system

Abstract

We provide the system. [Solution] A means of receiving questions from users, A means of analyzing received questions using natural language processing algorithms to extract intent and entities, A means of searching for related information from a database based on the analysis results, A means of generating appropriate answers based on search results, A means of displaying the generated response to the user, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, the method including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In modern society, improving the efficiency of customer support is very important, and in particular, a quick and accurate response to a large number of inquiries is required. In conventional support chats, many rely on human responses, which has led to problems such as response delays and lack of consistency in answers. In addition, there are few means to effectively utilize a vast amount of information, and it has been difficult to appropriately search for and use past inquiries and related data. The present invention aims to solve these problems and improve the efficiency and quality of support chats.

Means for Solving the Problems

[0005] The present invention is a system that includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, means for generating appropriate answers based on the search results, and means for displaying the generated answers to the user. In this system, the system also includes means for periodically collecting information from databases and means for searching for and integrating information from multiple databases based on the analysis results, thereby improving the comprehensiveness of information and the accuracy of searches, and enabling the provision of quick and accurate answers.

[0006] "Means for receiving user questions" refers to the means by which questions entered by users through the support chat interface are sent to and received by the server.

[0007] A "natural language processing algorithm" is an algorithm that analyzes text data received from a user and extracts intent and entities within that text.

[0008] "Intention" refers to the purpose or requirement that the user is trying to achieve, as extracted from the user's questions and statements.

[0009] An "entity" refers to information that is extracted from a user's questions or statements, representing a specific subject or keyword.

[0010] A "means for searching related information" refers to a method for searching for related information within a database based on the analyzed intent or entity.

[0011] A "database" is a system or location that systematically stores information related to a question and past Q&A, and keeps it in a searchable format.

[0012] "Answer generation means" refers to a means of creating an appropriate answer to provide to the user based on the relevant information that has been searched.

[0013] A "means for displaying questions" refers to a means of displaying the generated answers on the user's device and providing them in a format that the user can verify.

[0014] "Information gathering methods" refer to means of periodically collecting new information from external sources and historical data to update databases.

[0015] "Information integration methods" refer to methods for integrating relevant information obtained from multiple databases and providing it to the user as a single, unified answer. [Brief explanation of the drawing]

[0016] [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] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when the emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when the emotion engine is combined.

Mode for Carrying Out the Invention

[0017] Hereinafter, an example of an embodiment of the system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0018] [[ID=2O]]First, the terms used in the following description will be described.

[0019] In the following embodiments, the numbered 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.

[0020] In the following embodiments, the numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0021] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0022] 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).

[0023] 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."

[0024] [First Embodiment]

[0025] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0026] 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.

[0027] 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).

[0028] 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.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] 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".

[0037] This invention relates to a method and system for responding quickly and accurately to user questions in a support chat system. In particular, this system improves the efficiency of support operations by analyzing user-entered questions, searching for relevant information from multiple databases, and generating and displaying answers.

[0038] System Overview

[0039] This system includes the following main means:

[0040] 1. Means of receiving questions from users

[0041] 2. Means for analyzing question content using natural language processing algorithms

[0042] 3. Means for searching for related information based on analysis results

[0043] 4. Means of generating answers based on search results

[0044] 5. Means for displaying the generated response to the user

[0045] Program processing flow

[0046] User

[0047] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the system receives it and proceeds to the next step.

[0048] terminal

[0049] The terminal receives input from the user and sends it to the server. The submitted questions are received and analyzed on the server side.

[0050] server

[0051] The server first analyzes the content of the question using a natural language processing (NLP) algorithm. As a result of this analysis, the intent of the question (e.g., "How to use the new Crew") and the entity (e.g., "Crew") are extracted.

[0052] Next, the server searches the database for relevant information based on the extracted intent and entities. The database includes, for example, crew navigation information and past Q&A data, and this information is retrieved quickly.

[0053] Based on the searched information, the server generates appropriate answers to provide to the user. These generated answers are then formatted and presented in a user-friendly manner.

[0054] terminal

[0055] The terminal receives the response generated from the server and displays it to the user. Based on that response, the user can obtain the necessary information.

[0056] Specific example

[0057] 1. The user types "How do I use the new crew?" into the support chat.

[0058] 2. The terminal sends user input to the server.

[0059] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[0060] 4. The server searches the database based on the analysis results and retrieves relevant information.

[0061] 5. The server generates a response that explains in detail "how to use the new crew" based on the search results.

[0062] 6. The device displays the generated response to the user, allowing the user to obtain the necessary information.

[0063] By implementing this invention, responses to user inquiries can be handled efficiently and consistently, improving the quality of support services.

[0064] The following describes the processing flow.

[0065] Step 1:

[0066] The user enters their question into the support chat interface. For example, they might type, "How do I use the new crew?"

[0067] Step 2:

[0068] The terminal receives input from the user and sends the question to the server. This transmission is done via an HTTP POST request.

[0069] Step 3:

[0070] The server receives the question sent from the terminal. The received question is taken as a string and proceeds to the next processing stage.

[0071] Step 4:

[0072] The server uses a natural language processing (NLP) algorithm to analyze the question. This analysis extracts the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew").

[0073] Step 5:

[0074] The server searches for relevant information from the database based on the analysis results. First, it searches the CrewNavi database, and then it searches the past Q&A database.

[0075] Step 6:

[0076] The server integrates the search results and generates an appropriate answer to provide to the user. This answer is then formatted in a human-readable format.

[0077] Step 7:

[0078] The server sends the generated response to the terminal. The terminal displays the received response to the user.

[0079] Step 8:

[0080] Users can check the answers displayed on their devices and obtain the necessary information, thereby resolving their questions.

[0081] This series of steps enables quick and accurate responses to user inquiries.

[0082] (Example 1)

[0083] 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."

[0084] Traditional support chat systems struggled to provide quick and accurate answers to user questions. This resulted in users spending a long time obtaining appropriate information, leading to decreased support efficiency. Furthermore, manual processes were involved in database retrieval and answer generation, consuming significant human resources.

[0085] 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.

[0086] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, and means for generating appropriate answers using an AI model based on the search results. This enables the provision of quick and accurate answers to user questions, thereby improving the efficiency and quality of support operations.

[0087] "Means for receiving user questions" refers to a function that incorporates user-entered questions into the system and plays the role of sending data to the server via a communication protocol.

[0088] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a function that uses natural language processing technology to analyze questions received from users and extract the main point and important information of those questions.

[0089] "A means of searching for relevant information from a database based on analysis results" refers to a function that efficiently searches for relevant information from an appropriate database based on the analysis results obtained through natural language processing.

[0090] "A means of generating appropriate answers using an AI model based on search results" refers to a function that uses an AI model to generate the optimal answer to a user's question based on information obtained from a database.

[0091] "Means for displaying generated answers to the user" refers to a function that displays the answers generated from the server on the user's terminal, allowing the user to easily check the answers.

[0092] "Means of regularly collecting information from a database" refers to a function that periodically updates or collects information from a database to keep the system up-to-date.

[0093] "A means of searching for and integrating information from multiple databases based on analysis results" refers to a function that searches multiple databases across different databases in accordance with the results of analysis using natural language processing, integrates the obtained information, and derives a single answer.

[0094] This invention relates to a support chat system for responding quickly and accurately to user inquiries. This system improves the efficiency of support operations by analyzing user-entered questions, searching for relevant information from multiple databases, and generating and displaying answers.

[0095] This system includes the following main components:

[0096] 1. Means of receiving questions from users

[0097] 2. Means for analyzing question content using natural language processing algorithms

[0098] 3. Means for searching for related information based on analysis results

[0099] 4. Generating answers based on search results: A method for generating answers using an AI model.

[0100] 5. Means for displaying the generated response to the user

[0101] Means for receiving questions from users

[0102] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the user's device sends the question data to the server. The user's input data is stored as a string in the device's local storage or memory.

[0103] A method for analyzing question content using a natural language processing algorithm.

[0104] The server receives question data sent from the terminal and analyzes the question content using natural language processing (NLP) algorithms. This analysis uses software such as Python's NLTK library, spaCy, or TENSORFLOW®. This analysis extracts the intent of the question (e.g., "How to use the new Crew") and entities (e.g., "Crew").

[0105] A means of searching for related information based on analysis results.

[0106] The server searches for relevant information from the database based on the analysis results obtained through natural language processing. The databases used here include, for example, MySQL®, PostgreSQL, or MongoDB. Search queries are dynamically generated according to the extracted intent and entities.

[0107] A method for generating answers based on search results using an AI model.

[0108] The server uses a generative AI model (e.g., OpenAI®'s GPT-3®) based on information retrieved from the database to generate responses for the user. The user inputs prompts to the AI ​​model, which then generates the optimal response. Examples of specific prompts are shown below:

[0109] User: How do I use the new crew?

[0110] System prompt: The user is asking how to use a new crew. Based on the following database information, generate an answer that the user will understand.

[0111] Database information:

[0112] New crew members can be used in the next step...

[0113] Points to note include...

[0114] Answer: To start using a new crew, first go to the "Settings" menu...

[0115] A means of displaying the generated response to the user.

[0116] The device receives the response generated from the server and displays it to the user. This display uses UI components from a web browser or native application to present the response in a user-friendly format.

[0117] By implementing this invention, it becomes possible to provide quick and accurate answers to user questions, thereby improving the efficiency and quality of support operations.

[0118] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0119] Step 1:

[0120] The user accesses the support chat interface and enters their question. The user types "How do I use the new crew?" into the input field and presses the submit button. The input data is stored as a string in the user's device's local storage or memory.

[0121] Input: User-entered question: "How do I use the new crew?"

[0122] Output: Temporary storage of question data on the terminal

[0123] Step 2:

[0124] The terminal receives the query data from the user and sends it to the server in JSON format. The communication protocol used is, for example, HTTP. The terminal includes the query data in the request body and sends a POST request to a specific URL on the server (e.g., https: / / support.example.com / query).

[0125] Input: Question data entered by the user on the device.

[0126] Output: HTTP POST request to the server

[0127] Step 3:

[0128] The server receives question data sent from the terminal. The server analyzes this JSON-formatted question data and uses natural language processing algorithms to extract the intent and entities of the question. The software used here includes Python's NLTK library, spaCy, and TensorFlow. As a result of the analysis, the server obtains the main point of the question and related keywords.

[0129] Input: Question data in JSON format sent from the device.

[0130] Output: Intent and entity of the extracted question (e.g., "How to use Crew", Entity: "Crew")

[0131] Step 4:

[0132] The server searches the database for relevant information based on the intent and entities extracted through natural language processing. The server executes dynamically generated SQL queries using databases such as MySQL, PostgreSQL, or MongoDB. The server searches the database for and retrieves information related to "how to use Crew."

[0133] Input: Extracted question intent and entities

[0134] Output: Relevant information retrieved from the database (e.g., "Crew Setup Guide")

[0135] Step 5:

[0136] The server generates user-facing responses using a generative AI model (e.g., OpenAI's GPT-3) based on the acquired data. The server inputs prompts into the generative AI model to generate the optimal response. An example of a prompt might be: "The user is asking, 'How do I use the new crew?' Based on the following database information, generate an answer that is easy for the user to understand."

[0137] Input: Related information retrieved from the database, prompt text

[0138] Output: Response generated by the generation AI model (Example: "To start using a new crew, first go to the settings menu...")

[0139] Step 6:

[0140] The terminal receives the response generated from the server and displays it to the user. The response is displayed in the terminal's chat interface, allowing the user to review and use the information.

[0141] Input: Response from a generative AI model sent from the server

[0142] Output: Display of responses to the user on the terminal (specific answers to the question "How do I use the new crew?" are displayed)

[0143] (Application Example 1)

[0144] 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."

[0145] Currently, customer support functions play a crucial role on many e-commerce sites. However, responding quickly and accurately to user inquiries is challenging, especially when it comes to providing appropriate answers to detailed product-related questions. To address this issue, a system is needed that understands user questions and provides relevant information immediately.

[0146] 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.

[0147] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, means for generating appropriate answers based on the search results, means for displaying the generated answers to the user, and means for identifying questions about products and searching for relevant information from a product database. This enables users to obtain quick and appropriate answers, thereby improving customer satisfaction and streamlining support operations.

[0148] "Means for receiving user questions" refers to a mechanism for receiving questions entered by users through an electronic interface.

[0149] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a mechanism that analyzes the content of received questions based on natural language processing technology to extract the intent of the question and related entities.

[0150] "Means for searching for relevant information from a database based on analysis results" refers to a mechanism for searching for relevant information from a database using analysis results obtained by a natural language processing algorithm.

[0151] "Means for generating appropriate answers based on search results" refers to a mechanism for generating answers to be provided to users based on search results obtained from a database.

[0152] "Means for displaying generated answers to the user" refers to a mechanism for displaying generated answers on an interface used by the user.

[0153] "Means for identifying product-related questions and retrieving relevant information from a product database" refers to a mechanism for identifying that a user's question relates to a product and for retrieving relevant information from a database containing information about that product.

[0154] This invention provides a system for quickly and accurately answering user questions on an e-commerce site. The system is comprised of several main means.

[0155] First, the user accesses the support chat via a device such as a smartphone, tablet, or PC and enters their question. Once this question is sent to the server, the server receives it.

[0156] The server then uses a natural language processing (NLP) algorithm to analyze the question received from the user. During this analysis, it extracts the intent of the question (e.g., how to use a product) and the entity (e.g., a specific product). For natural language processing, the 'natural' library is used as an example.

[0157] Based on the analysis results, the server searches the database for relevant information. For product-related questions, it retrieves the necessary information from the product database. A database system such as PostgreSQL is used for this process.

[0158] Based on the search results, the server generates an appropriate answer. The generated answer is then formatted to be more user-friendly.

[0159] Finally, the generated response is sent from the server to the terminal, which then displays it to the user. This allows the user to quickly obtain the necessary information.

[0160] As a concrete example, consider a case where a user types "How do I use this product?" into the support chat of an online shopping site. This question is processed as follows:

[0161] 1. The server receives the user's question.

[0162] 2. The server analyzes the question using a natural language processing algorithm, extracting "how to use the product" as the intent and "this product" as the entity.

[0163] 3. The server searches the product database and retrieves information on how to use the relevant product.

[0164] 4. The server generates a response based on the acquired information, such as, "To use this product, first turn on the power and press the button."

[0165] 5. The terminal displays the generated response to the user.

[0166] An example of a prompt message is, "Please tell me how to use this product."

[0167] By implementing this invention, users can obtain quick and accurate answers, improving customer satisfaction on e-commerce sites. The server is equipped with various means for receiving and analyzing user questions, searching for relevant information, and generating and displaying appropriate answers. This leads to increased efficiency and improved quality in support operations.

[0168] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0169] Step 1:

[0170] Users access the support chat from their smartphone, tablet, or PC and enter their questions. The questions consist of text data, such as "How do I use this product?"

[0171] Step 2:

[0172] The terminal receives a question from the user and sends the text data to the server. Here, the data is transmitted to the server via the network. The input is the user's question text, and the output is the text data sent to the server.

[0173] Step 3:

[0174] The server receives questions from users. The received questions are parsed using natural language processing algorithms. Specifically, the 'natural' library is used for tokenization and storming to extract the intent of the question (e.g., "how to use") and entities (e.g., "product"). The input is the question text submitted by the user, and the output is the parsed intent and entity data.

[0175] Step 4:

[0176] Based on the analysis results, the server searches the product database to retrieve relevant information. It queries the PostgreSQL database to obtain information related to the relevant product. The input is the analysis results, and the output is the product information retrieved from the database.

[0177] Step 5:

[0178] The server generates appropriate answers to user questions based on information obtained from the database. The answers are generated based on previously detected intents and entities. For example, they might be generated as text including specific steps, such as, "To use this product, first turn on the power and press the button." The input is product information retrieved from the database, and the output is the generated answer text.

[0179] Step 6:

[0180] The server sends the generated response to the terminal. The response text is transmitted to the terminal via the network. The input is the generated response text, and the output is the text data sent to the terminal.

[0181] Step 7:

[0182] The terminal displays the response received from the server to the user. The displayed content is the response text generated by the server. The input is the response text sent from the server, and the output is the text data displayed to the user.

[0183] As an example of a prompt, consider "How do I use this product?". Each processing step will explain in detail how this prompt is input, parsed, the answer is generated, and the display is shown. This is the specific flow of processing a user question.

[0184] 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.

