System

Abstract

A system is provided.SOLUTION: A system, comprising: means for collecting data using an application programming interface; means for summarizing the collected data using a natural language processing model; means for automatically generating a report based on the summary; and means for automatically delivering the generated report to a predetermined recipient.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] For modern companies, collecting and analyzing customer feedback (VOC) and quickly and efficiently incorporating it into their services is crucial to maintaining and improving competitiveness. However, collecting and analyzing VOC is extremely costly, and manual work requires time and effort. Using external tools also incurs high costs and reduces the time available for staff to focus on other tasks. Furthermore, summarizing collected data and creating reports is labor-intensive and requires specialized knowledge. There is a need for a system that can solve these issues, collect VOC efficiently and at low cost, and automate analysis and report creation. [Means for solving the problem]

[0005] The present invention solves the above problems by the following means. First, it provides a means for collecting data from multiple platforms using an application programming interface. Next, it provides a means for summarizing the collected data using a natural language processing model, automatically extracting the key points of the data. It also provides a means for automatically generating reports based on the summary results, allowing high-quality reports to be created even without specialized knowledge. Finally, it includes a means for automatically distributing the generated reports to specified recipients, allowing personnel to regularly understand user needs. It also provides a means for extracting important key phrases from the collected data and a means for composing reports based on templates, allowing each personnel to efficiently obtain the information they need.

[0006] An "application programming interface" is an interface that allows different software programs to share functions.

[0007] "Means for collecting data" refers to methods and devices for obtaining necessary information from multiple platforms.

[0008] A "natural language processing model" is an algorithm or technology that allows a computer to understand and generate human language.

[0009] A "summarizing means" is a method or device for extracting important parts of collected text data and summarizing them briefly.

[0010] "Means for automatically generating a report" refers to a method or device for automatically creating a report based on the summary results.

[0011] "Means for automatically delivering to a predetermined recipient" refers to a method or device for automatically sending a generated report to a pre-specified recipient.

[0012] A "key phrase extraction means" is a method or device for finding important words or phrases from text data.

[0013] A "template-based construction means" is a method or apparatus for generating reports according to a predetermined format or structure. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0015] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0018] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0019] 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.

[0020] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] 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.

[0027] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0032] 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.

[0033] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0034] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0035] This invention relates to a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, generates reports, and distributes them to designated recipients, thereby enabling companies to quickly and inexpensively grasp user needs. This system is realized by automating each stage of processing, centered on a server.

[0036] Data collection

[0037] The server first collects data using application programming interfaces (APIs) and other methods, such as obtaining user reviews, comments, and feedback data from social media platforms, app stores, and customer service platforms. This is done using API access and scraping techniques, and the collected data is then stored in an intermediate database.

[0038] Specific examples

[0039] The server uses the social media platform's API to collect tweets with the hashtag "product review" and store them in a database.

[0040] Data processing and analysis

[0041] The server then analyzes the collected data using a natural language processing model. Specifically, it utilizes the latest generative AI models (e.g., deep learning models) to summarize the vast amount of collected text data. This allows it to concisely organize user opinions and feedback and generate summaries. Models such as the BERT and GPT series are used in this summarization process.

[0042] Specific examples

[0043] The server inputs the collected tweets into a summary model and obtains summary results such as "A large number of users are requesting improvements to the app's user interface."

[0044] Report Generation

[0045] The server automatically generates reports based on the summarized data. The generated summaries and extracted key phrases are arranged in a template format to create a report that clearly shows important user needs. This makes it easy for anyone, even those without specialized knowledge, to generate high-quality reports.

[0046] Specific examples

[0047] The server organizes the collected summaries according to a template and creates a report that states "over 50% of users are satisfied with the new feature" as a "highlight of the week."

[0048] Report Distribution

[0049] The server automatically delivers the final report to the designated recipients. The report is periodically sent to the relevant departments involved in the work. The delivery method is via email or the company's internal messaging system.

[0050] Specific examples

[0051] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the main user feedback for the current week and suggestions based on it.

[0052] In this way, the system of the present invention automates the entire process of data collection, analysis, report generation, and distribution, making it possible to grasp user needs efficiently and at low cost. This system allows service providers to accurately and quickly implement improvement measures that reflect user needs.

[0053] The processing flow will be explained below.

[0054] Step 1:

[0055] The server uses the application programming interface (API) to authenticate for data collection. Specifically, it accesses the Twitter API and sets authentication information using the API key, API secret, access token, and access token secret.

[0056] Step 2:

[0057] The server uses the authentication information to collect tweets containing the target hashtags and keywords. It connects to the Twitter API using the Tweepy library and retrieves 100 tweets containing the specific hashtag "product review."

[0058] Step 3:

[0059] The server converts the collected tweet data (text, posting date and time, etc.) into a Pandas data frame and stores it in an intermediate database, thereby preparing the data in a format that can be used for subsequent analysis.

[0060] Step 4:

[0061] The server inputs the text data stored in the intermediate database into a natural language processing model, specifically using the Transformers library and generative AI models such as BERT and GPT to generate summaries.

[0062] Step 5:

[0063] The server extracts the generated summary text and selects important key phrases. It uses TF-IDF and key phrase extraction techniques to identify key words and phrases from the summarized text.

[0064] Step 6:

[0065] The server automatically generates reports based on the summaries and key phrases, organized according to predefined templates and containing key user needs and statistics.

[0066] Step 7:

[0067] The server prepares the generated report for delivery to the intended recipient by configuring the SMTP server (address, port, username, password) and converting the report content to email format.

[0068] Step 8:

[0069] The server connects to the SMTP server and automatically sends the generated report to the relevant service personnel. Specifically, it sends a "report on interface improvement requests" to the personnel in charge by email every week.

[0070] The above steps automate the entire process from collecting VOCs to generating and distributing reports, making it possible to efficiently understand user needs.

[0071] Example 1

[0072] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0073] Companies are seeking a way to quickly and low-costly understand user needs. However, collecting and analyzing vast amounts of information from social media, review sites, and customer service platforms, generating useful reports based on that information, and then distributing those reports to relevant parties in a timely manner requires a great deal of time and effort. There is a need to automate these processes and provide a way to efficiently understand user needs.

[0074] 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.

[0075] In this invention, the server includes means for collecting data using an application programming interface, means for storing the collected data in an intermediate database, means for analyzing the stored data using a natural language processing model and generating a summary, means for automatically generating a report based on the summary result, and means for automatically distributing the generated report to a predetermined recipient. This automates the process of data collection, analysis, report generation, and distribution, making it possible to grasp user needs quickly and at low cost.

[0076] An "application programming interface" is a set of definitions and protocols that provide an interface through which software components can communicate with other software.

[0077] "Data collection means" are the methods and mechanisms used to obtain the required data from external sources.

[0078] The "intermediate database" is a database for temporarily storing collected data and transferring it to subsequent processing.

[0079] A "natural language processing model" is an algorithm or system that uses machine learning technology to understand and analyze human language.

[0080] A "means for generating a summary" is a method or technology for extracting important information from a huge amount of text data and summarizing it concisely.

[0081] "Means for automatically generating a report" refers to a method or system for automatically creating a report in a predetermined format based on the analysis results or summary.

[0082] "Means for automatically delivering to recipients" refers to methods or techniques for automatically sending the generated report to pre-designated recipients.

[0083] This invention relates to a system that enables companies to quickly and inexpensively grasp user needs by efficiently collecting voice of customer (VOC), automatically analyzing and summarizing it, generating reports, and distributing them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0084] First, the server uses application programming interfaces (APIs) and scraping technology to collect user reviews, comments, and feedback data from social media, review sites, and customer service platforms. The collected data is then stored in an intermediate database. For example, the server uses the API of a social media platform to collect tweets with the hashtag "product review" and store them in a database.

[0085] The server then sends the collected text data to a natural language processing model (NLP model) to analyze user opinions and feedback. Specifically, it uses deep learning models (e.g., BERT and GPT series) to extract summaries and key phrases. This allows it to organize important information from large amounts of data and convert it into an interpretable format. For example, the server inputs the collected tweets into a summarization model and obtains a summary result such as, "Many users are requesting improvements to the app's user interface."

[0086] The server then automatically generates reports based on the analyzed data. By inserting summarized data and extracted key phrases into pre-defined templates, reports that clearly indicate user needs are created. For example, the server might create a "Highlights of the Week" report that states, "More than 50% of users are satisfied with the new features."

[0087] Finally, the server automatically distributes the generated report to the designated recipients via email or the company's internal messaging system. For example, every Monday the server sends a "Report on Interface Improvement Requests" to the email address of a service representative. This email includes the main user feedback for the current week and suggestions based on that feedback.

[0088] Specific examples

[0089] The server sends a request to the SNS API (e.g., Twitter API) at a specific time every day, retrieves posts containing a specific hashtag such as "product review," and stores them in an intermediate database.

[0090] The server inputs the reviews and comments stored in the database into an NLP model (for example, the BERT model in Hugging Face's Transformers library) and stores the results as summaries and key phrases in a separate table.

[0091] The server extracts the analysis results, embeds the data in a template called "Highlights of the Week," and generates a report such as "More than 50% of users are satisfied with the new feature."

[0092] The server emails the generated report based on a pre-specified list of recipients.

[0093] Prompt Sentence Examples

[0094] This week's highlights:

[0095] Summary: Many users are requesting improvements to the app's user interface.

[0096] In this way, the system of the present invention automates the process of data collection, analysis, report generation and distribution, making it possible to grasp user needs quickly and at low cost.

[0097] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0098] Step 1: Data collection

[0099] The server uses application programming interfaces (APIs) and scraping techniques to collect user reviews, comments, and feedback data from social media platforms, review sites, and customer service platforms. Specifically, the APIs are used to retrieve posts with specific hashtags (e.g., "product reviews") from social media platforms. The server then stores the retrieved data in an intermediary database.

[0100] Input: URL or API endpoint of social media or review site

[0101] Output: Feedback data stored in an intermediate database

[0102] Step 2: Save data

[0103] The server structures and stores the collected raw data in an intermediate database, which organizes the data and prepares it for further processing. In this step, data insertion operations are performed using SQL.

[0104] Input: Raw data collected

[0105] Output: An intermediate database containing structured data

[0106] Step 3: Analyze the data

[0107] The server sends the stored text data to a natural language processing (NLP) model to analyze user opinions and feedback. Specifically, it uses deep learning models (e.g., BERT and GPT series) to extract summaries and key phrases, making it easier to grasp the gist of the data.

[0108] Input: Text data stored in the intermediate database

[0109] Output: Summary sentences and key phrases analyzed by the NLP model

[0110] Specific operation: The server retrieves the text data one by one, inputs prompts such as "What is the main opinion of this review?" into the NLP model, and saves the results in another table.

[0111] Step 4: Generate a report

[0112] The server automatically generates reports based on the analyzed data. It inserts summarized data and extracted key phrases into pre-prepared template formats to create reports that clearly show the user's needs. A template engine is used in this process.

[0113] Input: Summary sentences and key phrases from the NLP model

[0114] Output: Formatted report

[0115] What happens: The server inserts the summary and key phrases into a template and generates a report as "Highlights of the Week."

[0116] Step 5: Report Delivery

[0117] The server automatically distributes the generated reports to the designated recipients via email or an internal messaging system, based on a pre-defined schedule.

[0118] Input: Generated report

[0119] Output: Report delivered to recipients

[0120] What it does: The server generates a report and sends it to the service representative's email address every Monday, containing key user feedback and suggestions based on it.

[0121] (Application example 1)

[0122] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0123] Traditional methods for collecting, analyzing, and generating reports on voice of customer (VOC) required a lot of manual work, which was time-consuming and costly, making it difficult to quickly and efficiently understand user needs. Furthermore, generating and distributing reports required human labor, which often led to a lack of timely information sharing. This made it difficult for service providers to quickly implement appropriate improvement measures, resulting in delays in improving customer satisfaction.

[0124] 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.

[0125] In this invention, the server includes means for collecting data using an application programming interface, means for summarizing the collected data using a natural language processing model, means for automatically generating a report based on the summary results, means for automatically distributing the generated report to a predetermined recipient, and means for summarizing the collected data based on its contents, generating a report according to a predetermined schedule, and distributing the report to a predetermined recipient via email, thereby enabling user needs to be grasped quickly and efficiently at low cost.

[0126] An "application programming interface" is something that provides an interface between software or applications, enabling the exchange of data and functions.

[0127] "Data" refers to information such as reviews, comments, and feedback obtained from customers.

[0128] A "natural language processing model" is a machine learning model for analyzing and summarizing text data, and includes the BERT and GPT series.

[0129] "Summarizing" means extracting important parts from multiple text data and summarizing them in a concise form.

[0130] A "report" is a document generated from summarized data that shows the results of organizing and analyzing user needs and feedback.

[0131] "Automatically generating" means that the system generates documents according to specified algorithms or templates without human intervention.

[0132] "Intended Recipient" means a person or department that has been pre-designated to receive the report.

[0133] "Automatically delivering" means that the system sends information as emails or messages based on set schedules or conditions, without manual intervention.

[0134] A "schedule" refers to a predetermined time and order plan for performing a series of tasks or processes.

[0135] "Email" is a means of sending and receiving text data and attachments in digital form over the Internet.

[0136] This invention relates to a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, generates reports, and distributes them to designated recipients, thereby enabling companies to quickly and inexpensively understand user needs. This system is targeted at collecting and analyzing customer feedback on online shopping sites.

[0137] The server first collects data using an application programming interface (API). Specifically, it retrieves customer reviews and comments from the review API of an online shopping site. Scraping technology can also be used, for example, to automatically collect reviews on a web page. The collected data is then stored in an intermediate database. Examples of databases include MongoDB and MySQL.

[0138] Next, the server inputs the collected data into a natural language processing model (NLP model) for analysis and summarization. This summarization process uses generative AI models such as BERT and the GPT series. Specifically, user reviews and comments are taken as input, and the summarization model concisely summarizes their content. For example, it uses a prompt such as, "Given the review: This product is easy to use and of high quality. The design is also good and I am very satisfied. It was a very good purchase. However, the delayed delivery was a negative point. Summarize it in 50 words or less."

[0139] The server automatically generates reports based on the summarized data. The generated summaries and extracted key phrases are arranged in a template format to create a report that clearly identifies important user needs. For example, a "Highlights of the Week" report might state that "Many customers praised the product design, but many complained about delivery speed." The templates are written in HTML or Markdown.

[0140] Finally, the server automatically distributes the generated reports to the designated recipients via email. A library for working with the SMTP protocol (e.g., smtplib) is used to send emails. A weekly report is then sent to each recipient. This report includes a summary of the customer feedback for the current week and recommendations based on that feedback.

[0141] In this way, the present invention automates the entire process of data collection, analysis, report generation, and distribution, making it possible to understand user needs efficiently and at low cost. Furthermore, because high-quality reports can be created without specialized knowledge, service providers can quickly implement improvement measures.

[0142] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0143] Step 1:

[0144] The server collects data using an application programming interface (API). Specifically, it obtains customer reviews and comments through the online shopping site's review API. The API call returns data in JSON format. The input is the API endpoint URL, and the output is customer reviews in JSON format. The specific operation of data collection is to send an HTTP GET request and receive a response.

[0145] Step 2:

[0146] The server stores the collected raw data in an intermediate database, which is implemented using MongoDB or MySQL. The input is customer reviews in JSON format, and the output is review information stored in the database. Specific operations include inserting the parsed JSON data into the database.

[0147] Step 3:

[0148] The server analyzes the collected data using a natural language processing model and generates a summary. The BERT and GPT series are used as generative AI models. The input is text data obtained from an intermediate database, and the output is summarized text. Specific operations include inputting the text data into the NLP model and obtaining the summary results. An example prompt is "Given the review: [review text], summarize it in 50 words or less."

[0149] Step 4:

[0150] The server automatically generates reports based on the summarized data. The reports are constructed based on a template format. The input is the summarized text, and the output is a structured report. Specifically, the server applies the summarized data to the template and assembles it in the required layout and format. The template can be in HTML or Markdown format.

[0151] Step 5:

[0152] The server automatically delivers the generated report to the specified recipient. The delivery method is email. The input is the generated report, and the output is the email sent to the recipient. Specific operations include sending email using an SMTP server. The smtplib library is used to send email.

