system

Abstract

Provide a system. [Solution] A means for receiving inquiry emails; means for analyzing the content of the inquiry email; means for generating an appropriate response based on the analyzed content; means for transmitting the generated response; A system including:

Description

[Technical Field]

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

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

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

[0004] In the past, responding to inquiries from companies required a lot of time and manpower, making it difficult to provide quick and accurate answers. Furthermore, each company representative had to manually research and create answers based on the content of each individual inquiry, which was inefficient and could lead to inconsistent response quality. This led to problems such as lower customer satisfaction and a deterioration in work efficiency. [Means for solving the problem]

[0005] The present invention provides a system that receives and analyzes inquiry emails, and automatically generates and sends appropriate responses. The system includes a means for receiving inquiry emails, a means for analyzing the content of the inquiry emails, a means for generating appropriate responses based on the analysis results, and a means for sending the generated responses, enabling prompt and accurate responses. Furthermore, the system uses natural language processing technology as a means for analysis, and retrieves necessary data from a database based on the analyzed content, improving the accuracy and consistency of responses.

[0006] An "inquiry email" is an email sent by a company or individual requesting specific information or assistance.

[0007] "Means for receiving" refers to the function of receiving email sent from outside using a device such as a server or mail server, or the software installed on such a device.

[0008] "Means of analysis" refers to the technology and algorithms used to process the content of received emails and understand their meaning, specifically natural language processing technology.

[0009] "Means of generation" refers to technology or algorithms that automatically create appropriate answers or information based on the analyzed content.

[0010] "Transmission means" refers to the technology and functionality for sending the generated response as email, and typically includes an SMTP server and corresponding communication protocols.

[0011] "Natural language processing technology" refers to technology that enables computers to understand, process, and generate human language, and includes tokenization, part-of-speech tagging, and semantic analysis of sentences.

[0012] A "database" is a system for efficiently storing, managing, and querying structured data, and is used to provide appropriate data or information in response to a query. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0014] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0015] First, the terms used in the following description will be explained.

[0016] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0017] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0018] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0019] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0020] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0021] [First embodiment]

[0022] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0023] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0024] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0025] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0026] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0027] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0028] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0029] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0030] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0031] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0032] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0033] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0034] This invention provides a system that responds quickly and appropriately to inquiry emails from companies. This system automatically performs a series of processes: receiving an inquiry email, analyzing its content, generating an appropriate response, and sending it as a reply email. The processing of the system's program is explained below in natural language.

[0035] System Configuration

[0036] The system includes the following components:

[0037] Server: Hosts the mail server and AI engine, and receives, analyzes, generates, and sends email inquiries.

[0038] Device: A PC or mobile device used by a user (company representative). Sends an inquiry email and receives an automatic reply email from the server.

[0039] User: Company representative. Responsible for sending inquiry emails to the system.

[0040] Program processing

[0041] 1. Server: Receives inquiry emails. Receives emails sent by users (company personnel) and converts the email text into text format.

[0042] 2. Server: Sends the email text to the AI ​​engine and analyzes the email content using natural language processing. Specifically, it performs the following processes:

[0043] Tokenization: Breaking down email content into words.

[0044] Part-of-speech tagging: tag each word with its part of speech.

[0045] Topic extraction: Extracting keywords and phrases to identify the subject of a query.

[0046] 3. Server: Based on the analysis results, the inquiry is categorized. Based on this categorization, the AI ​​engine generates the optimal answer corresponding to the inquiry.

[0047] 4. Server: Retrieves relevant data from the database as needed, for example, stock information, pricing information, or other data depending on the subject of the query.

[0048] 5. Server: Using templates, embed the acquired data and generate a reply email in a natural style. Template engines such as Jinja2 can be used.

[0049] 6. Server: The generated response email is sent to the user (company representative). The sending process is performed using an SMTP server.

[0050] Specific examples of processing

[0051] Example 1: Product inventory inquiry

[0052] User: A company representative sends an email inquiring about the stock status of Product A.

[0053] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0054] AI engine: Analyzes the email content and determines that the main question is about stock status. It retrieves stock information for product A from the database and generates a response based on that information.

[0055] Generated answer:

[0056] "Product A's current stock status is as follows:

[0057] Number in stock: 150

[0058] Next arrival: October 15, 2023

[0059] Thank you for your confirmation."

[0060] This email will include proposal materials such as an operation manual for the inventory management system and a detailed catalog for Product A.

[0061] Server: The generated email is automatically sent to the company representative.

[0062] Example 2: Price enquiry

[0063] User: A company representative sends an email inquiring about the price of Product B.

[0064] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0065] AI engine: Analyzes the email content and determines that the main question is about price. It retrieves the latest price information for Product B from a price information database and generates a response based on that information.

[0066] Generated answer:

[0067] "Product B's latest price information is as follows:

[0068] Unit price: 5,000 yen

[0069] Discount information: 10% off this month only

[0070] Please refer to the price list for details. If you have any questions, please let us know.

[0071] This email will include proposal materials such as a price list and details of discount campaigns.

[0072] Server: The generated email is automatically sent to the company representative.

[0073] This allows the system of the present invention to respond quickly and appropriately to inquiry emails from companies.

[0074] The processing flow will be explained below.

[0075] Step 1:

[0076] User: A company representative creates and sends an inquiry email.

[0077] Step 2:

[0078] Server: The mail server receives the inquiry email.

[0079] The SMTP server listens for emails on port 25, stores them in the inbox, and converts the email content into text format.

[0080] Step 3:

[0081] Server: Sends the email converted to text format to the AI ​​engine.

[0082] The email text is sent as a REST API request to the AI ​​engine for analysis.

[0083] Step 4:

[0084] AI engine: Analyzes email content using natural language processing technology.

[0085] It performs tokenization (dividing email content into words), part-of-speech tagging (tagging each word with its part of speech), and subject extraction (extracting keywords and phrases to identify the subject of the inquiry), using technologies such as morphological analyzers, POS taggers, TF-IDF, and BERT.

[0086] Step 5:

[0087] Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry.

[0088] Based on the analysis results, the products are divided into categories such as "inventory," "price," and "specifications."

[0089] Step 6:

[0090] Server: Select the appropriate template according to the category.

[0091] For example, an "inventory" inquiry uses an availability response template, and a "price" inquiry uses a price information response template.

[0092] Step 7:

[0093] Server: Retrieves the required data from the database.

[0094] Use Python scripts and SQL queries to pull relevant data from the database, such as inventory information, pricing information, etc.

[0095] Step 8:

[0096] Server: The acquired data is embedded in the template and a reply email is generated.

[0097] Use a template engine such as Jinja2 to create emails in a natural writing style.

[0098] Step 9:

[0099] Server: The generated response email is sent to the user.

[0100] Use an email sending module (such as smtplib) to send an automatic reply email to the user's email address via the SMTP server.

[0101] Step 10:

[0102] User: The company representative receives the automatic reply email and checks the contents.

[0103] Example 1

[0104] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0105] A problem with conventional inquiry response systems is that it takes a great deal of time and effort to analyze the content of received emails and then generate and send appropriate responses based on that information. Furthermore, manual responses are prone to human error and inconsistencies, making it difficult to provide effective and consistent responses. Furthermore, there is a need for system automation to quickly obtain appropriate data and reflect it in responses.

[0106] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0107] In this invention, the server includes a means for receiving an inquiry email, a means for analyzing the content of the inquiry email using natural language processing technology, a means for categorizing the inquiry content based on the analysis result, a means for retrieving related data from a database according to the categorized content, a means for generating an appropriate response using the retrieved data, and a means for transmitting the generated response. This automates the inquiry response process, enabling quick and accurate responses. Furthermore, the ability to generate consistent responses improves the quality of a company's inquiry responses.

[0108] An "inquiry email" is an email containing a question or request that a user sends to the system.

[0109] "Means of receiving" refers to the functions and processes by which the server receives and stores inquiry emails.

[0110] "Means for analysis" refers to the functions and processes for analyzing the content of received inquiry emails using natural language processing technology and extracting important information.

[0111] "Natural language processing technology" is a field of computer science that involves understanding and analyzing the meaning of text data, and is a technology that performs tasks such as subject extraction and part-of-speech tagging.

[0112] A "generative AI model" is an artificial intelligence model that generates answers in natural language based on input data.

[0113] A "prompt" is an instruction or cue provided to an AI model when it performs analysis.

[0114] "Means of categorization" refers to the functions and processes for classifying the content of inquiries based on the analysis results and determining the appropriate processing method.

[0115] A database is a system for organizing and storing large amounts of information and quickly retrieving data when needed.

[0116] "Means of obtaining relevant data" refers to the functions and processes for extracting the necessary information from the database in response to the inquiry.

[0117] "Answer generation means" refers to the functions and processes that automatically generate an appropriate reply to the user based on the relevant data obtained.

[0118] A "template engine" is software that uses template files to embed data in order to generate dynamic content.

[0119] The "means of sending" refers to the function or process for delivering the generated reply email to the user, and uses an SMTP server or the like.

[0120] A "system" is an integrated device or software that automatically performs a series of processes including receiving, analyzing, classifying, acquiring data, generating responses, and sending inquiries.

[0121] This invention is an automated system for responding promptly and appropriately to inquiry emails from companies. This system performs a series of automated processes: receiving inquiry emails, analyzing their contents, generating responses, and sending the responses.

[0122] System Configuration

[0123] The system includes the following components:

[0124] Server: Hosts the mail server and natural language processing engine, receives and analyzes inquiry emails, generates responses, and sends them. Postfix is ​​used as the mail server, and OpenAI (registered trademark) GPT-3 (registered trademark) is used as the natural language processing engine.

[0125] Device: A PC or mobile device used by a user (company representative). For example, a Windows PC or iPhone (registered trademark).

[0126] User: Company representative. Responsible for sending inquiry emails to the system.

[0127] Program processing

[0128] The server responds to the inquiry email by following these steps:

[0129] 1. Receiving an inquiry email:

[0130] A user sends an inquiry email using a PC or mobile device. The server receives and stores the inquiry email using the Postfix mail server.

[0131] 2. Email content analysis:

[0132] The server sends the received email text to an AI engine for natural language processing (OpenAI GPT-3) using a prompt. An example of a specific prompt is "What is the subject of this email? I would like to check the stock of product A." The natural language processing engine performs the following process:

[0133] Tokenization: Splitting the text of an email into words. For example, "I would like to check the stock of product A" would be split into "I would like to check the stock of product A."

[0134] Part-of-speech tagging: tag each word with its part of speech, for example, "Product A / noun | of / particle | inventory / noun | wo / particle | confirm / verb | shi / auxiliary verb | tai / auxiliary verb | desu / auxiliary verb."

[0135] Subject Extraction: Use TF-IDF and other natural language processing algorithms to extract the subject of the query. For example, identify that "stock availability" is the subject.

[0136] 3. Categorize your inquiry:

[0137] Based on the analysis results obtained from the AI ​​engine, the server classifies the inquiry into appropriate categories (e.g., stock confirmation, price confirmation).

[0138] 4. Obtaining relevant data:

[0139] The server sends a query to the MySQL® database according to the categorized inquiry content and retrieves the appropriate information. For example, it executes a query such as "SELECT stock FROM products WHERE product_name="Product A"" to retrieve stock information for Product A.

[0140] 5. Generate reply email:

[0141] The server uses the Jinja2 template engine to embed the acquired data into a template and generate a reply email in a natural style. For example, the stock quantity data is embedded in a template such as "The stock quantity is {{ stock}} units."

[0142] 6. Sending response email:

[0143] The server sends the generated response email to the user (company representative) using the SMTP server (Postfix). If the email was sent successfully, it is recorded in the log.

[0144] Specific examples

[0145] Example 1: Product inventory inquiry

[0146] User: A company representative sends an email inquiring about the stock status of Product A.

[0147] Server: The Postfix mail server receives the email, converts it to email text, and then sends the email text to the AI ​​engine with the following prompt:

[0148] "What is the subject of this email?: I would like to check the availability of product A."

[0149] AI engine: Analyzes the email content and determines that the subject is "stock status."

[0150] Server: Based on the analysis results of the AI ​​engine, classify this inquiry into the "Stock Check" category. Then, send a query to the database saying "Provide the stock quantity of product A."

[0151] Server: Retrieve stock information for Product A (for example, stock quantity 150) from the database and generate a response email using the Jinja2 template engine.

[0152] "Product A's current stock status is as follows:

[0153] Number in stock: 150

[0154] Next arrival: October 15, 2023

[0155] Thank you for your confirmation."

[0156] Server: The generated response email is sent to the user (company representative) using an SMTP server.

[0157] Example 2: Price enquiry

[0158] User: A company representative sends an email inquiring about the price of Product B.

[0159] Server: The Postfix mail server receives the email, converts it to email text, and then sends the email text to the AI ​​engine with the following prompt:

[0160] "What is the subject of this email? I would like to know the price of product B."

[0161] AI engine: Analyzes the content of the email and determines that the subject is "price."

[0162] Server: Based on the analysis results of the AI ​​engine, the server classifies this inquiry into the "price check" category. It then sends a query to the database saying, "Provide the price of product B."

[0163] Server: Retrieve the price information of Product B (for example, unit price 5,000 yen, discount 10%) from the database and generate a reply email using the Jinja2 template engine.

[0164] "Product B's latest price information is as follows:

[0165] Unit price: 5,000 yen

[0166] Discount information: 10% off this month only

[0167] Please refer to the price list for details. If you have any questions, please let us know.

[0168] Server: The generated response email is sent to the user (company representative) using an SMTP server.

[0169] In this way, the system of the present invention can respond quickly and appropriately to inquiry emails from companies.

[0170] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0171] Step 1: Receiving an inquiry email

[0172] Server: A user (company employee) sends an inquiry email. The Postfix mail server receives the email. The input is the email sent by the user, and the output is the email data stored on the mail server. Specifically, the server converts the received email into a specific format (e.g., a text file).

[0173] Step 2: Analyzing the email content

[0174] Server: Sends the text of the received email to an AI engine for natural language processing (OpenAI GPT-3). The input is the text content of the email, and the output is the analysis result. The specific operation is as follows:

[0175] Tokenization: Splitting text into words (e.g., "I would like to check the stock of product A" → "I would like to check the stock of product A").

[0176] Part-of-speech tagging: Tag each word with its part of speech (e.g., "Product A / noun | of / particle | inventory / noun | wo / particle | confirm / verb | shi / auxiliary verb | tai / auxiliary verb | desu / auxiliary verb").

[0177] Topic extraction: TF-IDF is used to extract keywords and identify topics (e.g., "stock status").

[0178] Step 3: Categorize the inquiry

[0179] Server: Categorizes the inquiry content based on the analysis results of the AI ​​engine. The input is the analysis result, and the output is category information (e.g., "Check stock availability" or "Check price"). The specific operation is to classify the inquiry into predefined categories based on the subject.

[0180] Step 4: Retrieve related data

[0181] Server: Sends a query to a database (e.g. MySQL) based on the categorization. The input is the category information and the associated data request, and the output is the specific information retrieved from the database (e.g., the number of items in stock for product A). The specific operation is as follows:

[0182] Create a query: Create an SQL query based on the subject of your inquiry (e.g., "SELECT stock FROM products WHERE product_name="Product A"").

[0183] Data Acquisition: Pulling the required information from the database.

[0184] Step 5: Generate a response email

[0185] Server: Generates a reply email using the Jinja2 template engine based on the retrieved data. The input is the information retrieved from the database and a template, and the output is a reply email. The specific operations are as follows:

[0186] Embed data: Embed data in the template (e.g., "Your stock is {{ stock}}").

[0187] Email formatting: Format your emails in a natural style.

[0188] Step 6: Sending a response email

[0189] Server: The generated reply email is sent to the user (company representative) using the SMTP server (Postfix). The input is the reply email, and the output is a confirmation of successful sending. The specific operations are as follows:

[0190] Sending email: Sends email using the SMTP protocol.

[0191] Log: Logs successful transmission.

[0192] (Application example 1)

[0193] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0194] Modern electronic payment services are receiving an increasing number of inquiries from users, and they need to respond to these inquiries quickly and accurately. However, because the content of inquiries varies widely, manually responding to all inquiries is difficult, time-consuming, and costly. Furthermore, there is a lack of systems that can provide appropriate transaction information based on the content of inquiries and avoid inappropriate responses. Therefore, there is a need to develop a system that automates and streamlines inquiry responses in electronic payment services.

