system
A system using natural language processing and generative AI models addresses inefficiencies in local government services by analyzing inquiries, retrieving information, and optimizing responses, enhancing user satisfaction and service quality.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- SOFTBANK GROUP CORP
- Filing Date
- 2024-12-05
- Publication Date
- 2026-06-17
AI Technical Summary
Existing systems face inefficiencies in responding to inquiries and providing information in local government services, leading to delays and a lack of immediate response, which burdens administrative staff and reduces user satisfaction.
A system utilizing natural language processing and generative AI models to analyze user inquiries, retrieve relevant information from databases, and generate optimized responses, while also collecting user feedback for continuous improvement.
This system enables efficient and timely information provision, reduces administrative complexity, and improves local government services by providing accurate and personalized responses based on user feedback.
Smart Images

Figure 2026098724000001_ABST
Abstract
Description
Technical Field
[0001] The technology of the present disclosure relates to a system.
Background Art
[0002] Patent Document 1 discloses a persona chatbot control method performed by at least one processor, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance as a response to the user utterance.
Prior Art Documents
Patent Documents
[0003]
Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] This invention aims to reduce the burden on users in administrative procedures and improve the efficiency of local government services. In particular, the problems are the delay in responding to inquiries and providing information manually in the past, and the inefficiency of collecting feedback from users. As a result, the shortage of administrative staff and the lack of immediate response to citizens have become problems.
Means for Solving the Problems
[0005] To address this challenge, the present invention provides a system that receives user inquiries, analyzes them through natural language processing, and generates responses using a generative AI model. Furthermore, it includes a function to generate optimized responses by referencing local government database information. This system can also collect user feedback and improve local government services through response generation. This will lead to more efficient administrative procedures and faster information provision to citizens.
[0006] A "user" refers to an individual or organization that uses the system to make an inquiry.
[0007] "Terminal means" refers to a device or interface for a user to input information and interact with the system.
[0008] "Computing means" refers to a program or device for processing queries received from users and performing natural language processing or other calculations.
[0009] "Generative model means" refers to a machine learning model that generates appropriate answers based on data analyzed by a computer.
[0010] "Information acquisition means" refers to a mechanism for obtaining data related to the generated response from an external database and updating the information to the latest version.
[0011] "Response generation means" refers to a process or device for generating a final response to be provided to the user based on acquired information and transmitting it to a terminal means.
[0012] "Feedback" refers to the opinions, requests, and suggestions that users provide to a system.
[0013] "Data mining techniques" refer to technologies or methods for analyzing policy information of local governments and extracting necessary data. [Brief explanation of the drawing]
[0014] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2, which combines an emotion engine. [Modes for carrying out the invention]
[0015] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.
[0016] First, the terms used in the following description will be explained.
[0017] In the following embodiments, a labeled processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.
[0018] In the following embodiments, a labeled RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.
[0019] In the following embodiments, a labeled storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.
[0020] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).
[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."
[0022] [First Embodiment]
[0023] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.
[0024] As shown in Figure 1, the 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 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).
[0026] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.
[0027] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and 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 perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.
[0029] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.
[0030] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.
[0031] As shown in Figure 2, in the data processing device 12, a specific processing 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" related to the technology of this 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 according to the specific processing program 56 executed on the RAM 30.
[0032] The storage 32 stores the data generation model 58 and the 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 processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.
[0034] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".
[0035] This invention provides a system that enables efficient and rapid information provision when users make inquiries related to government services. The main components of the system include a terminal used by the user, a server for data processing, and a generative AI model.
[0036] Users can use a terminal to input questions regarding administrative procedures. This terminal has the functionality to send data to a server via an internet connection and receive responses from the server.
[0037] When the server receives query data sent by a user, it analyzes the data using natural language processing techniques. This analysis process identifies the intent of the query and retrieves relevant information from the database.
[0038] Generative AI models are responsible for generating appropriate answers based on the analyzed information. These AI models are trained on large historical datasets, enabling them to create natural and accurate responses to user questions.
[0039] For example, if a user asks, "I want to change my address, how do I do that?", the server analyzes this inquiry and collects information about the relevant procedures and required documents from its database. The generative AI model then uses this information to construct an answer that includes specific steps.
[0040] The server then formats the generated response and sends it to the user's device. The user can then review this information displayed on their device screen and complete the necessary procedures online.
[0041] This invention reduces the complexity of administrative procedures, allowing users to efficiently obtain accurate information. Furthermore, user feedback is collected via the server and used to improve the system in the future. This enables local governments to provide higher quality services.
[0042] The following describes the processing flow.
[0043] Step 1:
[0044] Users input questions and inquiries about procedures using a terminal. In this process, they enter the specific question in text format into the interface.
[0045] Step 2:
[0046] The terminal receives data entered by the user and sends query data to the server via the internet. A protocol is used to ensure security during data transmission.
[0047] Step 3:
[0048] The server receives query data sent from the terminal. The received data is temporarily stored in a data buffer.
[0049] Step 4:
[0050] The server uses a natural language processing (NLP) module to analyze incoming queries. Specifically, it extracts keywords and context from text data to understand the user's intent.
[0051] Step 5:
[0052] The server launches a generative AI model based on the analyzed information. The generative AI model uses a pre-trained algorithm to construct the optimal answer based on the analysis results.
[0053] Step 6:
[0054] The server reviews the generated response and queries the database as needed. It retrieves relevant information and supplements or updates the response.
[0055] Step 7:
[0056] The server formats the final response into a user-friendly format. At this stage, the response is arranged in a natural language structure and is suitable for the user interface.
[0057] Step 8:
[0058] The server sends a formatted response to the terminal. This response is optimized to be real-time or near real-time.
[0059] Step 9:
[0060] The terminal displays the response received from the server in the user interface. Based on the displayed information, the user can determine the next steps to take or ask further questions.
[0061] Step 10:
[0062] Users can send feedback via their devices as needed, suggesting system improvements or making additional inquiries. This feedback will be used to inform subsequent data analysis.
[0063] (Example 1)
[0064] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."
[0065] In inquiries related to government services, users face the challenge of difficulty in obtaining quick and accurate information. Traditional systems can be inefficient in their process of appropriately retrieving and clearly presenting information based on the content of the inquiry. Furthermore, there are insufficient means to efficiently collect user feedback and utilize it for system improvement. Additionally, there is a need to optimize responses by taking into account local government policy information.
[0066] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.
[0067] In this invention, the server includes communication means for receiving inquiries from users, a processing unit for analyzing inquiries received from the communication means and performing natural language processing, information retrieval means for searching a database and extracting information based on the intent identified by the processing unit, response model means for generating answers using a generation AI based on the extracted information, and transmission means for formatting the generated answers to aid user understanding and transmitting them to a terminal. This enables users to obtain information on administrative procedures quickly and accurately. Furthermore, user feedback can be accepted and used to optimize the system, enabling the generation of appropriate answers based on policy information from public institutions.
[0068] "Communication means" refers to devices or interfaces used to receive inquiries from users and send data to a server.
[0069] A "processing unit" is a computing mechanism that analyzes received queries and performs natural language processing.
[0070] An "information retrieval tool" is a function that searches a database and extracts relevant information based on the user's intent identified through analysis.
[0071] A "response model means" is a system that uses a generation AI based on extracted information to generate appropriate responses for the user.
[0072] A "transmission method" is a mechanism for formatting the generated response and accurately transmitting it to the user's device.
[0073] A "feedback management system" is a mechanism for receiving feedback from users and using it to improve the system.
[0074] An "optimization method" is a technology that appropriately adjusts and optimizes the generated responses based on policy information from public institutions.
[0075] This invention provides a system that enables users to receive timely and accurate information when making inquiries related to government services. The system includes a terminal used by the user, a server for data processing, and a generative AI model.
[0076] Users can input questions about administrative procedures through the terminal's interface. This terminal is responsible for transmitting the inquiry to the server via an internet connection.
[0077] The server has a processing unit for analyzing received queries using natural language processing techniques to identify the user's intent. Natural language processing techniques used for analysis include, for example, tokenization, syntactic analysis, and intent classification algorithms.
[0078] After the processing unit identifies the intent, it uses information retrieval means to search the database based on that information and extract relevant information. This information is often obtained from databases of documents and procedures related to administrative processes.
[0079] Based on the extracted information, a generative AI model generates appropriate answers to questions as a response model. This AI model is trained using deep learning and large-scale language models, and creates natural and accurate responses based on a vast amount of historical data. For example, in response to a prompt such as "Please tell me what documents and procedures are required to renew my driver's license," it can provide detailed information about the necessary documents, locations, and procedures.
[0080] The server then formats the generated response using a communication method and accurately sends it to the user's device. The user can then review the information displayed on their device and complete the necessary procedures online.
[0081] This system reduces the complexity of administrative procedures, allowing users to efficiently obtain accurate information. Furthermore, user feedback is collected as a feedback management tool, contributing to future improvements of the system. This enables public institutions to provide higher quality services.
[0082] The flow of the specific processing in Example 1 will be explained using Figure 11.
[0083] Step 1:
[0084] The user enters questions about administrative procedures through the terminal's interface. The terminal converts the entered questions into digital data and sends them to a server via the internet. The input is a user query in text format, and the output is the query sent to the server.
[0085] Step 2:
[0086] The server receives user query data from the terminal into its processing unit. Natural language processing (NLP) techniques are used to analyze this query data. Specifically, the query is divided into words using tokenization, and the role of each word is identified through part-of-speech tagging. As a result of this analysis, the intent of the query is identified, and a structured format of the query data is output.
[0087] Step 3:
[0088] Using information retrieval tools within the server, information related to the intent identified from the analyzed data is retrieved from the database. This process involves database access via SQL queries and other methods to extract relevant information and items. The input is structured intent data, and the output is related information data.
[0089] Step 4:
[0090] The generative AI model generates a response based on the extracted information. The model is trained using a deep learning algorithm and leverages knowledge gained from large datasets to generate natural-sounding sentences. The input is relevant information data, and the output is the generated response text.
[0091] Step 5:
[0092] The server formats the responses generated by the generative AI model and sends them to the user's terminal. The formatting process adjusts the text to be human-readable and optimizes the visual arrangement of information. The input is the generated response text, and the output is the formatted response data.
[0093] Step 6:
[0094] The terminal receives formatted responses sent from the server and displays them to the user. The user reviews this information on the terminal and identifies the actions necessary to proceed with administrative procedures. The input is formatted response data, and the output is the display of information to the user.
[0095] (Application Example 1)
[0096] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."
[0097] There is a need to alleviate the complexity of information retrieval and the difficulty in understanding procedural processes that citizens face in administrative procedures. In particular, a lack of understanding of administrative procedures and required documents can be a significant burden for citizens. Furthermore, if administrative services are not sufficiently digitized, the time required to obtain information and complete procedures increases, reducing convenience for citizens. By solving these problems, it is expected that the quality of administrative services will improve and citizens' lives will become smoother.
[0098] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.
[0099] In this invention, the server includes an information terminal for receiving inquiries from users, a computing device for performing natural language processing based on the inquiries received from the information terminal, and a generative model for generating responses based on the natural language data processed by the computing device. This makes it possible to analyze voice input from citizens and provide accurate and rapid guidance on procedures and necessary materials.
[0100] An "information terminal" is a device used to receive inquiries from users and enables voice input and text input.
[0101] A "computational device" is a computer system that analyzes inquiries received from information terminals using natural language processing technology and interprets the user's intent.
[0102] A "generative model" is an artificial intelligence model that generates appropriate answers based on data analyzed by a computing device.
[0103] An "information acquisition device" is a system that updates the responses generated by a generative model with reference data and aggregates the necessary information.
[0104] A "response generation device" is a device that has the function of transmitting information integrated by an information acquisition device to an information terminal used by the user.
[0105] A "decision-making device" is a system that analyzes voice input from citizens and provides specific guidance on procedures and necessary documents.
[0106] A "knowledge extraction device" is a device that optimizes the answers output by a generative model based on the policy information of government agencies at that time.
[0107] A system for implementing this invention consists of the following components: an information terminal, a computing device, a generative model, an information acquisition device, a response generation device, and a decision-making device.
[0108] The information terminals used will be smartphones and smart glasses that allow users to input inquiries via voice or text. These terminals communicate with a server via the internet and transmit user inquiries to a computing device. Voice input will be converted to text using the Google® Speech-to-Text API.
[0109] The server, as a computing device, is built using the Flask framework and processes user input using natural language processing. The spaCy library is used for this processing, understanding the intent of queries and identifying the next data to reference. Furthermore, the GPT-3® generative model is incorporated, generating natural and accurate responses based on the analysis results. This generative model learns from large datasets and provides answers based on real-world context.
[0110] The information acquisition device compares the responses generated by the generation model with existing databases and extracts and updates the necessary information. The response generation device formats this information and transmits it to the user's information terminal. Ultimately, the user receives guidance on procedures and required documents through the information terminal, enabling citizens to smoothly proceed with the necessary procedures.
