query content.

[1463] Example 2: Collecting and analyzing information on the Internet

[1464] 1. The server uses the Twitter API to collect tweets related to specific keywords.

[1465] 2. The server stores the collected tweets in Amazon S3.

[1466] 3. The server preprocesses the stored tweets using NLTK and performs sentiment analysis.

[1467] 4. The server inputs the preprocessed data into BERT to classify the content of the tweets.

[1468] Examples of prompt statements

[1469] Prompt 1: Analyze the query history

[1470] "Based on your inquiry history over the past year, please tell us the most common inquiries you have made."

[1471] Prompt 2: Collecting and analyzing information on the Internet

[1472] "Collect tweets related to a specific keyword and tell me the percentage of positive and negative tweets."

[1473] In this way, the server can efficiently collect information, analyze it, and devise specific improvement methods to increase customer satisfaction. It can also reduce unnecessary calls to the call center, contributing to companies in terms of both quality and cost.

[1474] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1475] Step 1:

[1476] The server collects information using web scraping tools and APIs. Specifically, it uses Python's BeautifulSoup to scrape inquiry history from websites and Twitter API to collect tweets related to specific keywords. The input is the website URL or API endpoint, and the output is the collected raw data.

[1477] Step 2:

[1478] The server stores the collected information as text data. Specifically, it stores inquiry history and tweets in a MySQL database. The input is the collected raw data, and the output is the text data stored in the database.

[1479] Step 3:

[1480] The server preprocesses the stored text data using a natural language processing library. Specifically, it uses spaCy to clean the text (e.g., remove HTML tags and special characters) and tokenize it. The input is the text data stored in the database, and the output is the preprocessed text data.

[1481] Step 4:

[1482] The server inputs the preprocessed data into the generative AI model for analysis. Specifically, it uses OpenAI's GPT-3 to analyze trends in inquiry content and Google's BERT to classify tweet content. The input is the preprocessed text data, and the output is the analysis result.

[1483] Step 5:

[1484] The server visualizes the factors that are lowering customer satisfaction based on the analysis results. Specifically, it uses a visualization tool to display the analysis results as graphs and charts. The input is the analysis results, and the output is the visualized data.

[1485] Step 6:

[1486] The server then creates improvement methods based on the visualized results. Specifically, it generates a specific action plan to improve customer satisfaction based on the visualized data. The input is the visualized data, and the output is a proposal for improvement methods.

[1487] (Application example 1)

[1488] Next, a description will be given of Application Example 1 of Form 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".

[1489] Conventional customer support systems have had problems with not being able to respond quickly and appropriately to user inquiries, leading to lower customer satisfaction and an increase in unnecessary calls to call centers. In addition, they were unable to effectively utilize online information, making it difficult to link it to service improvements. This resulted in issues such as a decline in the company's quality and efficiency in terms of costs.

[1490] 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.

[1491] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, a means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for integrating the collected information and generating answers to customer inquiries in real time using generative AI, and a means for presenting the generated answers to customer support staff. This enables prompt and appropriate responses to inquiries from users, improving customer satisfaction and reducing unnecessary calls to the call center. Furthermore, by effectively utilizing information on the Internet, it is possible to improve services, improving the quality and cost efficiency of the company.

[1492] "Inquiry history to the company" refers to a record of questions, requests, complaints, etc. that a company has received from customers.

[1493] "Online information" refers to text data such as reviews, social media posts, and blog articles that exist on the Internet.

[1494] "Means of collection" refers to the method of obtaining information using web scraping tools or APIs.

[1495] "Means of analysis" refers to the method of inputting collected data into an AI model and extracting patterns and trends.

[1496] "Planning improvements to our own services" is the process of devising ways to improve services based on the results of analysis.

[1497] A "visualization means" is a method for displaying the analysis results in the form of graphs, charts, etc.

[1498] "Means for planning improvement methods" is a method for formulating specific improvement measures based on visualized data.

[1499] "Generative AI" is an artificial intelligence technology that generates text data using natural language processing.

[1500] "A means of generating answers to customer inquiries in real time" refers to a method of using generative AI to instantly create appropriate answers.

[1501] "Means for presentation to customer support representative" refers to the manner in which the generated answers are displayed to support staff.

[1502] The following system configuration and processing procedure will be described as an embodiment of the present invention.

[1503] System Configuration

[1504] The system consists of the following main components:

[1505] 1. Server: Collects data, analyzes it, and generates answers using generative AI.