[0195] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0196] In this invention, the server includes means for receiving an inquiry message, means for analyzing the contents of the inquiry message, means for generating an appropriate response based on the analyzed content, means for transmitting the generated response, means for retrieving transaction information from a database based on the analyzed content, and means for customizing templates for the response, thereby enabling prompt and accurate responses to a variety of inquiries in electronic payment services.

[0197] An "inquiry message" is a message containing a question or confirmation that a user makes to a service provider.

[0198] "Analysis" refers to the methods and results used to examine and understand the content of given data or messages.

[0199] A "response" is a reply or answer provided to an inquiry or question.

[0200] "Sending" is the act of delivering data or messages to another device or network destination.

[0201] "Transaction information" refers to detailed data about a specific transaction, primarily in electronic payment services.

[0202] A "database" is an organized collection of data and a system that allows for efficient searching and retrieval.

[0203] A "template" is a document or file with a prescribed format or pattern, designed so that parts of its content can be easily changed.

[0204] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language.

[0205] A "generative AI model" is an artificial intelligence algorithm that generates optimal outputs for a task based on input data.

[0206] A "prompt sentence" is a sentence that contains instructions or a request to perform a specific task or action.

[0207] The present invention relates to a system that automates inquiries about electronic payment services and provides prompt and accurate responses. This system is comprised of multiple servers, terminals, and users, and is described in detail below.

[0208] System Configuration

[0209] The system includes the following components:

[0210] Server: The central component that receives query messages, analyzes their content, generates appropriate responses, and sends them. The server also hosts an AI engine and has the ability to analyze the content of messages using natural language processing techniques.

[0211] Terminal: A communication device such as a smartphone or PC used by a user, whose main role is to input inquiry messages and receive automatic responses from the server.

[0212] User: A person or company representative who uses electronic payment services and sends inquiry messages.

[0213] Technology and software used

[0214] The server performs natural language processing using a generative AI model based on Hugging Face's Transformers library, and uses an SQLite database to manage necessary data such as transaction and price information, and retrieves the appropriate data depending on the inquiry.

[0215] Program processing

[0216] When the server receives a query message, it converts it into text format and sends it to the AI ​​engine. The AI ​​engine analyzes the content of the query message and performs appropriate tokenization, part-of-speech tagging, and topic extraction. Based on the analysis results, it then classifies the query content into an appropriate category and retrieves related data (e.g., transaction information and price information) from the database.

[0217] The server then combines the retrieved data with the template to generate a natural-sounding response, possibly using a template engine (e.g., Jinja2), which is then automatically sent to the user's device.

[0218] Examples and prompts

[0219] As a concrete example, if a user makes a query such as "What is the transaction status of product A?", the server will act as follows:

[0220] 1. User: Enter "Please tell me the transaction status of product A" on your smartphone and send it.

[0221] 2. Server: Receives the message and analyzes its contents. The AI ​​engine determines that the subject is "transaction status."

[0222] 3. Server: Retrieve the transaction status related to product A from the database and generate a natural response based on it.

[0223] 4. Server: Send the generated response "The latest transaction status for product A is 'Completed'" to the user's terminal.

[0224] Prompt Sentence Examples

[0225] (Prompt example 1)

[0226] User: Please tell me the price of product A.

[0227] System: The price of product A is 5,000 yen. Please check the details.

[0228] (Prompt example 2)

[0229] User: What is the transaction status for Item B?

[0230] System: The latest transaction status for Item B is 'Pending'. We will notify you as soon as there is an update.

[0231] In this way, the system can provide quick and accurate responses to various inquiries about electronic payment services.

[0232] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0233] Step 1:

[0234] Users input and send inquiry messages from their smartphones, PCs, or other devices. The input messages are then sent from the user's device to the server.

[0235] Step 2:

[0236] The server receives the query message and converts the data sent in the form of an email message or chat message into a text format, resulting in text data that can be parsed.

[0237] Step 3:

[0238] The server sends the received text data to the AI ​​engine (generative AI model) and analyzes the content. In this step, the following processes are performed using natural language processing technology:

[0239] Tokenization: Breaking the message content down into words.

[0240] Part-of-speech tagging: tag each word with its part of speech.

[0241] Thematic extraction: Identifying the main topics and keywords of a message.

[0242] The analysis results in themes and keywords related to the content of the query.

[0243] Step 4:

[0244] Based on the analysis, the server classifies the query into an appropriate category, such as "price inquiry," "transaction status inquiry," etc. This classification serves as a preliminary step in generating an appropriate response.

[0245] Step 5:

[0246] The server retrieves the necessary information from the database based on the category. For example, if the query is "price inquiry," it retrieves the relevant product price information from the database. Specifically, it executes an SQL query to extract the necessary data, such as price information and transaction status.

[0247] Step 6:

[0248] The server uses the retrieved data and the template to generate a response in a natural style. Using a template engine (such as Jinja2), the server embeds the data in the template to create a response. This response is an appropriate response to the query.

[0249] Step 7:

[0250] The server then sends the generated response to the user's device. The response message is delivered to the user using the SMTP protocol or the chat application's API. This process allows the user to receive a prompt and appropriate response.

[0251] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0252] The present invention provides a system that responds quickly and appropriately to inquiry emails from companies. This system automatically performs a series of processes: receiving an inquiry email, analyzing its content, generating an appropriate response, and sending it as a reply email. Furthermore, by combining this system with an emotion engine that recognizes the user's emotions, the system can adjust the tone and content of the response to match the user's emotions. The system's program processing is explained below in natural language.

[0253] System Configuration

[0254] The system includes the following components:

[0255] Server: Hosts the mail server and AI engine, receives and analyzes inquiry emails, generates and sends responses, and is also equipped with an emotion engine to analyze user emotions.

[0256] Device: A PC or mobile device used by a user (company representative). Sends an inquiry email and receives an automatic reply email from the server.

[0257] User: Company representative. Responsible for sending inquiry emails to the system.

[0258] Program processing

[0259] 1. Server: Receives inquiry emails. Receives emails sent by users (company personnel) and converts the email text into text format.

[0260] 2. Server: Sends the email text to the AI ​​engine and analyzes the email content using natural language processing. Specifically, it performs the following processes:

[0261] Tokenization: Breaking down email content into words.

[0262] Part-of-speech tagging: tag each word with its part of speech.

[0263] Topic extraction: Extracting keywords and phrases to identify the subject of a query.

[0264] 3. Server: Analyzes the user's emotions from the email content using an emotion engine. Specifically, it uses emotion analysis technology to determine the emotional state, such as positive, negative, or neutral.

[0265] 4. Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry. Based on this categorization, the AI ​​engine generates the optimal answer corresponding to the inquiry.

[0266] 5. Server: Based on the analyzed content, the server retrieves relevant data from the database, such as inventory information, price information, etc., depending on the subject of the inquiry.

[0267] 6. Server: Generates a response email by embedding the acquired data using a template. Template engines such as Jinja2 can be used. Furthermore, the tone and content of the response email are adjusted based on the analysis results of the emotion engine.

[0268] 7. Server: Sends the generated response email to the user. An automatic reply email is sent to the user's email address using an SMTP server.

[0269] Specific examples of processing

[0270] Example 1: Product inventory inquiry

[0271] User: A company representative sends an email inquiring about the stock status of Product A.

[0272] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0273] AI engine: Analyzes the email content and determines that the main question is about stock status. The emotion engine recognizes the user's emotion as negative. It retrieves stock information for product A from the database and generates an answer based on that information.

[0274] Generated answer:

[0275] "Product A's current stock status is as follows:

[0276] Number in stock: 150

[0277] Next arrival: October 15, 2023

[0278] We apologize for keeping you waiting and appreciate your understanding."

[0279] This email will include proposal materials such as an operation manual for the inventory management system and a detailed catalog for Product A.

[0280] Server: The generated email is automatically sent to the company representative.

[0281] Example 2: Price enquiry

[0282] User: A company representative sends an email inquiring about the price of Product B.

[0283] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0284] AI engine: Analyzes the email content and determines that the main topic of the inquiry is "price." The emotion engine recognizes the user's emotion as "positive." It retrieves the latest price information for Product B from the price information database and generates an answer based on that information.

[0285] Generated answer:

[0286] "Product B's latest price information is as follows:

[0287] Unit price: 5,000 yen

[0288] Discount information: 10% off this month only

[0289] Please refer to the price list for details. We look forward to your continued support.

[0290] This email will include proposal materials such as a price list and details of discount campaigns.

[0291] Server: The generated email is automatically sent to the company representative.

[0292] This allows the system of the present invention to respond to inquiries from companies in a prompt and appropriate manner, taking into consideration the feelings of the user.

[0293] The processing flow will be explained below.

[0294] Step 1:

[0295] User: A company representative creates and sends an inquiry email.

[0296] Step 2:

[0297] Server: The mail server receives the inquiry email.

[0298] The SMTP server listens for emails on port 25, stores them in the inbox, and converts the email content into text format.

[0299] Step 3:

[0300] Server: Sends the email converted to text format to the AI ​​engine.

[0301] The email text is sent as a REST API request to the AI ​​engine for analysis.

[0302] Step 4:

[0303] AI engine: Analyzes email content using natural language processing technology.

[0304] Tokenization: Breaking down email content into words.

[0305] Part-of-speech tagging: tag each word with its part of speech.

[0306] Topic extraction: Extracting keywords and phrases to identify the subject of a query, using techniques such as morphological analyzers, POS taggers, TF-IDF, and BERT.

[0307] Step 5:

[0308] Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry.

[0309] Based on the analysis results, the products are divided into categories such as "inventory," "price," and "specifications."

[0310] Step 6:

[0311] Server: Sends text formatted email to the emotion engine and analyzes the user's emotions.

[0312] Sentiment analysis techniques are used to determine emotional states such as positive, negative, or neutral.

[0313] Step 7:

[0314] Server: Select the appropriate template according to the category.

[0315] For example, an "inventory" inquiry uses an availability response template, and a "price" inquiry uses a price information response template.

[0316] Step 8:

[0317] Server: Retrieves the required data from the database.

[0318] Use Python scripts and SQL queries to pull relevant data from the database, such as inventory information, pricing information, etc.

[0319] Step 9:

[0320] Server: The acquired data is embedded in the template and a reply email is generated.

[0321] Use a template engine such as Jinja2 to create emails in a natural style. Adjust the tone and content of the response email based on the analysis results of the emotion engine.

[0322] Step 10:

[0323] Server: The generated response email is sent to the user.

[0324] Use an email sending module (such as smtplib) to send an automatic reply email to the user's email address via the SMTP server.

[0325] Step 11:

[0326] User: The company representative receives the automatic reply email and checks the contents.

[0327] Example 2

[0328] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0329] In conventional inquiry email response systems, analyzing email content and generating appropriate responses is often done manually, making it difficult to respond quickly and accurately.In addition, it is not possible to adjust the tone or content of responses according to the user's emotions, making it difficult to improve user satisfaction.

[0330] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0331] In this invention, the server includes means for receiving an inquiry email, means for analyzing the content of the inquiry email using natural language processing technology, means for recognizing the emotional state from the email content, means for retrieving necessary data from a database based on the analyzed content, means for generating an appropriate response based on the analyzed content and the retrieved data, and means for transmitting the generated response. This enables quick and accurate response to inquiries and can generate responses that correspond to the user's emotions, which is expected to improve user satisfaction.

[0332] An "inquiry email" is an email sent by a user (company representative) to inquire about information from the system.

[0333] "Natural language processing technology" refers to a set of techniques and algorithms that enable computers to understand and process human language.

[0334] The "emotional state" refers to the emotional state that the user is in when sending the inquiry email, and includes positive, negative, neutral, and the like.

[0335] A "database" is a structured collection of data that allows a system to store and retrieve required information.

[0336] An "appropriate response" is an email that is automatically generated based on the contents of the inquiry email and provides answers and information to the user's questions.

[0337] "Adjusting tone and expression" means changing the content of the reply email depending on the user's emotional state and providing information in an appropriate manner.

[0338] "Emotion analysis technology" is a technology for identifying and classifying user emotions from text data.

[0339] This invention is a system for responding promptly and appropriately to inquiries from companies via email. This system fully automates the process of receiving, analyzing, generating responses, and replying to emails. It also has the ability to recognize the user's emotional state and adjust the tone and content of the response based on that state.

[0340] System Configuration

[0341] The system includes the following components:

[0342] Server: Hosts the mail server and AI engine, receives and analyzes inquiry emails, generates and sends responses, and is also equipped with a sentiment analysis engine to analyze user sentiment.

[0343] Terminal: A personal computer or mobile terminal used by a user (company representative). Sends inquiry emails and receives automatic reply emails from the server.

[0344] User: A company representative who is responsible for sending inquiry emails to the system.

[0345] Program processing

[0346] The system processes an inquiry email as follows:

[0347] Hardware and Software

[0348] Server: An IMAP server is used to receive inquiry emails. Natural language processing technology (e.g., spaCy) is used to analyze email content, and a sentiment analysis engine (e.g., VADER) is used for sentiment analysis. A database management system such as MySQL is used for the database, and a template engine (e.g., Jinja2) is used to generate response emails. An SMTP server is used to send emails.

[0349] Specific examples

[0350] Example 1: Product inventory inquiry

[0351] User: A company representative sends an email inquiring about the stock status of Product A.

[0352] Server: Receives the inquiry email and sends the email text to the natural language processing engine.

[0353] Natural language processing engine: Analyzes the email content and determines that the main question is about stock status. The sentiment analysis engine recognizes the user's sentiment as negative. The engine retrieves stock information for product A from the database and generates a response based on that information.

[0354] Generated answer:

[0355] text

[0356] The current stock status of product A is as follows:

[0357] Quantity in stock: 150

[0358] Next expected arrival date: October 15, 2023

[0359] We apologize for the wait and appreciate your understanding.

[0360] This email will include proposal materials such as an inventory management system operation manual and a detailed catalog for product A.

[0361] Server: The generated email is automatically sent to the company representative.

[0362] Example 2: Price enquiry

[0363] User: A company representative sends an email inquiring about the price of Product B.

[0364] Server: Receives the inquiry email and sends the email text to the natural language processing engine.

[0365] Natural language processing engine: Analyzes the email content and determines that "price" is the main topic of the inquiry. The sentiment analysis engine recognizes the user's sentiment as "positive." The latest price information for Product B is retrieved from the price information database and an answer is generated based on that information.

[0366] Generated answer:

[0367] text

[0368] The latest price information for product B is as follows:

[0369] Unit price: 5,000 yen

[0370] Discount information: 10% off this month only

[0371] Please refer to the price list for details. We look forward to your continued support.

[0372] This email will include a price list and details of discount campaigns as proposal materials.

[0373] Server: The generated email is automatically sent to the company representative.

[0374] Prompt Sentence Examples

[0375] Below are some example prompts to input to a generative AI model:

[0376] text

[0377] Design a system that analyzes user inquiry emails, recognizes their emotional state, and generates the best possible response. However, use a template engine to generate the response email and adjust the tone depending on the emotional state.

[0378] This allows the system of the present invention to respond to inquiry emails from companies quickly and accurately while taking into consideration the user's feelings.

[0379] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0380] Step 1: Receiving an inquiry email

[0381] Server: Receives inquiry emails sent by users (company personnel) via the IMAP server. The emails are retrieved from a specific mailbox and the character code is converted to UTF-8.

[0382] Input: New emails from the IMAP mail server

[0383] Output: UTF-8 formatted email text

[0384] Specific operation: Connect to the IMAP server, check for new emails, and retrieve them. Convert the character code to UTF-8.

[0385] Step 2: Analyzing the email content

[0386] Server: The received email text is sent to a natural language processing engine (e.g., spaCy) to analyze the query. The analysis involves tokenization, part-of-speech tagging, and topic extraction.

[0387] Input: UTF-8 formatted email text

[0388] Output: Parsed email content (tokens, part-of-speech tags, subject)

[0389] What it does: Tokenizes email text, tags each word with a part-of-speech tag, and extracts key keywords and phrases.

[0390] Step 3: Sentiment Analysis

[0391] Server: The analyzed email content is sent to an emotion analysis engine (e.g., VADER) to analyze the user's emotional state.

