system

Abstract

We provide the system. [Solution] A means for receiving user input, Means for sending the user's input to the server, The server provides means for analyzing the user's input, A means for searching for relevant information from a database based on the aforementioned analysis results, Means for generating a response based on the search results, Means for sending 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 persona chatbot control method performed by at least one processor, the method including: receiving a user utterance; adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character; encoding the prompt; and inputting the encoded prompt into a language model to generate a chatbot utterance 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] Current in-house explanatory materials contain a large amount of information, making it difficult to quickly access the necessary information. Also, when the information described in the in-house materials is inaccurate in a specific situation, irregular handling is required. As a result, there are problems such as a decrease in work efficiency and an increase in stress among employees. To solve these problems, there is a need for a system that can immediately provide accurate information in response to user questions and can also handle irregular situations.

Means for Solving the Problems

[0005] The present invention is a system comprising means for receiving user input, means for transmitting the user input to a server, means for the server to analyze the user input, means for searching for relevant information from a database based on the analysis results, means for generating a response based on the search results, and means for transmitting the generated response to the user. The system further comprises means for filtering information that matches specific conditions based on the analysis results, and further comprises means for performing inference using an AI model when the specific conditions are met. This configuration allows users to quickly access the information they need and to respond accurately to irregular situations.

[0006] A "user" is a person or entity that inputs questions or requests to a system.

[0007] A "server" is a computer or system that receives input from a user and performs processing such as analysis, database searches, and response generation.

[0008] "Means of receiving input" refers to a function that provides an interface for users to input questions or requests.

[0009] "Means for sending input" refers to a function for sending questions or requests entered by the user to the server.

[0010] "Means for analyzing input" refers to the function where a server analyzes user input using technologies such as natural language processing and extracts important keywords and context.

[0011] "Means of searching for relevant information from a database" refers to a function that searches for and retrieves relevant information within a database based on the results of analysis performed by the server.

[0012] "Means for generating responses" refers to the function by which a server generates an appropriate response to a user's question based on search results.

[0013] "Means for sending a response" refers to a function for sending the generated response to the user.

[0014] A "filtering method" is a function that selects information that matches specific conditions based on the analysis results.

[0015] "Means of inference using AI models" refers to a function that uses AI models to obtain missing information or perform inferences when certain conditions are met.

[0016] "Specific conditions" refer to pre-established regulations, rules, or judgment criteria depending on the situation. [Brief explanation of the drawing]

[0017] [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] It is a sequence diagram showing the processing flow of the data processing system in Embodiment 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 Embodiment 2 when combined with an emotion engine. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when combined with an emotion engine.

Mode for Carrying Out the Invention

[0018] 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.

[0019] First, the terms used in the following description will be explained.

[0020] In the following embodiments, a 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.

[0021] In the following embodiments, a numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0022] 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.

[0023] 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).

[0024] 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."

[0025] [First Embodiment]

[0026] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0027] 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.

[0028] 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).

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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".

[0038] This invention relates to an FAQ AI Bot system that allows users to quickly access the information they need and respond accurately to irregular situations. This system can receive user questions as input, generate appropriate answers, and provide them.

[0039] System Configuration

[0040] User's terminal

[0041] Users access the FAQ AI Bot using devices such as PCs or smartphones. These devices provide an interface for users to enter questions, typically through a text form.

[0042] server

[0043] The server forms the core of the system and has the following main functions:

[0044] 1. Analyze user input

[0045] 2. Search for relevant information in the database.

[0046] 3. Generate a response

[0047] 4. Use AI models when performing filtering or inference.

[0048] Program processing

[0049] Enter and submit your question

[0050] 1. The user enters their question on their device. Example: "Please provide me with the latest information regarding company regulations."

[0051] 2. The terminal sends this input to the server. The server receives the HTTP request and extracts the text of the question.

[0052] Question analysis

[0053] 3. The server uses an NLP engine (e.g., BERT or GPT) to analyze the question and extract important keywords and context. In this example, "company regulations" and "latest information" are the main keywords.

[0054] Searching Databases

[0055] 4. The server searches the database for relevant information based on the analysis results. The database contains information such as past FAQs, manuals, and company regulations.

[0056] Response generation

[0057] 5. Based on the information acquired by the server, an NLP model generates a natural response. In most cases, information obtained from the database is sufficient, but if it does not meet specific conditions, an AI model is used to infer a response.

[0058] 6. The server sends the generated response to the terminal. The terminal displays this response in its user interface. Example: "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments."

[0059] Specific example

[0060] Example 1: Standard Question

[0061] 1. The user enters "Please explain the leave request process" into the terminal.

[0062] 2. The device sends this question to the server.

[0063] 3. The server uses an NLP engine to extract the keywords "leave request" and "flow".

[0064] 4. The server searches the database and retrieves relevant information (e.g., steps for requesting leave).

[0065] 5. The server generates a response stating, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department."

[0066] 6. The device displays this response, and the user confirms it.

[0067] Example 2: Questions regarding handling irregular situations

[0068] 1. The user asks, "Are there any exceptions for leave requests during emergencies?"

[0069] 2. The device sends this question to the server.

[0070] 3. The server uses an NLP engine to extract the keywords "emergency," "leave request," and "exception."

[0071] 4. The server searches the database, but if it cannot find a matching entry, it performs inference using the AI ​​model.

[0072] 5. The server generates a response stating, "In emergency situations, a retroactive application may be permitted as an exception."

[0073] 6. The device displays this response, and the user confirms it.

[0074] This configuration allows users to quickly obtain the information they need and to expect accurate responses even in specific situations.

[0075] The following describes the processing flow.

[0076] Step 1:

[0077] The user uses their device to enter a question into the FAQ AI Bot. For example, they might enter, "Please tell me the latest information regarding company regulations."

[0078] Step 2:

[0079] The device sends the user's question to the server. Specifically, it formats the question as an HTTP request and sends it to the server's API endpoint.

[0080] Step 3:

[0081] The server receives a question from the terminal. It parses the received HTTP request and extracts the question text.

[0082] Step 4:

[0083] The server inputs the received question into a natural language processing (NLP) engine. Specifically, it uses models such as BERT and GPT to analyze the question and extract important keywords and context.

[0084] Step 5:

[0085] The server searches the database for relevant information based on the analysis results. Specifically, it uses SQL queries and NoSQL queries to search the database and retrieve the relevant entries.

[0086] Step 6:

[0087] The server filters the search results if necessary. Filtering is performed based on pre-configured rules and parameters to select information that matches specific criteria.

[0088] Step 7:

[0089] If the server meets certain conditions, an AI model is used to infer missing information. For example, if information about special cases related to emergencies is missing, the AI ​​model will perform the inference.

[0090] Step 8:

[0091] The server generates the final response. Based on the acquired information and inference results, an NLP model generates a natural-sounding response. For example, it might generate a response such as, "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments."

[0092] Step 9:

[0093] The server sends the generated response to the terminal. The generated response is formatted as an HTTP response and sent to the terminal.

[0094] Step 10:

[0095] The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[0096] Step 11:

[0097] The user confirms the response displayed on their device. They can review the displayed information and quickly obtain the necessary information. For example, they might confirm information such as, "The latest version of the company regulations is as follows: Revised April 2023. There are exceptions for certain departments."

[0098] (Example 1)

[0099] 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."

[0100] Traditional FAQ systems sometimes made it difficult for users to quickly and accurately access the information they needed. Furthermore, they struggled to generate appropriate responses to irregular questions. There is a need to address these challenges and provide a system that allows users to obtain the information they need more efficiently.

[0101] 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.

[0102] In this invention, the server includes means for receiving user input, means for transmitting the user input to a communication device, means for the communication device to analyze the user input using a natural language processing device, means for searching for relevant information from a storage device based on the analysis results, means for generating a response based on the search results, and means for transmitting the generated response to a user device. This allows users to quickly access the information they need, as well as to respond accurately to irregular questions.

[0103] "Means for receiving user input" refers to the interface through which users input questions and information into the FAQ AI Bot system. This is generally a text input field found on devices such as personal computers and smartphones.

[0104] "Means for transmitting user input to a communication device" refers to a function that transmits user-entered questions or information to a server or communication network. This is typically done via HTTP requests or other protocols.

[0105] "Means by which a communication device analyzes the user's input using a natural language processing device" refers to a method in which a server or a specific computer device receives user input and analyzes the text information using natural language processing technology (e.g., an NLP engine). This extracts keywords and contextual information.

[0106] "Means for searching for relevant information from a storage device based on the analysis results" refers to the process of searching for relevant information from databases and other storage devices based on keywords and contextual information obtained from the analysis results. This includes SQL queries and NoSQL searches.

[0107] "Means for generating a response based on the search results" refers to a function that generates a natural language response to the user using information obtained from a storage device. This includes cases where an NLP model (e.g., GPT or BERT) is used.

[0108] "Means for sending the generated response to the user device" refers to the process of sending the generated response to the user's terminal using a communication protocol such as an HTTP response. The user terminal receives this and displays it on its interface.

[0109] This invention relates to an FAQ AI Bot system that allows users to quickly access necessary information and respond accurately to irregular situations. This system can receive user questions as input, generate appropriate answers, and provide them.

[0110] System Configuration

[0111] This system consists of user terminals, servers, and a communication network to connect them.

[0112] User's terminal

[0113] Users access the FAQ AI Bot using devices such as PCs and smartphones. These devices provide an interface for users to enter questions. Typically, this is a text input field where the user enters their question.

[0114] server

[0115] The server forms the core of the system and has the following main functions:

[0116] 1. Analyze user input

[0117] 2. Search for relevant information in the database.

[0118] 3. Generate a response

[0119] 4. Use AI models for filtering and inference.

[0120] To implement these functions, the server uses a natural language processing (NLP) engine (e.g., BERT or GPT), a database (e.g., MySQL®, MongoDB), and a generative AI model (e.g., GPT-3®).

[0121] Program processing

[0122] The following describes the specific operation of the system.

[0123] 1. The user enters a question into the FAQ AI Bot using a device such as a PC or smartphone. For example, they might enter, "Please tell me the latest information regarding company regulations."

[0124] 2. The terminal converts this input to JSON format and sends it to the server as an HTTP request.

[0125] 3. The server analyzes the received request and processes the question text using a natural language processing engine (BERT or GPT). The analysis extracts keywords such as "company regulations" and "latest information."

[0126] 4. The server uses these keywords to search databases (SQL databases or NoSQL databases) and retrieve relevant information.

[0127] 5. The server generates a response using natural language based on the information it has obtained. This may involve using an NLP model again. For example, a response such as, "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments," might be generated.

[0128] 6. The server converts this response to JSON format and sends it to the terminal as an HTTP response.

[0129] 7. The response received by the terminal is displayed on the user interface for the user to confirm.

[0130] Specific example

[0131] Example 1: Standard Question

[0132] 1. The user enters "Please explain the leave request process" into the terminal.

[0133] 2. The device sends this question to the server.

[0134] 3. The server uses an NLP engine to extract the keywords "leave request" and "flow".

[0135] 4. The server searches the database and retrieves relevant information (e.g., steps for requesting leave).

[0136] 5. The server generates a response stating, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department."

[0137] 6. The device displays this response, and the user confirms it.

[0138] Example 2: Questions regarding handling irregular situations

[0139] 1. The user asks, "Are there any exceptions for leave requests during emergencies?"

[0140] 2. The device sends this question to the server.

[0141] 3. The server uses an NLP engine to extract the keywords "emergency," "leave request," and "exception."

[0142] 4. The server searches the database, but if it cannot find a matching entry, it performs inference using the AI ​​model.

[0143] 5. The server generates a response stating, "In emergency situations, a retroactive application may be permitted as an exception."

[0144] 6. The device displays this response, and the user confirms it.

[0145] Examples of prompt statements

[0146] When using a generative AI model, you can use prompt statements like the following:

[0147] "User question: 'Please tell me about the procedures for business trips.'"

[0148] "No relevant information was found. Please generate an appropriate response."

[0149] Thus, the present invention provides a system that allows users to quickly obtain information and to respond accurately even in specific situations.

[0150] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0151] Step 1:

[0152] Users input questions into the FAQ AI Bot using devices such as PCs or smartphones. The input is in text format; typically, a user might enter "Please provide the latest information regarding company regulations" into the text box. The input data is text.

[0153] Step 2:

[0154] The terminal receives the user's input, converts it to JSON format, and sends it to the server as an HTTP request. The input is the text entered by the user, and the output is a JSON object containing this text. The data processing performed here is the structuring of the text data.

[0155] Step 3:

[0156] The server parses the JSON from the received HTTP request and extracts the text data. The input here is the JSON object sent in step 2, and the output is the extracted raw text data. This process involves obtaining the necessary information from the HTTP request headers and body.

[0157] Step 4:

[0158] The server analyzes the extracted text using a natural language processing (NLP) engine (e.g., BERT or GPT). The input is the text data obtained in step 3, and the output is the analysis result, which includes important keywords and contextual information. Specifically, the process involves tokenizing the text and extracting important keywords.

[0159] Step 5:

[0160] The server searches the database based on the analysis results. The input here is the analyzed keywords, and the output is the relevant information retrieved from the database. Specifically, this involves creating SQL or NoSQL queries and searching the database.

[0161] Step 6:

[0162] The server generates a natural response based on information retrieved from the database. Typically, the database provides sufficient information to answer the question, but if the information is unavailable, an AI model is used to infer the answer. The input for this step is the database search results, and the output is the generated response text.

[0163] Step 7:

[0164] This function converts the server-generated response into JSON format and sends it to the terminal as an HTTP response. The input is the generated response text, and the output is a JSON object for the response. The specific actions here are converting the text to JSON and setting the response headers.

[0165] Step 8:

[0166] The terminal parses the response received from the server and displays it in the user interface. The input is a JSON object sent from the server, and the output is the response text to be displayed. This process includes parsing the JSON and updating the UI for text display.

[0167] Thus, this system seamlessly handles the entire process from user input to response generation, enabling the rapid and accurate provision of information.

[0168] (Application Example 1)

[0169] 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."

[0170] In modern brick-and-mortar stores, it is crucial to answer a wide range of customer questions quickly and accurately. However, conventional FAQ systems have been inadequate in handling specific situations and responding to voice input. This has made it difficult for store employees, especially those assisting customers, to answer customer questions in real time. To address this challenge, there is a need for a system that accurately analyzes customer voice questions and provides quick and accurate responses.

[0171] 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.

[0172] In this invention, the server includes means for receiving user input, means for transmitting the user input to the server, means for the server to analyze the user input, means for searching for relevant information from a database based on the analysis results, means for generating a response based on the search results, means for transmitting the generated response to the user, means for receiving customer questions as voice using a speech recognition device and converting them to text, means for transmitting the textened questions to the server along with the analysis results, means for the server to analyze the textened questions using a natural language processing engine and search for information from a database, and means for displaying the response on a display device. This enables faster and more accurate customer service in physical stores, and allows for appropriate responses to specific conditions and voice input.

[0173] "Means for receiving user input" refers to a device equipped with an interface for recognizing voice questions and text input from customers and incorporating them into the system.

[0174] "Means for sending to the server" refers to communication methods provided for sending user input data to a server via the internet or a local network.

[0175] "Means of analysis performed by the server" refers to algorithms and software that utilize a natural language processing engine to analyze user input on the server and extract keywords and context.

[0176] "Means for searching for relevant information from a database" refers to the means by which a server quickly searches for and retrieves appropriate information from within a database based on the analysis results.

[0177] "Means for generating responses" refers to a natural language generation engine that generates responses in a natural form in human language based on acquired information.

[0178] "Means for sending the generated response to the user" refers to communication means for returning the generated response to the user's device and displaying it.

[0179] A "speech recognition device" is a device and software that recognizes user-input speech and converts it into text.

[0180] "Means for converting to text" refers to speech recognition algorithms and software for converting speech data captured by a speech recognition device into text data.

[0181] A "natural language processing engine" is an algorithm and software that analyzes text data, extracts context and keywords, and searches for relevant information.

[0182] A "display device" is hardware such as a display that provides the user with a visual representation of the response generated on the device.

[0183] This invention is a system that allows users to ask voice questions in a physical store and generates quick and accurate responses based on those questions. This system consists of the following main elements:

[0184] Main components:

[0185] 1. Voice recognition device:

[0186] The system captures user voice questions and converts them to text. Specifically, it uses a microphone built into a wearable device such as smart glasses and speech recognition software such as Google® Speech-to-Text API.

[0187] 2. Means for receiving user input:

[0188] It accepts audio data captured through a speech recognition device as text data. It uses the input interface of smart glasses or a smartphone.

[0189] 3. Means of sending text data to the server:

[0190] Text data is sent to a server via the internet. Communication is conducted using the HTTP protocol.

[0191] 4. Means the server uses for analysis:

[0192] The server receives text data and analyzes it using a natural language processing engine (e.g., BERT, GPT). It extracts context and keywords and performs preprocessing to generate an appropriate response.

[0193] 5. Means of retrieving information from a database:

[0194] The server searches the database for relevant information based on the analysis results. The database contains information such as FAQs, product features, and usage instructions.

[0195] 6. Means for generating a response:

[0196] Based on information retrieved from a database, an automated response generation engine generates natural-sounding responses. Typically, a generative AI model (e.g., GPT-3) is used for this purpose.

[0197] 7. Means for sending the generated response to the user:

[0198] The server sends the generated response as text data to the user's terminal. The HTTP protocol is used again.

[0199] 8. Display device:

[0200] The response is displayed on the user's device. The response is displayed on the smart glasses display or the smartphone screen.

[0201] Processing examples and specific examples:

[0202] Customer Questions and Answers

[0203] Customer question: "How do I wash this jacket?"

[0204] Speech recognition: The speech recognition device converts this question into text.

[0205] Sending to the server: Send the converted text "Please tell me how to wash this jacket" to the server.

[0206] Analysis and Retrieval: The server's natural language processing engine extracts the keywords "jacket" and "washing instructions," and searches the database for related information.

[0207] Response generation: Generate the response, "Hand washing is recommended for this jacket. Please wash it in lukewarm water below 30 degrees Celsius."

[0208] Response display: The response is displayed on the smart glasses' screen, providing the customer with appropriate information.

[0209] Example of a prompt:

[0210] 1. User's question: How do I wash this jacket?

[0211] 2. Response to be generated: Hand washing is recommended for this jacket. Please wash in lukewarm water below 30 degrees Celsius.

[0212] In this way, a system can be realized that provides smooth and accurate responses to customer questions in physical stores.

[0213] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0214] Step 1:

[0215] The user performs voice input.

[0216] Specific operation: The user wears smart glasses in a physical store and asks a voice question, for example, "How do I wash this jacket?"

[0217] Input: Customer voice question.

[0218] Output: Audio data captured by the smart glasses' microphone.

[0219] Step 2:

[0220] The device converts the audio data into text.

[0221] Specific operation: The microphone built into the smart glasses picks up speech, and the Google Speech-to-Text API is used to convert the speech to text in real time.

[0222] Input: Captured audio data.

[0223] Output: Text data (e.g., "Please tell me how to wash this jacket").

[0224] Step 3:

[0225] The terminal sends text data to the server.

[0226] Specific operation: This text data is sent to the server using the HTTP protocol via the smart glasses' communication module.

[0227] Input: Text data.

[0228] Output: HTTP request sent to the server.

[0229] Step 4:

[0230] The server receives and analyzes text data.

[0231] Specific operation: The server extracts text data from the HTTP request and performs contextual analysis and keyword extraction using a natural language processing engine (e.g., BERT or GPT).