[0111] For example, if a citizen asks via voice, "Please tell me how to register my new address," the system will analyze the inquiry and guide them to the necessary documents and submission locations. The information the user needs will be displayed on the terminal, and a path will be set up to complete the procedure online.
[0112] As an example of a prompt, a user query is sent to the system in the format "User Query: 'Please tell me how to register a new address.'" Based on this, the request to the generative model is expressed as "Explain the steps, required documents, and possible online submission options for new address registration."
[0113] The flow of a specific process in Application Example 1 will be explained using Figure 12.
[0114] Step 1:
[0115] The user uses a smartphone or smart glasses to input their inquiry via voice or text. The input voice is converted into text data using the Google Speech-to-Text API and sent from the smart device to the server.
[0116] Step 2:
[0117] The server is built using Flask and receives input data from users. This data is processed using the spaCy library to analyze the intent of the query. Based on the analyzed intent, the next necessary data is identified. The input is the user's query, and the output is the analysis result including the identified intent.
[0118] Step 3:
[0119] The server sends the analysis results to the generative AI model. The GPT-3 generative model receives the intent in the form of a prompt and generates a relevant response. The prompt is sent in the form of "User Query: 'Please tell me how to register a new address.'" The input to the model is the analysis results, and the output is the generated natural language response.
[0120] Step 4:
[0121] The server receives the response from the generative model, and the information retrieval device accesses the database to collect necessary additional information based on the generated response. Database matching updates and integrates the information. The input is the response from the generative model, and the output is the integrated additional information.
[0122] Step 5:
[0123] The response generation device formats the integrated information and sends the response to the user's information terminal. The user can then easily access information about the procedure and necessary documents on their terminal. The input is integrated information, and the output is a formatted response.
[0124] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.
[0125] This invention relates to a system that receives inquiries from users, analyzes those emotions using an emotion engine, and generates appropriate responses through a generative AI model. The aim of this system is to improve the user experience and the quality of personalized responses.
[0126] The system's main components include terminals for users to input inquiries, servers for data processing and sentiment analysis, and generative AI models and sentiment engines. These elements communicate over the internet and exchange information in real time.
[0127] The user enters text about questions or problems through their terminal. The entered data is sent to the server, which holds the received data in a buffer.
[0128] On the server, the natural language processing module is executed first to analyze the user's query. This analysis extracts the specific information the user intends to convey. Next, the emotion engine uses the same data to determine the user's emotional state (e.g., joy, anger, sadness). The emotion engine uses sophisticated algorithms to estimate emotions based on factors such as tone of voice and text, and keywords.
[0129] The generative AI model considers the results of natural language processing and emotion engine analysis to generate responses tailored to specific emotions. The tone and expression of the response are adjusted according to the emotion. For example, if an inquiry includes dissatisfaction, the generative model will generate a response in milder language that includes countermeasures that take this into account.
[0130] The server scrutinizes the generated response by linking it with the database and updates the information as needed. The response is then formatted and sent to the terminal.
[0131] The device displays responses in a user-optimized format, making it easier for users to understand the next steps. For example, if a user is dissatisfied with an administrative service, the system will return a response that emphasizes explanations of countermeasures, including expressions of gratitude and apology.
[0132] This invention enables the system to respond sensitively to user emotions and provide more personalized services. Furthermore, it is expected that administrative services will be further improved through continuous improvement based on feedback.
[0133] The following describes the processing flow.
[0134] Step 1:
[0135] Users can input questions and comments about administrative procedures and services through their terminals. For example, they can input something like, "My address change procedure is delayed."
[0136] Step 2:
[0137] The terminal processes the input data and sends it to the server using a secure protocol. The data is properly structured according to the transmission format.
[0138] Step 3:
[0139] The server temporarily stores the query received from the terminal in a data buffer. This data is then used in the subsequent analysis process.
[0140] Step 4:
[0141] The server activates a natural language processing module to analyze the query. It extracts keywords and context from the text to identify the information the user intends to receive.
[0142] Step 5:
[0143] The server uses an emotion engine to analyze the sentiment in the query text. For example, it can determine from the text content whether the user is feeling frustrated.
[0144] Step 6:
[0145] The server calls a generative AI model based on the analyzed inquiry content and sentiment data. The AI model generates an answer that reflects both of these factors.
[0146] Step 7:
[0147] The server uses information retrieval tools to obtain relevant details from the database regarding the generated response, and then uses those details to supplement or update the response.
[0148] Step 8:
[0149] The server adjusts the tone and style of the emotionally generated responses, refining the formatting. For example, it might add nuances of apology or change the language to be more polite.
[0150] Step 9:
[0151] The server sends the completed response to the terminal. This process is carried out quickly to ensure the user has the optimal reaction time.
[0152] Step 10:
[0153] The device displays the answer to the user, allowing the user to obtain specific actions or further guidance regarding their inquiry.
[0154] Step 11:
[0155] Users can provide feedback on the answers. This feedback is returned to the system and stored as data for future improvements.
[0156] (Example 2)
[0157] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".
[0158] Existing customer service systems often provide uniform answers without adequately considering user emotions, resulting in a poor user experience. Furthermore, the quality of the generated answers is not optimized for the user's specific emotional state, leading to insufficient individualized support.
[0159] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.
[0160] In this invention, the server includes means for performing natural language analysis using a computing device, means for inferring the user's emotional state using an emotion analysis device, and means for generating an emotion-based response using a generative model device. This enables the generation of sophisticated responses that take emotional states into consideration.
[0161] A "terminal device" is a device that receives inquiries from users and sends data to a server.
[0162] A "processing unit" is a device equipped with computing resources for analyzing data obtained from terminal devices.
[0163] An "emotion analysis device" is a device that has the function of inferring the emotional state of a user from their inquiry data.
[0164] A "generative model device" is a device that generates appropriate responses based on analyzed data and emotional states.
[0165] An "information management device" is a device that compares the generated responses with a storage medium and updates the information as needed.
[0166] A "response generation device" is a device that has the function of transmitting information acquired by an information management device to a terminal device.
[0167] A "data analysis device" is a device used to apply optimization based on the user's emotional state to policy information.
[0168] This invention is a system comprising a terminal device for the user to input inquiries, a computing device for analyzing data, an emotion analysis device for analyzing the user's emotions, a generative model device for generating answers using a generative model, an information management device for scrutinizing answers and managing information, and a response creation device for finally creating a response and sending it to the user's terminal.
[0169] The terminal device receives text input from the user and transmits it to the server as a digital signal. Typical personal computers and smartphones are used.
[0170] The computing units deployed on the server are equipped with high-performance CPUs and memory, and analyze data received from users. Here, natural language processing technology is used to enable the analysis of user intent.
[0171] The emotion analysis device extracts text tone and keywords from the text data entered by the user to infer the user's emotional state. By applying a highly accurate algorithm, it can accurately determine emotions such as joy, anger, and sadness.
[0172] The generative modeling device generates responses in a tone appropriate to the user's emotional state, based on the analysis results. The generative AI model used in this process employs the latest natural language generation technology.
[0173] The information management device compares the generated responses with the storage medium to verify the information's up-to-dateness. Information is updated as needed, ensuring that accurate data is always maintained.
[0174] The response generation device sends the reviewed response to the user's terminal device and formats it for appropriate display. This allows the user to logically select their next action.
[0175] For example, if a user uses the prompt "Tell me about recent improvements in government services," the system will analyze the user's intent and return a response that provides detailed information while taking emotions into consideration. This is expected to improve the user experience and dramatically enhance the quality of personalized responses.
[0176] The flow of the specific processing in Example 2 will be explained using Figure 13.
[0177] Step 1:
[0178] The user enters their inquiry in text format using a terminal device. The entered text data serves as a prompt to the system. For example, the user might enter, "Please tell me about recent government services." The terminal then converts this text data into a digital signal and sends it to the server.
[0179] Step 2:
[0180] The server stores the text data received from the terminal in a receive buffer for analysis. It decodes the digital signal as input into text, preparing it for analysis. This data is then sent to the processing unit.
[0181] Step 3:
[0182] The server passes the text data received via the computing unit to a natural language processing module for analysis. The input is raw text data from the user, and the output is keywords related to the analyzed intent and topic. This analysis clarifies the user's specific information needs.
[0183] Step 4:
[0184] The server passes the analyzed results to an emotion analysis device, which evaluates the emotional state based on the text data. The analyzed keywords are used as input, and the user's emotional state (e.g., joy, anger, sadness) is obtained as output. Emotion analysis is performed based on the tone and keywords of the text.
[0185] Step 5:
[0186] The server activates its generative AI model function and generates responses using the results of natural language processing and sentiment analysis. The input here consists of two data points: the user's intent and emotion, and the output is a natural-sounding response adjusted to match the emotion. As a result, responses that are both reliable and approachable are obtained.
[0187] Step 6:
[0188] The server sends the generated responses to the information management device, where they are compared with the database to verify the timeliness and accuracy of the information. Here, the generated responses are checked against the internal data repository to confirm that the information is up-to-date, and then the results are converted into a format that can be sent to the terminal.
[0189] Step 7:
[0190] The server sends the generated response to the terminal device via the response generation device. Finally, the terminal displays this response to the user. The output displays an optimized answer to the user's question and provides information to prompt the user's next action.
[0191] Through these steps, the system provides appropriate information based on the user's emotions in real time, improving the user experience.
[0192] (Application Example 2)
[0193] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".
[0194] Customer support on modern e-commerce sites presents a challenging task, requiring a thorough understanding and appropriate response to user emotions. Uniform, emotion-insensitive responses can degrade the user experience. Technologies are needed to improve this.
[0195] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.
[0196] In this invention, the server includes an input device means for receiving inquiries from users, a computing device means for performing natural language processing based on the inquiries received from the input device means, and an emotion analysis means for performing emotion analysis based on the natural language data processed by the computing device means. This makes it possible to generate responses that take the user's emotions into consideration.
[0197] "An input device for receiving inquiries from users" refers to a device that has the function of allowing users to input inquiries to the system.
[0198] A "computation device means" is a device that has computational functions for performing natural language processing based on data received from an input device means.
[0199] An "emotion analysis device" is a device that has the function of analyzing a user's emotional state from natural language data processed by a computing device.
[0200] A "generative model device means" is a device that has the function of automatically generating appropriate answers to user inquiries based on the results of sentiment analysis means.
[0201] "Information acquisition means" refers to a device that has the function of referencing the response generated by the generation model device means with information from a database and updating it as necessary.
[0202] A "response generation means" is a device that has the function of sending a response to the user in an appropriate format based on the information obtained by the information acquisition means.
[0203] This invention provides a system for improving the response accuracy of customer support on e-commerce websites. The system consists of various components and is implemented by an input device used by the user and multiple modules that operate on a server.
[0204] The user enters their inquiry in text format using an input device. This input device is a common computer terminal such as a smartphone or personal computer. The data entered by the user is transmitted to the server via the internet.
[0205] The server first uses a natural language processing engine, positioned as a computing device, to analyze the user's query. Existing libraries such as "SpaCy" and "Transformers" can be used for natural language processing. This process extracts the specific content intended by the user.
[0206] Next, the emotion analysis system analyzes the user's emotional state based on the text data. This emotion analysis uses algorithms to analyze the tone of the text and specific language patterns. As a result, the user's emotions are recognized as "joy," "anger," etc.
[0207] The generative model device takes the results of natural language processing and sentiment analysis as input and generates the optimal response in plain text form. The generated response is compared with a database by the information acquisition means, and the information is updated as needed. This database referencing ensures that the response to the user is based on the latest information at the present time.
[0208] Finally, the response generation mechanism sends the generated answer to the user's input device in an appropriate format, allowing the user to receive a response to their inquiry. This is displayed on the terminal screen, making it easy for the user to understand what action to take next.
[0209] For example, if a user inquires that they are dissatisfied because their delivery is delayed, the system will use sentiment analysis to read the dissatisfaction, and a generative model will generate a response that includes an apology and a solution that takes the user's feelings into consideration. The response might be: "We apologize for the delay. We are currently checking the delivery status and will contact you shortly."
[0210] An example of a prompt message input to the generating AI model is: "User inquiry: 'My delivery is delayed. Please take action.' Sentiment analysis result: 'Angry' Please provide an appropriate response."
[0211] The flow of a specific process in Application Example 2 will be explained using Figure 14.
[0212] Step 1:
[0213] The user uses an input device to enter their inquiry in text format and presses the submit button. The entered text data is transmitted to the server via the internet. At this point, the input is raw data from the user.
[0214] Step 2:
[0215] The server's computing system initiates natural language processing using the received text data. Using natural language processing libraries (e.g., SpaCy or Transformers), it extracts the specific intent of the query and related information from the text. The input is the user's raw text data, and the output is the parsed structured data.
[0216] Step 3:
[0217] The server uses sentiment analysis tools to estimate the user's emotional state based on structured data. The sentiment analysis algorithm analyzes the tone and keywords of the text and classifies emotions as "joy" or "anger." In this step, the input is output data from natural language processing, and the emotional state is output.