[0392] Input: Parsed email content (tokens, part-of-speech tags, subject)

[0393] Output: Emotional state (positive, negative, neutral)

[0394] Specific operation: The analyzed email content is input into an emotion analysis engine to determine the emotional state.

[0395] Step 4: Categorize the inquiry

[0396] Server: Based on the analysis results of the natural language processing engine, the inquiry content is categorized (e.g., "stock status," "price," etc.).

[0397] Input: Parsed email content, emotional state

[0398] Output: Category information (availability, price, etc.)

[0399] What it does: Maps email content to specific categories and selects the appropriate response template.

[0400] Step 5: Retrieve related data

[0401] Server: Based on the categorized inquiry, retrieves the necessary information from the database, such as stock information and price information.

[0402] Input: Category information

[0403] Output: Retrieved data (stock information, price information, etc.)

[0404] Specific operation: Executes a database query, retrieves the corresponding data, and stores the retrieved data in internal memory.

[0405] Step 6: Generate a response email

[0406] Server: Use a template engine (e.g., Jinja2) to embed the acquired data into the response email. Adjust the tone and content of the response email based on the results of sentiment analysis.

[0407] Input: Template, captured data, emotional state

[0408] Output: Generated response email

[0409] What it does: Embed data in templates to tailor the wording and tone of emails based on emotional state.

[0410] Step 7: Sending an email

[0411] Server: The generated response email is sent to the user via the SMTP server.

[0412] Input: Generated response email

[0413] Output: Email sent

[0414] Specific operation: Connect to the SMTP server, add the generated email to the sending queue, and send it to the user's email address.

[0415] (Application example 2)

[0416] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0417] Current inquiry response systems are required to respond quickly and appropriately to inquiry emails from users. However, previous systems provided mechanical responses without taking emotions into account, which often resulted in a poor user experience. Particularly in the field of e-commerce, flexible responses that match the user's emotions are important, and this aspect needs to be improved.

[0418] The specific processing 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 receiving an inquiry email, means for analyzing the content of the inquiry email, means for generating an appropriate response, means for sending the generated response, means for analyzing the user's emotions, and means for adjusting the tone and content of the response based on the analyzed user's emotions. This enables a quick and appropriate response that takes the user's emotions into consideration.

[0419] An "inquiry email" is an electronic message sent by a user to a business or organization seeking information or support.

[0420] "Means of analysis" refers to the processes and techniques used to understand the content of received inquiry emails and identify their meaning and intent.

[0421] "Natural language processing technology" refers to a series of technologies and methods that allow computers to understand, analyze, and generate human language.

[0422] "Means for generating an appropriate answer" refers to a process for automatically creating the most appropriate answer to a user's inquiry based on the analysis results.

[0423] "Means for analyzing user emotions" refers to a technology that determines the user's emotional state from the content of the inquiry email and classifies it as a numerical value or category.

[0424] "Methods for adjusting the tone and content of responses" refers to techniques for changing the wording and expression of responses to match the analyzed emotional state.

[0425] A "database" is an information system used to manage information related to inquiries (e.g., product inventory, pricing information, etc.).

[0426] "Mobile devices such as smartphones" refers to devices that are portable and can connect to the Internet and use applications.

[0427] MODE FOR CARRYING OUT THE INVENTION

[0428] This invention is an automatic response system for responding quickly and accurately to inquiry emails from companies. This system automatically performs a series of processes, from receiving the inquiry email, analyzing its content, generating and sending a response, and even adjusting the tone of the response to take the user's emotions into consideration.

[0429] System Configuration

[0430] The system includes the following components:

[0431] 1. Server: Equipped with an emotion engine that receives and analyzes inquiry emails, generates and sends responses, and analyzes user emotions.

[0432] 2. Terminal: A mobile terminal such as a PC or smartphone used by a user (company representative). It sends an inquiry email and receives an automatic reply email from the server.

[0433] 3. User: Company representative. Responsible for sending inquiry emails to the system.

[0434] Hardware and software used

[0435] Mail server: Gmail API

[0436] AI engine: Google® Cloud Natural Language API

[0437] Sentiment analysis engine: IBM Watson(R) Natural Language Understanding API

[0438] Template engine: Jinja2

[0439] Database: PostgreSQL

[0440] System Operation

[0441] The server first receives the inquiry email. At this time, it retrieves the email using the Gmail API, etc. The received email text is analyzed using the Google Cloud Natural Language API, and tokenization, part-of-speech tagging, and topic extraction are performed.

[0442] The emotional state of the received email is then analyzed using the IBM Watson Natural Language Understanding API, which can determine whether the email content carries a positive, negative, or neutral sentiment.

[0443] Based on the analysis results, an appropriate answer is generated. The answer is generated using the Jinja2 template engine, retrieving the necessary data from the database (PostgreSQL) according to the inquiry content. The tone of the generated answer is adjusted based on the results of sentiment analysis, so that the content takes the user's emotions into consideration.

[0444] Finally, the generated response is automatically sent to the user using an SMTP server, allowing the user to receive a prompt and appropriate response.

[0445] Specific examples

[0446] Example 1: Product inventory inquiry

[0447] A user sends an email inquiry about the stock status of product A. The server receives the inquiry email and sends the email text to the AI ​​engine for analysis. The AI ​​engine determines that "stock status" is the main content of the inquiry, and the emotion engine recognizes the user's emotion as "negative." The server retrieves stock information for product A from the database and generates an answer based on that information. The generated answer includes the following:

[0448] "Product A's current stock status is as follows:

[0449] Number in stock: 150

[0450] Next arrival: October 15, 2023

[0451] We apologize for keeping you waiting and appreciate your understanding."

[0452] Example 2: Price enquiry

[0453] A user sends an email inquiry about the price of product B. The server receives the inquiry email and sends the email text to the AI ​​engine for analysis. The AI ​​engine determines that "price" is the main topic of the inquiry, and the emotion engine recognizes the user's emotion as "positive." The latest price information for product B is retrieved from the price information database, and an answer is generated based on that information. The generated answer includes the following:

[0454] "Product B's latest price information is as follows:

[0455] Unit price: 5,000 yen

[0456] Discount information: 10% off this month only

[0457] Please refer to the price list for details. We look forward to your continued support.

[0458] Prompt Sentence Examples

[0459] "Generate an auto-reply message based on the email sent by the user. The content of the email should be, 'Please let me know if the product is in stock.'"

[0460] This allows the system of the present invention to respond quickly and appropriately while taking into consideration the user's feelings.

[0461] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0462] Step 1:

[0463] The server receives an inquiry email from a user using the Gmail API. At this time, it obtains the email content in text format. The input is the inquiry email sent by the user, and the output is the email content in text format.

[0464] Step 2:

[0465] The server sends the received email text to the Google Cloud Natural Language API for natural language processing. Specific operations include tokenizing the text, tagging parts of speech, and extracting themes to analyze the meaning and intent of the email. The input is the email content in text format, and the output is the analysis results: keywords and themes.

[0466] Step 3:

[0467] The server uses the IBM Watson Natural Language Understanding API to analyze the user's emotions from the email content. Specifically, it uses sentiment analysis technology to determine emotional states such as positive, negative, and neutral as numerical values ​​or categories. The input is the email content in text format, and the output is the result of the sentiment analysis.

[0468] Step 4:

[0469] The server categorizes the query based on the analysis results and retrieves the necessary data from the database. For example, it retrieves inventory and price information from a PostgreSQL database. The input is the subject and keywords resulting from the analysis, and the output is the retrieved related information.

[0470] Step 5:

[0471] The server uses the Jinja2 template engine to embed the retrieved data and generate an appropriate reply email. Furthermore, it adjusts the tone and content of the reply email based on the analysis results of the emotion engine. Specific operations include modifying the email text based on the emotion analysis. The input is the retrieved related information and the emotion analysis results, and the output is the generated reply email.

[0472] Step 6:

[0473] The server automatically sends the generated reply email to the user using the SMTP server. Specifically, it sets the email address, subject, and body, and performs the sending process via the SMTP server. The input is the generated reply email, and the output is the email sent to the user.

[0474] This allows the system to respond to inquiries promptly and appropriately, taking into consideration the user's feelings.

[0475] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0476] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0477] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0478] [Second embodiment]

[0479] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0480] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0481] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0482] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0483] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0484] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0485] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0486] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0487] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0488] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0489] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0490] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0491] This invention provides a system that responds quickly and appropriately to inquiry emails from companies. This system automatically performs a series of processes: receiving an inquiry email, analyzing its content, generating an appropriate response, and sending it as a reply email. The processing of the system's program is explained below in natural language.

[0492] System Configuration

[0493] The system includes the following components:

[0494] Server: Hosts the mail server and AI engine, and receives, analyzes, generates, and sends email inquiries.

[0495] Device: A PC or mobile device used by a user (company representative). Sends an inquiry email and receives an automatic reply email from the server.

[0496] User: Company representative. Responsible for sending inquiry emails to the system.

[0497] Program processing

[0498] 1. Server: Receives inquiry emails. Receives emails sent by users (company personnel) and converts the email text into text format.

[0499] 2. Server: Sends the email text to the AI ​​engine and analyzes the email content using natural language processing. Specifically, it performs the following processes:

[0500] Tokenization: Breaking down email content into words.

[0501] Part-of-speech tagging: tag each word with its part of speech.

[0502] Topic extraction: Extracting keywords and phrases to identify the subject of a query.

[0503] 3. Server: Based on the analysis results, the inquiry is categorized. Based on this categorization, the AI ​​engine generates the optimal answer corresponding to the inquiry.

[0504] 4. Server: Retrieves relevant data from the database as needed, for example, stock information, pricing information, or other data depending on the subject of the query.

[0505] 5. Server: Using templates, embed the acquired data and generate a reply email in a natural style. Template engines such as Jinja2 can be used.

[0506] 6. Server: The generated response email is sent to the user (company representative). The sending process is performed using an SMTP server.

[0507] Specific examples of processing

[0508] Example 1: Product inventory inquiry

[0509] User: A company representative sends an email inquiring about the stock status of Product A.

[0510] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0511] AI engine: Analyzes the email content and determines that the main question is about stock status. It retrieves stock information for product A from the database and generates a response based on that information.

[0512] Generated answer:

[0513] "Product A's current stock status is as follows:

[0514] Number in stock: 150

[0515] Next arrival: October 15, 2023

[0516] Thank you for your confirmation."

[0517] This email will include proposal materials such as an operation manual for the inventory management system and a detailed catalog for Product A.

[0518] Server: The generated email is automatically sent to the company representative.

[0519] Example 2: Price enquiry

[0520] User: A company representative sends an email inquiring about the price of Product B.

[0521] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0522] AI engine: Analyzes the email content and determines that the main question is about price. It retrieves the latest price information for Product B from a price information database and generates a response based on that information.

[0523] Generated answer:

[0524] "Product B's latest price information is as follows:

[0525] Unit price: 5,000 yen

[0526] Discount information: 10% off this month only

[0527] Please refer to the price list for details. If you have any questions, please let us know.

[0528] This email will include proposal materials such as a price list and details of discount campaigns.

[0529] Server: The generated email is automatically sent to the company representative.

[0530] This allows the system of the present invention to respond quickly and appropriately to inquiry emails from companies.

[0531] The processing flow will be explained below.

[0532] Step 1:

[0533] User: A company representative creates and sends an inquiry email.

[0534] Step 2:

[0535] Server: The mail server receives the inquiry email.

[0536] The SMTP server listens for emails on port 25, stores them in the inbox, and converts the email content into text format.

[0537] Step 3:

[0538] Server: Sends the email converted to text format to the AI ​​engine.

[0539] The email text is sent as a REST API request to the AI ​​engine for analysis.

[0540] Step 4:

[0541] AI engine: Analyzes email content using natural language processing technology.

[0542] It performs tokenization (dividing email content into words), part-of-speech tagging (tagging each word with its part of speech), and subject extraction (extracting keywords and phrases to identify the subject of the inquiry), using technologies such as morphological analyzers, POS taggers, TF-IDF, and BERT.

[0543] Step 5:

[0544] Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry.

[0545] Based on the analysis results, the products are divided into categories such as "inventory," "price," and "specifications."

[0546] Step 6:

[0547] Server: Select the appropriate template according to the category.

[0548] For example, an "inventory" inquiry uses an availability response template, and a "price" inquiry uses a price information response template.

[0549] Step 7:

[0550] Server: Retrieves the required data from the database.

[0551] Use Python scripts and SQL queries to pull relevant data from the database, such as inventory information, pricing information, etc.

[0552] Step 8:

[0553] Server: The acquired data is embedded in the template and a reply email is generated.

[0554] Use a template engine such as Jinja2 to create emails in a natural writing style.

[0555] Step 9:

[0556] Server: The generated response email is sent to the user.

[0557] Use an email sending module (such as smtplib) to send an automatic reply email to the user's email address via the SMTP server.

[0558] Step 10:

[0559] User: The company representative receives the automatic reply email and checks the contents.

[0560] Example 1

[0561] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0562] A problem with conventional inquiry response systems is that it takes a great deal of time and effort to analyze the content of received emails and then generate and send appropriate responses based on that information. Furthermore, manual responses are prone to human error and inconsistencies, making it difficult to provide effective and consistent responses. Furthermore, there is a need for system automation to quickly obtain appropriate data and reflect it in responses.

[0563] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0564] In this invention, the server includes a means for receiving an inquiry email, a means for analyzing the content of the inquiry email using natural language processing technology, a means for categorizing the inquiry content based on the analysis result, a means for retrieving related data from a database according to the categorized content, a means for generating an appropriate response using the retrieved data, and a means for transmitting the generated response. This automates the inquiry response process, enabling quick and accurate responses. Furthermore, the ability to generate consistent responses improves the quality of a company's inquiry responses.

[0565] An "inquiry email" is an email containing a question or request that a user sends to the system.

[0566] "Means of receiving" refers to the functions and processes by which the server receives and stores inquiry emails.

[0567] "Means for analysis" refers to the functions and processes for analyzing the content of received inquiry emails using natural language processing technology and extracting important information.

[0568] "Natural language processing technology" is a field of computer science that involves understanding and analyzing the meaning of text data, and is a technology that performs tasks such as subject extraction and part-of-speech tagging.

[0569] A "generative AI model" is an artificial intelligence model that generates answers in natural language based on input data.

[0570] A "prompt" is an instruction or cue provided to an AI model when it performs analysis.

[0571] "Means of categorization" refers to the functions and processes for classifying the content of inquiries based on the analysis results and determining the appropriate processing method.

[0572] A database is a system for organizing and storing large amounts of information and quickly retrieving data when needed.

[0573] "Means of obtaining relevant data" refers to the functions and processes for extracting the necessary information from the database in response to the inquiry.

[0574] "Answer generation means" refers to the functions and processes that automatically generate an appropriate reply to the user based on the relevant data obtained.

[0575] A "template engine" is software that uses template files to embed data in order to generate dynamic content.

[0576] The "means of sending" refers to the function or process for delivering the generated reply email to the user, and uses an SMTP server or the like.

[0577] A "system" is an integrated device or software that automatically performs a series of processes including receiving, analyzing, classifying, acquiring data, generating responses, and sending inquiries.

[0578] This invention is an automated system for responding promptly and appropriately to inquiry emails from companies. This system performs a series of automated processes: receiving inquiry emails, analyzing their contents, generating responses, and sending the responses.

[0579] System Configuration

[0580] The system includes the following components:

[0581] Server: Hosts the mail server and natural language processing engine, receives and analyzes inquiry emails, generates responses, and sends them. Postfix is ​​used as the mail server, and OpenAI GPT-3 is used as the natural language processing engine.

[0582] Device: A computer or mobile device used by a user (company representative). For example, a Windows PC or iPhone.

[0583] User: Company representative. Responsible for sending inquiry emails to the system.

[0584] Program processing

[0585] The server responds to the inquiry email by following these steps:

[0586] 1. Receiving an inquiry email:

[0587] A user sends an inquiry email using a PC or mobile device. The server receives and stores the inquiry email using the Postfix mail server.

[0588] 2. Email content analysis:

[0589] The server sends the received email text to an AI engine for natural language processing (OpenAI GPT-3) using a prompt. An example of a specific prompt is "What is the subject of this email? I would like to check the stock of product A." The natural language processing engine performs the following process:

[0590] Tokenization: Splitting the text of an email into words. For example, "I would like to check the stock of product A" would be split into "I would like to check the stock of product A."