[0153] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0154] This invention relates to a system that enables companies to quickly and inexpensively grasp user needs and emotions by efficiently collecting voice of customer (VOC), automatically analyzing and summarizing it, and combining it with an emotion engine that recognizes user emotions to generate reports and distribute them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0155] Data collection

[0156] The server uses an Application Programming Interface (API) to authenticate and collect data, specifically, API keys, API secrets, access tokens, and access token secrets to retrieve user reviews, comments, and feedback data from, for example, social media platforms, app stores, and customer service platforms, and stores this data in an intermediary database.

[0157] Specific examples

[0158] The server uses the SNS API to collect tweets with the hashtag "product review" and stores the data in an intermediate database.

[0159] Data processing and analysis

[0160] The server inputs the collected data into a natural language processing model for analysis. Specifically, it uses the Transformers library to utilize the latest generative AI models (e.g., BERT and GPT series) to summarize the collected text data.

[0161] Specific examples

[0162] The server inputs the collected tweets into a summarization model and obtains summary results, such as a large number of users requesting improvements to the app's user interface.

[0163] Emotion recognition

[0164] The server uses an emotion engine to recognize user emotions from the collected data, analyzes user comments and reviews, and assigns emotion tags such as positive, negative, and neutral.

[0165] Specific examples

[0166] The server processes tweets with a sentiment engine and classifies them as "positive" tweets such as "I really love the new design" and "I wish the app would stop crashing so often" as "negative" tweets.

[0167] Report Generation

[0168] The server automatically generates reports based on the summarized data and the user's emotional information. The generated summaries, key phrases, and emotional information are organized according to a template format, allowing users to create high-quality reports without specialized knowledge.

[0169] Specific examples

[0170] The server organizes the results of the summarization and sentiment analysis according to templates and generates a "Highlights of the Week" report such as "60% of users are happy with the new feature, but 30% have reported bugs."

[0171] Report Distribution

[0172] The server automatically delivers the generated reports to the designated recipients. It configures the SMTP server (address, port, username, password) and converts the report content into email format. The reports are sent to the relevant departments on a weekly basis.

[0173] Specific examples

[0174] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the current week's key user feedback and the results of sentiment analysis.

[0175] In this way, the system of the present invention automates the entire process of data collection, emotion recognition, analysis, report generation, and distribution, making it possible to grasp user needs and emotions efficiently and at low cost. This system allows service providers to quickly and accurately implement improvement measures that reflect users' emotions and needs.

[0176] The processing flow will be explained below.

[0177] Step 1:

[0178] The server uses the application programming interface (API) to authenticate for data collection. Specifically, it uses the API key, API secret, access token, and access token secret to access the API of Twitter and other SNSs and set authentication information.

[0179] Step 2:

[0180] The server uses the authentication information to collect tweets containing the target hashtags and keywords. It then connects to the Twitter API using the Tweepy library, retrieves 100 tweets containing "product reviews," and extracts information such as the text and posting date and time.

[0181] Step 3:

[0182] The server converts the collected tweet data into a Pandas data frame and stores it in an intermediate database, thereby preparing the data in a format that can be used for subsequent analysis.

[0183] Step 4:

[0184] The server then inputs the stored data into a natural language processing model to analyze the collected text data. Specifically, it uses the Transformers library to generate summaries using generative AI models such as BERT and GPT, which extract the key points of user opinions and feedback.

[0185] Step 5:

[0186] The server inputs the summarized text data into an emotion engine to recognize the user's emotions, and the emotion engine assigns emotion tags such as positive, negative, or neutral to each text.

[0187] Step 6:

[0188] The server extracts important key phrases from the collected data, using TF-IDF and other key phrase extraction techniques to identify key words and phrases in the summarized text.

[0189] Step 7:

[0190] The server automatically generates a report based on the summary, key phrases, and sentiment information. The generated summary, key phrases, and sentiment information are organized according to a template format, and a report is created that clearly shows important user needs and sentiment.

[0191] Step 8:

[0192] The server configures the SMTP server (address, port, username, password) and converts the generated reports into email format, ensuring that the reports contain the required information and analysis results.

[0193] Step 9:

[0194] The server connects to an SMTP server and automatically sends the generated report to a designated recipient, such as a service representative, on a weekly or set schedule.

[0195] The above steps automate the collection of VOC, emotion recognition, analysis, report generation, and distribution, making it possible to grasp user needs and emotions efficiently and at low cost. This system allows service providers to quickly and accurately implement service improvement measures that reflect user requests and emotions.

[0196] Example 2

[0197] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0198] It is difficult for companies to efficiently collect voice of customer (VOC) information quickly and at low cost, and analyze the data to understand user needs and emotions. Furthermore, since it takes a great deal of time and effort to extract the necessary information from the collected data, organize it, and gain insights, there is a need for an automated system that can be easily used even by personnel without specialized knowledge.

[0199] 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.

[0200] In this invention, the server includes means for collecting data using an application programming interface, means for storing the collected data in an intermediate database, means for summarizing the collected data using a natural language processing model, means for recognizing emotions from the summary results and the collected data, means for automatically generating a report based on the summary results and the emotion recognition results, and means for automatically delivering the generated report to a predetermined recipient. This efficiently automates the process from data collection to emotion recognition, and report generation and delivery, enabling companies to grasp user needs and emotions quickly and at low cost.

[0201] An "application programming interface" is a defined interface through which different software systems communicate with each other and exchange data.

[0202] "Means of collecting data" refers to the technologies and methods for obtaining information from various data sources, and typically involves using APIs.

[0203] An "intermediate database" is a database that temporarily stores collected data and uses it in subsequent processing.

[0204] A "natural language processing model" is an algorithm or model that uses machine learning or deep learning techniques to analyze and understand human language.

[0205] "Means of summarization" refers to techniques and methods for extracting the main content from long text data and displaying it in a short form.

[0206] "Means of emotion recognition" refers to algorithms and technologies for extracting emotional information such as positive, negative, and neutral from text data.

[0207] A "summary result" is a short document or phrase generated by a natural language processing model that succinctly conveys the main content of the text data.

[0208] The "emotion recognition result" is the result of emotion analysis on text data, and is classification information of the emotion that each text has.

[0209] "Means for automatically generating reports" refers to technologies and methods for automatically creating report-format documents based on analytical data such as summary results and emotion recognition results.

[0210] "Specified recipient" refers to a predetermined recipient of the report, meaning a specific department or person.

[0211] "Means for automated report delivery" refers to any technology or method for automatically sending generated reports to intended recipients.

[0212] This invention is a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, and then generates reports by combining it with an emotion engine that recognizes user emotions, and delivers them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0213] First, in the data collection stage, the server collects data using the application programming interface (API). Specifically, it uses the API key, API secret, access token, and access token secret to retrieve user reviews, comments, and feedback data from social media, app stores, and customer service platforms. This collected data is then stored in an intermediate database.

[0214] Specific examples

[0215] The server uses the SNS API to collect tweets with the hashtag "product review" and stores the data in an intermediate database.

[0216] The server then analyzes the collected data by inputting it into a natural language processing model. Specifically, it uses the Transformers library to utilize generative AI models (such as the BERT and GPT series). This process summarizes the collected text data.

[0217] Specific examples

[0218] The server inputs the collected tweets into a summarization model and summarizes the requests of many users for improvements to the app's user interface.

[0219] The server then uses an emotion engine to recognize user emotions from the collected data, performs sentiment analysis on user comments and reviews, and assigns emotion tags such as positive, negative, or neutral.

[0220] Specific examples

[0221] The server processes tweets with a sentiment engine and classifies them as "positive" tweets such as "I really love the new design" and "I wish the app would stop crashing so often" as "negative" tweets.

[0222] The server then automatically generates a report based on the summarized data and the user's emotional information. The report is organized according to a template format, allowing high-quality reports to be created without specialized knowledge.

[0223] Specific examples

[0224] The server organizes the results of the summarization and sentiment analysis according to templates and generates a "Highlights of the Week" report such as "60% of users are happy with the new feature, but 30% have reported bugs."

[0225] The server then automatically delivers the generated reports to the intended recipients using an SMTP server, with settings including address, port, username and password.

[0226] Specific examples

[0227] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the current week's key user feedback and the results of sentiment analysis.

[0228] Prompt Sentence Examples

[0229] 1. "Extract and summarize key points from customer reviews and feedback."

[0230] 2. "Analyze the sentiment of the collected tweets and categorize them as positive or negative."

[0231] 3. "Generate a report based on this week's user feedback and sentiment analysis results."

[0232] This system allows service providers to quickly and accurately implement improvement measures that reflect users' feelings and needs.

[0233] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0234] Step 1:

[0235] Data collection

[0236] The server authenticates to collect data using an Application Programming Interface (API), requiring API key, API secret, access token, and access token secret as inputs, and uses these to retrieve user reviews, comments, and feedback data from social media, app stores, and customer service platforms, and obtains raw data as output.

[0237] Specific actions

[0238] The server configures the Twitter API to collect tweets with the hashtag "product reviews." If authentication is successful, the tweet data (username, text, timestamp, etc.) is retrieved.

[0239] Step 2:

[0240] Storing data in an intermediate database

[0241] The server stores the collected data in an intermediate database. As input, the raw data collected in step 1 is used. As output, these data are stored in appropriate tables. The database uses an SQL database (e.g., MySQL or PostgreSQL).

[0242] Specific actions

[0243] The server uses SQL commands to insert the collected tweets into a table called "tweets," storing each tweet's ID, username, text, and timestamp.

[0244] Step 3:

[0245] Data analysis using natural language processing

[0246] The server inputs the data stored in the intermediate database into a natural language processing model for analysis. The input is text data retrieved from the database. The output is a summary result. This process uses the Transformers library and generative AI models (e.g., BERT, GPT series).

[0247] Specific actions

[0248] The server retrieves specific tweet text from the database and inputs it into the summarization model, which then analyzes it to obtain summaries, which are then stored in a new table called "summaries."

[0249] Step 4:

[0250] Emotion recognition

[0251] The server performs emotion recognition on the collected data. Text data is used as input. Emotion tags such as positive, negative, and neutral are generated as output. A sentiment analysis engine (e.g., VADER, TextBlob, etc.) is used.

[0252] Specific actions

[0253] The server feeds the tweet data into a sentiment analysis engine, which classifies each tweet into a sentiment tag (positive, negative, neutral), and adds the results to a new column in the "tweets" table.

[0254] Step 5:

[0255] Automatic report generation

[0256] The server automatically generates a report based on the summary results and emotion recognition results. The analysis results and emotion tags obtained in steps 3 and 4 are used as input. A template-based report is generated as output. A template engine (e.g., Jinja2) is used.

[0257] Specific actions

[0258] The server uses a template engine to embed the analysis results and sentiment tags into reports, creating templates that can be used to generate reports, such as "This week's highlights" that include "60% of users are happy with the new features, but 30% have reported bugs."

[0259] Step 6:

[0260] Automatic report delivery

[0261] The server automatically delivers the generated report to the specified recipients. As input, it uses the generated report and the SMTP server configuration information (address, port, username, password). As output, it sends an email to the recipients.

[0262] Specific actions

[0263] The server connects to the SMTP server, converts the generated report into an email format, and sends it to the email address of the service representative. For example, a schedule can be set to send a "report on interface improvement requests" every Monday.

[0264] By going through each step in this way, the process from data collection to emotion recognition, report generation and distribution is efficiently automated.

[0265] (Application example 2)

[0266] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0267] In recent years, collecting and analyzing voice of the customer (VOC) in brick-and-mortar stores has become essential for improving store operations and customer satisfaction. However, conventional methods have difficulty efficiently collecting and analyzing large volumes of data, and are unable to identify emotions, making it difficult to identify specific areas for improvement or take prompt action. This has made it difficult for brick-and-mortar store managers and staff to accurately understand customers' true needs and emotions.

[0268] 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.

[0269] In this invention, the server includes means for collecting data using an application programming interface, means for summarizing the collected data using a natural language processing model, means for identifying user emotions from the collected data using an emotion engine that recognizes user emotions, means for automatically generating a report based on the summary results and the emotion identification results, and means for automatically delivering the generated report to a specified recipient. This makes it possible to efficiently collect customer feedback and perform emotion analysis, thereby quickly identifying and addressing specific areas for improvement.

[0270] "Application Programming Interface" refers to the well-defined rules or protocols by which software applications communicate with each other.

[0271] "Data collection methods" refers to mechanisms for gathering information and feedback from multiple sources.

[0272] A "natural language processing model" refers to algorithms and technologies that allow computers to understand, analyze, and generate human language.

[0273] "Summarization methods" refers to the processes and techniques used to summarize collected information succinctly.

[0274] "Emotion engine" refers to algorithms and technologies for identifying user emotions from text data.

[0275] "Means for identifying emotions" refers to processes and techniques for extracting and classifying user emotions from text data.

[0276] "Means for automatically generating reports" refers to processes or technologies for automatically creating reports based on summary results and emotion identification results.

[0277] "Means for delivering reports" refers to the process or technology used to transmit generated reports to intended recipients.

[0278] "Key phrase extraction methods" refers to the processes and techniques used to identify and extract important words and phrases from text data.

[0279] "Template-based organization" refers to the process or technique of organizing and structuring information according to a predefined format.

[0280] The present invention is a system that efficiently collects customer feedback (VOC) from physical stores, automatically summarizes and analyzes the feedback, and compiles the results into a report and distributes it to relevant parties. This system is implemented as follows.

[0281] First, the server collects feedback data using an application programming interface (API), using authentication information such as an API key, API secret, or access token. Data collection targets include social media, feedback terminals, and customer service platforms.

[0282] The collected data is stored in an intermediate database. The server then summarizes this data using natural language processing models, including generative AI models such as BERT and the GPT series from the Transformers library. The summarized data is further tagged with sentiment tags, such as positive, negative, or neutral, using a sentiment engine. The sentiment engine, which is also part of natural language processing, uses the Transformers library for data analysis.

[0283] Once the specific data has been summarized and sentiment analyzed, a report is automatically generated based on it. The report contains summarized data, key phrases, and sentiment information. The report generation is based on templates, which organize the information and make the report content easy to read and understand.

[0284] The generated reports are automatically delivered to the specified recipients by the server. The SMTP server settings (address, port, username, password) are configured, the report contents are converted into email format, and sent to the specified email address. Reports are generally delivered weekly, allowing relevant staff to regularly check customer feedback.

[0285] Specific examples

[0286] For example, if a customer uses a feedback terminal to comment, "The product description is difficult to understand," the system collects this comment and stores it in an intermediate database. The server then inputs this data into a natural language processing model and outputs a summary, "Product description improvement request." The emotion engine then analyzes this summary and assigns the emotion tag "negative." Finally, a report is generated based on this information and automatically sent to the store manager.

[0287] Prompt Sentence Examples

[0288] Based on the following example prompt, the system will perform appropriate data collection, summarization, and sentiment analysis:

[0289] 1. Data Collection Prompt: "What is an effective way to collect customer reviews? Specifically, can you provide an example of using an API key to retrieve social media data?"

[0290] 2. Data Summarization Prompt: "What is the latest generative AI model that summarizes collected review data? Please include a Python code example."

[0291] 3. Sentiment Analysis Prompt: "What generative AI model would be suitable for performing sentiment analysis on review data and how would you use it?"

[0292] 4. Report Generation Prompt: "How can I automatically generate a report based on the results of my analysis? Can you give me an example of a template I could use?"

[0293] 5. Report Delivery Prompt: "How can I automatically deliver generated reports to a specified email address? Please provide a Python code example."

[0294] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0295] Step 1: Data collection

[0296] The server collects feedback data using an application programming interface (API). Specifically, it uses the API of social media and feedback devices to obtain customer comments and reviews using authentication information (API key, API secret, access token, etc.). At this time, the data is obtained in JSON format and stored in an intermediate database.

[0297] Input: API key, secret, access token

[0298] Output: Feedback data stored in an intermediate database

[0299] Step 2: Data Summarization

[0300] The server inputs the collected feedback data into a natural language processing model to summarize it. The generative AI models used here are the BERT and GPT series from the Transformers library. The model extracts important parts of the input text data and outputs a concise summary.

[0301] Input: Feedback data stored in the intermediate database

[0302] Output: Summarized text data

[0303] Step 3: Emotion Identification

[0304] The server then inputs the summarized text data into an emotion engine for sentiment analysis, which also uses the Transformers library to tag each piece of text with an emotion tag, such as positive, negative, or neutral.

[0305] Input: Summarized text data

[0306] Output: Text data with emotion tags

[0307] Step 4: Generate a report

[0308] The server automatically generates a report based on the summary and emotion identification results. This report is based on a template and organizes summaries, emotion tags, key phrases, etc.