[0232] Input: Text data sent to the server.

[0233] Output: Contextual analysis and keyword extraction results (e.g., "jacket", "washing instructions").

[0234] Step 5:

[0235] The server searches the database for relevant information.

[0236] Specific operation: Search the database for relevant information such as FAQs, product descriptions, and usage instructions based on keywords.

[0237] Input: Keyword extraction results.

[0238] Output: Relevant information retrieved from the database (e.g., "Hand washing is recommended, wash with lukewarm water below 30 degrees Celsius").

[0239] Step 6:

[0240] The server generates a response based on the information it has obtained.

[0241] Specific operation: Use a natural language generation engine on the server to assemble the retrieved information into a response in a natural language format.

[0242] Input: Information retrieved from the database.

[0243] Output: Generated response (Example: "Hand washing is recommended for this jacket. Please wash it in lukewarm water below 30 degrees Celsius.").

[0244] Step 7:

[0245] The server sends the generated response to the terminal.

[0246] Specific operation: The server sends the generated response to the smart glasses as an HTTP response.

[0247] Input: The generated response.

[0248] Output: HTTP response sent from the server.

[0249] Step 8:

[0250] The terminal displays a response.

[0251] Specific operation: The terminal (smart glasses) displays the response message received from the server on the screen. The store clerk confirms the content and conveys it to the customer.

[0252] Input: The response message received from the server.

[0253] Output: The response message displayed on the smart glasses' screen.

[0254] This series of processes enables quick and accurate responses to customer voice inquiries in physical stores.

[0255] 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.

[0256] This invention relates to an FAQ AI Bot system that generates and provides responses to user questions quickly and accurately. In addition to standard functions that analyze user input and retrieve relevant information from a database to generate responses, the system incorporates an emotion engine that recognizes user emotions and adjusts responses accordingly.

[0257] System Configuration

[0258] User's terminal

[0259] Users access the FAQ AI Bot using a device (such as a PC or smartphone). The device provides an interface for users to enter questions. These questions are typically entered into a text form.

[0260] server

[0261] The server forms the core of the system and has the following main functions:

[0262] 1. Analyze user input

[0263] 2. Search for relevant information in the database.

[0264] 3. Generate a response

[0265] 4. Use AI models for filtering and inference.

[0266] 5. Use an emotion engine to recognize the user's emotions and adjust responses accordingly.

[0267] Program processing

[0268] Enter and submit your question

[0269] 1. The user enters a question on their device. For example, they might enter, "Please tell me the company's leave request process."

[0270] 2. The terminal sends this input to the server. The server receives the HTTP request and extracts the text of the question.

[0271] Question analysis

[0272] 3. The server uses a natural language processing (NLP) engine to analyze the question and extract important keywords and context. In this example, "leave request" and "flow" are the main keywords.

[0273] Searching Databases

[0274] 4. The server searches the database for relevant information based on the analysis results. The database contains information such as company regulations and FAQs.

[0275] Recognition and regulation of emotional responses

[0276] 5. The server's emotion engine recognizes the user's emotions. It recognizes whether the user is stressed, in a hurry, etc., based on the text of the question, the voice spoken during input, and the typing speed.

[0277] 6. The server adjusts its response based on emotions. For example, if the user is feeling stressed, the response will be more polite and detailed.

[0278] Response generation

[0279] 7. The server generates a final response. Based on the acquired information and the results of the emotion engine, a natural response is generated using the NLP model. For example, a response such as "For the vacation application process, please first obtain approval from your supervisor and then submit the application form to the Personnel Department. If it is an emergency, post-application may be permitted as an exception in some cases. Please confirm with the Personnel Department for details." is generated.

[0280] Transmission of the response

[0281] 8. The server transmits the generated response to the terminal. The generated response is formatted as an HTTP response and transmitted to the terminal.

[0282] Display of the response

[0283] 9. The terminal receives the response from the server and displays it on the user interface. Specifically, the received HTTP response is analyzed and the response text is displayed.

[0284] Confirmation of the user's response

[0285] 10. The user confirms the response displayed on the terminal. For example, information such as "For the vacation application process, please first obtain approval from your supervisor and then submit the application form to the Personnel Department. If it is an emergency, post-application may be permitted as an exception in some cases. Please confirm with the Personnel Department for details." is confirmed.

[0286] With this configuration, the present invention enables the user to quickly and accurately access the required information, and by further adjusting the response according to the user's emotion, it is possible to improve user satisfaction and business efficiency.

[0287] The following describes the processing flow.

[0288] Step 1:

[0289] The user uses the terminal to input a question to the FAQ AI Bot. For example, input "Please tell me the in-company vacation application process."

[0290] Step 2:

[0291] The device sends this input to the server. Specifically, it formats the question as an HTTP request and sends it to the server's API endpoint.

[0292] Step 3:

[0293] The server receives a question from the terminal. It parses the received HTTP request and extracts the question text.

[0294] Step 4:

[0295] The server uses an NLP engine to analyze the question text. Specifically, it uses natural language processing models such as BERT and GPT to extract important keywords and context. In this example, "leave request" and "flow" are the main keywords.

[0296] Step 5:

[0297] The server searches the database for relevant information based on the analysis results. Specifically, it uses SQL or NoSQL queries to search the database and retrieve the relevant entries.

[0298] Step 6:

[0299] The server generates a response based on the data it has acquired. However, at this point, a standard response is generated that does not take into account the results of the emotion engine.

[0300] Step 7:

[0301] The server's emotion engine recognizes the user's emotions. Specifically, it analyzes parameters such as emotional expressions in the question text and the user's response time to determine whether the user is stressed or in a hurry.

[0302] Step 8:

[0303] The server adjusts the response based on the results of the emotion engine. For example, if the user is feeling stressed, the response is changed to be more polite and detailed. Specifically, a response such as "For the vacation application process, first obtain approval from your supervisor and then submit the application form to the personnel department. If it is urgent, as an exception, retroactive applications may be permitted in some cases. Please confirm with the personnel department for details." is generated.

[0304] Step 9:

[0305] The server finally sends the generated response to the terminal. The response is formatted as an HTTP response and sent to the terminal.

[0306] Step 10:

[0307] The terminal receives the response from the server and displays it on the user interface. Specifically, the received HTTP response is analyzed and the response text is displayed.

[0308] Step 11:

[0309] The user checks the response displayed on the terminal. For example, information such as "For the vacation application process, first obtain approval from your supervisor and then submit the application form to the personnel department. If it is urgent, as an exception, retroactive applications may be permitted in some cases. Please confirm with the personnel department for details." is checked.

[0310] In this way, this system can generate responses quickly and accurately to the user's questions, and further improve user satisfaction by adjusting the responses based on the user's emotions.

[0311] (Example 2)

[0312] Next, Example 2 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart device 14 is referred to as the "terminal".

[0313] Conventional FAQ systems generate a fixed response to user questions, but struggle to generate appropriate responses that take into account the user's emotions and the context of the question. As a result, they fail to adequately enhance user satisfaction, and inappropriate responses may be provided, especially in situations involving urgency or stress. This invention aims to solve this problem and provide users with more appropriate and efficient responses.

[0314] 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.

[0315] In this invention, the server includes means for using a natural language processing engine to analyze user input, means for recognizing the user's emotions using an emotion engine, and means for adjusting the response content based on the user's emotions. This enables the generation of appropriate responses that take into account the user's emotions and context. Furthermore, means for sending the generated response to the user enable the provision of information quickly and accurately.

[0316] "Means for receiving user input" refers to an interface for incorporating questions and requests entered by the user through a terminal into the system.

[0317] "Means for sending user input to the server" refers to communication means for transferring data entered by the user from the terminal to the server.

[0318] "Means by which the server uses a natural language processing engine to analyze the user's input" refers to a technique for the server to analyze the user's input using natural language processing technology and extract keywords and context.

[0319] "Means for searching for relevant information from a database based on the analysis results" refers to means for querying a database and obtaining relevant information based on the analysis results obtained by the natural language processing engine.

[0320] "Methods for recognizing user emotions using an emotion engine" refer to algorithms and technologies that analyze user input and related data to detect the user's emotional state.

[0321] "Means for adjusting response content based on user emotions" refers to means for adjusting the tone and level of detail of the generated response based on the recognized emotional state of the user.

[0322] "Means for generating a final response based on the search results and adjusted response content" refers to a technology that creates a final answer to present to the user based on the acquired information and adjusted response content.

[0323] "Means for sending the generated response to the user" refers to communication means for transferring the response generated by the server to the user's terminal and displaying it.

[0324] Modes for carrying out the invention

[0325] This invention relates to an FAQ AI Bot system that generates and provides responses to user questions quickly and accurately. In addition to standard functions that analyze user input and retrieve relevant information from a database to generate responses, the system combines an emotion engine to recognize the user's emotions and adjust the response accordingly.

[0326] System Configuration

[0327] User's terminal

[0328] Users access the FAQ AI Bot using a device (such as a PC or smartphone). The device provides an interface for users to enter questions. These questions are typically entered into a text form. For example, a specific question might be, "Could you please explain the company's leave request process?"

[0329] server

[0330] The server forms the core of the system and has the following main functions:

[0331] 1. Use a natural language processing (NLP) engine to analyze user input.

[0332] Examples of software that can be used include spaCy and NLTK.

[0333] For example, "leave request" and "flow" might be extracted as keywords.

[0334] 2. Search the database for relevant information.

[0335] The databases that can be used include MySQL and PostgreSQL.

[0336] Based on the analysis results, records related to the "Leave Request Flow" are searched.

[0337] 3. Use an emotion engine to recognize the user's emotions.

[0338] The software used includes IBM Watson® and Microsoft® Azure® Text Analytics.

[0339] The system analyzes the user's input content and input speed to determine whether the user is stressed or in a hurry.

[0340] 4. Adjust responses based on the user's emotions.

[0341] For example, provide detailed and attentive responses to users who are experiencing stress.

[0342] 5. Generate the final response.

[0343] The generative AI model used is GPT-3.

[0344] For example, a response like this might be generated: "To request leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[0345] 6. Send the generated response to the terminal.

[0346] The server formats the generated response as an HTTP response and sends it to the user's terminal.

[0347] Specific example

[0348] As a concrete example, consider a case where a user enters the question, "I want to know the training schedule for new employees."

[0349] The question is sent to the server via the terminal.

[0350] The server uses an NLP engine to analyze the data and extract keywords such as "new employees," "training," and "schedule."

[0351] The system searches for relevant information in the database and uses an emotion engine to recognize the user's emotions.

[0352] Finally, the NLP model generates a response such as, "The training schedule for new employees is as follows: Day 1 is orientation, Day 2 is an introduction to each department, and Day 3 is on-the-job training," and provides it to the user.

[0353] Example of a prompt

[0354] Examples of prompt statements to input into a generative AI model (e.g., GPT-3) include the following:

[0355] A user is asking, "I want to know the training schedule for new employees." Please generate an appropriate response to this question. The response should be polite and detailed.

[0356] The introduction of this system will improve user satisfaction by allowing users to access the information they need quickly and accurately, and by providing appropriate responses that are tailored to their emotions.

[0357] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0358] The processing flow of this system's program

[0359] Step 1:

[0360] The user enters the question on the device.

[0361] Input: The user enters the question into a text box on the terminal interface.

[0362] Operation: Enter a phrase such as "Please tell me the internal leave request flow" into the input field displayed on the interface.

[0363] Output: The question text is saved to the device.

[0364] Step 2:

[0365] The terminal sends this input to the server.

[0366] Input: The question text entered by the user.

[0367] Operation: The terminal converts the user's input into an HTTP request and sends it to the server. The request is formatted in JSON format and sent as an HTTP POST request.

[0368] Output: The server receives an HTTP request.

[0369] Step 3:

[0370] The server uses a natural language processing (NLP) engine to analyze the question.

[0371] Input: Received question text.

[0372] Operation: The server uses an NLP engine (e.g., spaCy, NLTK) to parse the question text. It tokenizes the question and extracts key keywords and context.

[0373] Output: Extracted keywords (e.g., "leave request", "flow") and contextual information.

[0374] Step 4:

[0375] The server searches the database for relevant information based on the analysis results.

[0376] Input: Extracted keywords and contextual information.

[0377] Operation: The server generates SQL queries and queries the database (e.g., MySQL, PostgreSQL). Specifically, it executes queries such as "SELECT FROM faq WHERE keywords LIKE '%leave request%' AND '%flow%'".

[0378] Output: Relevant information returned from the database.

[0379] Step 5:

[0380] The server's emotion engine recognizes the user's emotions.

[0381] Input: Received question text and related input patterns (spoken audio data, typing speed, etc.).

[0382] Operation: The server analyzes the user's emotional state using an emotion engine (e.g., IBM Watson, Microsoft Azure Text Analytics). It detects emotions such as stress and urgency.

[0383] Output: User's emotional state (e.g., stressed, in a hurry).

[0384] Step 6:

[0385] The server adjusts its response based on emotion.

[0386] Input: User's emotional state and related information retrieved from the database.

[0387] Operation: The server adjusts the tone and level of detail of its responses based on the perceived emotional state. For example, it generates more polite and detailed responses for users who are feeling stressed.

[0388] Output: Information on the adjusted response.

[0389] Step 7:

[0390] The server generates the final response.

[0391] Input: Information about the adjusted response and related information.

[0392] Operation: The server uses an NLP model (e.g., GPT-3) to generate natural-sounding responses. For example, it might generate a response like, "To request leave, first obtain approval from your supervisor, then submit the application to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[0393] Output: The generated response text.

[0394] Step 8:

[0395] The server sends the generated response to the terminal.

[0396] Input: The generated response text.

[0397] Operation: The server formats the generated response as an HTTP response and sends it to the user's terminal.

[0398] Output: The terminal receives an HTTP response.

[0399] Step 9:

[0400] The terminal receives a response from the server and displays it in the user interface.

[0401] Input: HTTP response received from the server.

[0402] Operation: The device analyzes the received response, extracts the response text, and displays it in the user interface. Specifically, it displays the response "The leave request flow is..." in the chat window.

[0403] Output: Explicit response information.

[0404] Step 10:

[0405] The user confirms the response displayed on the device.

[0406] Input: Response information displayed on the terminal.

[0407] Operation: The user checks the response displayed through the terminal and obtains the necessary information. For example, they might check information such as "The leave request flow is..."

[0408] Output: Information obtained by the user.

[0409] (Application Example 2)

[0410] 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 device 14 will be referred to as the "terminal."

[0411] Traditional FAQ systems can provide standard answers to user questions, but they cannot generate responses that take user emotions into account, making it difficult to provide sufficient satisfaction, especially when users are in a hurry or feeling stressed. Furthermore, in physical stores, there is a need for efficient and effective support when store staff quickly and accurately answer customer questions. To solve these problems, an FAQ system with the ability to recognize and respond to user emotions is necessary.

[0412] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for analyzing user input, means for searching for relevant information from a database based on the analysis results, means for generating a response based on the search results, means for recognizing the user's emotions, and means for adjusting the response content based on the emotion recognition results. This makes it possible to provide a higher level of satisfaction by adjusting the response according to the user's emotions.

[0413] "Means for receiving user input" refers to devices or software that provide an interface for users to input questions or requests into a system.

[0414] "Means for sending user input to the server" refers to protocols or mechanisms for sending information entered by a user to a server via the internet or a network.

[0415] "Means by which the server analyzes the user's input" refers to a system that analyzes data such as text and audio sent by the user using natural language processing technology.

[0416] "Means for searching for relevant information from a database based on the analysis results" refers to algorithms or search engines that find appropriate information from databases within or outside the system based on the analyzed input data.

[0417] "Means for generating a response based on the search results" refers to a program or AI model that generates an appropriate answer for the user based on the searched information.

[0418] "Means for sending the generated response to the user" refers to a system that sends the generated response to the user's terminal via communication means.

[0419] "Means of recognizing user emotions" refers to emotion recognition technology that identifies a user's emotional state from input text, voice, facial expressions, etc.

[0420] "Means for adjusting the response content based on the emotion recognition result" refers to an algorithm or program that adjusts the content and tone of the generated response based on the recognized emotion information.

[0421] "Means for analyzing the user's voice input and converting it into text" refers to speech recognition technology that converts the content entered by the user via voice into text information.

[0422] "Means for recognizing user emotions from the aforementioned voice input" refers to acoustic analysis techniques and algorithms for analyzing voice data to recognize the emotions the user is experiencing.

[0423] This invention relates to an FAQ system that generates quick and accurate responses to user questions and provides responses that take emotions into consideration. This system can be implemented, in particular, as a customer support system using smart glasses for use in physical stores.

[0424] System Configuration

[0425] hardware

[0426] This system uses the following hardware:

[0427] Smart glasses: Equipped with a camera, microphone, and speaker, used by staff to communicate with customers.

[0428] Server: Forms the core of the system, handling data processing, emotion recognition, and response generation.

[0429] software

[0430] This system uses the following software:

[0431] Speech recognition software: Use the speech_recognition library to convert customer speech to text.

[0432] Emotion recognition software: Using EmotionRecognition, we recognize customer emotions from audio and images.

[0433] The NLP engine transformers' pipeline is used to perform natural language processing to generate FAQ responses.

[0434] Database access: Use the requests library to retrieve information from the database.

[0435] Process Overview

[0436] 1. The user (customer) asks questions via voice to a staff member wearing smart glasses.

[0437] 2. A microphone built into the smart glasses worn by the staff captures the audio.

[0438] 3. Speech recognition software converts the speech into text.

[0439] 4. The server analyzes the converted text and extracts important keywords.

[0440] 5. The server searches the database based on the extracted keywords and retrieves relevant information.

[0441] 6. Simultaneously, emotion recognition software recognizes the user's emotions from audio data and camera footage.

[0442] 7. The server uses the emotion recognition results to adjust the content of the response it generates.

[0443] 8. The NLP engine generates the final response and communicates it to the staff through the smart glasses' display and speaker.

[0444] 9. The staff member then responds to the customer based on that response.

[0445] Specific example

[0446] For example, if a customer asks, "What is the warranty period for this product?", the system will operate as follows:

[0447] The smart glasses capture the voice, and the voice recognition software transcribes the question into text, such as "What is the warranty period for this product?"

[0448] The server searches the database for relevant information using "product warranty period" as a keyword and retrieves the results.

[0449] Emotion recognition software recognizes emotions from a customer's tone of voice and facial expressions, and determines, for example, that they are "feeling stressed."

[0450] Based on this sentiment information, the server generates a response saying, "My apologies. I will answer your question promptly. This product has a one-year warranty."

[0451] The staff member verifies this answer through smart glasses and immediately relays it to the customer.

[0452] Example of a prompt

[0453] "The customer has submitted a question via voice. Please analyze the question appropriately, recognize the customer's emotions, and then generate a response."

[0454] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0455] Step 1:

[0456] A user asks a staff member wearing smart glasses a question using voice. A microphone built into the smart glasses captures the voice data. The user's question is then input into the system as voice data. Input: Voice data. Output: Voice data.

[0457] Step 2:

[0458] The device (smart glasses) sends audio data to speech recognition software, which then converts the audio data into text. The speech recognition software uses the speech_recognition library to analyze the audio data and convert it back into text. Input: Audio data. Output: Text data.

[0459] Step 3:

[0460] The server receives text data and performs analysis using a natural language processing engine. Specifically, it extracts important keywords and context from the text. Here, the transformers library's pipeline is used. Input: Text data. Output: Keywords and context information.