[0591] Part-of-speech tagging: tag each word with its part of speech, for example, "Product A / noun | of / particle | inventory / noun | wo / particle | confirm / verb | shi / auxiliary verb | tai / auxiliary verb | desu / auxiliary verb."

[0592] Subject Extraction: Use TF-IDF and other natural language processing algorithms to extract the subject of the query. For example, identify that "stock availability" is the subject.

[0593] 3. Categorize your inquiry:

[0594] Based on the analysis results obtained from the AI ​​engine, the server classifies the inquiry into appropriate categories (e.g., stock confirmation, price confirmation).

[0595] 4. Obtaining relevant data:

[0596] The server sends a query to the MySQL database based on the categorized inquiry to retrieve the appropriate information. For example, it executes a query such as "SELECT stock FROM products WHERE product_name="Product A"" to retrieve stock information for Product A.

[0597] 5. Generate reply email:

[0598] The server uses the Jinja2 template engine to embed the acquired data into a template and generate a reply email in a natural style. For example, the stock quantity data is embedded in a template such as "The stock quantity is {{ stock}} units."

[0599] 6. Sending response email:

[0600] The server sends the generated response email to the user (company representative) using the SMTP server (Postfix). If the email was sent successfully, it is recorded in the log.

[0601] Specific examples

[0602] Example 1: Product inventory inquiry

[0603] User: A company representative sends an email inquiring about the stock status of Product A.

[0604] Server: The Postfix mail server receives the email, converts it to email text, and then sends the email text to the AI ​​engine with the following prompt:

[0605] "What is the subject of this email?: I would like to check the availability of product A."

[0606] AI engine: Analyzes the email content and determines that the subject is "stock status."

[0607] Server: Based on the analysis results of the AI ​​engine, classify this inquiry into the "Stock Check" category. Then, send a query to the database saying "Provide the stock quantity of product A."

[0608] Server: Retrieve stock information for Product A (for example, stock quantity 150) from the database and generate a response email using the Jinja2 template engine.

[0609] "Product A's current stock status is as follows:

[0610] Number in stock: 150

[0611] Next arrival: October 15, 2023

[0612] Thank you for your confirmation."

[0613] Server: The generated response email is sent to the user (company representative) using an SMTP server.

[0614] Example 2: Price enquiry

[0615] User: A company representative sends an email inquiring about the price of Product B.

[0616] Server: The Postfix mail server receives the email, converts it to email text, and then sends the email text to the AI ​​engine with the following prompt:

[0617] "What is the subject of this email? I would like to know the price of product B."

[0618] AI engine: Analyzes the content of the email and determines that the subject is "price."

[0619] Server: Based on the analysis results of the AI ​​engine, the server classifies this inquiry into the "price check" category. It then sends a query to the database saying, "Provide the price of product B."

[0620] Server: Retrieve the price information of Product B (for example, unit price 5,000 yen, discount 10%) from the database and generate a reply email using the Jinja2 template engine.

[0621] "Product B's latest price information is as follows:

[0622] Unit price: 5,000 yen

[0623] Discount information: 10% off this month only

[0624] Please refer to the price list for details. If you have any questions, please let us know.

[0625] Server: The generated response email is sent to the user (company representative) using an SMTP server.

[0626] In this way, the system of the present invention can respond quickly and appropriately to inquiry emails from companies.

[0627] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0628] Step 1: Receiving an inquiry email

[0629] Server: A user (company employee) sends an inquiry email. The Postfix mail server receives the email. The input is the email sent by the user, and the output is the email data stored on the mail server. Specifically, the server converts the received email into a specific format (e.g., a text file).

[0630] Step 2: Analyzing the email content

[0631] Server: Sends the text of the received email to an AI engine for natural language processing (OpenAI GPT-3). The input is the text content of the email, and the output is the analysis result. The specific operation is as follows:

[0632] Tokenization: Splitting text into words (e.g., "I would like to check the stock of product A" → "I would like to check the stock of product A").

[0633] Part-of-speech tagging: Tag each word with its part of speech (e.g., "Product A / noun | of / particle | inventory / noun | wo / particle | confirm / verb | shi / auxiliary verb | tai / auxiliary verb | desu / auxiliary verb").

[0634] Topic extraction: TF-IDF is used to extract keywords and identify topics (e.g., "stock status").

[0635] Step 3: Categorize the inquiry

[0636] Server: Categorizes the inquiry content based on the analysis results of the AI ​​engine. The input is the analysis result, and the output is category information (e.g., "Check stock availability" or "Check price"). The specific operation is to classify the inquiry into predefined categories based on the subject.

[0637] Step 4: Retrieve related data

[0638] Server: Sends a query to a database (e.g. MySQL) based on the categorization. The input is the category information and the associated data request, and the output is the specific information retrieved from the database (e.g., the number of items in stock for product A). The specific operation is as follows:

[0639] Create a query: Create an SQL query based on the subject of your inquiry (e.g., "SELECT stock FROM products WHERE product_name="Product A"").

[0640] Data Acquisition: Pulling the required information from the database.

[0641] Step 5: Generate a response email

[0642] Server: Generates a reply email using the Jinja2 template engine based on the retrieved data. The input is the information retrieved from the database and a template, and the output is a reply email. The specific operations are as follows:

[0643] Embed data: Embed data in the template (e.g., "Your stock is {{ stock}}").

[0644] Email formatting: Format your emails in a natural style.

[0645] Step 6: Sending a response email

[0646] Server: The generated reply email is sent to the user (company representative) using the SMTP server (Postfix). The input is the reply email, and the output is a confirmation of successful sending. The specific operations are as follows:

[0647] Sending email: Sends email using the SMTP protocol.

[0648] Log: Logs successful transmission.

[0649] (Application example 1)

[0650] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0651] Modern electronic payment services are receiving an increasing number of inquiries from users, and they need to respond to these inquiries quickly and accurately. However, because the content of inquiries varies widely, manually responding to all inquiries is difficult, time-consuming, and costly. Furthermore, there is a lack of systems that can provide appropriate transaction information based on the content of inquiries and avoid inappropriate responses. Therefore, there is a need to develop a system that automates and streamlines inquiry responses in electronic payment services.

[0652] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0653] In this invention, the server includes means for receiving an inquiry message, means for analyzing the contents of the inquiry message, means for generating an appropriate response based on the analyzed content, means for transmitting the generated response, means for retrieving transaction information from a database based on the analyzed content, and means for customizing templates for the response, thereby enabling prompt and accurate responses to a variety of inquiries in electronic payment services.

[0654] An "inquiry message" is a message containing a question or confirmation that a user makes to a service provider.

[0655] "Analysis" refers to the methods and results used to examine and understand the content of given data or messages.

[0656] A "response" is a reply or answer provided to an inquiry or question.

[0657] "Sending" is the act of delivering data or messages to another device or network destination.

[0658] "Transaction information" refers to detailed data about a specific transaction, primarily in electronic payment services.

[0659] A "database" is an organized collection of data and a system that allows for efficient searching and retrieval.

[0660] A "template" is a document or file with a prescribed format or pattern, designed so that parts of its content can be easily changed.

[0661] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language.

[0662] A "generative AI model" is an artificial intelligence algorithm that generates optimal outputs for a task based on input data.

[0663] A "prompt sentence" is a sentence that contains instructions or a request to perform a specific task or action.

[0664] The present invention relates to a system that automates inquiries about electronic payment services and provides prompt and accurate responses. This system is comprised of multiple servers, terminals, and users, and is described in detail below.

[0665] System Configuration

[0666] The system includes the following components:

[0667] Server: The central component that receives query messages, analyzes their content, generates appropriate responses, and sends them. The server also hosts an AI engine and has the ability to analyze the content of messages using natural language processing techniques.

[0668] Terminal: A communication device such as a smartphone or PC used by a user, whose main role is to input inquiry messages and receive automatic responses from the server.

[0669] User: A person or company representative who uses electronic payment services and sends inquiry messages.

[0670] Technology and software used

[0671] The server performs natural language processing using a generative AI model based on Hugging Face's Transformers library, and uses an SQLite database to manage necessary data such as transaction and price information, and retrieves the appropriate data depending on the inquiry.

[0672] Program processing

[0673] When the server receives a query message, it converts it into text format and sends it to the AI ​​engine. The AI ​​engine analyzes the content of the query message and performs appropriate tokenization, part-of-speech tagging, and topic extraction. Based on the analysis results, it then classifies the query content into an appropriate category and retrieves related data (e.g., transaction information and price information) from the database.

[0674] The server then combines the retrieved data with the template to generate a natural-sounding response, possibly using a template engine (e.g., Jinja2), which is then automatically sent to the user's device.

[0675] Examples and prompts

[0676] As a concrete example, if a user makes a query such as "What is the transaction status of product A?", the server will act as follows:

[0677] 1. User: Enter "Please tell me the transaction status of product A" on your smartphone and send it.

[0678] 2. Server: Receives the message and analyzes its contents. The AI ​​engine determines that the subject is "transaction status."

[0679] 3. Server: Retrieve the transaction status related to product A from the database and generate a natural response based on it.

[0680] 4. Server: Send the generated response "The latest transaction status for product A is 'Completed'" to the user's terminal.

[0681] Prompt Sentence Examples

[0682] (Prompt example 1)

[0683] User: Please tell me the price of product A.

[0684] System: The price of product A is 5,000 yen. Please check the details.

[0685] (Prompt example 2)

[0686] User: What is the transaction status for Item B?

[0687] System: The latest transaction status for Item B is 'Pending'. We will notify you as soon as there is an update.

[0688] In this way, the system can provide quick and accurate responses to various inquiries about electronic payment services.

[0689] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0690] Step 1:

[0691] Users input and send inquiry messages from their smartphones, PCs, or other devices. The input messages are then sent from the user's device to the server.

[0692] Step 2:

[0693] The server receives the query message and converts the data sent in the form of an email message or chat message into a text format, resulting in text data that can be parsed.

[0694] Step 3:

[0695] The server sends the received text data to the AI ​​engine (generative AI model) and analyzes the content. In this step, the following processes are performed using natural language processing technology:

[0696] Tokenization: Breaking the message content down into words.

[0697] Part-of-speech tagging: tag each word with its part of speech.

[0698] Thematic extraction: Identifying the main topics and keywords of a message.

[0699] The analysis results in themes and keywords related to the content of the query.

[0700] Step 4:

[0701] Based on the analysis, the server classifies the query into an appropriate category, such as "price inquiry," "transaction status inquiry," etc. This classification serves as a preliminary step in generating an appropriate response.

[0702] Step 5:

[0703] The server retrieves the necessary information from the database based on the category. For example, if the query is "price inquiry," it retrieves the relevant product price information from the database. Specifically, it executes an SQL query to extract the necessary data, such as price information and transaction status.

[0704] Step 6:

[0705] The server uses the retrieved data and the template to generate a response in a natural style. Using a template engine (such as Jinja2), the server embeds the data in the template to create a response. This response is an appropriate response to the query.

[0706] Step 7:

[0707] The server then sends the generated response to the user's device. The response message is delivered to the user using the SMTP protocol or the chat application's API. This process allows the user to receive a prompt and appropriate response.

[0708] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0709] The present invention provides a system that responds quickly and appropriately to inquiry emails from companies. This system automatically performs a series of processes: receiving an inquiry email, analyzing its content, generating an appropriate response, and sending it as a reply email. Furthermore, by combining this system with an emotion engine that recognizes the user's emotions, the system can adjust the tone and content of the response to match the user's emotions. The system's program processing is explained below in natural language.

[0710] System Configuration

[0711] The system includes the following components:

[0712] Server: Hosts the mail server and AI engine, receives and analyzes inquiry emails, generates and sends responses, and is also equipped with an emotion engine to analyze user emotions.

[0713] Device: A PC or mobile device used by a user (company representative). Sends an inquiry email and receives an automatic reply email from the server.

[0714] User: Company representative. Responsible for sending inquiry emails to the system.

[0715] Program processing

[0716] 1. Server: Receives inquiry emails. Receives emails sent by users (company personnel) and converts the email text into text format.

[0717] 2. Server: Sends the email text to the AI ​​engine and analyzes the email content using natural language processing. Specifically, it performs the following processes:

[0718] Tokenization: Breaking down email content into words.

[0719] Part-of-speech tagging: tag each word with its part of speech.

[0720] Topic extraction: Extracting keywords and phrases to identify the subject of a query.

[0721] 3. Server: Analyzes the user's emotions from the email content using an emotion engine. Specifically, it uses emotion analysis technology to determine the emotional state, such as positive, negative, or neutral.

[0722] 4. Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry. Based on this categorization, the AI ​​engine generates the optimal answer corresponding to the inquiry.

[0723] 5. Server: Based on the analyzed content, the server retrieves relevant data from the database, such as inventory information, price information, etc., depending on the subject of the inquiry.

[0724] 6. Server: Generates a response email by embedding the acquired data using a template. Template engines such as Jinja2 can be used. Furthermore, the tone and content of the response email are adjusted based on the analysis results of the emotion engine.

[0725] 7. Server: Sends the generated response email to the user. An automatic reply email is sent to the user's email address using an SMTP server.

[0726] Specific examples of processing

[0727] Example 1: Product inventory inquiry

[0728] User: A company representative sends an email inquiring about the stock status of Product A.

[0729] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0730] AI engine: Analyzes the email content and determines that the main question is about stock status. The emotion engine recognizes the user's emotion as negative. It retrieves stock information for product A from the database and generates an answer based on that information.

[0731] Generated answer:

[0732] "Product A's current stock status is as follows:

[0733] Number in stock: 150

[0734] Next arrival: October 15, 2023

[0735] We apologize for keeping you waiting and appreciate your understanding."

[0736] This email will include proposal materials such as an operation manual for the inventory management system and a detailed catalog for Product A.

[0737] Server: The generated email is automatically sent to the company representative.

[0738] Example 2: Price enquiry

[0739] User: A company representative sends an email inquiring about the price of Product B.

[0740] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0741] AI engine: Analyzes the email content and determines that the main topic of the inquiry is "price." The emotion engine recognizes the user's emotion as "positive." It retrieves the latest price information for Product B from the price information database and generates an answer based on that information.

[0742] Generated answer:

[0743] "Product B's latest price information is as follows:

[0744] Unit price: 5,000 yen

[0745] Discount information: 10% off this month only

[0746] Please refer to the price list for details. We look forward to your continued support.

[0747] This email will include proposal materials such as a price list and details of discount campaigns.

[0748] Server: The generated email is automatically sent to the company representative.

[0749] This allows the system of the present invention to respond to inquiries from companies in a prompt and appropriate manner, taking into consideration the feelings of the user.

[0750] The processing flow will be explained below.

[0751] Step 1:

[0752] User: A company representative creates and sends an inquiry email.

[0753] Step 2:

[0754] Server: The mail server receives the inquiry email.

[0755] The SMTP server listens for emails on port 25, stores them in the inbox, and converts the email content into text format.

[0756] Step 3:

[0757] Server: Sends the email converted to text format to the AI ​​engine.

[0758] The email text is sent as a REST API request to the AI ​​engine for analysis.

[0759] Step 4:

[0760] AI engine: Analyzes email content using natural language processing technology.

[0761] Tokenization: Breaking down email content into words.

[0762] Part-of-speech tagging: tag each word with its part of speech.

[0763] Topic extraction: Extracting keywords and phrases to identify the subject of a query, using techniques such as morphological analyzers, POS taggers, TF-IDF, and BERT.

[0764] Step 5:

[0765] Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry.

[0766] Based on the analysis results, the products are divided into categories such as "inventory," "price," and "specifications."

[0767] Step 6:

[0768] Server: Sends text formatted email to the emotion engine and analyzes the user's emotions.

[0769] Sentiment analysis techniques are used to determine emotional states such as positive, negative, or neutral.

[0770] Step 7:

[0771] Server: Select the appropriate template according to the category.

[0772] For example, an "inventory" inquiry uses an availability response template, and a "price" inquiry uses a price information response template.

[0773] Step 8:

[0774] Server: Retrieves the required data from the database.

[0775] Use Python scripts and SQL queries to pull relevant data from the database, such as inventory information, pricing information, etc.

[0776] Step 9:

[0777] Server: The acquired data is embedded in the template and a reply email is generated.

[0778] Use a template engine such as Jinja2 to create emails in a natural style. Adjust the tone and content of the response email based on the analysis results of the emotion engine.