[0309] Input: Text data with emotion tags

[0310] Output: Organized reports in template format

[0311] Step 5: Report Delivery

[0312] The server automatically delivers the generated report to the specified recipient by configuring the SMTP server settings (address, port, username, password), converting the report content into email format, and sending it.

[0313] Input: Reports organized in template format

[0314] Output: Report sent to recipient's (e.g. store manager) email address

[0315] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0316] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0317] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0318] [Second embodiment]

[0319] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0320] 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.

[0321] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0322] 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.

[0323] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0324] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0325] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0326] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0327] 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.

[0328] 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.

[0329] 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.

[0330] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0331] This invention relates to a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, generates reports, and distributes them to designated recipients, thereby enabling companies to quickly and inexpensively grasp user needs. This system is realized by automating each stage of processing, centered on a server.

[0332] Data collection

[0333] The server first collects data using application programming interfaces (APIs) and other methods, such as obtaining user reviews, comments, and feedback data from social media platforms, app stores, and customer service platforms. This is done using API access and scraping techniques, and the collected data is then stored in an intermediate database.

[0334] Specific examples

[0335] The server uses the social media platform's API to collect tweets with the hashtag "product review" and store them in a database.

[0336] Data processing and analysis

[0337] The server then analyzes the collected data using a natural language processing model. Specifically, it utilizes the latest generative AI models (e.g., deep learning models) to summarize the vast amount of collected text data. This allows it to concisely organize user opinions and feedback and generate summaries. Models such as the BERT and GPT series are used in this summarization process.

[0338] Specific examples

[0339] The server inputs the collected tweets into a summary model and obtains summary results such as "A large number of users are requesting improvements to the app's user interface."

[0340] Report Generation

[0341] The server automatically generates reports based on the summarized data. The generated summaries and extracted key phrases are arranged in a template format to create a report that clearly shows important user needs. This makes it easy for anyone, even those without specialized knowledge, to generate high-quality reports.

[0342] Specific examples

[0343] The server organizes the collected summaries according to a template and creates a report that states "over 50% of users are satisfied with the new feature" as a "highlight of the week."

[0344] Report Distribution

[0345] The server automatically delivers the final report to the designated recipients. The report is periodically sent to the relevant departments involved in the work. The delivery method is via email or the company's internal messaging system.

[0346] Specific examples

[0347] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the main user feedback for the current week and suggestions based on it.

[0348] In this way, the system of the present invention automates the entire process of data collection, analysis, report generation, and distribution, making it possible to grasp user needs efficiently and at low cost. This system allows service providers to accurately and quickly implement improvement measures that reflect user needs.

[0349] The processing flow will be explained below.

[0350] Step 1:

[0351] The server uses the application programming interface (API) to authenticate for data collection. Specifically, it accesses the Twitter API and sets authentication information using the API key, API secret, access token, and access token secret.

[0352] Step 2:

[0353] The server uses the authentication information to collect tweets containing the target hashtags and keywords. It connects to the Twitter API using the Tweepy library and retrieves 100 tweets containing the specific hashtag "product review."

[0354] Step 3:

[0355] The server converts the collected tweet data (text, posting date and time, etc.) into a Pandas data frame and stores it in an intermediate database, thereby preparing the data in a format that can be used for subsequent analysis.

[0356] Step 4:

[0357] The server inputs the text data stored in the intermediate database into a natural language processing model, specifically using the Transformers library and generative AI models such as BERT and GPT to generate summaries.

[0358] Step 5:

[0359] The server extracts the generated summary text and selects important key phrases. It uses TF-IDF and key phrase extraction techniques to identify key words and phrases from the summarized text.

[0360] Step 6:

[0361] The server automatically generates reports based on the summaries and key phrases, organized according to predefined templates and containing key user needs and statistics.

[0362] Step 7:

[0363] The server prepares the generated report for delivery to the intended recipient by configuring the SMTP server (address, port, username, password) and converting the report content to email format.

[0364] Step 8:

[0365] The server connects to the SMTP server and automatically sends the generated report to the relevant service personnel. Specifically, it sends a "report on interface improvement requests" to the personnel in charge by email every week.

[0366] The above steps automate the entire process from collecting VOCs to generating and distributing reports, making it possible to efficiently understand user needs.

[0367] Example 1

[0368] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0369] Companies are seeking a way to quickly and low-costly understand user needs. However, collecting and analyzing vast amounts of information from social media, review sites, and customer service platforms, generating useful reports based on that information, and then distributing those reports to relevant parties in a timely manner requires a great deal of time and effort. There is a need to automate these processes and provide a way to efficiently understand user needs.

[0370] 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.

[0371] In this invention, the server includes means for collecting data using an application programming interface, means for storing the collected data in an intermediate database, means for analyzing the stored data using a natural language processing model and generating a summary, means for automatically generating a report based on the summary result, and means for automatically distributing the generated report to a predetermined recipient. This automates the process of data collection, analysis, report generation, and distribution, making it possible to grasp user needs quickly and at low cost.

[0372] An "application programming interface" is a set of definitions and protocols that provide an interface through which software components can communicate with other software.

[0373] "Data collection means" are the methods and mechanisms used to obtain the required data from external sources.

[0374] The "intermediate database" is a database for temporarily storing collected data and transferring it to subsequent processing.

[0375] A "natural language processing model" is an algorithm or system that uses machine learning technology to understand and analyze human language.

[0376] A "means for generating a summary" is a method or technology for extracting important information from a huge amount of text data and summarizing it concisely.

[0377] "Means for automatically generating a report" refers to a method or system for automatically creating a report in a predetermined format based on the analysis results or summary.

[0378] "Means for automatically delivering to recipients" refers to methods or techniques for automatically sending the generated report to pre-designated recipients.

[0379] This invention relates to a system that enables companies to quickly and inexpensively grasp user needs by efficiently collecting voice of customer (VOC), automatically analyzing and summarizing it, generating reports, and distributing them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0380] First, the server uses application programming interfaces (APIs) and scraping technology to collect user reviews, comments, and feedback data from social media, review sites, and customer service platforms. The collected data is then stored in an intermediate database. For example, the server uses the API of a social media platform to collect tweets with the hashtag "product review" and store them in a database.

[0381] The server then sends the collected text data to a natural language processing model (NLP model) to analyze user opinions and feedback. Specifically, it uses deep learning models (e.g., BERT and GPT series) to extract summaries and key phrases. This allows it to organize important information from large amounts of data and convert it into an interpretable format. For example, the server inputs the collected tweets into a summarization model and obtains a summary result such as, "Many users are requesting improvements to the app's user interface."

[0382] The server then automatically generates reports based on the analyzed data. By inserting summarized data and extracted key phrases into pre-defined templates, reports that clearly indicate user needs are created. For example, the server might create a "Highlights of the Week" report that states, "More than 50% of users are satisfied with the new features."

[0383] Finally, the server automatically distributes the generated report to the designated recipients via email or the company's internal messaging system. For example, every Monday the server sends a "Report on Interface Improvement Requests" to the email address of a service representative. This email includes the main user feedback for the current week and suggestions based on that feedback.

[0384] Specific examples

[0385] The server sends a request to the SNS API (e.g., Twitter API) at a specific time every day, retrieves posts containing a specific hashtag such as "product review," and stores them in an intermediate database.

[0386] The server inputs the reviews and comments stored in the database into an NLP model (for example, the BERT model in Hugging Face's Transformers library) and stores the results as summaries and key phrases in a separate table.

[0387] The server extracts the analysis results, embeds the data in a template called "Highlights of the Week," and generates a report such as "More than 50% of users are satisfied with the new feature."

[0388] The server emails the generated report based on a pre-specified list of recipients.

[0389] Prompt Sentence Examples

[0390] This week's highlights:

[0391] Summary: Many users are requesting improvements to the app's user interface.

[0392] In this way, the system of the present invention automates the process of data collection, analysis, report generation and distribution, making it possible to grasp user needs quickly and at low cost.

[0393] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0394] Step 1: Data collection

[0395] The server uses application programming interfaces (APIs) and scraping techniques to collect user reviews, comments, and feedback data from social media platforms, review sites, and customer service platforms. Specifically, the APIs are used to retrieve posts with specific hashtags (e.g., "product reviews") from social media platforms. The server then stores the retrieved data in an intermediary database.

[0396] Input: URL or API endpoint of social media or review site

[0397] Output: Feedback data stored in an intermediate database

[0398] Step 2: Save data

[0399] The server structures and stores the collected raw data in an intermediate database, which organizes the data and prepares it for further processing. In this step, data insertion operations are performed using SQL.

[0400] Input: Raw data collected

[0401] Output: An intermediate database containing structured data

[0402] Step 3: Analyze the data

[0403] The server sends the stored text data to a natural language processing (NLP) model to analyze user opinions and feedback. Specifically, it uses deep learning models (e.g., BERT and GPT series) to extract summaries and key phrases, making it easier to grasp the gist of the data.

[0404] Input: Text data stored in the intermediate database

[0405] Output: Summary sentences and key phrases analyzed by the NLP model

[0406] Specific operation: The server retrieves the text data one by one, inputs prompts such as "What is the main opinion of this review?" into the NLP model, and saves the results in another table.

[0407] Step 4: Generate a report

[0408] The server automatically generates reports based on the analyzed data. It inserts summarized data and extracted key phrases into pre-prepared template formats to create reports that clearly show the user's needs. A template engine is used in this process.

[0409] Input: Summary sentences and key phrases from the NLP model

[0410] Output: Formatted report

[0411] What happens: The server inserts the summary and key phrases into a template and generates a report as "Highlights of the Week."

[0412] Step 5: Report Delivery

[0413] The server automatically distributes the generated reports to the designated recipients via email or an internal messaging system, based on a pre-defined schedule.

[0414] Input: Generated report

[0415] Output: Report delivered to recipients

[0416] What it does: The server generates a report and sends it to the service representative's email address every Monday, containing key user feedback and suggestions based on it.

[0417] (Application example 1)

[0418] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0419] Traditional methods for collecting, analyzing, and generating reports on voice of customer (VOC) required a lot of manual work, which was time-consuming and costly, making it difficult to quickly and efficiently understand user needs. Furthermore, generating and distributing reports required human labor, which often led to a lack of timely information sharing. This made it difficult for service providers to quickly implement appropriate improvement measures, resulting in delays in improving customer satisfaction.

[0420] 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.

[0421] In this invention, the server includes means for collecting data using an application programming interface, means for summarizing the collected data using a natural language processing model, means for automatically generating a report based on the summary results, means for automatically distributing the generated report to a predetermined recipient, and means for summarizing the collected data based on its contents, generating a report according to a predetermined schedule, and distributing the report to a predetermined recipient via email, thereby enabling user needs to be grasped quickly and efficiently at low cost.

[0422] An "application programming interface" is something that provides an interface between software or applications, enabling the exchange of data and functions.

[0423] "Data" refers to information such as reviews, comments, and feedback obtained from customers.

[0424] A "natural language processing model" is a machine learning model for analyzing and summarizing text data, and includes the BERT and GPT series.

[0425] "Summarizing" means extracting important parts from multiple text data and summarizing them in a concise form.

[0426] A "report" is a document generated from summarized data that shows the results of organizing and analyzing user needs and feedback.

[0427] "Automatically generating" means that the system generates documents according to specified algorithms or templates without human intervention.

[0428] "Intended Recipient" means a person or department that has been pre-designated to receive the report.

[0429] "Automatically delivering" means that the system sends information as emails or messages based on set schedules or conditions, without manual intervention.

[0430] A "schedule" refers to a predetermined time and order plan for performing a series of tasks or processes.

[0431] "Email" is a means of sending and receiving text data and attachments in digital form over the Internet.

[0432] This invention relates to a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, generates reports, and distributes them to designated recipients, thereby enabling companies to quickly and inexpensively understand user needs. This system is targeted at collecting and analyzing customer feedback on online shopping sites.

[0433] The server first collects data using an application programming interface (API). Specifically, it retrieves customer reviews and comments from the review API of an online shopping site. Scraping technology can also be used, for example, to automatically collect reviews on a web page. The collected data is then stored in an intermediate database. Examples of databases include MongoDB and MySQL.

[0434] Next, the server inputs the collected data into a natural language processing model (NLP model) for analysis and summarization. This summarization process uses generative AI models such as BERT and the GPT series. Specifically, user reviews and comments are taken as input, and the summarization model concisely summarizes their content. For example, it uses a prompt such as, "Given the review: This product is easy to use and of high quality. The design is also good and I am very satisfied. It was a very good purchase. However, the delayed delivery was a negative point. Summarize it in 50 words or less."

[0435] The server automatically generates reports based on the summarized data. The generated summaries and extracted key phrases are arranged in a template format to create a report that clearly identifies important user needs. For example, a "Highlights of the Week" report might state that "Many customers praised the product design, but many complained about delivery speed." The templates are written in HTML or Markdown.

[0436] Finally, the server automatically distributes the generated reports to the designated recipients via email. A library for working with the SMTP protocol (e.g., smtplib) is used to send emails. A weekly report is then sent to each recipient. This report includes a summary of the customer feedback for the current week and recommendations based on that feedback.

[0437] In this way, the present invention automates the entire process of data collection, analysis, report generation, and distribution, making it possible to understand user needs efficiently and at low cost. Furthermore, because high-quality reports can be created without specialized knowledge, service providers can quickly implement improvement measures.

[0438] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0439] Step 1:

[0440] The server collects data using an application programming interface (API). Specifically, it obtains customer reviews and comments through the online shopping site's review API. The API call returns data in JSON format. The input is the API endpoint URL, and the output is customer reviews in JSON format. The specific operation of data collection is to send an HTTP GET request and receive a response.

[0441] Step 2:

[0442] The server stores the collected raw data in an intermediate database, which is implemented using MongoDB or MySQL. The input is customer reviews in JSON format, and the output is review information stored in the database. Specific operations include inserting the parsed JSON data into the database.

[0443] Step 3:

[0444] The server analyzes the collected data using a natural language processing model and generates a summary. The BERT and GPT series are used as generative AI models. The input is text data obtained from an intermediate database, and the output is summarized text. Specific operations include inputting the text data into the NLP model and obtaining the summary results. An example prompt is "Given the review: [review text], summarize it in 50 words or less."

[0445] Step 4:

[0446] The server automatically generates reports based on the summarized data. The reports are constructed based on a template format. The input is the summarized text, and the output is a structured report. Specifically, the server applies the summarized data to the template and assembles it in the required layout and format. The template can be in HTML or Markdown format.

[0447] Step 5:

[0448] The server automatically delivers the generated report to the specified recipient. The delivery method is email. The input is the generated report, and the output is the email sent to the recipient. Specific operations include sending email using an SMTP server. The smtplib library is used to send email.

[0449] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0450] This invention relates to a system that enables companies to quickly and inexpensively grasp user needs and emotions by efficiently collecting voice of customer (VOC), automatically analyzing and summarizing it, and combining it with an emotion engine that recognizes user emotions to generate reports and distribute them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0451] Data collection

[0452] The server uses an Application Programming Interface (API) to authenticate and collect data, specifically, API keys, API secrets, access tokens, and access token secrets to retrieve user reviews, comments, and feedback data from, for example, social media platforms, app stores, and customer service platforms, and stores this data in an intermediary database.

[0453] Specific examples

[0454] The server uses the SNS API to collect tweets with the hashtag "product review" and stores the data in an intermediate database.

[0455] Data processing and analysis

[0456] The server inputs the collected data into a natural language processing model for analysis. Specifically, it uses the Transformers library to utilize the latest generative AI models (e.g., BERT and GPT series) to summarize the collected text data.

[0457] Specific examples

[0458] The server inputs the collected tweets into a summarization model and obtains summary results, such as a large number of users requesting improvements to the app's user interface.

[0459] Emotion recognition

[0460] The server uses an emotion engine to recognize user emotions from the collected data, analyzes user comments and reviews, and assigns emotion tags such as positive, negative, and neutral.

[0461] Specific examples

[0462] The server processes tweets with a sentiment engine and classifies them as "positive" tweets such as "I really love the new design" and "I wish the app would stop crashing so often" as "negative" tweets.

[0463] Report Generation

[0464] The server automatically generates reports based on the summarized data and the user's emotional information. The generated summaries, key phrases, and emotional information are organized according to a template format, allowing users to create high-quality reports without specialized knowledge.