[0461] Step 4:

[0462] The server searches the database for relevant information based on the analysis results. Here, the requests library is used to search for FAQ information stored in the database. Input: Keywords and contextual information. Output: Relevant information (FAQ responses).

[0463] Step 5:

[0464] Based on the FAQ responses obtained by the server, emotion recognition software is used to recognize the user's emotions. Emotion recognition includes voice tone and speed, and facial expression analysis. The EmotionRecognition engine is used here. Input: Audio data and camera video. Output: Emotion data.

[0465] Step 6:

[0466] The server uses sentiment data to adjust FAQ responses. For example, if a user is stressed, the response will be more polite; if they are in a hurry, the response will be more concise. Input: FAQ response and sentiment data. Output: Adjusted response.

[0467] Step 7:

[0468] The server generates the final response. It uses an NLP engine (the transformers library's pipeline) to generate a refined response in a natural way. Input: Data based on the refined response. Output: Final response text.

[0469] Step 8:

[0470] The server generates a final response text and sends it to the smart glasses, which then communicates it to the staff through the smart glasses' display and speaker. Input: Final response text. Output: Response information presented to the staff.

[0471] Step 9:

[0472] Staff members view the smart glasses display, confirm the response through the speaker, and provide the appropriate answer to the customer via voice. Input: Staff confirmation. Output: Response to the customer.

[0473] 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.

[0474] 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.

[0475] 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.

[0476] [Second Embodiment]

[0477] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0478] 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.

[0479] 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).

[0480] 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.

[0481] 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.

[0482] 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).

[0483] 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.

[0484] 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.

[0485] 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.

[0486] 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.

[0487] 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.

[0488] 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".

[0489] This invention relates to an FAQ AI Bot system that allows users to quickly access the information they need and respond accurately to irregular situations. This system can receive user questions as input, generate appropriate answers, and provide them.

[0490] System Configuration

[0491] User's terminal

[0492] Users access the FAQ AI Bot using devices such as PCs or smartphones. These devices provide an interface for users to enter questions, typically through a text form.

[0493] server

[0494] The server forms the core of the system and has the following main functions:

[0495] 1. Analyze user input

[0496] 2. Search for relevant information in the database.

[0497] 3. Generate a response

[0498] 4. Use AI models when performing filtering or inference.

[0499] Program processing

[0500] Enter and submit your question

[0501] 1. The user enters their question on their device. Example: "Please provide me with the latest information regarding company regulations."

[0502] 2. The terminal sends this input to the server. The server receives the HTTP request and extracts the text of the question.

[0503] Question analysis

[0504] 3. The server uses an NLP engine (e.g., BERT or GPT) to analyze the question and extract important keywords and context. In this example, "company regulations" and "latest information" are the main keywords.

[0505] Searching Databases

[0506] 4. The server searches the database for relevant information based on the analysis results. The database contains information such as past FAQs, manuals, and company regulations.

[0507] Response generation

[0508] 5. Based on the information acquired by the server, an NLP model generates a natural response. In most cases, information obtained from the database is sufficient, but if it does not meet specific conditions, an AI model is used to infer a response.

[0509] 6. The server sends the generated response to the terminal. The terminal displays this response in its user interface. Example: "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments."

[0510] Specific example

[0511] Example 1: Standard Question

[0512] 1. The user enters "Please explain the leave request process" into the terminal.

[0513] 2. The device sends this question to the server.

[0514] 3. The server uses an NLP engine to extract the keywords "leave request" and "flow".

[0515] 4. The server searches the database and retrieves relevant information (e.g., steps for requesting leave).

[0516] 5. The server generates a response stating, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department."

[0517] 6. The device displays this response, and the user confirms it.

[0518] Example 2: Questions regarding handling irregular situations

[0519] 1. The user asks, "Are there any exceptions for leave requests during emergencies?"

[0520] 2. The device sends this question to the server.

[0521] 3. The server uses an NLP engine to extract the keywords "emergency," "leave request," and "exception."

[0522] 4. The server searches the database, but if it cannot find a matching entry, it performs inference using the AI ​​model.

[0523] 5. The server generates a response stating, "In emergency situations, a retroactive application may be permitted as an exception."

[0524] 6. The device displays this response, and the user confirms it.

[0525] This configuration allows users to quickly obtain the information they need and to expect accurate responses even in specific situations.

[0526] The following describes the processing flow.

[0527] Step 1:

[0528] The user uses their device to enter a question into the FAQ AI Bot. For example, they might enter, "Please tell me the latest information regarding company regulations."

[0529] Step 2:

[0530] The device sends the user's question to the server. Specifically, it formats the question as an HTTP request and sends it to the server's API endpoint.

[0531] Step 3:

[0532] The server receives a question from the terminal. It parses the received HTTP request and extracts the question text.

[0533] Step 4:

[0534] The server inputs the received question into a natural language processing (NLP) engine. Specifically, it uses models such as BERT and GPT to analyze the question and extract important keywords and context.

[0535] Step 5:

[0536] The server searches the database for relevant information based on the analysis results. Specifically, it uses SQL queries and NoSQL queries to search the database and retrieve the relevant entries.

[0537] Step 6:

[0538] The server filters the search results if necessary. Filtering is performed based on pre-configured rules and parameters to select information that matches specific criteria.

[0539] Step 7:

[0540] If the server meets certain conditions, an AI model is used to infer missing information. For example, if information about special cases related to emergencies is missing, the AI ​​model will perform the inference.

[0541] Step 8:

[0542] The server generates the final response. Based on the acquired information and inference results, an NLP model generates a natural-sounding response. For example, it might generate a response such as, "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments."

[0543] Step 9:

[0544] The server sends the generated response to the terminal. The generated response is formatted as an HTTP response and sent to the terminal.

[0545] Step 10:

[0546] The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[0547] Step 11:

[0548] The user confirms the response displayed on their device. They can review the displayed information and quickly obtain the necessary information. For example, they might confirm information such as, "The latest version of the company regulations is as follows: Revised April 2023. There are exceptions for certain departments."

[0549] (Example 1)

[0550] 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."

[0551] Traditional FAQ systems sometimes made it difficult for users to quickly and accurately access the information they needed. Furthermore, they struggled to generate appropriate responses to irregular questions. There is a need to address these challenges and provide a system that allows users to obtain the information they need more efficiently.

[0552] 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.

[0553] In this invention, the server includes means for receiving user input, means for transmitting the user input to a communication device, means for the communication device to analyze the user input using a natural language processing device, means for searching for relevant information from a storage device based on the analysis results, means for generating a response based on the search results, and means for transmitting the generated response to a user device. This allows users to quickly access the information they need, as well as to respond accurately to irregular questions.

[0554] "Means for receiving user input" refers to the interface through which users input questions and information into the FAQ AI Bot system. This is generally a text input field found on devices such as personal computers and smartphones.

[0555] "Means for transmitting user input to a communication device" refers to a function that transmits user-entered questions or information to a server or communication network. This is typically done via HTTP requests or other protocols.

[0556] "Means by which a communication device analyzes the user's input using a natural language processing device" refers to a method in which a server or a specific computer device receives user input and analyzes the text information using natural language processing technology (e.g., an NLP engine). This extracts keywords and contextual information.

[0557] "Means for searching for relevant information from a storage device based on the analysis results" refers to the process of searching for relevant information from databases and other storage devices based on keywords and contextual information obtained from the analysis results. This includes SQL queries and NoSQL searches.

[0558] "Means for generating a response based on the search results" refers to a function that generates a natural language response to the user using information obtained from a storage device. This includes cases where an NLP model (e.g., GPT or BERT) is used.

[0559] "Means for sending the generated response to the user device" refers to the process of sending the generated response to the user's terminal using a communication protocol such as an HTTP response. The user terminal receives this and displays it on its interface.

[0560] This invention relates to an FAQ AI Bot system that allows users to quickly access necessary information and respond accurately to irregular situations. This system can receive user questions as input, generate appropriate answers, and provide them.

[0561] System Configuration

[0562] This system consists of user terminals, servers, and a communication network to connect them.

[0563] User's terminal

[0564] Users access the FAQ AI Bot using devices such as PCs and smartphones. These devices provide an interface for users to enter questions. Typically, this is a text input field where the user enters their question.

[0565] server

[0566] The server forms the core of the system and has the following main functions:

[0567] 1. Analyze user input

[0568] 2. Search for relevant information in the database.

[0569] 3. Generate a response

[0570] 4. Use AI models for filtering and inference.

[0571] To implement these functions, the server uses a natural language processing (NLP) engine (e.g., BERT or GPT), a database (e.g., MySQL, MongoDB), and a generative AI model (e.g., GPT-3).

[0572] Program processing

[0573] The following describes the specific operation of the system.

[0574] 1. The user enters a question into the FAQ AI Bot using a device such as a PC or smartphone. For example, they might enter, "Please tell me the latest information regarding company regulations."

[0575] 2. The terminal converts this input to JSON format and sends it to the server as an HTTP request.

[0576] 3. The server analyzes the received request and processes the question text using a natural language processing engine (BERT or GPT). The analysis extracts keywords such as "company regulations" and "latest information."

[0577] 4. The server uses these keywords to search databases (SQL databases or NoSQL databases) and retrieve relevant information.

[0578] 5. The server generates a response using natural language based on the information it has obtained. This may involve using an NLP model again. For example, a response such as, "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments," might be generated.

[0579] 6. The server converts this response to JSON format and sends it to the terminal as an HTTP response.

[0580] 7. The response received by the terminal is displayed on the user interface for the user to confirm.

[0581] Specific example

[0582] Example 1: Standard Question

[0583] 1. The user enters "Please explain the leave request process" into the terminal.

[0584] 2. The device sends this question to the server.

[0585] 3. The server uses an NLP engine to extract the keywords "leave request" and "flow".

[0586] 4. The server searches the database and retrieves relevant information (e.g., steps for requesting leave).

[0587] 5. The server generates a response stating, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department."

[0588] 6. The device displays this response, and the user confirms it.

[0589] Example 2: Questions regarding handling irregular situations

[0590] 1. The user asks, "Are there any exceptions for leave requests during emergencies?"

[0591] 2. The device sends this question to the server.

[0592] 3. The server uses an NLP engine to extract the keywords "emergency," "leave request," and "exception."

[0593] 4. The server searches the database, but if it cannot find a matching entry, it performs inference using the AI ​​model.

[0594] 5. The server generates a response stating, "In emergency situations, a retroactive application may be permitted as an exception."

[0595] 6. The device displays this response, and the user confirms it.

[0596] Examples of prompt statements

[0597] When using a generative AI model, you can use prompt statements like the following:

[0598] "User question: 'Please tell me about the procedures for business trips.'"

[0599] "No relevant information was found. Please generate an appropriate response."

[0600] Thus, the present invention provides a system that allows users to quickly obtain information and to respond accurately even in specific situations.

[0601] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0602] Step 1:

[0603] Users input questions into the FAQ AI Bot using devices such as PCs or smartphones. The input is in text format; typically, a user might enter "Please provide the latest information regarding company regulations" into the text box. The input data is text.

[0604] Step 2:

[0605] The terminal receives the user's input, converts it to JSON format, and sends it to the server as an HTTP request. The input is the text entered by the user, and the output is a JSON object containing this text. The data processing performed here is the structuring of the text data.

[0606] Step 3:

[0607] The server parses the JSON from the received HTTP request and extracts the text data. The input here is the JSON object sent in step 2, and the output is the extracted raw text data. This process involves obtaining the necessary information from the HTTP request headers and body.

[0608] Step 4:

[0609] The server analyzes the extracted text using a natural language processing (NLP) engine (e.g., BERT or GPT). The input is the text data obtained in step 3, and the output is the analysis result including important keywords and contextual information. Specifically, the process involves tokenizing the text and extracting important keywords.

[0610] Step 5:

[0611] The server searches the database based on the analysis results. The input here is the analyzed keywords, and the output is the relevant information retrieved from the database. Specifically, this involves creating SQL or NoSQL queries and searching the database.

[0612] Step 6:

[0613] The server generates a natural response based on information retrieved from the database. Typically, the database provides sufficient information to answer the question; however, if the information is unavailable, an AI model is used to infer the answer. The input for this step is the database search results, and the output is the generated response text.

[0614] Step 7:

[0615] This function converts the server-generated response into JSON format and sends it to the terminal as an HTTP response. The input is the generated response text, and the output is a JSON object for the response. The specific actions here are converting the text to JSON and setting the response headers.

[0616] Step 8:

[0617] The terminal parses the response received from the server and displays it in the user interface. The input is a JSON object sent from the server, and the output is the response text to be displayed. This process includes parsing the JSON and updating the UI for text display.

[0618] Thus, this system seamlessly handles the entire process from user input to response generation, enabling the rapid and accurate provision of information.

[0619] (Application Example 1)

[0620] 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."

[0621] In modern brick-and-mortar stores, it is crucial to answer a wide range of customer questions quickly and accurately. However, conventional FAQ systems have been inadequate in handling specific situations and responding to voice input. This has made it difficult for store employees, especially those assisting customers, to answer customer questions in real time. To address this challenge, there is a need for a system that accurately analyzes customer voice questions and provides quick and accurate responses.

[0622] 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.

[0623] In this invention, the server includes means for receiving user input, means for transmitting the user input to the server, means for the server to analyze the user input, means for searching for relevant information from a database based on the analysis results, means for generating a response based on the search results, means for transmitting the generated response to the user, means for receiving customer questions as voice using a speech recognition device and converting them to text, means for transmitting the textened questions to the server along with the analysis results, means for the server to analyze the textened questions using a natural language processing engine and search for information from a database, and means for displaying the response on a display device. This enables faster and more accurate customer service in physical stores, and allows for appropriate responses to specific conditions and voice input.

[0624] "Means for receiving user input" refers to a device equipped with an interface for recognizing voice questions and text input from customers and incorporating them into the system.

[0625] "Means of sending to the server" refers to communication methods provided for sending user input data to a server via the internet or a local network.

[0626] "Means of analysis performed by the server" refers to algorithms and software that utilize a natural language processing engine to analyze user input on the server and extract keywords and context.

[0627] "Means for searching for relevant information from a database" refers to the means by which a server quickly searches for and retrieves appropriate information from within a database based on the analysis results.

[0628] "Means for generating responses" refers to a natural language generation engine that generates responses in a natural form in human language based on acquired information.

[0629] "Means for sending the generated response to the user" refers to communication means for returning the generated response to the user's device and displaying it.

[0630] A "speech recognition device" is a device and software that recognizes user-input speech and converts it into text.

[0631] "Means for converting to text" refers to speech recognition algorithms and software for converting speech data captured by a speech recognition device into text data.

[0632] A "natural language processing engine" is an algorithm and software that analyzes text data, extracts context and keywords, and searches for relevant information.

[0633] A "display device" is hardware such as a display that provides the user with a visual representation of the response generated on the device.

[0634] This invention is a system that allows users to ask voice questions in a physical store and generates quick and accurate responses based on those questions. This system consists of the following main elements:

[0635] Main components:

[0636] 1. Voice recognition device:

[0637] The system captures user voice questions and converts them to text. Specifically, it uses microphones built into wearable devices such as smart glasses and speech recognition software such as the Google Speech-to-Text API.

[0638] 2. Means for receiving user input:

[0639] It accepts audio data captured through a speech recognition device as text data. It uses the input interface of smart glasses or a smartphone.

[0640] 3. Means of sending text data to the server:

[0641] Text data is sent to a server via the internet. Communication is conducted using the HTTP protocol.

[0642] 4. Means the server uses for analysis:

[0643] The server receives text data and analyzes it using a natural language processing engine (e.g., BERT, GPT). It extracts context and keywords and performs preprocessing to generate an appropriate response.

[0644] 5. Means of retrieving information from a database:

[0645] The server searches the database for relevant information based on the analysis results. The database contains information such as FAQs, product features, and usage instructions.

[0646] 6. Means for generating a response:

[0647] Based on information retrieved from a database, an automated response generation engine generates natural-sounding responses. Typically, a generative AI model (e.g., GPT-3) is used for this purpose.

[0648] 7. Means for sending the generated response to the user:

[0649] The server sends the generated response as text data to the user's terminal. The HTTP protocol is used again.

[0650] 8. Display device:

[0651] The response is displayed on the user's device. The response is displayed on the smart glasses display or the smartphone screen.

[0652] Processing examples and specific examples:

[0653] Customer Questions and Answers

[0654] Customer question: "How do I wash this jacket?"

[0655] Speech recognition: The speech recognition device converts this question into text.

[0656] Sending to the server: Send the converted text "Please tell me how to wash this jacket" to the server.

[0657] Analysis and Retrieval: The server's natural language processing engine extracts the keywords "jacket" and "washing instructions," and searches the database for related information.

[0658] Response generation: Generate the response, "Hand washing is recommended for this jacket. Please wash it in lukewarm water below 30 degrees Celsius."

[0659] Response display: The response is displayed on the smart glasses' screen, providing the customer with appropriate information.

[0660] Example of a prompt:

[0661] 1. User's question: How do I wash this jacket?

[0662] 2. Response to be generated: Hand washing is recommended for this jacket. Please wash in lukewarm water below 30 degrees Celsius.

[0663] In this way, a system can be realized that provides smooth and accurate responses to customer questions in physical stores.

[0664] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0665] Step 1:

[0666] The user performs voice input.

[0667] Specific operation: The user wears smart glasses in a physical store and asks a voice question, for example, "How do I wash this jacket?"

[0668] Input: Customer voice question.

[0669] Output: Audio data captured by the smart glasses' microphone.

[0670] Step 2:

[0671] The device converts the audio data into text.

[0672] Specific operation: The microphone built into the smart glasses picks up speech, and the Google Speech-to-Text API is used to convert the speech to text in real time.

[0673] Input: Captured audio data.

[0674] Output: Text data (e.g., "Please tell me how to wash this jacket").

[0675] Step 3:

[0676] The terminal sends text data to the server.

[0677] Specific operation: This text data is sent to the server using the HTTP protocol via the smart glasses' communication module.

[0678] Input: Text data.

[0679] Output: HTTP request sent to the server.

[0680] Step 4:

[0681] The server receives and analyzes text data.

[0682] Specific operation: The server extracts text data from the HTTP request and performs contextual analysis and keyword extraction using a natural language processing engine (e.g., BERT or GPT).

[0683] Input: Text data sent to the server.

[0684] Output: Contextual analysis and keyword extraction results (e.g., "jacket", "washing instructions").

[0685] Step 5:

[0686] The server searches the database for relevant information.

[0687] Specific operation: Search the database for relevant information such as FAQs, product descriptions, and usage instructions based on keywords.

[0688] Input: Keyword extraction results.

[0689] Output: Relevant information retrieved from the database (e.g., "Hand washing is recommended, wash with lukewarm water below 30 degrees Celsius").

[0690] Step 6:

[0691] The server generates a response based on the information it has obtained.

[0692] Specific operation: Use a natural language generation engine on the server to assemble the retrieved information into a response in a natural language format.

[0693] Input: Information retrieved from the database.

[0694] Output: Generated response (Example: "Hand washing is recommended for this jacket. Please wash it in lukewarm water below 30 degrees Celsius.").

[0695] Step 7:

[0696] The server sends the generated response to the terminal.

[0697] Specific operation: The server sends the generated response to the smart glasses as an HTTP response.

[0698] Input: The generated response.

[0699] Output: HTTP response sent from the server.

[0700] Step 8:

[0701] The terminal displays a response.

[0702] Specific operation: The terminal (smart glasses) displays the response message received from the server on the screen. The store clerk confirms the content and conveys it to the customer.