[0185] This invention provides a method and system for responding quickly and accurately to user questions in a support chat system. In particular, this system improves the efficiency of support operations by analyzing user-inputted questions, searching for relevant information from multiple databases, and generating and displaying answers. Furthermore, this invention can achieve even more advanced support by incorporating an emotion engine that recognizes user emotions and adjusts responses accordingly.

[0186] System Overview

[0187] This system includes the following main means:

[0188] 1. Means of receiving questions from users

[0189] 2. Means for analyzing question content using natural language processing algorithms

[0190] 3. Means for searching for related information based on analysis results

[0191] 4. Means of generating answers based on search results

[0192] 5. Means for displaying the generated response to the user

[0193] 6. Emotion engine that recognizes emotions from user input

[0194] 7. Means for adjusting responses based on emotion recognition results

[0195] 8. A means of selecting different response tones according to the user's emotions.

[0196] This system not only enables quick and appropriate responses to user inquiries, but also allows for flexible responses that take user emotions into consideration.

[0197] Program processing flow

[0198] User

[0199] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the system receives it and proceeds to the next step.

[0200] terminal

[0201] The terminal receives input from the user and sends it to the server. The submitted questions are received and analyzed on the server side.

[0202] server

[0203] The server first analyzes the content of the question using a natural language processing (NLP) algorithm. As a result of this analysis, the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew") are extracted.

[0204] Next, the server uses an emotion engine to recognize emotions from the user's input. This engine analyzes the emotional tone and meaning within the text and determines the user's emotional state based on that. For example, it might recognize "distressed" or "frustrated."

[0205] The server searches for relevant information from the database based on the analysis results and emotion recognition results. First, it searches the CrewNavi database, and then the past Q&A database. It integrates the retrieved information and generates an appropriate answer to provide to the user. This answer is adjusted in tone and content based on the user's emotions.

[0206] terminal

[0207] The terminal receives the response generated from the server and displays it to the user. The displayed response includes a flexible approach that takes the user's emotions into consideration. Based on the response, the user can obtain the necessary information.

[0208] Specific example

[0209] 1. The user types "How do I use the new crew?" into the support chat.

[0210] 2. The terminal sends user input to the server.

[0211] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[0212] 4. The server uses an emotion engine to recognize the user's emotions and, for example, determine that the user is "distressed."

[0213] 5. The server searches the database based on the analysis results and emotion recognition results and retrieves relevant information.

[0214] 6. The server generates a detailed response explaining "how to use the new crew" based on the search results. At this time, the text is adjusted to a gentle tone based on the sentiment recognition results.

[0215] 7. The device displays the generated response to the user, allowing the user to obtain the necessary information in an easy-to-understand and reassuring way.

[0216] By implementing this invention, not only will user inquiries be handled efficiently and consistently, but flexible support that takes user emotions into consideration will be provided, dramatically improving the quality of support services.

[0217] The following describes the processing flow.

[0218] Step 1:

[0219] The user enters their question into the support chat interface. For example, they might type, "How do I use the new crew?"

[0220] Step 2:

[0221] The terminal receives input from the user and sends the question to the server. This transmission is done via an HTTP POST request.

[0222] Step 3:

[0223] The server receives the question sent from the terminal. The received question is taken as a string and proceeds to the next processing stage.

[0224] Step 4:

[0225] The server uses a natural language processing (NLP) algorithm to analyze the question. This analysis extracts the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew").

[0226] Step 5:

[0227] The server uses an emotion engine to recognize emotions from user input. This engine analyzes the emotional tone and keywords in the text to determine the user's emotional state. For example, it can extract emotions such as "troubled" or "frustrated" from the text.

[0228] Step 6:

[0229] The server searches the database for relevant information based on the analysis results and sentiment recognition results. First, it searches the CrewNavi database, and then the past Q&A database. The search query is generated based on the extracted intent and entities.

[0230] Step 7:

[0231] The server integrates the search results and generates an appropriate response to provide to the user. This response is tailored based on the user's sentiment perception. For example, if the user is judged to be "distressed," the response will be structured in a more helpful and reassuring tone.

[0232] Step 8:

[0233] The server sends the generated response to the terminal. The terminal receives it and displays it to the user.

[0234] Step 9:

[0235] Users can check the answers displayed on their devices and obtain the necessary information, thereby resolving their questions.

[0236] This series of steps enables quick and accurate responses to user questions, while also providing flexible support that takes user emotions into consideration.

[0237] (Example 2)

[0238] 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".

[0239] Conventional support chat systems often fail to respond to user inquiries quickly and appropriately. Furthermore, they struggle to provide flexible responses that take user emotions into account, resulting in a poor user experience. This invention aims to solve these problems by providing prompt and accurate answers to user inquiries while also enabling responses that consider user emotions.

[0240] 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.

[0241] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, and means for searching for relevant information from a database based on the analysis results and sentiment recognition results. This makes it possible to generate and display quick and accurate answers based on the content of the questions and the user's sentiment.

[0242] 1. "Means for receiving questions from users" refers to a function that retrieves questions entered by the user within the system and passes them on to the next processing step.

[0243] 2. "Means for analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a function that uses natural language processing technology to identify and extract the purpose and main elements of a question from received text.

[0244] 3. "Means for searching for relevant information from a database based on analysis results and emotion recognition results" refers to a function that searches for and retrieves relevant information from a database, taking into account the analysis results of the question content and the user's emotional state as determined by emotion recognition.

[0245] 4. "Means for integrating searched information and generating responses based on the user's emotional state" refers to a function that comprehensively processes the acquired information and creates responses with appropriate tone and content to match the user's emotional state.

[0246] 5. "Means for displaying generated answers to the user" refers to a function that displays answers created within the system on the user's screen and provides them to the user.

[0247] 6. "Means for periodically collecting information from a database" refers to a function in which the system retrieves the latest information from the database at regular intervals and always maintains up-to-date data.

[0248] 7. "Means for searching and integrating information from multiple databases based on analysis results" refers to a function that searches multiple different databases based on analysis and centralizes and integrates the obtained information.

[0249] This invention provides a method and system for responding quickly and accurately to user questions in a support chat system. This system recognizes the user's emotions and adjusts the response accordingly, thereby achieving a higher level of support.

[0250] This system is broadly composed of the following elements:

[0251] 1. Means of receiving questions from users

[0252] 2. Means for analyzing question content using natural language processing algorithms

[0253] 3. A means of searching for relevant information from a database based on the analysis results and emotion recognition results.

[0254] 4. Means for integrating retrieved information and generating responses based on the user's emotional state.

[0255] 5. Means for displaying the generated response to the user

[0256] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" When the user submits the question, the input is received on the device and sent to the server.

[0257] The terminal sends user questions as text data to the server. This text data is sent to the server using the HTTP protocol. The server receives the sent questions and proceeds to the analysis process.

[0258] The server first analyzes the content of the question using a natural language processing algorithm. Specifically, generative AI models such as Google's BERT and OpenAI's GPT-4 are available. Based on this analysis, the intent and entities of the question are extracted. For example, the intent "How to use the new Crew" and the entity "Crew" are extracted.

[0259] Next, the server uses an emotion recognition algorithm to recognize the user's emotions. Tools such as IBM's Tone Analyzer are used to analyze the emotional tone and meaning within the text. This allows for the recognition of states such as "distressed" or "frustrated."

[0260] The server searches databases to retrieve relevant information based on the analysis results and sentiment recognition results. It executes SQL queries against the CrewNavi database and past Q&A databases to extract the necessary information. The retrieved information is integrated through a response generation mechanism. Here, the response generation uses models such as the GPT-4 model, and the tone and content are adjusted based on the user's sentiment.

[0261] The generated response is sent back to the device. This response is sent as text data and displayed to the user on the device. The displayed response includes flexible responses that take the user's emotions into consideration.

[0262] Specific example

[0263] 1. The user types "How do I use the new crew?" into the support chat.

[0264] 2. The terminal sends user input to the server.

[0265] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[0266] 4. The server uses an emotion engine to recognize the user's emotions and, for example, determine that the user is "distressed."

[0267] 5. The server searches the database based on the analysis results and emotion recognition results and retrieves relevant information.

[0268] 6. The server generates a detailed response explaining "how to use the new crew" based on the search results. At this time, the text is adjusted to a gentle tone based on the sentiment recognition results.

[0269] 7. The device displays the generated response to the user, allowing the user to obtain the necessary information in an easy-to-understand and reassuring way.

[0270] Example of a prompt

[0271] By inputting prompts like the following into the AI ​​model, it is possible to obtain specific answers to user questions.

[0272] "Please explain how to use the new crew. Our users seem to be having trouble."

[0273] By implementing this invention, not only will user inquiries be handled efficiently and consistently, but flexible support that takes user emotions into consideration will be provided, dramatically improving the quality of support services.

[0274] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0275] Step 1:

[0276] The user enters a question.

[0277] Input: The user accesses the support chat interface and enters their question in text format.

[0278] Specific operation: The user enters "Please teach me how to use the new crew" and clicks the send button.

[0279] Output: The input text data is sent to the terminal.

[0280] Step 2:

[0281] The terminal sends the user's question to the server.

[0282] Input: The text data entered and sent by the user in Step 1.

[0283] Specific operation: The terminal sends this text data to the server as an HTTP POST request.

[0284] Output: The text data reaches the server.

[0285] Step 3:

[0286] The server receives the question.

[0287] Input: The text data sent from the terminal in Step 2.

[0288] Specific operation: The server receives the HTTP POST request and extracts the question text.

[0289] Output: The extracted text data is passed to the next analysis process.

[0290] Step 4:

[0291] The server analyzes the question content using a natural language processing algorithm.

[0292] Input: The question text extracted in Step 3.

[0293] Specific operation: The server uses generative AI models such as BERT and GPT-4 to analyze text and extract the intent of the question and its entities.

[0294] Output: As an analysis result, for example, the intention "how to use the crew" and the entity "crew" are extracted.

[0295] Step 5:

[0296] The server recognizes the user's emotions using an emotion engine.

[0297] Input: Question text analyzed in Step 4.

[0298] Specific operation: The server uses an emotion recognition algorithm (e.g., IBM's Tone Analyzer) to recognize the user's emotional state from the text.

[0299] Output: As an emotion recognition result, the user's emotion, such as "distressed," is determined.

[0300] Step 6:

[0301] The server searches the database for relevant information based on the analysis results and emotion recognition results.

[0302] Input: Analysis results from Step 4 and emotion recognition results from Step 5.

[0303] Specific operation: The server uses SQL queries to retrieve relevant information from the CrewNavi database and past Q&A database based on the analysis results and sentiment recognition results.

[0304] Output: Relevant information retrieved from the database.

[0305] Step 7:

[0306] The server integrates the searched information and generates responses based on the user's emotional state.

[0307] Input: The related information obtained in Step 6.

[0308] Specific operation: The server integrates the obtained information and generates an answer in a tone suitable for the user's sentiment using the GPT-4 model.

[0309] Output: The final answer text to be provided to the user.

[0310] Step 8:

[0311] The server sends the generated answer to the terminal.

[0312] Input: The answer text generated in Step 7.

[0313] Specific operation: The server sends the generated answer text to the terminal as an HTTP POST request.

[0314] Output: The answer text reaches the terminal.

[0315] Step 9:

[0316] The terminal displays the answer to the user.

[0317] Input: The answer text sent from the server in Step 8.

[0318] Specific operation: The terminal displays the received answer text in the chat window and provides it to the user.

[0319] Output: The user can confirm an answer in a friendly tone such as "The usage of the new crew is as follows..." and obtain the necessary information.

[0320] (Application Example 2)

[0321] 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".

[0322] Customer support on e-commerce sites requires quick and appropriate responses to the content and tone of inquiries, but traditional systems struggle to provide flexible responses that take user emotions into consideration. Furthermore, there is a challenge in providing answers that are intuitively easy to understand and reassuring.

[0323] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for receiving questions from the user, means for analyzing the received questions with a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, means for generating an appropriate answer based on the search results, means for displaying the generated answer to the user, emotion recognition engine means for recognizing emotions from the user's input text, and means for adjusting the generated answer based on the user's emotions. This enables flexible and prompt customer support that takes the user's emotions into consideration, and allows for responses that give the user a sense of security.

[0324] "Means for receiving user inquiries" refers to interface devices or software that receive inquiries and questions made by users to the system.

[0325] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to natural language processing technology used to analyze text entered by a user and identify the main intent and related elements (entities) within it.

[0326] "Means for searching for relevant information from a database based on analysis results" refers to a function that searches for and retrieves appropriate information from databases located on a server or network based on the results of natural language processing.

[0327] "Means for generating appropriate answers based on search results" refers to algorithms and programs that automatically create specific answers to user questions based on relevant information obtained from a database.

[0328] "Means of displaying generated answers to the user" refers to display devices or user interfaces that display generated answers in a format that the user can view.

[0329] An "emotion recognition engine that recognizes emotions from user input text" refers to an algorithm or model that analyzes and identifies the emotional state of a user at that time (for example, confusion, frustration, joy, etc.) from the text they input.

[0330] "Means for adjusting generated responses based on user emotions" refers to methods or programs for adjusting the tone and expression of responses in accordance with the user's emotions identified by the emotion recognition engine.

[0331] This invention is a system for providing customer support on an e-commerce site and includes the following components.

[0332] The system's main components are users, terminals, and servers. The following outlines how each component interacts and the specific steps taken to achieve its objectives.

[0333] 1. User

[0334] The user first accesses the support chat interface and enters their question. This question might be something like, "How do I return an item?"

[0335] 2. Terminal

[0336] The terminal receives input from the user and sends it to the server. The terminal includes a user interface, through which it displays the response to the user. The terminal is typically a communication device such as a smartphone or computer.

[0337] 3. Server

[0338] The server first analyzes the user's submitted question using a natural language processing (NLP) algorithm. This analysis extracts the question's main intent and related entities. For example, a natural language processing library such as "nlp_library" is used for this analysis. Next, the server uses the sentiment recognition engine "sentiment_analysis_library" to determine the user's emotional state from the user's input text. Based on this sentiment recognition result, the server searches the database for relevant information.

[0339] The database contains past Q&A data and product information, and is accessed using "database_connector". Based on the search results, an algorithm is activated to generate appropriate answers, and the answers are generated.

[0340] The generated responses are adjusted in tone and expression according to the user's emotions. For example, if the emotion recognition engine determines that the user is "distressed," an explanation will be provided in a gentle tone.

[0341] The terminal displays the final generated response to the user through the user interface.

[0342] Specific example

[0343] For example, a user might type "How do I return this item?" into a support chat. This question is received by the device and sent to the server. The server uses an NLP algorithm to extract the intent "return" and the entity "method," and its sentiment recognition engine determines that the user is "in distress." It then searches its database for information on "return procedures" and generates a gentle-toned response based on that information, such as "Please don't worry. We will explain the return process in detail."

[0344] Examples of input prompts for a generative AI model:

[0345] Prompt: "How do I return this item?"

[0346] Expected output: "Please rest assured. We will explain the return process in detail."

[0347] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0348] Step 1: The user enters their question in the support chat.

[0349] Specific action: The user accesses the support chat interface and enters a question such as, "How do I return this item?"

[0350] Input: User's question text

[0351] Output: Question text sent to the terminal

[0352] Step 2: The device sends the question to the server.

[0353] Specific operation: The terminal sends the question text received from the user to the server. HTTP / HTTPS is used as the communication protocol.

[0354] Input: Question text

[0355] Output: Question text sent to the server

[0356] Step 3: The server analyzes the question using a natural language processing (NLP) algorithm.

[0357] Specific operation: The server uses "nlp_library" to parse the received question text and extract the main intent and related elements (entities).

[0358] Input: Question text

[0359] Output: Intent (intent) and Entity (related element)

[0360] Step 4: The server recognizes the user's emotions.

[0361] Specific operation: The server uses the "sentiment_analysis_library" to recognize the user's emotional state from the question text, such as confusion or frustration.

[0362] Input: Question text

[0363] Output: User's emotion (e.g., troubled)

[0364] Step 5: The server searches the database for relevant information.

[0365] Specific operation: The server uses "database_connector" to search for relevant information from the database based on the analysis results. The database to be searched includes past Q&A data and product information.

[0366] Input: Intent and entity

[0367] Output: Related information (e.g., details on how to return the item)

[0368] Step 6: The server generates the response and adjusts it based on sentiment.

[0369] Specific operation: The server generates a response based on the relevant information it has acquired, and adjusts the tone and expression based on the emotions recognized by the "emotion recognition engine." For example, if the user is "distressed," the response tone will become softer.