[0465] Specific examples

[0466] The server organizes the results of the summarization and sentiment analysis according to templates and generates a "Highlights of the Week" report such as "60% of users are happy with the new feature, but 30% have reported bugs."

[0467] Report Distribution

[0468] The server automatically delivers the generated reports to the designated recipients. It configures the SMTP server (address, port, username, password) and converts the report content into email format. The reports are sent to the relevant departments on a weekly basis.

[0469] Specific examples

[0470] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the current week's key user feedback and the results of sentiment analysis.

[0471] In this way, the system of the present invention automates the entire process of data collection, emotion recognition, analysis, report generation, and distribution, making it possible to grasp user needs and emotions efficiently and at low cost. This system allows service providers to quickly and accurately implement improvement measures that reflect users' emotions and needs.

[0472] The processing flow will be explained below.

[0473] Step 1:

[0474] The server uses the application programming interface (API) to authenticate for data collection. Specifically, it uses the API key, API secret, access token, and access token secret to access the API of Twitter and other SNSs and set authentication information.

[0475] Step 2:

[0476] The server uses the authentication information to collect tweets containing the target hashtags and keywords. It then connects to the Twitter API using the Tweepy library, retrieves 100 tweets containing "product reviews," and extracts information such as the text and posting date and time.

[0477] Step 3:

[0478] The server converts the collected tweet data into a Pandas data frame and stores it in an intermediate database, thereby preparing the data in a format that can be used for subsequent analysis.

[0479] Step 4:

[0480] The server then inputs the stored data into a natural language processing model to analyze the collected text data. Specifically, it uses the Transformers library to generate summaries using generative AI models such as BERT and GPT, which extract the key points of user opinions and feedback.

[0481] Step 5:

[0482] The server inputs the summarized text data into an emotion engine to recognize the user's emotions, and the emotion engine assigns emotion tags such as positive, negative, or neutral to each text.

[0483] Step 6:

[0484] The server extracts important key phrases from the collected data, using TF-IDF and other key phrase extraction techniques to identify key words and phrases in the summarized text.

[0485] Step 7:

[0486] The server automatically generates a report based on the summary, key phrases, and sentiment information. The generated summary, key phrases, and sentiment information are organized according to a template format, and a report is created that clearly shows important user needs and sentiment.

[0487] Step 8:

[0488] The server configures the SMTP server (address, port, username, password) and converts the generated reports into email format, ensuring that the reports contain the required information and analysis results.

[0489] Step 9:

[0490] The server connects to an SMTP server and automatically sends the generated report to a designated recipient, such as a service representative, on a weekly or set schedule.

[0491] The above steps automate the collection of VOC, emotion recognition, analysis, report generation, and distribution, making it possible to grasp user needs and emotions efficiently and at low cost. This system allows service providers to quickly and accurately implement service improvement measures that reflect user requests and emotions.

[0492] Example 2

[0493] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0494] It is difficult for companies to efficiently collect voice of customer (VOC) information quickly and at low cost, and analyze the data to understand user needs and emotions. Furthermore, since it takes a great deal of time and effort to extract the necessary information from the collected data, organize it, and gain insights, there is a need for an automated system that can be easily used even by personnel without specialized knowledge.

[0495] 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.

[0496] In this invention, the server includes means for collecting data using an application programming interface, means for storing the collected data in an intermediate database, means for summarizing the collected data using a natural language processing model, means for recognizing emotions from the summary results and the collected data, means for automatically generating a report based on the summary results and the emotion recognition results, and means for automatically delivering the generated report to a predetermined recipient. This efficiently automates the process from data collection to emotion recognition, and report generation and delivery, enabling companies to grasp user needs and emotions quickly and at low cost.

[0497] An "application programming interface" is a defined interface through which different software systems communicate with each other and exchange data.

[0498] "Means of collecting data" refers to the technologies and methods for obtaining information from various data sources, and typically involves using APIs.

[0499] An "intermediate database" is a database that temporarily stores collected data and uses it in subsequent processing.

[0500] A "natural language processing model" is an algorithm or model that uses machine learning or deep learning techniques to analyze and understand human language.

[0501] "Means of summarization" refers to techniques and methods for extracting the main content from long text data and displaying it in a short form.

[0502] "Means of emotion recognition" refers to algorithms and technologies for extracting emotional information such as positive, negative, and neutral from text data.

[0503] A "summary result" is a short document or phrase generated by a natural language processing model that succinctly conveys the main content of the text data.

[0504] The "emotion recognition result" is the result of emotion analysis on text data, and is classification information of the emotion that each text has.

[0505] "Means for automatically generating reports" refers to technologies and methods for automatically creating report-format documents based on analytical data such as summary results and emotion recognition results.

[0506] "Specified recipient" refers to a predetermined recipient of the report, meaning a specific department or person.

[0507] "Means for automated report delivery" refers to any technology or method for automatically sending generated reports to intended recipients.

[0508] This invention is a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, and then generates reports by combining it with an emotion engine that recognizes user emotions, and delivers them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0509] First, in the data collection stage, the server collects data using the application programming interface (API). Specifically, it uses the API key, API secret, access token, and access token secret to retrieve user reviews, comments, and feedback data from social media, app stores, and customer service platforms. This collected data is then stored in an intermediate database.

[0510] Specific examples

[0511] The server uses the SNS API to collect tweets with the hashtag "product review" and stores the data in an intermediate database.

[0512] The server then analyzes the collected data by inputting it into a natural language processing model. Specifically, it uses the Transformers library to utilize generative AI models (such as the BERT and GPT series). This process summarizes the collected text data.

[0513] Specific examples

[0514] The server inputs the collected tweets into a summarization model and summarizes the requests of many users for improvements to the app's user interface.

[0515] The server then uses an emotion engine to recognize user emotions from the collected data, performs sentiment analysis on user comments and reviews, and assigns emotion tags such as positive, negative, or neutral.

[0516] Specific examples

[0517] The server processes tweets with a sentiment engine and classifies them as "positive" tweets such as "I really love the new design" and "I wish the app would stop crashing so often" as "negative" tweets.

[0518] The server then automatically generates a report based on the summarized data and the user's emotional information. The report is organized according to a template format, allowing high-quality reports to be created without specialized knowledge.

[0519] Specific examples

[0520] The server organizes the results of the summarization and sentiment analysis according to templates and generates a "Highlights of the Week" report such as "60% of users are happy with the new feature, but 30% have reported bugs."

[0521] The server then automatically delivers the generated reports to the intended recipients using an SMTP server, with settings including address, port, username and password.

[0522] Specific examples

[0523] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the current week's key user feedback and the results of sentiment analysis.

[0524] Prompt Sentence Examples

[0525] 1. "Extract and summarize key points from customer reviews and feedback."

[0526] 2. "Analyze the sentiment of the collected tweets and categorize them as positive or negative."

[0527] 3. "Generate a report based on this week's user feedback and sentiment analysis results."

[0528] This system allows service providers to quickly and accurately implement improvement measures that reflect users' feelings and needs.

[0529] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0530] Step 1:

[0531] Data collection

[0532] The server authenticates to collect data using an Application Programming Interface (API), requiring API key, API secret, access token, and access token secret as inputs, and uses these to retrieve user reviews, comments, and feedback data from social media, app stores, and customer service platforms, and obtains raw data as output.

[0533] Specific actions

[0534] The server configures the Twitter API to collect tweets with the hashtag "product reviews." If authentication is successful, the tweet data (username, text, timestamp, etc.) is retrieved.

[0535] Step 2:

[0536] Storing data in an intermediate database

[0537] The server stores the collected data in an intermediate database. As input, the raw data collected in step 1 is used. As output, these data are stored in appropriate tables. The database uses an SQL database (e.g., MySQL or PostgreSQL).

[0538] Specific actions

[0539] The server uses SQL commands to insert the collected tweets into a table called "tweets," storing each tweet's ID, username, text, and timestamp.

[0540] Step 3:

[0541] Data analysis using natural language processing

[0542] The server inputs the data stored in the intermediate database into a natural language processing model for analysis. The input is text data retrieved from the database. The output is a summary result. This process uses the Transformers library and generative AI models (e.g., BERT, GPT series).

[0543] Specific actions

[0544] The server retrieves specific tweet text from the database and inputs it into the summarization model, which then analyzes it to obtain summaries, which are then stored in a new table called "summaries."

[0545] Step 4:

[0546] Emotion recognition

[0547] The server performs emotion recognition on the collected data. Text data is used as input. Emotion tags such as positive, negative, and neutral are generated as output. A sentiment analysis engine (e.g., VADER, TextBlob, etc.) is used.

[0548] Specific actions

[0549] The server feeds the tweet data into a sentiment analysis engine, which classifies each tweet into a sentiment tag (positive, negative, neutral), and adds the results to a new column in the "tweets" table.

[0550] Step 5:

[0551] Automatic report generation

[0552] The server automatically generates a report based on the summary results and emotion recognition results. The analysis results and emotion tags obtained in steps 3 and 4 are used as input. A template-based report is generated as output. A template engine (e.g., Jinja2) is used.

[0553] Specific actions

[0554] The server uses a template engine to embed the analysis results and sentiment tags into reports, creating templates that can be used to generate reports, such as "This week's highlights" that include "60% of users are happy with the new features, but 30% have reported bugs."

[0555] Step 6:

[0556] Automatic report delivery

[0557] The server automatically delivers the generated report to the specified recipients. As input, it uses the generated report and the SMTP server configuration information (address, port, username, password). As output, it sends an email to the recipients.

[0558] Specific actions

[0559] The server connects to the SMTP server, converts the generated report into an email format, and sends it to the email address of the service representative. For example, a schedule can be set to send a "report on interface improvement requests" every Monday.

[0560] By going through each step in this way, the process from data collection to emotion recognition, report generation and distribution is efficiently automated.

[0561] (Application example 2)

[0562] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0563] In recent years, collecting and analyzing voice of the customer (VOC) in brick-and-mortar stores has become essential for improving store operations and customer satisfaction. However, conventional methods have difficulty efficiently collecting and analyzing large volumes of data, and are unable to identify emotions, making it difficult to identify specific areas for improvement or take prompt action. This has made it difficult for brick-and-mortar store managers and staff to accurately understand customers' true needs and emotions.

[0564] 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.

[0565] In this invention, the server includes means for collecting data using an application programming interface, means for summarizing the collected data using a natural language processing model, means for identifying user emotions from the collected data using an emotion engine that recognizes user emotions, means for automatically generating a report based on the summary results and the emotion identification results, and means for automatically delivering the generated report to a specified recipient. This makes it possible to efficiently collect customer feedback and perform emotion analysis, thereby quickly identifying and addressing specific areas for improvement.

[0566] "Application Programming Interface" refers to the well-defined rules or protocols by which software applications communicate with each other.

[0567] "Data collection methods" refers to mechanisms for gathering information and feedback from multiple sources.

[0568] A "natural language processing model" refers to algorithms and technologies that allow computers to understand, analyze, and generate human language.

[0569] "Summarization methods" refers to the processes and techniques used to summarize collected information succinctly.

[0570] "Emotion engine" refers to algorithms and technologies for identifying user emotions from text data.

[0571] "Means for identifying emotions" refers to processes and techniques for extracting and classifying user emotions from text data.

[0572] "Means for automatically generating reports" refers to processes or technologies for automatically creating reports based on summary results and emotion identification results.

[0573] "Means for delivering reports" refers to the process or technology used to transmit generated reports to intended recipients.

[0574] "Key phrase extraction methods" refers to the processes and techniques used to identify and extract important words and phrases from text data.

[0575] "Template-based organization" refers to the process or technique of organizing and structuring information according to a predefined format.

[0576] The present invention is a system that efficiently collects customer feedback (VOC) from physical stores, automatically summarizes and analyzes the feedback, and compiles the results into a report and distributes it to relevant parties. This system is implemented as follows.

[0577] First, the server collects feedback data using an application programming interface (API), using authentication information such as an API key, API secret, or access token. Data collection targets include social media, feedback terminals, and customer service platforms.

[0578] The collected data is stored in an intermediate database. The server then summarizes this data using natural language processing models, including generative AI models such as BERT and the GPT series from the Transformers library. The summarized data is further tagged with sentiment tags, such as positive, negative, or neutral, using a sentiment engine. The sentiment engine, which is also part of natural language processing, uses the Transformers library for data analysis.

[0579] Once the specific data has been summarized and sentiment analyzed, a report is automatically generated based on it. The report contains summarized data, key phrases, and sentiment information. The report generation is based on templates, which organize the information and make the report content easy to read and understand.

[0580] The generated reports are automatically delivered to the specified recipients by the server. The SMTP server settings (address, port, username, password) are configured, the report contents are converted into email format, and sent to the specified email address. Reports are generally delivered weekly, allowing relevant staff to regularly check customer feedback.

[0581] Specific examples

[0582] For example, if a customer uses a feedback terminal to comment, "The product description is difficult to understand," the system collects this comment and stores it in an intermediate database. The server then inputs this data into a natural language processing model and outputs a summary, "Product description improvement request." The emotion engine then analyzes this summary and assigns the emotion tag "negative." Finally, a report is generated based on this information and automatically sent to the store manager.

[0583] Prompt Sentence Examples

[0584] Based on the following example prompt, the system will perform appropriate data collection, summarization, and sentiment analysis:

[0585] 1. Data Collection Prompt: "What is an effective way to collect customer reviews? Specifically, can you provide an example of using an API key to retrieve social media data?"

[0586] 2. Data Summarization Prompt: "What is the latest generative AI model that summarizes collected review data? Please include a Python code example."

[0587] 3. Sentiment Analysis Prompt: "What generative AI model would be suitable for performing sentiment analysis on review data and how would you use it?"

[0588] 4. Report Generation Prompt: "How can I automatically generate a report based on the results of my analysis? Can you give me an example of a template I could use?"

[0589] 5. Report Delivery Prompt: "How can I automatically deliver generated reports to a specified email address? Please provide a Python code example."

[0590] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0591] Step 1: Data collection

[0592] The server collects feedback data using an application programming interface (API). Specifically, it uses the API of social media and feedback devices to obtain customer comments and reviews using authentication information (API key, API secret, access token, etc.). At this time, the data is obtained in JSON format and stored in an intermediate database.

[0593] Input: API key, secret, access token

[0594] Output: Feedback data stored in an intermediate database

[0595] Step 2: Data Summarization

[0596] The server inputs the collected feedback data into a natural language processing model to summarize it. The generative AI models used here are the BERT and GPT series from the Transformers library. The model extracts important parts of the input text data and outputs a concise summary.

[0597] Input: Feedback data stored in the intermediate database

[0598] Output: Summarized text data

[0599] Step 3: Emotion Identification

[0600] The server then inputs the summarized text data into an emotion engine for sentiment analysis, which also uses the Transformers library to tag each piece of text with an emotion tag, such as positive, negative, or neutral.

[0601] Input: Summarized text data

[0602] Output: Text data with emotion tags

[0603] Step 4: Generate a report

[0604] The server automatically generates a report based on the summary and emotion identification results. This report is based on a template and organizes summaries, emotion tags, key phrases, etc.

[0605] Input: Text data with emotion tags

[0606] Output: Organized reports in template format

[0607] Step 5: Report Delivery

[0608] The server automatically delivers the generated report to the specified recipient by configuring the SMTP server settings (address, port, username, password), converting the report content into email format, and sending it.

[0609] Input: Reports organized in template format

[0610] Output: Report sent to recipient's (e.g. store manager) email address

[0611] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0612] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0613] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0614] [Third embodiment]

[0615] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0616] 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.

[0617] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0618] 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.

[0619] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0620] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0621] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0622] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0623] 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.

[0624] 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.

[0625] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0626] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0627] This invention relates to a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, generates reports, and distributes them to designated recipients, thereby enabling companies to quickly and inexpensively grasp user needs. This system is realized by automating each stage of processing, centered on a server.

[0628] Data collection

[0629] The server first collects data using application programming interfaces (APIs) and other methods, such as obtaining user reviews, comments, and feedback data from social media platforms, app stores, and customer service platforms. This is done using API access and scraping techniques, and the collected data is then stored in an intermediate database.

[0630] Specific examples

[0631] The server uses the social media platform's API to collect tweets with the hashtag "product review" and store them in a database.

[0632] Data processing and analysis

[0633] The server then analyzes the collected data using a natural language processing model. Specifically, it utilizes the latest generative AI models (e.g., deep learning models) to summarize the vast amount of collected text data. This allows it to concisely organize user opinions and feedback and generate summaries. Models such as the BERT and GPT series are used in this summarization process.