[1506] 2. Terminal: The device used by the customer support representative (e.g. PC, smartphone).

[1507] 3. User: The customer making the enquiry.

[1508] Hardware and Software Used

[1509] Hardware: Servers, PCs, smartphones

[1510] Software: Python, BeautifulSoup, requests, OpenAI API

[1511] Data processing and calculation

[1512] 1. Data collection:

[1513] The server uses a web scraping tool (BeautifulSoup) and an API (requests) to collect inquiry history and online information for the company.

[1514] The collected data is saved in text format.

[1515] 2. Data Analysis:

[1516] The server inputs the collected text data into a generative AI (OpenAI API) for analysis.

[1517] The analysis results identify and visualize the factors that are lowering customer satisfaction.

[1518] 3. Answer generation:

[1519] The server uses generative AI to generate answers to customer inquiries in real time.

[1520] The generated answer is presented to the customer support representative's terminal.

[1521] Examples

[1522] For example, if a user asks "Do you have product A in stock?", the server will input the following prompt to the generative AI:

[1523] Analyze user inquiries and generate the best answers. Here are some of the inquiries we collected:

[1524] 1. Do you have product A in stock?

[1525] 2. Please tell me how to return product B.

[1526] 3. I would like to see reviews of product C.

[1527] The generative AI generates an appropriate response based on this prompt text and displays it on the customer support representative's device, enabling the representative to respond quickly and appropriately.

[1528] This system enables prompt and appropriate responses to user inquiries, improving customer satisfaction and reducing unnecessary calls to the call center. In addition, by effectively utilizing information on the Internet, it is possible to improve services, improving the quality and cost efficiency of companies.

[1529] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1530] Step 1:

[1531] The server uses a web scraping tool (BeautifulSoup) to collect inquiry history and related information from a specified website. The website URL is given as input, and the collected text data is obtained as output. Specifically, the server sends an HTTP request and analyzes the obtained HTML data to extract text.

[1532] Step 2:

[1533] The server uses API requests to collect inquiry history and related information from the specified API endpoint. The API URL is given as input, and the collected data in JSON format is obtained as output. In concrete terms, the server sends an API request, analyzes the obtained JSON data, and extracts the required information.

[1534] Step 3:

[1535] The server integrates the text data and JSON data collected in steps 1 and 2. Text data and JSON data are given as input, and the integrated text data is obtained as output. In concrete terms, the server combines the text data and JSON data and saves them as a single text data.

[1536] Step 4:

[1537] The server inputs the integrated text data into the generative AI (OpenAI API) and performs analysis. The integrated text data is given as input, and the analysis results are obtained as output. In concrete terms, the server sends a prompt to the generative AI and receives a response from the AI.

[1538] Step 5:

[1539] The server generates answers to customer inquiries in real time based on the analysis results obtained from the generative AI. The analysis results are given as input, and the generated answers are obtained as output. In concrete terms, the server analyzes the analysis results and generates appropriate answers.

[1540] Step 6:

[1541] The server presents the generated answer to the customer support representative's terminal. The generated answer is given as an input, and the answer displayed on the representative's terminal is obtained as an output. In concrete terms, the server transmits the generated answer to the representative's terminal and displays it.

[1542] Step 7:

[1543] The customer support representative responds to the user's inquiry based on the presented answer. The input is the answer presented by the server, and the output is a response to the user. In concrete terms, the representative checks the presented answer and gives an appropriate response to the user.

[1544] Example 2

[1545] Next, a second embodiment of the second form will be described. 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".

[1546] With conventional customer satisfaction improvement systems, it was difficult to effectively analyze customer feedback and reviews and extract specific areas for improvement. In addition, the process of visualizing the analysis results and proposing specific improvement methods was time-consuming and laborious, making it difficult to respond quickly. Furthermore, it was not possible to reduce unnecessary calls to the call center, and there was a need to improve efficiency in both quality and cost.

[1547] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1548] In this invention, the server

[1549] A means of collecting inquiry history and online information for our company,

[1550] A means for pre-processing the collected information;

[1551] A means for analyzing the preprocessed information using a generative AI model; and

[1552] A means for extracting factors that are lowering customer satisfaction based on the analysis results;

[1553] A means for visualizing the extracted factors;

[1554] A means to plan improvement methods based on the visualized results;

[1555] This makes it possible to effectively analyze customer feedback and reviews and quickly extract specific areas for improvement. In addition, visualizing the analysis results makes it easier to plan specific improvement methods, which reduces unnecessary calls to the call center, resulting in improved efficiency in both quality and cost.