[0370] Input: Related information, user sentiment

[0371] Output: Adjusted response text

[0372] Step 7: The server sends the generated response to the terminal.

[0373] Specific operation: The server sends the adjusted response text to the terminal.

[0374] Input: Adjusted response text

[0375] Output: Answer text sent to the terminal

[0376] Step 8: The device displays the answer to the user.

[0377] Specific operation: The terminal displays the received response text in the user interface, making it viewable by the user.

[0378] Input: Response text sent from the server

[0379] Output: Answer displayed in the user interface

[0380] 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.

[0381] 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.

[0382] 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.

[0383] [Second Embodiment]

[0384] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0385] 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.

[0386] 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).

[0387] 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.

[0388] 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.

[0389] 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).

[0390] 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.

[0391] 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.

[0392] 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.

[0393] 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.

[0394] 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.

[0395] 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".

[0396] This invention relates to a method and system for responding quickly and accurately to user questions in a support chat system. In particular, this system improves the efficiency of support operations by analyzing user-entered questions, searching for relevant information from multiple databases, and generating and displaying answers.

[0397] System Overview

[0398] This system includes the following main means:

[0399] 1. Means of receiving questions from users

[0400] 2. Means for analyzing question content using natural language processing algorithms

[0401] 3. Means for searching for related information based on analysis results

[0402] 4. Means of generating answers based on search results

[0403] 5. Means for displaying the generated response to the user

[0404] Program processing flow

[0405] User

[0406] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the system receives it and proceeds to the next step.

[0407] terminal

[0408] The terminal receives input from the user and sends it to the server. The submitted questions are received and analyzed on the server side.

[0409] server

[0410] The server first analyzes the content of the question using a natural language processing (NLP) algorithm. As a result of this analysis, the intent of the question (e.g., "How to use the new Crew") and the entity (e.g., "Crew") are extracted.

[0411] Next, the server searches the database for relevant information based on the extracted intent and entities. The database includes, for example, crew navigation information and past Q&A data, and this information is retrieved quickly.

[0412] Based on the searched information, the server generates appropriate answers to provide to the user. These generated answers are then formatted and presented in a user-friendly manner.

[0413] terminal

[0414] The terminal receives the response generated from the server and displays it to the user. Based on that response, the user can obtain the necessary information.

[0415] Specific example

[0416] 1. The user types "How do I use the new crew?" into the support chat.

[0417] 2. The terminal sends user input to the server.

[0418] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[0419] 4. The server searches the database based on the analysis results and retrieves relevant information.

[0420] 5. The server generates a response that explains in detail "how to use the new crew" based on the search results.

[0421] 6. The device displays the generated response to the user, allowing the user to obtain the necessary information.

[0422] By implementing this invention, responses to user inquiries can be handled efficiently and consistently, improving the quality of support services.

[0423] The following describes the processing flow.

[0424] Step 1:

[0425] The user enters their question into the support chat interface. For example, they might type, "How do I use the new crew?"

[0426] Step 2:

[0427] The terminal receives input from the user and sends the question to the server. This transmission is done via an HTTP POST request.

[0428] Step 3:

[0429] The server receives the question sent from the terminal. The received question is taken as a string and proceeds to the next processing stage.

[0430] Step 4:

[0431] The server uses a natural language processing (NLP) algorithm to analyze the question. This analysis extracts the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew").

[0432] Step 5:

[0433] The server searches for relevant information from the database based on the analysis results. First, it searches the CrewNavi database, and then it searches the past Q&A database.

[0434] Step 6:

[0435] The server integrates the search results and generates an appropriate answer to provide to the user. This answer is then formatted in a human-readable format.

[0436] Step 7:

[0437] The server sends the generated response to the terminal. The terminal displays the received response to the user.

[0438] Step 8:

[0439] Users can check the answers displayed on their devices and obtain the necessary information, thereby resolving their questions.

[0440] This series of steps enables quick and accurate responses to user inquiries.

[0441] (Example 1)

[0442] 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."

[0443] Traditional support chat systems struggled to provide quick and accurate answers to user questions. This resulted in users spending a long time obtaining appropriate information, leading to decreased support efficiency. Furthermore, manual processes were involved in database retrieval and answer generation, consuming significant human resources.

[0444] 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.

[0445] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, and means for generating appropriate answers using an AI model based on the search results. This enables the provision of quick and accurate answers to user questions, thereby improving the efficiency and quality of support operations.

[0446] "Means for receiving user questions" refers to a function that incorporates user-entered questions into the system and plays the role of sending data to the server via a communication protocol.

[0447] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a function that uses natural language processing technology to analyze questions received from users and extract the main point and important information of those questions.

[0448] "A means of searching for relevant information from a database based on analysis results" refers to a function that efficiently searches for relevant information from an appropriate database based on the analysis results obtained through natural language processing.

[0449] "A means of generating appropriate answers using an AI model based on search results" refers to a function that uses an AI model to generate the optimal answer to a user's question based on information obtained from a database.

[0450] "Means for displaying generated answers to the user" refers to a function that displays the answers generated from the server on the user's terminal, allowing the user to easily check the answers.

[0451] "Means of regularly collecting information from a database" refers to a function that periodically updates or collects information from a database to keep the system up-to-date.

[0452] "A means of searching for and integrating information from multiple databases based on analysis results" refers to a function that searches multiple databases across different databases in accordance with the results of analysis using natural language processing, integrates the obtained information, and derives a single answer.

[0453] This invention relates to a support chat system for responding quickly and accurately to user inquiries. This system improves the efficiency of support operations by analyzing user-entered questions, searching for relevant information from multiple databases, and generating and displaying answers.

[0454] This system includes the following main components:

[0455] 1. Means of receiving questions from users

[0456] 2. Means for analyzing question content using natural language processing algorithms

[0457] 3. Means for searching for related information based on analysis results

[0458] 4. Generating answers based on search results: A method for generating answers using an AI model.

[0459] 5. Means for displaying the generated response to the user

[0460] Means for receiving questions from users

[0461] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the user's device sends the question data to the server. The user's input data is stored as a string in the device's local storage or memory.

[0462] A method for analyzing question content using a natural language processing algorithm.

[0463] The server receives question data sent from the terminal and analyzes the question content using natural language processing (NLP) algorithms. This analysis uses software such as Python's NLTK library, spaCy, or TensorFlow. This analysis extracts the intent of the question (e.g., "How to use the new Crew") and entities (e.g., "Crew").

[0464] A means of searching for related information based on analysis results.

[0465] The server searches for relevant information from the database based on the analysis results obtained from natural language processing. The database used here could be, for example, MySQL, PostgreSQL, or MongoDB. Search queries are dynamically generated based on the extracted intent and entities.

[0466] A method for generating answers based on search results using an AI model.

[0467] The server uses a generative AI model (e.g., OpenAI's GPT-3) to generate user-facing responses based on information retrieved from the database. The user inputs prompts to the AI ​​model, which then generates the optimal response. Examples of specific prompts are shown below:

[0468] User: How do I use the new crew?

[0469] System prompt: The user is asking how to use a new crew. Based on the following database information, generate an answer that the user will understand.

[0470] Database information:

[0471] New crew members can be used in the next step...

[0472] Points to note include...

[0473] Answer: To start using a new crew, first go to the "Settings" menu...

[0474] A means of displaying the generated response to the user.

[0475] The device receives the response generated from the server and displays it to the user. This display uses UI components from a web browser or native application to present the response in a user-friendly format.

[0476] By implementing this invention, it becomes possible to provide quick and accurate answers to user questions, thereby improving the efficiency and quality of support operations.

[0477] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0478] Step 1:

[0479] The user accesses the support chat interface and enters their question. The user types "How do I use the new crew?" into the input field and presses the submit button. The input data is stored as a string in the user's device's local storage or memory.

[0480] Input: User-entered question: "How do I use the new crew?"

[0481] Output: Temporary storage of question data on the terminal

[0482] Step 2:

[0483] The terminal receives the query data from the user and sends it to the server in JSON format. The communication protocol used is, for example, HTTP. The terminal includes the query data in the request body and sends a POST request to a specific URL on the server (e.g., https: / / support.example.com / query).

[0484] Input: Question data entered by the user on the device.

[0485] Output: HTTP POST request to the server

[0486] Step 3:

[0487] The server receives question data sent from the terminal. The server analyzes this JSON-formatted question data and uses natural language processing algorithms to extract the intent and entities of the question. The software used here includes Python's NLTK library, spaCy, and TensorFlow. As a result of the analysis, the server obtains the main point of the question and related keywords.

[0488] Input: Question data in JSON format sent from the device.

[0489] Output: Intent and entity of the extracted question (e.g., "How to use Crew", Entity: "Crew")

[0490] Step 4:

[0491] The server searches the database for relevant information based on the intent and entities extracted through natural language processing. The server executes dynamically generated SQL queries using databases such as MySQL, PostgreSQL, or MongoDB. The server searches the database for and retrieves information related to "how to use Crew."

[0492] Input: Extracted question intent and entities

[0493] Output: Relevant information retrieved from the database (e.g., "Crew Setup Guide")

[0494] Step 5:

[0495] The server generates user-facing responses using a generative AI model (e.g., OpenAI's GPT-3) based on the acquired data. The server inputs prompts into the generative AI model to generate the optimal response. An example of a prompt might be: "The user is asking, 'How do I use the new crew?' Based on the following database information, generate an answer that is easy for the user to understand."

[0496] Input: Related information retrieved from the database, prompt text

[0497] Output: Response generated by the generation AI model (Example: "To start using a new crew, first go to the settings menu...")

[0498] Step 6:

[0499] The terminal receives the response generated from the server and displays it to the user. The response is displayed in the terminal's chat interface, allowing the user to review and use the information.

[0500] Input: Response from a generative AI model sent from the server

[0501] Output: Display of responses to the user on the terminal (specific answers to the question "How do I use the new crew?" are displayed)

[0502] (Application Example 1)

[0503] 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."

[0504] Currently, customer support functions play a crucial role on many e-commerce sites. However, responding quickly and accurately to user inquiries is challenging, especially when it comes to providing appropriate answers to detailed product-related questions. To address this issue, a system is needed that understands user questions and provides relevant information immediately.

[0505] 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.

[0506] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, means for generating appropriate answers based on the search results, means for displaying the generated answers to the user, and means for identifying questions about products and searching for relevant information from a product database. This enables users to obtain quick and appropriate answers, thereby improving customer satisfaction and streamlining support operations.

[0507] "Means for receiving user questions" refers to a mechanism for receiving questions entered by users through an electronic interface.

[0508] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a mechanism that analyzes the content of received questions based on natural language processing technology to extract the intent of the question and related entities.

[0509] "Means for searching for relevant information from a database based on analysis results" refers to a mechanism for searching for relevant information from a database using analysis results obtained by a natural language processing algorithm.

[0510] "Means for generating appropriate answers based on search results" refers to a mechanism for generating answers to be provided to users based on search results obtained from a database.

[0511] "Means for displaying generated answers to the user" refers to a mechanism for displaying generated answers on an interface used by the user.

[0512] "Means for identifying product-related questions and retrieving relevant information from a product database" refers to a mechanism for identifying that a user's question relates to a product and for retrieving relevant information from a database containing information about that product.

[0513] This invention provides a system for quickly and accurately answering user questions on an e-commerce site. The system is comprised of several main means.

[0514] First, the user accesses the support chat via a device such as a smartphone, tablet, or PC and enters their question. Once this question is sent to the server, the server receives it.

[0515] The server then uses a natural language processing (NLP) algorithm to analyze the question received from the user. During this analysis, it extracts the intent of the question (e.g., how to use a product) and the entity (e.g., a specific product). For natural language processing, the 'natural' library is used as an example.

[0516] Based on the analysis results, the server searches the database for relevant information. For product-related questions, it retrieves the necessary information from the product database. A database system such as PostgreSQL is used for this process.

[0517] Based on the search results, the server generates an appropriate answer. The generated answer is then formatted to be more user-friendly.

[0518] Finally, the generated response is sent from the server to the terminal, which then displays it to the user. This allows the user to quickly obtain the necessary information.

[0519] As a concrete example, consider a case where a user types "How do I use this product?" into the support chat of an online shopping site. This question is processed as follows:

[0520] 1. The server receives the user's question.

[0521] 2. The server analyzes the question using a natural language processing algorithm, extracting "how to use the product" as the intent and "this product" as the entity.

[0522] 3. The server searches the product database and retrieves information on how to use the relevant product.

[0523] 4. The server generates a response based on the acquired information, such as, "To use this product, first turn on the power and press the button."

[0524] 5. The terminal displays the generated response to the user.

[0525] An example of a prompt message is, "Please tell me how to use this product."

[0526] By implementing this invention, users can obtain quick and accurate answers, improving customer satisfaction on e-commerce sites. The server is equipped with various means for receiving and analyzing user questions, searching for relevant information, and generating and displaying appropriate answers. This leads to increased efficiency and improved quality in support operations.

[0527] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0528] Step 1:

[0529] Users access the support chat from their smartphone, tablet, or PC and enter their questions. The questions consist of text data, such as "How do I use this product?"

[0530] Step 2:

[0531] The terminal receives a question from the user and sends the text data to the server. Here, the data is transmitted to the server via the network. The input is the user's question text, and the output is the text data sent to the server.

[0532] Step 3:

[0533] The server receives questions from users. The received questions are parsed using natural language processing algorithms. Specifically, the 'natural' library is used for tokenization and storming to extract the intent of the question (e.g., "how to use") and entities (e.g., "product"). The input is the question text submitted by the user, and the output is the parsed intent and entity data.

[0534] Step 4:

[0535] Based on the analysis results, the server searches the product database to retrieve relevant information. It queries the PostgreSQL database to obtain information related to the relevant product. The input is the analysis results, and the output is the product information retrieved from the database.

[0536] Step 5:

[0537] The server generates appropriate answers to user questions based on information obtained from the database. The answers are generated based on previously detected intents and entities. For example, they might be generated as text including specific steps, such as, "To use this product, first turn on the power and press the button." The input is product information retrieved from the database, and the output is the generated answer text.

[0538] Step 6:

[0539] The server sends the generated response to the terminal. The response text is transmitted to the terminal via the network. The input is the generated response text, and the output is the text data sent to the terminal.

[0540] Step 7:

[0541] The terminal displays the response received from the server to the user. The displayed content is the response text generated by the server. The input is the response text sent from the server, and the output is the text data displayed to the user.

[0542] As an example of a prompt, consider "How do I use this product?". Each processing step will explain in detail how this prompt is input, parsed, the answer is generated, and the display is shown. This is the specific flow of processing a user question.

[0543] 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.

[0544] This invention provides a method and system for responding quickly and accurately to user questions in a support chat system. In particular, this system improves the efficiency of support operations by analyzing user-inputted questions, searching for relevant information from multiple databases, and generating and displaying answers. Furthermore, this invention can achieve even more advanced support by incorporating an emotion engine that recognizes user emotions and adjusts responses accordingly.

[0545] System Overview

[0546] This system includes the following main means:

[0547] 1. Means of receiving questions from users

[0548] 2. Means for analyzing question content using natural language processing algorithms

[0549] 3. Means for searching for related information based on analysis results

[0550] 4. Means of generating answers based on search results

[0551] 5. Means for displaying the generated response to the user

[0552] 6. Emotion engine that recognizes emotions from user input

[0553] 7. Means for adjusting responses based on emotion recognition results

[0554] 8. A means of selecting different response tones according to the user's emotions.

[0555] This system not only enables quick and appropriate responses to user inquiries, but also allows for flexible responses that take user emotions into consideration.

[0556] Program processing flow

[0557] User

[0558] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the system receives it and proceeds to the next step.

[0559] terminal

[0560] The terminal receives input from the user and sends it to the server. The submitted questions are received and analyzed on the server side.

[0561] server

[0562] The server first analyzes the content of the question using a natural language processing (NLP) algorithm. As a result of this analysis, the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew") are extracted.

[0563] Next, the server uses an emotion engine to recognize emotions from the user's input. This engine analyzes the emotional tone and meaning within the text and determines the user's emotional state based on that. For example, it might recognize "distressed" or "frustrated."

[0564] The server searches for relevant information from the database based on the analysis results and emotion recognition results. First, it searches the CrewNavi database, and then the past Q&A database. It integrates the retrieved information and generates an appropriate answer to provide to the user. This answer is adjusted in tone and content based on the user's emotions.