[0634] Specific examples

[0635] The server inputs the collected tweets into a summary model and obtains summary results such as "A large number of users are requesting improvements to the app's user interface."

[0636] Report Generation

[0637] The server automatically generates reports based on the summarized data. The generated summaries and extracted key phrases are arranged in a template format to create a report that clearly shows important user needs. This makes it easy for anyone, even those without specialized knowledge, to generate high-quality reports.

[0638] Specific examples

[0639] The server organizes the collected summaries according to a template and creates a report that states "over 50% of users are satisfied with the new feature" as a "highlight of the week."

[0640] Report Distribution

[0641] The server automatically delivers the final report to the designated recipients. The report is periodically sent to the relevant departments involved in the work. The delivery method is via email or the company's internal messaging system.

[0642] Specific examples

[0643] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the main user feedback for the current week and suggestions based on it.

[0644] In this way, the system of the present invention automates the entire process of data collection, analysis, report generation, and distribution, making it possible to grasp user needs efficiently and at low cost. This system allows service providers to accurately and quickly implement improvement measures that reflect user needs.

[0645] The processing flow will be explained below.

[0646] Step 1:

[0647] The server uses the application programming interface (API) to authenticate for data collection. Specifically, it accesses the Twitter API and sets authentication information using the API key, API secret, access token, and access token secret.

[0648] Step 2:

[0649] The server uses the authentication information to collect tweets containing the target hashtags and keywords. It connects to the Twitter API using the Tweepy library and retrieves 100 tweets containing the specific hashtag "product review."

[0650] Step 3:

[0651] The server converts the collected tweet data (text, posting date and time, etc.) into a Pandas data frame and stores it in an intermediate database, thereby preparing the data in a format that can be used for subsequent analysis.

[0652] Step 4:

[0653] The server inputs the text data stored in the intermediate database into a natural language processing model, specifically using the Transformers library and generative AI models such as BERT and GPT to generate summaries.

[0654] Step 5:

[0655] The server extracts the generated summary text and selects important key phrases. It uses TF-IDF and key phrase extraction techniques to identify key words and phrases from the summarized text.

[0656] Step 6:

[0657] The server automatically generates reports based on the summaries and key phrases, organized according to predefined templates and containing key user needs and statistics.

[0658] Step 7:

[0659] The server prepares the generated report for delivery to the intended recipient by configuring the SMTP server (address, port, username, password) and converting the report content to email format.

[0660] Step 8:

[0661] The server connects to the SMTP server and automatically sends the generated report to the relevant service personnel. Specifically, it sends a "report on interface improvement requests" to the personnel in charge by email every week.

[0662] The above steps automate the entire process from collecting VOCs to generating and distributing reports, making it possible to efficiently understand user needs.

[0663] Example 1

[0664] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0665] Companies are seeking a way to quickly and low-costly understand user needs. However, collecting and analyzing vast amounts of information from social media, review sites, and customer service platforms, generating useful reports based on that information, and then distributing those reports to relevant parties in a timely manner requires a great deal of time and effort. There is a need to automate these processes and provide a way to efficiently understand user needs.

[0666] 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.

[0667] In this invention, the server includes means for collecting data using an application programming interface, means for storing the collected data in an intermediate database, means for analyzing the stored data using a natural language processing model and generating a summary, means for automatically generating a report based on the summary result, and means for automatically distributing the generated report to a predetermined recipient. This automates the process of data collection, analysis, report generation, and distribution, making it possible to grasp user needs quickly and at low cost.

[0668] An "application programming interface" is a set of definitions and protocols that provide an interface through which software components can communicate with other software.

[0669] "Data collection means" are the methods and mechanisms used to obtain the required data from external sources.

[0670] The "intermediate database" is a database for temporarily storing collected data and transferring it to subsequent processing.

[0671] A "natural language processing model" is an algorithm or system that uses machine learning technology to understand and analyze human language.

[0672] A "means for generating a summary" is a method or technology for extracting important information from a huge amount of text data and summarizing it concisely.

[0673] "Means for automatically generating a report" refers to a method or system for automatically creating a report in a predetermined format based on the analysis results or summary.

[0674] "Means for automatically delivering to recipients" refers to methods or techniques for automatically sending the generated report to pre-designated recipients.

[0675] This invention relates to a system that enables companies to quickly and inexpensively grasp user needs by efficiently collecting voice of customer (VOC), automatically analyzing and summarizing it, generating reports, and distributing them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0676] First, the server uses application programming interfaces (APIs) and scraping technology to collect user reviews, comments, and feedback data from social media, review sites, and customer service platforms. The collected data is then stored in an intermediate database. For example, the server uses the API of a social media platform to collect tweets with the hashtag "product review" and store them in a database.

[0677] The server then sends the collected text data to a natural language processing model (NLP model) to analyze user opinions and feedback. Specifically, it uses deep learning models (e.g., BERT and GPT series) to extract summaries and key phrases. This allows it to organize important information from large amounts of data and convert it into an interpretable format. For example, the server inputs the collected tweets into a summarization model and obtains a summary result such as, "Many users are requesting improvements to the app's user interface."

[0678] The server then automatically generates reports based on the analyzed data. By inserting summarized data and extracted key phrases into pre-defined templates, reports that clearly indicate user needs are created. For example, the server might create a "Highlights of the Week" report that states, "More than 50% of users are satisfied with the new features."

[0679] Finally, the server automatically distributes the generated report to the designated recipients via email or the company's internal messaging system. For example, every Monday the server sends a "Report on Interface Improvement Requests" to the email address of a service representative. This email includes the main user feedback for the current week and suggestions based on that feedback.

[0680] Specific examples

[0681] The server sends a request to the SNS API (e.g., Twitter API) at a specific time every day, retrieves posts containing a specific hashtag such as "product review," and stores them in an intermediate database.

[0682] The server inputs the reviews and comments stored in the database into an NLP model (for example, the BERT model in Hugging Face's Transformers library) and stores the results as summaries and key phrases in a separate table.

[0683] The server extracts the analysis results, embeds the data in a template called "Highlights of the Week," and generates a report such as "More than 50% of users are satisfied with the new feature."

[0684] The server emails the generated report based on a pre-specified list of recipients.

[0685] Prompt Sentence Examples

[0686] This week's highlights:

[0687] Summary: Many users are requesting improvements to the app's user interface.

[0688] In this way, the system of the present invention automates the process of data collection, analysis, report generation and distribution, making it possible to grasp user needs quickly and at low cost.

[0689] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0690] Step 1: Data collection

[0691] The server uses application programming interfaces (APIs) and scraping techniques to collect user reviews, comments, and feedback data from social media platforms, review sites, and customer service platforms. Specifically, the APIs are used to retrieve posts with specific hashtags (e.g., "product reviews") from social media platforms. The server then stores the retrieved data in an intermediary database.

[0692] Input: URL or API endpoint of social media or review site

[0693] Output: Feedback data stored in an intermediate database

[0694] Step 2: Save data

[0695] The server structures and stores the collected raw data in an intermediate database, which organizes the data and prepares it for further processing. In this step, data insertion operations are performed using SQL.

[0696] Input: Raw data collected

[0697] Output: An intermediate database containing structured data

[0698] Step 3: Analyze the data

[0699] The server sends the stored text data to a natural language processing (NLP) model to analyze user opinions and feedback. Specifically, it uses deep learning models (e.g., BERT and GPT series) to extract summaries and key phrases, making it easier to grasp the gist of the data.

[0700] Input: Text data stored in the intermediate database

[0701] Output: Summary sentences and key phrases analyzed by the NLP model

[0702] Specific operation: The server retrieves the text data one by one, inputs prompts such as "What is the main opinion of this review?" into the NLP model, and saves the results in another table.

[0703] Step 4: Generate a report

[0704] The server automatically generates reports based on the analyzed data. It inserts summarized data and extracted key phrases into pre-prepared template formats to create reports that clearly show the user's needs. A template engine is used in this process.

[0705] Input: Summary sentences and key phrases from the NLP model

[0706] Output: Formatted report

[0707] What happens: The server inserts the summary and key phrases into a template and generates a report as "Highlights of the Week."

[0708] Step 5: Report Delivery

[0709] The server automatically distributes the generated reports to the designated recipients via email or an internal messaging system, based on a pre-defined schedule.

[0710] Input: Generated report

[0711] Output: Report delivered to recipients

[0712] What it does: The server generates a report and sends it to the service representative's email address every Monday, containing key user feedback and suggestions based on it.

[0713] (Application example 1)

[0714] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0715] Traditional methods for collecting, analyzing, and generating reports on voice of customer (VOC) required a lot of manual work, which was time-consuming and costly, making it difficult to quickly and efficiently understand user needs. Furthermore, generating and distributing reports required human labor, which often led to a lack of timely information sharing. This made it difficult for service providers to quickly implement appropriate improvement measures, resulting in delays in improving customer satisfaction.

[0716] 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.

[0717] In this invention, the server includes means for collecting data using an application programming interface, means for summarizing the collected data using a natural language processing model, means for automatically generating a report based on the summary results, means for automatically distributing the generated report to a predetermined recipient, and means for summarizing the collected data based on its contents, generating a report according to a predetermined schedule, and distributing the report to a predetermined recipient via email, thereby enabling user needs to be grasped quickly and efficiently at low cost.

[0718] An "application programming interface" is something that provides an interface between software or applications, enabling the exchange of data and functions.

[0719] "Data" refers to information such as reviews, comments, and feedback obtained from customers.

[0720] A "natural language processing model" is a machine learning model for analyzing and summarizing text data, and includes the BERT and GPT series.

[0721] "Summarizing" means extracting important parts from multiple text data and summarizing them in a concise form.

[0722] A "report" is a document generated from summarized data that shows the results of organizing and analyzing user needs and feedback.

[0723] "Automatically generating" means that the system generates documents according to specified algorithms or templates without human intervention.

[0724] "Intended Recipient" means a person or department that has been pre-designated to receive the report.

[0725] "Automatically delivering" means that the system sends information as emails or messages based on set schedules or conditions, without manual intervention.

[0726] A "schedule" refers to a predetermined time and order plan for performing a series of tasks or processes.

[0727] "Email" is a means of sending and receiving text data and attachments in digital form over the Internet.

[0728] This invention relates to a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, generates reports, and distributes them to designated recipients, thereby enabling companies to quickly and inexpensively understand user needs. This system is targeted at collecting and analyzing customer feedback on online shopping sites.

[0729] The server first collects data using an application programming interface (API). Specifically, it retrieves customer reviews and comments from the review API of an online shopping site. Scraping technology can also be used, for example, to automatically collect reviews on a web page. The collected data is then stored in an intermediate database. Examples of databases include MongoDB and MySQL.

[0730] Next, the server inputs the collected data into a natural language processing model (NLP model) for analysis and summarization. This summarization process uses generative AI models such as BERT and the GPT series. Specifically, user reviews and comments are taken as input, and the summarization model concisely summarizes their content. For example, it uses a prompt such as, "Given the review: This product is easy to use and of high quality. The design is also good and I am very satisfied. It was a very good purchase. However, the delayed delivery was a negative point. Summarize it in 50 words or less."

[0731] The server automatically generates reports based on the summarized data. The generated summaries and extracted key phrases are arranged in a template format to create a report that clearly identifies important user needs. For example, a "Highlights of the Week" report might state that "Many customers praised the product design, but many complained about delivery speed." The templates are written in HTML or Markdown.

[0732] Finally, the server automatically distributes the generated reports to the designated recipients via email. A library for working with the SMTP protocol (e.g., smtplib) is used to send emails. A weekly report is then sent to each recipient. This report includes a summary of the customer feedback for the current week and recommendations based on that feedback.

[0733] In this way, the present invention automates the entire process of data collection, analysis, report generation, and distribution, making it possible to understand user needs efficiently and at low cost. Furthermore, because high-quality reports can be created without specialized knowledge, service providers can quickly implement improvement measures.

[0734] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0735] Step 1:

[0736] The server collects data using an application programming interface (API). Specifically, it obtains customer reviews and comments through the online shopping site's review API. The API call returns data in JSON format. The input is the API endpoint URL, and the output is customer reviews in JSON format. The specific operation of data collection is to send an HTTP GET request and receive a response.

[0737] Step 2:

[0738] The server stores the collected raw data in an intermediate database, which is implemented using MongoDB or MySQL. The input is customer reviews in JSON format, and the output is review information stored in the database. Specific operations include inserting the parsed JSON data into the database.

[0739] Step 3:

[0740] The server analyzes the collected data using a natural language processing model and generates a summary. The BERT and GPT series are used as generative AI models. The input is text data obtained from an intermediate database, and the output is summarized text. Specific operations include inputting the text data into the NLP model and obtaining the summary results. An example prompt is "Given the review: [review text], summarize it in 50 words or less."

[0741] Step 4:

[0742] The server automatically generates reports based on the summarized data. The reports are constructed based on a template format. The input is the summarized text, and the output is a structured report. Specifically, the server applies the summarized data to the template and assembles it in the required layout and format. The template can be in HTML or Markdown format.

[0743] Step 5:

[0744] The server automatically delivers the generated report to the specified recipient. The delivery method is email. The input is the generated report, and the output is the email sent to the recipient. Specific operations include sending email using an SMTP server. The smtplib library is used to send email.

[0745] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0746] This invention relates to a system that enables companies to quickly and inexpensively grasp user needs and emotions by efficiently collecting voice of customer (VOC), automatically analyzing and summarizing it, and combining it with an emotion engine that recognizes user emotions to generate reports and distribute them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0747] Data collection

[0748] The server uses an Application Programming Interface (API) to authenticate and collect data, specifically, API keys, API secrets, access tokens, and access token secrets to retrieve user reviews, comments, and feedback data from, for example, social media platforms, app stores, and customer service platforms, and stores this data in an intermediary database.

[0749] Specific examples

[0750] The server uses the SNS API to collect tweets with the hashtag "product review" and stores the data in an intermediate database.

[0751] Data processing and analysis

[0752] The server inputs the collected data into a natural language processing model for analysis. Specifically, it uses the Transformers library to utilize the latest generative AI models (e.g., BERT and GPT series) to summarize the collected text data.

[0753] Specific examples

[0754] The server inputs the collected tweets into a summarization model and obtains summary results, such as a large number of users requesting improvements to the app's user interface.

[0755] Emotion recognition

[0756] The server uses an emotion engine to recognize user emotions from the collected data, analyzes user comments and reviews, and assigns emotion tags such as positive, negative, and neutral.

[0757] Specific examples

[0758] The server processes tweets with a sentiment engine and classifies them as "positive" tweets such as "I really love the new design" and "I wish the app would stop crashing so often" as "negative" tweets.

[0759] Report Generation

[0760] The server automatically generates reports based on the summarized data and the user's emotional information. The generated summaries, key phrases, and emotional information are organized according to a template format, allowing users to create high-quality reports without specialized knowledge.

[0761] Specific examples

[0762] The server organizes the results of the summarization and sentiment analysis according to templates and generates a "Highlights of the Week" report such as "60% of users are happy with the new feature, but 30% have reported bugs."

[0763] Report Distribution

[0764] The server automatically delivers the generated reports to the designated recipients. It configures the SMTP server (address, port, username, password) and converts the report content into email format. The reports are sent to the relevant departments on a weekly basis.

[0765] Specific examples

[0766] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the current week's key user feedback and the results of sentiment analysis.

[0767] In this way, the system of the present invention automates the entire process of data collection, emotion recognition, analysis, report generation, and distribution, making it possible to grasp user needs and emotions efficiently and at low cost. This system allows service providers to quickly and accurately implement improvement measures that reflect users' emotions and needs.

[0768] The processing flow will be explained below.

[0769] Step 1:

[0770] The server uses the application programming interface (API) to authenticate for data collection. Specifically, it uses the API key, API secret, access token, and access token secret to access the API of Twitter and other SNSs and set authentication information.

[0771] Step 2:

[0772] The server uses the authentication information to collect tweets containing the target hashtags and keywords. It then connects to the Twitter API using the Tweepy library, retrieves 100 tweets containing "product reviews," and extracts information such as the text and posting date and time.

[0773] Step 3:

[0774] The server converts the collected tweet data into a Pandas data frame and stores it in an intermediate database, thereby preparing the data in a format that can be used for subsequent analysis.