[1556] "Inquiry history to our company" refers to a record of inquiries made by customers to our company.

[1557] "Online information" refers to text data such as customer feedback, reviews, and comments that exist on the Internet.

[1558] "Means of collection" refers to methods and technologies for automatically obtaining inquiry history and online information.

[1559] "Preprocessing means" refers to methods or techniques for converting collected text data into a format that is easy to analyze.

[1560] A "generative AI model" is an artificial intelligence model that has the ability to learn from large amounts of text data and generate new text.

[1561] "Means of analysis" refers to the methods and techniques used to analyze the pre-processed data using the generative AI model and extract meaningful information.

[1562] "Factors that reduce customer satisfaction" are elements that cause customer dissatisfaction and negative feedback.

[1563] "Means of extraction" refers to methods or techniques for extracting specific factors from the analysis results.

[1564] A "visualization means" is a method or technique for displaying the extracted factors in a visual format, such as a graph or chart.

[1565] "Means for planning improvement methods" are methods and techniques for devising specific improvement measures based on the visualized results.

[1566] The present invention is a system for improving customer satisfaction, in which a server, a terminal, and a user work together. A specific embodiment of this system will be described below.

[1567] 1. System Configuration

[1568] The server has the following main functions:

[1569] A means of collecting inquiry history and online information for your company

[1570] A means of pre-processing the collected information

[1571] A means of analyzing pre-processed information using generative AI models

[1572] A means of extracting factors that are reducing customer satisfaction based on analysis results

[1573] A means to visualize the extracted factors

[1574] A means to plan improvement methods based on visualized results

[1575] 2. Hardware and Software Used

[1576] The server is a computer system equipped with a high-performance processor and a large amount of memory. The following software is used:

[1577] Web scraping tools and APIs for data collection

[1578] Natural language processing libraries for data preprocessing (e.g. NLTK, spaCy)

[1579] Generative AI models (e.g. GPT-3)

[1580] Tools for data visualization (e.g. Matplotlib, D3.js)

[1581] 3. Program Processing

[1582] The server first collects contact history and online information, including website forms, emails, and social media comments, and then preprocesses the collected text data. Specifically, it cleans, tokenizes, and normalizes the text.

[1583] The preprocessed data is input into a generative AI model (GPT-3). This model has the ability to learn from large amounts of text data and generate new text. The generative AI model extracts factors that are lowering customer satisfaction from the input data. Specifically, it identifies negative feedback and dissatisfaction and lists them as factors.

[1584] 4. Specific Examples

[1585] As a concrete example, the following prompt sentence is input into the generative AI model.

[1586] Example prompt:

[1587] "Please extract the factors that are causing low customer satisfaction from the following text data. Text data: 'Poor product quality. Slow delivery. Unfriendly customer service.'"

[1588] The output of the generative AI model is, "The factors that are lowering customer satisfaction are product quality, delivery delays, and unfriendly customer service."

[1589] 5. Visualize and plan ways to improve

[1590] The server visualizes the extracted factors in visual formats such as graphs and charts. Based on the visualized results, specific improvement measures are proposed. For example, improving product quality, improving the delivery system, or customer service training could be considered.

[1591] In this way, the server can extract factors that are lowering customer satisfaction from the information collected using the generative AI model and quickly devise specific improvement measures.

[1592] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1593] Step 1:

[1594] The server collects the company's inquiry history and other online information. As input, it obtains text data such as website forms, emails, and SNS comments. Specifically, the server automatically collects data using web scraping tools and APIs. The output is the collected raw text data.

[1595] Step 2:

[1596] The server preprocesses the collected text data. As input, it uses the raw text data collected in step 1. Specifically, the server uses natural language processing libraries (e.g., NLTK, spaCy) to clean the text (remove unnecessary characters and symbols), tokenize (split into words and phrases), and normalize (convert into a unified format). The output is the preprocessed, clean text data.

[1597] Step 3:

[1598] The server inputs the preprocessed data into a generative AI model (e.g. GPT-3). The clean text data preprocessed in step 2 is used as input. Specifically, the server divides the preprocessed text data into batches and inputs them into the generative AI model. The generative AI model has the ability to learn from large amounts of text data and generate new text. The output is the result of analysis by the generative AI model.