[0218] Step 4:
[0219] The generative model device generates the optimal response based on the sentiment analysis results and structured data. Using the generative AI model, text with appropriate wording corresponding to the inquiry content and the user's emotions is created. The input here is emotions and structured data, and the output is the response text.
[0220] Step 5:
[0221] The server compares the generated response with the database through the information retrieval mechanism and updates the information as needed. Database lookups are performed to ensure that the response is based on the latest information. The input for this step is the generated response, and the output is the verified response.
[0222] Step 6:
[0223] The response generation mechanism sends the confirmed response text to the user's input device. The display on the terminal is in a format that is easy for the user to understand. The final input is the confirmed response, and the output is the final response displayed on the user's screen.
[0224] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating 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.
[0225] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.
[0226] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.
[0227] [Second Embodiment]
[0228] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.
[0229] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.
[0230] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).
[0231] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.
[0232] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.
[0233] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).
[0234] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.
[0235] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.
[0236] The specific processing program 56 is an example of a "program" relating to the technology of this 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.
[0237] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.
[0238] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.
[0239] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. 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".
[0240] This invention provides a system that enables efficient and rapid information provision when users make inquiries related to government services. The main components of the system include a terminal used by the user, a server for data processing, and a generative AI model.
[0241] Users can use a terminal to input questions regarding administrative procedures. This terminal has the functionality to send data to a server via an internet connection and receive responses from the server.
[0242] When the server receives query data sent by a user, it analyzes the data using natural language processing techniques. This analysis process identifies the intent of the query and retrieves relevant information from the database.
[0243] Generative AI models are responsible for generating appropriate answers based on the analyzed information. These AI models are trained on large historical datasets, enabling them to create natural and accurate responses to user questions.
[0244] For example, if a user asks, "I want to change my address, how do I do that?", the server analyzes this inquiry and collects information about the relevant procedures and required documents from its database. The generative AI model then uses this information to construct an answer that includes specific steps.
[0245] The server then formats the generated response and sends it to the user's device. The user can then review this information displayed on their device screen and complete the necessary procedures online.
[0246] This invention reduces the complexity of administrative procedures, allowing users to efficiently obtain accurate information. Furthermore, user feedback is collected via the server and used to improve the system in the future. This enables local governments to provide higher quality services.
[0247] The following describes the processing flow.
[0248] Step 1:
[0249] Users input questions and inquiries about procedures using a terminal. In this process, they enter the specific question in text format into the interface.
[0250] Step 2:
[0251] The terminal receives data entered by the user and sends query data to the server via the internet. A protocol is used to ensure security during data transmission.
[0252] Step 3:
[0253] The server receives query data sent from the terminal. The received data is temporarily stored in a data buffer.
[0254] Step 4:
[0255] The server uses a natural language processing (NLP) module to analyze incoming queries. Specifically, it extracts keywords and context from text data to understand the user's intent.
[0256] Step 5:
[0257] The server launches a generative AI model based on the analyzed information. The generative AI model uses a pre-trained algorithm to construct the optimal answer based on the analysis results.
[0258] Step 6:
[0259] The server reviews the generated response and queries the database as needed. It retrieves relevant information and supplements or updates the response.
[0260] Step 7:
[0261] The server formats the final response into a user-friendly format. At this stage, the response is arranged in a natural language structure and is suitable for the user interface.
[0262] Step 8:
[0263] The server sends a formatted response to the terminal. This response is optimized to be real-time or near real-time.
[0264] Step 9:
[0265] The terminal displays the response received from the server in the user interface. Based on the displayed information, the user can determine the next steps to take or ask further questions.
[0266] Step 10:
[0267] Users can send feedback via their devices as needed, suggesting system improvements or making additional inquiries. This feedback will be used to inform subsequent data analysis.
[0268] (Example 1)
[0269] Next, we will describe Example 1. 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."
[0270] In inquiries related to government services, users face the challenge of difficulty in obtaining quick and accurate information. Traditional systems can be inefficient in their process of appropriately retrieving and clearly presenting information based on the content of the inquiry. Furthermore, there are insufficient means to efficiently collect user feedback and utilize it for system improvement. Additionally, there is a need to optimize responses by taking into account local government policy information.
[0271] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.
[0272] In this invention, the server includes communication means for receiving inquiries from users, a processing unit for analyzing inquiries received from the communication means and performing natural language processing, information retrieval means for searching a database and extracting information based on the intent identified by the processing unit, response model means for generating answers using a generation AI based on the extracted information, and transmission means for formatting the generated answers to aid user understanding and transmitting them to a terminal. This enables users to obtain information on administrative procedures quickly and accurately. Furthermore, user feedback can be accepted and used to optimize the system, enabling the generation of appropriate answers based on policy information from public institutions.
[0273] "Communication means" refers to devices or interfaces used to receive inquiries from users and send data to a server.
[0274] A "processing unit" is a computing mechanism that analyzes received queries and performs natural language processing.
[0275] An "information retrieval tool" is a function that searches a database and extracts relevant information based on the user's intent identified through analysis.
[0276] A "response model means" is a system that uses a generation AI based on extracted information to generate appropriate responses for the user.
[0277] A "transmission method" is a mechanism for formatting the generated response and accurately transmitting it to the user's device.
[0278] A "feedback management system" is a mechanism for receiving feedback from users and using it to improve the system.
[0279] An "optimization method" is a technology that appropriately adjusts and optimizes the generated responses based on policy information from public institutions.
[0280] This invention provides a system that enables quick and accurate information provision when a user makes an administrative-related inquiry. The system includes a terminal used by the user, a server for data processing, and a generative AI model.
[0281] The user can input questions regarding administrative procedures through the interface of the terminal. This terminal serves the role of transmitting the inquiry content to the server via an internet connection as a communication means.
[0282] The server has a processing device for analyzing the received inquiry using natural language processing technology to identify the user's intention. Natural language processing technologies used for the analysis include, for example, algorithms for tokenization, syntax analysis, and intention classification.
[0283] After the processing device identifies the intention, it searches the database using information reference means based on that information and extracts relevant information. The information is often obtained from databases of documents and procedures related to administrative procedures.
[0284] Based on the extracted information, the generative AI model, as response model means, generates an appropriate answer to the question. This AI model is trained by deep learning and large language models and creates natural and accurate responses based on a vast dataset of past data. For example, for an example prompt sentence like "Please tell me the required documents and procedures for renewing a driver's license," it is possible to provide detailed information about the required documents, locations, and procedures.
[0285] After that, the server formats the generated answer using transmission means and accurately transmits it to the user's terminal. The user can check the information displayed on the terminal and complete the necessary procedures online.
[0286] This system reduces the complexity in administrative procedures, enabling users to efficiently obtain accurate information. Furthermore, feedback from users is collected as feedback management means and can contribute to future improvements of the system. As a result, public institutions can provide higher-quality services.
[0287] The flow of a specific process in Example 1 will be described with reference to FIG. 11.
[0288] Step 1:
[0289] The user inputs a question regarding an administrative procedure through the interface of the terminal. The terminal converts the input question into digital data and transmits it to the server via the Internet. The input is a text-formatted user query, and the output is the query transmission to the server.
[0290] Step 2:
[0291] The server receives the user's query data received from the terminal by the processing device. To analyze this query data, natural language processing (NLP) technology is used. Specifically, the query is split into words by tokenization, and the role of each word is identified by part-of-speech tagging. As a result of this analysis, the intention of the query is identified, and a structured form of the query data is output.
[0292] Step 3:
[0293] Using the information reference means in the server, information related to the intention identified from the analyzed data is retrieved from the database. In this process, database access is performed through an SQL query or the like to extract relevant information and items. The input is structured intention data, and the output is relevant information data.
[0294] Step 4:
[0295] The generative AI model generates a response based on the extracted information. The model is trained using a deep learning algorithm and leverages knowledge gained from large datasets to generate natural-sounding sentences. The input is relevant information data, and the output is the generated response text.
[0296] Step 5:
[0297] The server formats the responses generated by the generative AI model and sends them to the user's terminal. The formatting process adjusts the text to be human-readable and optimizes the visual arrangement of information. The input is the generated response text, and the output is the formatted response data.
[0298] Step 6:
[0299] The terminal receives formatted responses sent from the server and displays them to the user. The user reviews this information on the terminal and identifies the actions necessary to proceed with administrative procedures. The input is formatted response data, and the output is the display of information to the user.
[0300] (Application Example 1)
[0301] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."
[0302] There is a need to alleviate the complexity of information retrieval and the difficulty in understanding procedural processes that citizens face in administrative procedures. In particular, a lack of understanding of administrative procedures and required documents can be a significant burden for citizens. Furthermore, if administrative services are not sufficiently digitized, the time required to obtain information and complete procedures increases, reducing convenience for citizens. By solving these problems, it is expected that the quality of administrative services will improve and citizens' lives will become smoother.
[0303] The specific processing by the specific processing unit 290 of the data processing apparatus 12 in Application Example 1 is realized by the following means.
[0304] In this invention, the server includes an information terminal that receives inquiries from users, a computing device that performs natural language processing based on the inquiries received from the information terminal, and a generation model that generates answers based on the natural language data processed by the computing device. Thereby, it becomes possible to analyze voice inputs from citizens and accurately and quickly guide procedures and necessary documents.
[0305] The "information terminal" is a device for receiving inquiries from users and enables voice input and text input.
[0306] The "computing device" is a computer system that analyzes inquiries received from an information terminal by natural language processing technology and interprets the intentions of users.
[0307] The "generation model" is an artificial intelligence model that creates appropriate answers based on data analyzed by a computing device.
[0308] The "information acquisition device" is a system that further updates the answers generated by the generation model with reference data and aggregates necessary information.
[0309] The "response generation device" is a device having a function of transmitting the information integrated by the information acquisition device to the information terminal used by the user.
[0310] The "decision-making device" is a system that analyzes voice inputs from citizens and specifically guides procedures and necessary documents.
[0311] The "knowledge extraction device" is a device that optimizes the answers output by the generation model based on the policy information of government agencies at the time.
[0312] A system for implementing this invention consists of the following components: an information terminal, a computing device, a generative model, an information acquisition device, a response generation device, and a decision-making device.
[0313] The information terminals used will be smartphones and smart glasses that allow users to input inquiries via voice or text. These terminals communicate with a server via the internet and transmit user inquiries to a computing device. Voice input will be converted to text using the Google Speech-to-Text API.
[0314] The server, as a computing device, is built using the Flask framework and processes user input using natural language processing. The spaCy library is used for this processing, understanding the intent of queries and identifying the next data to reference. Furthermore, the GPT-3 generative model is incorporated, generating natural and accurate responses based on the analysis results. This generative model learns from large datasets and provides answers based on real-world context.
[0315] The information acquisition device compares the responses generated by the generation model with existing databases and extracts and updates the necessary information. The response generation device formats this information and transmits it to the user's information terminal. Ultimately, the user receives guidance on procedures and required documents through the information terminal, enabling citizens to smoothly proceed with the necessary procedures.
[0316] For example, if a citizen asks via voice, "Please tell me how to register my new address," the system will analyze the inquiry and guide them to the necessary documents and submission locations. The information the user needs will be displayed on the terminal, and a path will be set up to complete the procedure online.
[0317] As an example of a prompt, a user query is sent to the system in the format "User Query: 'Please tell me how to register a new address.'" Based on this, the request to the generative model is expressed as "Explain the steps, required documents, and possible online submission options for new address registration."
[0318] The flow of a specific process in Application Example 1 will be explained using Figure 12.
[0319] Step 1:
[0320] The user uses a smartphone or smart glasses to input their inquiry via voice or text. The input voice is converted into text data using the Google Speech-to-Text API and sent from the smart device to the server.
[0321] Step 2:
[0322] The server is built using Flask and receives input data from users. This data is processed using the spaCy library to analyze the intent of the query. Based on the analyzed intent, the next necessary data is identified. The input is the user's query, and the output is the analysis result including the identified intent.
[0323] Step 3:
[0324] The server sends the analysis results to the generative AI model. The GPT-3 generative model receives the intent in the form of a prompt and generates a relevant response. The prompt is sent in the form of "User Query: 'Please tell me how to register a new address.'" The input to the model is the analysis results, and the output is the generated natural language response.
[0325] Step 4:
[0326] The server receives the response from the generative model, and the information retrieval device accesses the database to collect necessary additional information based on the generated response. Database matching updates and integrates the information. The input is the response from the generative model, and the output is the integrated additional information.
[0327] Step 5:
[0328] The response generation device formats the integrated information and sends the response to the user's information terminal. The user can then easily access information about the procedure and necessary documents on their terminal. The input is integrated information, and the output is a formatted response.
[0329] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.
[0330] This invention relates to a system that receives inquiries from users, analyzes those emotions using an emotion engine, and generates appropriate responses through a generative AI model. The aim of this system is to improve the user experience and the quality of personalized responses.