[0775] Step 4:

[0776] The server then inputs the stored data into a natural language processing model to analyze the collected text data. Specifically, it uses the Transformers library to generate summaries using generative AI models such as BERT and GPT, which extract the key points of user opinions and feedback.

[0777] Step 5:

[0778] The server inputs the summarized text data into an emotion engine to recognize the user's emotions, and the emotion engine assigns emotion tags such as positive, negative, or neutral to each text.

[0779] Step 6:

[0780] The server extracts important key phrases from the collected data, using TF-IDF and other key phrase extraction techniques to identify key words and phrases in the summarized text.

[0781] Step 7:

[0782] The server automatically generates a report based on the summary, key phrases, and sentiment information. The generated summary, key phrases, and sentiment information are organized according to a template format, and a report is created that clearly shows important user needs and sentiment.

[0783] Step 8:

[0784] The server configures the SMTP server (address, port, username, password) and converts the generated reports into email format, ensuring that the reports contain the required information and analysis results.

[0785] Step 9:

[0786] The server connects to an SMTP server and automatically sends the generated report to a designated recipient, such as a service representative, on a weekly or set schedule.

[0787] The above steps automate the collection of VOC, emotion recognition, analysis, report generation, and distribution, making it possible to grasp user needs and emotions efficiently and at low cost. This system allows service providers to quickly and accurately implement service improvement measures that reflect user requests and emotions.

[0788] Example 2

[0789] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0790] It is difficult for companies to efficiently collect voice of customer (VOC) information quickly and at low cost, and analyze the data to understand user needs and emotions. Furthermore, since it takes a great deal of time and effort to extract the necessary information from the collected data, organize it, and gain insights, there is a need for an automated system that can be easily used even by personnel without specialized knowledge.

[0791] 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.

[0792] In this invention, the server includes means for collecting data using an application programming interface, means for storing the collected data in an intermediate database, means for summarizing the collected data using a natural language processing model, means for recognizing emotions from the summary results and the collected data, means for automatically generating a report based on the summary results and the emotion recognition results, and means for automatically delivering the generated report to a predetermined recipient. This efficiently automates the process from data collection to emotion recognition, and report generation and delivery, enabling companies to grasp user needs and emotions quickly and at low cost.

[0793] An "application programming interface" is a defined interface through which different software systems communicate with each other and exchange data.

[0794] "Means of collecting data" refers to the technologies and methods for obtaining information from various data sources, and typically involves using APIs.

[0795] An "intermediate database" is a database that temporarily stores collected data and uses it in subsequent processing.

[0796] A "natural language processing model" is an algorithm or model that uses machine learning or deep learning techniques to analyze and understand human language.

[0797] "Means of summarization" refers to techniques and methods for extracting the main content from long text data and displaying it in a short form.

[0798] "Means of emotion recognition" refers to algorithms and technologies for extracting emotional information such as positive, negative, and neutral from text data.

[0799] A "summary result" is a short document or phrase generated by a natural language processing model that succinctly conveys the main content of the text data.

[0800] The "emotion recognition result" is the result of emotion analysis on text data, and is classification information of the emotion that each text has.

[0801] "Means for automatically generating reports" refers to technologies and methods for automatically creating report-format documents based on analytical data such as summary results and emotion recognition results.

[0802] "Specified recipient" refers to a predetermined recipient of the report, meaning a specific department or person.

[0803] "Means for automated report delivery" refers to any technology or method for automatically sending generated reports to intended recipients.

[0804] This invention is a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, and then generates reports by combining it with an emotion engine that recognizes user emotions, and delivers them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0805] First, in the data collection stage, the server collects data using the application programming interface (API). Specifically, it uses the API key, API secret, access token, and access token secret to retrieve user reviews, comments, and feedback data from social media, app stores, and customer service platforms. This collected data is then stored in an intermediate database.

[0806] Specific examples

[0807] The server uses the SNS API to collect tweets with the hashtag "product review" and stores the data in an intermediate database.

[0808] The server then analyzes the collected data by inputting it into a natural language processing model. Specifically, it uses the Transformers library to utilize generative AI models (such as the BERT and GPT series). This process summarizes the collected text data.

[0809] Specific examples

[0810] The server inputs the collected tweets into a summarization model and summarizes the requests of many users for improvements to the app's user interface.

[0811] The server then uses an emotion engine to recognize user emotions from the collected data, performs sentiment analysis on user comments and reviews, and assigns emotion tags such as positive, negative, or neutral.

[0812] Specific examples

[0813] The server processes tweets with a sentiment engine and classifies them as "positive" tweets such as "I really love the new design" and "I wish the app would stop crashing so often" as "negative" tweets.

[0814] The server then automatically generates a report based on the summarized data and the user's emotional information. The report is organized according to a template format, allowing high-quality reports to be created without specialized knowledge.

[0815] Specific examples

[0816] The server organizes the results of the summarization and sentiment analysis according to templates and generates a "Highlights of the Week" report such as "60% of users are happy with the new feature, but 30% have reported bugs."

[0817] The server then automatically delivers the generated reports to the intended recipients using an SMTP server, with settings including address, port, username and password.

[0818] Specific examples

[0819] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the current week's key user feedback and the results of sentiment analysis.

[0820] Prompt Sentence Examples

[0821] 1. "Extract and summarize key points from customer reviews and feedback."

[0822] 2. "Analyze the sentiment of the collected tweets and categorize them as positive or negative."

[0823] 3. "Generate a report based on this week's user feedback and sentiment analysis results."

[0824] This system allows service providers to quickly and accurately implement improvement measures that reflect users' feelings and needs.

[0825] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0826] Step 1:

[0827] Data collection

[0828] The server authenticates to collect data using an Application Programming Interface (API), requiring API key, API secret, access token, and access token secret as inputs, and uses these to retrieve user reviews, comments, and feedback data from social media, app stores, and customer service platforms, and obtains raw data as output.

[0829] Specific actions

[0830] The server configures the Twitter API to collect tweets with the hashtag "product reviews." If authentication is successful, the tweet data (username, text, timestamp, etc.) is retrieved.

[0831] Step 2:

[0832] Storing data in an intermediate database

[0833] The server stores the collected data in an intermediate database. As input, the raw data collected in step 1 is used. As output, these data are stored in appropriate tables. The database uses an SQL database (e.g., MySQL or PostgreSQL).

[0834] Specific actions

[0835] The server uses SQL commands to insert the collected tweets into a table called "tweets," storing each tweet's ID, username, text, and timestamp.

[0836] Step 3:

[0837] Data analysis using natural language processing

[0838] The server inputs the data stored in the intermediate database into a natural language processing model for analysis. The input is text data retrieved from the database. The output is a summary result. This process uses the Transformers library and generative AI models (e.g., BERT, GPT series).

[0839] Specific actions

[0840] The server retrieves specific tweet text from the database and inputs it into the summarization model, which then analyzes it to obtain summaries, which are then stored in a new table called "summaries."

[0841] Step 4:

[0842] Emotion recognition

[0843] The server performs emotion recognition on the collected data. Text data is used as input. Emotion tags such as positive, negative, and neutral are generated as output. A sentiment analysis engine (e.g., VADER, TextBlob, etc.) is used.

[0844] Specific actions

[0845] The server feeds the tweet data into a sentiment analysis engine, which classifies each tweet into a sentiment tag (positive, negative, neutral), and adds the results to a new column in the "tweets" table.

[0846] Step 5:

[0847] Automatic report generation

[0848] The server automatically generates a report based on the summary results and emotion recognition results. The analysis results and emotion tags obtained in steps 3 and 4 are used as input. A template-based report is generated as output. A template engine (e.g., Jinja2) is used.

[0849] Specific actions

[0850] The server uses a template engine to embed the analysis results and sentiment tags into reports, creating templates that can be used to generate reports, such as "This week's highlights" that include "60% of users are happy with the new features, but 30% have reported bugs."

[0851] Step 6:

[0852] Automatic report delivery

[0853] The server automatically delivers the generated report to the specified recipients. As input, it uses the generated report and the SMTP server configuration information (address, port, username, password). As output, it sends an email to the recipients.

[0854] Specific actions

[0855] The server connects to the SMTP server, converts the generated report into an email format, and sends it to the email address of the service representative. For example, a schedule can be set to send a "report on interface improvement requests" every Monday.

[0856] By going through each step in this way, the process from data collection to emotion recognition, report generation and distribution is efficiently automated.

[0857] (Application example 2)

[0858] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0859] In recent years, collecting and analyzing voice of the customer (VOC) in brick-and-mortar stores has become essential for improving store operations and customer satisfaction. However, conventional methods have difficulty efficiently collecting and analyzing large volumes of data, and are unable to identify emotions, making it difficult to identify specific areas for improvement or take prompt action. This has made it difficult for brick-and-mortar store managers and staff to accurately understand customers' true needs and emotions.

[0860] 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.

[0861] In this invention, the server includes means for collecting data using an application programming interface, means for summarizing the collected data using a natural language processing model, means for identifying user emotions from the collected data using an emotion engine that recognizes user emotions, means for automatically generating a report based on the summary results and the emotion identification results, and means for automatically delivering the generated report to a specified recipient. This makes it possible to efficiently collect customer feedback and perform emotion analysis, thereby quickly identifying and addressing specific areas for improvement.

[0862] "Application Programming Interface" refers to the well-defined rules or protocols by which software applications communicate with each other.

[0863] "Data collection methods" refers to mechanisms for gathering information and feedback from multiple sources.

[0864] A "natural language processing model" refers to algorithms and technologies that allow computers to understand, analyze, and generate human language.

[0865] "Summarization methods" refers to the processes and techniques used to summarize collected information succinctly.

[0866] "Emotion engine" refers to algorithms and technologies for identifying user emotions from text data.

[0867] "Means for identifying emotions" refers to processes and techniques for extracting and classifying user emotions from text data.

[0868] "Means for automatically generating reports" refers to processes or technologies for automatically creating reports based on summary results and emotion identification results.

[0869] "Means for delivering reports" refers to the process or technology used to transmit generated reports to intended recipients.

[0870] "Key phrase extraction methods" refers to the processes and techniques used to identify and extract important words and phrases from text data.

[0871] "Template-based organization" refers to the process or technique of organizing and structuring information according to a predefined format.

[0872] The present invention is a system that efficiently collects customer feedback (VOC) from physical stores, automatically summarizes and analyzes the feedback, and compiles the results into a report and distributes it to relevant parties. This system is implemented as follows.

[0873] First, the server collects feedback data using an application programming interface (API), using authentication information such as an API key, API secret, or access token. Data collection targets include social media, feedback terminals, and customer service platforms.

[0874] The collected data is stored in an intermediate database. The server then summarizes this data using natural language processing models, including generative AI models such as BERT and the GPT series from the Transformers library. The summarized data is further tagged with sentiment tags, such as positive, negative, or neutral, using a sentiment engine. The sentiment engine, which is also part of natural language processing, uses the Transformers library for data analysis.

[0875] Once the specific data has been summarized and sentiment analyzed, a report is automatically generated based on it. The report contains summarized data, key phrases, and sentiment information. The report generation is based on templates, which organize the information and make the report content easy to read and understand.

[0876] The generated reports are automatically delivered to the specified recipients by the server. The SMTP server settings (address, port, username, password) are configured, the report contents are converted into email format, and sent to the specified email address. Reports are generally delivered weekly, allowing relevant staff to regularly check customer feedback.

[0877] Specific examples

[0878] For example, if a customer uses a feedback terminal to comment, "The product description is difficult to understand," the system collects this comment and stores it in an intermediate database. The server then inputs this data into a natural language processing model and outputs a summary, "Product description improvement request." The emotion engine then analyzes this summary and assigns the emotion tag "negative." Finally, a report is generated based on this information and automatically sent to the store manager.

[0879] Prompt Sentence Examples

[0880] Based on the following example prompt, the system will perform appropriate data collection, summarization, and sentiment analysis:

[0881] 1. Data Collection Prompt: "What is an effective way to collect customer reviews? Specifically, can you provide an example of using an API key to retrieve social media data?"

[0882] 2. Data Summarization Prompt: "What is the latest generative AI model that summarizes collected review data? Please include a Python code example."

[0883] 3. Sentiment Analysis Prompt: "What generative AI model would be suitable for performing sentiment analysis on review data and how would you use it?"

[0884] 4. Report Generation Prompt: "How can I automatically generate a report based on the results of my analysis? Can you give me an example of a template I could use?"

[0885] 5. Report Delivery Prompt: "How can I automatically deliver generated reports to a specified email address? Please provide a Python code example."

[0886] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0887] Step 1: Data collection

[0888] The server collects feedback data using an application programming interface (API). Specifically, it uses the API of social media and feedback devices to obtain customer comments and reviews using authentication information (API key, API secret, access token, etc.). At this time, the data is obtained in JSON format and stored in an intermediate database.

[0889] Input: API key, secret, access token

[0890] Output: Feedback data stored in an intermediate database

[0891] Step 2: Data Summarization

[0892] The server inputs the collected feedback data into a natural language processing model to summarize it. The generative AI models used here are the BERT and GPT series from the Transformers library. The model extracts important parts of the input text data and outputs a concise summary.

[0893] Input: Feedback data stored in the intermediate database

[0894] Output: Summarized text data

[0895] Step 3: Emotion Identification

[0896] The server then inputs the summarized text data into an emotion engine for sentiment analysis, which also uses the Transformers library to tag each piece of text with an emotion tag, such as positive, negative, or neutral.

[0897] Input: Summarized text data

[0898] Output: Text data with emotion tags

[0899] Step 4: Generate a report

[0900] The server automatically generates a report based on the summary and emotion identification results. This report is based on a template and organizes summaries, emotion tags, key phrases, etc.

[0901] Input: Text data with emotion tags

[0902] Output: Organized reports in template format

[0903] Step 5: Report Delivery

[0904] The server automatically delivers the generated report to the specified recipient by configuring the SMTP server settings (address, port, username, password), converting the report content into email format, and sending it.

[0905] Input: Reports organized in template format

[0906] Output: Report sent to recipient's (e.g. store manager) email address

[0907] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0908] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0909] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[0910] [Fourth embodiment]

[0911] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0912] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0913] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0914] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[0915] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0916] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0917] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0918] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[0919] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0920] 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.

[0921] 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.

[0922] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0923] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[0924] This invention relates to a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, generates reports, and distributes them to designated recipients, thereby enabling companies to quickly and inexpensively grasp user needs. This system is realized by automating each stage of processing, centered on a server.

[0925] Data collection

[0926] The server first collects data using application programming interfaces (APIs) and other methods, such as obtaining user reviews, comments, and feedback data from social media platforms, app stores, and customer service platforms. This is done using API access and scraping techniques, and the collected data is then stored in an intermediate database.

[0927] Specific examples

[0928] The server uses the social media platform's API to collect tweets with the hashtag "product review" and store them in a database.

[0929] Data processing and analysis

[0930] The server then analyzes the collected data using a natural language processing model. Specifically, it utilizes the latest generative AI models (e.g., deep learning models) to summarize the vast amount of collected text data. This allows it to concisely organize user opinions and feedback and generate summaries. Models such as the BERT and GPT series are used in this summarization process.

[0931] Specific examples

[0932] The server inputs the collected tweets into a summary model and obtains summary results such as "A large number of users are requesting improvements to the app's user interface."

[0933] Report Generation

[0934] The server automatically generates reports based on the summarized data. The generated summaries and extracted key phrases are arranged in a template format to create a report that clearly shows important user needs. This makes it easy for anyone, even those without specialized knowledge, to generate high-quality reports.

[0935] Specific examples

[0936] The server organizes the collected summaries according to a template and creates a report that states "over 50% of users are satisfied with the new feature" as a "highlight of the week."

[0937] Report Distribution

[0938] The server automatically delivers the final report to the designated recipients. The report is periodically sent to the relevant departments involved in the work. The delivery method is via email or the company's internal messaging system.

[0939] Specific examples

[0940] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the main user feedback for the current week and suggestions based on it.

[0941] In this way, the system of the present invention automates the entire process of data collection, analysis, report generation, and distribution, making it possible to grasp user needs efficiently and at low cost. This system allows service providers to accurately and quickly implement improvement measures that reflect user needs.

[0942] The processing flow will be explained below.

[0943] Step 1:

[0944] The server uses the application programming interface (API) to authenticate for data collection. Specifically, it accesses the Twitter API and sets authentication information using the API key, API secret, access token, and access token secret.

[0945] Step 2:

[0946] The server uses the authentication information to collect tweets containing the target hashtags and keywords. It connects to the Twitter API using the Tweepy library and retrieves 100 tweets containing the specific hashtag "product review."