[1599] Step 4:

[1600] The server extracts factors that are reducing customer satisfaction from the output of the generative AI model. As input, it uses the analysis results obtained from the generative AI model in step 3. Specifically, the server analyzes the output of the generative AI model and extracts negative keywords and phrases. The output is a list of factors that are reducing customer satisfaction.

[1601] Step 5:

[1602] The server visualizes the extracted factors. As input, it uses the list of factors extracted in step 4. Specifically, the server uses a data visualization tool (e.g., Matplotlib, D3.js) to display the factors in a visual form such as a graph or chart. The output is a graph or chart of the visualized factors.

[1603] Step 6:

[1604] The server plans improvement methods based on the visualization results. As input, it uses graphs and charts of the factors visualized in step 5. In terms of specific operations, the server devises specific improvement measures based on the visualized data. For example, improving product quality, improving the delivery system, customer service training, etc. are possible. The output is a list of specific improvement methods.

[1605] (Application example 2)

[1606] Next, a description will be given of Application Example 2 of Form 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".

[1607] With conventional customer satisfaction analysis systems, it was difficult to efficiently collect and analyze customer reviews and feedback and identify specific areas for improvement. In addition, there was a lack of a way to visualize the analysis results and quickly notify managers, which meant it took a long time to improve customer satisfaction. Furthermore, there was a lack of specific measures to reduce unnecessary calls to the call center.

[1608] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1609] In this invention, the server includes a means for collecting inquiry history and online information, an AI means for analyzing the collected information and planning improvements to the company's services, a means for visualizing factors that lower customer satisfaction based on the analysis results, a means for planning improvement methods based on the visualization results, a means for collecting customer reviews and feedback from mail order sites, a means for extracting factors that lower customer satisfaction using generative AI, and a means for notifying the manager of the extracted factors. This makes it possible to quickly identify specific factors that lower customer satisfaction and notify the manager. In addition, it is possible to take specific measures to reduce unnecessary calls to the call center.

[1610] An "inquiry history" is a record of inquiries made by a customer to a company.

[1611] "Online information" refers to text data such as reviews and feedback that exist on the Internet.

[1612] "Means of collection" refers to the methods and technologies used to collect inquiry history and online information.

[1613] "Means of analysis" are the methods and techniques used to analyze collected information and convert it into meaningful data.

[1614] "AI means for proposing improvements" refers to artificial intelligence technology that proposes improvement measures for services based on the results of analysis.

[1615] A "visualization means" is a method or technique for displaying the results of an analysis in a visual format, such as a graph or chart.

[1616] "Means for planning improvement methods" refers to methods and techniques for devising specific improvement measures based on the visualized results.

[1617] An "online shopping site" is a website that sells products and services over the Internet.

[1618] A "customer review" is an evaluation or comment made by a customer about a product or service.

[1619] "Feedback" refers to opinions and impressions provided by customers.

[1620] "Generative AI" is an artificial intelligence technology that has the ability to learn from large amounts of text data and generate new text.

[1621] "Means of extracting factors" refers to methods and techniques that use generative AI to identify the causes of declining customer satisfaction.

[1622] "Means of notifying the administrator" refers to methods or techniques for notifying the administrator of the extracted causes.

[1623] A system for carrying out this invention has the following configuration: The server includes means for collecting the history of inquiries made to the company and information on the internet, AI means for analyzing the collected information and proposing improvements to the company's services, means for visualizing factors that are lowering customer satisfaction based on the analysis results, means for proposing improvement methods based on the visualization results, means for collecting customer reviews and feedback from mail order sites, means for extracting factors that are lowering customer satisfaction using generative AI, and means for notifying an administrator of the extracted factors.

[1624] Hardware and Software Used

[1625] Hardware: Servers, smartphones

[1626] Software: Python, TensorFlow, Transformers library (Hugging Face), Flask (backend), React Native (frontend)

[1627] Data processing and calculation

[1628] Data collection

[1629] The server collects customer reviews and feedback from online shopping sites through the API, allowing for efficient collection of customer opinions and feedback.

[1630] Data Preprocessing

[1631] The collected text data is cleaned and tokenized, which converts the data into a form suitable for analysis.