[0565] terminal

[0566] The terminal receives the response generated from the server and displays it to the user. The displayed response includes a flexible approach that takes the user's emotions into consideration. Based on the response, the user can obtain the necessary information.

[0567] Specific example

[0568] 1. The user types "How do I use the new crew?" into the support chat.

[0569] 2. The terminal sends user input to the server.

[0570] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[0571] 4. The server uses an emotion engine to recognize the user's emotions and, for example, determine that the user is "distressed."

[0572] 5. The server searches the database based on the analysis results and emotion recognition results and retrieves relevant information.

[0573] 6. The server generates a detailed response explaining "how to use the new crew" based on the search results. At this time, the text is adjusted to a gentle tone based on the sentiment recognition results.

[0574] 7. The device displays the generated response to the user, allowing the user to obtain the necessary information in an easy-to-understand and reassuring way.

[0575] By implementing this invention, not only will user inquiries be handled efficiently and consistently, but flexible support that takes user emotions into consideration will be provided, dramatically improving the quality of support services.

[0576] The following describes the processing flow.

[0577] Step 1:

[0578] The user enters their question into the support chat interface. For example, they might type, "How do I use the new crew?"

[0579] Step 2:

[0580] The terminal receives input from the user and sends the question to the server. This transmission is done via an HTTP POST request.

[0581] Step 3:

[0582] The server receives the question sent from the terminal. The received question is taken as a string and proceeds to the next processing stage.

[0583] Step 4:

[0584] The server uses a natural language processing (NLP) algorithm to analyze the question. This analysis extracts the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew").

[0585] Step 5:

[0586] The server uses an emotion engine to recognize emotions from user input. This engine analyzes the emotional tone and keywords in the text to determine the user's emotional state. For example, it can extract emotions such as "troubled" or "frustrated" from the text.

[0587] Step 6:

[0588] The server searches the database for relevant information based on the analysis results and sentiment recognition results. First, it searches the CrewNavi database, and then the past Q&A database. The search query is generated based on the extracted intent and entities.

[0589] Step 7:

[0590] The server integrates the search results and generates an appropriate response to provide to the user. This response is tailored based on the user's sentiment perception. For example, if the user is judged to be "distressed," the response will be structured in a more helpful and reassuring tone.

[0591] Step 8:

[0592] The server sends the generated response to the terminal. The terminal receives it and displays it to the user.

[0593] Step 9:

[0594] Users can check the answers displayed on their devices and obtain the necessary information, thereby resolving their questions.

[0595] This series of steps enables quick and accurate responses to user questions, while also providing flexible support that takes user emotions into consideration.

[0596] (Example 2)

[0597] 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".

[0598] Conventional support chat systems often fail to respond to user inquiries quickly and appropriately. Furthermore, they struggle to provide flexible responses that take user emotions into account, resulting in a poor user experience. This invention aims to solve these problems by providing prompt and accurate answers to user inquiries while also enabling responses that consider user emotions.

[0599] 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.

[0600] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, and means for searching for relevant information from a database based on the analysis results and sentiment recognition results. This makes it possible to generate and display quick and accurate answers based on the content of the questions and the user's sentiment.

[0601] 1. "Means for receiving questions from users" refers to a function that retrieves questions entered by the user within the system and passes them on to the next processing step.

[0602] 2. "Means for analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a function that uses natural language processing technology to identify and extract the purpose and main elements of a question from received text.

[0603] 3. "Means for searching for relevant information from a database based on analysis results and emotion recognition results" refers to a function that searches for and retrieves relevant information from a database, taking into account the analysis results of the question content and the user's emotional state as determined by emotion recognition.

[0604] 4. "Means for integrating searched information and generating responses based on the user's emotional state" refers to a function that comprehensively processes the acquired information and creates responses with appropriate tone and content to match the user's emotional state.

[0605] 5. "Means for displaying generated answers to the user" refers to a function that displays answers created within the system on the user's screen and provides them to the user.

[0606] 6. "Means for periodically collecting information from a database" refers to a function in which the system retrieves the latest information from the database at regular intervals and always maintains up-to-date data.

[0607] 7. "Means for searching and integrating information from multiple databases based on analysis results" refers to a function that searches multiple different databases based on analysis and centralizes and integrates the obtained information.

[0608] This invention provides a method and system for responding quickly and accurately to user questions in a support chat system. This system recognizes the user's emotions and adjusts the response accordingly, thereby achieving a higher level of support.

[0609] This system is broadly composed of the following elements:

[0610] 1. Means of receiving questions from users

[0611] 2. Means for analyzing question content using natural language processing algorithms

[0612] 3. A means of searching for relevant information from a database based on the analysis results and emotion recognition results.

[0613] 4. Means for integrating retrieved information and generating responses based on the user's emotional state.

[0614] 5. Means for displaying the generated response to the user

[0615] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" When the user submits the question, the input is received on the device and sent to the server.

[0616] The terminal sends user questions as text data to the server. This text data is sent to the server using the HTTP protocol. The server receives the sent questions and proceeds to the analysis process.

[0617] The server first analyzes the content of the question using a natural language processing algorithm. Specifically, generative AI models such as Google's BERT and OpenAI's GPT-4 are available. Based on the results of this analysis, the intent and entities of the question are extracted. For example, the intent "How to use the new Crew" and the entity "Crew" are extracted.

[0618] Next, the server uses an emotion recognition algorithm to recognize the user's emotions. Tools such as IBM's Tone Analyzer are used to analyze the emotional tone and meaning within the text. This allows for the recognition of states such as "distressed" or "frustrated."

[0619] The server searches databases to retrieve relevant information based on the analysis results and sentiment recognition results. It executes SQL queries against the CrewNavi database and past Q&A databases to extract the necessary information. The retrieved information is integrated through a response generation mechanism. Here, the response generation uses models such as the GPT-4 model, and the tone and content are adjusted based on the user's sentiment.

[0620] The generated response is sent back to the device. This response is sent as text data and displayed to the user on the device. The displayed response includes flexible responses that take the user's emotions into consideration.

[0621] Specific example

[0622] 1. The user types "How do I use the new crew?" into the support chat.

[0623] 2. The terminal sends user input to the server.

[0624] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[0625] 4. The server uses an emotion engine to recognize the user's emotions and, for example, determine that the user is "distressed."

[0626] 5. The server searches the database based on the analysis results and emotion recognition results and retrieves relevant information.

[0627] 6. The server generates a detailed response explaining "how to use the new crew" based on the search results. At this time, the text is adjusted to a gentle tone based on the sentiment recognition results.

[0628] 7. The device displays the generated response to the user, allowing the user to obtain the necessary information in an easy-to-understand and reassuring way.

[0629] Example of a prompt

[0630] By inputting prompts like the following into the AI ​​model, it is possible to obtain specific answers to user questions.

[0631] "Please explain how to use the new crew. Our users seem to be having trouble."

[0632] By implementing this invention, not only will user inquiries be handled efficiently and consistently, but flexible support that takes user emotions into consideration will be provided, dramatically improving the quality of support services.

[0633] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0634] Step 1:

[0635] The user enters a question.

[0636] Input: The user accesses the support chat interface and enters their question in text format.

[0637] Specific action: The user types "Please tell me how to use the new crew" and clicks the submit button.

[0638] Output: The entered text data is sent to the terminal.

[0639] Step 2:

[0640] The terminal sends the user's question to the server.

[0641] Input: Text data entered and submitted by the user in Step 1.

[0642] Specific action: The device sends this text data to the server as an HTTP POST request.

[0643] Output: Text data reaches the server.

[0644] Step 3:

[0645] The server receives the question.

[0646] Input: Text data sent from the device in Step 2.

[0647] Specific operation: The server receives an HTTP POST request and extracts the question text.

[0648] Output: The extracted text data is passed on to the next analysis process.

[0649] Step 4:

[0650] The server analyzes the question using a natural language processing algorithm.

[0651] Input: Question text extracted in Step 3.

[0652] Specific operation: The server uses generative AI models such as BERT and GPT-4 to analyze text and extract the intent of the question and its entities.

[0653] Output: As an analysis result, for example, the intention "how to use the crew" and the entity "crew" are extracted.

[0654] Step 5:

[0655] The server recognizes the user's emotions using an emotion engine.

[0656] Input: Question text analyzed in Step 4.

[0657] Specific operation: The server uses an emotion recognition algorithm (e.g., IBM's Tone Analyzer) to recognize the user's emotional state from the text.

[0658] Output: As an emotion recognition result, the user's emotion, such as "distressed," is determined.

[0659] Step 6:

[0660] The server searches the database for relevant information based on the analysis results and emotion recognition results.

[0661] Input: Analysis results from Step 4 and emotion recognition results from Step 5.

[0662] Specific operation: The server uses SQL queries to retrieve relevant information from the CrewNavi database and past Q&A database based on the analysis results and sentiment recognition results.

[0663] Output: Relevant information retrieved from the database.

[0664] Step 7:

[0665] The server integrates the searched information and generates responses based on the user's emotional state.

[0666] Input: Relevant information obtained in Step 6.

[0667] Specific operation: The server integrates the acquired information and uses the GPT-4 model to generate a response in a tone appropriate to the user's emotions.

[0668] Output: The final answer text to be provided to the user.

[0669] Step 8:

[0670] The server sends the generated response to the terminal.

[0671] Input: The answer text generated in Step 7.

[0672] Specific operation: The server sends the generated response text to the terminal as an HTTP POST request.

[0673] Output: The response text reaches the terminal.

[0674] Step 9:

[0675] The device displays the answer to the user.

[0676] Input: The response text sent from the server in Step 8.

[0677] Specific operation: The device displays the received response text in the chat window and provides it to the user.

[0678] Output: The user can see a gentle response such as "Here's how to use the new crew..." and obtain the necessary information.

[0679] (Application Example 2)

[0680] 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."

[0681] Customer support on e-commerce sites requires quick and appropriate responses to the content and tone of inquiries, but traditional systems struggle to provide flexible responses that take user emotions into consideration. Furthermore, there is a challenge in providing answers that are intuitively easy to understand and reassuring.

[0682] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for receiving questions from the user, means for analyzing the received questions with a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, means for generating an appropriate answer based on the search results, means for displaying the generated answer to the user, emotion recognition engine means for recognizing emotions from the user's input text, and means for adjusting the generated answer based on the user's emotions. This enables flexible and prompt customer support that takes the user's emotions into consideration, and allows for responses that give the user a sense of security.

[0683] "Means for receiving user inquiries" refers to interface devices or software that receive inquiries and questions made by users to the system.

[0684] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to natural language processing technology used to analyze text entered by a user and identify the main intent and related elements (entities) within it.

[0685] "Means for searching for relevant information from a database based on analysis results" refers to a function that searches for and retrieves appropriate information from databases located on a server or network based on the results of natural language processing.

[0686] "Means for generating appropriate answers based on search results" refers to algorithms and programs that automatically create specific answers to user questions based on relevant information obtained from a database.

[0687] "Means of displaying generated answers to the user" refers to display devices or user interfaces that display generated answers in a format that the user can view.

[0688] An "emotion recognition engine that recognizes emotions from user input text" refers to an algorithm or model that analyzes and identifies the emotional state of a user at that time (for example, confusion, frustration, joy, etc.) from the text they input.

[0689] "Means for adjusting generated responses based on user emotions" refers to methods or programs for adjusting the tone and expression of responses in accordance with the user's emotions identified by the emotion recognition engine.

[0690] This invention is a system for providing customer support on an e-commerce site and includes the following components.

[0691] The system's main components are users, terminals, and servers. The following outlines how each component interacts and the specific steps taken to achieve its objectives.

[0692] 1. User

[0693] The user first accesses the support chat interface and enters their question. This question might be something like, "How do I return an item?"

[0694] 2. Terminal

[0695] The terminal receives input from the user and sends it to the server. The terminal includes a user interface, through which it displays the response to the user. The terminal is typically a communication device such as a smartphone or computer.

[0696] 3. Server

[0697] The server first analyzes the user's submitted question using a natural language processing (NLP) algorithm. This analysis extracts the question's main intent and related entities. For example, a natural language processing library such as "nlp_library" is used for this analysis. Next, the server uses the sentiment recognition engine "sentiment_analysis_library" to determine the user's emotional state from the user's input text. Based on this sentiment recognition result, the server searches the database for relevant information.

[0698] The database contains past Q&A data and product information, and is accessed using "database_connector". Based on the search results, an algorithm is activated to generate appropriate answers, and the answers are generated.

[0699] The generated responses are adjusted in tone and expression according to the user's emotions. For example, if the emotion recognition engine determines that the user is "distressed," an explanation will be provided in a gentle tone.

[0700] The terminal displays the final generated response to the user through the user interface.

[0701] Specific example

[0702] For example, a user might type "How do I return this item?" into a support chat. This question is received by the device and sent to the server. The server uses an NLP algorithm to extract the intent "return" and the entity "method," and its sentiment recognition engine determines that the user is "in distress." It then searches its database for information on "return procedures" and generates a gentle-toned response based on that information, such as "Please don't worry. We will explain the return process in detail."

[0703] Examples of input prompts for a generative AI model:

[0704] Prompt: "How do I return this item?"

[0705] Expected output: "Please rest assured. We will explain the return process in detail."

[0706] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0707] Step 1: The user enters their question in the support chat.

[0708] Specific action: The user accesses the support chat interface and enters a question such as, "How do I return this item?"

[0709] Input: User's question text

[0710] Output: Question text sent to the terminal

[0711] Step 2: The device sends the question to the server.

[0712] Specific operation: The terminal sends the question text received from the user to the server. HTTP / HTTPS is used as the communication protocol.

[0713] Input: Question text

[0714] Output: Question text sent to the server

[0715] Step 3: The server analyzes the question using a natural language processing (NLP) algorithm.

[0716] Specific operation: The server uses "nlp_library" to parse the received question text and extract the main intent and related elements (entities).

[0717] Input: Question text

[0718] Output: Intent (intent) and Entity (related element)

[0719] Step 4: The server recognizes the user's emotions.

[0720] Specific operation: The server uses the "sentiment_analysis_library" to recognize the user's emotional state from the question text, such as confusion or frustration.

[0721] Input: Question text

[0722] Output: User's emotion (e.g., troubled)

[0723] Step 5: The server searches the database for relevant information.

[0724] Specific operation: The server uses "database_connector" to search for relevant information from the database based on the analysis results. The database to be searched includes past Q&A data and product information.

[0725] Input: Intent and entity

[0726] Output: Related information (e.g., details on how to return the item)

[0727] Step 6: The server generates the response and adjusts it based on sentiment.

[0728] Specific operation: The server generates a response based on the relevant information it has acquired, and adjusts the tone and expression based on the emotions recognized by the "emotion recognition engine." For example, if the user is "distressed," the response tone will become softer.

[0729] Input: Related information, user sentiment

[0730] Output: Adjusted response text

[0731] Step 7: The server sends the generated response to the terminal.

[0732] Specific operation: The server sends the adjusted response text to the terminal.

[0733] Input: Adjusted response text

[0734] Output: Answer text sent to the terminal

[0735] Step 8: The device displays the answer to the user.

[0736] Specific operation: The terminal displays the received response text in the user interface, making it viewable by the user.

[0737] Input: Response text sent from the server

[0738] Output: Answer displayed in the user interface

[0739] 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.

[0740] 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.

[0741] 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.

[0742] [Third Embodiment]

[0743] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0744] 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.

[0745] 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).

[0746] 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.

[0747] 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.

[0748] 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).

[0749] 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.

[0750] 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.

[0751] 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.

[0752] 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.

[0753] 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.

[0754] 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".

[0755] This invention relates to a method and system for responding quickly and accurately to user questions in a support chat system. In particular, this system improves the efficiency of support operations by analyzing user-entered questions, searching for relevant information from multiple databases, and generating and displaying answers.

[0756] System Overview

[0757] This system includes the following main means:

[0758] 1. Means of receiving questions from users

[0759] 2. Means for analyzing question content using natural language processing algorithms

[0760] 3. Means for searching for related information based on analysis results

[0761] 4. Means of generating answers based on search results

[0762] 5. Means for displaying the generated response to the user

[0763] Program processing flow

[0764] User

[0765] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the system receives it and proceeds to the next step.

[0766] terminal

[0767] The terminal receives input from the user and sends it to the server. The submitted questions are received and analyzed on the server side.

[0768] server

[0769] The server first analyzes the content of the question using a natural language processing (NLP) algorithm. As a result of this analysis, the intent of the question (e.g., "How to use the new Crew") and the entity (e.g., "Crew") are extracted.