[0331] The system's main components include terminals for users to input inquiries, servers for data processing and sentiment analysis, and generative AI models and sentiment engines. These elements communicate over the internet and exchange information in real time.
[0332] The user enters text about questions or problems through their terminal. The entered data is sent to the server, which holds the received data in a buffer.
[0333] On the server, the natural language processing module is executed first to analyze the user's query. This analysis extracts the specific information the user intends to convey. Next, the emotion engine uses the same data to determine the user's emotional state (e.g., joy, anger, sadness). The emotion engine uses sophisticated algorithms to estimate emotions based on factors such as tone of voice and text, and keywords.
[0334] The generative AI model considers the results of natural language processing and emotion engine analysis to generate responses tailored to specific emotions. The tone and expression of the response are adjusted according to the emotion. For example, if an inquiry includes dissatisfaction, the generative model will generate a response in milder language that includes countermeasures that take this into account.
[0335] The server scrutinizes the generated response by linking it with the database and updates the information as needed. The response is then formatted and sent to the terminal.
[0336] The device displays responses in a user-optimized format, making it easier for users to understand the next steps. For example, if a user is dissatisfied with an administrative service, the system will return a response that emphasizes explanations of countermeasures, including expressions of gratitude and apology.
[0337] This invention enables the system to respond sensitively to user emotions and provide more personalized services. Furthermore, it is expected that administrative services will be further improved through continuous improvement based on feedback.
[0338] The following describes the processing flow.
[0339] Step 1:
[0340] Users can input questions and comments about administrative procedures and services through their terminals. For example, they can input something like, "My address change procedure is delayed."
[0341] Step 2:
[0342] The terminal processes the input data and sends it to the server using a secure protocol. The data is properly structured according to the transmission format.
[0343] Step 3:
[0344] The server temporarily stores the query received from the terminal in a data buffer. This data is then used in the subsequent analysis process.
[0345] Step 4:
[0346] The server activates a natural language processing module to analyze the query. It extracts keywords and context from the text to identify the information the user intends to receive.
[0347] Step 5:
[0348] The server uses an emotion engine to analyze the sentiment in the query text. For example, it can determine from the text content whether the user is feeling frustrated.
[0349] Step 6:
[0350] The server calls a generative AI model based on the analyzed inquiry content and sentiment data. The AI model generates an answer that reflects both of these factors.
[0351] Step 7:
[0352] The server uses information retrieval tools to obtain relevant details from the database regarding the generated response, and then uses those details to supplement or update the response.
[0353] Step 8:
[0354] The server adjusts the tone and style of the emotionally generated responses, refining the formatting. For example, it might add nuances of apology or change the language to be more polite.
[0355] Step 9:
[0356] The server sends the completed response to the terminal. This process is carried out quickly to ensure the user has the optimal reaction time.
[0357] Step 10:
[0358] The device displays the answer to the user, allowing the user to obtain specific actions or further guidance regarding their inquiry.
[0359] Step 11:
[0360] Users can provide feedback on the answers. This feedback is returned to the system and stored as data for future improvements.
[0361] (Example 2)
[0362] Next, we will describe Example 2. 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".
[0363] Existing customer service systems often provide uniform answers without adequately considering user emotions, resulting in a poor user experience. Furthermore, the quality of the generated answers is not optimized for the user's specific emotional state, leading to insufficient individualized support.
[0364] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.
[0365] In this invention, the server includes means for performing natural language analysis using a computing device, means for inferring the user's emotional state using an emotion analysis device, and means for generating an emotion-based response using a generative model device. This enables the generation of sophisticated responses that take emotional states into consideration.
[0366] A "terminal device" is a device that receives inquiries from users and sends data to a server.
[0367] A "processing unit" is a device equipped with computing resources for analyzing data obtained from terminal devices.
[0368] An "emotion analysis device" is a device that has the function of inferring the emotional state of a user from their inquiry data.
[0369] A "generative model device" is a device that generates appropriate responses based on analyzed data and emotional states.
[0370] An "information management device" is a device that compares the generated responses with a storage medium and updates the information as needed.
[0371] A "response generation device" is a device that has the function of transmitting information acquired by an information management device to a terminal device.
[0372] A "data analysis device" is a device used to apply optimization based on the user's emotional state to policy information.
[0373] This invention is a system comprising a terminal device for the user to input inquiries, a computing device for analyzing data, an emotion analysis device for analyzing the user's emotions, a generative model device for generating answers using a generative model, an information management device for scrutinizing answers and managing information, and a response creation device for finally creating a response and sending it to the user's terminal.
[0374] The terminal device receives text input from the user and transmits it to the server as a digital signal. Typical personal computers and smartphones are used.
[0375] The computing units deployed on the server are equipped with high-performance CPUs and memory, and analyze data received from users. Here, natural language processing technology is used to enable the analysis of user intent.
[0376] The emotion analysis device extracts text tone and keywords from the text data entered by the user to infer the user's emotional state. By applying a highly accurate algorithm, it can accurately determine emotions such as joy, anger, and sadness.
[0377] The generative modeling device generates responses in a tone appropriate to the user's emotional state, based on the analysis results. The generative AI model used in this process employs the latest natural language generation technology.
[0378] The information management device compares the generated responses with the storage medium to verify the information's up-to-dateness. Information is updated as needed, ensuring that accurate data is always maintained.
[0379] The response generation device sends the reviewed response to the user's terminal device and formats it for appropriate display. This allows the user to logically select their next action.
[0380] For example, if a user uses the prompt "Tell me about recent improvements in government services," the system will analyze the user's intent and return a response that provides detailed information while taking emotions into consideration. This is expected to improve the user experience and dramatically enhance the quality of personalized responses.
[0381] The flow of the specific processing in Example 2 will be explained using Figure 13.
[0382] Step 1:
[0383] The user enters their inquiry in text format using a terminal device. The entered text data serves as a prompt to the system. For example, the user might enter, "Please tell me about recent government services." The terminal then converts this text data into a digital signal and sends it to the server.
[0384] Step 2:
[0385] The server stores the text data received from the terminal in a receive buffer for analysis. It decodes the digital signal as input into text, preparing it for analysis. This data is then sent to the processing unit.
[0386] Step 3:
[0387] The server passes the text data received via the computing unit to a natural language processing module for analysis. The input is raw text data from the user, and the output is keywords related to the analyzed intent and topic. This analysis clarifies the user's specific information needs.
[0388] Step 4:
[0389] The server passes the analyzed results to an emotion analysis device, which evaluates the emotional state based on the text data. The analyzed keywords are used as input, and the user's emotional state (e.g., joy, anger, sadness) is obtained as output. Emotion analysis is performed based on the tone and keywords of the text.
[0390] Step 5:
[0391] The server activates its generative AI model function and generates responses using the results of natural language processing and sentiment analysis. The input here consists of two data points: the user's intent and emotion, and the output is a natural-sounding response adjusted to match the emotion. As a result, responses that are both reliable and approachable are obtained.
[0392] Step 6:
[0393] The server sends the generated responses to the information management device, where they are compared with the database to verify the timeliness and accuracy of the information. Here, the generated responses are checked against the internal data repository to confirm that the information is up-to-date, and then the results are converted into a format that can be sent to the terminal.
[0394] Step 7:
[0395] The server sends the generated response to the terminal device via the response generation device. Finally, the terminal displays this response to the user. The output displays an optimized answer to the user's question and provides information to prompt the user's next action.
[0396] Through these steps, the system provides appropriate information based on the user's emotions in real time, improving the user experience.
[0397] (Application Example 2)
[0398] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."
[0399] Customer support on modern e-commerce sites presents a challenging task, requiring a thorough understanding and appropriate response to user emotions. Uniform, emotion-insensitive responses can degrade the user experience. Technologies are needed to improve this.
[0400] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.
[0401] In this invention, the server includes an input device means for receiving inquiries from users, a computing device means for performing natural language processing based on the inquiries received from the input device means, and an emotion analysis means for performing emotion analysis based on the natural language data processed by the computing device means. This makes it possible to generate responses that take the user's emotions into consideration.
[0402] "An input device for receiving inquiries from users" refers to a device that has the function of allowing users to input inquiries to the system.
[0403] A "computation device means" is a device that has computational functions for performing natural language processing based on data received from an input device means.
[0404] An "emotion analysis device" is a device that has the function of analyzing a user's emotional state from natural language data processed by a computing device.
[0405] A "generative model device means" is a device that has the function of automatically generating appropriate answers to user inquiries based on the results of sentiment analysis means.
[0406] "Information acquisition means" refers to a device that has the function of referencing the response generated by the generation model device means with information from a database and updating it as necessary.
[0407] A "response generation means" is a device that has the function of sending a response to the user in an appropriate format based on the information obtained by the information acquisition means.
[0408] This invention provides a system for improving the response accuracy of customer support on e-commerce websites. The system consists of various components and is implemented by an input device used by the user and multiple modules that operate on a server.
[0409] The user enters their inquiry in text format using an input device. This input device is a common computer terminal such as a smartphone or personal computer. The data entered by the user is transmitted to the server via the internet.
[0410] The server first uses a natural language processing engine, positioned as a computing device, to analyze the user's query. Existing libraries such as "SpaCy" and "Transformers" can be used for natural language processing. This process extracts the specific content intended by the user.
[0411] Next, the emotion analysis system analyzes the user's emotional state based on the text data. This emotion analysis uses algorithms to analyze the tone of the text and specific language patterns. As a result, the user's emotions are recognized as "joy," "anger," etc.
[0412] The generative model device takes the results of natural language processing and sentiment analysis as input and generates the optimal response in plain text form. The generated response is compared with a database by the information acquisition means, and the information is updated as needed. This database referencing ensures that the response to the user is based on the latest information at the present time.
[0413] Finally, the response generation mechanism sends the generated answer to the user's input device in an appropriate format, allowing the user to receive a response to their inquiry. This is displayed on the terminal screen, making it easy for the user to understand what action to take next.
[0414] For example, if a user inquires that they are dissatisfied because their delivery is delayed, the system will use sentiment analysis to read the dissatisfaction, and a generative model will generate a response that includes an apology and a solution that takes the user's feelings into consideration. The response might be: "We apologize for the delay. We are currently checking the delivery status and will contact you shortly."
[0415] An example of a prompt message input to the generating AI model is: "User inquiry: 'My delivery is delayed. Please take action.' Sentiment analysis result: 'Angry' Please provide an appropriate response."
[0416] The flow of a specific process in Application Example 2 will be explained using Figure 14.
[0417] Step 1:
[0418] The user uses an input device to enter their inquiry in text format and presses the submit button. The entered text data is transmitted to the server via the internet. At this point, the input is raw data from the user.
[0419] Step 2:
[0420] The server's computing system initiates natural language processing using the received text data. Using natural language processing libraries (e.g., SpaCy or Transformers), it extracts the specific intent of the query and related information from the text. The input is the user's raw text data, and the output is the parsed structured data.
[0421] Step 3:
[0422] The server uses sentiment analysis tools to estimate the user's emotional state based on structured data. The sentiment analysis algorithm analyzes the tone and keywords of the text and classifies emotions as "joy" or "anger." In this step, the input is output data from natural language processing, and the emotional state is output.
[0423] Step 4:
[0424] The generative model device generates the optimal response based on the sentiment analysis results and structured data. Using the generative AI model, text with appropriate wording corresponding to the inquiry content and the user's emotions is created. The input here is emotions and structured data, and the output is the response text.
[0425] Step 5:
[0426] The server compares the generated response with the database through the information retrieval mechanism and updates the information as needed. Database lookups are performed to ensure that the response is based on the latest information. The input for this step is the generated response, and the output is the verified response.
[0427] Step 6:
[0428] The response generation mechanism sends the confirmed response text to the user's input device. The display on the terminal is in a format that is easy for the user to understand. The final input is the confirmed response, and the output is the final response displayed on the user's screen.
[0429] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.
[0430] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.
[0431] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.
[0432] [Third Embodiment]
[0433] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.
[0434] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.
[0435] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).
[0436] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.
[0437] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.
[0438] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).
[0439] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.
[0440] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.
[0441] The specific processing program 56 is an example of a "program" relating to the technology of this 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.
[0442] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.
[0443] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.
[0444] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".
[0445] This invention provides a system that enables efficient and rapid information provision when users make inquiries related to government services. The main components of the system include a terminal used by the user, a server for data processing, and a generative AI model.
[0446] Users can use a terminal to input questions regarding administrative procedures. This terminal has the functionality to send data to a server via an internet connection and receive responses from the server.
[0447] When the server receives query data sent by a user, it analyzes the data using natural language processing techniques. This analysis process identifies the intent of the query and retrieves relevant information from the database.
[0448] Generative AI models are responsible for generating appropriate answers based on the analyzed information. These AI models are trained on large historical datasets, enabling them to create natural and accurate responses to user questions.
[0449] For example, if a user asks, "I want to change my address, how do I do that?", the server analyzes this inquiry and collects information about the relevant procedures and required documents from its database. The generative AI model then uses this information to construct an answer that includes specific steps.