[1632] Application of generative AI models

[1633] The server uses generative AI (such as GPT-3) to analyze the text data and extract factors that are reducing customer satisfaction. Generative AI has the ability to learn from large amounts of text data and generate new text.

[1634] Viewing the results

[1635] The extracted factors are visualized on the dashboard of a smartphone app, allowing managers to quickly grasp the specific points of dissatisfaction.

[1636] Administrator Notification

[1637] The server notifies the administrator of the extracted cause, allowing the administrator to take prompt action.

[1638] Examples

[1639] For example, by inputting a prompt sentence like the one below into a generative AI model, the AI ​​can analyze the review and extract specific points of dissatisfaction.

[1640] Example prompt:

[1641] "Analyze the following customer reviews to identify factors that are causing low customer satisfaction. Review: 'The product was delivered late and the quality was below expectations.'"

[1642] By inputting this prompt into a generative AI model, the AI ​​can analyze the reviews and extract specific complaints such as "The product was delivered late" or "The quality was below expectations." This allows managers to take swift action to address the complaints, helping to improve customer satisfaction.

[1643] The flow of the specific process in the application example 2 will be described with reference to FIG.

[1644] Step 1:

[1645] The server collects customer reviews and feedback from the online shopping site through an API. The input is the API endpoint of the online shopping site, and the output is the text data of the collected customer reviews and feedback. Specifically, the server periodically calls the API to obtain new reviews and feedback.

[1646] Step 2:

[1647] The server preprocesses the collected text data. The input is the text data collected in step 1, and the output is the cleaned and tokenized data. Specifically, the server removes unnecessary characters and spaces and splits the text into tokens.

[1648] Step 3:

[1649] The server uses a generative AI model (such as GPT-3) to analyze the preprocessed data. The input is the data preprocessed in step 2, and the output is the analysis result including the factors that are reducing customer satisfaction. Specifically, the server inputs a prompt sentence into the generative AI model and obtains the analysis result.

[1650] Step 4:

[1651] The server visualizes the analysis results. The input is the analysis results obtained in step 3, and the output is the data displayed in a visual format such as a graph or chart. Specifically, the server converts the analysis results into graphs and charts for display on the dashboard.

[1652] Step 5:

[1653] The server then proposes an improvement method based on the visualized results. The input is the data visualized in step 4, and the output is a specific improvement measure. Specifically, the server analyzes the visualized data and proposes an improvement measure.

[1654] Step 6:

[1655] The server notifies the administrator of the extracted causes and remedial measures. The input is the remedial measures proposed in step 5, and the output is the notification sent to the administrator. Specifically, the server sends information to the administrator via email or push notification.

[1656] Step 7:

[1657] The administrator receives the notification and takes specific corrective action. The input is the notification received in step 6, and the output is the corrective action taken. Specifically, the administrator makes the necessary changes or actions based on the content of the notification.

[1658] Example 3

[1659] Next, a third embodiment of the third embodiment will be described. 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".

[1660] With conventional customer satisfaction improvement systems, it was difficult to effectively collect and analyze customer inquiries and online evaluation data and quickly propose specific improvement measures. In addition, the process of creating specific action plans based on visualized data was complicated, making it difficult for companies to respond quickly. This led to issues such as insufficient improvement of customer satisfaction and reduction of unnecessary calls to call centers.

[1661] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1662] In this invention, the server includes a means for collecting inquiry history and online information, an artificial intelligence means for analyzing the collected information and proposing improvements to the company's services, a means for visualizing factors lowering customer satisfaction based on the analysis results, a means for proposing improvement methods based on the visualization results, a means for inputting a prompt sentence into the generative AI model to have it propose improvement measures, and a means for creating a specific action plan based on the proposed improvement measures. This makes it possible to effectively collect and analyze customer inquiry contents and online evaluation data, quickly propose specific improvement measures, and create a specific action plan based on the visualized data.

[1663] "Inquiry history to our company" refers to a record of the content, date and time, frequency, etc. of inquiries made by customers to a company.

[1664] "Online information" refers to data such as customer ratings, reviews, and comments that exist on the Internet.

[1665] "Means of collection" refers to the methods and technologies used to obtain the necessary information using databases or APIs.

[1666] "Analytical means" are the methods and techniques used to analyze collected data and extract meaningful information.

[1667] An "artificial intelligence means" is a system that uses techniques such as machine learning and natural language processing to analyze data and gain insights.