[0770] Next, the server searches the database for relevant information based on the extracted intent and entities. The database includes, for example, crew navigation information and past Q&A data, and this information is retrieved quickly.

[0771] Based on the searched information, the server generates appropriate answers to provide to the user. These generated answers are then formatted and presented in a user-friendly manner.

[0772] terminal

[0773] The terminal receives the response generated from the server and displays it to the user. Based on that response, the user can obtain the necessary information.

[0774] Specific example

[0775] 1. The user types "How do I use the new crew?" into the support chat.

[0776] 2. The terminal sends user input to the server.

[0777] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[0778] 4. The server searches the database based on the analysis results and retrieves relevant information.

[0779] 5. The server generates a response that explains in detail "how to use the new crew" based on the search results.

[0780] 6. The device displays the generated response to the user, allowing the user to obtain the necessary information.

[0781] By implementing this invention, responses to user inquiries can be handled efficiently and consistently, improving the quality of support services.

[0782] The following describes the processing flow.

[0783] Step 1:

[0784] The user enters their question into the support chat interface. For example, they might type, "How do I use the new crew?"

[0785] Step 2:

[0786] The terminal receives input from the user and sends the question to the server. This transmission is done via an HTTP POST request.

[0787] Step 3:

[0788] The server receives the question sent from the terminal. The received question is taken as a string and proceeds to the next processing stage.

[0789] Step 4:

[0790] The server uses a natural language processing (NLP) algorithm to analyze the question. This analysis extracts the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew").

[0791] Step 5:

[0792] The server searches for relevant information from the database based on the analysis results. First, it searches the CrewNavi database, and then it searches the past Q&A database.

[0793] Step 6:

[0794] The server integrates the search results and generates an appropriate answer to provide to the user. This answer is then formatted in a human-readable format.

[0795] Step 7:

[0796] The server sends the generated response to the terminal. The terminal displays the received response to the user.

[0797] Step 8:

[0798] Users can check the answers displayed on their devices and obtain the necessary information, thereby resolving their questions.

[0799] This series of steps enables quick and accurate responses to user inquiries.

[0800] (Example 1)

[0801] 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."

[0802] Traditional support chat systems struggled to provide quick and accurate answers to user questions. This resulted in users spending a long time obtaining appropriate information, leading to decreased support efficiency. Furthermore, manual processes were involved in database retrieval and answer generation, consuming significant human resources.

[0803] 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.

[0804] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, and means for generating appropriate answers using an AI model based on the search results. This enables the provision of quick and accurate answers to user questions, thereby improving the efficiency and quality of support operations.

[0805] "Means for receiving user questions" refers to a function that incorporates user-entered questions into the system and plays the role of sending data to the server via a communication protocol.

[0806] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a function that uses natural language processing technology to analyze questions received from users and extract the main point and important information of those questions.

[0807] "A means of searching for relevant information from a database based on analysis results" refers to a function that efficiently searches for relevant information from an appropriate database based on the analysis results obtained through natural language processing.

[0808] "A means of generating appropriate answers using an AI model based on search results" refers to a function that uses an AI model to generate the optimal answer to a user's question based on information obtained from a database.

[0809] "Means for displaying generated answers to the user" refers to a function that displays the answers generated from the server on the user's terminal, allowing the user to easily check the answers.

[0810] "Means of regularly collecting information from a database" refers to a function that periodically updates or collects information from a database to keep the system up-to-date.

[0811] "A means of searching for and integrating information from multiple databases based on analysis results" refers to a function that searches multiple databases across different databases in accordance with the results of analysis using natural language processing, integrates the obtained information, and derives a single answer.

[0812] This invention relates to a support chat system for responding quickly and accurately to user inquiries. This system improves the efficiency of support operations by analyzing user-entered questions, searching for relevant information from multiple databases, and generating and displaying answers.

[0813] This system includes the following main components:

[0814] 1. Means of receiving questions from users

[0815] 2. Means for analyzing question content using natural language processing algorithms

[0816] 3. Means for searching for related information based on analysis results

[0817] 4. Generating answers based on search results: A method for generating answers using an AI model.

[0818] 5. Means for displaying the generated response to the user

[0819] Means for receiving questions from users

[0820] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the user's device sends the question data to the server. The user's input data is stored as a string in the device's local storage or memory.

[0821] A method for analyzing question content using a natural language processing algorithm.

[0822] The server receives question data sent from the terminal and analyzes the question content using natural language processing (NLP) algorithms. This analysis uses software such as Python's NLTK library, spaCy, or TensorFlow. This analysis extracts the intent of the question (e.g., "How to use the new Crew") and entities (e.g., "Crew").

[0823] A means of searching for related information based on analysis results.

[0824] The server searches for relevant information from the database based on the analysis results obtained from natural language processing. The database used here could be, for example, MySQL, PostgreSQL, or MongoDB. Search queries are dynamically generated based on the extracted intent and entities.

[0825] A method for generating answers based on search results using an AI model.

[0826] The server uses a generative AI model (e.g., OpenAI's GPT-3) to generate user-facing responses based on information retrieved from the database. The user inputs prompts to the AI ​​model, which then generates the optimal response. Examples of specific prompts are shown below:

[0827] User: How do I use the new crew?

[0828] System prompt: The user is asking how to use a new crew. Based on the following database information, generate an answer that the user will understand.

[0829] Database information:

[0830] New crew members can be used in the next step...

[0831] Points to note include...

[0832] Answer: To start using a new crew, first go to the "Settings" menu...

[0833] A means of displaying the generated response to the user.

[0834] The device receives the response generated from the server and displays it to the user. This display uses UI components from a web browser or native application to present the response in a user-friendly format.

[0835] By implementing this invention, it becomes possible to provide quick and accurate answers to user questions, thereby improving the efficiency and quality of support operations.

[0836] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0837] Step 1:

[0838] The user accesses the support chat interface and enters their question. The user types "How do I use the new crew?" into the input field and presses the submit button. The input data is stored as a string in the user's device's local storage or memory.

[0839] Input: User-entered question: "How do I use the new crew?"

[0840] Output: Temporary storage of question data on the terminal

[0841] Step 2:

[0842] The terminal receives the query data from the user and sends it to the server in JSON format. The communication protocol used is, for example, HTTP. The terminal includes the query data in the request body and sends a POST request to a specific URL on the server (e.g., https: / / support.example.com / query).

[0843] Input: Question data entered by the user on the device.

[0844] Output: HTTP POST request to the server

[0845] Step 3:

[0846] The server receives question data sent from the terminal. The server analyzes this JSON-formatted question data and uses natural language processing algorithms to extract the intent and entities of the question. The software used here includes Python's NLTK library, spaCy, and TensorFlow. As a result of the analysis, the server obtains the main point of the question and related keywords.

[0847] Input: Question data in JSON format sent from the device.

[0848] Output: Intent and entity of the extracted question (e.g., "How to use Crew", Entity: "Crew")

[0849] Step 4:

[0850] The server searches the database for relevant information based on the intent and entities extracted through natural language processing. The server executes dynamically generated SQL queries using databases such as MySQL, PostgreSQL, or MongoDB. The server searches the database for and retrieves information related to "how to use Crew."

[0851] Input: Extracted question intent and entities

[0852] Output: Relevant information retrieved from the database (e.g., "Crew Setup Guide")

[0853] Step 5:

[0854] The server generates user-facing responses using a generative AI model (e.g., OpenAI's GPT-3) based on the acquired data. The server inputs prompts into the generative AI model to generate the optimal response. An example of a prompt might be: "The user is asking, 'How do I use the new crew?' Based on the following database information, generate an answer that is easy for the user to understand."

[0855] Input: Related information retrieved from the database, prompt text

[0856] Output: Response generated by the generation AI model (Example: "To start using a new crew, first go to the settings menu...")

[0857] Step 6:

[0858] The terminal receives the response generated from the server and displays it to the user. The response is displayed in the terminal's chat interface, allowing the user to review and use the information.

[0859] Input: Response from a generative AI model sent from the server

[0860] Output: Display of responses to the user on the terminal (specific answers to the question "How do I use the new crew?" are displayed)

[0861] (Application Example 1)

[0862] 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."

[0863] Currently, customer support functions play a crucial role on many e-commerce sites. However, responding quickly and accurately to user inquiries is challenging, especially when it comes to providing appropriate answers to detailed product-related questions. To address this issue, a system is needed that understands user questions and provides relevant information immediately.

[0864] 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.

[0865] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, means for generating appropriate answers based on the search results, means for displaying the generated answers to the user, and means for identifying questions about products and searching for relevant information from a product database. This enables users to obtain quick and appropriate answers, thereby improving customer satisfaction and streamlining support operations.

[0866] "Means for receiving user questions" refers to a mechanism for receiving questions entered by users through an electronic interface.

[0867] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a mechanism that analyzes the content of received questions based on natural language processing technology to extract the intent of the question and related entities.

[0868] "Means for searching for relevant information from a database based on analysis results" refers to a mechanism for searching for relevant information from a database using analysis results obtained by a natural language processing algorithm.

[0869] "Means for generating appropriate answers based on search results" refers to a mechanism for generating answers to be provided to users based on search results obtained from a database.

[0870] "Means for displaying generated answers to the user" refers to a mechanism for displaying generated answers on an interface used by the user.

[0871] "Means for identifying product-related questions and retrieving relevant information from a product database" refers to a mechanism for identifying that a user's question relates to a product and for retrieving relevant information from a database containing information about that product.

[0872] This invention provides a system for quickly and accurately answering user questions on an e-commerce site. The system is comprised of several main means.

[0873] First, the user accesses the support chat via a device such as a smartphone, tablet, or PC and enters their question. Once this question is sent to the server, the server receives it.

[0874] The server then uses a natural language processing (NLP) algorithm to analyze the question received from the user. During this analysis, it extracts the intent of the question (e.g., how to use a product) and the entity (e.g., a specific product). For natural language processing, the 'natural' library is used as an example.

[0875] Based on the analysis results, the server searches the database for relevant information. For product-related questions, it retrieves the necessary information from the product database. A database system such as PostgreSQL is used for this process.

[0876] Based on the search results, the server generates an appropriate answer. The generated answer is then formatted to be more user-friendly.

[0877] Finally, the generated response is sent from the server to the terminal, which then displays it to the user. This allows the user to quickly obtain the necessary information.

[0878] As a concrete example, consider a case where a user types "How do I use this product?" into the support chat of an online shopping site. This question is processed as follows:

[0879] 1. The server receives the user's question.

[0880] 2. The server analyzes the question using a natural language processing algorithm, extracting "how to use the product" as the intent and "this product" as the entity.

[0881] 3. The server searches the product database and retrieves information on how to use the relevant product.

[0882] 4. The server generates a response based on the acquired information, such as, "To use this product, first turn on the power and press the button."

[0883] 5. The terminal displays the generated response to the user.

[0884] An example of a prompt message is, "Please tell me how to use this product."

[0885] By implementing this invention, users can obtain quick and accurate answers, improving customer satisfaction on e-commerce sites. The server is equipped with various means for receiving and analyzing user questions, searching for relevant information, and generating and displaying appropriate answers. This leads to increased efficiency and improved quality in support operations.

[0886] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0887] Step 1:

[0888] Users access the support chat from their smartphone, tablet, or PC and enter their questions. The questions consist of text data, such as "How do I use this product?"

[0889] Step 2:

[0890] The terminal receives a question from the user and sends the text data to the server. Here, the data is transmitted to the server via the network. The input is the user's question text, and the output is the text data sent to the server.

[0891] Step 3:

[0892] The server receives questions from users. The received questions are parsed using natural language processing algorithms. Specifically, the 'natural' library is used for tokenization and storming to extract the intent of the question (e.g., "how to use") and entities (e.g., "product"). The input is the question text submitted by the user, and the output is the parsed intent and entity data.

[0893] Step 4:

[0894] Based on the analysis results, the server searches the product database to retrieve relevant information. It queries the PostgreSQL database to obtain information related to the relevant product. The input is the analysis results, and the output is the product information retrieved from the database.

[0895] Step 5:

[0896] The server generates appropriate answers to user questions based on information obtained from the database. The answers are generated based on previously detected intents and entities. For example, they might be generated as text including specific steps, such as, "To use this product, first turn on the power and press the button." The input is product information retrieved from the database, and the output is the generated answer text.

[0897] Step 6:

[0898] The server sends the generated response to the terminal. The response text is transmitted to the terminal via the network. The input is the generated response text, and the output is the text data sent to the terminal.

[0899] Step 7:

[0900] The terminal displays the response received from the server to the user. The displayed content is the response text generated by the server. The input is the response text sent from the server, and the output is the text data displayed to the user.

[0901] As an example of a prompt, consider "How do I use this product?". Each processing step will explain in detail how this prompt is input, parsed, the answer is generated, and the display is shown. This is the specific flow of processing a user question.

[0902] 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.

[0903] This invention provides a method and system for responding quickly and accurately to user questions in a support chat system. In particular, this system improves the efficiency of support operations by analyzing user-inputted questions, searching for relevant information from multiple databases, and generating and displaying answers. Furthermore, this invention can achieve even more advanced support by incorporating an emotion engine that recognizes user emotions and adjusts responses accordingly.

[0904] System Overview

[0905] This system includes the following main means:

[0906] 1. Means of receiving questions from users

[0907] 2. Means for analyzing question content using natural language processing algorithms

[0908] 3. Means for searching for related information based on analysis results

[0909] 4. Means of generating answers based on search results

[0910] 5. Means for displaying the generated response to the user

[0911] 6. Emotion engine that recognizes emotions from user input

[0912] 7. Means for adjusting responses based on emotion recognition results

[0913] 8. A means of selecting different response tones according to the user's emotions.

[0914] This system not only enables quick and appropriate responses to user inquiries, but also allows for flexible responses that take user emotions into consideration.

[0915] Program processing flow

[0916] User

[0917] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the system receives it and proceeds to the next step.

[0918] terminal

[0919] The terminal receives input from the user and sends it to the server. The submitted questions are received and analyzed on the server side.

[0920] server

[0921] The server first analyzes the content of the question using a natural language processing (NLP) algorithm. As a result of this analysis, the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew") are extracted.

[0922] Next, the server uses an emotion engine to recognize emotions from the user's input. This engine analyzes the emotional tone and meaning within the text and determines the user's emotional state based on that. For example, it might recognize "distressed" or "frustrated."

[0923] The server searches for relevant information from the database based on the analysis results and emotion recognition results. First, it searches the CrewNavi database, and then the past Q&A database. It integrates the retrieved information and generates an appropriate answer to provide to the user. This answer is adjusted in tone and content based on the user's emotions.

[0924] terminal

[0925] The terminal receives the response generated from the server and displays it to the user. The displayed response includes a flexible approach that takes the user's emotions into consideration. Based on the response, the user can obtain the necessary information.

[0926] Specific example

[0927] 1. The user types "How do I use the new crew?" into the support chat.

[0928] 2. The terminal sends user input to the server.

[0929] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[0930] 4. The server uses an emotion engine to recognize the user's emotions and, for example, determine that the user is "distressed."

[0931] 5. The server searches the database based on the analysis results and emotion recognition results and retrieves relevant information.

[0932] 6. The server generates a detailed response explaining "how to use the new crew" based on the search results. At this time, the text is adjusted to a gentle tone based on the sentiment recognition results.

[0933] 7. The device displays the generated response to the user, allowing the user to obtain the necessary information in an easy-to-understand and reassuring way.

[0934] By implementing this invention, not only will user inquiries be handled efficiently and consistently, but flexible support that takes user emotions into consideration will be provided, dramatically improving the quality of support services.

[0935] The following describes the processing flow.

[0936] Step 1:

[0937] The user enters their question into the support chat interface. For example, they might type, "How do I use the new crew?"

[0938] Step 2:

[0939] The terminal receives input from the user and sends the question to the server. This transmission is done via an HTTP POST request.

[0940] Step 3:

[0941] The server receives the question sent from the terminal. The received question is taken as a string and proceeds to the next processing stage.

[0942] Step 4:

[0943] The server uses a natural language processing (NLP) algorithm to analyze the question. This analysis extracts the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew").

[0944] Step 5:

[0945] The server uses an emotion engine to recognize emotions from user input. This engine analyzes the emotional tone and keywords in the text to determine the user's emotional state. For example, it can extract emotions such as "troubled" or "frustrated" from the text.

[0946] Step 6:

[0947] The server searches the database for relevant information based on the analysis results and sentiment recognition results. First, it searches the CrewNavi database, and then the past Q&A database. The search query is generated based on the extracted intent and entities.