[0450] The server then formats the generated response and sends it to the user's device. The user can then review this information displayed on their device screen and complete the necessary procedures online.
[0451] This invention reduces the complexity of administrative procedures, allowing users to efficiently obtain accurate information. Furthermore, user feedback is collected via the server and used to improve the system in the future. This enables local governments to provide higher quality services.
[0452] The following describes the processing flow.
[0453] Step 1:
[0454] Users input questions and inquiries about procedures using a terminal. In this process, they enter the specific question in text format into the interface.
[0455] Step 2:
[0456] The terminal receives data entered by the user and sends query data to the server via the internet. A protocol is used to ensure security during data transmission.
[0457] Step 3:
[0458] The server receives query data sent from the terminal. The received data is temporarily stored in a data buffer.
[0459] Step 4:
[0460] The server uses a natural language processing (NLP) module to analyze incoming queries. Specifically, it extracts keywords and context from text data to understand the user's intent.
[0461] Step 5:
[0462] The server launches a generative AI model based on the analyzed information. The generative AI model uses a pre-trained algorithm to construct the optimal answer based on the analysis results.
[0463] Step 6:
[0464] The server reviews the generated response and queries the database as needed. It retrieves relevant information and supplements or updates the response.
[0465] Step 7:
[0466] The server formats the final response into a user-friendly format. At this stage, the response is arranged in a natural language structure and is suitable for the user interface.
[0467] Step 8:
[0468] The server sends a formatted response to the terminal. This response is optimized to be real-time or near real-time.
[0469] Step 9:
[0470] The terminal displays the response received from the server in the user interface. Based on the displayed information, the user can determine the next steps to take or ask further questions.
[0471] Step 10:
[0472] Users can send feedback via their devices as needed, suggesting system improvements or making additional inquiries. This feedback will be used to inform subsequent data analysis.
[0473] (Example 1)
[0474] Next, we will describe Example 1. 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."
[0475] In inquiries related to government services, users face the challenge of difficulty in obtaining quick and accurate information. Traditional systems can be inefficient in their process of appropriately retrieving and clearly presenting information based on the content of the inquiry. Furthermore, there are insufficient means to efficiently collect user feedback and utilize it for system improvement. Additionally, there is a need to optimize responses by taking into account local government policy information.
[0476] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.
[0477] In this invention, the server includes communication means for receiving inquiries from users, a processing unit for analyzing inquiries received from the communication means and performing natural language processing, information retrieval means for searching a database and extracting information based on the intent identified by the processing unit, response model means for generating answers using a generation AI based on the extracted information, and transmission means for formatting the generated answers to aid user understanding and transmitting them to a terminal. This enables users to obtain information on administrative procedures quickly and accurately. Furthermore, user feedback can be accepted and used to optimize the system, enabling the generation of appropriate answers based on policy information from public institutions.
[0478] "Communication means" refers to devices or interfaces used to receive inquiries from users and send data to a server.
[0479] A "processing unit" is a computing mechanism that analyzes received queries and performs natural language processing.
[0480] An "information retrieval tool" is a function that searches a database and extracts relevant information based on the user's intent identified through analysis.
[0481] A "response model means" is a system that uses a generation AI based on extracted information to generate appropriate responses for the user.
[0482] A "transmission method" is a mechanism for formatting the generated response and accurately transmitting it to the user's device.
[0483] A "feedback management system" is a mechanism for receiving feedback from users and using it to improve the system.
[0484] An "optimization method" is a technology that appropriately adjusts and optimizes the generated responses based on policy information from public institutions.
[0485] This invention provides a system that enables users to receive timely and accurate information when making inquiries related to government services. The system includes a terminal used by the user, a server for data processing, and a generative AI model.
[0486] Users can input questions about administrative procedures through the terminal's interface. This terminal is responsible for transmitting the inquiry to the server via an internet connection.
[0487] The server has a processing unit for analyzing received queries using natural language processing techniques to identify the user's intent. Natural language processing techniques used for analysis include, for example, tokenization, syntactic analysis, and intent classification algorithms.
[0488] After the processing unit identifies the intent, it uses information retrieval means to search the database based on that information and extract relevant information. This information is often obtained from databases of documents and procedures related to administrative processes.
[0489] Based on the extracted information, a generative AI model generates appropriate answers to questions as a response model. This AI model is trained using deep learning and large-scale language models, and creates natural and accurate responses based on a vast amount of historical data. For example, in response to a prompt such as "Please tell me what documents and procedures are required to renew my driver's license," it can provide detailed information about the necessary documents, locations, and procedures.
[0490] The server then formats the generated response using a communication method and accurately sends it to the user's device. The user can then review the information displayed on their device and complete the necessary procedures online.
[0491] This system reduces the complexity of administrative procedures, allowing users to efficiently obtain accurate information. Furthermore, user feedback is collected as a feedback management tool, contributing to future improvements of the system. This enables public institutions to provide higher quality services.
[0492] The flow of the specific processing in Example 1 will be explained using Figure 11.
[0493] Step 1:
[0494] The user enters questions about administrative procedures through the terminal's interface. The terminal converts the entered questions into digital data and sends them to a server via the internet. The input is a user query in text format, and the output is the query sent to the server.
[0495] Step 2:
[0496] The server receives user query data from the terminal into its processing unit. Natural language processing (NLP) techniques are used to analyze this query data. Specifically, the query is divided into words using tokenization, and the role of each word is identified through part-of-speech tagging. As a result of this analysis, the intent of the query is identified, and a structured format of the query data is output.
[0497] Step 3:
[0498] Using information retrieval tools within the server, information related to the intent identified from the analyzed data is retrieved from the database. This process involves database access via SQL queries and other methods to extract relevant information and items. The input is structured intent data, and the output is related information data.
[0499] Step 4:
[0500] The generative AI model generates a response based on the extracted information. The model is trained using a deep learning algorithm and leverages knowledge gained from large datasets to generate natural-sounding sentences. The input is relevant information data, and the output is the generated response text.
[0501] Step 5:
[0502] The server formats the responses generated by the generative AI model and sends them to the user's terminal. The formatting process adjusts the text to be human-readable and optimizes the visual arrangement of information. The input is the generated response text, and the output is the formatted response data.
[0503] Step 6:
[0504] The terminal receives formatted responses sent from the server and displays them to the user. The user reviews this information on the terminal and identifies the actions necessary to proceed with administrative procedures. The input is formatted response data, and the output is the display of information to the user.
[0505] (Application Example 1)
[0506] Next, we will explain Application Example 1. In the following explanation, 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."
[0507] There is a need to alleviate the complexity of information retrieval and the difficulty in understanding procedural processes that citizens face in administrative procedures. In particular, a lack of understanding of administrative procedures and required documents can be a significant burden for citizens. Furthermore, if administrative services are not sufficiently digitized, the time required to obtain information and complete procedures increases, reducing convenience for citizens. By solving these problems, it is expected that the quality of administrative services will improve and citizens' lives will become smoother.
[0508] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.
[0509] In this invention, the server includes an information terminal for receiving inquiries from users, a computing device for performing natural language processing based on the inquiries received from the information terminal, and a generative model for generating responses based on the natural language data processed by the computing device. This makes it possible to analyze voice input from citizens and provide accurate and rapid guidance on procedures and necessary materials.
[0510] An "information terminal" is a device used to receive inquiries from users and enables voice input and text input.
[0511] A "computational device" is a computer system that analyzes inquiries received from information terminals using natural language processing technology and interprets the user's intent.
[0512] A "generative model" is an artificial intelligence model that generates appropriate answers based on data analyzed by a computing device.
[0513] An "information acquisition device" is a system that updates the responses generated by a generative model with reference data and aggregates the necessary information.
[0514] A "response generation device" is a device that has the function of transmitting information integrated by an information acquisition device to an information terminal used by the user.
[0515] A "decision-making device" is a system that analyzes voice input from citizens and provides specific guidance on procedures and necessary documents.
[0516] A "knowledge extraction device" is a device that optimizes the answers output by a generative model based on the policy information of government agencies at that time.
[0517] A system for implementing this invention consists of the following components: an information terminal, a computing device, a generative model, an information acquisition device, a response generation device, and a decision-making device.
[0518] The information terminals used will be smartphones and smart glasses that allow users to input inquiries via voice or text. These terminals communicate with a server via the internet and transmit user inquiries to a computing device. Voice input will be converted to text using the Google Speech-to-Text API.
[0519] The server, as a computing device, is built using the Flask framework and processes user input using natural language processing. The spaCy library is used for this processing, understanding the intent of queries and identifying the next data to reference. Furthermore, the GPT-3 generative model is incorporated, generating natural and accurate responses based on the analysis results. This generative model learns from large datasets and provides answers based on real-world context.
[0520] The information acquisition device compares the responses generated by the generation model with existing databases and extracts and updates the necessary information. The response generation device formats this information and transmits it to the user's information terminal. Ultimately, the user receives guidance on procedures and required documents through the information terminal, enabling citizens to smoothly proceed with the necessary procedures.
[0521] For example, if a citizen asks via voice, "Please tell me how to register my new address," the system will analyze the inquiry and guide them to the necessary documents and submission locations. The information the user needs will be displayed on the terminal, and a path will be set up to complete the procedure online.
[0522] As an example of a prompt, a user query is sent to the system in the format "User Query: 'Please tell me how to register a new address.'" Based on this, the request to the generative model is expressed as "Explain the steps, required documents, and possible online submission options for new address registration."
[0523] The flow of a specific process in Application Example 1 will be explained using Figure 12.
[0524] Step 1:
[0525] The user uses a smartphone or smart glasses to input their inquiry via voice or text. The input voice is converted into text data using the Google Speech-to-Text API and sent from the smart device to the server.
[0526] Step 2:
[0527] The server is built using Flask and receives input data from users. This data is processed using the spaCy library to analyze the intent of the query. Based on the analyzed intent, the next necessary data is identified. The input is the user's query, and the output is the analysis result including the identified intent.
[0528] Step 3:
[0529] The server sends the analysis results to the generative AI model. The GPT-3 generative model receives the intent in the form of a prompt and generates a relevant response. The prompt is sent in the form of "User Query: 'Please tell me how to register a new address.'" The input to the model is the analysis results, and the output is the generated natural language response.
[0530] Step 4:
[0531] The server receives the response from the generative model, and the information retrieval device accesses the database to collect necessary additional information based on the generated response. Database matching updates and integrates the information. The input is the response from the generative model, and the output is the integrated additional information.
[0532] Step 5:
[0533] The response generation device formats the integrated information and sends the response to the user's information terminal. The user can then easily access information about the procedure and necessary documents on their terminal. The input is integrated information, and the output is a formatted response.
[0534] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.
[0535] This invention relates to a system that receives inquiries from users, analyzes those emotions using an emotion engine, and generates appropriate responses through a generative AI model. The aim of this system is to improve the user experience and the quality of personalized responses.
[0536] The system's main components include terminals for users to input inquiries, servers for data processing and sentiment analysis, and generative AI models and sentiment engines. These elements communicate over the internet and exchange information in real time.
[0537] The user enters text about questions or problems through their terminal. The entered data is sent to the server, which holds the received data in a buffer.
[0538] On the server, the natural language processing module is executed first to analyze the user's query. This analysis extracts the specific information the user intends to convey. Next, the emotion engine uses the same data to determine the user's emotional state (e.g., joy, anger, sadness). The emotion engine uses sophisticated algorithms to estimate emotions based on factors such as tone of voice and text, and keywords.
[0539] The generative AI model considers the results of natural language processing and emotion engine analysis to generate responses tailored to specific emotions. The tone and expression of the response are adjusted according to the emotion. For example, if an inquiry includes dissatisfaction, the generative model will generate a response in milder language that includes countermeasures that take this into account.
[0540] The server scrutinizes the generated response by linking it with the database and updates the information as needed. The response is then formatted and sent to the terminal.
[0541] The device displays responses in a user-optimized format, making it easier for users to understand the next steps. For example, if a user is dissatisfied with an administrative service, the system will return a response that emphasizes explanations of countermeasures, including expressions of gratitude and apology.
[0542] This invention enables the system to respond sensitively to user emotions and provide more personalized services. Furthermore, it is expected that administrative services will be further improved through continuous improvement based on feedback.
[0543] The following describes the processing flow.
[0544] Step 1:
[0545] Users can input questions and comments about administrative procedures and services through their terminals. For example, they can input something like, "My address change procedure is delayed."
[0546] Step 2:
[0547] The terminal processes the input data and sends it to the server using a secure protocol. The data is properly structured according to the transmission format.
[0548] Step 3:
[0549] The server temporarily stores the query received from the terminal in a data buffer. This data is then used in the subsequent analysis process.
[0550] Step 4:
[0551] The server activates a natural language processing module to analyze the query. It extracts keywords and context from the text to identify the information the user intends to receive.
[0552] Step 5:
[0553] The server uses an emotion engine to analyze the sentiment in the query text. For example, it can determine from the text content whether the user is feeling frustrated.