[0703] Input: The response message received from the server.

[0704] Output: The response message displayed on the smart glasses' screen.

[0705] This series of processes enables quick and accurate responses to customer voice inquiries in physical stores.

[0706] 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.

[0707] This invention relates to an FAQ AI Bot system that generates and provides responses to user questions quickly and accurately. In addition to standard functions that analyze user input and retrieve relevant information from a database to generate responses, the system incorporates an emotion engine that recognizes user emotions and adjusts responses accordingly.

[0708] System Configuration

[0709] User's terminal

[0710] Users access the FAQ AI Bot using a device (such as a PC or smartphone). The device provides an interface for users to enter questions. These questions are typically entered into a text form.

[0711] server

[0712] The server forms the core of the system and has the following main functions:

[0713] 1. Analyze user input

[0714] 2. Search for relevant information in the database.

[0715] 3. Generate a response

[0716] 4. Use AI models for filtering and inference.

[0717] 5. Use an emotion engine to recognize the user's emotions and adjust responses accordingly.

[0718] Program processing

[0719] Enter and submit your question

[0720] 1. The user enters a question on their device. For example, they might enter, "Please tell me the company's leave request process."

[0721] 2. The terminal sends this input to the server. The server receives the HTTP request and extracts the text of the question.

[0722] Question analysis

[0723] 3. The server uses a natural language processing (NLP) engine to analyze the question and extract important keywords and context. In this example, "leave request" and "flow" are the main keywords.

[0724] Searching Databases

[0725] 4. The server searches the database for relevant information based on the analysis results. The database contains information such as company regulations and FAQs.

[0726] Recognition and regulation of emotional responses

[0727] 5. The server's emotion engine recognizes the user's emotions. It recognizes whether the user is stressed, in a hurry, etc., based on the text of the question, the voice spoken during input, and the typing speed.

[0728] 6. The server adjusts its response based on emotions. For example, if the user is stressed, the response will be more polite and detailed.

[0729] Response generation

[0730] 7. The server generates the final response. Based on the acquired information and the results of the emotion engine, the NLP model generates a natural response. For example, it might generate a response such as, "To request leave, first obtain approval from your supervisor, then submit the application to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[0731] Sending a response

[0732] 8. The server sends the generated response to the terminal. The generated response is formatted as an HTTP response and sent to the terminal.

[0733] Display of response

[0734] 9. The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[0735] User response confirmation

[0736] 10. Confirm the response displayed on the user's device. For example, confirm the information that says, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please contact the Human Resources Department for details."

[0737] This configuration allows the present invention to enable users to quickly and accurately access the information they need, and further improve user satisfaction and operational efficiency by adjusting responses according to the user's emotions.

[0738] The following describes the processing flow.

[0739] Step 1:

[0740] The user enters a question into the FAQ AI Bot using their device. For example, they might enter, "Please tell me the company's leave request process."

[0741] Step 2:

[0742] The device sends this input to the server. Specifically, it formats the question as an HTTP request and sends it to the server's API endpoint.

[0743] Step 3:

[0744] The server receives a question from the terminal. It parses the received HTTP request and extracts the question text.

[0745] Step 4:

[0746] The server analyzes the question text using an NLP engine. Specifically, it uses natural language processing models such as BERT and GPT to extract important keywords and context. In this example, "leave request" and "flow" are the main keywords.

[0747] Step 5:

[0748] The server searches the database for relevant information based on the analysis results. Specifically, it uses SQL or NoSQL queries to search the database and retrieve the relevant entries.

[0749] Step 6:

[0750] The server generates a response based on the data it has acquired. However, at this point, a standard response is generated that does not take into account the results of the emotion engine.

[0751] Step 7:

[0752] The server's emotion engine recognizes the user's emotions. Specifically, it analyzes parameters such as emotional expressions in the question text and the user's response time to determine whether the user is stressed or in a hurry.

[0753] Step 8:

[0754] The server adjusts its response based on the results of the emotion engine. For example, if the user is stressed, the response will be changed to be more polite and detailed. Specifically, it might generate a response such as, "To request leave, you must first obtain approval from your supervisor and then submit the application to HR. In urgent cases, a later application may be permitted as an exception. Please check with HR for details."

[0755] Step 9:

[0756] The server sends the final response it generates to the terminal. The response is formatted as an HTTP response and sent to the terminal.

[0757] Step 10:

[0758] The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[0759] Step 11:

[0760] The user checks the response displayed on their device. For example, they might check information such as, "To request leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[0761] In this way, this system can generate quick and accurate responses to user questions, and further improve user satisfaction by adjusting the responses based on the user's emotions.

[0762] (Example 2)

[0763] 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".

[0764] Conventional FAQ systems generate a fixed response to user questions, but struggle to generate appropriate responses that take into account the user's emotions and the context of the question. As a result, they fail to adequately enhance user satisfaction, and inappropriate responses may be provided, especially in situations involving urgency or stress. This invention aims to solve this problem and provide users with more appropriate and efficient responses.

[0765] 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.

[0766] In this invention, the server includes means for using a natural language processing engine to analyze user input, means for recognizing the user's emotions using an emotion engine, and means for adjusting the response content based on the user's emotions. This enables the generation of appropriate responses that take into account the user's emotions and context. Furthermore, means for sending the generated response to the user enable the provision of information quickly and accurately.

[0767] "Means for receiving user input" refers to an interface for incorporating questions and requests entered by the user through a terminal into the system.

[0768] "Means for sending user input to the server" refers to communication means for transferring data entered by the user from the terminal to the server.

[0769] "Means by which the server uses a natural language processing engine to analyze the user's input" refers to a technique for the server to analyze the user's input using natural language processing technology and extract keywords and context.

[0770] "Means for searching for relevant information from a database based on the analysis results" refers to means for querying a database and obtaining relevant information based on the analysis results obtained by the natural language processing engine.

[0771] "Methods for recognizing user emotions using an emotion engine" refer to algorithms and technologies that analyze user input and related data to detect the user's emotional state.

[0772] "Means for adjusting response content based on user emotions" refers to means for adjusting the tone and level of detail of the generated response based on the recognized emotional state of the user.

[0773] "Means for generating a final response based on the search results and adjusted response content" refers to a technology that creates a final answer to present to the user based on the acquired information and adjusted response content.

[0774] "Means for sending the generated response to the user" refers to communication means for transferring the response generated by the server to the user's terminal and displaying it.

[0775] Modes for carrying out the invention

[0776] This invention relates to an FAQ AI Bot system that generates and provides responses to user questions quickly and accurately. In addition to standard functions that analyze user input and retrieve relevant information from a database to generate responses, the system combines an emotion engine to recognize the user's emotions and adjust the response accordingly.

[0777] System Configuration

[0778] User's terminal

[0779] Users access the FAQ AI Bot using a device (such as a PC or smartphone). The device provides an interface for users to enter questions. These questions are typically entered into a text form. For example, a specific question might be, "Could you please explain the company's leave request process?"

[0780] server

[0781] The server forms the core of the system and has the following main functions:

[0782] 1. Use a natural language processing (NLP) engine to analyze user input.

[0783] Examples of software that can be used include spaCy and NLTK.

[0784] For example, "leave request" and "flow" might be extracted as keywords.

[0785] 2. Search the database for relevant information.

[0786] The databases that can be used include MySQL and PostgreSQL.

[0787] Based on the analysis results, records related to the "Leave Request Flow" are searched.

[0788] 3. Use an emotion engine to recognize the user's emotions.

[0789] The software used includes IBM Watson and Microsoft Azure Text Analytics.

[0790] The system analyzes the user's input content and input speed to determine whether the user is stressed or in a hurry.

[0791] 4. Adjust responses based on the user's emotions.

[0792] For example, provide detailed and attentive responses to users who are experiencing stress.

[0793] 5. Generate the final response.

[0794] The generative AI model used is GPT-3.

[0795] For example, a response like this might be generated: "To request leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[0796] 6. Send the generated response to the terminal.

[0797] The server formats the generated response as an HTTP response and sends it to the user's terminal.

[0798] Specific example

[0799] As a concrete example, consider a case where a user enters the question, "I want to know the training schedule for new employees."

[0800] The question is sent to the server via the terminal.

[0801] The server uses an NLP engine to analyze the data and extract keywords such as "new employees," "training," and "schedule."

[0802] The system searches for relevant information in the database and uses an emotion engine to recognize the user's emotions.

[0803] Finally, the NLP model generates a response such as, "The training schedule for new employees is as follows: Day 1 is orientation, Day 2 is an introduction to each department, and Day 3 is on-the-job training," and provides it to the user.

[0804] Example of a prompt

[0805] Examples of prompt statements to input into a generative AI model (e.g., GPT-3) include the following:

[0806] A user is asking, "I want to know the training schedule for new employees." Please generate an appropriate response to this question. The response should be polite and detailed.

[0807] The introduction of this system will improve user satisfaction by allowing users to access the information they need quickly and accurately, and by providing appropriate responses that are tailored to their emotions.

[0808] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0809] The processing flow of this system's program

[0810] Step 1:

[0811] The user enters the question on the device.

[0812] Input: The user enters the question into a text box on the terminal interface.

[0813] Operation: Enter a phrase such as "Please tell me the internal leave request flow" into the input field displayed on the interface.

[0814] Output: The question text is saved to the device.

[0815] Step 2:

[0816] The terminal sends this input to the server.

[0817] Input: The question text entered by the user.

[0818] Operation: The terminal converts the user's input into an HTTP request and sends it to the server. The request is formatted in JSON format and sent as an HTTP POST request.

[0819] Output: The server receives an HTTP request.

[0820] Step 3:

[0821] The server uses a natural language processing (NLP) engine to analyze the question.

[0822] Input: Received question text.

[0823] Operation: The server uses an NLP engine (e.g., spaCy, NLTK) to parse the question text. It tokenizes the question and extracts key keywords and context.

[0824] Output: Extracted keywords (e.g., "leave request", "flow") and contextual information.

[0825] Step 4:

[0826] The server searches the database for relevant information based on the analysis results.

[0827] Input: Extracted keywords and contextual information.

[0828] Operation: The server generates SQL queries and queries the database (e.g., MySQL, PostgreSQL). Specifically, it executes queries such as "SELECT FROM faq WHERE keywords LIKE '%leave request%' AND '%flow%'".

[0829] Output: Relevant information returned from the database.

[0830] Step 5:

[0831] The server's emotion engine recognizes the user's emotions.

[0832] Input: Received question text and related input patterns (spoken audio data, typing speed, etc.).

[0833] Operation: The server analyzes the user's emotional state using an emotion engine (e.g., IBM Watson, Microsoft Azure Text Analytics). It detects emotions such as stress and urgency.

[0834] Output: User's emotional state (e.g., stressed, in a hurry).

[0835] Step 6:

[0836] The server adjusts its response based on emotion.

[0837] Input: User's emotional state and related information retrieved from the database.

[0838] Operation: The server adjusts the tone and level of detail of its responses based on the perceived emotional state. For example, it generates more polite and detailed responses for users who are feeling stressed.

[0839] Output: Information on the adjusted response.

[0840] Step 7:

[0841] The server generates the final response.

[0842] Input: Information about the adjusted response and related information.

[0843] Operation: The server uses an NLP model (e.g., GPT-3) to generate natural-sounding responses. For example, it might generate a response like, "To request leave, first obtain approval from your supervisor, then submit the application to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[0844] Output: The generated response text.

[0845] Step 8:

[0846] The server sends the generated response to the terminal.

[0847] Input: The generated response text.

[0848] Operation: The server formats the generated response as an HTTP response and sends it to the user's terminal.

[0849] Output: The terminal receives an HTTP response.

[0850] Step 9:

[0851] The terminal receives a response from the server and displays it in the user interface.

[0852] Input: HTTP response received from the server.

[0853] Operation: The device analyzes the received response, extracts the response text, and displays it in the user interface. Specifically, it displays the response "The leave request flow is..." in the chat window.

[0854] Output: Explicit response information.

[0855] Step 10:

[0856] The user confirms the response displayed on the device.

[0857] Input: Response information displayed on the terminal.

[0858] Operation: The user checks the response displayed through the terminal and obtains the necessary information. For example, they might check information such as "The leave request flow is..."

[0859] Output: Information obtained by the user.

[0860] (Application Example 2)

[0861] 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."

[0862] Traditional FAQ systems can provide standard answers to user questions, but they cannot generate responses that take user emotions into account, making it difficult to provide sufficient satisfaction, especially when users are in a hurry or feeling stressed. Furthermore, in physical stores, there is a need for efficient and effective support when store staff quickly and accurately answer customer questions. To solve these problems, an FAQ system with the ability to recognize and respond to user emotions is necessary.

[0863] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for analyzing user input, means for searching for relevant information from a database based on the analysis results, means for generating a response based on the search results, means for recognizing the user's emotions, and means for adjusting the response content based on the emotion recognition results. This makes it possible to provide a higher level of satisfaction by adjusting the response according to the user's emotions.

[0864] "Means for receiving user input" refers to devices or software that provide an interface for users to input questions or requests into a system.

[0865] "Means for sending user input to the server" refers to protocols or mechanisms for sending information entered by a user to a server via the internet or a network.

[0866] "Means by which the server analyzes the user's input" refers to a system that analyzes data such as text and voice sent by the user using natural language processing technology.

[0867] "Means for searching for relevant information from a database based on the analysis results" refers to algorithms or search engines that find appropriate information from databases within or outside the system based on the analyzed input data.

[0868] "Means for generating a response based on the search results" refers to a program or AI model that generates an appropriate answer for the user based on the searched information.

[0869] "Means for sending the generated response to the user" refers to a system that sends the generated response to the user's terminal via communication means.

[0870] "Means of recognizing user emotions" refers to emotion recognition technology that identifies a user's emotional state from input text, voice, facial expressions, etc.

[0871] "Means for adjusting the response content based on the emotion recognition result" refers to an algorithm or program that adjusts the content and tone of the generated response based on the recognized emotion information.

[0872] "Means for analyzing the user's voice input and converting it into text" refers to speech recognition technology that converts the content entered by the user via voice into text information.

[0873] "Means for recognizing user emotions from the aforementioned voice input" refers to acoustic analysis techniques and algorithms for analyzing voice data to recognize the emotions the user is experiencing.

[0874] This invention relates to an FAQ system that generates quick and accurate responses to user questions and provides responses that take emotions into consideration. This system can be implemented, in particular, as a customer support system using smart glasses for use in physical stores.

[0875] System Configuration

[0876] hardware

[0877] This system uses the following hardware:

[0878] Smart glasses: Equipped with a camera, microphone, and speaker, used by staff to communicate with customers.

[0879] Server: Forms the core of the system, handling data processing, emotion recognition, and response generation.

[0880] software

[0881] This system uses the following software:

[0882] Speech recognition software: Use the speech_recognition library to convert customer speech to text.

[0883] Emotion recognition software: Using EmotionRecognition, we recognize customer emotions from voice and images.

[0884] The NLP engine transformers' pipeline is used to perform natural language processing to generate FAQ responses.

[0885] Database access: Use the requests library to retrieve information from the database.

[0886] Process Overview

[0887] 1. The user (customer) asks questions via voice to a staff member wearing smart glasses.

[0888] 2. A microphone built into the smart glasses worn by the staff captures the audio.

[0889] 3. Speech recognition software converts the speech into text.

[0890] 4. The server analyzes the converted text and extracts important keywords.

[0891] 5. The server searches the database based on the extracted keywords and retrieves relevant information.

[0892] 6. Simultaneously, emotion recognition software recognizes the user's emotions from audio data and camera footage.

[0893] 7. The server uses the emotion recognition results to adjust the content of the response it generates.

[0894] 8. The NLP engine generates the final response and communicates it to the staff through the smart glasses' display and speaker.

[0895] 9. The staff member then responds to the customer based on that response.

[0896] Specific example

[0897] For example, if a customer asks, "What is the warranty period for this product?", the system will operate as follows:

[0898] The smart glasses capture the voice, and the voice recognition software transcribes the question into text, such as "What is the warranty period for this product?"

[0899] The server searches the database for relevant information using "product warranty period" as a keyword and retrieves the results.

[0900] Emotion recognition software recognizes emotions from a customer's tone of voice and facial expressions, and determines, for example, that they are "feeling stressed."

[0901] Based on this sentiment information, the server generates a response saying, "My apologies. I will answer your question promptly. This product has a one-year warranty."

[0902] The staff member verifies this answer through smart glasses and immediately relays it to the customer.

[0903] Example of a prompt

[0904] "The customer has submitted a question via voice. Please analyze the question appropriately, recognize the customer's emotions, and then generate a response."

[0905] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0906] Step 1:

[0907] A user asks a staff member wearing smart glasses a question using voice. A microphone built into the smart glasses captures the voice data. The user's question is then input into the system as voice data. Input: Voice data. Output: Voice data.

[0908] Step 2:

[0909] The device (smart glasses) sends audio data to speech recognition software, which then converts the audio data into text. The speech recognition software uses the speech_recognition library to analyze the audio data and convert it back into text. Input: Audio data. Output: Text data.

[0910] Step 3:

[0911] The server receives text data and performs analysis using a natural language processing engine. Specifically, it extracts important keywords and context from the text. Here, the transformers library's pipeline is used. Input: Text data. Output: Keywords and context information.

[0912] Step 4:

[0913] The server searches the database for relevant information based on the analysis results. Here, the requests library is used to search for FAQ information stored in the database. Input: Keywords and contextual information. Output: Relevant information (FAQ responses).

[0914] Step 5:

[0915] Based on the FAQ responses obtained by the server, emotion recognition software is used to recognize the user's emotions. Emotion recognition includes voice tone and speed, and facial expression analysis. The EmotionRecognition engine is used here. Input: Audio data and camera video. Output: Emotion data.

[0916] Step 6:

[0917] The server uses sentiment data to adjust FAQ responses. For example, if a user is stressed, the response will be more polite; if they are in a hurry, the response will be more concise. Input: FAQ response and sentiment data. Output: Adjusted response.

[0918] Step 7:

[0919] The server generates the final response. It uses an NLP engine (the transformers library's pipeline) to generate a refined response in a natural way. Input: Data based on the refined response. Output: Final response text.

[0920] Step 8:

[0921] The server generates a final response text and sends it to the smart glasses, which then communicates it to the staff through the smart glasses' display and speaker. Input: Final response text. Output: Response information presented to the staff.

[0922] Step 9:

[0923] Staff members view the smart glasses display, confirm the response through the speaker, and provide the appropriate answer to the customer verbally. Input: Staff confirmation. Output: Response to the customer.

[0924] 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.

[0925] 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.

[0926] 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.

[0927] [Third Embodiment]

[0928] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0929] 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.

[0930] 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).

[0931] 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.

[0932] 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.

[0933] 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).

[0934] 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.

[0935] 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.

[0936] 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.

[0937] 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.

[0938] 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.

[0939] 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".

[0940] This invention relates to an FAQ AI Bot system that allows users to quickly access the information they need and respond accurately to irregular situations. This system can receive user questions as input, generate appropriate answers, and provide them.

[0941] System Configuration

[0942] User's terminal

[0943] Users access the FAQ AI Bot using devices such as PCs or smartphones. These devices provide an interface for users to enter questions, typically through a text form.

[0944] server

[0945] The server forms the core of the system and has the following main functions:

[0946] 1. Analyze user input

[0947] 2. Search for relevant information in the database.

[0948] 3. Generate a response

[0949] 4. Use AI models when performing filtering or inference.