[0948] Step 7:

[0949] The server integrates the search results and generates an appropriate response to provide to the user. This response is tailored based on the user's sentiment perception. For example, if the user is judged to be "distressed," the response will be structured in a more helpful and reassuring tone.

[0950] Step 8:

[0951] The server sends the generated response to the terminal. The terminal receives it and displays it to the user.

[0952] Step 9:

[0953] Users can check the answers displayed on their devices and obtain the necessary information, thereby resolving their questions.

[0954] This series of steps enables quick and accurate responses to user questions, while also providing flexible support that takes user emotions into consideration.

[0955] (Example 2)

[0956] 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."

[0957] Conventional support chat systems often fail to respond to user inquiries quickly and appropriately. Furthermore, they struggle to provide flexible responses that take user emotions into account, resulting in a poor user experience. This invention aims to solve these problems by providing prompt and accurate answers to user inquiries while also enabling responses that consider user emotions.

[0958] 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.

[0959] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, and means for searching for relevant information from a database based on the analysis results and sentiment recognition results. This makes it possible to generate and display quick and accurate answers based on the content of the questions and the user's sentiment.

[0960] 1. "Means for receiving questions from users" refers to a function that retrieves questions entered by the user within the system and passes them on to the next processing step.

[0961] 2. "Means for analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a function that uses natural language processing technology to identify and extract the purpose and main elements of a question from received text.

[0962] 3. "Means for searching for relevant information from a database based on analysis results and emotion recognition results" refers to a function that searches for and retrieves relevant information from a database, taking into account the analysis results of the question content and the user's emotional state as determined by emotion recognition.

[0963] 4. "Means for integrating searched information and generating responses based on the user's emotional state" refers to a function that comprehensively processes the acquired information and creates responses with appropriate tone and content to match the user's emotional state.

[0964] 5. "Means for displaying generated answers to the user" refers to a function that displays answers created within the system on the user's screen and provides them to the user.

[0965] 6. "Means for periodically collecting information from a database" refers to a function in which the system retrieves the latest information from the database at regular intervals and always maintains up-to-date data.

[0966] 7. "Means for searching and integrating information from multiple databases based on analysis results" refers to a function that searches multiple different databases based on analysis and centralizes and integrates the obtained information.

[0967] This invention provides a method and system for responding quickly and accurately to user questions in a support chat system. This system recognizes the user's emotions and adjusts the response accordingly, thereby achieving a higher level of support.

[0968] This system is broadly composed of the following elements:

[0969] 1. Means of receiving questions from users

[0970] 2. Means for analyzing question content using natural language processing algorithms

[0971] 3. A means of searching for relevant information from a database based on the analysis results and emotion recognition results.

[0972] 4. Means for integrating retrieved information and generating responses based on the user's emotional state.

[0973] 5. Means for displaying the generated response to the user

[0974] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" When the user submits the question, the input is received on the device and sent to the server.

[0975] The terminal sends user questions as text data to the server. This text data is sent to the server using the HTTP protocol. The server receives the sent questions and proceeds to the analysis process.

[0976] The server first analyzes the content of the question using a natural language processing algorithm. Specifically, generative AI models such as Google's BERT and OpenAI's GPT-4 are available. Based on the results of this analysis, the intent and entities of the question are extracted. For example, the intent "How to use the new Crew" and the entity "Crew" are extracted.

[0977] Next, the server uses an emotion recognition algorithm to recognize the user's emotions. Tools such as IBM's Tone Analyzer are used to analyze the emotional tone and meaning within the text. This allows for the recognition of states such as "distressed" or "frustrated."

[0978] The server searches databases to retrieve relevant information based on the analysis results and sentiment recognition results. It executes SQL queries against the CrewNavi database and past Q&A databases to extract the necessary information. The retrieved information is integrated through a response generation mechanism. Here, the response generation uses models such as the GPT-4 model, and the tone and content are adjusted based on the user's sentiment.

[0979] The generated response is sent back to the device. This response is sent as text data and displayed to the user on the device. The displayed response includes flexible responses that take the user's emotions into consideration.

[0980] Specific example

[0981] 1. The user types "How do I use the new crew?" into the support chat.

[0982] 2. The terminal sends user input to the server.

[0983] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[0984] 4. The server uses an emotion engine to recognize the user's emotions and, for example, determine that the user is "distressed."

[0985] 5. The server searches the database based on the analysis results and emotion recognition results and retrieves relevant information.

[0986] 6. The server generates a detailed response explaining "how to use the new crew" based on the search results. At this time, the text is adjusted to a gentle tone based on the sentiment recognition results.

[0987] 7. The device displays the generated response to the user, allowing the user to obtain the necessary information in an easy-to-understand and reassuring way.

[0988] Example of a prompt

[0989] By inputting prompts like the following into the AI ​​model, it is possible to obtain specific answers to user questions.

[0990] "Please explain how to use the new crew. Our users seem to be having trouble."

[0991] By implementing this invention, not only will user inquiries be handled efficiently and consistently, but flexible support that takes user emotions into consideration will be provided, dramatically improving the quality of support services.

[0992] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0993] Step 1:

[0994] The user enters a question.

[0995] Input: The user accesses the support chat interface and enters their question in text format.

[0996] Specific action: The user types "Please tell me how to use the new crew" and clicks the submit button.

[0997] Output: The entered text data is sent to the terminal.

[0998] Step 2:

[0999] The terminal sends the user's question to the server.

[1000] Input: Text data entered and submitted by the user in Step 1.

[1001] Specific action: The device sends this text data to the server as an HTTP POST request.

[1002] Output: Text data reaches the server.

[1003] Step 3:

[1004] The server receives the question.

[1005] Input: Text data sent from the device in Step 2.

[1006] Specific operation: The server receives an HTTP POST request and extracts the question text.

[1007] Output: The extracted text data is passed on to the next analysis process.

[1008] Step 4:

[1009] The server analyzes the question using a natural language processing algorithm.

[1010] Input: Question text extracted in Step 3.

[1011] Specific operation: The server uses generative AI models such as BERT and GPT-4 to analyze text and extract the intent of the question and its entities.

[1012] Output: As an analysis result, for example, the intention "how to use the crew" and the entity "crew" are extracted.

[1013] Step 5:

[1014] The server recognizes the user's emotions using an emotion engine.

[1015] Input: Question text analyzed in Step 4.

[1016] Specific operation: The server uses an emotion recognition algorithm (e.g., IBM's Tone Analyzer) to recognize the user's emotional state from the text.

[1017] Output: As an emotion recognition result, the user's emotion, such as "distressed," is determined.

[1018] Step 6:

[1019] The server searches the database for relevant information based on the analysis results and emotion recognition results.

[1020] Input: Analysis results from Step 4 and emotion recognition results from Step 5.

[1021] Specific operation: The server uses SQL queries to retrieve relevant information from the CrewNavi database and past Q&A database based on the analysis results and sentiment recognition results.

[1022] Output: Relevant information retrieved from the database.

[1023] Step 7:

[1024] The server integrates the searched information and generates responses based on the user's emotional state.

[1025] Input: Relevant information obtained in Step 6.

[1026] Specific operation: The server integrates the acquired information and uses the GPT-4 model to generate a response in a tone appropriate to the user's emotions.

[1027] Output: The final answer text to be provided to the user.

[1028] Step 8:

[1029] The server sends the generated response to the terminal.

[1030] Input: The answer text generated in Step 7.

[1031] Specific operation: The server sends the generated response text to the terminal as an HTTP POST request.

[1032] Output: The response text reaches the terminal.

[1033] Step 9:

[1034] The device displays the answer to the user.

[1035] Input: The response text sent from the server in Step 8.

[1036] Specific operation: The device displays the received response text in the chat window and provides it to the user.

[1037] Output: The user can see a gentle response such as "Here's how to use the new crew..." and obtain the necessary information.

[1038] (Application Example 2)

[1039] 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."

[1040] Customer support on e-commerce sites requires quick and appropriate responses to the content and tone of inquiries, but traditional systems struggle to provide flexible responses that take user emotions into consideration. Furthermore, there is a challenge in providing answers that are intuitively easy to understand and reassuring.

[1041] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for receiving questions from the user, means for analyzing the received questions with a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, means for generating an appropriate answer based on the search results, means for displaying the generated answer to the user, emotion recognition engine means for recognizing emotions from the user's input text, and means for adjusting the generated answer based on the user's emotions. This enables flexible and prompt customer support that takes the user's emotions into consideration, and allows for responses that give the user a sense of security.

[1042] "Means for receiving user inquiries" refers to interface devices or software that receive inquiries and questions made by users to the system.

[1043] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to natural language processing technology used to analyze text entered by a user and identify the main intent and related elements (entities) within it.

[1044] "Means for searching for relevant information from a database based on analysis results" refers to a function that searches for and retrieves appropriate information from databases located on a server or network based on the results of natural language processing.

[1045] "Means for generating appropriate answers based on search results" refers to algorithms and programs that automatically create specific answers to user questions based on relevant information obtained from a database.

[1046] "Means of displaying generated answers to the user" refers to display devices or user interfaces that display generated answers in a format that the user can view.

[1047] An "emotion recognition engine that recognizes emotions from user input text" refers to an algorithm or model that analyzes and identifies the emotional state of a user at that time (for example, confusion, frustration, joy, etc.) from the text they input.

[1048] "Means for adjusting generated responses based on user emotions" refers to methods or programs for adjusting the tone and expression of responses in accordance with the user's emotions identified by the emotion recognition engine.

[1049] This invention is a system for providing customer support on an e-commerce site and includes the following components.

[1050] The system's main components are users, terminals, and servers. The following outlines how each component interacts and the specific steps taken to achieve its objectives.

[1051] 1. User

[1052] The user first accesses the support chat interface and enters their question. This question might be something like, "How do I return an item?"

[1053] 2. Terminal

[1054] The terminal receives input from the user and sends it to the server. The terminal includes a user interface, through which it displays the response to the user. The terminal is typically a communication device such as a smartphone or computer.

[1055] 3. Server

[1056] The server first analyzes the user's submitted question using a natural language processing (NLP) algorithm. This analysis extracts the question's main intent and related entities. For example, a natural language processing library such as "nlp_library" is used for this analysis. Next, the server uses the sentiment recognition engine "sentiment_analysis_library" to determine the user's emotional state from the user's input text. Based on this sentiment recognition result, the server searches the database for relevant information.

[1057] The database contains past Q&A data and product information, and is accessed using "database_connector". Based on the search results, an algorithm is activated to generate appropriate answers, and the answers are generated.

[1058] The generated responses are adjusted in tone and expression according to the user's emotions. For example, if the emotion recognition engine determines that the user is "distressed," an explanation will be provided in a gentle tone.

[1059] The terminal displays the final generated response to the user through the user interface.

[1060] Specific example

[1061] For example, a user might type "How do I return this item?" into a support chat. This question is received by the device and sent to the server. The server uses an NLP algorithm to extract the intent "return" and the entity "method," and its sentiment recognition engine determines that the user is "in distress." It then searches its database for information on "return procedures" and generates a gentle-toned response based on that information, such as "Please don't worry. We will explain the return process in detail."

[1062] Examples of input prompts for a generative AI model:

[1063] Prompt: "How do I return this item?"

[1064] Expected output: "Please rest assured. We will explain the return process in detail."

[1065] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1066] Step 1: The user enters their question in the support chat.

[1067] Specific action: The user accesses the support chat interface and enters a question such as, "How do I return this item?"

[1068] Input: User's question text

[1069] Output: Question text sent to the terminal

[1070] Step 2: The device sends the question to the server.

[1071] Specific operation: The terminal sends the question text received from the user to the server. HTTP / HTTPS is used as the communication protocol.

[1072] Input: Question text

[1073] Output: Question text sent to the server

[1074] Step 3: The server analyzes the question using a natural language processing (NLP) algorithm.

[1075] Specific operation: The server uses "nlp_library" to parse the received question text and extract the main intent and related elements (entities).

[1076] Input: Question text

[1077] Output: Intent (intent) and Entity (related element)

[1078] Step 4: The server recognizes the user's emotions.

[1079] Specific operation: The server uses the "sentiment_analysis_library" to recognize the user's emotional state from the question text, such as confusion or frustration.

[1080] Input: Question text

[1081] Output: User's emotion (e.g., troubled)

[1082] Step 5: The server searches the database for relevant information.

[1083] Specific operation: The server uses "database_connector" to search for relevant information from the database based on the analysis results. The database to be searched includes past Q&A data and product information.

[1084] Input: Intent and entity

[1085] Output: Related information (e.g., details on how to return the item)

[1086] Step 6: The server generates the response and adjusts it based on sentiment.

[1087] Specific operation: The server generates a response based on the relevant information it has acquired, and adjusts the tone and expression based on the emotions recognized by the "emotion recognition engine." For example, if the user is "distressed," the response tone will become softer.

[1088] Input: Related information, user sentiment

[1089] Output: Adjusted response text

[1090] Step 7: The server sends the generated response to the terminal.

[1091] Specific operation: The server sends the adjusted response text to the terminal.

[1092] Input: Adjusted response text

[1093] Output: Answer text sent to the terminal

[1094] Step 8: The device displays the answer to the user.

[1095] Specific operation: The terminal displays the received response text in the user interface, making it viewable by the user.

[1096] Input: Response text sent from the server

[1097] Output: Answer displayed in the user interface

[1098] 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.

[1099] 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.

[1100] 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.

[1101] [Fourth Embodiment]

[1102] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1103] 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.

[1104] 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).

[1105] 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.

[1106] 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.

[1107] 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).

[1108] 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.

[1109] 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.

[1110] 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.

[1111] 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.

[1112] 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.

[1113] 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.

[1114] 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".

[1115] This invention relates to a method and system for responding quickly and accurately to user questions in a support chat system. In particular, this system improves the efficiency of support operations by analyzing user-entered questions, searching for relevant information from multiple databases, and generating and displaying answers.

[1116] System Overview

[1117] This system includes the following main means:

[1118] 1. Means of receiving questions from users

[1119] 2. Means for analyzing question content using natural language processing algorithms

[1120] 3. Means for searching for related information based on analysis results

[1121] 4. Means of generating answers based on search results

[1122] 5. Means for displaying the generated response to the user

[1123] Program processing flow

[1124] User

[1125] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the system receives it and proceeds to the next step.

[1126] terminal

[1127] The terminal receives input from the user and sends it to the server. The submitted questions are received and analyzed on the server side.

[1128] server

[1129] The server first analyzes the content of the question using a natural language processing (NLP) algorithm. As a result of this analysis, the intent of the question (e.g., "How to use the new Crew") and the entity (e.g., "Crew") are extracted.

[1130] Next, the server searches the database for relevant information based on the extracted intent and entities. The database includes, for example, crew navigation information and past Q&A data, and this information is retrieved quickly.

[1131] Based on the searched information, the server generates appropriate answers to provide to the user. These generated answers are then formatted and presented in a user-friendly manner.

[1132] terminal

[1133] The terminal receives the response generated from the server and displays it to the user. Based on that response, the user can obtain the necessary information.

[1134] Specific example

[1135] 1. The user types "How do I use the new crew?" into the support chat.

[1136] 2. The terminal sends user input to the server.

[1137] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[1138] 4. The server searches the database based on the analysis results and retrieves relevant information.

[1139] 5. The server generates a response that explains in detail "how to use the new crew" based on the search results.

[1140] 6. The device displays the generated response to the user, allowing the user to obtain the necessary information.

[1141] By implementing this invention, responses to user inquiries can be handled efficiently and consistently, improving the quality of support services.

[1142] The following describes the processing flow.

[1143] Step 1:

[1144] The user enters their question into the support chat interface. For example, they might type, "How do I use the new crew?"

[1145] Step 2:

[1146] The terminal receives input from the user and sends the question to the server. This transmission is done via an HTTP POST request.

[1147] Step 3:

[1148] The server receives the question sent from the terminal. The received question is taken as a string and proceeds to the next processing stage.

[1149] Step 4:

[1150] The server uses a natural language processing (NLP) algorithm to analyze the question. This analysis extracts the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew").

[1151] Step 5:

[1152] The server searches for relevant information from the database based on the analysis results. First, it searches the CrewNavi database, and then it searches the past Q&A database.

[1153] Step 6:

[1154] The server integrates the search results and generates an appropriate answer to provide to the user. This answer is then formatted in a human-readable format.

[1155] Step 7:

[1156] The server sends the generated response to the terminal. The terminal displays the received response to the user.