[0554] Step 6:
[0555] The server calls a generative AI model based on the analyzed inquiry content and sentiment data. The AI model generates an answer that reflects both of these factors.
[0556] Step 7:
[0557] The server uses information retrieval tools to obtain relevant details from the database regarding the generated response, and then uses those details to supplement or update the response.
[0558] Step 8:
[0559] The server adjusts the tone and style of the emotionally generated responses, refining the formatting. For example, it might add nuances of apology or change the language to be more polite.
[0560] Step 9:
[0561] The server sends the completed response to the terminal. This process is carried out quickly to ensure the user has the optimal reaction time.
[0562] Step 10:
[0563] The device displays the answer to the user, allowing the user to obtain specific actions or further guidance regarding their inquiry.
[0564] Step 11:
[0565] Users can provide feedback on the answers. This feedback is returned to the system and stored as data for future improvements.
[0566] (Example 2)
[0567] Next, we will describe Example 2. 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."
[0568] Existing customer service systems often provide uniform answers without adequately considering user emotions, resulting in a poor user experience. Furthermore, the quality of the generated answers is not optimized for the user's specific emotional state, leading to insufficient individualized support.
[0569] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.
[0570] In this invention, the server includes means for performing natural language analysis using a computing device, means for inferring the user's emotional state using an emotion analysis device, and means for generating an emotion-based response using a generative model device. This enables the generation of sophisticated responses that take emotional states into consideration.
[0571] A "terminal device" is a device that receives inquiries from users and sends data to a server.
[0572] A "processing unit" is a device equipped with computing resources for analyzing data obtained from terminal devices.
[0573] An "emotion analysis device" is a device that has the function of inferring the emotional state of a user from their inquiry data.
[0574] A "generative model device" is a device that generates appropriate responses based on analyzed data and emotional states.
[0575] An "information management device" is a device that compares the generated responses with a storage medium and updates the information as needed.
[0576] A "response generation device" is a device that has the function of transmitting information acquired by an information management device to a terminal device.
[0577] A "data analysis device" is a device used to apply optimization based on the user's emotional state to policy information.
[0578] This invention is a system comprising a terminal device for the user to input inquiries, a computing device for analyzing data, an emotion analysis device for analyzing the user's emotions, a generative model device for generating answers using a generative model, an information management device for scrutinizing answers and managing information, and a response creation device for finally creating a response and sending it to the user's terminal.
[0579] The terminal device receives text input from the user and transmits it to the server as a digital signal. Typical personal computers and smartphones are used.
[0580] The computing units deployed on the server are equipped with high-performance CPUs and memory, and analyze data received from users. Here, natural language processing technology is used to enable the analysis of user intent.
[0581] The emotion analysis device extracts text tone and keywords from the text data entered by the user to infer the user's emotional state. By applying a highly accurate algorithm, it can accurately determine emotions such as joy, anger, and sadness.
[0582] The generative modeling device generates responses in a tone appropriate to the user's emotional state, based on the analysis results. The generative AI model used in this process employs the latest natural language generation technology.
[0583] The information management device compares the generated responses with the storage medium to verify the information's up-to-dateness. Information is updated as needed, ensuring that accurate data is always maintained.
[0584] The response generation device sends the reviewed response to the user's terminal device and formats it for appropriate display. This allows the user to logically select their next action.
[0585] For example, if a user uses the prompt "Tell me about recent improvements in government services," the system will analyze the user's intent and return a response that provides detailed information while taking emotions into consideration. This is expected to improve the user experience and dramatically enhance the quality of personalized responses.
[0586] The flow of the specific processing in Example 2 will be explained using Figure 13.
[0587] Step 1:
[0588] The user enters their inquiry in text format using a terminal device. The entered text data serves as a prompt to the system. For example, the user might enter, "Please tell me about recent government services." The terminal then converts this text data into a digital signal and sends it to the server.
[0589] Step 2:
[0590] The server stores the text data received from the terminal in a receive buffer for analysis. It decodes the digital signal as input into text, preparing it for analysis. This data is then sent to the processing unit.
[0591] Step 3:
[0592] The server passes the text data received via the computing unit to a natural language processing module for analysis. The input is raw text data from the user, and the output is keywords related to the analyzed intent and topic. This analysis clarifies the user's specific information needs.
[0593] Step 4:
[0594] The server passes the analyzed results to an emotion analysis device, which evaluates the emotional state based on the text data. The analyzed keywords are used as input, and the user's emotional state (e.g., joy, anger, sadness) is obtained as output. Emotion analysis is performed based on the tone and keywords of the text.
[0595] Step 5:
[0596] The server activates its generative AI model function and generates responses using the results of natural language processing and sentiment analysis. The input here consists of two data points: the user's intent and emotion, and the output is a natural-sounding response adjusted to match the emotion. As a result, responses that are both reliable and approachable are obtained.
[0597] Step 6:
[0598] The server sends the generated responses to the information management device, where they are compared with the database to verify the timeliness and accuracy of the information. Here, the generated responses are checked against the internal data repository to confirm that the information is up-to-date, and then the results are converted into a format that can be sent to the terminal.
[0599] Step 7:
[0600] The server sends the generated response to the terminal device via the response generation device. Finally, the terminal displays this response to the user. The output displays an optimized answer to the user's question and provides information to prompt the user's next action.
[0601] Through these steps, the system provides appropriate information based on the user's emotions in real time, improving the user experience.
[0602] (Application Example 2)
[0603] Next, we will explain application example 2. In the following explanation, 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."
[0604] Customer support on modern e-commerce sites presents a challenging task, requiring a thorough understanding and appropriate response to user emotions. Uniform, emotion-insensitive responses can degrade the user experience. Technologies are needed to improve this.
[0605] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.
[0606] In this invention, the server includes an input device means for receiving inquiries from users, a computing device means for performing natural language processing based on the inquiries received from the input device means, and an emotion analysis means for performing emotion analysis based on the natural language data processed by the computing device means. This makes it possible to generate responses that take the user's emotions into consideration.
[0607] "An input device for receiving inquiries from users" refers to a device that has the function of allowing users to input inquiries to the system.
[0608] A "computation device means" is a device that has computational functions for performing natural language processing based on data received from an input device means.
[0609] An "emotion analysis device" is a device that has the function of analyzing a user's emotional state from natural language data processed by a computing device.
[0610] A "generative model device means" is a device that has the function of automatically generating appropriate answers to user inquiries based on the results of sentiment analysis means.
[0611] "Information acquisition means" refers to a device that has the function of referencing the response generated by the generation model device means with information from a database and updating it as necessary.
[0612] A "response generation means" is a device that has the function of sending a response to the user in an appropriate format based on the information obtained by the information acquisition means.
[0613] This invention provides a system for improving the response accuracy of customer support on e-commerce websites. The system consists of various components and is implemented by an input device used by the user and multiple modules that operate on a server.
[0614] The user enters their inquiry in text format using an input device. This input device is a common computer terminal such as a smartphone or personal computer. The data entered by the user is transmitted to the server via the internet.
[0615] The server first uses a natural language processing engine, positioned as a computing device, to analyze the user's query. Existing libraries such as "SpaCy" and "Transformers" can be used for natural language processing. This process extracts the specific content intended by the user.
[0616] Next, the emotion analysis system analyzes the user's emotional state based on the text data. This emotion analysis uses algorithms to analyze the tone of the text and specific language patterns. As a result, the user's emotions are recognized as "joy," "anger," etc.
[0617] The generative model device takes the results of natural language processing and sentiment analysis as input and generates the optimal response in plain text form. The generated response is compared with a database by the information acquisition means, and the information is updated as needed. This database referencing ensures that the response to the user is based on the latest information at the present time.
[0618] Finally, the response generation mechanism sends the generated answer to the user's input device in an appropriate format, allowing the user to receive a response to their inquiry. This is displayed on the terminal screen, making it easy for the user to understand what action to take next.
[0619] For example, if a user inquires that they are dissatisfied because their delivery is delayed, the system will use sentiment analysis to read the dissatisfaction, and a generative model will generate a response that includes an apology and a solution that takes the user's feelings into consideration. The response might be: "We apologize for the delay. We are currently checking the delivery status and will contact you shortly."
[0620] An example of a prompt message input to the generating AI model is: "User inquiry: 'My delivery is delayed. Please take action.' Sentiment analysis result: 'Angry' Please provide an appropriate response."
[0621] The flow of a specific process in Application Example 2 will be explained using Figure 14.
[0622] Step 1:
[0623] The user uses an input device to enter their inquiry in text format and presses the submit button. The entered text data is transmitted to the server via the internet. At this point, the input is raw data from the user.
[0624] Step 2:
[0625] The server's computing system initiates natural language processing using the received text data. Using natural language processing libraries (e.g., SpaCy or Transformers), it extracts the specific intent of the query and related information from the text. The input is the user's raw text data, and the output is the parsed structured data.
[0626] Step 3:
[0627] The server uses sentiment analysis tools to estimate the user's emotional state based on structured data. The sentiment analysis algorithm analyzes the tone and keywords of the text and classifies emotions as "joy" or "anger." In this step, the input is output data from natural language processing, and the emotional state is output.
[0628] Step 4:
[0629] The generative model device generates the optimal response based on the sentiment analysis results and structured data. Using the generative AI model, text with appropriate wording corresponding to the inquiry content and the user's emotions is created. The input here is emotions and structured data, and the output is the response text.
[0630] Step 5:
[0631] The server compares the generated response with the database through the information retrieval mechanism and updates the information as needed. Database lookups are performed to ensure that the response is based on the latest information. The input for this step is the generated response, and the output is the verified response.
[0632] Step 6:
[0633] The response generation mechanism sends the confirmed response text to the user's input device. The display on the terminal is in a format that is easy for the user to understand. The final input is the confirmed response, and the output is the final response displayed on the user's screen.
[0634] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.
[0635] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.
[0636] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.
[0637] [Fourth Embodiment]
[0638] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.
[0639] As shown in Figure 7, the 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.
[0640] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).
[0641] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.
[0642] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.
[0643] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).
[0644] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.
[0645] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive 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 robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.
[0646] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.
[0647] The specific processing program 56 is an example of a "program" relating to the technology of this 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.
[0648] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.
[0649] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.
[0650] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".
[0651] This invention provides a system that enables efficient and rapid information provision when users make inquiries related to government services. The main components of the system include a terminal used by the user, a server for data processing, and a generative AI model.
[0652] Users can use a terminal to input questions regarding administrative procedures. This terminal has the functionality to send data to a server via an internet connection and receive responses from the server.
[0653] When the server receives query data sent by a user, it analyzes the data using natural language processing techniques. This analysis process identifies the intent of the query and retrieves relevant information from the database.
[0654] Generative AI models are responsible for generating appropriate answers based on the analyzed information. These AI models are trained on large historical datasets, enabling them to create natural and accurate responses to user questions.
[0655] For example, if a user asks, "I want to change my address, how do I do that?", the server analyzes this inquiry and collects information about the relevant procedures and required documents from its database. The generative AI model then uses this information to construct an answer that includes specific steps.
[0656] The server then formats the generated response and sends it to the user's device. The user can then review this information displayed on their device screen and complete the necessary procedures online.
[0657] This invention reduces the complexity of administrative procedures, allowing users to efficiently obtain accurate information. Furthermore, user feedback is collected via the server and used to improve the system in the future. This enables local governments to provide higher quality services.
[0658] The following describes the processing flow.
[0659] Step 1:
[0660] Users input questions and inquiries about procedures using a terminal. In this process, they enter the specific question in text format into the interface.
[0661] Step 2:
[0662] The terminal receives data entered by the user and sends query data to the server via the internet. A protocol is used to ensure security during data transmission.
[0663] Step 3:
[0664] The server receives query data sent from the terminal. The received data is temporarily stored in a data buffer.
[0665] Step 4:
[0666] The server uses a natural language processing (NLP) module to analyze incoming queries. Specifically, it extracts keywords and context from text data to understand the user's intent.
[0667] Step 5:
[0668] The server launches a generative AI model based on the analyzed information. The generative AI model uses a pre-trained algorithm to construct the optimal answer based on the analysis results.
[0669] Step 6:
[0670] The server reviews the generated response and queries the database as needed. It retrieves relevant information and supplements or updates the response.
[0671] Step 7:
[0672] The server formats the final response into a user-friendly format. At this stage, the response is arranged in a natural language structure and is suitable for the user interface.
[0673] Step 8:
[0674] The server sends a formatted response to the terminal. This response is optimized to be real-time or near real-time.
[0675] Step 9:
[0676] The terminal displays the response received from the server in the user interface. Based on the displayed information, the user can determine the next steps to take or ask further questions.
[0677] Step 10:
[0678] Users can send feedback via their devices as needed, suggesting system improvements or making additional inquiries. This feedback will be used to inform subsequent data analysis.
[0679] (Example 1)
[0680] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".