[0950] Program processing

[0951] Enter and submit your question

[0952] 1. The user enters their question on their device. Example: "Please provide me with the latest information regarding company regulations."

[0953] 2. The terminal sends this input to the server. The server receives the HTTP request and extracts the text of the question.

[0954] Question analysis

[0955] 3. The server uses an NLP engine (e.g., BERT or GPT) to analyze the question and extract important keywords and context. In this example, "company regulations" and "latest information" are the main keywords.

[0956] Searching Databases

[0957] 4. The server searches the database for relevant information based on the analysis results. The database contains information such as past FAQs, manuals, and company regulations.

[0958] Response generation

[0959] 5. Based on the information acquired by the server, an NLP model generates a natural response. In most cases, information obtained from the database is sufficient, but if it does not meet specific conditions, an AI model is used to infer a response.

[0960] 6. The server sends the generated response to the terminal. The terminal displays this response in its user interface. Example: "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments."

[0961] Specific example

[0962] Example 1: Standard Question

[0963] 1. The user enters "Please explain the leave request process" into the terminal.

[0964] 2. The device sends this question to the server.

[0965] 3. The server uses an NLP engine to extract the keywords "leave request" and "flow".

[0966] 4. The server searches the database and retrieves relevant information (e.g., steps for requesting leave).

[0967] 5. The server generates a response stating, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department."

[0968] 6. The device displays this response, and the user confirms it.

[0969] Example 2: Questions regarding handling irregular situations

[0970] 1. The user asks, "Are there any exceptions for leave requests during emergencies?"

[0971] 2. The device sends this question to the server.

[0972] 3. The server uses an NLP engine to extract the keywords "emergency," "leave request," and "exception."

[0973] 4. The server searches the database, but if it cannot find a matching entry, it performs inference using the AI ​​model.

[0974] 5. The server generates a response stating, "In emergency situations, a retroactive application may be permitted as an exception."

[0975] 6. The device displays this response, and the user confirms it.

[0976] This configuration allows users to quickly obtain the information they need and to expect accurate responses even in specific situations.

[0977] The following describes the processing flow.

[0978] Step 1:

[0979] The user uses their device to enter a question into the FAQ AI Bot. For example, they might enter, "Please tell me the latest information regarding company regulations."

[0980] Step 2:

[0981] The device sends the user's question to the server. Specifically, it formats the question as an HTTP request and sends it to the server's API endpoint.

[0982] Step 3:

[0983] The server receives a question from the terminal. It parses the received HTTP request and extracts the question text.

[0984] Step 4:

[0985] The server inputs the received question into a natural language processing (NLP) engine. Specifically, it uses models such as BERT and GPT to analyze the question and extract important keywords and context.

[0986] Step 5:

[0987] The server searches the database for relevant information based on the analysis results. Specifically, it uses SQL queries and NoSQL queries to search the database and retrieve the relevant entries.

[0988] Step 6:

[0989] The server filters the search results if necessary. Filtering is performed based on pre-configured rules and parameters to select information that matches specific criteria.

[0990] Step 7:

[0991] If the server meets certain conditions, an AI model is used to infer missing information. For example, if information about special cases related to emergencies is missing, the AI ​​model will perform the inference.

[0992] Step 8:

[0993] The server generates the final response. Based on the acquired information and inference results, an NLP model generates a natural-sounding response. For example, it might generate a response such as, "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments."

[0994] Step 9:

[0995] The server sends the generated response to the terminal. The generated response is formatted as an HTTP response and sent to the terminal.

[0996] Step 10:

[0997] The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[0998] Step 11:

[0999] The user confirms the response displayed on their device. They can review the displayed information and quickly obtain the necessary information. For example, they might confirm information such as, "The latest version of the company regulations is as follows: Revised April 2023. There are exceptions for certain departments."

[1000] (Example 1)

[1001] 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."

[1002] Traditional FAQ systems sometimes made it difficult for users to quickly and accurately access the information they needed. Furthermore, they struggled to generate appropriate responses to irregular questions. There is a need to address these challenges and provide a system that allows users to obtain the information they need more efficiently.

[1003] 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.

[1004] In this invention, the server includes means for receiving user input, means for transmitting the user input to a communication device, means for the communication device to analyze the user input using a natural language processing device, means for searching for relevant information from a storage device based on the analysis results, means for generating a response based on the search results, and means for transmitting the generated response to a user device. This allows users to quickly access the information they need, as well as to respond accurately to irregular questions.

[1005] "Means for receiving user input" refers to the interface through which users input questions and information into the FAQ AI Bot system. This is generally a text input field found on devices such as personal computers and smartphones.

[1006] "Means for transmitting user input to a communication device" refers to a function that transmits user-entered questions or information to a server or communication network. This is typically done via HTTP requests or other protocols.

[1007] "Means by which a communication device analyzes the user's input using a natural language processing device" refers to a method in which a server or a specific computer device receives user input and analyzes the text information using natural language processing technology (e.g., an NLP engine). This extracts keywords and contextual information.

[1008] "Means for searching for relevant information from a storage device based on the analysis results" refers to the process of searching for relevant information from databases and other storage devices based on keywords and contextual information obtained from the analysis results. This includes SQL queries and NoSQL searches.

[1009] "Means for generating a response based on the search results" refers to a function that generates a natural language response to the user using information obtained from a storage device. This includes cases where an NLP model (e.g., GPT or BERT) is used.

[1010] "Means for sending the generated response to the user device" refers to the process of sending the generated response to the user's terminal using a communication protocol such as an HTTP response. The user terminal receives this and displays it on its interface.

[1011] This invention relates to an FAQ AI Bot system that allows users to quickly access necessary information and respond accurately to irregular situations. This system can receive user questions as input, generate appropriate answers, and provide them.

[1012] System Configuration

[1013] This system consists of user terminals, servers, and a communication network to connect them.

[1014] User's terminal

[1015] Users access the FAQ AI Bot using devices such as PCs and smartphones. These devices provide an interface for users to enter questions. Typically, this is a text input field where the user enters their question.

[1016] server

[1017] The server forms the core of the system and has the following main functions:

[1018] 1. Analyze user input

[1019] 2. Search for relevant information in the database.

[1020] 3. Generate a response

[1021] 4. Use AI models for filtering and inference.

[1022] To implement these functions, the server uses a natural language processing (NLP) engine (e.g., BERT or GPT), a database (e.g., MySQL, MongoDB), and a generative AI model (e.g., GPT-3).

[1023] Program processing

[1024] The following describes the specific operation of the system.

[1025] 1. The user enters a question into the FAQ AI Bot using a device such as a PC or smartphone. For example, they might enter, "Please tell me the latest information regarding company regulations."

[1026] 2. The terminal converts this input to JSON format and sends it to the server as an HTTP request.

[1027] 3. The server analyzes the received request and processes the question text using a natural language processing engine (BERT or GPT). The analysis extracts keywords such as "company regulations" and "latest information."

[1028] 4. The server uses these keywords to search databases (SQL databases or NoSQL databases) and retrieve relevant information.

[1029] 5. The server generates a response using natural language based on the information it has obtained. This may involve using an NLP model again. For example, a response such as, "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments," might be generated.

[1030] 6. The server converts this response to JSON format and sends it to the terminal as an HTTP response.

[1031] 7. The response received by the terminal is displayed on the user interface for the user to confirm.

[1032] Specific example

[1033] Example 1: Standard Question

[1034] 1. The user enters "Please explain the leave request process" into the terminal.

[1035] 2. The device sends this question to the server.

[1036] 3. The server uses an NLP engine to extract the keywords "leave request" and "flow".

[1037] 4. The server searches the database and retrieves relevant information (e.g., steps for requesting leave).

[1038] 5. The server generates a response stating, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department."

[1039] 6. The device displays this response, and the user confirms it.

[1040] Example 2: Questions regarding handling irregular situations

[1041] 1. The user asks, "Are there any exceptions for leave requests during emergencies?"

[1042] 2. The device sends this question to the server.

[1043] 3. The server uses an NLP engine to extract the keywords "emergency," "leave request," and "exception."

[1044] 4. The server searches the database, but if it cannot find a matching entry, it performs inference using the AI ​​model.

[1045] 5. The server generates a response stating, "In emergency situations, a retroactive application may be permitted as an exception."

[1046] 6. The device displays this response, and the user confirms it.

[1047] Examples of prompt statements

[1048] When using a generative AI model, you can use prompt statements like the following:

[1049] "User question: 'Please tell me about the procedures for business trips.'"

[1050] "No relevant information was found. Please generate an appropriate response."

[1051] Thus, the present invention provides a system that allows users to quickly obtain information and to respond accurately even in specific situations.

[1052] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1053] Step 1:

[1054] Users input questions into the FAQ AI Bot using devices such as PCs or smartphones. The input is in text format; typically, a user might enter "Please provide the latest information regarding company regulations" into the text box. The input data is text.

[1055] Step 2:

[1056] The terminal receives the user's input, converts it to JSON format, and sends it to the server as an HTTP request. The input is the text entered by the user, and the output is a JSON object containing this text. The data processing performed here is the structuring of the text data.

[1057] Step 3:

[1058] The server parses the JSON from the received HTTP request and extracts the text data. The input here is the JSON object sent in step 2, and the output is the extracted raw text data. This process involves obtaining the necessary information from the HTTP request headers and body.

[1059] Step 4:

[1060] The server analyzes the extracted text using a natural language processing (NLP) engine (e.g., BERT or GPT). The input is the text data obtained in step 3, and the output is the analysis result including important keywords and contextual information. Specifically, the process involves tokenizing the text and extracting important keywords.

[1061] Step 5:

[1062] The server searches the database based on the analysis results. The input here is the analyzed keywords, and the output is the relevant information retrieved from the database. Specifically, this involves creating SQL or NoSQL queries and searching the database.

[1063] Step 6:

[1064] The server generates a natural response based on information retrieved from the database. Typically, the database provides sufficient information to answer the question; however, if the information is unavailable, an AI model is used to infer the answer. The input for this step is the database search results, and the output is the generated response text.

[1065] Step 7:

[1066] This function converts the server-generated response into JSON format and sends it to the terminal as an HTTP response. The input is the generated response text, and the output is a JSON object for the response. The specific actions here are converting the text to JSON and setting the response headers.

[1067] Step 8:

[1068] The terminal parses the response received from the server and displays it in the user interface. The input is a JSON object sent from the server, and the output is the response text to be displayed. This process includes parsing the JSON and updating the UI for text display.

[1069] Thus, this system seamlessly handles the entire process from user input to response generation, enabling the rapid and accurate provision of information.

[1070] (Application Example 1)

[1071] 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."

[1072] In modern brick-and-mortar stores, it is crucial to answer a wide range of customer questions quickly and accurately. However, conventional FAQ systems have been inadequate in handling specific situations and responding to voice input. This has made it difficult for store employees, especially those assisting customers, to answer customer questions in real time. To address this challenge, there is a need for a system that accurately analyzes customer voice questions and provides quick and accurate responses.

[1073] 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.

[1074] In this invention, the server includes means for receiving user input, means for transmitting the user input to the server, means for the server to analyze the user input, means for searching for relevant information from a database based on the analysis results, means for generating a response based on the search results, means for transmitting the generated response to the user, means for receiving customer questions as voice using a speech recognition device and converting them to text, means for transmitting the textened questions to the server along with the analysis results, means for the server to analyze the textened questions using a natural language processing engine and search for information from a database, and means for displaying the response on a display device. This enables faster and more accurate customer service in physical stores, and allows for appropriate responses to specific conditions and voice input.

[1075] "Means for receiving user input" refers to a device equipped with an interface for recognizing voice questions and text input from customers and incorporating them into the system.

[1076] "Means of sending to the server" refers to communication methods provided for sending user input data to a server via the internet or a local network.

[1077] "Means of analysis performed by the server" refers to algorithms and software that utilize a natural language processing engine to analyze user input on the server and extract keywords and context.

[1078] "Means for searching for relevant information from a database" refers to the means by which a server quickly searches for and retrieves appropriate information from within a database based on the analysis results.

[1079] "Means for generating responses" refers to a natural language generation engine that generates responses in a natural form in human language based on acquired information.

[1080] "Means for sending the generated response to the user" refers to communication means for returning the generated response to the user's device and displaying it.

[1081] A "speech recognition device" is a device and software that recognizes user-input speech and converts it into text.

[1082] "Means for converting to text" refers to speech recognition algorithms and software for converting speech data captured by a speech recognition device into text data.

[1083] A "natural language processing engine" is an algorithm and software that analyzes text data, extracts context and keywords, and searches for relevant information.

[1084] A "display device" is hardware such as a display that provides the user with a visual representation of the response generated on the device.

[1085] This invention is a system that allows users to ask voice questions in a physical store and generates quick and accurate responses based on those questions. This system consists of the following main elements:

[1086] Main components:

[1087] 1. Voice recognition device:

[1088] The system captures user voice questions and converts them to text. Specifically, it uses microphones built into wearable devices such as smart glasses and speech recognition software such as the Google Speech-to-Text API.

[1089] 2. Means for receiving user input:

[1090] It accepts audio data captured through a speech recognition device as text data. It uses the input interface of smart glasses or a smartphone.

[1091] 3. Means of sending text data to the server:

[1092] Text data is sent to a server via the internet. Communication is conducted using the HTTP protocol.

[1093] 4. Means the server uses for analysis:

[1094] The server receives text data and analyzes it using a natural language processing engine (e.g., BERT, GPT). It extracts context and keywords and performs preprocessing to generate an appropriate response.

[1095] 5. Means of retrieving information from a database:

[1096] The server searches the database for relevant information based on the analysis results. The database contains information such as FAQs, product features, and usage instructions.

[1097] 6. Means for generating a response:

[1098] Based on information retrieved from a database, an automated response generation engine generates natural-sounding responses. Typically, a generative AI model (e.g., GPT-3) is used for this purpose.

[1099] 7. Means for sending the generated response to the user:

[1100] The server sends the generated response as text data to the user's terminal. The HTTP protocol is used again.

[1101] 8. Display device:

[1102] The response is displayed on the user's device. The response is displayed on the smart glasses display or the smartphone screen.

[1103] Processing examples and specific examples:

[1104] Customer Questions and Answers

[1105] Customer question: "How do I wash this jacket?"

[1106] Speech recognition: The speech recognition device converts this question into text.

[1107] Sending to the server: Send the converted text "Please tell me how to wash this jacket" to the server.

[1108] Analysis and Retrieval: The server's natural language processing engine extracts the keywords "jacket" and "washing instructions," and searches the database for related information.

[1109] Response generation: Generate the response, "Hand washing is recommended for this jacket. Please wash it in lukewarm water below 30 degrees Celsius."

[1110] Response display: The response is displayed on the smart glasses' screen, providing the customer with appropriate information.

[1111] Example of a prompt:

[1112] 1. User's question: How do I wash this jacket?

[1113] 2. Response to be generated: Hand washing is recommended for this jacket. Please wash in lukewarm water below 30 degrees Celsius.

[1114] In this way, a system can be realized that provides smooth and accurate responses to customer questions in physical stores.

[1115] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1116] Step 1:

[1117] The user performs voice input.

[1118] Specific operation: The user wears smart glasses in a physical store and asks a voice question, for example, "How do I wash this jacket?"

[1119] Input: Customer voice question.

[1120] Output: Audio data captured by the smart glasses' microphone.

[1121] Step 2:

[1122] The device converts the audio data into text.

[1123] Specific operation: The microphone built into the smart glasses picks up speech, and the Google Speech-to-Text API is used to convert the speech to text in real time.

[1124] Input: Captured audio data.

[1125] Output: Text data (e.g., "Please tell me how to wash this jacket").

[1126] Step 3:

[1127] The terminal sends text data to the server.

[1128] Specific operation: This text data is sent to the server using the HTTP protocol via the smart glasses' communication module.

[1129] Input: Text data.

[1130] Output: HTTP request sent to the server.

[1131] Step 4:

[1132] The server receives and analyzes text data.

[1133] Specific operation: The server extracts text data from the HTTP request and performs contextual analysis and keyword extraction using a natural language processing engine (e.g., BERT or GPT).

[1134] Input: Text data sent to the server.

[1135] Output: Contextual analysis and keyword extraction results (e.g., "jacket", "washing instructions").

[1136] Step 5:

[1137] The server searches the database for relevant information.

[1138] Specific operation: Search the database for relevant information such as FAQs, product descriptions, and usage instructions based on keywords.

[1139] Input: Keyword extraction results.

[1140] Output: Relevant information retrieved from the database (e.g., "Hand washing is recommended, wash with lukewarm water below 30 degrees Celsius").

[1141] Step 6:

[1142] The server generates a response based on the information it has obtained.

[1143] Specific operation: Use a natural language generation engine on the server to assemble the retrieved information into a response in a natural language format.

[1144] Input: Information retrieved from the database.

[1145] Output: Generated response (Example: "Hand washing is recommended for this jacket. Please wash it in lukewarm water below 30 degrees Celsius.").

[1146] Step 7:

[1147] The server sends the generated response to the terminal.

[1148] Specific operation: The server sends the generated response to the smart glasses as an HTTP response.

[1149] Input: The generated response.

[1150] Output: HTTP response sent from the server.

[1151] Step 8:

[1152] The terminal displays a response.

[1153] Specific operation: The terminal (smart glasses) displays the response message received from the server on the screen. The store clerk confirms the content and conveys it to the customer.

[1154] Input: The response message received from the server.

[1155] Output: The response message displayed on the smart glasses' screen.

[1156] This series of processes enables quick and accurate responses to customer voice inquiries in physical stores.

[1157] 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.

[1158] This invention relates to an FAQ AI Bot system that generates and provides responses to user questions quickly and accurately. In addition to standard functions that analyze user input and retrieve relevant information from a database to generate responses, the system incorporates an emotion engine that recognizes user emotions and adjusts responses accordingly.

[1159] System Configuration

[1160] User's terminal

[1161] Users access the FAQ AI Bot using a device (such as a PC or smartphone). The device provides an interface for users to enter questions. These questions are typically entered into a text form.

[1162] server

[1163] The server forms the core of the system and has the following main functions:

[1164] 1. Analyze user input

[1165] 2. Search for relevant information in the database.

[1166] 3. Generate a response

[1167] 4. Use AI models for filtering and inference.

[1168] 5. Use an emotion engine to recognize the user's emotions and adjust responses accordingly.

[1169] Program processing

[1170] Enter and submit your question

[1171] 1. The user enters a question on their device. For example, they might enter, "Please tell me the company's leave request process."

[1172] 2. The terminal sends this input to the server. The server receives the HTTP request and extracts the text of the question.

[1173] Question analysis

[1174] 3. The server uses a natural language processing (NLP) engine to analyze the question and extract important keywords and context. In this example, "leave request" and "flow" are the main keywords.

[1175] Searching Databases

[1176] 4. The server searches the database for relevant information based on the analysis results. The database contains information such as company regulations and FAQs.

[1177] Recognition and regulation of emotional responses

[1178] 5. The server's emotion engine recognizes the user's emotions. It recognizes whether the user is stressed, in a hurry, etc., based on the text of the question, the voice spoken during input, and the typing speed.

[1179] 6. The server adjusts its response based on emotions. For example, if the user is stressed, the response will be more polite and detailed.

[1180] Response generation

[1181] 7. The server generates the final response. Based on the acquired information and the results of the emotion engine, the NLP model generates a natural response. For example, it might generate a response such as, "To request leave, first obtain approval from your supervisor, then submit the application to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[1182] Sending a response

[1183] 8. The server sends the generated response to the terminal. The generated response is formatted as an HTTP response and sent to the terminal.