[0779] Step 10:

[0780] Server: The generated response email is sent to the user.

[0781] Use an email sending module (such as smtplib) to send an automatic reply email to the user's email address via the SMTP server.

[0782] Step 11:

[0783] User: The company representative receives the automatic reply email and checks the contents.

[0784] Example 2

[0785] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0786] In conventional inquiry email response systems, analyzing email content and generating appropriate responses is often done manually, making it difficult to respond quickly and accurately.In addition, it is not possible to adjust the tone or content of responses according to the user's emotions, making it difficult to improve user satisfaction.

[0787] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0788] In this invention, the server includes means for receiving an inquiry email, means for analyzing the content of the inquiry email using natural language processing technology, means for recognizing the emotional state from the email content, means for retrieving necessary data from a database based on the analyzed content, means for generating an appropriate response based on the analyzed content and the retrieved data, and means for transmitting the generated response. This enables quick and accurate response to inquiries and can generate responses that correspond to the user's emotions, which is expected to improve user satisfaction.

[0789] An "inquiry email" is an email sent by a user (company representative) to inquire about information from the system.

[0790] "Natural language processing technology" refers to a set of techniques and algorithms that enable computers to understand and process human language.

[0791] The "emotional state" refers to the emotional state that the user is in when sending the inquiry email, and includes positive, negative, neutral, and the like.

[0792] A "database" is a structured collection of data that allows a system to store and retrieve required information.

[0793] An "appropriate response" is an email that is automatically generated based on the contents of the inquiry email and provides answers and information to the user's questions.

[0794] "Adjusting tone and expression" means changing the content of the reply email depending on the user's emotional state and providing information in an appropriate manner.

[0795] "Emotion analysis technology" is a technology for identifying and classifying user emotions from text data.

[0796] This invention is a system for responding promptly and appropriately to inquiries from companies via email. This system fully automates the process of receiving, analyzing, generating responses, and replying to emails. It also has the ability to recognize the user's emotional state and adjust the tone and content of the response based on that state.

[0797] System Configuration

[0798] The system includes the following components:

[0799] Server: Hosts the mail server and AI engine, receives and analyzes inquiry emails, generates and sends responses, and is also equipped with a sentiment analysis engine to analyze user sentiment.

[0800] Terminal: A personal computer or mobile terminal used by a user (company representative). Sends inquiry emails and receives automatic reply emails from the server.

[0801] User: A company representative who is responsible for sending inquiry emails to the system.

[0802] Program processing

[0803] The system processes an inquiry email as follows:

[0804] Hardware and Software

[0805] Server: An IMAP server is used to receive inquiry emails. Natural language processing technology (e.g., spaCy) is used to analyze email content, and a sentiment analysis engine (e.g., VADER) is used for sentiment analysis. A database management system such as MySQL is used for the database, and a template engine (e.g., Jinja2) is used to generate response emails. An SMTP server is used to send emails.

[0806] Specific examples

[0807] Example 1: Product inventory inquiry

[0808] User: A company representative sends an email inquiring about the stock status of Product A.

[0809] Server: Receives the inquiry email and sends the email text to the natural language processing engine.

[0810] Natural language processing engine: Analyzes the email content and determines that the main question is about stock status. The sentiment analysis engine recognizes the user's sentiment as negative. The engine retrieves stock information for product A from the database and generates a response based on that information.

[0811] Generated answer:

[0812] text

[0813] The current stock status of product A is as follows:

[0814] Quantity in stock: 150

[0815] Next expected arrival date: October 15, 2023

[0816] We apologize for the wait and appreciate your understanding.

[0817] This email will include proposal materials such as an inventory management system operation manual and a detailed catalog for product A.

[0818] Server: The generated email is automatically sent to the company representative.

[0819] Example 2: Price enquiry

[0820] User: A company representative sends an email inquiring about the price of Product B.

[0821] Server: Receives the inquiry email and sends the email text to the natural language processing engine.

[0822] Natural language processing engine: Analyzes the email content and determines that "price" is the main topic of the inquiry. The sentiment analysis engine recognizes the user's sentiment as "positive." The latest price information for Product B is retrieved from the price information database and an answer is generated based on that information.

[0823] Generated answer:

[0824] text

[0825] The latest price information for product B is as follows:

[0826] Unit price: 5,000 yen

[0827] Discount information: 10% off this month only

[0828] Please refer to the price list for details. We look forward to your continued support.

[0829] This email will include a price list and details of discount campaigns as proposal materials.

[0830] Server: The generated email is automatically sent to the company representative.

[0831] Prompt Sentence Examples

[0832] Below are some example prompts to input to a generative AI model:

[0833] text

[0834] Design a system that analyzes user inquiry emails, recognizes their emotional state, and generates the best possible response. However, use a template engine to generate the response email and adjust the tone depending on the emotional state.

[0835] This allows the system of the present invention to respond to inquiry emails from companies quickly and accurately while taking into consideration the user's feelings.

[0836] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0837] Step 1: Receiving an inquiry email

[0838] Server: Receives inquiry emails sent by users (company personnel) via the IMAP server. The emails are retrieved from a specific mailbox and the character code is converted to UTF-8.

[0839] Input: New emails from the IMAP mail server

[0840] Output: UTF-8 formatted email text

[0841] Specific operation: Connect to the IMAP server, check for new emails, and retrieve them. Convert the character code to UTF-8.

[0842] Step 2: Analyzing the email content

[0843] Server: The received email text is sent to a natural language processing engine (e.g., spaCy) to analyze the query. The analysis involves tokenization, part-of-speech tagging, and topic extraction.

[0844] Input: UTF-8 formatted email text

[0845] Output: Parsed email content (tokens, part-of-speech tags, subject)

[0846] What it does: Tokenizes email text, tags each word with a part-of-speech tag, and extracts key keywords and phrases.

[0847] Step 3: Sentiment Analysis

[0848] Server: The analyzed email content is sent to an emotion analysis engine (e.g., VADER) to analyze the user's emotional state.

[0849] Input: Parsed email content (tokens, part-of-speech tags, subject)

[0850] Output: Emotional state (positive, negative, neutral)

[0851] Specific operation: The analyzed email content is input into an emotion analysis engine to determine the emotional state.

[0852] Step 4: Categorize the inquiry

[0853] Server: Based on the analysis results of the natural language processing engine, the inquiry content is categorized (e.g., "stock status," "price," etc.).

[0854] Input: Parsed email content, emotional state

[0855] Output: Category information (availability, price, etc.)

[0856] What it does: Maps email content to specific categories and selects the appropriate response template.

[0857] Step 5: Retrieve related data

[0858] Server: Based on the categorized inquiry, retrieves the necessary information from the database, such as stock information and price information.

[0859] Input: Category information

[0860] Output: Retrieved data (stock information, price information, etc.)

[0861] Specific operation: Executes a database query, retrieves the corresponding data, and stores the retrieved data in internal memory.

[0862] Step 6: Generate a response email

[0863] Server: Use a template engine (e.g., Jinja2) to embed the acquired data into the response email. Adjust the tone and content of the response email based on the results of sentiment analysis.

[0864] Input: Template, captured data, emotional state

[0865] Output: Generated response email

[0866] What it does: Embed data in templates to tailor the wording and tone of emails based on emotional state.

[0867] Step 7: Sending an email

[0868] Server: The generated response email is sent to the user via the SMTP server.

[0869] Input: Generated response email

[0870] Output: Email sent

[0871] Specific operation: Connect to the SMTP server, add the generated email to the sending queue, and send it to the user's email address.

[0872] (Application example 2)

[0873] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0874] Current inquiry response systems are required to respond quickly and appropriately to inquiry emails from users. However, previous systems provided mechanical responses without taking emotions into account, which often resulted in a poor user experience. Particularly in the field of e-commerce, flexible responses that match the user's emotions are important, and this aspect needs to be improved.

[0875] The specific processing 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 receiving an inquiry email, means for analyzing the content of the inquiry email, means for generating an appropriate response, means for sending the generated response, means for analyzing the user's emotions, and means for adjusting the tone and content of the response based on the analyzed user's emotions. This enables a quick and appropriate response that takes the user's emotions into consideration.

[0876] An "inquiry email" is an electronic message sent by a user to a business or organization seeking information or support.

[0877] "Means of analysis" refers to the processes and techniques used to understand the content of received inquiry emails and identify their meaning and intent.

[0878] "Natural language processing technology" refers to a series of technologies and methods that allow computers to understand, analyze, and generate human language.

[0879] "Means for generating an appropriate answer" refers to a process for automatically creating the most appropriate answer to a user's inquiry based on the analysis results.

[0880] "Means for analyzing user emotions" refers to a technology that determines the user's emotional state from the content of the inquiry email and classifies it as a numerical value or category.

[0881] "Methods for adjusting the tone and content of responses" refers to techniques for changing the wording and expression of responses to match the analyzed emotional state.

[0882] A "database" is an information system used to manage information related to inquiries (e.g., product inventory, pricing information, etc.).

[0883] "Mobile devices such as smartphones" refers to devices that are portable and can connect to the Internet and use applications.

[0884] MODE FOR CARRYING OUT THE INVENTION

[0885] This invention is an automatic response system for responding quickly and accurately to inquiry emails from companies. This system automatically performs a series of processes, from receiving the inquiry email, analyzing its content, generating and sending a response, and even adjusting the tone of the response to take the user's emotions into consideration.

[0886] System Configuration

[0887] The system includes the following components:

[0888] 1. Server: Equipped with an emotion engine that receives and analyzes inquiry emails, generates and sends responses, and analyzes user emotions.

[0889] 2. Terminal: A mobile terminal such as a PC or smartphone used by a user (company representative). It sends an inquiry email and receives an automatic reply email from the server.

[0890] 3. User: Company representative. Responsible for sending inquiry emails to the system.

[0891] Hardware and software used

[0892] Mail server: Gmail API

[0893] AI engine: Google Cloud Natural Language API

[0894] Sentiment analysis engine: IBM Watson Natural Language Understanding API

[0895] Template engine: Jinja2

[0896] Database: PostgreSQL

[0897] System Operation

[0898] The server first receives the inquiry email. At this time, it retrieves the email using the Gmail API, etc. The received email text is analyzed using the Google Cloud Natural Language API, and tokenization, part-of-speech tagging, and topic extraction are performed.

[0899] The emotional state of the received email is then analyzed using the IBM Watson Natural Language Understanding API, which can determine whether the email content carries a positive, negative, or neutral sentiment.

[0900] Based on the analysis results, an appropriate answer is generated. The answer is generated using the Jinja2 template engine, retrieving the necessary data from the database (PostgreSQL) according to the inquiry content. The tone of the generated answer is adjusted based on the results of sentiment analysis, so that the content takes the user's emotions into consideration.

[0901] Finally, the generated response is automatically sent to the user using an SMTP server, allowing the user to receive a prompt and appropriate response.

[0902] Specific examples

[0903] Example 1: Product inventory inquiry

[0904] A user sends an email inquiry about the stock status of product A. The server receives the inquiry email and sends the email text to the AI ​​engine for analysis. The AI ​​engine determines that "stock status" is the main content of the inquiry, and the emotion engine recognizes the user's emotion as "negative." The server retrieves stock information for product A from the database and generates an answer based on that information. The generated answer includes the following:

[0905] "Product A's current stock status is as follows:

[0906] Number in stock: 150

[0907] Next arrival: October 15, 2023

[0908] We apologize for keeping you waiting and appreciate your understanding."

[0909] Example 2: Price enquiry

[0910] A user sends an email inquiry about the price of product B. The server receives the inquiry email and sends the email text to the AI ​​engine for analysis. The AI ​​engine determines that "price" is the main topic of the inquiry, and the emotion engine recognizes the user's emotion as "positive." The latest price information for product B is retrieved from the price information database, and an answer is generated based on that information. The generated answer includes the following:

[0911] "Product B's latest price information is as follows:

[0912] Unit price: 5,000 yen

[0913] Discount information: 10% off this month only

[0914] Please refer to the price list for details. We look forward to your continued support.

[0915] Prompt Sentence Examples

[0916] "Generate an auto-reply message based on the email sent by the user. The content of the email should be, 'Please let me know if the product is in stock.'"

[0917] This allows the system of the present invention to respond quickly and appropriately while taking into consideration the user's feelings.

[0918] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0919] Step 1:

[0920] The server receives an inquiry email from a user using the Gmail API. At this time, it obtains the email content in text format. The input is the inquiry email sent by the user, and the output is the email content in text format.

[0921] Step 2:

[0922] The server sends the received email text to the Google Cloud Natural Language API for natural language processing. Specific operations include tokenizing the text, tagging parts of speech, and extracting themes to analyze the meaning and intent of the email. The input is the email content in text format, and the output is the analysis results: keywords and themes.

[0923] Step 3:

[0924] The server uses the IBM Watson Natural Language Understanding API to analyze the user's emotions from the email content. Specifically, it uses sentiment analysis technology to determine emotional states such as positive, negative, and neutral as numerical values ​​or categories. The input is the email content in text format, and the output is the result of the sentiment analysis.

[0925] Step 4:

[0926] The server categorizes the query based on the analysis results and retrieves the necessary data from the database. For example, it retrieves inventory and price information from a PostgreSQL database. The input is the subject and keywords resulting from the analysis, and the output is the retrieved related information.

[0927] Step 5:

[0928] The server uses the Jinja2 template engine to embed the retrieved data and generate an appropriate reply email. Furthermore, it adjusts the tone and content of the reply email based on the analysis results of the emotion engine. Specific operations include modifying the email text based on the emotion analysis. The input is the retrieved related information and the emotion analysis results, and the output is the generated reply email.

[0929] Step 6:

[0930] The server automatically sends the generated reply email to the user using the SMTP server. Specifically, it sets the email address, subject, and body, and performs the sending process via the SMTP server. The input is the generated reply email, and the output is the email sent to the user.

[0931] This allows the system to respond to inquiries promptly and appropriately, taking into consideration the user's feelings.

[0932] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0933] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0934] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0935] [Third embodiment]

[0936] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0937] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0938] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0939] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0940] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0941] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0942] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0943] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0944] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0945] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0946] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0947] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0948] This invention provides a system that responds quickly and appropriately to inquiry emails from companies. This system automatically performs a series of processes: receiving an inquiry email, analyzing its content, generating an appropriate response, and sending it as a reply email. The processing of the system's program is explained below in natural language.

[0949] System Configuration

[0950] The system includes the following components:

[0951] Server: Hosts the mail server and AI engine, and receives, analyzes, generates, and sends email inquiries.

[0952] Device: A PC or mobile device used by a user (company representative). Sends an inquiry email and receives an automatic reply email from the server.

[0953] User: Company representative. Responsible for sending inquiry emails to the system.

[0954] Program processing

[0955] 1. Server: Receives inquiry emails. Receives emails sent by users (company personnel) and converts the email text into text format.

[0956] 2. Server: Sends the email text to the AI ​​engine and analyzes the email content using natural language processing. Specifically, it performs the following processes:

[0957] Tokenization: Breaking down email content into words.

[0958] Part-of-speech tagging: tag each word with its part of speech.

[0959] Topic extraction: Extracting keywords and phrases to identify the subject of a query.

[0960] 3. Server: Based on the analysis results, the inquiry is categorized. Based on this categorization, the AI ​​engine generates the optimal answer corresponding to the inquiry.

[0961] 4. Server: Retrieves relevant data from the database as needed, for example, stock information, pricing information, or other data depending on the subject of the query.

[0962] 5. Server: Using templates, embed the acquired data and generate a reply email in a natural style. Template engines such as Jinja2 can be used.

[0963] 6. Server: The generated response email is sent to the user (company representative). The sending process is performed using an SMTP server.

[0964] Specific examples of processing

[0965] Example 1: Product inventory inquiry

[0966] User: A company representative sends an email inquiring about the stock status of Product A.

[0967] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0968] AI engine: Analyzes the email content and determines that the main question is about stock status. It retrieves stock information for product A from the database and generates a response based on that information.

[0969] Generated answer:

[0970] "Product A's current stock status is as follows:

[0971] Number in stock: 150

[0972] Next arrival: October 15, 2023

[0973] Thank you for your confirmation."

[0974] This email will include proposal materials such as an operation manual for the inventory management system and a detailed catalog for Product A.

[0975] Server: The generated email is automatically sent to the company representative.