[1668] "Visualization means" refers to methods and technologies for displaying analysis results in the form of graphs, dashboards, etc., making them intuitively understandable.

[1669] A "generative AI model" is an artificial intelligence model that is trained to generate appropriate outputs for specific inputs.

[1670] A "prompt" is an instruction entered into a generative AI model to cause it to perform a specific task.

[1671] "Means for proposing improvement measures" refers to methods or technologies that use generative AI models to propose specific improvement measures based on the analysis results.

[1672] "Means for creating an action plan" refers to methods or techniques for formulating a specific action plan based on proposed improvement measures.

[1673] The present invention is directed to a system that aims to improve customer satisfaction and reduce unnecessary calls to a call center. A specific embodiment of this system will be described below.

[1674] First, the server collects the history of inquiries to the company and information on the internet. To do this, a database service or API is used. For example, a general cloud database is used as the database service, and data is obtained from customer inquiry systems and SNS platforms via API.

[1675] Next, the server analyzes the collected data. This analysis is performed using programming languages ​​such as Python and R. Specifically, natural language processing (NLP) libraries (e.g., NLTK and spaCy) are used to analyze the text data and perform frequent keyword and sentiment analysis. Machine learning models are also used to perform data clustering and anomaly detection.

[1676] The analysis results are displayed as dashboards and graphs using visualization tools such as Tableau and Microsoft Power BI. This allows users to view the dashboards on a browser and intuitively understand the analysis results.

[1677] Based on the visualization results, the user inputs a prompt sentence into the generative AI model to suggest improvement measures. For example, the following prompt sentence can be input into the generative AI model:

[1678] "Please suggest improvement measures to increase customer satisfaction based on the content and frequency of inquiries from customers and online evaluation data."

[1679] The server sends this prompt to the generative AI model and receives improvement measures proposed by the AI. The user creates a specific action plan based on the proposed improvement measures. For example, if the AI ​​suggests an improvement measure such as "There are many complaints about a specific product, so we suggest that you conduct additional tests to improve the quality of the product," the user creates and executes a specific test plan based on the proposal.

[1680] In this way, companies can improve customer satisfaction and reduce unnecessary calls to call centers. This system can consistently perform data collection, analysis, visualization, improvement proposals, and action plan creation, enabling quick and effective response. The flow of the identification process in the third embodiment will be described with reference to FIG. 15.

[1681] Step 1: Data collection

[1682] The server collects the company's inquiry history and online information. As input, it obtains data from customer inquiry systems and social media platforms. Specifically, the server obtains data from these platforms via APIs and stores it in a database. As output, the content of inquiries, their frequency, and online evaluation data are stored in the database.

[1683] Step 2: Data analysis

[1684] The server analyzes the collected data. As input, it uses inquiry content stored in a database and online evaluation data. Specifically, the server uses a natural language processing (NLP) library (e.g., NLTK or spaCy) in Python or R to analyze the text data. This provides the results of frequent keywords and sentiment analysis. It also uses machine learning models to perform data clustering and anomaly detection. The analysis results are obtained as output.

[1685] Step 3: Data visualization

[1686] The server visualizes the analysis results. It uses the results of the data analysis as input. Specifically, the server exports the analysis results in JSON format and imports them into a visualization tool (e.g. Tableau or Microsoft Power BI), which displays the analysis results as dashboards and graphs. As output, it provides the visualized data.

[1687] Step 4: Propose improvements

[1688] Based on the visualization results, the user inputs a prompt sentence into the generative AI model to have it suggest improvements. The visualized data and the prompt sentence are used as input. In concrete terms, the user inputs the following prompt sentence into the generative AI model: "Please suggest improvements to improve customer satisfaction based on the content and frequency of customer inquiries, as well as online evaluation data." The server sends this prompt sentence to the generative AI model and receives the improvements suggested by the AI. The output is the improvements suggested by the AI.

[1689] Step 5: Create an action plan

[1690] The user creates a specific action plan based on the proposed improvement measures. The improvement measures proposed by the AI ​​are used as input. In concrete terms, the user consults with relevant departments within the company based on the proposed improvement measures (e.g., "There are many complaints about a particular product, so we propose that additional testing be carried out to improve the quality of the product") and formulates a specific action plan. A specific action plan is obtained as output.

[1691] In this way, companies can improve customer satisfaction and reduce unnecessary calls to their call centers.

[1692] (Application example 3)

[1693] Next, a description will be given of Application Example 3 of Form Example 3. 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".