[1184] Display of response

[1185] 9. The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[1186] User response confirmation

[1187] 10. Confirm the response displayed on the user's device. For example, confirm the information that says, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please contact the Human Resources Department for details."

[1188] This configuration allows the present invention to enable users to quickly and accurately access the information they need, and further improve user satisfaction and operational efficiency by adjusting responses according to the user's emotions.

[1189] The following describes the processing flow.

[1190] Step 1:

[1191] The user enters a question into the FAQ AI Bot using their device. For example, they might enter, "Please tell me the company's leave request process."

[1192] Step 2:

[1193] The device sends this input to the server. Specifically, it formats the question as an HTTP request and sends it to the server's API endpoint.

[1194] Step 3:

[1195] The server receives a question from the terminal. It parses the received HTTP request and extracts the question text.

[1196] Step 4:

[1197] The server analyzes the question text using an NLP engine. Specifically, it uses natural language processing models such as BERT and GPT to extract important keywords and context. In this example, "leave request" and "flow" are the main keywords.

[1198] Step 5:

[1199] The server searches the database for relevant information based on the analysis results. Specifically, it uses SQL or NoSQL queries to search the database and retrieve the relevant entries.

[1200] Step 6:

[1201] The server generates a response based on the data it has acquired. However, at this point, a standard response is generated that does not take into account the results of the emotion engine.

[1202] Step 7:

[1203] The server's emotion engine recognizes the user's emotions. Specifically, it analyzes parameters such as emotional expressions in the question text and the user's response time to determine whether the user is stressed or in a hurry.

[1204] Step 8:

[1205] The server adjusts its response based on the results of the emotion engine. For example, if the user is stressed, the response will be changed to be more polite and detailed. Specifically, it might generate a response such as, "To request leave, you must first obtain approval from your supervisor and then submit the application to HR. In urgent cases, a later application may be permitted as an exception. Please check with HR for details."

[1206] Step 9:

[1207] The server sends the final response it generates to the terminal. The response is formatted as an HTTP response and sent to the terminal.

[1208] Step 10:

[1209] The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[1210] Step 11:

[1211] The user checks the response displayed on their device. For example, they might check information such as, "To request leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[1212] In this way, this system can generate quick and accurate responses to user questions, and further improve user satisfaction by adjusting the responses based on the user's emotions.

[1213] (Example 2)

[1214] 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."

[1215] Conventional FAQ systems generate a fixed response to user questions, but struggle to generate appropriate responses that take into account the user's emotions and the context of the question. As a result, they fail to adequately enhance user satisfaction, and inappropriate responses may be provided, especially in situations involving urgency or stress. This invention aims to solve this problem and provide users with more appropriate and efficient responses.

[1216] 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.

[1217] In this invention, the server includes means for using a natural language processing engine to analyze user input, means for recognizing the user's emotions using an emotion engine, and means for adjusting the response content based on the user's emotions. This enables the generation of appropriate responses that take into account the user's emotions and context. Furthermore, means for sending the generated response to the user enable the provision of information quickly and accurately.

[1218] "Means for receiving user input" refers to an interface for incorporating questions and requests entered by the user through a terminal into the system.

[1219] "Means for sending user input to the server" refers to communication means for transferring data entered by the user from the terminal to the server.

[1220] "Means by which the server uses a natural language processing engine to analyze the user's input" refers to a technique for the server to analyze the user's input using natural language processing technology and extract keywords and context.

[1221] "Means for searching for relevant information from a database based on the analysis results" refers to means for querying a database and obtaining relevant information based on the analysis results obtained by the natural language processing engine.

[1222] "Methods for recognizing user emotions using an emotion engine" refer to algorithms and technologies that analyze user input and related data to detect the user's emotional state.

[1223] "Means for adjusting response content based on user emotions" refers to means for adjusting the tone and level of detail of the generated response based on the recognized emotional state of the user.

[1224] "Means for generating a final response based on the search results and adjusted response content" refers to a technology that creates a final answer to present to the user based on the acquired information and adjusted response content.

[1225] "Means for sending the generated response to the user" refers to communication means for transferring the response generated by the server to the user's terminal and displaying it.

[1226] Modes for carrying out the invention

[1227] This invention relates to an FAQ AI Bot system that generates and provides responses to user questions quickly and accurately. In addition to standard functions that analyze user input and retrieve relevant information from a database to generate responses, the system combines an emotion engine to recognize the user's emotions and adjust the response accordingly.

[1228] System Configuration

[1229] User's terminal

[1230] Users access the FAQ AI Bot using a device (such as a PC or smartphone). The device provides an interface for users to enter questions. These questions are typically entered into a text form. For example, a specific question might be, "Could you please explain the company's leave request process?"

[1231] server

[1232] The server forms the core of the system and has the following main functions:

[1233] 1. Use a natural language processing (NLP) engine to analyze user input.

[1234] Examples of software that can be used include spaCy and NLTK.

[1235] For example, "leave request" and "flow" might be extracted as keywords.

[1236] 2. Search the database for relevant information.

[1237] The databases that can be used include MySQL and PostgreSQL.

[1238] Based on the analysis results, records related to the "Leave Request Flow" are searched.

[1239] 3. Use an emotion engine to recognize the user's emotions.

[1240] The software used includes IBM Watson and Microsoft Azure Text Analytics.

[1241] The system analyzes the user's input content and input speed to determine whether the user is stressed or in a hurry.

[1242] 4. Adjust responses based on the user's emotions.

[1243] For example, provide detailed and attentive responses to users who are experiencing stress.

[1244] 5. Generate the final response.

[1245] The generative AI model used is GPT-3.

[1246] For example, a response like this might be generated: "To request leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[1247] 6. Send the generated response to the terminal.

[1248] The server formats the generated response as an HTTP response and sends it to the user's terminal.

[1249] Specific example

[1250] As a concrete example, consider a case where a user enters the question, "I want to know the training schedule for new employees."

[1251] The question is sent to the server via the terminal.

[1252] The server uses an NLP engine to analyze the data and extract keywords such as "new employees," "training," and "schedule."

[1253] The system searches for relevant information in the database and uses an emotion engine to recognize the user's emotions.

[1254] Finally, the NLP model generates a response such as, "The training schedule for new employees is as follows: Day 1 is orientation, Day 2 is an introduction to each department, and Day 3 is on-the-job training," and provides it to the user.

[1255] Example of a prompt

[1256] Examples of prompt statements to input into a generative AI model (e.g., GPT-3) include the following:

[1257] A user is asking, "I want to know the training schedule for new employees." Please generate an appropriate response to this question. The response should be polite and detailed.

[1258] The introduction of this system will improve user satisfaction by allowing users to access the information they need quickly and accurately, and by providing appropriate responses that are tailored to their emotions.

[1259] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1260] The processing flow of this system's program

[1261] Step 1:

[1262] The user enters the question on the device.

[1263] Input: The user enters the question into a text box on the terminal interface.

[1264] Operation: Enter a phrase such as "Please tell me the internal leave request flow" into the input field displayed on the interface.

[1265] Output: The question text is saved to the device.

[1266] Step 2:

[1267] The terminal sends this input to the server.

[1268] Input: The question text entered by the user.

[1269] Operation: The terminal converts the user's input into an HTTP request and sends it to the server. The request is formatted in JSON format and sent as an HTTP POST request.

[1270] Output: The server receives an HTTP request.

[1271] Step 3:

[1272] The server uses a natural language processing (NLP) engine to analyze the question.

[1273] Input: Received question text.

[1274] Operation: The server uses an NLP engine (e.g., spaCy, NLTK) to parse the question text. It tokenizes the question and extracts key keywords and context.

[1275] Output: Extracted keywords (e.g., "leave request", "flow") and contextual information.

[1276] Step 4:

[1277] The server searches the database for relevant information based on the analysis results.

[1278] Input: Extracted keywords and contextual information.

[1279] Operation: The server generates SQL queries and queries the database (e.g., MySQL, PostgreSQL). Specifically, it executes queries such as "SELECT FROM faq WHERE keywords LIKE '%leave request%' AND '%flow%'".

[1280] Output: Relevant information returned from the database.

[1281] Step 5:

[1282] The server's emotion engine recognizes the user's emotions.

[1283] Input: Received question text and related input patterns (spoken audio data, typing speed, etc.).

[1284] Operation: The server analyzes the user's emotional state using an emotion engine (e.g., IBM Watson, Microsoft Azure Text Analytics). It detects emotions such as stress and urgency.

[1285] Output: User's emotional state (e.g., stressed, in a hurry).

[1286] Step 6:

[1287] The server adjusts its response based on emotion.

[1288] Input: User's emotional state and related information retrieved from the database.

[1289] Operation: The server adjusts the tone and level of detail of its responses based on the perceived emotional state. For example, it generates more polite and detailed responses for users who are feeling stressed.

[1290] Output: Information on the adjusted response.

[1291] Step 7:

[1292] The server generates the final response.

[1293] Input: Information about the adjusted response and related information.

[1294] Operation: The server uses an NLP model (e.g., GPT-3) to generate natural-sounding responses. For example, it might generate a response like, "To request leave, first obtain approval from your supervisor, then submit the application to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[1295] Output: The generated response text.

[1296] Step 8:

[1297] The server sends the generated response to the terminal.

[1298] Input: The generated response text.

[1299] Operation: The server formats the generated response as an HTTP response and sends it to the user's terminal.

[1300] Output: The terminal receives an HTTP response.

[1301] Step 9:

[1302] The terminal receives a response from the server and displays it in the user interface.

[1303] Input: HTTP response received from the server.

[1304] Operation: The device analyzes the received response, extracts the response text, and displays it in the user interface. Specifically, it displays the response "The leave request flow is..." in the chat window.

[1305] Output: Explicit response information.

[1306] Step 10:

[1307] The user confirms the response displayed on the device.

[1308] Input: Response information displayed on the terminal.

[1309] Operation: The user checks the response displayed through the terminal and obtains the necessary information. For example, they might check information such as "The leave request flow is..."

[1310] Output: Information obtained by the user.

[1311] (Application Example 2)

[1312] 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."

[1313] Traditional FAQ systems can provide standard answers to user questions, but they cannot generate responses that take user emotions into account, making it difficult to provide sufficient satisfaction, especially when users are in a hurry or feeling stressed. Furthermore, in physical stores, there is a need for efficient and effective support when store staff quickly and accurately answer customer questions. To solve these problems, an FAQ system with the ability to recognize and respond to user emotions is necessary.

[1314] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for analyzing user input, means for searching for relevant information from a database based on the analysis results, means for generating a response based on the search results, means for recognizing the user's emotions, and means for adjusting the response content based on the emotion recognition results. This makes it possible to provide a higher level of satisfaction by adjusting the response according to the user's emotions.

[1315] "Means for receiving user input" refers to devices or software that provide an interface for users to input questions or requests into a system.

[1316] "Means for sending user input to the server" refers to protocols or mechanisms for sending information entered by a user to a server via the internet or a network.

[1317] "Means by which the server analyzes the user's input" refers to a system that analyzes data such as text and voice sent by the user using natural language processing technology.

[1318] "Means for searching for relevant information from a database based on the analysis results" refers to algorithms or search engines that find appropriate information from databases within or outside the system based on the analyzed input data.

[1319] "Means for generating a response based on the search results" refers to a program or AI model that generates an appropriate answer for the user based on the searched information.

[1320] "Means for sending the generated response to the user" refers to a system that sends the generated response to the user's terminal via communication means.

[1321] "Means of recognizing user emotions" refers to emotion recognition technology that identifies a user's emotional state from input text, voice, facial expressions, etc.

[1322] "Means for adjusting the response content based on the emotion recognition result" refers to an algorithm or program that adjusts the content and tone of the generated response based on the recognized emotion information.

[1323] "Means for analyzing the user's voice input and converting it into text" refers to speech recognition technology that converts the content entered by the user via voice into text information.

[1324] "Means for recognizing user emotions from the aforementioned voice input" refers to acoustic analysis techniques and algorithms for analyzing voice data to recognize the emotions the user is experiencing.

[1325] This invention relates to an FAQ system that generates quick and accurate responses to user questions and provides responses that take emotions into consideration. This system can be implemented, in particular, as a customer support system using smart glasses for use in physical stores.

[1326] System Configuration

[1327] hardware

[1328] This system uses the following hardware:

[1329] Smart glasses: Equipped with a camera, microphone, and speaker, used by staff to communicate with customers.

[1330] Server: Forms the core of the system, handling data processing, emotion recognition, and response generation.

[1331] software

[1332] This system uses the following software:

[1333] Speech recognition software: Use the speech_recognition library to convert customer speech to text.

[1334] Emotion recognition software: Using EmotionRecognition, we recognize customer emotions from voice and images.

[1335] The NLP engine transformers' pipeline is used to perform natural language processing to generate FAQ responses.

[1336] Database access: Use the requests library to retrieve information from the database.

[1337] Process Overview

[1338] 1. The user (customer) asks questions via voice to a staff member wearing smart glasses.

[1339] 2. A microphone built into the smart glasses worn by the staff captures the audio.

[1340] 3. Speech recognition software converts the speech into text.

[1341] 4. The server analyzes the converted text and extracts important keywords.

[1342] 5. The server searches the database based on the extracted keywords and retrieves relevant information.

[1343] 6. Simultaneously, emotion recognition software recognizes the user's emotions from audio data and camera footage.

[1344] 7. The server uses the emotion recognition results to adjust the content of the response it generates.

[1345] 8. The NLP engine generates the final response and communicates it to the staff through the smart glasses' display and speaker.

[1346] 9. The staff member then responds to the customer based on that response.

[1347] Specific example

[1348] For example, if a customer asks, "What is the warranty period for this product?", the system will operate as follows:

[1349] The smart glasses capture the voice, and the voice recognition software transcribes the question into text, such as "What is the warranty period for this product?"

[1350] The server searches the database for relevant information using "product warranty period" as a keyword and retrieves the results.

[1351] Emotion recognition software recognizes emotions from a customer's tone of voice and facial expressions, and determines, for example, that they are "feeling stressed."

[1352] Based on this sentiment information, the server generates a response saying, "My apologies. I will answer your question promptly. This product has a one-year warranty."

[1353] The staff member verifies this answer through smart glasses and immediately relays it to the customer.

[1354] Example of a prompt

[1355] "The customer has submitted a question via voice. Please analyze the question appropriately, recognize the customer's emotions, and then generate a response."

[1356] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1357] Step 1:

[1358] A user asks a staff member wearing smart glasses a question using voice. A microphone built into the smart glasses captures the voice data. The user's question is then input into the system as voice data. Input: Voice data. Output: Voice data.

[1359] Step 2:

[1360] The device (smart glasses) sends audio data to speech recognition software, which then converts the audio data into text. The speech recognition software uses the speech_recognition library to analyze the audio data and convert it back into text. Input: Audio data. Output: Text data.

[1361] Step 3:

[1362] The server receives text data and performs analysis using a natural language processing engine. Specifically, it extracts important keywords and context from the text. Here, the transformers library's pipeline is used. Input: Text data. Output: Keywords and context information.

[1363] Step 4:

[1364] The server searches the database for relevant information based on the analysis results. Here, the requests library is used to search for FAQ information stored in the database. Input: Keywords and contextual information. Output: Relevant information (FAQ responses).

[1365] Step 5:

[1366] Based on the FAQ responses obtained by the server, emotion recognition software is used to recognize the user's emotions. Emotion recognition includes voice tone and speed, and facial expression analysis. The EmotionRecognition engine is used here. Input: Audio data and camera video. Output: Emotion data.

[1367] Step 6:

[1368] The server uses sentiment data to adjust FAQ responses. For example, if a user is stressed, the response will be more polite; if they are in a hurry, the response will be more concise. Input: FAQ response and sentiment data. Output: Adjusted response.

[1369] Step 7:

[1370] The server generates the final response. It uses an NLP engine (the transformers library's pipeline) to generate a refined response in a natural way. Input: Data based on the refined response. Output: Final response text.

[1371] Step 8:

[1372] The server generates a final response text and sends it to the smart glasses, which then communicates it to the staff through the smart glasses' display and speaker. Input: Final response text. Output: Response information presented to the staff.

[1373] Step 9:

[1374] Staff members view the smart glasses display, confirm the response through the speaker, and provide the appropriate answer to the customer verbally. Input: Staff confirmation. Output: Response to the customer.

[1375] 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.

[1376] 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.

[1377] 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.

[1378] [Fourth Embodiment]

[1379] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1380] 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.

[1381] 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).

[1382] 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.

[1383] 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.

[1384] 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).

[1385] 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.

[1386] 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.

[1387] 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.

[1388] 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.

[1389] 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.

[1390] 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.

[1391] 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".

[1392] This invention relates to an FAQ AI Bot system that allows users to quickly access the information they need and respond accurately to irregular situations. This system can receive user questions as input, generate appropriate answers, and provide them.

[1393] System Configuration

[1394] User's terminal

[1395] Users access the FAQ AI Bot using devices such as PCs or smartphones. These devices provide an interface for users to enter questions, typically through a text form.

[1396] server

[1397] The server forms the core of the system and has the following main functions:

[1398] 1. Analyze user input

[1399] 2. Search for relevant information in the database.

[1400] 3. Generate a response

[1401] 4. Use AI models when performing filtering or inference.

[1402] Program processing

[1403] Enter and submit your question

[1404] 1. The user enters their question on their device. Example: "Please provide me with the latest information regarding company regulations."

[1405] 2. The terminal sends this input to the server. The server receives the HTTP request and extracts the text of the question.

[1406] Question analysis

[1407] 3. The server uses an NLP engine (e.g., BERT or GPT) to analyze the question and extract important keywords and context. In this example, "company regulations" and "latest information" are the main keywords.

[1408] Searching Databases

[1409] 4. The server searches the database for relevant information based on the analysis results. The database contains information such as past FAQs, manuals, and company regulations.

[1410] Response generation

[1411] 5. Based on the information acquired by the server, an NLP model generates a natural response. In most cases, information obtained from the database is sufficient, but if it does not meet specific conditions, an AI model is used to infer a response.

[1412] 6. The server sends the generated response to the terminal. The terminal displays this response in its user interface. Example: "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments."

[1413] Specific example

[1414] Example 1: Standard Question

[1415] 1. The user enters "Please explain the leave request process" into the terminal.

[1416] 2. The device sends this question to the server.

[1417] 3. The server uses an NLP engine to extract the keywords "leave request" and "flow".

[1418] 4. The server searches the database and retrieves relevant information (e.g., steps for requesting leave).

[1419] 5. The server generates a response stating, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department."

[1420] 6. The device displays this response, and the user confirms it.

[1421] Example 2: Questions regarding handling irregular situations

[1422] 1. The user asks, "Are there any exceptions for leave requests during emergencies?"

[1423] 2. The device sends this question to the server.

[1424] 3. The server uses an NLP engine to extract the keywords "emergency," "leave request," and "exception."

[1425] 4. The server searches the database, but if it cannot find a matching entry, it performs inference using the AI ​​model.

[1426] 5. The server generates a response stating, "In emergency situations, a retroactive application may be permitted as an exception."

[1427] 6. The device displays this response, and the user confirms it.

[1428] This configuration allows users to quickly obtain the information they need and to expect accurate responses even in specific situations.

[1429] The following describes the processing flow.

[1430] Step 1:

[1431] The user uses their device to enter a question into the FAQ AI Bot. For example, they might enter, "Please tell me the latest information regarding company regulations."