[0947] Step 3:

[0948] The server converts the collected tweet data (text, posting date and time, etc.) into a Pandas data frame and stores it in an intermediate database, thereby preparing the data in a format that can be used for subsequent analysis.

[0949] Step 4:

[0950] The server inputs the text data stored in the intermediate database into a natural language processing model, specifically using the Transformers library and generative AI models such as BERT and GPT to generate summaries.

[0951] Step 5:

[0952] The server extracts the generated summary text and selects important key phrases. It uses TF-IDF and key phrase extraction techniques to identify key words and phrases from the summarized text.

[0953] Step 6:

[0954] The server automatically generates reports based on the summaries and key phrases, organized according to predefined templates and containing key user needs and statistics.

[0955] Step 7:

[0956] The server prepares the generated report for delivery to the intended recipient by configuring the SMTP server (address, port, username, password) and converting the report content to email format.

[0957] Step 8:

[0958] The server connects to the SMTP server and automatically sends the generated report to the relevant service personnel. Specifically, it sends a "report on interface improvement requests" to the personnel in charge by email every week.

[0959] The above steps automate the entire process from collecting VOCs to generating and distributing reports, making it possible to efficiently understand user needs.

[0960] Example 1

[0961] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[0962] Companies are seeking a way to quickly and low-costly understand user needs. However, collecting and analyzing vast amounts of information from social media, review sites, and customer service platforms, generating useful reports based on that information, and then distributing those reports to relevant parties in a timely manner requires a great deal of time and effort. There is a need to automate these processes and provide a way to efficiently understand user needs.

[0963] 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.

[0964] In this invention, the server includes means for collecting data using an application programming interface, means for storing the collected data in an intermediate database, means for analyzing the stored data using a natural language processing model and generating a summary, means for automatically generating a report based on the summary result, and means for automatically distributing the generated report to a predetermined recipient. This automates the process of data collection, analysis, report generation, and distribution, making it possible to grasp user needs quickly and at low cost.

[0965] An "application programming interface" is a set of definitions and protocols that provide an interface through which software components can communicate with other software.

[0966] "Data collection means" are the methods and mechanisms used to obtain the required data from external sources.

[0967] The "intermediate database" is a database for temporarily storing collected data and transferring it to subsequent processing.

[0968] A "natural language processing model" is an algorithm or system that uses machine learning technology to understand and analyze human language.

[0969] A "means for generating a summary" is a method or technology for extracting important information from a huge amount of text data and summarizing it concisely.

[0970] "Means for automatically generating a report" refers to a method or system for automatically creating a report in a predetermined format based on the analysis results or summary.

[0971] "Means for automatically delivering to recipients" refers to methods or techniques for automatically sending the generated report to pre-designated recipients.

[0972] This invention relates to a system that enables companies to quickly and inexpensively grasp user needs by efficiently collecting voice of customer (VOC), automatically analyzing and summarizing it, generating reports, and distributing them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[0973] First, the server uses application programming interfaces (APIs) and scraping technology to collect user reviews, comments, and feedback data from social media, review sites, and customer service platforms. The collected data is then stored in an intermediate database. For example, the server uses the API of a social media platform to collect tweets with the hashtag "product review" and store them in a database.

[0974] The server then sends the collected text data to a natural language processing model (NLP model) to analyze user opinions and feedback. Specifically, it uses deep learning models (e.g., BERT and GPT series) to extract summaries and key phrases. This allows it to organize important information from large amounts of data and convert it into an interpretable format. For example, the server inputs the collected tweets into a summarization model and obtains a summary result such as, "Many users are requesting improvements to the app's user interface."

[0975] The server then automatically generates reports based on the analyzed data. By inserting summarized data and extracted key phrases into pre-defined templates, reports that clearly indicate user needs are created. For example, the server might create a "Highlights of the Week" report that states, "More than 50% of users are satisfied with the new features."

[0976] Finally, the server automatically distributes the generated report to the designated recipients via email or the company's internal messaging system. For example, every Monday the server sends a "Report on Interface Improvement Requests" to the email address of a service representative. This email includes the main user feedback for the current week and suggestions based on that feedback.

[0977] Specific examples

[0978] The server sends a request to the SNS API (e.g., Twitter API) at a specific time every day, retrieves posts containing a specific hashtag such as "product review," and stores them in an intermediate database.

[0979] The server inputs the reviews and comments stored in the database into an NLP model (for example, the BERT model in Hugging Face's Transformers library) and stores the results as summaries and key phrases in a separate table.

[0980] The server extracts the analysis results, embeds the data in a template called "Highlights of the Week," and generates a report such as "More than 50% of users are satisfied with the new feature."

[0981] The server emails the generated report based on a pre-specified list of recipients.

[0982] Prompt Sentence Examples

[0983] This week's highlights:

[0984] Summary: Many users are requesting improvements to the app's user interface.

[0985] In this way, the system of the present invention automates the process of data collection, analysis, report generation and distribution, making it possible to grasp user needs quickly and at low cost.

[0986] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0987] Step 1: Data collection

[0988] The server uses application programming interfaces (APIs) and scraping techniques to collect user reviews, comments, and feedback data from social media platforms, review sites, and customer service platforms. Specifically, the APIs are used to retrieve posts with specific hashtags (e.g., "product reviews") from social media platforms. The server then stores the retrieved data in an intermediary database.

[0989] Input: URL or API endpoint of social media or review site

[0990] Output: Feedback data stored in an intermediate database

[0991] Step 2: Save data

[0992] The server structures and stores the collected raw data in an intermediate database, which organizes the data and prepares it for further processing. In this step, data insertion operations are performed using SQL.

[0993] Input: Raw data collected

[0994] Output: An intermediate database containing structured data

[0995] Step 3: Analyze the data

[0996] The server sends the stored text data to a natural language processing (NLP) model to analyze user opinions and feedback. Specifically, it uses deep learning models (e.g., BERT and GPT series) to extract summaries and key phrases, making it easier to grasp the gist of the data.

[0997] Input: Text data stored in the intermediate database

[0998] Output: Summary sentences and key phrases analyzed by the NLP model

[0999] Specific operation: The server retrieves the text data one by one, inputs prompts such as "What is the main opinion of this review?" into the NLP model, and saves the results in another table.

[1000] Step 4: Generate a report

[1001] The server automatically generates reports based on the analyzed data. It inserts summarized data and extracted key phrases into pre-prepared template formats to create reports that clearly show the user's needs. A template engine is used in this process.

[1002] Input: Summary sentences and key phrases from the NLP model

[1003] Output: Formatted report

[1004] What happens: The server inserts the summary and key phrases into a template and generates a report as "Highlights of the Week."

[1005] Step 5: Report Delivery

[1006] The server automatically distributes the generated reports to the designated recipients via email or an internal messaging system, based on a pre-defined schedule.

[1007] Input: Generated report

[1008] Output: Report delivered to recipients

[1009] What it does: The server generates a report and sends it to the service representative's email address every Monday, containing key user feedback and suggestions based on it.

[1010] (Application example 1)

[1011] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1012] Traditional methods for collecting, analyzing, and generating reports on voice of customer (VOC) required a lot of manual work, which was time-consuming and costly, making it difficult to quickly and efficiently understand user needs. Furthermore, generating and distributing reports required human labor, which often led to a lack of timely information sharing. This made it difficult for service providers to quickly implement appropriate improvement measures, resulting in delays in improving customer satisfaction.

[1013] 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.

[1014] In this invention, the server includes means for collecting data using an application programming interface, means for summarizing the collected data using a natural language processing model, means for automatically generating a report based on the summary results, means for automatically distributing the generated report to a predetermined recipient, and means for summarizing the collected data based on its contents, generating a report according to a predetermined schedule, and distributing the report to a predetermined recipient via email, thereby enabling user needs to be grasped quickly and efficiently at low cost.

[1015] An "application programming interface" is something that provides an interface between software or applications, enabling the exchange of data and functions.

[1016] "Data" refers to information such as reviews, comments, and feedback obtained from customers.

[1017] A "natural language processing model" is a machine learning model for analyzing and summarizing text data, and includes the BERT and GPT series.

[1018] "Summarizing" means extracting important parts from multiple text data and summarizing them in a concise form.

[1019] A "report" is a document generated from summarized data that shows the results of organizing and analyzing user needs and feedback.

[1020] "Automatically generating" means that the system generates documents according to specified algorithms or templates without human intervention.

[1021] "Intended Recipient" means a person or department that has been pre-designated to receive the report.

[1022] "Automatically delivering" means that the system sends information as emails or messages based on set schedules or conditions, without manual intervention.

[1023] A "schedule" refers to a predetermined time and order plan for performing a series of tasks or processes.

[1024] "Email" is a means of sending and receiving text data and attachments in digital form over the Internet.

[1025] This invention relates to a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, generates reports, and distributes them to designated recipients, thereby enabling companies to quickly and inexpensively understand user needs. This system is targeted at collecting and analyzing customer feedback on online shopping sites.

[1026] The server first collects data using an application programming interface (API). Specifically, it retrieves customer reviews and comments from the review API of an online shopping site. Scraping technology can also be used, for example, to automatically collect reviews on a web page. The collected data is then stored in an intermediate database. Examples of databases include MongoDB and MySQL.

[1027] Next, the server inputs the collected data into a natural language processing model (NLP model) for analysis and summarization. This summarization process uses generative AI models such as BERT and the GPT series. Specifically, user reviews and comments are taken as input, and the summarization model concisely summarizes their content. For example, it uses a prompt such as, "Given the review: This product is easy to use and of high quality. The design is also good and I am very satisfied. It was a very good purchase. However, the delayed delivery was a negative point. Summarize it in 50 words or less."

[1028] The server automatically generates reports based on the summarized data. The generated summaries and extracted key phrases are arranged in a template format to create a report that clearly identifies important user needs. For example, a "Highlights of the Week" report might state that "Many customers praised the product design, but many complained about delivery speed." The templates are written in HTML or Markdown.

[1029] Finally, the server automatically distributes the generated reports to the designated recipients via email. A library for working with the SMTP protocol (e.g., smtplib) is used to send emails. A weekly report is then sent to each recipient. This report includes a summary of the customer feedback for the current week and recommendations based on that feedback.

[1030] In this way, the present invention automates the entire process of data collection, analysis, report generation, and distribution, making it possible to understand user needs efficiently and at low cost. Furthermore, because high-quality reports can be created without specialized knowledge, service providers can quickly implement improvement measures.

[1031] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1032] Step 1:

[1033] The server collects data using an application programming interface (API). Specifically, it obtains customer reviews and comments through the online shopping site's review API. The API call returns data in JSON format. The input is the API endpoint URL, and the output is customer reviews in JSON format. The specific operation of data collection is to send an HTTP GET request and receive a response.

[1034] Step 2:

[1035] The server stores the collected raw data in an intermediate database, which is implemented using MongoDB or MySQL. The input is customer reviews in JSON format, and the output is review information stored in the database. Specific operations include inserting the parsed JSON data into the database.

[1036] Step 3:

[1037] The server analyzes the collected data using a natural language processing model and generates a summary. The BERT and GPT series are used as generative AI models. The input is text data obtained from an intermediate database, and the output is summarized text. Specific operations include inputting the text data into the NLP model and obtaining the summary results. An example prompt is "Given the review: [review text], summarize it in 50 words or less."

[1038] Step 4:

[1039] The server automatically generates reports based on the summarized data. The reports are constructed based on a template format. The input is the summarized text, and the output is a structured report. Specifically, the server applies the summarized data to the template and assembles it in the required layout and format. The template can be in HTML or Markdown format.

[1040] Step 5:

[1041] The server automatically delivers the generated report to the specified recipient. The delivery method is email. The input is the generated report, and the output is the email sent to the recipient. Specific operations include sending email using an SMTP server. The smtplib library is used to send email.

[1042] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1043] This invention relates to a system that enables companies to quickly and inexpensively grasp user needs and emotions by efficiently collecting voice of customer (VOC), automatically analyzing and summarizing it, and combining it with an emotion engine that recognizes user emotions to generate reports and distribute them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[1044] Data collection

[1045] The server uses an Application Programming Interface (API) to authenticate and collect data, specifically, API keys, API secrets, access tokens, and access token secrets to retrieve user reviews, comments, and feedback data from, for example, social media platforms, app stores, and customer service platforms, and stores this data in an intermediary database.

[1046] Specific examples

[1047] The server uses the SNS API to collect tweets with the hashtag "product review" and stores the data in an intermediate database.

[1048] Data processing and analysis

[1049] The server inputs the collected data into a natural language processing model for analysis. Specifically, it uses the Transformers library to utilize the latest generative AI models (e.g., BERT and GPT series) to summarize the collected text data.

[1050] Specific examples

[1051] The server inputs the collected tweets into a summarization model and obtains summary results, such as a large number of users requesting improvements to the app's user interface.

[1052] Emotion recognition

[1053] The server uses an emotion engine to recognize user emotions from the collected data, analyzes user comments and reviews, and assigns emotion tags such as positive, negative, and neutral.

[1054] Specific examples

[1055] The server processes tweets with a sentiment engine and classifies them as "positive" tweets such as "I really love the new design" and "I wish the app would stop crashing so often" as "negative" tweets.

[1056] Report Generation

[1057] The server automatically generates reports based on the summarized data and the user's emotional information. The generated summaries, key phrases, and emotional information are organized according to a template format, allowing users to create high-quality reports without specialized knowledge.

[1058] Specific examples

[1059] The server organizes the results of the summarization and sentiment analysis according to templates and generates a "Highlights of the Week" report such as "60% of users are happy with the new feature, but 30% have reported bugs."

[1060] Report Distribution

[1061] The server automatically delivers the generated reports to the designated recipients. It configures the SMTP server (address, port, username, password) and converts the report content into email format. The reports are sent to the relevant departments on a weekly basis.

[1062] Specific examples

[1063] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the current week's key user feedback and the results of sentiment analysis.

[1064] In this way, the system of the present invention automates the entire process of data collection, emotion recognition, analysis, report generation, and distribution, making it possible to grasp user needs and emotions efficiently and at low cost. This system allows service providers to quickly and accurately implement improvement measures that reflect users' emotions and needs.

[1065] The processing flow will be explained below.

[1066] Step 1:

[1067] The server uses the application programming interface (API) to authenticate for data collection. Specifically, it uses the API key, API secret, access token, and access token secret to access the API of Twitter and other SNSs and set authentication information.

[1068] Step 2:

[1069] The server uses the authentication information to collect tweets containing the target hashtags and keywords. It then connects to the Twitter API using the Tweepy library, retrieves 100 tweets containing "product reviews," and extracts information such as the text and posting date and time.

[1070] Step 3:

[1071] The server converts the collected tweet data into a Pandas data frame and stores it in an intermediate database, thereby preparing the data in a format that can be used for subsequent analysis.

[1072] Step 4:

[1073] The server then inputs the stored data into a natural language processing model to analyze the collected text data. Specifically, it uses the Transformers library to generate summaries using generative AI models such as BERT and GPT, which extract the key points of user opinions and feedback.

[1074] Step 5:

[1075] The server inputs the summarized text data into an emotion engine to recognize the user's emotions, and the emotion engine assigns emotion tags such as positive, negative, or neutral to each text.

[1076] Step 6:

[1077] The server extracts important key phrases from the collected data, using TF-IDF and other key phrase extraction techniques to identify key words and phrases in the summarized text.

[1078] Step 7:

[1079] The server automatically generates a report based on the summary, key phrases, and sentiment information. The generated summary, key phrases, and sentiment information are organized according to a template format, and a report is created that clearly shows important user needs and sentiment.

[1080] Step 8:

[1081] The server configures the SMTP server (address, port, username, password) and converts the generated reports into email format, ensuring that the reports contain the required information and analysis results.

[1082] Step 9:

[1083] The server connects to an SMTP server and automatically sends the generated report to a designated recipient, such as a service representative, on a weekly or set schedule.

[1084] The above steps automate the collection of VOC, emotion recognition, analysis, report generation, and distribution, making it possible to grasp user needs and emotions efficiently and at low cost. This system allows service providers to quickly and accurately implement service improvement measures that reflect user requests and emotions.

[1085] Example 2

[1086] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1087] It is difficult for companies to efficiently collect voice of customer (VOC) information quickly and at low cost, and analyze the data to understand user needs and emotions. Furthermore, since it takes a great deal of time and effort to extract the necessary information from the collected data, organize it, and gain insights, there is a need for an automated system that can be easily used even by personnel without specialized knowledge.