[0976] Example 2: Price enquiry

[0977] User: A company representative sends an email inquiring about the price of Product B.

[0978] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[0979] AI engine: Analyzes the email content and determines that the main question is about price. It retrieves the latest price information for Product B from a price information database and generates a response based on that information.

[0980] Generated answer:

[0981] "Product B's latest price information is as follows:

[0982] Unit price: 5,000 yen

[0983] Discount information: 10% off this month only

[0984] Please refer to the price list for details. If you have any questions, please let us know.

[0985] This email will include proposal materials such as a price list and details of discount campaigns.

[0986] Server: The generated email is automatically sent to the company representative.

[0987] This allows the system of the present invention to respond quickly and appropriately to inquiry emails from companies.

[0988] The processing flow will be explained below.

[0989] Step 1:

[0990] User: A company representative creates and sends an inquiry email.

[0991] Step 2:

[0992] Server: The mail server receives the inquiry email.

[0993] The SMTP server listens for emails on port 25, stores them in the inbox, and converts the email content into text format.

[0994] Step 3:

[0995] Server: Sends the email converted to text format to the AI ​​engine.

[0996] The email text is sent as a REST API request to the AI ​​engine for analysis.

[0997] Step 4:

[0998] AI engine: Analyzes email content using natural language processing technology.

[0999] It performs tokenization (dividing email content into words), part-of-speech tagging (tagging each word with its part of speech), and subject extraction (extracting keywords and phrases to identify the subject of the inquiry), using technologies such as morphological analyzers, POS taggers, TF-IDF, and BERT.

[1000] Step 5:

[1001] Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry.

[1002] Based on the analysis results, the products are divided into categories such as "inventory," "price," and "specifications."

[1003] Step 6:

[1004] Server: Select the appropriate template according to the category.

[1005] For example, an "inventory" inquiry uses an availability response template, and a "price" inquiry uses a price information response template.

[1006] Step 7:

[1007] Server: Retrieves the required data from the database.

[1008] Use Python scripts and SQL queries to pull relevant data from the database, such as inventory information, pricing information, etc.

[1009] Step 8:

[1010] Server: The acquired data is embedded in the template and a reply email is generated.

[1011] Use a template engine such as Jinja2 to create emails in a natural writing style.

[1012] Step 9:

[1013] Server: The generated response email is sent to the user.

[1014] Use an email sending module (such as smtplib) to send an automatic reply email to the user's email address via the SMTP server.

[1015] Step 10:

[1016] User: The company representative receives the automatic reply email and checks the contents.

[1017] Example 1

[1018] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1019] A problem with conventional inquiry response systems is that it takes a great deal of time and effort to analyze the content of received emails and then generate and send appropriate responses based on that information. Furthermore, manual responses are prone to human error and inconsistencies, making it difficult to provide effective and consistent responses. Furthermore, there is a need for system automation to quickly obtain appropriate data and reflect it in responses.

[1020] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1021] In this invention, the server includes a means for receiving an inquiry email, a means for analyzing the content of the inquiry email using natural language processing technology, a means for categorizing the inquiry content based on the analysis result, a means for retrieving related data from a database according to the categorized content, a means for generating an appropriate response using the retrieved data, and a means for transmitting the generated response. This automates the inquiry response process, enabling quick and accurate responses. Furthermore, the ability to generate consistent responses improves the quality of a company's inquiry responses.

[1022] An "inquiry email" is an email containing a question or request that a user sends to the system.

[1023] "Means of receiving" refers to the functions and processes by which the server receives and stores inquiry emails.

[1024] "Means for analysis" refers to the functions and processes for analyzing the content of received inquiry emails using natural language processing technology and extracting important information.

[1025] "Natural language processing technology" is a field of computer science that involves understanding and analyzing the meaning of text data, and is a technology that performs tasks such as subject extraction and part-of-speech tagging.

[1026] A "generative AI model" is an artificial intelligence model that generates answers in natural language based on input data.

[1027] A "prompt" is an instruction or cue provided to an AI model when it performs analysis.

[1028] "Means of categorization" refers to the functions and processes for classifying the content of inquiries based on the analysis results and determining the appropriate processing method.

[1029] A database is a system for organizing and storing large amounts of information and quickly retrieving data when needed.

[1030] "Means of obtaining relevant data" refers to the functions and processes for extracting the necessary information from the database in response to the inquiry.

[1031] "Answer generation means" refers to the functions and processes that automatically generate an appropriate reply to the user based on the relevant data obtained.

[1032] A "template engine" is software that uses template files to embed data in order to generate dynamic content.

[1033] The "means of sending" refers to the function or process for delivering the generated reply email to the user, and uses an SMTP server or the like.

[1034] A "system" is an integrated device or software that automatically performs a series of processes including receiving, analyzing, classifying, acquiring data, generating responses, and sending inquiries.

[1035] This invention is an automated system for responding promptly and appropriately to inquiry emails from companies. This system performs a series of automated processes: receiving inquiry emails, analyzing their contents, generating responses, and sending the responses.

[1036] System Configuration

[1037] The system includes the following components:

[1038] Server: Hosts the mail server and natural language processing engine, receives and analyzes inquiry emails, generates responses, and sends them. Postfix is ​​used as the mail server, and OpenAI GPT-3 is used as the natural language processing engine.

[1039] Device: A computer or mobile device used by a user (company representative). For example, a Windows PC or iPhone.

[1040] User: Company representative. Responsible for sending inquiry emails to the system.

[1041] Program processing

[1042] The server responds to the inquiry email by following these steps:

[1043] 1. Receiving an inquiry email:

[1044] A user sends an inquiry email using a PC or mobile device. The server receives and stores the inquiry email using the Postfix mail server.

[1045] 2. Email content analysis:

[1046] The server sends the received email text to an AI engine for natural language processing (OpenAI GPT-3) using a prompt. An example of a specific prompt is "What is the subject of this email? I would like to check the stock of product A." The natural language processing engine performs the following process:

[1047] Tokenization: Splitting the text of an email into words. For example, "I would like to check the stock of product A" would be split into "I would like to check the stock of product A."

[1048] Part-of-speech tagging: tag each word with its part of speech, for example, "Product A / noun | of / particle | inventory / noun | wo / particle | confirm / verb | shi / auxiliary verb | tai / auxiliary verb | desu / auxiliary verb."

[1049] Subject Extraction: Use TF-IDF and other natural language processing algorithms to extract the subject of the query. For example, identify that "stock availability" is the subject.

[1050] 3. Categorize your inquiry:

[1051] Based on the analysis results obtained from the AI ​​engine, the server classifies the inquiry into appropriate categories (e.g., stock confirmation, price confirmation).

[1052] 4. Obtaining relevant data:

[1053] The server sends a query to the MySQL database based on the categorized inquiry to retrieve the appropriate information. For example, it executes a query such as "SELECT stock FROM products WHERE product_name="Product A"" to retrieve stock information for Product A.

[1054] 5. Generate reply email:

[1055] The server uses the Jinja2 template engine to embed the acquired data into a template and generate a reply email in a natural style. For example, the stock quantity data is embedded in a template such as "The stock quantity is {{ stock}} units."

[1056] 6. Sending response email:

[1057] The server sends the generated response email to the user (company representative) using the SMTP server (Postfix). If the email was sent successfully, it is recorded in the log.

[1058] Specific examples

[1059] Example 1: Product inventory inquiry

[1060] User: A company representative sends an email inquiring about the stock status of Product A.

[1061] Server: The Postfix mail server receives the email, converts it to email text, and then sends the email text to the AI ​​engine with the following prompt:

[1062] "What is the subject of this email?: I would like to check the availability of product A."

[1063] AI engine: Analyzes the email content and determines that the subject is "stock status."

[1064] Server: Based on the analysis results of the AI ​​engine, classify this inquiry into the "Stock Check" category. Then, send a query to the database saying "Provide the stock quantity of product A."

[1065] Server: Retrieve stock information for Product A (for example, stock quantity 150) from the database and generate a response email using the Jinja2 template engine.

[1066] "Product A's current stock status is as follows:

[1067] Number in stock: 150

[1068] Next arrival: October 15, 2023

[1069] Thank you for your confirmation."

[1070] Server: The generated response email is sent to the user (company representative) using an SMTP server.

[1071] Example 2: Price enquiry

[1072] User: A company representative sends an email inquiring about the price of Product B.

[1073] Server: The Postfix mail server receives the email, converts it to email text, and then sends the email text to the AI ​​engine with the following prompt:

[1074] "What is the subject of this email? I would like to know the price of product B."

[1075] AI engine: Analyzes the content of the email and determines that the subject is "price."

[1076] Server: Based on the analysis results of the AI ​​engine, the server classifies this inquiry into the "price check" category. It then sends a query to the database saying, "Provide the price of product B."

[1077] Server: Retrieve the price information of Product B (for example, unit price 5,000 yen, discount 10%) from the database and generate a reply email using the Jinja2 template engine.

[1078] "Product B's latest price information is as follows:

[1079] Unit price: 5,000 yen

[1080] Discount information: 10% off this month only

[1081] Please refer to the price list for details. If you have any questions, please let us know.

[1082] Server: The generated response email is sent to the user (company representative) using an SMTP server.

[1083] In this way, the system of the present invention can respond quickly and appropriately to inquiry emails from companies.

[1084] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1085] Step 1: Receiving an inquiry email

[1086] Server: A user (company employee) sends an inquiry email. The Postfix mail server receives the email. The input is the email sent by the user, and the output is the email data stored on the mail server. Specifically, the server converts the received email into a specific format (e.g., a text file).

[1087] Step 2: Analyzing the email content

[1088] Server: Sends the text of the received email to an AI engine for natural language processing (OpenAI GPT-3). The input is the text content of the email, and the output is the analysis result. The specific operation is as follows:

[1089] Tokenization: Splitting text into words (e.g., "I would like to check the stock of product A" → "I would like to check the stock of product A").

[1090] Part-of-speech tagging: Tag each word with its part of speech (e.g., "Product A / noun | of / particle | inventory / noun | wo / particle | confirm / verb | shi / auxiliary verb | tai / auxiliary verb | desu / auxiliary verb").

[1091] Topic extraction: TF-IDF is used to extract keywords and identify topics (e.g., "stock status").

[1092] Step 3: Categorize the inquiry

[1093] Server: Categorizes the inquiry content based on the analysis results of the AI ​​engine. The input is the analysis result, and the output is category information (e.g., "Check stock availability" or "Check price"). The specific operation is to classify the inquiry into predefined categories based on the subject.

[1094] Step 4: Retrieve related data

[1095] Server: Sends a query to a database (e.g. MySQL) based on the categorization. The input is the category information and the associated data request, and the output is the specific information retrieved from the database (e.g., the number of items in stock for product A). The specific operation is as follows:

[1096] Create a query: Create an SQL query based on the subject of your inquiry (e.g., "SELECT stock FROM products WHERE product_name="Product A"").

[1097] Data Acquisition: Pulling the required information from the database.

[1098] Step 5: Generate a response email

[1099] Server: Generates a reply email using the Jinja2 template engine based on the retrieved data. The input is the information retrieved from the database and a template, and the output is a reply email. The specific operations are as follows:

[1100] Embed data: Embed data in the template (e.g., "Your stock is {{ stock}}").

[1101] Email formatting: Format your emails in a natural style.

[1102] Step 6: Sending a response email

[1103] Server: The generated reply email is sent to the user (company representative) using the SMTP server (Postfix). The input is the reply email, and the output is a confirmation of successful sending. The specific operations are as follows:

[1104] Sending email: Sends email using the SMTP protocol.

[1105] Log: Logs successful transmission.

[1106] (Application example 1)

[1107] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1108] Modern electronic payment services are receiving an increasing number of inquiries from users, and they need to respond to these inquiries quickly and accurately. However, because the content of inquiries varies widely, manually responding to all inquiries is difficult, time-consuming, and costly. Furthermore, there is a lack of systems that can provide appropriate transaction information based on the content of inquiries and avoid inappropriate responses. Therefore, there is a need to develop a system that automates and streamlines inquiry responses in electronic payment services.

[1109] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1110] In this invention, the server includes means for receiving an inquiry message, means for analyzing the contents of the inquiry message, means for generating an appropriate response based on the analyzed content, means for transmitting the generated response, means for retrieving transaction information from a database based on the analyzed content, and means for customizing templates for the response, thereby enabling prompt and accurate responses to a variety of inquiries in electronic payment services.

[1111] An "inquiry message" is a message containing a question or confirmation that a user makes to a service provider.

[1112] "Analysis" refers to the methods and results used to examine and understand the content of given data or messages.

[1113] A "response" is a reply or answer provided to an inquiry or question.

[1114] "Sending" is the act of delivering data or messages to another device or network destination.

[1115] "Transaction information" refers to detailed data about a specific transaction, primarily in electronic payment services.

[1116] A "database" is an organized collection of data and a system that allows for efficient searching and retrieval.

[1117] A "template" is a document or file with a prescribed format or pattern, designed so that parts of its content can be easily changed.

[1118] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language.

[1119] A "generative AI model" is an artificial intelligence algorithm that generates optimal outputs for a task based on input data.

[1120] A "prompt sentence" is a sentence that contains instructions or a request to perform a specific task or action.

[1121] The present invention relates to a system that automates inquiries about electronic payment services and provides prompt and accurate responses. This system is comprised of multiple servers, terminals, and users, and is described in detail below.

[1122] System Configuration

[1123] The system includes the following components:

[1124] Server: The central component that receives query messages, analyzes their content, generates appropriate responses, and sends them. The server also hosts an AI engine and has the ability to analyze the content of messages using natural language processing techniques.

[1125] Terminal: A communication device such as a smartphone or PC used by a user, whose main role is to input inquiry messages and receive automatic responses from the server.

[1126] User: A person or company representative who uses electronic payment services and sends inquiry messages.

[1127] Technology and software used

[1128] The server performs natural language processing using a generative AI model based on Hugging Face's Transformers library, and uses an SQLite database to manage necessary data such as transaction and price information, and retrieves the appropriate data depending on the inquiry.

[1129] Program processing

[1130] When the server receives a query message, it converts it into text format and sends it to the AI ​​engine. The AI ​​engine analyzes the content of the query message and performs appropriate tokenization, part-of-speech tagging, and topic extraction. Based on the analysis results, it then classifies the query content into an appropriate category and retrieves related data (e.g., transaction information and price information) from the database.

[1131] The server then combines the retrieved data with the template to generate a natural-sounding response, possibly using a template engine (e.g., Jinja2), which is then automatically sent to the user's device.

[1132] Examples and prompts

[1133] As a concrete example, if a user makes a query such as "What is the transaction status of product A?", the server will act as follows:

[1134] 1. User: Enter "Please tell me the transaction status of product A" on your smartphone and send it.

[1135] 2. Server: Receives the message and analyzes its contents. The AI ​​engine determines that the subject is "transaction status."

[1136] 3. Server: Retrieve the transaction status related to product A from the database and generate a natural response based on it.

[1137] 4. Server: Send the generated response "The latest transaction status for product A is 'Completed'" to the user's terminal.

[1138] Prompt Sentence Examples

[1139] (Prompt example 1)

[1140] User: Please tell me the price of product A.

[1141] System: The price of product A is 5,000 yen. Please check the details.

[1142] (Prompt example 2)

[1143] User: What is the transaction status for Item B?

[1144] System: The latest transaction status for Item B is 'Pending'. We will notify you as soon as there is an update.

[1145] In this way, the system can provide quick and accurate responses to various inquiries about electronic payment services.

[1146] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1147] Step 1:

[1148] Users input and send inquiry messages from their smartphones, PCs, or other devices. The input messages are then sent from the user's device to the server.

[1149] Step 2:

[1150] The server receives the query message and converts the data sent in the form of an email message or chat message into a text format, resulting in text data that can be parsed.

[1151] Step 3:

[1152] The server sends the received text data to the AI ​​engine (generative AI model) and analyzes the content. In this step, the following processes are performed using natural language processing technology:

[1153] Tokenization: Breaking the message content down into words.

[1154] Part-of-speech tagging: tag each word with its part of speech.

[1155] Thematic extraction: Identifying the main topics and keywords of a message.

[1156] The analysis results in themes and keywords related to the content of the query.

[1157] Step 4:

[1158] Based on the analysis, the server classifies the query into an appropriate category, such as "price inquiry," "transaction status inquiry," etc. This classification serves as a preliminary step in generating an appropriate response.