[1432] Step 2:

[1433] The device sends the user's question to the server. Specifically, it formats the question as an HTTP request and sends it to the server's API endpoint.

[1434] Step 3:

[1435] The server receives a question from the terminal. It parses the received HTTP request and extracts the question text.

[1436] Step 4:

[1437] The server inputs the received question into a natural language processing (NLP) engine. Specifically, it uses models such as BERT and GPT to analyze the question and extract important keywords and context.

[1438] Step 5:

[1439] The server searches the database for relevant information based on the analysis results. Specifically, it uses SQL queries and NoSQL queries to search the database and retrieve the relevant entries.

[1440] Step 6:

[1441] The server filters the search results if necessary. Filtering is performed based on pre-configured rules and parameters to select information that matches specific criteria.

[1442] Step 7:

[1443] If the server meets certain conditions, an AI model is used to infer missing information. For example, if information about special cases related to emergencies is missing, the AI ​​model will perform the inference.

[1444] Step 8:

[1445] The server generates the final response. Based on the acquired information and inference results, an NLP model generates a natural-sounding response. For example, it might generate a response such as, "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments."

[1446] Step 9:

[1447] The server sends the generated response to the terminal. The generated response is formatted as an HTTP response and sent to the terminal.

[1448] Step 10:

[1449] The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[1450] Step 11:

[1451] The user confirms the response displayed on their device. They can review the displayed information and quickly obtain the necessary information. For example, they might confirm information such as, "The latest version of the company regulations is as follows: Revised April 2023. There are exceptions for certain departments."

[1452] (Example 1)

[1453] 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".

[1454] Traditional FAQ systems sometimes made it difficult for users to quickly and accurately access the information they needed. Furthermore, they struggled to generate appropriate responses to irregular questions. There is a need to address these challenges and provide a system that allows users to obtain the information they need more efficiently.

[1455] 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.

[1456] In this invention, the server includes means for receiving user input, means for transmitting the user input to a communication device, means for the communication device to analyze the user input using a natural language processing device, means for searching for relevant information from a storage device based on the analysis results, means for generating a response based on the search results, and means for transmitting the generated response to a user device. This allows users to quickly access the information they need, as well as to respond accurately to irregular questions.

[1457] "Means for receiving user input" refers to the interface through which users input questions and information into the FAQ AI Bot system. This is generally a text input field found on devices such as personal computers and smartphones.

[1458] "Means for transmitting user input to a communication device" refers to a function that transmits user-entered questions or information to a server or communication network. This is typically done via HTTP requests or other protocols.

[1459] "Means by which a communication device analyzes the user's input using a natural language processing device" refers to a method in which a server or a specific computer device receives user input and analyzes the text information using natural language processing technology (e.g., an NLP engine). This extracts keywords and contextual information.

[1460] "Means for searching for relevant information from a storage device based on the analysis results" refers to the process of searching for relevant information from databases and other storage devices based on keywords and contextual information obtained from the analysis results. This includes SQL queries and NoSQL searches.

[1461] "Means for generating a response based on the search results" refers to a function that generates a natural language response to the user using information obtained from a storage device. This includes cases where an NLP model (e.g., GPT or BERT) is used.

[1462] "Means for sending the generated response to the user device" refers to the process of sending the generated response to the user's terminal using a communication protocol such as an HTTP response. The user terminal receives this and displays it on its interface.

[1463] This invention relates to an FAQ AI Bot system that allows users to quickly access necessary information and respond accurately to irregular situations. This system can receive user questions as input, generate appropriate answers, and provide them.

[1464] System Configuration

[1465] This system consists of user terminals, servers, and a communication network to connect them.

[1466] User's terminal

[1467] Users access the FAQ AI Bot using devices such as PCs and smartphones. These devices provide an interface for users to enter questions. Typically, this is a text input field where the user enters their question.

[1468] server

[1469] The server forms the core of the system and has the following main functions:

[1470] 1. Analyze user input

[1471] 2. Search for relevant information in the database.

[1472] 3. Generate a response

[1473] 4. Use AI models for filtering and inference.

[1474] To implement these functions, the server uses a natural language processing (NLP) engine (e.g., BERT or GPT), a database (e.g., MySQL, MongoDB), and a generative AI model (e.g., GPT-3).

[1475] Program processing

[1476] The following describes the specific operation of the system.

[1477] 1. The user enters a question into the FAQ AI Bot using a device such as a PC or smartphone. For example, they might enter, "Please tell me the latest information regarding company regulations."

[1478] 2. The terminal converts this input to JSON format and sends it to the server as an HTTP request.

[1479] 3. The server analyzes the received request and processes the question text using a natural language processing engine (BERT or GPT). The analysis extracts keywords such as "company regulations" and "latest information."

[1480] 4. The server uses these keywords to search databases (SQL databases or NoSQL databases) and retrieve relevant information.

[1481] 5. The server generates a response using natural language based on the information it has obtained. This may involve using an NLP model again. For example, a response such as, "The latest version of the company regulations is the April 2023 revision. There are exceptions for certain departments," might be generated.

[1482] 6. The server converts this response to JSON format and sends it to the terminal as an HTTP response.

[1483] 7. The response received by the terminal is displayed on the user interface for the user to confirm.

[1484] Specific example

[1485] Example 1: Standard Question

[1486] 1. The user enters "Please explain the leave request process" into the terminal.

[1487] 2. The device sends this question to the server.

[1488] 3. The server uses an NLP engine to extract the keywords "leave request" and "flow".

[1489] 4. The server searches the database and retrieves relevant information (e.g., steps for requesting leave).

[1490] 5. The server generates a response stating, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department."

[1491] 6. The device displays this response, and the user confirms it.

[1492] Example 2: Questions regarding handling irregular situations

[1493] 1. The user asks, "Are there any exceptions for leave requests during emergencies?"

[1494] 2. The device sends this question to the server.

[1495] 3. The server uses an NLP engine to extract the keywords "emergency," "leave request," and "exception."

[1496] 4. The server searches the database, but if it cannot find a matching entry, it performs inference using the AI ​​model.

[1497] 5. The server generates a response stating, "In emergency situations, a retroactive application may be permitted as an exception."

[1498] 6. The device displays this response, and the user confirms it.

[1499] Examples of prompt statements

[1500] When using a generative AI model, you can use prompt statements like the following:

[1501] "User question: 'Please tell me about the procedures for business trips.'"

[1502] "No relevant information was found. Please generate an appropriate response."

[1503] Thus, the present invention provides a system that allows users to quickly obtain information and to respond accurately even in specific situations.

[1504] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1505] Step 1:

[1506] Users input questions into the FAQ AI Bot using devices such as PCs or smartphones. The input is in text format; typically, a user might enter "Please provide the latest information regarding company regulations" into the text box. The input data is text.

[1507] Step 2:

[1508] The terminal receives the user's input, converts it to JSON format, and sends it to the server as an HTTP request. The input is the text entered by the user, and the output is a JSON object containing this text. The data processing performed here is the structuring of the text data.

[1509] Step 3:

[1510] The server parses the JSON from the received HTTP request and extracts the text data. The input here is the JSON object sent in step 2, and the output is the extracted raw text data. This process involves obtaining the necessary information from the HTTP request headers and body.

[1511] Step 4:

[1512] The server analyzes the extracted text using a natural language processing (NLP) engine (e.g., BERT or GPT). The input is the text data obtained in step 3, and the output is the analysis result, which includes important keywords and contextual information. Specifically, the process involves tokenizing the text and extracting important keywords.

[1513] Step 5:

[1514] The server searches the database based on the analysis results. The input here is the analyzed keywords, and the output is the relevant information retrieved from the database. Specifically, this involves creating SQL or NoSQL queries and searching the database.

[1515] Step 6:

[1516] The server generates a natural response based on information retrieved from the database. Typically, the database provides sufficient information to answer the question, but if the information is unavailable, an AI model is used to infer the answer. The input for this step is the database search results, and the output is the generated response text.

[1517] Step 7:

[1518] This function converts the server-generated response into JSON format and sends it to the terminal as an HTTP response. The input is the generated response text, and the output is a JSON object for the response. The specific actions here are converting the text to JSON and setting the response headers.

[1519] Step 8:

[1520] The terminal parses the response received from the server and displays it in the user interface. The input is a JSON object sent from the server, and the output is the response text to be displayed. This process includes parsing the JSON and updating the UI for text display.

[1521] Thus, this system seamlessly handles the entire process from user input to response generation, enabling the rapid and accurate provision of information.

[1522] (Application Example 1)

[1523] 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".

[1524] In modern brick-and-mortar stores, it is crucial to answer a wide range of customer questions quickly and accurately. However, conventional FAQ systems have been inadequate in handling specific situations and responding to voice input. This has made it difficult for store employees, especially those assisting customers, to answer customer questions in real time. To address this challenge, there is a need for a system that accurately analyzes customer voice questions and provides quick and accurate responses.

[1525] 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.

[1526] In this invention, the server includes means for receiving user input, means for transmitting the user input to the server, means for the server to analyze the user input, means for searching for relevant information from a database based on the analysis results, means for generating a response based on the search results, means for transmitting the generated response to the user, means for receiving customer questions as voice using a speech recognition device and converting them to text, means for transmitting the textened questions to the server along with the analysis results, means for the server to analyze the textened questions using a natural language processing engine and search for information from a database, and means for displaying the response on a display device. This enables faster and more accurate customer service in physical stores, and allows for appropriate responses to specific conditions and voice input.

[1527] "Means for receiving user input" refers to a device equipped with an interface for recognizing voice questions and text input from customers and incorporating them into the system.

[1528] "Means for sending to the server" refers to communication methods provided for sending user input data to a server via the internet or a local network.

[1529] "Means of analysis performed by the server" refers to algorithms and software that utilize a natural language processing engine to analyze user input on the server and extract keywords and context.

[1530] "Means for searching for relevant information from a database" refers to the means by which a server quickly searches for and retrieves appropriate information from within a database based on the analysis results.

[1531] "Means for generating responses" refers to a natural language generation engine that generates responses in a natural form in human language based on acquired information.

[1532] "Means for sending the generated response to the user" refers to communication means for returning the generated response to the user's device and displaying it.

[1533] A "speech recognition device" is a device and software that recognizes user-input speech and converts it into text.

[1534] "Means for converting to text" refers to speech recognition algorithms and software for converting speech data captured by a speech recognition device into text data.

[1535] A "natural language processing engine" is an algorithm and software that analyzes text data, extracts context and keywords, and searches for relevant information.

[1536] A "display device" is hardware such as a display that provides the user with a visual representation of the response generated on the device.

[1537] This invention is a system that allows users to ask voice questions in a physical store and generates quick and accurate responses based on those questions. This system consists of the following main elements:

[1538] Main components:

[1539] 1. Voice recognition device:

[1540] The system captures user voice questions and converts them to text. Specifically, it uses microphones built into wearable devices such as smart glasses and speech recognition software such as the Google Speech-to-Text API.

[1541] 2. Means for receiving user input:

[1542] It accepts audio data captured through a speech recognition device as text data. It uses the input interface of smart glasses or a smartphone.

[1543] 3. Means of sending text data to the server:

[1544] Text data is sent to a server via the internet. Communication is conducted using the HTTP protocol.

[1545] 4. Means the server uses for analysis:

[1546] The server receives text data and analyzes it using a natural language processing engine (e.g., BERT, GPT). It extracts context and keywords and performs preprocessing to generate an appropriate response.

[1547] 5. Means of retrieving information from a database:

[1548] The server searches the database for relevant information based on the analysis results. The database contains information such as FAQs, product features, and usage instructions.

[1549] 6. Means for generating a response:

[1550] Based on information retrieved from a database, an automated response generation engine generates natural-sounding responses. Typically, a generative AI model (e.g., GPT-3) is used for this purpose.

[1551] 7. Means for sending the generated response to the user:

[1552] The server sends the generated response as text data to the user's terminal. The HTTP protocol is used again.

[1553] 8. Display device:

[1554] The response is displayed on the user's device. The response is displayed on the smart glasses display or the smartphone screen.

[1555] Processing examples and specific examples:

[1556] Customer Questions and Answers

[1557] Customer question: "How do I wash this jacket?"

[1558] Speech recognition: The speech recognition device converts this question into text.

[1559] Sending to the server: Send the converted text "Please tell me how to wash this jacket" to the server.

[1560] Analysis and Retrieval: The server's natural language processing engine extracts the keywords "jacket" and "washing instructions," and searches the database for related information.

[1561] Response generation: Generate the response, "Hand washing is recommended for this jacket. Please wash it in lukewarm water below 30 degrees Celsius."

[1562] Response display: The response is displayed on the smart glasses' screen, providing the customer with appropriate information.

[1563] Example of a prompt:

[1564] 1. User's question: How do I wash this jacket?

[1565] 2. Response to be generated: Hand washing is recommended for this jacket. Please wash in lukewarm water below 30 degrees Celsius.

[1566] In this way, a system can be realized that provides smooth and accurate responses to customer questions in physical stores.

[1567] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1568] Step 1:

[1569] The user performs voice input.

[1570] Specific operation: The user wears smart glasses in a physical store and asks a voice question, for example, "How do I wash this jacket?"

[1571] Input: Customer voice question.

[1572] Output: Audio data captured by the smart glasses' microphone.

[1573] Step 2:

[1574] The device converts the audio data into text.

[1575] Specific operation: The microphone built into the smart glasses picks up speech, and the Google Speech-to-Text API is used to convert the speech to text in real time.

[1576] Input: Captured audio data.

[1577] Output: Text data (e.g., "Please tell me how to wash this jacket").

[1578] Step 3:

[1579] The terminal sends text data to the server.

[1580] Specific operation: This text data is sent to the server using the HTTP protocol via the smart glasses' communication module.

[1581] Input: Text data.

[1582] Output: HTTP request sent to the server.

[1583] Step 4:

[1584] The server receives and analyzes text data.

[1585] Specific operation: The server extracts text data from the HTTP request and performs contextual analysis and keyword extraction using a natural language processing engine (e.g., BERT or GPT).

[1586] Input: Text data sent to the server.

[1587] Output: Contextual analysis and keyword extraction results (e.g., "jacket", "washing instructions").

[1588] Step 5:

[1589] The server searches the database for relevant information.

[1590] Specific operation: Search the database for relevant information such as FAQs, product descriptions, and usage instructions based on keywords.

[1591] Input: Keyword extraction results.

[1592] Output: Relevant information retrieved from the database (e.g., "Hand washing is recommended, wash with lukewarm water below 30 degrees Celsius").

[1593] Step 6:

[1594] The server generates a response based on the information it has obtained.

[1595] Specific operation: Use a natural language generation engine on the server to assemble the retrieved information into a response in a natural language format.

[1596] Input: Information retrieved from the database.

[1597] Output: Generated response (Example: "Hand washing is recommended for this jacket. Please wash it in lukewarm water below 30 degrees Celsius.").

[1598] Step 7:

[1599] The server sends the generated response to the terminal.

[1600] Specific operation: The server sends the generated response to the smart glasses as an HTTP response.

[1601] Input: The generated response.

[1602] Output: HTTP response sent from the server.

[1603] Step 8:

[1604] The terminal displays a response.

[1605] Specific operation: The terminal (smart glasses) displays the response message received from the server on the screen. The store clerk confirms the content and conveys it to the customer.

[1606] Input: The response message received from the server.

[1607] Output: The response message displayed on the smart glasses' screen.

[1608] This series of processes enables quick and accurate responses to customer voice inquiries in physical stores.

[1609] 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.

[1610] This invention relates to an FAQ AI Bot system that generates and provides responses to user questions quickly and accurately. In addition to standard functions that analyze user input and retrieve relevant information from a database to generate responses, the system incorporates an emotion engine that recognizes user emotions and adjusts responses accordingly.

[1611] System Configuration

[1612] User's terminal

[1613] Users access the FAQ AI Bot using a device (such as a PC or smartphone). The device provides an interface for users to enter questions. These questions are typically entered into a text form.

[1614] server

[1615] The server forms the core of the system and has the following main functions:

[1616] 1. Analyze user input

[1617] 2. Search for relevant information in the database.

[1618] 3. Generate a response

[1619] 4. Use AI models for filtering and inference.

[1620] 5. Use an emotion engine to recognize the user's emotions and adjust responses accordingly.

[1621] Program processing

[1622] Enter and submit your question

[1623] 1. The user enters a question on their device. For example, they might enter, "Please tell me the company's leave request process."

[1624] 2. The terminal sends this input to the server. The server receives the HTTP request and extracts the text of the question.

[1625] Question analysis

[1626] 3. The server uses a natural language processing (NLP) engine to analyze the question and extract important keywords and context. In this example, "leave request" and "flow" are the main keywords.

[1627] Searching Databases

[1628] 4. The server searches the database for relevant information based on the analysis results. The database contains information such as company regulations and FAQs.

[1629] Recognition and regulation of emotional responses

[1630] 5. The server's emotion engine recognizes the user's emotions. It recognizes whether the user is stressed, in a hurry, etc., based on the text of the question, the voice spoken during input, and the typing speed.

[1631] 6. The server adjusts its response based on emotions. For example, if the user is feeling stressed, the response will be more polite and detailed.

[1632] Response generation

[1633] 7. The server generates the final response. Based on the acquired information and the results of the emotion engine, the NLP model generates a natural response. For example, it might generate a response such as, "To request leave, first obtain approval from your supervisor, then submit the application to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[1634] Sending a response

[1635] 8. The server sends the generated response to the terminal. The generated response is formatted as an HTTP response and sent to the terminal.

[1636] Display of response

[1637] 9. The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[1638] User response confirmation

[1639] 10. Confirm the response displayed on the user's device. For example, confirm the information that says, "To apply for leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please contact the Human Resources Department for details."

[1640] This configuration allows the present invention to enable users to quickly and accurately access the information they need, and further improve user satisfaction and operational efficiency by adjusting responses according to the user's emotions.

[1641] The following describes the processing flow.

[1642] Step 1:

[1643] The user enters a question into the FAQ AI Bot using their device. For example, they might enter, "Please tell me the company's leave request process."

[1644] Step 2:

[1645] The device sends this input to the server. Specifically, it formats the question as an HTTP request and sends it to the server's API endpoint.

[1646] Step 3:

[1647] The server receives a question from the terminal. It parses the received HTTP request and extracts the question text.

[1648] Step 4:

[1649] The server uses an NLP engine to analyze the question text. Specifically, it uses natural language processing models such as BERT and GPT to extract important keywords and context. In this example, "leave request" and "flow" are the main keywords.

[1650] Step 5:

[1651] The server searches the database for relevant information based on the analysis results. Specifically, it uses SQL or NoSQL queries to search the database and retrieve the relevant entries.

[1652] Step 6:

[1653] The server generates a response based on the data it has acquired. However, at this point, a standard response is generated that does not take into account the results of the emotion engine.

[1654] Step 7:

[1655] The server's emotion engine recognizes the user's emotions. Specifically, it analyzes parameters such as emotional expressions in the question text and the user's response time to determine whether the user is stressed or in a hurry.

[1656] Step 8:

[1657] The server adjusts its response based on the results of the emotion engine. For example, if the user is stressed, the response will be changed to be more polite and detailed. Specifically, it might generate a response such as, "To request leave, you must first obtain approval from your supervisor and then submit the application to HR. In urgent cases, a later application may be permitted as an exception. Please check with HR for details."

[1658] Step 9:

[1659] The server sends the final response it generates to the terminal. The response is formatted as an HTTP response and sent to the terminal.