[1157] Step 8:

[1158] Users can check the answers displayed on their devices and obtain the necessary information, thereby resolving their questions.

[1159] This series of steps enables quick and accurate responses to user inquiries.

[1160] (Example 1)

[1161] 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".

[1162] Traditional support chat systems struggled to provide quick and accurate answers to user questions. This resulted in users spending a long time obtaining appropriate information, leading to decreased support efficiency. Furthermore, manual processes were involved in database retrieval and answer generation, consuming significant human resources.

[1163] 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.

[1164] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, and means for generating appropriate answers using an AI model based on the search results. This enables the provision of quick and accurate answers to user questions, thereby improving the efficiency and quality of support operations.

[1165] "Means for receiving user questions" refers to a function that incorporates user-entered questions into the system and plays the role of sending data to the server via a communication protocol.

[1166] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a function that uses natural language processing technology to analyze questions received from users and extract the main point and important information of those questions.

[1167] "A means of searching for relevant information from a database based on analysis results" refers to a function that efficiently searches for relevant information from an appropriate database based on the analysis results obtained through natural language processing.

[1168] "A means of generating appropriate answers using an AI model based on search results" refers to a function that uses an AI model to generate the optimal answer to a user's question based on information obtained from a database.

[1169] "Means for displaying generated answers to the user" refers to a function that displays the answers generated from the server on the user's terminal, allowing the user to easily check the answers.

[1170] "Means of regularly collecting information from a database" refers to a function that periodically updates or collects information from a database to keep the system up-to-date.

[1171] "A means of searching for and integrating information from multiple databases based on analysis results" refers to a function that searches multiple databases across different databases in accordance with the results of analysis using natural language processing, integrates the obtained information, and derives a single answer.

[1172] This invention relates to a support chat system for responding quickly and accurately to user inquiries. This system improves the efficiency of support operations by analyzing user-entered questions, searching for relevant information from multiple databases, and generating and displaying answers.

[1173] This system includes the following main components:

[1174] 1. Means of receiving questions from users

[1175] 2. Means for analyzing question content using natural language processing algorithms

[1176] 3. Means for searching for related information based on analysis results

[1177] 4. Generating answers based on search results: A method for generating answers using an AI model.

[1178] 5. Means for displaying the generated response to the user

[1179] Means for receiving questions from users

[1180] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the user's device sends the question data to the server. The user's input data is stored as a string in the device's local storage or memory.

[1181] A method for analyzing question content using a natural language processing algorithm.

[1182] The server receives question data sent from the terminal and analyzes the question content using natural language processing (NLP) algorithms. This analysis uses software such as Python's NLTK library, spaCy, or TensorFlow. This analysis extracts the intent of the question (e.g., "How to use the new Crew") and entities (e.g., "Crew").

[1183] A means of searching for related information based on analysis results.

[1184] The server searches for relevant information from the database based on the analysis results obtained from natural language processing. The database used here could be, for example, MySQL, PostgreSQL, or MongoDB. Search queries are dynamically generated based on the extracted intent and entities.

[1185] A method for generating answers based on search results using an AI model.

[1186] The server uses a generative AI model (e.g., OpenAI's GPT-3) to generate user-facing responses based on information retrieved from the database. The user inputs prompts to the AI ​​model, which then generates the optimal response. Examples of specific prompts are shown below:

[1187] User: How do I use the new crew?

[1188] System prompt: The user is asking how to use a new crew. Based on the following database information, generate an answer that the user will understand.

[1189] Database information:

[1190] New crew members can be used in the next step...

[1191] Points to note include...

[1192] Answer: To start using a new crew, first go to the "Settings" menu...

[1193] A means of displaying the generated response to the user.

[1194] The device receives the response generated from the server and displays it to the user. This display uses UI components from a web browser or native application to present the response in a user-friendly format.

[1195] By implementing this invention, it becomes possible to provide quick and accurate answers to user questions, thereby improving the efficiency and quality of support operations.

[1196] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1197] Step 1:

[1198] The user accesses the support chat interface and enters their question. The user types "How do I use the new crew?" into the input field and presses the submit button. The input data is stored as a string in the user's device's local storage or memory.

[1199] Input: User-entered question: "How do I use the new crew?"

[1200] Output: Temporary storage of question data on the terminal

[1201] Step 2:

[1202] The terminal receives the query data from the user and sends it to the server in JSON format. The communication protocol used is, for example, HTTP. The terminal includes the query data in the request body and sends a POST request to a specific URL on the server (e.g., https: / / support.example.com / query).

[1203] Input: Question data entered by the user on the device.

[1204] Output: HTTP POST request to the server

[1205] Step 3:

[1206] The server receives question data sent from the terminal. The server analyzes this JSON-formatted question data and uses natural language processing algorithms to extract the intent and entities of the question. The software used here includes Python's NLTK library, spaCy, and TensorFlow. As a result of the analysis, the server obtains the main point of the question and related keywords.

[1207] Input: Question data in JSON format sent from the device.

[1208] Output: Intent and entity of the extracted question (e.g., "How to use Crew", Entity: "Crew")

[1209] Step 4:

[1210] The server searches the database for relevant information based on the intent and entities extracted through natural language processing. The server executes dynamically generated SQL queries using databases such as MySQL, PostgreSQL, or MongoDB. The server searches the database for and retrieves information related to "how to use Crew."

[1211] Input: Extracted question intent and entities

[1212] Output: Relevant information retrieved from the database (e.g., "Crew Setup Guide")

[1213] Step 5:

[1214] The server generates user-facing responses using a generative AI model (e.g., OpenAI's GPT-3) based on the acquired data. The server inputs prompts into the generative AI model to generate the optimal response. An example of a prompt might be: "The user is asking, 'How do I use the new crew?' Based on the following database information, generate an answer that is easy for the user to understand."

[1215] Input: Related information retrieved from the database, prompt text

[1216] Output: Response generated by the generation AI model (Example: "To start using a new crew, first go to the settings menu...")

[1217] Step 6:

[1218] The terminal receives the response generated from the server and displays it to the user. The response is displayed in the terminal's chat interface, allowing the user to review and use the information.

[1219] Input: Response from a generative AI model sent from the server

[1220] Output: Display of responses to the user on the terminal (specific answers to the question "How do I use the new crew?" are displayed)

[1221] (Application Example 1)

[1222] 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".

[1223] Currently, customer support functions play a crucial role on many e-commerce sites. However, responding quickly and accurately to user inquiries is challenging, especially when it comes to providing appropriate answers to detailed product-related questions. To address this issue, a system is needed that understands user questions and provides relevant information immediately.

[1224] 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.

[1225] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, means for generating appropriate answers based on the search results, means for displaying the generated answers to the user, and means for identifying questions about products and searching for relevant information from a product database. This enables users to obtain quick and appropriate answers, thereby improving customer satisfaction and streamlining support operations.

[1226] "Means for receiving user questions" refers to a mechanism for receiving questions entered by users through an electronic interface.

[1227] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a mechanism that analyzes the content of received questions based on natural language processing technology to extract the intent of the question and related entities.

[1228] "Means for searching for relevant information from a database based on analysis results" refers to a mechanism for searching for relevant information from a database using analysis results obtained by a natural language processing algorithm.

[1229] "Means for generating appropriate answers based on search results" refers to a mechanism for generating answers to be provided to users based on search results obtained from a database.

[1230] "Means for displaying generated answers to the user" refers to a mechanism for displaying generated answers on an interface used by the user.

[1231] "Means for identifying product-related questions and retrieving relevant information from a product database" refers to a mechanism for identifying that a user's question relates to a product and for retrieving relevant information from a database containing information about that product.

[1232] This invention provides a system for quickly and accurately answering user questions on an e-commerce site. The system is comprised of several main means.

[1233] First, the user accesses the support chat via a device such as a smartphone, tablet, or PC and enters their question. Once this question is sent to the server, the server receives it.

[1234] The server then uses a natural language processing (NLP) algorithm to analyze the question received from the user. During this analysis, it extracts the intent of the question (e.g., how to use a product) and the entity (e.g., a specific product). For natural language processing, the 'natural' library is used as an example.

[1235] Based on the analysis results, the server searches the database for relevant information. For product-related questions, it retrieves the necessary information from the product database. A database system such as PostgreSQL is used for this process.

[1236] Based on the search results, the server generates an appropriate answer. The generated answer is then formatted to be more user-friendly.

[1237] Finally, the generated response is sent from the server to the terminal, which then displays it to the user. This allows the user to quickly obtain the necessary information.

[1238] As a concrete example, consider a case where a user types "How do I use this product?" into the support chat of an online shopping site. This question is processed as follows:

[1239] 1. The server receives the user's question.

[1240] 2. The server analyzes the question using a natural language processing algorithm, extracting "how to use the product" as the intent and "this product" as the entity.

[1241] 3. The server searches the product database and retrieves information on how to use the relevant product.

[1242] 4. The server generates a response based on the acquired information, such as, "To use this product, first turn on the power and press the button."

[1243] 5. The terminal displays the generated response to the user.

[1244] An example of a prompt message is, "Please tell me how to use this product."

[1245] By implementing this invention, users can obtain quick and accurate answers, improving customer satisfaction on e-commerce sites. The server is equipped with various means for receiving and analyzing user questions, searching for relevant information, and generating and displaying appropriate answers. This leads to increased efficiency and improved quality in support operations.

[1246] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1247] Step 1:

[1248] Users access the support chat from their smartphone, tablet, or PC and enter their questions. The questions consist of text data, such as "How do I use this product?"

[1249] Step 2:

[1250] The terminal receives a question from the user and sends the text data to the server. Here, the data is transmitted to the server via the network. The input is the user's question text, and the output is the text data sent to the server.

[1251] Step 3:

[1252] The server receives questions from users. The received questions are parsed using natural language processing algorithms. Specifically, the 'natural' library is used for tokenization and storming to extract the intent of the question (e.g., "how to use") and entities (e.g., "product"). The input is the question text submitted by the user, and the output is the parsed intent and entity data.

[1253] Step 4:

[1254] Based on the analysis results, the server searches the product database to retrieve relevant information. It queries the PostgreSQL database to obtain information related to the relevant product. The input is the analysis results, and the output is the product information retrieved from the database.

[1255] Step 5:

[1256] The server generates appropriate answers to user questions based on information obtained from the database. The answers are generated based on previously detected intents and entities. For example, they might be generated as text including specific steps, such as, "To use this product, first turn on the power and press the button." The input is product information retrieved from the database, and the output is the generated answer text.

[1257] Step 6:

[1258] The server sends the generated response to the terminal. The response text is transmitted to the terminal via the network. The input is the generated response text, and the output is the text data sent to the terminal.

[1259] Step 7:

[1260] The terminal displays the response received from the server to the user. The displayed content is the response text generated by the server. The input is the response text sent from the server, and the output is the text data displayed to the user.

[1261] As an example of a prompt, consider "How do I use this product?". Each processing step will explain in detail how this prompt is input, parsed, the answer is generated, and the display is shown. This is the specific flow of processing a user question.

[1262] 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.

[1263] This invention provides a method and system for responding quickly and accurately to user questions in a support chat system. In particular, this system improves the efficiency of support operations by analyzing user-inputted questions, searching for relevant information from multiple databases, and generating and displaying answers. Furthermore, this invention can achieve even more advanced support by incorporating an emotion engine that recognizes user emotions and adjusts responses accordingly.

[1264] System Overview

[1265] This system includes the following main means:

[1266] 1. Means of receiving questions from users

[1267] 2. Means for analyzing question content using natural language processing algorithms

[1268] 3. Means for searching for related information based on analysis results

[1269] 4. Means of generating answers based on search results

[1270] 5. Means for displaying the generated response to the user

[1271] 6. Emotion engine that recognizes emotions from user input

[1272] 7. Means for adjusting responses based on emotion recognition results

[1273] 8. A means of selecting different response tones according to the user's emotions.

[1274] This system not only enables quick and appropriate responses to user inquiries, but also allows for flexible responses that take user emotions into consideration.

[1275] Program processing flow

[1276] User

[1277] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" Once the question is submitted, the system receives it and proceeds to the next step.

[1278] terminal

[1279] The terminal receives input from the user and sends it to the server. The submitted questions are received and analyzed on the server side.

[1280] server

[1281] The server first analyzes the content of the question using a natural language processing (NLP) algorithm. As a result of this analysis, the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew") are extracted.

[1282] Next, the server uses an emotion engine to recognize emotions from the user's input. This engine analyzes the emotional tone and meaning within the text and determines the user's emotional state based on that. For example, it might recognize "distressed" or "frustrated."

[1283] The server searches for relevant information from the database based on the analysis results and emotion recognition results. First, it searches the CrewNavi database, and then the past Q&A database. It integrates the retrieved information and generates an appropriate answer to provide to the user. This answer is adjusted in tone and content based on the user's emotions.

[1284] terminal

[1285] The terminal receives the response generated from the server and displays it to the user. The displayed response includes a flexible approach that takes the user's emotions into consideration. Based on the response, the user can obtain the necessary information.

[1286] Specific example

[1287] 1. The user types "How do I use the new crew?" into the support chat.

[1288] 2. The terminal sends user input to the server.

[1289] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[1290] 4. The server uses an emotion engine to recognize the user's emotions and, for example, determine that the user is "distressed."

[1291] 5. The server searches the database based on the analysis results and emotion recognition results and retrieves relevant information.

[1292] 6. The server generates a detailed response explaining "how to use the new crew" based on the search results. At this time, the text is adjusted to a gentle tone based on the sentiment recognition results.

[1293] 7. The device displays the generated response to the user, allowing the user to obtain the necessary information in an easy-to-understand and reassuring way.

[1294] By implementing this invention, not only will user inquiries be handled efficiently and consistently, but flexible support that takes user emotions into consideration will be provided, dramatically improving the quality of support services.

[1295] The following describes the processing flow.

[1296] Step 1:

[1297] The user enters their question into the support chat interface. For example, they might type, "How do I use the new crew?"

[1298] Step 2:

[1299] The terminal receives input from the user and sends the question to the server. This transmission is done via an HTTP POST request.

[1300] Step 3:

[1301] The server receives the question sent from the terminal. The received question is taken as a string and proceeds to the next processing stage.

[1302] Step 4:

[1303] The server uses a natural language processing (NLP) algorithm to analyze the question. This analysis extracts the intent of the question (e.g., "How to use Crew") and the entity (e.g., "Crew").

[1304] Step 5:

[1305] The server uses an emotion engine to recognize emotions from user input. This engine analyzes the emotional tone and keywords in the text to determine the user's emotional state. For example, it can extract emotions such as "troubled" or "frustrated" from the text.

[1306] Step 6:

[1307] The server searches the database for relevant information based on the analysis results and sentiment recognition results. First, it searches the CrewNavi database, and then the past Q&A database. The search query is generated based on the extracted intent and entities.

[1308] Step 7:

[1309] The server integrates the search results and generates an appropriate response to provide to the user. This response is tailored based on the user's sentiment perception. For example, if the user is judged to be "distressed," the response will be structured in a more helpful and reassuring tone.

[1310] Step 8:

[1311] The server sends the generated response to the terminal. The terminal receives it and displays it to the user.

[1312] Step 9:

[1313] Users can check the answers displayed on their devices and obtain the necessary information, thereby resolving their questions.

[1314] This series of steps enables quick and accurate responses to user questions, while also providing flexible support that takes user emotions into consideration.

[1315] (Example 2)

[1316] 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".

[1317] Conventional support chat systems often fail to respond to user inquiries quickly and appropriately. Furthermore, they struggle to provide flexible responses that take user emotions into account, resulting in a poor user experience. This invention aims to solve these problems by providing prompt and accurate answers to user inquiries while also enabling responses that consider user emotions.

[1318] 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.

[1319] In this invention, the server includes means for receiving questions from users, means for analyzing the received questions using a natural language processing algorithm to extract intent and entities, and means for searching for relevant information from a database based on the analysis results and sentiment recognition results. This makes it possible to generate and display quick and accurate answers based on the content of the questions and the user's sentiment.

[1320] 1. "Means for receiving questions from users" refers to a function that retrieves questions entered by the user within the system and passes them on to the next processing step.

[1321] 2. "Means for analyzing received questions using natural language processing algorithms to extract intent and entities" refers to a function that uses natural language processing technology to identify and extract the purpose and main elements of a question from received text.

[1322] 3. "Means for searching for relevant information from a database based on analysis results and emotion recognition results" refers to a function that searches for and retrieves relevant information from a database, taking into account the analysis results of the question content and the user's emotional state as determined by emotion recognition.