[1159] Step 5:

[1160] The server retrieves the necessary information from the database based on the category. For example, if the query is "price inquiry," it retrieves the relevant product price information from the database. Specifically, it executes an SQL query to extract the necessary data, such as price information and transaction status.

[1161] Step 6:

[1162] The server uses the retrieved data and the template to generate a response in a natural style. Using a template engine (such as Jinja2), the server embeds the data in the template to create a response. This response is an appropriate response to the query.

[1163] Step 7:

[1164] The server then sends the generated response to the user's device. The response message is delivered to the user using the SMTP protocol or the chat application's API. This process allows the user to receive a prompt and appropriate response.

[1165] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1166] The present invention provides a system that responds quickly and appropriately to inquiry emails from companies. This system automatically performs a series of processes: receiving an inquiry email, analyzing its content, generating an appropriate response, and sending it as a reply email. Furthermore, by combining this system with an emotion engine that recognizes the user's emotions, the system can adjust the tone and content of the response to match the user's emotions. The system's program processing is explained below in natural language.

[1167] System Configuration

[1168] The system includes the following components:

[1169] Server: Hosts the mail server and AI engine, receives and analyzes inquiry emails, generates and sends responses, and is also equipped with an emotion engine to analyze user emotions.

[1170] Device: A PC or mobile device used by a user (company representative). Sends an inquiry email and receives an automatic reply email from the server.

[1171] User: Company representative. Responsible for sending inquiry emails to the system.

[1172] Program processing

[1173] 1. Server: Receives inquiry emails. Receives emails sent by users (company personnel) and converts the email text into text format.

[1174] 2. Server: Sends the email text to the AI ​​engine and analyzes the email content using natural language processing. Specifically, it performs the following processes:

[1175] Tokenization: Breaking down email content into words.

[1176] Part-of-speech tagging: tag each word with its part of speech.

[1177] Topic extraction: Extracting keywords and phrases to identify the subject of a query.

[1178] 3. Server: Analyzes the user's emotions from the email content using an emotion engine. Specifically, it uses emotion analysis technology to determine the emotional state, such as positive, negative, or neutral.

[1179] 4. Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry. Based on this categorization, the AI ​​engine generates the optimal answer corresponding to the inquiry.

[1180] 5. Server: Based on the analyzed content, the server retrieves relevant data from the database, such as inventory information, price information, etc., depending on the subject of the inquiry.

[1181] 6. Server: Generates a response email by embedding the acquired data using a template. Template engines such as Jinja2 can be used. Furthermore, the tone and content of the response email are adjusted based on the analysis results of the emotion engine.

[1182] 7. Server: Sends the generated response email to the user. An automatic reply email is sent to the user's email address using an SMTP server.

[1183] Specific examples of processing

[1184] Example 1: Product inventory inquiry

[1185] User: A company representative sends an email inquiring about the stock status of Product A.

[1186] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[1187] AI engine: Analyzes the email content and determines that the main question is about stock status. The emotion engine recognizes the user's emotion as negative. It retrieves stock information for product A from the database and generates an answer based on that information.

[1188] Generated answer:

[1189] "Product A's current stock status is as follows:

[1190] Number in stock: 150

[1191] Next arrival: October 15, 2023

[1192] We apologize for keeping you waiting and appreciate your understanding."

[1193] This email will include proposal materials such as an operation manual for the inventory management system and a detailed catalog for Product A.

[1194] Server: The generated email is automatically sent to the company representative.

[1195] Example 2: Price enquiry

[1196] User: A company representative sends an email inquiring about the price of Product B.

[1197] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[1198] AI engine: Analyzes the email content and determines that the main topic of the inquiry is "price." The emotion engine recognizes the user's emotion as "positive." It retrieves the latest price information for Product B from the price information database and generates an answer based on that information.

[1199] Generated answer:

[1200] "Product B's latest price information is as follows:

[1201] Unit price: 5,000 yen

[1202] Discount information: 10% off this month only

[1203] Please refer to the price list for details. We look forward to your continued support.

[1204] This email will include proposal materials such as a price list and details of discount campaigns.

[1205] Server: The generated email is automatically sent to the company representative.

[1206] This allows the system of the present invention to respond to inquiries from companies in a prompt and appropriate manner, taking into consideration the feelings of the user.

[1207] The processing flow will be explained below.

[1208] Step 1:

[1209] User: A company representative creates and sends an inquiry email.

[1210] Step 2:

[1211] Server: The mail server receives the inquiry email.

[1212] The SMTP server listens for emails on port 25, stores them in the inbox, and converts the email content into text format.

[1213] Step 3:

[1214] Server: Sends the email converted to text format to the AI ​​engine.

[1215] The email text is sent as a REST API request to the AI ​​engine for analysis.

[1216] Step 4:

[1217] AI engine: Analyzes email content using natural language processing technology.

[1218] Tokenization: Breaking down email content into words.

[1219] Part-of-speech tagging: tag each word with its part of speech.

[1220] Topic extraction: Extracting keywords and phrases to identify the subject of a query, using techniques such as morphological analyzers, POS taggers, TF-IDF, and BERT.

[1221] Step 5:

[1222] Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry.

[1223] Based on the analysis results, the products are divided into categories such as "inventory," "price," and "specifications."

[1224] Step 6:

[1225] Server: Sends text formatted email to the emotion engine and analyzes the user's emotions.

[1226] Sentiment analysis techniques are used to determine emotional states such as positive, negative, or neutral.

[1227] Step 7:

[1228] Server: Select the appropriate template according to the category.

[1229] For example, an "inventory" inquiry uses an availability response template, and a "price" inquiry uses a price information response template.

[1230] Step 8:

[1231] Server: Retrieves the required data from the database.

[1232] Use Python scripts and SQL queries to pull relevant data from the database, such as inventory information, pricing information, etc.

[1233] Step 9:

[1234] Server: The acquired data is embedded in the template and a reply email is generated.

[1235] Use a template engine such as Jinja2 to create emails in a natural style. Adjust the tone and content of the response email based on the analysis results of the emotion engine.

[1236] Step 10:

[1237] Server: The generated response email is sent to the user.

[1238] Use an email sending module (such as smtplib) to send an automatic reply email to the user's email address via the SMTP server.

[1239] Step 11:

[1240] User: The company representative receives the automatic reply email and checks the contents.

[1241] Example 2

[1242] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1243] In conventional inquiry email response systems, analyzing email content and generating appropriate responses is often done manually, making it difficult to respond quickly and accurately.In addition, it is not possible to adjust the tone or content of responses according to the user's emotions, making it difficult to improve user satisfaction.

[1244] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1245] In this invention, the server includes means for receiving an inquiry email, means for analyzing the content of the inquiry email using natural language processing technology, means for recognizing the emotional state from the email content, means for retrieving necessary data from a database based on the analyzed content, means for generating an appropriate response based on the analyzed content and the retrieved data, and means for transmitting the generated response. This enables quick and accurate response to inquiries and can generate responses that correspond to the user's emotions, which is expected to improve user satisfaction.

[1246] An "inquiry email" is an email sent by a user (company representative) to inquire about information from the system.

[1247] "Natural language processing technology" refers to a set of techniques and algorithms that enable computers to understand and process human language.

[1248] The "emotional state" refers to the emotional state that the user is in when sending the inquiry email, and includes positive, negative, neutral, and the like.

[1249] A "database" is a structured collection of data that allows a system to store and retrieve required information.

[1250] An "appropriate response" is an email that is automatically generated based on the contents of the inquiry email and provides answers and information to the user's questions.

[1251] "Adjusting tone and expression" means changing the content of the reply email depending on the user's emotional state and providing information in an appropriate manner.

[1252] "Emotion analysis technology" is a technology for identifying and classifying user emotions from text data.

[1253] This invention is a system for responding promptly and appropriately to inquiries from companies via email. This system fully automates the process of receiving, analyzing, generating responses, and replying to emails. It also has the ability to recognize the user's emotional state and adjust the tone and content of the response based on that state.

[1254] System Configuration

[1255] The system includes the following components:

[1256] Server: Hosts the mail server and AI engine, receives and analyzes inquiry emails, generates and sends responses, and is also equipped with a sentiment analysis engine to analyze user sentiment.

[1257] Terminal: A personal computer or mobile terminal used by a user (company representative). Sends inquiry emails and receives automatic reply emails from the server.

[1258] User: A company representative who is responsible for sending inquiry emails to the system.

[1259] Program processing

[1260] The system processes an inquiry email as follows:

[1261] Hardware and Software

[1262] Server: An IMAP server is used to receive inquiry emails. Natural language processing technology (e.g., spaCy) is used to analyze email content, and a sentiment analysis engine (e.g., VADER) is used for sentiment analysis. A database management system such as MySQL is used for the database, and a template engine (e.g., Jinja2) is used to generate response emails. An SMTP server is used to send emails.

[1263] Specific examples

[1264] Example 1: Product inventory inquiry

[1265] User: A company representative sends an email inquiring about the stock status of Product A.

[1266] Server: Receives the inquiry email and sends the email text to the natural language processing engine.

[1267] Natural language processing engine: Analyzes the email content and determines that the main question is about stock status. The sentiment analysis engine recognizes the user's sentiment as negative. The engine retrieves stock information for product A from the database and generates a response based on that information.

[1268] Generated answer:

[1269] text

[1270] The current stock status of product A is as follows:

[1271] Quantity in stock: 150

[1272] Next expected arrival date: October 15, 2023

[1273] We apologize for the wait and appreciate your understanding.

[1274] This email will include proposal materials such as an inventory management system operation manual and a detailed catalog for product A.

[1275] Server: The generated email is automatically sent to the company representative.

[1276] Example 2: Price enquiry

[1277] User: A company representative sends an email inquiring about the price of Product B.

[1278] Server: Receives the inquiry email and sends the email text to the natural language processing engine.

[1279] Natural language processing engine: Analyzes the email content and determines that "price" is the main topic of the inquiry. The sentiment analysis engine recognizes the user's sentiment as "positive." The latest price information for Product B is retrieved from the price information database and an answer is generated based on that information.

[1280] Generated answer:

[1281] text

[1282] The latest price information for product B is as follows:

[1283] Unit price: 5,000 yen

[1284] Discount information: 10% off this month only

[1285] Please refer to the price list for details. We look forward to your continued support.

[1286] This email will include a price list and details of discount campaigns as proposal materials.

[1287] Server: The generated email is automatically sent to the company representative.

[1288] Prompt Sentence Examples

[1289] Below are some example prompts to input to a generative AI model:

[1290] text

[1291] Design a system that analyzes user inquiry emails, recognizes their emotional state, and generates the best possible response. However, use a template engine to generate the response email and adjust the tone depending on the emotional state.

[1292] This allows the system of the present invention to respond to inquiry emails from companies quickly and accurately while taking into consideration the user's feelings.

[1293] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1294] Step 1: Receiving an inquiry email

[1295] Server: Receives inquiry emails sent by users (company personnel) via the IMAP server. The emails are retrieved from a specific mailbox and the character code is converted to UTF-8.

[1296] Input: New emails from the IMAP mail server

[1297] Output: UTF-8 formatted email text

[1298] Specific operation: Connect to the IMAP server, check for new emails, and retrieve them. Convert the character code to UTF-8.

[1299] Step 2: Analyzing the email content

[1300] Server: The received email text is sent to a natural language processing engine (e.g., spaCy) to analyze the query. The analysis involves tokenization, part-of-speech tagging, and topic extraction.

[1301] Input: UTF-8 formatted email text

[1302] Output: Parsed email content (tokens, part-of-speech tags, subject)

[1303] What it does: Tokenizes email text, tags each word with a part-of-speech tag, and extracts key keywords and phrases.

[1304] Step 3: Sentiment Analysis

[1305] Server: The analyzed email content is sent to an emotion analysis engine (e.g., VADER) to analyze the user's emotional state.

[1306] Input: Parsed email content (tokens, part-of-speech tags, subject)

[1307] Output: Emotional state (positive, negative, neutral)

[1308] Specific operation: The analyzed email content is input into an emotion analysis engine to determine the emotional state.

[1309] Step 4: Categorize the inquiry

[1310] Server: Based on the analysis results of the natural language processing engine, the inquiry content is categorized (e.g., "stock status," "price," etc.).

[1311] Input: Parsed email content, emotional state

[1312] Output: Category information (availability, price, etc.)

[1313] What it does: Maps email content to specific categories and selects the appropriate response template.

[1314] Step 5: Retrieve related data

[1315] Server: Based on the categorized inquiry, retrieves the necessary information from the database, such as stock information and price information.

[1316] Input: Category information

[1317] Output: Retrieved data (stock information, price information, etc.)

[1318] Specific operation: Executes a database query, retrieves the corresponding data, and stores the retrieved data in internal memory.

[1319] Step 6: Generate a response email

[1320] Server: Use a template engine (e.g., Jinja2) to embed the acquired data into the response email. Adjust the tone and content of the response email based on the results of sentiment analysis.

[1321] Input: Template, captured data, emotional state

[1322] Output: Generated response email

[1323] What it does: Embed data in templates to tailor the wording and tone of emails based on emotional state.

[1324] Step 7: Sending an email

[1325] Server: The generated response email is sent to the user via the SMTP server.

[1326] Input: Generated response email

[1327] Output: Email sent

[1328] Specific operation: Connect to the SMTP server, add the generated email to the sending queue, and send it to the user's email address.

[1329] (Application example 2)

[1330] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1331] Current inquiry response systems are required to respond quickly and appropriately to inquiry emails from users. However, previous systems provided mechanical responses without taking emotions into account, which often resulted in a poor user experience. Particularly in the field of e-commerce, flexible responses that match the user's emotions are important, and this aspect needs to be improved.

[1332] The specific processing 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 receiving an inquiry email, means for analyzing the content of the inquiry email, means for generating an appropriate response, means for sending the generated response, means for analyzing the user's emotions, and means for adjusting the tone and content of the response based on the analyzed user's emotions. This enables a quick and appropriate response that takes the user's emotions into consideration.

[1333] An "inquiry email" is an electronic message sent by a user to a business or organization seeking information or support.

[1334] "Means of analysis" refers to the processes and techniques used to understand the content of received inquiry emails and identify their meaning and intent.

[1335] "Natural language processing technology" refers to a series of technologies and methods that allow computers to understand, analyze, and generate human language.

[1336] "Means for generating an appropriate answer" refers to a process for automatically creating the most appropriate answer to a user's inquiry based on the analysis results.

[1337] "Means for analyzing user emotions" refers to a technology that determines the user's emotional state from the content of the inquiry email and classifies it as a numerical value or category.

[1338] "Methods for adjusting the tone and content of responses" refers to techniques for changing the wording and expression of responses to match the analyzed emotional state.

[1339] A "database" is an information system used to manage information related to inquiries (e.g., product inventory, pricing information, etc.).

[1340] "Mobile devices such as smartphones" refers to devices that are portable and can connect to the Internet and use applications.

[1341] MODE FOR CARRYING OUT THE INVENTION

[1342] This invention is an automatic response system for responding quickly and accurately to inquiry emails from companies. This system automatically performs a series of processes, from receiving the inquiry email, analyzing its content, generating and sending a response, and even adjusting the tone of the response to take the user's emotions into consideration.

[1343] System Configuration

[1344] The system includes the following components:

[1345] 1. Server: Equipped with an emotion engine that receives and analyzes inquiry emails, generates and sends responses, and analyzes user emotions.

[1346] 2. Terminal: A mobile terminal such as a PC or smartphone used by a user (company representative). It sends an inquiry email and receives an automatic reply email from the server.

[1347] 3. User: Company representative. Responsible for sending inquiry emails to the system.

[1348] Hardware and software used

[1349] Mail server: Gmail API

[1350] AI engine: Google Cloud Natural Language API

[1351] Sentiment analysis engine: IBM Watson Natural Language Understanding API

[1352] Template engine: Jinja2

[1353] Database: PostgreSQL

[1354] System Operation

[1355] The server first receives the inquiry email. At this time, it retrieves the email using the Gmail API, etc. The received email text is analyzed using the Google Cloud Natural Language API, and tokenization, part-of-speech tagging, and topic extraction are performed.

[1356] The emotional state of the received email is then analyzed using the IBM Watson Natural Language Understanding API, which can determine whether the email content carries a positive, negative, or neutral sentiment.