[0681] In inquiries related to government services, users face the challenge of difficulty in obtaining quick and accurate information. Traditional systems can be inefficient in their process of appropriately retrieving and clearly presenting information based on the content of the inquiry. Furthermore, there are insufficient means to efficiently collect user feedback and utilize it for system improvement. Additionally, there is a need to optimize responses by taking into account local government policy information.
[0682] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.
[0683] In this invention, the server includes communication means for receiving inquiries from users, a processing unit for analyzing inquiries received from the communication means and performing natural language processing, information retrieval means for searching a database and extracting information based on the intent identified by the processing unit, response model means for generating answers using a generation AI based on the extracted information, and transmission means for formatting the generated answers to aid user understanding and transmitting them to a terminal. This enables users to obtain information on administrative procedures quickly and accurately. Furthermore, user feedback can be accepted and used to optimize the system, enabling the generation of appropriate answers based on policy information from public institutions.
[0684] "Communication means" refers to devices or interfaces used to receive inquiries from users and send data to a server.
[0685] A "processing unit" is a computing mechanism that analyzes received queries and performs natural language processing.
[0686] An "information retrieval tool" is a function that searches a database and extracts relevant information based on the user's intent identified through analysis.
[0687] A "response model means" is a system that uses a generation AI based on extracted information to generate appropriate responses for the user.
[0688] A "transmission method" is a mechanism for formatting the generated response and accurately transmitting it to the user's device.
[0689] A "feedback management system" is a mechanism for receiving feedback from users and using it to improve the system.
[0690] An "optimization method" is a technology that appropriately adjusts and optimizes the generated responses based on policy information from public institutions.
[0691] This invention provides a system that enables users to receive timely and accurate information when making inquiries related to government services. The system includes a terminal used by the user, a server for data processing, and a generative AI model.
[0692] Users can input questions about administrative procedures through the terminal's interface. This terminal is responsible for transmitting the inquiry to the server via an internet connection.
[0693] The server has a processing unit for analyzing received queries using natural language processing techniques to identify the user's intent. Natural language processing techniques used for analysis include, for example, tokenization, syntactic analysis, and intent classification algorithms.
[0694] After the processing unit identifies the intent, it uses information retrieval means to search the database based on that information and extract relevant information. This information is often obtained from databases of documents and procedures related to administrative processes.
[0695] Based on the extracted information, a generative AI model generates appropriate answers to questions as a response model. This AI model is trained using deep learning and large-scale language models, and creates natural and accurate responses based on a vast amount of historical data. For example, in response to a prompt such as "Please tell me what documents and procedures are required to renew my driver's license," it can provide detailed information about the necessary documents, locations, and procedures.
[0696] The server then formats the generated response using a communication method and accurately sends it to the user's device. The user can then review the information displayed on their device and complete the necessary procedures online.
[0697] This system reduces the complexity of administrative procedures, allowing users to efficiently obtain accurate information. Furthermore, user feedback is collected as a feedback management tool, contributing to future improvements of the system. This enables public institutions to provide higher quality services.
[0698] The flow of the specific processing in Example 1 will be explained using Figure 11.
[0699] Step 1:
[0700] The user enters questions about administrative procedures through the terminal's interface. The terminal converts the entered questions into digital data and sends them to a server via the internet. The input is a user query in text format, and the output is the query sent to the server.
[0701] Step 2:
[0702] The server receives user query data from the terminal into its processing unit. Natural language processing (NLP) techniques are used to analyze this query data. Specifically, the query is divided into words using tokenization, and the role of each word is identified through part-of-speech tagging. As a result of this analysis, the intent of the query is identified, and a structured format of the query data is output.
[0703] Step 3:
[0704] Using information retrieval tools within the server, information related to the intent identified from the analyzed data is retrieved from the database. This process involves database access via SQL queries and other methods to extract relevant information and items. The input is structured intent data, and the output is related information data.
[0705] Step 4:
[0706] The generative AI model generates a response based on the extracted information. The model is trained using a deep learning algorithm and leverages knowledge gained from large datasets to generate natural-sounding sentences. The input is relevant information data, and the output is the generated response text.
[0707] Step 5:
[0708] The server formats the responses generated by the generative AI model and sends them to the user's terminal. The formatting process adjusts the text to be human-readable and optimizes the visual arrangement of information. The input is the generated response text, and the output is the formatted response data.
[0709] Step 6:
[0710] The terminal receives formatted responses sent from the server and displays them to the user. The user reviews this information on the terminal and identifies the actions necessary to proceed with administrative procedures. The input is formatted response data, and the output is the display of information to the user.
[0711] (Application Example 1)
[0712] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".
[0713] There is a need to alleviate the complexity of information retrieval and the difficulty in understanding procedural processes that citizens face in administrative procedures. In particular, a lack of understanding of administrative procedures and required documents can be a significant burden for citizens. Furthermore, if administrative services are not sufficiently digitized, the time required to obtain information and complete procedures increases, reducing convenience for citizens. By solving these problems, it is expected that the quality of administrative services will improve and citizens' lives will become smoother.
[0714] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.
[0715] In this invention, the server includes an information terminal for receiving inquiries from users, a computing device for performing natural language processing based on the inquiries received from the information terminal, and a generative model for generating responses based on the natural language data processed by the computing device. This makes it possible to analyze voice input from citizens and provide accurate and rapid guidance on procedures and necessary materials.
[0716] An "information terminal" is a device used to receive inquiries from users and enables voice input and text input.
[0717] A "computational device" is a computer system that analyzes inquiries received from information terminals using natural language processing technology and interprets the user's intent.
[0718] A "generative model" is an artificial intelligence model that generates appropriate answers based on data analyzed by a computing device.
[0719] An "information acquisition device" is a system that updates the responses generated by a generative model with reference data and aggregates the necessary information.
[0720] A "response generation device" is a device that has the function of transmitting information integrated by an information acquisition device to an information terminal used by the user.
[0721] A "decision-making device" is a system that analyzes voice input from citizens and provides specific guidance on procedures and necessary documents.
[0722] A "knowledge extraction device" is a device that optimizes the answers output by a generative model based on the policy information of government agencies at that time.
[0723] A system for implementing this invention consists of the following components: an information terminal, a computing device, a generative model, an information acquisition device, a response generation device, and a decision-making device.
[0724] The information terminals used will be smartphones and smart glasses that allow users to input inquiries via voice or text. These terminals communicate with a server via the internet and transmit user inquiries to a computing device. Voice input will be converted to text using the Google Speech-to-Text API.
[0725] The server, as a computing device, is built using the Flask framework and processes user input using natural language processing. The spaCy library is used for this processing, understanding the intent of queries and identifying the next data to reference. Furthermore, the GPT-3 generative model is incorporated, generating natural and accurate responses based on the analysis results. This generative model learns from large datasets and provides answers based on real-world context.
[0726] The information acquisition device compares the responses generated by the generation model with existing databases and extracts and updates the necessary information. The response generation device formats this information and transmits it to the user's information terminal. Ultimately, the user receives guidance on procedures and required documents through the information terminal, enabling citizens to smoothly proceed with the necessary procedures.
[0727] For example, if a citizen asks via voice, "Please tell me how to register my new address," the system will analyze the inquiry and guide them to the necessary documents and submission locations. The information the user needs will be displayed on the terminal, and a path will be set up to complete the procedure online.
[0728] As an example of a prompt, a user query is sent to the system in the format "User Query: 'Please tell me how to register a new address.'" Based on this, the request to the generative model is expressed as "Explain the steps, required documents, and possible online submission options for new address registration."
[0729] The flow of a specific process in Application Example 1 will be explained using Figure 12.
[0730] Step 1:
[0731] The user uses a smartphone or smart glasses to input their inquiry via voice or text. The input voice is converted into text data using the Google Speech-to-Text API and sent from the smart device to the server.
[0732] Step 2:
[0733] The server is built using Flask and receives input data from users. This data is processed using the spaCy library to analyze the intent of the query. Based on the analyzed intent, the next necessary data is identified. The input is the user's query, and the output is the analysis result including the identified intent.
[0734] Step 3:
[0735] The server sends the analysis results to the generative AI model. The GPT-3 generative model receives the intent in the form of a prompt and generates a relevant response. The prompt is sent in the form of "User Query: 'Please tell me how to register a new address.'" The input to the model is the analysis results, and the output is the generated natural language response.
[0736] Step 4:
[0737] The server receives the response from the generative model, and the information retrieval device accesses the database to collect necessary additional information based on the generated response. Database matching updates and integrates the information. The input is the response from the generative model, and the output is the integrated additional information.
[0738] Step 5:
[0739] The response generation device formats the integrated information and sends the response to the user's information terminal. The user can then easily access information about the procedure and necessary documents on their terminal. The input is integrated information, and the output is a formatted response.
[0740] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.
[0741] This invention relates to a system that receives inquiries from users, analyzes those emotions using an emotion engine, and generates appropriate responses through a generative AI model. The aim of this system is to improve the user experience and the quality of personalized responses.
[0742] The system's main components include terminals for users to input inquiries, servers for data processing and sentiment analysis, and generative AI models and sentiment engines. These elements communicate over the internet and exchange information in real time.
[0743] The user enters text about questions or problems through their terminal. The entered data is sent to the server, which holds the received data in a buffer.
[0744] On the server, the natural language processing module is executed first to analyze the user's query. This analysis extracts the specific information the user intends to convey. Next, the emotion engine uses the same data to determine the user's emotional state (e.g., joy, anger, sadness). The emotion engine uses sophisticated algorithms to estimate emotions based on factors such as tone of voice and text, and keywords.
[0745] The generative AI model considers the results of natural language processing and emotion engine analysis to generate responses tailored to specific emotions. The tone and expression of the response are adjusted according to the emotion. For example, if an inquiry includes dissatisfaction, the generative model will generate a response in milder language that includes countermeasures that take this into account.
[0746] The server scrutinizes the generated response by linking it with the database and updates the information as needed. The response is then formatted and sent to the terminal.
[0747] The device displays responses in a user-optimized format, making it easier for users to understand the next steps. For example, if a user is dissatisfied with an administrative service, the system will return a response that emphasizes explanations of countermeasures, including expressions of gratitude and apology.
[0748] This invention enables the system to respond sensitively to user emotions and provide more personalized services. Furthermore, it is expected that administrative services will be further improved through continuous improvement based on feedback.
[0749] The following describes the processing flow.
[0750] Step 1:
[0751] Users can input questions and comments about administrative procedures and services through their terminals. For example, they can input something like, "My address change procedure is delayed."
[0752] Step 2:
[0753] The terminal processes the input data and sends it to the server using a secure protocol. The data is properly structured according to the transmission format.
[0754] Step 3:
[0755] The server temporarily stores the query received from the terminal in a data buffer. This data is then used in the subsequent analysis process.
[0756] Step 4:
[0757] The server activates a natural language processing module to analyze the query. It extracts keywords and context from the text to identify the information the user intends to receive.
[0758] Step 5:
[0759] The server uses an emotion engine to analyze the sentiment in the query text. For example, it can determine from the text content whether the user is feeling frustrated.
[0760] Step 6:
[0761] The server calls a generative AI model based on the analyzed inquiry content and sentiment data. The AI model generates an answer that reflects both of these factors.
[0762] Step 7:
[0763] The server uses information retrieval tools to obtain relevant details from the database regarding the generated response, and then uses those details to supplement or update the response.
[0764] Step 8:
[0765] The server adjusts the tone and style of the emotionally generated responses, refining the formatting. For example, it might add nuances of apology or change the language to be more polite.
[0766] Step 9:
[0767] The server sends the completed response to the terminal. This process is carried out quickly to ensure the user has the optimal reaction time.
[0768] Step 10:
[0769] The device displays the answer to the user, allowing the user to obtain specific actions or further guidance regarding their inquiry.
[0770] Step 11:
[0771] Users can provide feedback on the answers. This feedback is returned to the system and stored as data for future improvements.
[0772] (Example 2)
[0773] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".
[0774] Existing customer service systems often provide uniform answers without adequately considering user emotions, resulting in a poor user experience. Furthermore, the quality of the generated answers is not optimized for the user's specific emotional state, leading to insufficient individualized support.
[0775] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.
[0776] In this invention, the server includes means for performing natural language analysis using a computing device, means for inferring the user's emotional state using an emotion analysis device, and means for generating an emotion-based response using a generative model device. This enables the generation of sophisticated responses that take emotional states into consideration.
[0777] A "terminal device" is a device that receives inquiries from users and sends data to a server.
[0778] A "processing unit" is a device equipped with computing resources for analyzing data obtained from terminal devices.
[0779] An "emotion analysis device" is a device that has the function of inferring the emotional state of a user from their inquiry data.
[0780] A "generative model device" is a device that generates appropriate responses based on analyzed data and emotional states.
[0781] An "information management device" is a device that compares the generated responses with a storage medium and updates the information as needed.
[0782] A "response generation device" is a device that has the function of transmitting information acquired by an information management device to a terminal device.
[0783] A "data analysis device" is a device used to apply optimization based on the user's emotional state to policy information.