[1323] 4. "Means for integrating searched information and generating responses based on the user's emotional state" refers to a function that comprehensively processes the acquired information and creates responses with appropriate tone and content to match the user's emotional state.

[1324] 5. "Means for displaying generated answers to the user" refers to a function that displays answers created within the system on the user's screen and provides them to the user.

[1325] 6. "Means for periodically collecting information from a database" refers to a function in which the system retrieves the latest information from the database at regular intervals and always maintains up-to-date data.

[1326] 7. "Means for searching and integrating information from multiple databases based on analysis results" refers to a function that searches multiple different databases based on analysis and centralizes and integrates the obtained information.

[1327] This invention provides a method and system for responding quickly and accurately to user questions in a support chat system. This system recognizes the user's emotions and adjusts the response accordingly, thereby achieving a higher level of support.

[1328] This system is broadly composed of the following elements:

[1329] 1. Means of receiving questions from users

[1330] 2. Means for analyzing question content using natural language processing algorithms

[1331] 3. A means of searching for relevant information from a database based on the analysis results and emotion recognition results.

[1332] 4. Means for integrating retrieved information and generating responses based on the user's emotional state.

[1333] 5. Means for displaying the generated response to the user

[1334] The user accesses the support chat interface and enters their question. For example, they might enter a specific question such as, "How do I use the new crew?" When the user submits the question, the input is received on the device and sent to the server.

[1335] The terminal sends user questions as text data to the server. This text data is sent to the server using the HTTP protocol. The server receives the sent questions and proceeds to the analysis process.

[1336] The server first analyzes the content of the question using a natural language processing algorithm. Specifically, generative AI models such as Google's BERT and OpenAI's GPT-4 are available. Based on the results of this analysis, the intent and entities of the question are extracted. For example, the intent "How to use the new Crew" and the entity "Crew" are extracted.

[1337] Next, the server uses an emotion recognition algorithm to recognize the user's emotions. Tools such as IBM's Tone Analyzer are used to analyze the emotional tone and meaning within the text. This allows for the recognition of states such as "distressed" or "frustrated."

[1338] The server searches databases to retrieve relevant information based on the analysis results and sentiment recognition results. It executes SQL queries against the CrewNavi database and past Q&A databases to extract the necessary information. The retrieved information is integrated through a response generation mechanism. Here, the response generation uses models such as the GPT-4 model, and the tone and content are adjusted based on the user's sentiment.

[1339] The generated response is sent back to the device. This response is sent as text data and displayed to the user on the device. The displayed response includes flexible responses that take the user's emotions into consideration.

[1340] Specific example

[1341] 1. The user types "How do I use the new crew?" into the support chat.

[1342] 2. The terminal sends user input to the server.

[1343] 3. The server receives the question and uses a natural language processing algorithm to extract the intent "how to use Crew" and the entity "Crew".

[1344] 4. The server uses an emotion engine to recognize the user's emotions and, for example, determine that the user is "distressed."

[1345] 5. The server searches the database based on the analysis results and emotion recognition results and retrieves relevant information.

[1346] 6. The server generates a detailed response explaining "how to use the new crew" based on the search results. At this time, the text is adjusted to a gentle tone based on the sentiment recognition results.

[1347] 7. The device displays the generated response to the user, allowing the user to obtain the necessary information in an easy-to-understand and reassuring way.

[1348] Example of a prompt

[1349] By inputting prompts like the following into the AI ​​model, it is possible to obtain specific answers to user questions.

[1350] "Please explain how to use the new crew. Our users seem to be having trouble."

[1351] By implementing this invention, not only will user inquiries be handled efficiently and consistently, but flexible support that takes user emotions into consideration will be provided, dramatically improving the quality of support services.

[1352] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1353] Step 1:

[1354] The user enters a question.

[1355] Input: The user accesses the support chat interface and enters their question in text format.

[1356] Specific action: The user types "Please tell me how to use the new crew" and clicks the submit button.

[1357] Output: The entered text data is sent to the terminal.

[1358] Step 2:

[1359] The terminal sends the user's question to the server.

[1360] Input: Text data entered and submitted by the user in Step 1.

[1361] Specific action: The device sends this text data to the server as an HTTP POST request.

[1362] Output: Text data reaches the server.

[1363] Step 3:

[1364] The server receives the question.

[1365] Input: Text data sent from the device in Step 2.

[1366] Specific operation: The server receives an HTTP POST request and extracts the question text.

[1367] Output: The extracted text data is passed on to the next analysis process.

[1368] Step 4:

[1369] The server analyzes the question using a natural language processing algorithm.

[1370] Input: Question text extracted in Step 3.

[1371] Specific operation: The server uses generative AI models such as BERT and GPT-4 to analyze text and extract the intent of the question and its entities.

[1372] Output: As an analysis result, for example, the intention "how to use the crew" and the entity "crew" are extracted.

[1373] Step 5:

[1374] The server recognizes the user's emotions using an emotion engine.

[1375] Input: Question text analyzed in Step 4.

[1376] Specific operation: The server uses an emotion recognition algorithm (e.g., IBM's Tone Analyzer) to recognize the user's emotional state from the text.

[1377] Output: As an emotion recognition result, the user's emotion, such as "distressed," is determined.

[1378] Step 6:

[1379] The server searches the database for relevant information based on the analysis results and emotion recognition results.

[1380] Input: Analysis results from Step 4 and emotion recognition results from Step 5.

[1381] Specific operation: The server uses SQL queries to retrieve relevant information from the CrewNavi database and past Q&A database based on the analysis results and sentiment recognition results.

[1382] Output: Relevant information retrieved from the database.

[1383] Step 7:

[1384] The server integrates the searched information and generates responses based on the user's emotional state.

[1385] Input: Relevant information obtained in Step 6.

[1386] Specific operation: The server integrates the acquired information and uses the GPT-4 model to generate a response in a tone appropriate to the user's emotions.

[1387] Output: The final answer text to be provided to the user.

[1388] Step 8:

[1389] The server sends the generated response to the terminal.

[1390] Input: The answer text generated in Step 7.

[1391] Specific operation: The server sends the generated response text to the terminal as an HTTP POST request.

[1392] Output: The response text reaches the terminal.

[1393] Step 9:

[1394] The device displays the answer to the user.

[1395] Input: The response text sent from the server in Step 8.

[1396] Specific operation: The device displays the received response text in the chat window and provides it to the user.

[1397] Output: The user can see a gentle response such as "Here's how to use the new crew..." and obtain the necessary information.

[1398] (Application Example 2)

[1399] 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".

[1400] Customer support on e-commerce sites requires quick and appropriate responses to the content and tone of inquiries, but traditional systems struggle to provide flexible responses that take user emotions into consideration. Furthermore, there is a challenge in providing answers that are intuitively easy to understand and reassuring.

[1401] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for receiving questions from the user, means for analyzing the received questions with a natural language processing algorithm to extract intent and entities, means for searching for relevant information from a database based on the analysis results, means for generating an appropriate answer based on the search results, means for displaying the generated answer to the user, emotion recognition engine means for recognizing emotions from the user's input text, and means for adjusting the generated answer based on the user's emotions. This enables flexible and prompt customer support that takes the user's emotions into consideration, and allows for responses that give the user a sense of security.

[1402] "Means for receiving user inquiries" refers to interface devices or software that receive inquiries and questions made by users to the system.

[1403] "A means of analyzing received questions using natural language processing algorithms to extract intent and entities" refers to natural language processing technology used to analyze text entered by a user and identify the main intent and related elements (entities) within it.

[1404] "Means for searching for relevant information from a database based on analysis results" refers to a function that searches for and retrieves appropriate information from databases located on a server or network based on the results of natural language processing.

[1405] "Means for generating appropriate answers based on search results" refers to algorithms and programs that automatically create specific answers to user questions based on relevant information obtained from a database.

[1406] "Means of displaying generated answers to the user" refers to display devices or user interfaces that display generated answers in a format that the user can view.

[1407] An "emotion recognition engine that recognizes emotions from user input text" refers to an algorithm or model that analyzes and identifies the emotional state of a user at that time (for example, confusion, frustration, joy, etc.) from the text they input.

[1408] "Means for adjusting generated responses based on user emotions" refers to methods or programs for adjusting the tone and expression of responses in accordance with the user's emotions identified by the emotion recognition engine.

[1409] This invention is a system for providing customer support on an e-commerce site and includes the following components.

[1410] The system's main components are users, terminals, and servers. The following outlines how each component interacts and the specific steps taken to achieve its objectives.

[1411] 1. User

[1412] The user first accesses the support chat interface and enters their question. This question might be something like, "How do I return an item?"

[1413] 2. Terminal

[1414] The terminal receives input from the user and sends it to the server. The terminal includes a user interface, through which it displays the response to the user. The terminal is typically a communication device such as a smartphone or computer.

[1415] 3. Server

[1416] The server first analyzes the user's submitted question using a natural language processing (NLP) algorithm. This analysis extracts the question's main intent and related entities. For example, a natural language processing library such as "nlp_library" is used for this analysis. Next, the server uses the sentiment recognition engine "sentiment_analysis_library" to determine the user's emotional state from the user's input text. Based on this sentiment recognition result, the server searches the database for relevant information.

[1417] The database contains past Q&A data and product information, and is accessed using "database_connector". Based on the search results, an algorithm is activated to generate appropriate answers, and the answers are generated.

[1418] The generated responses are adjusted in tone and expression according to the user's emotions. For example, if the emotion recognition engine determines that the user is "distressed," an explanation will be provided in a gentle tone.

[1419] The terminal displays the final generated response to the user through the user interface.

[1420] Specific example

[1421] For example, a user might type "How do I return this item?" into a support chat. This question is received by the device and sent to the server. The server uses an NLP algorithm to extract the intent "return" and the entity "method," and its sentiment recognition engine determines that the user is "in distress." It then searches its database for information on "return procedures" and generates a gentle-toned response based on that information, such as "Please don't worry. We will explain the return process in detail."

[1422] Examples of input prompts for a generative AI model:

[1423] Prompt: "How do I return this item?"

[1424] Expected output: "Please rest assured. We will explain the return process in detail."

[1425] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1426] Step 1: The user enters their question in the support chat.

[1427] Specific action: The user accesses the support chat interface and enters a question such as, "How do I return this item?"

[1428] Input: User's question text

[1429] Output: Question text sent to the terminal

[1430] Step 2: The device sends the question to the server.

[1431] Specific operation: The terminal sends the question text received from the user to the server. HTTP / HTTPS is used as the communication protocol.

[1432] Input: Question text

[1433] Output: Question text sent to the server

[1434] Step 3: The server analyzes the question using a natural language processing (NLP) algorithm.

[1435] Specific operation: The server uses "nlp_library" to parse the received question text and extract the main intent and related elements (entities).

[1436] Input: Question text

[1437] Output: Intent (intent) and Entity (related element)

[1438] Step 4: The server recognizes the user's emotions.

[1439] Specific operation: The server uses the "sentiment_analysis_library" to recognize the user's emotional state from the question text, such as confusion or frustration.

[1440] Input: Question text

[1441] Output: User's emotion (e.g., troubled)

[1442] Step 5: The server searches the database for relevant information.

[1443] Specific operation: The server uses "database_connector" to search for relevant information from the database based on the analysis results. The database to be searched includes past Q&A data and product information.

[1444] Input: Intent and entity

[1445] Output: Related information (e.g., details on how to return the item)

[1446] Step 6: The server generates the response and adjusts it based on sentiment.

[1447] Specific operation: The server generates a response based on the relevant information it has acquired, and adjusts the tone and expression based on the emotions recognized by the "emotion recognition engine." For example, if the user is "distressed," the response tone will become softer.

[1448] Input: Related information, user sentiment

[1449] Output: Adjusted response text

[1450] Step 7: The server sends the generated response to the terminal.

[1451] Specific operation: The server sends the adjusted response text to the terminal.

[1452] Input: Adjusted response text

[1453] Output: Answer text sent to the terminal

[1454] Step 8: The device displays the answer to the user.

[1455] Specific operation: The terminal displays the received response text in the user interface, making it viewable by the user.

[1456] Input: Response text sent from the server

[1457] Output: Answer displayed in the user interface

[1458] 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.

[1459] 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.

[1460] 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 robot 414.

[1461] 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.

[1462] 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.

[1463] 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.

[1464] 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.

[1465] 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.

[1466] 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."

[1467] 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.

[1468] 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.

[1469] 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.

[1470] 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.

[1471] 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.

[1472] 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.

[1473] 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.

[1474] 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.

[1475] 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.

[1476] 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.

[1477] 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.

[1478] 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.

[1479] The following is further disclosed regarding the embodiments described above.

[1480] (Claim 1)

[1481] A means of receiving questions from users,

[1482] A means of analyzing received questions using natural language processing algorithms to extract intent and entities,

[1483] A means of searching for related information from a database based on the analysis results,

[1484] A means of generating appropriate answers based on search results,

[1485] A means of displaying the generated response to the user,

[1486] A system that includes this.

[1487] (Claim 2)

[1488] The system according to claim 1, further comprising means for periodically collecting information from a database.

[1489] (Claim 3)

[1490] The system according to claim 1, further comprising means for retrieving and integrating information from multiple databases based on the analysis results.

[1491] "Example 1"

[1492] (Claim 1)

[1493] A means of receiving questions from users,

[1494] A means of analyzing received questions using natural language processing algorithms to extract intent and entities,

[1495] A means of searching for related information from a database based on the analysis results,

[1496] A method for generating appropriate answers based on search results using an AI model,

[1497] A means of displaying the generated response to the user,

[1498] A system that includes this.

[1499] (Claim 2)

[1500] The system according to claim 1, further comprising means for periodically collecting information from a database.

[1501] (Claim 3)

[1502] The system according to claim 1, further comprising means for retrieving and integrating information from multiple databases based on the analysis results.

[1503] "Application Example 1"

[1504] (Claim 1)

[1505] A means of receiving questions from users,

[1506] A means of analyzing received questions using natural language processing algorithms to extract intent and entities,

[1507] A means of searching for related information from a database based on the analysis results,

[1508] A means of generating appropriate answers based on search results,

[1509] A means of displaying the generated response to the user,

[1510] A means to identify questions about a product and search for related information in the product database,

[1511] A system that includes this.

[1512] (Claim 2)

[1513] The system according to claim 1, further comprising means for periodically collecting information from a database.

[1514] (Claim 3)

[1515] The system according to claim 1, further comprising means for retrieving and integrating information from multiple databases based on the analysis results.

[1516] "Example 2 of combining an emotion engine"

[1517] (Claim 1)

[1518] A means of receiving questions from users,

[1519] A means of analyzing received questions using natural language processing algorithms to extract intent and entities,

[1520] A means of searching for relevant information from a database based on the analysis results and emotion recognition results,

[1521] A means for integrating searched information and generating responses based on the user's emotional state,

[1522] A means of displaying the generated response to the user,

[1523] A system that includes this.

[1524] (Claim 2)

[1525] [Regularly collect information from the database.]

[1526] The system according to claim 1, further comprising means.

[1527] (Claim 3)

[1528] [Search and integrate information from multiple databases based on the analysis results.]

[1529] The system according to claim 1, further comprising means.

[1530] "Application example 2 when combining with an emotional engine"

[1531] (Claim 1)

[1532] A means of receiving questions from users,

[1533] A means of analyzing received questions using natural language processing algorithms to extract intent and entities,

[1534] A means of searching for related information from a database based on the analysis results,

[1535] A means of generating appropriate answers based on search results,

[1536] A means of displaying the generated response to the user,

[1537] An emotion recognition engine means that recognizes emotions from user input text,

[1538] A means of adjusting the generated response based on the user's emotions,

[1539] A system that includes this.

[1540] (Claim 2)

[1541] The system according to claim 1, further comprising means for periodically collecting information from a database.

[1542] (Claim 3)

[1543] The system according to claim 1, further comprising means for retrieving and integrating information from multiple databases based on the analysis results. [Explanation of symbols]

[1544] 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 means of receiving questions from users, A means of analyzing received questions using natural language processing algorithms to extract intent and entities, A means of searching for related information from a database based on the analysis results, A means of generating appropriate answers based on search results, A means of displaying the generated response to the user, A system that includes this.

2. The system according to claim 1, further comprising means for periodically collecting information from a database.

3. The system according to claim 1, further comprising means for retrieving and integrating information from multiple databases based on the analysis results.

Images