[1088] 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.

[1089] In this invention, the server includes means for collecting data using an application programming interface, means for storing the collected data in an intermediate database, means for summarizing the collected data using a natural language processing model, means for recognizing emotions from the summary results and the collected data, means for automatically generating a report based on the summary results and the emotion recognition results, and means for automatically delivering the generated report to a predetermined recipient. This efficiently automates the process from data collection to emotion recognition, and report generation and delivery, enabling companies to grasp user needs and emotions quickly and at low cost.

[1090] An "application programming interface" is a defined interface through which different software systems communicate with each other and exchange data.

[1091] "Means of collecting data" refers to the technologies and methods for obtaining information from various data sources, and typically involves using APIs.

[1092] An "intermediate database" is a database that temporarily stores collected data and uses it in subsequent processing.

[1093] A "natural language processing model" is an algorithm or model that uses machine learning or deep learning techniques to analyze and understand human language.

[1094] "Means of summarization" refers to techniques and methods for extracting the main content from long text data and displaying it in a short form.

[1095] "Means of emotion recognition" refers to algorithms and technologies for extracting emotional information such as positive, negative, and neutral from text data.

[1096] A "summary result" is a short document or phrase generated by a natural language processing model that succinctly conveys the main content of the text data.

[1097] The "emotion recognition result" is the result of emotion analysis on text data, and is classification information of the emotion that each text has.

[1098] "Means for automatically generating reports" refers to technologies and methods for automatically creating report-format documents based on analytical data such as summary results and emotion recognition results.

[1099] "Specified recipient" refers to a predetermined recipient of the report, meaning a specific department or person.

[1100] "Means for automated report delivery" refers to any technology or method for automatically sending generated reports to intended recipients.

[1101] This invention is a system that efficiently collects voice of customer (VOC), automatically analyzes and summarizes it, and then generates reports by combining it with an emotion engine that recognizes user emotions, and delivers them to designated recipients. This system is realized by automating each stage of processing, centered on a server.

[1102] First, in the data collection stage, the server collects data using the application programming interface (API). Specifically, it uses the API key, API secret, access token, and access token secret to retrieve user reviews, comments, and feedback data from social media, app stores, and customer service platforms. This collected data is then stored in an intermediate database.

[1103] Specific examples

[1104] The server uses the SNS API to collect tweets with the hashtag "product review" and stores the data in an intermediate database.

[1105] The server then analyzes the collected data by inputting it into a natural language processing model. Specifically, it uses the Transformers library to utilize generative AI models (such as the BERT and GPT series). This process summarizes the collected text data.

[1106] Specific examples

[1107] The server inputs the collected tweets into a summarization model and summarizes the requests of many users for improvements to the app's user interface.

[1108] The server then uses an emotion engine to recognize user emotions from the collected data, performs sentiment analysis on user comments and reviews, and assigns emotion tags such as positive, negative, or neutral.

[1109] Specific examples

[1110] The server processes tweets with a sentiment engine and classifies them as "positive" tweets such as "I really love the new design" and "I wish the app would stop crashing so often" as "negative" tweets.

[1111] The server then automatically generates a report based on the summarized data and the user's emotional information. The report is organized according to a template format, allowing high-quality reports to be created without specialized knowledge.

[1112] Specific examples

[1113] The server organizes the results of the summarization and sentiment analysis according to templates and generates a "Highlights of the Week" report such as "60% of users are happy with the new feature, but 30% have reported bugs."

[1114] The server then automatically delivers the generated reports to the intended recipients using an SMTP server, with settings including address, port, username and password.

[1115] Specific examples

[1116] Every Monday, the server sends an "Interface Improvement Request Report" to the service representative's email address, containing the current week's key user feedback and the results of sentiment analysis.

[1117] Prompt Sentence Examples

[1118] 1. "Extract and summarize key points from customer reviews and feedback."

[1119] 2. "Analyze the sentiment of the collected tweets and categorize them as positive or negative."

[1120] 3. "Generate a report based on this week's user feedback and sentiment analysis results."

[1121] This system allows service providers to quickly and accurately implement improvement measures that reflect users' feelings and needs.

[1122] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1123] Step 1:

[1124] Data collection

[1125] The server authenticates to collect data using an Application Programming Interface (API), requiring API key, API secret, access token, and access token secret as inputs, and uses these to retrieve user reviews, comments, and feedback data from social media, app stores, and customer service platforms, and obtains raw data as output.

[1126] Specific actions

[1127] The server configures the Twitter API to collect tweets with the hashtag "product reviews." If authentication is successful, the tweet data (username, text, timestamp, etc.) is retrieved.

[1128] Step 2:

[1129] Storing data in an intermediate database

[1130] The server stores the collected data in an intermediate database. As input, the raw data collected in step 1 is used. As output, these data are stored in appropriate tables. The database uses an SQL database (e.g., MySQL or PostgreSQL).

[1131] Specific actions

[1132] The server uses SQL commands to insert the collected tweets into a table called "tweets," storing each tweet's ID, username, text, and timestamp.

[1133] Step 3:

[1134] Data analysis using natural language processing

[1135] The server inputs the data stored in the intermediate database into a natural language processing model for analysis. The input is text data retrieved from the database. The output is a summary result. This process uses the Transformers library and generative AI models (e.g., BERT, GPT series).

[1136] Specific actions

[1137] The server retrieves specific tweet text from the database and inputs it into the summarization model, which then analyzes it to obtain summaries, which are then stored in a new table called "summaries."

[1138] Step 4:

[1139] Emotion recognition

[1140] The server performs emotion recognition on the collected data. Text data is used as input. Emotion tags such as positive, negative, and neutral are generated as output. A sentiment analysis engine (e.g., VADER, TextBlob, etc.) is used.

[1141] Specific actions

[1142] The server feeds the tweet data into a sentiment analysis engine, which classifies each tweet into a sentiment tag (positive, negative, neutral), and adds the results to a new column in the "tweets" table.

[1143] Step 5:

[1144] Automatic report generation

[1145] The server automatically generates a report based on the summary results and emotion recognition results. The analysis results and emotion tags obtained in steps 3 and 4 are used as input. A template-based report is generated as output. A template engine (e.g., Jinja2) is used.

[1146] Specific actions

[1147] The server uses a template engine to embed the analysis results and sentiment tags into reports, creating templates that can be used to generate reports, such as "This week's highlights" that include "60% of users are happy with the new features, but 30% have reported bugs."

[1148] Step 6:

[1149] Automatic report delivery

[1150] The server automatically delivers the generated report to the specified recipients. As input, it uses the generated report and the SMTP server configuration information (address, port, username, password). As output, it sends an email to the recipients.

[1151] Specific actions

[1152] The server connects to the SMTP server, converts the generated report into an email format, and sends it to the email address of the service representative. For example, a schedule can be set to send a "report on interface improvement requests" every Monday.

[1153] By going through each step in this way, the process from data collection to emotion recognition, report generation and distribution is efficiently automated.

[1154] (Application example 2)

[1155] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1156] In recent years, collecting and analyzing voice of the customer (VOC) in brick-and-mortar stores has become essential for improving store operations and customer satisfaction. However, conventional methods have difficulty efficiently collecting and analyzing large volumes of data, and are unable to identify emotions, making it difficult to identify specific areas for improvement or take prompt action. This has made it difficult for brick-and-mortar store managers and staff to accurately understand customers' true needs and emotions.

[1157] 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.

[1158] In this invention, the server includes means for collecting data using an application programming interface, means for summarizing the collected data using a natural language processing model, means for identifying user emotions from the collected data using an emotion engine that recognizes user emotions, means for automatically generating a report based on the summary results and the emotion identification results, and means for automatically delivering the generated report to a specified recipient. This makes it possible to efficiently collect customer feedback and perform emotion analysis, thereby quickly identifying and addressing specific areas for improvement.

[1159] "Application Programming Interface" refers to the well-defined rules or protocols by which software applications communicate with each other.

[1160] "Data collection methods" refers to mechanisms for gathering information and feedback from multiple sources.

[1161] A "natural language processing model" refers to algorithms and technologies that allow computers to understand, analyze, and generate human language.

[1162] "Summarization methods" refers to the processes and techniques used to summarize collected information succinctly.

[1163] "Emotion engine" refers to algorithms and technologies for identifying user emotions from text data.

[1164] "Means for identifying emotions" refers to processes and techniques for extracting and classifying user emotions from text data.

[1165] "Means for automatically generating reports" refers to processes or technologies for automatically creating reports based on summary results and emotion identification results.

[1166] "Means for delivering reports" refers to the process or technology used to transmit generated reports to intended recipients.

[1167] "Key phrase extraction methods" refers to the processes and techniques used to identify and extract important words and phrases from text data.

[1168] "Template-based organization" refers to the process or technique of organizing and structuring information according to a predefined format.

[1169] The present invention is a system that efficiently collects customer feedback (VOC) from physical stores, automatically summarizes and analyzes the feedback, and compiles the results into a report and distributes it to relevant parties. This system is implemented as follows.

[1170] First, the server collects feedback data using an application programming interface (API), using authentication information such as an API key, API secret, or access token. Data collection targets include social media, feedback terminals, and customer service platforms.

[1171] The collected data is stored in an intermediate database. The server then summarizes this data using natural language processing models, including generative AI models such as BERT and the GPT series from the Transformers library. The summarized data is further tagged with sentiment tags, such as positive, negative, or neutral, using a sentiment engine. The sentiment engine, which is also part of natural language processing, uses the Transformers library for data analysis.

[1172] Once the specific data has been summarized and sentiment analyzed, a report is automatically generated based on it. The report contains summarized data, key phrases, and sentiment information. The report generation is based on templates, which organize the information and make the report content easy to read and understand.

[1173] The generated reports are automatically delivered to the specified recipients by the server. The SMTP server settings (address, port, username, password) are configured, the report contents are converted into email format, and sent to the specified email address. Reports are generally delivered weekly, allowing relevant staff to regularly check customer feedback.

[1174] Specific examples

[1175] For example, if a customer uses a feedback terminal to comment, "The product description is difficult to understand," the system collects this comment and stores it in an intermediate database. The server then inputs this data into a natural language processing model and outputs a summary, "Product description improvement request." The emotion engine then analyzes this summary and assigns the emotion tag "negative." Finally, a report is generated based on this information and automatically sent to the store manager.

[1176] Prompt Sentence Examples

[1177] Based on the following example prompt, the system will perform appropriate data collection, summarization, and sentiment analysis:

[1178] 1. Data Collection Prompt: "What is an effective way to collect customer reviews? Specifically, can you provide an example of using an API key to retrieve social media data?"

[1179] 2. Data Summarization Prompt: "What is the latest generative AI model that summarizes collected review data? Please include a Python code example."

[1180] 3. Sentiment Analysis Prompt: "What generative AI model would be suitable for performing sentiment analysis on review data and how would you use it?"

[1181] 4. Report Generation Prompt: "How can I automatically generate a report based on the results of my analysis? Can you give me an example of a template I could use?"

[1182] 5. Report Delivery Prompt: "How can I automatically deliver generated reports to a specified email address? Please provide a Python code example."

[1183] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1184] Step 1: Data collection

[1185] The server collects feedback data using an application programming interface (API). Specifically, it uses the API of social media and feedback devices to obtain customer comments and reviews using authentication information (API key, API secret, access token, etc.). At this time, the data is obtained in JSON format and stored in an intermediate database.

[1186] Input: API key, secret, access token

[1187] Output: Feedback data stored in an intermediate database

[1188] Step 2: Data Summarization

[1189] The server inputs the collected feedback data into a natural language processing model to summarize it. The generative AI models used here are the BERT and GPT series from the Transformers library. The model extracts important parts of the input text data and outputs a concise summary.

[1190] Input: Feedback data stored in the intermediate database

[1191] Output: Summarized text data

[1192] Step 3: Emotion Identification

[1193] The server then inputs the summarized text data into an emotion engine for sentiment analysis, which also uses the Transformers library to tag each piece of text with an emotion tag, such as positive, negative, or neutral.

[1194] Input: Summarized text data

[1195] Output: Text data with emotion tags

[1196] Step 4: Generate a report

[1197] The server automatically generates a report based on the summary and emotion identification results. This report is based on a template and organizes summaries, emotion tags, key phrases, etc.

[1198] Input: Text data with emotion tags

[1199] Output: Organized reports in template format

[1200] Step 5: Report Delivery

[1201] The server automatically delivers the generated report to the specified recipient by configuring the SMTP server settings (address, port, username, password), converting the report content into email format, and sending it.

[1202] Input: Reports organized in template format

[1203] Output: Report sent to recipient's (e.g. store manager) email address

[1204] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1205] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1206] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1207] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1208] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1209] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1210] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1211] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1212] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1213] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1214] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1215] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1216] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1217] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1218] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1219] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1220] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1221] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1222] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1223] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1224] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1225] The following is further disclosed regarding the above embodiment.

[1226] (Claim 1)

[1227] a means for collecting data using an application programming interface;

[1228] A means for summarizing the collected data using a natural language processing model;

[1229] A means for automatically generating reports based on the summary results;

[1230] means for automatically delivering the generated reports to intended recipients;

[1231] A system including:

[1232] (Claim 2)

[1233] 10. The system of claim 1, further comprising: means for extracting key phrases from the collected data.

[1234] (Claim 3)

[1235] 10. The system of claim 1, further comprising means for configuring the generated report based on a template.

[1236] "Example 1"

[1237] (Claim 1)

[1238] a means for collecting data using an application programming interface;

[1239] means for storing the collected data in an intermediate database;

[1240] means for analyzing the stored data using a natural language processing model to generate a summary;

[1241] A means for automatically generating reports based on the summary results;

[1242] means for automatically delivering the generated reports to intended recipients;

[1243] A system including:

[1244] (Claim 2)

[1245] 10. The system of claim 1, further comprising: means for extracting key phrases from the collected data.

[1246] (Claim 3)

[1247] 10. The system of claim 1, further comprising means for configuring the generated report based on a template.

[1248] "Application Example 1"

[1249] (Claim 1)

[1250] a means for collecting data using an application programming interface;

[1251] A means for summarizing the collected data using a natural language processing model;

[1252] A means for automatically generating reports based on the summary results;

[1253] means for automatically delivering the generated reports to intended recipients;

[1254] means for summarizing the collected data based on its contents, generating reports according to a predetermined schedule, and distributing the reports via email to predetermined recipients;

[1255] A system including:

[1256] (Claim 2)

[1257] 10. The system of claim 1, further comprising: means for extracting key phrases from the collected data.

[1258] (Claim 3)

[1259] 10. The system of claim 1, further comprising means for configuring the generated report based on a template.

[1260] "Example 2: Combining Emotion Engines"

[1261] (Claim 1)

[1262] a means for collecting data using an application programming interface;

[1263] means for storing the collected data in an intermediate database;

[1264] A means for summarizing the collected data using a natural language processing model;

[1265] a means for recognizing emotions from the summary results and collected data;

[1266] A means for automatically generating a report based on the summary result and the emotion recognition result;

[1267] means for automatically delivering the generated reports to intended recipients;

[1268] A system including:

[1269] (Claim 2)

[1270] 10. The system of claim 1, further comprising: means for extracting key phrases from the collected data.

[1271] (Claim 3)

[1272] 10. The system of claim 1, further comprising means for configuring the generated report based on a template.

[1273] "Application example 2 when combining emotion engines"

[1274] (Claim 1)

[1275] a means for collecting data using an application programming interface;

[1276] A means for summarizing the collected data using a natural language processing model;

[1277] means for identifying a user's emotion from the collected data using an emotion engine that recognizes the user's emotion;

[1278] A means for automatically generating a report based on the summary result and the emotion identification result;

[1279] means for automatically delivering the generated reports to intended recipients;

[1280] A system including:

[1281] (Claim 2)

[1282] 10. The system of claim 1, further comprising: means for extracting key phrases from the collected data.

[1283] (Claim 3)

[1284] 10. The system of claim 1, further comprising means for configuring the generated report based on a template. [Explanation of symbols]

[1285] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for collecting data using an application programming interface; A means for summarizing the collected data using a natural language processing model; A means for automatically generating reports based on the summary results; means for automatically delivering the generated reports to intended recipients; A system including:

2. The system of claim 1 further comprising means for extracting key phrases from the collected data.

3. The system of claim 1 further comprising means for configuring the generated report based on a template.

Images