[1660] Step 10:

[1661] The terminal receives a response from the server and displays it on the user interface. Specifically, it parses the received HTTP response and displays the response text.

[1662] Step 11:

[1663] The user checks the response displayed on their device. For example, they might check for information such as, "To request leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[1664] In this way, this system can generate quick and accurate responses to user questions, and further improve user satisfaction by adjusting the responses based on the user's emotions.

[1665] (Example 2)

[1666] 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".

[1667] Conventional FAQ systems generate a fixed response to user questions, but struggle to generate appropriate responses that take into account the user's emotions and the context of the question. As a result, they fail to adequately enhance user satisfaction, and inappropriate responses may be provided, especially in situations involving urgency or stress. This invention aims to solve this problem and provide users with more appropriate and efficient responses.

[1668] 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.

[1669] In this invention, the server includes means for using a natural language processing engine to analyze user input, means for recognizing the user's emotions using an emotion engine, and means for adjusting the response content based on the user's emotions. This enables the generation of appropriate responses that take into account the user's emotions and context. Furthermore, means for sending the generated response to the user enable the provision of information quickly and accurately.

[1670] "Means for receiving user input" refers to an interface for incorporating questions and requests entered by the user through a terminal into the system.

[1671] "Means for sending user input to the server" refers to communication means for transferring data entered by the user from the terminal to the server.

[1672] "Means by which the server uses a natural language processing engine to analyze the user's input" refers to a technique for the server to analyze the user's input using natural language processing technology and extract keywords and context.

[1673] "Means for searching for relevant information from a database based on the analysis results" refers to means for querying a database and obtaining relevant information based on the analysis results obtained by the natural language processing engine.

[1674] "Methods for recognizing user emotions using an emotion engine" refer to algorithms and technologies that analyze user input and related data to detect the user's emotional state.

[1675] "Means for adjusting response content based on user emotions" refers to means for adjusting the tone and level of detail of the generated response based on the recognized emotional state of the user.

[1676] "Means for generating a final response based on the search results and adjusted response content" refers to a technology that creates a final answer to present to the user based on the acquired information and adjusted response content.

[1677] "Means for sending the generated response to the user" refers to communication means for transferring the response generated by the server to the user's terminal and displaying it.

[1678] Modes for carrying out the invention

[1679] This invention relates to an FAQ AI Bot system that generates and provides responses to user questions quickly and accurately. In addition to standard functions that analyze user input and retrieve relevant information from a database to generate responses, the system combines an emotion engine to recognize the user's emotions and adjust the response accordingly.

[1680] System Configuration

[1681] User's terminal

[1682] Users access the FAQ AI Bot using a device (such as a PC or smartphone). The device provides an interface for users to enter questions. These questions are typically entered into a text form. For example, a specific question might be, "Could you please explain the company's leave request process?"

[1683] server

[1684] The server forms the core of the system and has the following main functions:

[1685] 1. Use a natural language processing (NLP) engine to analyze user input.

[1686] Examples of software that can be used include spaCy and NLTK.

[1687] For example, "leave request" and "flow" might be extracted as keywords.

[1688] 2. Search the database for relevant information.

[1689] The databases that can be used include MySQL and PostgreSQL.

[1690] Based on the analysis results, records related to the "Leave Request Flow" are searched.

[1691] 3. Use an emotion engine to recognize the user's emotions.

[1692] The software used includes IBM Watson and Microsoft Azure Text Analytics.

[1693] The system analyzes the user's input content and input speed to determine whether the user is stressed or in a hurry.

[1694] 4. Adjust responses based on the user's emotions.

[1695] For example, provide detailed and attentive responses to users who are experiencing stress.

[1696] 5. Generate the final response.

[1697] The generative AI model used is GPT-3.

[1698] For example, a response like this might be generated: "To request leave, first obtain approval from your supervisor, then submit the application form to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[1699] 6. Send the generated response to the terminal.

[1700] The server formats the generated response as an HTTP response and sends it to the user's terminal.

[1701] Specific example

[1702] As a concrete example, consider a case where a user enters the question, "I want to know the training schedule for new employees."

[1703] The question is sent to the server via the terminal.

[1704] The server uses an NLP engine to analyze the data and extract keywords such as "new employees," "training," and "schedule."

[1705] The system searches for relevant information in the database and uses an emotion engine to recognize the user's emotions.

[1706] Finally, the NLP model generates a response such as, "The training schedule for new employees is as follows: Day 1 is orientation, Day 2 is an introduction to each department, and Day 3 is on-the-job training," and provides it to the user.

[1707] Example of a prompt

[1708] Examples of prompt statements to input into a generative AI model (e.g., GPT-3) include the following:

[1709] A user is asking, "I want to know the training schedule for new employees." Please generate an appropriate response to this question. The response should be polite and detailed.

[1710] The introduction of this system will improve user satisfaction by allowing users to access the information they need quickly and accurately, and by providing appropriate responses that are tailored to their emotions.

[1711] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1712] The processing flow of this system's program

[1713] Step 1:

[1714] The user enters the question on the device.

[1715] Input: The user enters the question into a text box on the terminal interface.

[1716] Operation: Enter a phrase such as "Please tell me the internal leave request flow" into the input field displayed on the interface.

[1717] Output: The question text is saved to the device.

[1718] Step 2:

[1719] The terminal sends this input to the server.

[1720] Input: The question text entered by the user.

[1721] Operation: The terminal converts the user's input into an HTTP request and sends it to the server. The request is formatted in JSON format and sent as an HTTP POST request.

[1722] Output: The server receives an HTTP request.

[1723] Step 3:

[1724] The server uses a natural language processing (NLP) engine to analyze the question.

[1725] Input: Received question text.

[1726] Operation: The server uses an NLP engine (e.g., spaCy, NLTK) to parse the question text. It tokenizes the question and extracts key keywords and context.

[1727] Output: Extracted keywords (e.g., "leave request", "flow") and contextual information.

[1728] Step 4:

[1729] The server searches the database for relevant information based on the analysis results.

[1730] Input: Extracted keywords and contextual information.

[1731] Operation: The server generates SQL queries and queries the database (e.g., MySQL, PostgreSQL). Specifically, it executes queries such as "SELECT FROM faq WHERE keywords LIKE '%leave request%' AND '%flow%'".

[1732] Output: Relevant information returned from the database.

[1733] Step 5:

[1734] The server's emotion engine recognizes the user's emotions.

[1735] Input: Received question text and related input patterns (spoken audio data, typing speed, etc.).

[1736] Operation: The server analyzes the user's emotional state using an emotion engine (e.g., IBM Watson, Microsoft Azure Text Analytics). It detects emotions such as stress and urgency.

[1737] Output: User's emotional state (e.g., stressed, in a hurry).

[1738] Step 6:

[1739] The server adjusts its response based on emotion.

[1740] Input: User's emotional state and related information retrieved from the database.

[1741] Operation: The server adjusts the tone and level of detail of its responses based on the perceived emotional state. For example, it generates more polite and detailed responses for users who are feeling stressed.

[1742] Output: Information on the adjusted response.

[1743] Step 7:

[1744] The server generates the final response.

[1745] Input: Information about the adjusted response and related information.

[1746] Operation: The server uses an NLP model (e.g., GPT-3) to generate natural-sounding responses. For example, it might generate a response like, "To request leave, first obtain approval from your supervisor, then submit the application to the Human Resources Department. In urgent cases, a later application may be permitted as an exception. Please check with the Human Resources Department for details."

[1747] Output: The generated response text.

[1748] Step 8:

[1749] The server sends the generated response to the terminal.

[1750] Input: The generated response text.

[1751] Operation: The server formats the generated response as an HTTP response and sends it to the user's terminal.

[1752] Output: The terminal receives an HTTP response.

[1753] Step 9:

[1754] The terminal receives a response from the server and displays it in the user interface.

[1755] Input: HTTP response received from the server.

[1756] Operation: The device analyzes the received response, extracts the response text, and displays it in the user interface. Specifically, it displays the response "The leave request flow is..." in the chat window.

[1757] Output: Explicit response information.

[1758] Step 10:

[1759] The user confirms the response displayed on the device.

[1760] Input: Response information displayed on the terminal.

[1761] Operation: The user checks the response displayed through the terminal and obtains the necessary information. For example, they might check information such as "The leave request flow is..."

[1762] Output: Information obtained by the user.

[1763] (Application Example 2)

[1764] 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".

[1765] Traditional FAQ systems can provide standard answers to user questions, but they cannot generate responses that take user emotions into account, making it difficult to provide sufficient satisfaction, especially when users are in a hurry or feeling stressed. Furthermore, in physical stores, there is a need for efficient and effective support when store staff quickly and accurately answer customer questions. To solve these problems, an FAQ system with the ability to recognize and respond to user emotions is necessary.

[1766] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for analyzing user input, means for searching for relevant information from a database based on the analysis results, means for generating a response based on the search results, means for recognizing the user's emotions, and means for adjusting the response content based on the emotion recognition results. This makes it possible to provide a higher level of satisfaction by adjusting the response according to the user's emotions.

[1767] "Means for receiving user input" refers to devices or software that provide an interface for users to input questions or requests into a system.

[1768] "Means for sending user input to the server" refers to protocols or mechanisms for sending information entered by a user to a server via the internet or a network.

[1769] "Means by which the server analyzes the user's input" refers to a system that analyzes data such as text and audio sent by the user using natural language processing technology.

[1770] "Means for searching for relevant information from a database based on the analysis results" refers to algorithms or search engines that find appropriate information from databases within or outside the system based on the analyzed input data.

[1771] "Means for generating a response based on the search results" refers to a program or AI model that generates an appropriate answer for the user based on the searched information.

[1772] "Means for sending the generated response to the user" refers to a system that sends the generated response to the user's terminal via communication means.

[1773] "Means of recognizing user emotions" refers to emotion recognition technology that identifies a user's emotional state from input text, voice, facial expressions, etc.

[1774] "Means for adjusting the response content based on the emotion recognition result" refers to an algorithm or program that adjusts the content and tone of the generated response based on the recognized emotion information.

[1775] "Means for analyzing the user's voice input and converting it into text" refers to speech recognition technology that converts the content entered by the user via voice into text information.

[1776] "Means for recognizing user emotions from the aforementioned voice input" refers to acoustic analysis techniques and algorithms for analyzing voice data to recognize the emotions the user is experiencing.

[1777] This invention relates to an FAQ system that generates quick and accurate responses to user questions and provides responses that take emotions into consideration. This system can be implemented, in particular, as a customer support system using smart glasses for use in physical stores.

[1778] System Configuration

[1779] hardware

[1780] This system uses the following hardware:

[1781] Smart glasses: Equipped with a camera, microphone, and speaker, used by staff to communicate with customers.

[1782] Server: Forms the core of the system, handling data processing, emotion recognition, and response generation.

[1783] software

[1784] This system uses the following software:

[1785] Speech recognition software: Use the speech_recognition library to convert customer speech to text.

[1786] Emotion recognition software: Using EmotionRecognition, we recognize customer emotions from voice and images.

[1787] The NLP engine transformers' pipeline is used to perform natural language processing to generate FAQ responses.

[1788] Database access: Use the requests library to retrieve information from the database.

[1789] Process Overview

[1790] 1. The user (customer) asks questions via voice to a staff member wearing smart glasses.

[1791] 2. A microphone built into the smart glasses worn by the staff captures the audio.

[1792] 3. Speech recognition software converts the speech into text.

[1793] 4. The server analyzes the converted text and extracts important keywords.

[1794] 5. The server searches the database based on the extracted keywords and retrieves relevant information.

[1795] 6. Simultaneously, emotion recognition software recognizes the user's emotions from audio data and camera footage.

[1796] 7. The server uses the emotion recognition results to adjust the content of the response it generates.

[1797] 8. The NLP engine generates the final response and communicates it to the staff through the smart glasses' display and speaker.

[1798] 9. The staff member then responds to the customer based on that response.

[1799] Specific example

[1800] For example, if a customer asks, "What is the warranty period for this product?", the system will operate as follows:

[1801] The smart glasses capture the voice, and the voice recognition software transcribes the question into text, such as "What is the warranty period for this product?"

[1802] The server searches the database for relevant information using "product warranty period" as a keyword and retrieves the results.

[1803] Emotion recognition software recognizes emotions from a customer's tone of voice and facial expressions, and determines, for example, that they are "feeling stressed."

[1804] Based on this sentiment information, the server generates a response saying, "My apologies. I will answer your question promptly. This product has a one-year warranty."

[1805] The staff member verifies this answer through smart glasses and immediately relays it to the customer.

[1806] Example of a prompt

[1807] "The customer has submitted a question via voice. Please analyze the question appropriately, recognize the customer's emotions, and then generate a response."

[1808] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1809] Step 1:

[1810] A user asks a staff member wearing smart glasses a question using voice. A microphone built into the smart glasses captures the voice data. The user's question is then input into the system as voice data. Input: Voice data. Output: Voice data.

[1811] Step 2:

[1812] The device (smart glasses) sends audio data to speech recognition software, which then converts the audio data into text. The speech recognition software uses the speech_recognition library to analyze the audio data and convert it back into text. Input: Audio data. Output: Text data.

[1813] Step 3:

[1814] The server receives text data and performs analysis using a natural language processing engine. Specifically, it extracts important keywords and context from the text. Here, the transformers library's pipeline is used. Input: Text data. Output: Keywords and context information.

[1815] Step 4:

[1816] The server searches the database for relevant information based on the analysis results. Here, the requests library is used to search for FAQ information stored in the database. Input: Keywords and contextual information. Output: Relevant information (FAQ responses).

[1817] Step 5:

[1818] Based on the FAQ responses obtained by the server, emotion recognition software is used to recognize the user's emotions. Emotion recognition includes voice tone and speed, and facial expression analysis. The EmotionRecognition engine is used here. Input: Audio data and camera video. Output: Emotion data.

[1819] Step 6:

[1820] The server uses sentiment data to adjust FAQ responses. For example, if a user is stressed, the response will be more polite; if they are in a hurry, the response will be more concise. Input: FAQ response and sentiment data. Output: Adjusted response.

[1821] Step 7:

[1822] The server generates the final response. It uses an NLP engine (the transformers library's pipeline) to generate a refined response in a natural way. Input: Data based on the refined response. Output: Final response text.

[1823] Step 8:

[1824] The server generates a final response text and sends it to the smart glasses, which then communicates it to the staff through the smart glasses' display and speaker. Input: Final response text. Output: Response information presented to the staff.

[1825] Step 9:

[1826] Staff members view the smart glasses display, confirm the response through the speaker, and provide the appropriate answer to the customer verbally. Input: Staff confirmation. Output: Response to the customer.

[1827] 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.

[1828] 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.

[1829] 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.

[1830] 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.

[1831] 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.

[1832] 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.

[1833] 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.

[1834] 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.

[1835] 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."

[1836] 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.

[1837] 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.

[1838] 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.

[1839] 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.

[1840] 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.

[1841] 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.

[1842] 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.

[1843] 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.

[1844] 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.

[1845] 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.

[1846] 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.

[1847] 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 as being incorporated by reference.

[1848] The following is further disclosed regarding the embodiments described above.

[1849] (Claim 1)

[1850] A means of receiving user input,

[1851] Means for sending the user's input to the server,

[1852] The server provides means for analyzing the user's input,

[1853] A means for searching for relevant information from a database based on the aforementioned analysis results,

[1854] Means for generating a response based on the search results,

[1855] Means for sending the generated response to the user,

[1856] A system that includes this.

[1857] (Claim 2)

[1858] The system according to claim 1, further comprising means for filtering information that matches specific conditions based on the analysis results.

[1859] (Claim 3)

[1860] The system according to claim 1, further comprising means for performing inference using an AI model when the aforementioned specific conditions are met.

[1861] "Example 1"

[1862] (Claim 1)

[1863] A means of receiving user input,

[1864] Means for transmitting the user's input to a communication device,

[1865] The communication device includes means for analyzing the user's input with a natural language processing device,

[1866] A means for searching for relevant information from a storage device based on the aforementioned analysis results,

[1867] Means for generating a response based on the search results,

[1868] Means for transmitting the generated response to the user device,

[1869] A system that includes this.

[1870] (Claim 2)

[1871] The system according to claim 1, which filters information that matches specific conditions based on the analysis results.

[1872] (Claim 3)

[1873] The system according to claim 1, which performs inference using an AI model when the aforementioned specific conditions are met.

[1874] "Application Example 1"

[1875] (Claim 1)

[1876] A means of receiving user input,

[1877] Means for sending the user's input to the server,

[1878] The server provides means for analyzing the user's input,

[1879] A means for searching for relevant information from a database based on the aforementioned analysis results,

[1880] Means for generating a response based on the search results,

[1881] Means for sending the generated response to the user,

[1882] A means of receiving customer questions by voice using a speech recognition device and converting them into text,

[1883] A means for sending the aforementioned text-based questions to a server along with the analysis results,

[1884] A server analyzes the question, which has been transcribed into text using a natural language processing engine, and retrieves information from a database.

[1885] means for displaying the response on a display device,

[1886] A system that includes this.

[1887] (Claim 2)

[1888] The system according to claim 1, further comprising means for filtering information that matches specific conditions based on the analysis results.

[1889] (Claim 3)

[1890] The system according to claim 1, further comprising means for performing inference using an AI model when the aforementioned specific conditions are met.

[1891] "Example 2 of combining an emotion engine"

[1892] (Claim 1)

[1893] A means of receiving user input,

[1894] Means for sending the user's input to the server,

[1895] The server uses a natural language processing engine to analyze the user's input,

[1896] A means for searching for relevant information from a database based on the aforementioned analysis results,

[1897] A means of recognizing a user's emotions using an emotion engine,

[1898] A means of adjusting the response based on the user's emotions,

[1899] Means for generating a final response based on the search results and the adjusted response content,

[1900] Means for sending the generated response to the user,

[1901] A system that includes this.

[1902] (Claim 2)

[1903] The system according to claim 1, further comprising means for filtering information that matches specific conditions based on the analysis results.

[1904] (Claim 3)

[1905] The system according to claim 1, further comprising means for performing inference using a generative AI model when the aforementioned specific conditions are met.

[1906] "Application example 2 when combining with an emotional engine"

[1907] (Claim 1)

[1908] A means of receiving user input,

[1909] Means for sending the user's input to the server,

[1910] The server provides means for analyzing the user's input,

[1911] A means for searching for relevant information from a database based on the aforementioned analysis results,

[1912] Means for generating a response based on the search results,

[1913] Means for sending the generated response to the user,

[1914] Means of recognizing user emotions,

[1915] Means for adjusting the response content based on the aforementioned emotion recognition results,

[1916] A system that includes this.

[1917] (Claim 2)

[1918] The system according to claim 1, further comprising means for filtering information that matches specific conditions based on the analysis results.

[1919] (Claim 3)

[1920] The system according to claim 1, further comprising means for performing inference using an AI model when the aforementioned specific conditions are met.

[1921] (Claim 4)

[1922] The means for analyzing the user's voice input and converting it into text,

[1923] The system according to claim 1, further comprising means for recognizing the user's emotions from the voice input. [Explanation of Symbols]

[1924] 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 user input, Means for sending the user's input to the server, The server provides means for analyzing the user's input, A means for searching for relevant information from a database based on the aforementioned analysis results, Means for generating a response based on the search results, Means for sending the generated response to the user, A system that includes this.

2. The system according to claim 1, further comprising means for filtering information that matches specific conditions based on the analysis results.

3. The system according to claim 1, further comprising means for performing inference using an AI model when the aforementioned specific conditions are met.

Images