[1694] With conventional customer support systems, it was difficult to effectively analyze and visualize customer inquiries and online reviews. In addition, there was a lack of means to plan specific improvement measures based on the analysis results, and it was not possible to fully improve customer satisfaction or reduce unnecessary calls to the call center. This caused companies to face challenges in both quality and cost.

[1695] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1696] In this invention, the server includes a means for collecting the history of inquiries to the company and information on the Internet, an AI means for analyzing the collected information and proposing improvements to the company's services, a means for visualizing factors that are lowering customer satisfaction based on the analysis results, a means for proposing improvement methods based on the visualization results, a means for proposing improvement measures based on specific inquiry contents using generative AI, and a means for displaying the proposed improvement measures. This makes it possible to effectively analyze and visualize the contents of inquiries from customers and online evaluations. In addition, specific improvement measures can be proposed based on the analysis results, which can improve customer satisfaction and reduce unnecessary calls to the call center.

[1697] "Inquiry history" refers to a record of customer inquiries, including their content, frequency, and date and time.

[1698] "Online information" refers to data such as customer ratings, reviews, and comments that exist on the Internet.

[1699] "Collection means" refers to methods or devices for acquiring inquiry history and online information.

[1700] An "analysis means" is a method or device for analyzing collected information and converting it into meaningful data.

[1701] A "visualization means" is a method or device that displays analysis results in the form of graphs, dashboards, etc., to enable intuitive understanding.

[1702] An "improvement planning means" is a method or device for devising improvement measures for a service based on the visualization results.

[1703] "Generative AI" is a type of artificial intelligence, a model that has the ability to generate appropriate outputs for specific inputs.

[1704] A "proposal means" is a method or device for using generative AI to propose improvement measures based on specific inquiry content.

[1705] A "display means" is a method or device for showing the suggested improvement to the user.

[1706] A system for implementing this invention includes a server, a user terminal, and a generative AI model. The server collects inquiry history and online information and analyzes this data. The analysis results are visualized and displayed in a format that the user can intuitively understand. Furthermore, the generative AI model is used to propose improvement measures based on the specific inquiry content and display them on the user terminal.

[1707] The server uses Python and Dash to build a dashboard to visualize the types of inquiries and their frequency. When a user selects a specific inquiry from a drop-down menu, the data related to that inquiry is displayed as a bar graph, allowing the user to see at a glance which inquiries are occurring frequently.

[1708] In addition, the server uses the OpenAI API to generate improvement measures suggested by the generative AI model based on the selected inquiry content. For example, if the user selects "Product defect," the server inputs the following prompt sentence to the generative AI model:

[1709] Example prompt:

[1710] If the customer inquiry is about a "defective product," please provide suggestions for improving the service.

[1711] Based on this prompt, the generative AI model generates "suggestions for reducing product defects," proposing specific improvement measures such as "strengthening quality control" and "simplifying return procedures." These suggestions are displayed on the user's device, allowing the user to use them as a reference when creating a specific action plan.

[1712] This system makes it possible to effectively analyze and visualize customer inquiries and online reviews, and also to devise specific improvement measures based on the analysis results, thereby improving customer satisfaction and reducing unnecessary calls to the call center.

[1713] The flow of the specific process in the application example 3 will be described with reference to FIG.

[1714] Step 1:

[1715] The server collects inquiry history and online information.

[1716] Input: Customer inquiries, online ratings and reviews

[1717] Data processing: Save inquiry details and evaluations in a database

[1718] Output: Collected dataset

[1719] Step 2:

[1720] The server analyzes the collected data.

[1721] Input: Collected dataset

[1722] Data calculations: data cleansing, classification, frequency analysis

[1723] Output: Analysis results (frequency of inquiries, evaluation trends, etc.)

[1724] Step 3:

[1725] The server visualizes the analysis results.

[1726] Input: Analysis results

[1727] Data processing: generating graphs and dashboards

[1728] Output: Visualized data (bar graphs, dashboards)

[1729] Step 4:

[1730] The user reviews the visualized data and selects a particular query.

[1731] Input: visualized data, user selection

[1732] Data processing: Extraction of selected inquiry details

[1733] Output: Selected inquiry details

[1734] Step 5:

[1735] The server uses a generative AI model to suggest improvement measures based on the selected inquiry content.