[0784] This invention is a system comprising a terminal device for the user to input inquiries, a computing device for analyzing data, an emotion analysis device for analyzing the user's emotions, a generative model device for generating answers using a generative model, an information management device for scrutinizing answers and managing information, and a response creation device for finally creating a response and sending it to the user's terminal.
[0785] The terminal device receives text input from the user and transmits it to the server as a digital signal. Typical personal computers and smartphones are used.
[0786] The computing units deployed on the server are equipped with high-performance CPUs and memory, and analyze data received from users. Here, natural language processing technology is used to enable the analysis of user intent.
[0787] The emotion analysis device extracts text tone and keywords from the text data entered by the user to infer the user's emotional state. By applying a highly accurate algorithm, it can accurately determine emotions such as joy, anger, and sadness.
[0788] The generative modeling device generates responses in a tone appropriate to the user's emotional state, based on the analysis results. The generative AI model used in this process employs the latest natural language generation technology.
[0789] The information management device compares the generated responses with the storage medium to verify the information's up-to-dateness. Information is updated as needed, ensuring that accurate data is always maintained.
[0790] The response generation device sends the reviewed response to the user's terminal device and formats it for appropriate display. This allows the user to logically select their next action.
[0791] For example, if a user uses the prompt "Tell me about recent improvements in government services," the system will analyze the user's intent and return a response that provides detailed information while taking emotions into consideration. This is expected to improve the user experience and dramatically enhance the quality of personalized responses.
[0792] The flow of the specific processing in Example 2 will be explained using Figure 13.
[0793] Step 1:
[0794] The user enters their inquiry in text format using a terminal device. The entered text data serves as a prompt to the system. For example, the user might enter, "Please tell me about recent government services." The terminal then converts this text data into a digital signal and sends it to the server.
[0795] Step 2:
[0796] The server stores the text data received from the terminal in a receive buffer for analysis. It decodes the digital signal as input into text, preparing it for analysis. This data is then sent to the processing unit.
[0797] Step 3:
[0798] The server passes the text data received via the computing unit to a natural language processing module for analysis. The input is raw text data from the user, and the output is keywords related to the analyzed intent and topic. This analysis clarifies the user's specific information needs.
[0799] Step 4:
[0800] The server passes the analyzed results to an emotion analysis device, which evaluates the emotional state based on the text data. The analyzed keywords are used as input, and the user's emotional state (e.g., joy, anger, sadness) is obtained as output. Emotion analysis is performed based on the tone and keywords of the text.
[0801] Step 5:
[0802] The server activates its generative AI model function and generates responses using the results of natural language processing and sentiment analysis. The input here consists of two data points: the user's intent and emotion, and the output is a natural-sounding response adjusted to match the emotion. As a result, responses that are both reliable and approachable are obtained.
[0803] Step 6:
[0804] The server sends the generated responses to the information management device, where they are compared with the database to verify the timeliness and accuracy of the information. Here, the generated responses are checked against the internal data repository to confirm that the information is up-to-date, and then the results are converted into a format that can be sent to the terminal.
[0805] Step 7:
[0806] The server sends the generated response to the terminal device via the response generation device. Finally, the terminal displays this response to the user. The output displays an optimized answer to the user's question and provides information to prompt the user's next action.
[0807] Through these steps, the system provides appropriate information based on the user's emotions in real time, improving the user experience.
[0808] (Application Example 2)
[0809] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".
[0810] Customer support on modern e-commerce sites presents a challenging task, requiring a thorough understanding and appropriate response to user emotions. Uniform, emotion-insensitive responses can degrade the user experience. Technologies are needed to improve this.
[0811] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.
[0812] In this invention, the server includes an input device means for receiving inquiries from users, a computing device means for performing natural language processing based on the inquiries received from the input device means, and an emotion analysis means for performing emotion analysis based on the natural language data processed by the computing device means. This makes it possible to generate responses that take the user's emotions into consideration.
[0813] "An input device for receiving inquiries from users" refers to a device that has the function of allowing users to input inquiries to the system.
[0814] A "computation device means" is a device that has computational functions for performing natural language processing based on data received from an input device means.
[0815] An "emotion analysis device" is a device that has the function of analyzing a user's emotional state from natural language data processed by a computing device.
[0816] A "generative model device means" is a device that has the function of automatically generating appropriate answers to user inquiries based on the results of sentiment analysis means.
[0817] "Information acquisition means" refers to a device that has the function of referencing the response generated by the generation model device means with information from a database and updating it as necessary.
[0818] A "response generation means" is a device that has the function of sending a response to the user in an appropriate format based on the information obtained by the information acquisition means.
[0819] This invention provides a system for improving the response accuracy of customer support on e-commerce websites. The system consists of various components and is implemented by an input device used by the user and multiple modules that operate on a server.
[0820] The user enters their inquiry in text format using an input device. This input device is a common computer terminal such as a smartphone or personal computer. The data entered by the user is transmitted to the server via the internet.
[0821] The server first uses a natural language processing engine, positioned as a computing device, to analyze the user's query. Existing libraries such as "SpaCy" and "Transformers" can be used for natural language processing. This process extracts the specific content intended by the user.
[0822] Next, the emotion analysis system analyzes the user's emotional state based on the text data. This emotion analysis uses algorithms to analyze the tone of the text and specific language patterns. As a result, the user's emotions are recognized as "joy," "anger," etc.
[0823] The generative model device takes the results of natural language processing and sentiment analysis as input and generates the optimal response in plain text form. The generated response is compared with a database by the information acquisition means, and the information is updated as needed. This database referencing ensures that the response to the user is based on the latest information at the present time.
[0824] Finally, the response generation mechanism sends the generated answer to the user's input device in an appropriate format, allowing the user to receive a response to their inquiry. This is displayed on the terminal screen, making it easy for the user to understand what action to take next.
[0825] For example, if a user inquires that they are dissatisfied because their delivery is delayed, the system will use sentiment analysis to read the dissatisfaction, and a generative model will generate a response that includes an apology and a solution that takes the user's feelings into consideration. The response might be: "We apologize for the delay. We are currently checking the delivery status and will contact you shortly."
[0826] An example of a prompt message input to the generating AI model is: "User inquiry: 'My delivery is delayed. Please take action.' Sentiment analysis result: 'Angry' Please provide an appropriate response."
[0827] The flow of a specific process in Application Example 2 will be explained using Figure 14.
[0828] Step 1:
[0829] The user uses an input device to enter their inquiry in text format and presses the submit button. The entered text data is transmitted to the server via the internet. At this point, the input is raw data from the user.
[0830] Step 2:
[0831] The server's computing system initiates natural language processing using the received text data. Using natural language processing libraries (e.g., SpaCy or Transformers), it extracts the specific intent of the query and related information from the text. The input is the user's raw text data, and the output is the parsed structured data.
[0832] Step 3:
[0833] The server uses sentiment analysis tools to estimate the user's emotional state based on structured data. The sentiment analysis algorithm analyzes the tone and keywords of the text and classifies emotions as "joy" or "anger." In this step, the input is output data from natural language processing, and the emotional state is output.
[0834] Step 4:
[0835] The generative model device generates the optimal response based on the sentiment analysis results and structured data. Using the generative AI model, text with appropriate wording corresponding to the inquiry content and the user's emotions is created. The input here is emotions and structured data, and the output is the response text.
[0836] Step 5:
[0837] The server compares the generated response with the database through the information retrieval mechanism and updates the information as needed. Database lookups are performed to ensure that the response is based on the latest information. The input for this step is the generated response, and the output is the verified response.
[0838] Step 6:
[0839] The response generation mechanism sends the confirmed response text to the user's input device. The display on the terminal is in a format that is easy for the user to understand. The final input is the confirmed response, and the output is the final response displayed on the user's screen.
[0840] 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 controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.
[0841] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.
[0842] 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 this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.
[0843] Furthermore, the emotion identification model 59, acting 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 a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.
[0844] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.
[0845] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.
[0846] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.
[0847] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.
[0848] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is 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 the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."
[0849] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.
[0850] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.
[0851] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.
[0852] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.
[0853] 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.
[0854] Furthermore, it is not necessary to store the entirety 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 the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.
[0855] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.
[0856] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of 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). Alternatively, the hardware resource that performs a specific process may consist of a single processor.
[0857] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.
[0858] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.
[0859] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.
[0860] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.
[0861] The following is further disclosed regarding the embodiments described above.
[0862] (Claim 1)
[0863] A terminal device for receiving inquiries from users,
[0864] A computer means that performs natural language processing based on an inquiry received from the terminal means,
[0865] A generative model means that generates an answer based on natural language data processed by the aforementioned computer means,
[0866] The aforementioned generation model means provides an information acquisition means that updates the information by referencing a database,
[0867] A response generation means that transmits the information integrated by the information acquisition means to a terminal means,
[0868] A system that includes this.
[0869] (Claim 2)
[0870] The system according to claim 1, further comprising terminal means for receiving user feedback.
[0871] (Claim 3)
[0872] The system according to claim 1, further comprising data mining means for optimizing the responses generated by the generation model means based on policy information of local governments.
[0873] "Example 1"
[0874] (Claim 1)
[0875] A means of communication for receiving inquiries from users,
[0876] A processing unit that analyzes inquiries received from the aforementioned communication means and performs natural language processing,
[0877] Information referencing means for searching a database and extracting information based on the intent identified by the aforementioned processing device,
[0878] A response model means that generates an answer using an AI based on extracted information,
[0879] A means of transmission that formats the generated response to help the user understand it and transmits it to the terminal,
[0880] A system that includes this.
[0881] (Claim 2)
[0882] The system according to claim 1, further comprising a feedback management means for accepting user feedback and optimizing the system accordingly.
[0883] (Claim 3)
[0884] The system according to claim 1, further comprising an optimization means for appropriately adjusting the response generated by the response model means based on policy information of public institutions.
[0885] "Application Example 1"
[0886] (Claim 1)
[0887] An information terminal that receives inquiries from users,
[0888] A computing device that performs natural language processing based on an inquiry received from the aforementioned information terminal,
[0889] A generative model that generates answers based on natural language data processed by the aforementioned computing device,
[0890] An information acquisition device that updates the information using reference data based on the response generated by the aforementioned generation model,
[0891] A response generation device that transmits the information integrated by the aforementioned information acquisition device to an information terminal,
[0892] A decision-making device that analyzes voice input from citizens and guides them through procedures and provides necessary documents,
[0893] A system that includes this.
[0894] (Claim 2)
[0895] The system according to claim 1, further comprising an information terminal for receiving user feedback.
[0896] (Claim 3)
[0897] The system according to claim 1, further comprising a knowledge extraction device that optimizes the responses generated by the generation model based on policy information of government agencies.
[0898] "Example 2 of combining an emotion engine"
[0899] (Claim 1)
[0900] A terminal device that receives inquiries from users,
[0901] A computing device that analyzes the data obtained from the terminal device,
[0902] An emotion analysis device that estimates the user's emotional state based on the query data analyzed by the aforementioned computing device,
[0903] A generative model device that generates a response using the output of the emotion analysis device and the natural language analysis results,
[0904] An information management device updates the information by referring to a storage medium to the response generated by the aforementioned generation model device,
[0905] A response creation device that transmits information acquired by the information management device to a terminal device,
[0906] A system that includes this.
[0907] (Claim 2)
[0908] The system according to claim 1, wherein the generative model device adjusts the response based on the user's emotional state.
[0909] (Claim 3)
[0910] The system according to claim 1, further comprising a data analysis device that applies optimization based on the user's emotional state to policy information.
[0911] "Application example 2 when combining with an emotional engine"
[0912] (Claim 1)
[0913] An input device means for receiving inquiries from users,
[0914] A computing device means that performs natural language processing based on a query received from the input device means,
[0915] A sentiment analysis means that performs sentiment analysis based on natural language data processed by the aforementioned computing device means,
[0916] A generative model device means that generates a response based on the results of the emotion analysis means,
[0917] The aforementioned generation model device means provides an information acquisition means that updates the information by referring to a database,
[0918] A response generation means that transmits the information integrated by the information acquisition means to an input device means,
[0919] A system that includes this.
[0920] (Claim 2)
[0921] The system according to claim 1, which receives feedback from a user through the input device means.
[0922] (Claim 3)
[0923] The system according to claim 1, further comprising data analysis means for optimizing the response generated by the generation model device means based on statistical information. [Explanation of symbols]
[0924] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>
Claims
1. A terminal device for receiving inquiries from users, A computer means that performs natural language processing based on an inquiry received from the terminal means, A generative model means that generates an answer based on natural language data processed by the aforementioned computer means, The aforementioned generation model means provides an information acquisition means that updates the information by referencing a database, A response generation means that transmits the information integrated by the information acquisition means to a terminal means, A system that includes this.
2. The system according to claim 1, further comprising terminal means for receiving user feedback.
3. The system according to claim 1, further comprising data mining means for optimizing the responses generated by the generation model means based on policy information of local governments.