[1736] Input: Selected inquiry content, prompt text

[1737] Data computation: Proposal generation using generative AI models

[1738] Output: Suggested remedy

[1739] Step 6:

[1740] The server displays the proposed remedy on the user terminal.

[1741] Input: Proposed remedy

[1742] Data processing: Formatting the proposal content

[1743] Output: Remediation measures displayed on the user's terminal

[1744] Step 7:

[1745] The user reviews the proposed improvements and creates a specific action plan.

[1746] Input: Proposed remedy

[1747] Data processing: Creating an action plan

[1748] Output: A concrete action plan

[1749] In addition, 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.

[1750] "Example 1"

[1751] One embodiment of the present invention is a system that incorporates an emotion engine. This system collects inquiry history and online information for the company, and analyzes that information. The analysis is performed by both the AI ​​means and the emotion engine. The AI ​​means understands the content of the information and plans improvements to the service. Meanwhile, the emotion engine analyzes the emotions of the user. For example, the emotional state of the user is estimated from the wording and tone used when the user makes an inquiry. This makes it possible to grasp not only the user's satisfaction level, but also the emotions behind it.

[1752] "Example 2"

[1753] The emotion analysis results obtained from the emotion engine are provided to the AI ​​means. The AI ​​means uses these results to plan improvements to the service. For example, if anger is detected from many users, the cause is identified and a remedial measure is proposed. This makes it possible to improve the service in response to the user's emotions, not just the content of the inquiry.

[1754] "Example 3"

[1755] The system of the present invention also includes a means for visualizing the analysis results. This makes it possible to grasp at a glance the factors that are lowering customer satisfaction and the emotional state of users. For example, the dashboard displays the most frequently inquired items and the distribution of emotional states. This allows companies to quickly grasp problems and devise appropriate improvement measures.

[1756] "Example 4"

[1757] The system of the present invention also includes a means for proposing improvement methods based on the visualization results. This makes it possible to create specific action plans and improve customer satisfaction. For example, based on information obtained from the emotion engine, improvement measures such as providing new services that resolve user dissatisfaction can be proposed. This not only improves user satisfaction, but also contributes to the company's quality and cost aspects, such as reducing unnecessary calls to call centers.

[1758] The process flow of each embodiment will be described below.

[1759] "Example 1"

[1760] Step 1: The system collects inquiry history and online information for the company.

[1761] Step 2: The collected information is analyzed by both AI means and the emotion engine.

[1762] Step 3: The emotion engine analyzes the user's emotions and provides the results to the AI ​​means.

[1763] "Example 2"

[1764] Step 1: The AI ​​means plans service improvements based on the emotion analysis results provided by the emotion engine.

[1765] Step 2: If anger is detected among many users, identify the cause and develop measures to improve it.

[1766] "Example 3"

[1767] Step 1: The system visualizes the analysis results.

[1768] Step 2: Display the most frequently inquired items and distribution of emotional states on the dashboard.

[1769] "Example 4"

[1770] Step 1: The system plans improvement measures based on the visualized results.

[1771] Step 2: Create a concrete action plan to improve customer satisfaction.

[1772] Step 3: Based on the information obtained from the emotion engine, develop improvement measures, such as providing new services to resolve user dissatisfaction.

[1773] Example 1

[1774] Next, a description will be given of Example 1 of the embodiment 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".

[1775] With the conventional system, it was difficult to not only collect the history of inquiries to the company and information on the Internet, but also to effectively analyze that information and understand the emotional state of the c...

Claims

1. an emotion engine that recognizes the emotion of a user based on the content of an inquiry from the user to the company; A planning means for inputting a specific prompt sentence into a generating AI model, extracting factors that are lowering customer satisfaction in the company's services according to the recognized emotion, and planning improvement measures for the company's services based on the factors that are lowering customer satisfaction; A means for feeding back the proposed improvement measures to the user; A system including:

2. 2. The system according to claim 1, further comprising a means for recognizing the emotions of the user by analyzing a history of inquiries made to the company by the user and information on a network, and for displaying, when an item about which users frequently inquire is identified, a ranking of the items about which users frequently inquire and a distribution of the emotional states of the user on a dashboard or in the form of a graph.

3. The system according to claim 1, wherein the planning means extracts delivery delay as the cause when the emotion includes anger from the wording and tone of the inquiry from the user, and plans a change in delivery company or a review of the delivery process as the remedial measure for the cause.

Images