[1357] Based on the analysis results, an appropriate answer is generated. The answer is generated using the Jinja2 template engine, retrieving the necessary data from the database (PostgreSQL) according to the inquiry content. The tone of the generated answer is adjusted based on the results of sentiment analysis, so that the content takes the user's emotions into consideration.

[1358] Finally, the generated response is automatically sent to the user using an SMTP server, allowing the user to receive a prompt and appropriate response.

[1359] Specific examples

[1360] Example 1: Product inventory inquiry

[1361] A user sends an email inquiry about the stock status of product A. The server receives the inquiry email and sends the email text to the AI ​​engine for analysis. The AI ​​engine determines that "stock status" is the main content of the inquiry, and the emotion engine recognizes the user's emotion as "negative." The server retrieves stock information for product A from the database and generates an answer based on that information. The generated answer includes the following:

[1362] "Product A's current stock status is as follows:

[1363] Number in stock: 150

[1364] Next arrival: October 15, 2023

[1365] We apologize for keeping you waiting and appreciate your understanding."

[1366] Example 2: Price enquiry

[1367] A user sends an email inquiry about the price of product B. The server receives the inquiry email and sends the email text to the AI ​​engine for analysis. The AI ​​engine determines that "price" is the main topic of the inquiry, and the emotion engine recognizes the user's emotion as "positive." The latest price information for product B is retrieved from the price information database, and an answer is generated based on that information. The generated answer includes the following:

[1368] "Product B's latest price information is as follows:

[1369] Unit price: 5,000 yen

[1370] Discount information: 10% off this month only

[1371] Please refer to the price list for details. We look forward to your continued support.

[1372] Prompt Sentence Examples

[1373] "Generate an auto-reply message based on the email sent by the user. The content of the email should be, 'Please let me know if the product is in stock.'"

[1374] This allows the system of the present invention to respond quickly and appropriately while taking into consideration the user's feelings.

[1375] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1376] Step 1:

[1377] The server receives an inquiry email from a user using the Gmail API. At this time, it obtains the email content in text format. The input is the inquiry email sent by the user, and the output is the email content in text format.

[1378] Step 2:

[1379] The server sends the received email text to the Google Cloud Natural Language API for natural language processing. Specific operations include tokenizing the text, tagging parts of speech, and extracting themes to analyze the meaning and intent of the email. The input is the email content in text format, and the output is the analysis results: keywords and themes.

[1380] Step 3:

[1381] The server uses the IBM Watson Natural Language Understanding API to analyze the user's emotions from the email content. Specifically, it uses sentiment analysis technology to determine emotional states such as positive, negative, and neutral as numerical values ​​or categories. The input is the email content in text format, and the output is the result of the sentiment analysis.

[1382] Step 4:

[1383] The server categorizes the query based on the analysis results and retrieves the necessary data from the database. For example, it retrieves inventory and price information from a PostgreSQL database. The input is the subject and keywords resulting from the analysis, and the output is the retrieved related information.

[1384] Step 5:

[1385] The server uses the Jinja2 template engine to embed the retrieved data and generate an appropriate reply email. Furthermore, it adjusts the tone and content of the reply email based on the analysis results of the emotion engine. Specific operations include modifying the email text based on the emotion analysis. The input is the retrieved related information and the emotion analysis results, and the output is the generated reply email.

[1386] Step 6:

[1387] The server automatically sends the generated reply email to the user using the SMTP server. Specifically, it sets the email address, subject, and body, and performs the sending process via the SMTP server. The input is the generated reply email, and the output is the email sent to the user.

[1388] This allows the system to respond to inquiries promptly and appropriately, taking into consideration the user's feelings.

[1389] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1390] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1391] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1392] [Fourth embodiment]

[1393] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1394] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1395] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1396] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1397] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1398] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1399] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1400] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1401] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1402] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1403] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1404] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1405] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1406] This invention provides a system that responds quickly and appropriately to inquiry emails from companies. This system automatically performs a series of processes: receiving an inquiry email, analyzing its content, generating an appropriate response, and sending it as a reply email. The processing of the system's program is explained below in natural language.

[1407] System Configuration

[1408] The system includes the following components:

[1409] Server: Hosts the mail server and AI engine, and receives, analyzes, generates, and sends email inquiries.

[1410] Device: A PC or mobile device used by a user (company representative). Sends an inquiry email and receives an automatic reply email from the server.

[1411] User: Company representative. Responsible for sending inquiry emails to the system.

[1412] Program processing

[1413] 1. Server: Receives inquiry emails. Receives emails sent by users (company personnel) and converts the email text into text format.

[1414] 2. Server: Sends the email text to the AI ​​engine and analyzes the email content using natural language processing. Specifically, it performs the following processes:

[1415] Tokenization: Breaking down email content into words.

[1416] Part-of-speech tagging: tag each word with its part of speech.

[1417] Topic extraction: Extracting keywords and phrases to identify the subject of a query.

[1418] 3. Server: Based on the analysis results, the inquiry is categorized. Based on this categorization, the AI ​​engine generates the optimal answer corresponding to the inquiry.

[1419] 4. Server: Retrieves relevant data from the database as needed, for example, stock information, pricing information, or other data depending on the subject of the query.

[1420] 5. Server: Using templates, embed the acquired data and generate a reply email in a natural style. Template engines such as Jinja2 can be used.

[1421] 6. Server: The generated response email is sent to the user (company representative). The sending process is performed using an SMTP server.

[1422] Specific examples of processing

[1423] Example 1: Product inventory inquiry

[1424] User: A company representative sends an email inquiring about the stock status of Product A.

[1425] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[1426] AI engine: Analyzes the email content and determines that the main question is about stock status. It retrieves stock information for product A from the database and generates a response based on that information.

[1427] Generated answer:

[1428] "Product A's current stock status is as follows:

[1429] Number in stock: 150

[1430] Next arrival: October 15, 2023

[1431] Thank you for your confirmation."

[1432] This email will include proposal materials such as an operation manual for the inventory management system and a detailed catalog for Product A.

[1433] Server: The generated email is automatically sent to the company representative.

[1434] Example 2: Price enquiry

[1435] User: A company representative sends an email inquiring about the price of Product B.

[1436] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[1437] AI engine: Analyzes the email content and determines that the main question is about price. It retrieves the latest price information for Product B from a price information database and generates a response based on that information.

[1438] Generated answer:

[1439] "Product B's latest price information is as follows:

[1440] Unit price: 5,000 yen

[1441] Discount information: 10% off this month only

[1442] Please refer to the price list for details. If you have any questions, please let us know.

[1443] This email will include proposal materials such as a price list and details of discount campaigns.

[1444] Server: The generated email is automatically sent to the company representative.

[1445] This allows the system of the present invention to respond quickly and appropriately to inquiry emails from companies.

[1446] The processing flow will be explained below.

[1447] Step 1:

[1448] User: A company representative creates and sends an inquiry email.

[1449] Step 2:

[1450] Server: The mail server receives the inquiry email.

[1451] The SMTP server listens for emails on port 25, stores them in the inbox, and converts the email content into text format.

[1452] Step 3:

[1453] Server: Sends the email converted to text format to the AI ​​engine.

[1454] The email text is sent as a REST API request to the AI ​​engine for analysis.

[1455] Step 4:

[1456] AI engine: Analyzes email content using natural language processing technology.

[1457] It performs tokenization (dividing email content into words), part-of-speech tagging (tagging each word with its part of speech), and subject extraction (extracting keywords and phrases to identify the subject of the inquiry), using technologies such as morphological analyzers, POS taggers, TF-IDF, and BERT.

[1458] Step 5:

[1459] Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry.

[1460] Based on the analysis results, the products are divided into categories such as "inventory," "price," and "specifications."

[1461] Step 6:

[1462] Server: Select the appropriate template according to the category.

[1463] For example, an "inventory" inquiry uses an availability response template, and a "price" inquiry uses a price information response template.

[1464] Step 7:

[1465] Server: Retrieves the required data from the database.

[1466] Use Python scripts and SQL queries to pull relevant data from the database, such as inventory information, pricing information, etc.

[1467] Step 8:

[1468] Server: The acquired data is embedded in the template and a reply email is generated.

[1469] Use a template engine such as Jinja2 to create emails in a natural writing style.

[1470] Step 9:

[1471] Server: The generated response email is sent to the user.

[1472] Use an email sending module (such as smtplib) to send an automatic reply email to the user's email address via the SMTP server.

[1473] Step 10:

[1474] User: The company representative receives the automatic reply email and checks the contents.

[1475] Example 1

[1476] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1477] A problem with conventional inquiry response systems is that it takes a great deal of time and effort to analyze the content of received emails and then generate and send appropriate responses based on that information. Furthermore, manual responses are prone to human error and inconsistencies, making it difficult to provide effective and consistent responses. Furthermore, there is a need for system automation to quickly obtain appropriate data and reflect it in responses.

[1478] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1479] In this invention, the server includes a means for receiving an inquiry email, a means for analyzing the content of the inquiry email using natural language processing technology, a means for categorizing the inquiry content based on the analysis result, a means for retrieving related data from a database according to the categorized content, a means for generating an appropriate response using the retrieved data, and a means for transmitting the generated response. This automates the inquiry response process, enabling quick and accurate responses. Furthermore, the ability to generate consistent responses improves the quality of a company's inquiry responses.

[1480] An "inquiry email" is an email containing a question or request that a user sends to the system.

[1481] "Means of receiving" refers to the functions and processes by which the server receives and stores inquiry emails.

[1482] "Means for analysis" refers to the functions and processes for analyzing the content of received inquiry emails using natural language processing technology and extracting important information.

[1483] "Natural language processing technology" is a field of computer science that involves understanding and analyzing the meaning of text data, and is a technology that performs tasks such as subject extraction and part-of-speech tagging.

[1484] A "generative AI model" is an artificial intelligence model that generates answers in natural language based on input data.

[1485] A "prompt" is an instruction or cue provided to an AI model when it performs analysis.

[1486] "Means of categorization" refers to the functions and processes for classifying the content of inquiries based on the analysis results and determining the appropriate processing method.

[1487] A database is a system for organizing and storing large amounts of information and quickly retrieving data when needed.

[1488] "Means of obtaining relevant data" refers to the functions and processes for extracting the necessary information from the database in response to the inquiry.

[1489] "Answer generation means" refers to the functions and processes that automatically generate an appropriate reply to the user based on the relevant data obtained.

[1490] A "template engine" is software that uses template files to embed data in order to generate dynamic content.

[1491] The "means of sending" refers to the function or process for delivering the generated reply email to the user, and uses an SMTP server or the like.

[1492] A "system" is an integrated device or software that automatically performs a series of processes including receiving, analyzing, classifying, acquiring data, generating responses, and sending inquiries.

[1493] This invention is an automated system for responding promptly and appropriately to inquiry emails from companies. This system performs a series of automated processes: receiving inquiry emails, analyzing their contents, generating responses, and sending the responses.

[1494] System Configuration

[1495] The system includes the following components:

[1496] Server: Hosts the mail server and natural language processing engine, receives and analyzes inquiry emails, generates responses, and sends them. Postfix is ​​used as the mail server, and OpenAI GPT-3 is used as the natural language processing engine.

[1497] Device: A computer or mobile device used by a user (company representative). For example, a Windows PC or iPhone.

[1498] User: Company representative. Responsible for sending inquiry emails to the system.

[1499] Program processing

[1500] The server responds to the inquiry email by following these steps:

[1501] 1. Receiving an inquiry email:

[1502] A user sends an inquiry email using a PC or mobile device. The server receives and stores the inquiry email using the Postfix mail server.

[1503] 2. Email content analysis:

[1504] The server sends the received email text to an AI engine for natural language processing (OpenAI GPT-3) using a prompt. An example of a specific prompt is "What is the subject of this email? I would like to check the stock of product A." The natural language processing engine performs the following process:

[1505] Tokenization: Splitting the text of an email into words. For example, "I would like to check the stock of product A" would be split into "I would like to check the stock of product A."

[1506] Part-of-speech tagging: tag each word with its part of speech, for example, "Product A / noun | of / particle | inventory / noun | wo / particle | confirm / verb | shi / auxiliary verb | tai / auxiliary verb | desu / auxiliary verb."

[1507] Subject Extraction: Use TF-IDF and other natural language processing algorithms to extract the subject of the query. For example, identify that "stock availability" is the subject.

[1508] 3. Categorize your inquiry:

[1509] Based on the analysis results obtained from the AI ​​engine, the server classifies the inquiry into appropriate categories (e.g., stock confirmation, price confirmation).

[1510] 4. Obtaining relevant data:

[1511] The server sends a query to the MySQL database based on the categorized inquiry to retrieve the appropriate information. For example, it executes a query such as "SELECT stock FROM products WHERE product_name="Product A"" to retrieve stock information for Product A.

[1512] 5. Generate reply email:

[1513] The server uses the Jinja2 template engine to embed the acquired data into a template and generate a reply email in a natural style. For example, the stock quantity data is embedded in a template such as "The stock quantity is {{ stock}} units."

[1514] 6. Sending response email:

[1515] The server sends the generated response email to the user (company representative) using the SMTP server (Postfix). If the email was sent successfully, it is recorded in the log.

[1516] Specific examples

[1517] Example 1: Product inventory inquiry

[1518] User: A company representative sends an email inquiring about the stock status of Product A.

[1519] Server: The Postfix mail server receives the email, converts it to email text, and then sends the email text to the AI ​​engine with the following prompt:

[1520] "What is the subject of this email?: I would like to check the availability of product A."

[1521] AI engine: Analyzes the email content and determines that the subject is "stock status."

[1522] Server: Based on the analysis results of the AI ​​engine, classify this inquiry into the "Stock Check" category. Then, send a query to the database saying "Provide the stock quantity of product A."

[1523] Server: Retrieve stock information for Product A (for example, stock quantity 150) from the database and generate a response email using the Jinja2 template engine.

[1524] "Product A's current stock status is as follows:

[1525] Number in stock: 150

[1526] Next arrival: October 15, 2023

[1527] Thank you for your confirmation."

[1528] Server: The generated response email is sent to the user (company representative) using an SMTP server.

[1529] Example 2: Price enquiry

[1530] User: A company representative sends an email inquiring about the price of Product B.

[1531] Server: The Postfix mail server receives the email, converts it to email text, and then sends the email text to the AI ​​engine with the following prompt:

[1532] "What is the subject of this email? I would like to know the price of product B."

[1533] AI engine: Analyzes the content of the email and determines that the subject is "price."

[1534] Server: Based on the analysis results of the AI ​​engine, the server classifies this inquiry into the "price check" category. It then sends a query to the database saying, "Provide the price of product B."

[1535] Server: Retrieve the price information of Product B (for example, unit price 5,000 yen, discount 10%) from the database and generate a reply email using the Jinja2 template engine.

[1536] "Product B's latest price information is as follows:

[1537] Unit price: 5,000 yen

[1538] Discount information: 10% off this month only

[1539] Please refer to the price list for details. If you have any questions, please let us know.

[1540] Server: The generated response email is sent to the user (company representative) using an SMTP server.

[1541] In this way, the system of the present invention can respond quickly and appropriately to inquiry emails from companies.

[1542] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1543] Step 1: Receiving an inquiry email

[1544] Server: A user (company employee) sends an inquiry email. The Postfix mail server receives the email. The input is the email sent by the user, and the output is the email data stored on the mail server. Specifically, the server converts the received email into a specific format (e.g., a text file).

[1545] Step 2: Analyzing the email content

[1546] Server: Sends the text of the received email to an AI engine for natural language processing (OpenAI GPT-3). The input is the text content of the email, and the output is the analysis result. The specific operation is as follows:

[1547] Tokenization: Splitting text into words (e.g., "I would like to check the stock of product A" → "I would like to check the stock of product A").

[1548] Part-of-speech tagging: Tag each word with its part of speech (e.g., "Product A / noun | of / particle | inventory / noun | wo / particle | confirm / verb | shi / auxiliary verb | tai / auxiliary verb | desu / auxiliary verb").

[1549] Topic extraction: TF-IDF is used to extract keywords and identify topics (e.g., "stock status").

[1550] Step 3: Categorize the inquiry

[1551] Server: Categorizes the inquiry content based on the analysis results of the AI ​​engine. The input is the analysis result, and the output is category information (e.g., "Check stock availability" or "Check price"). The specific operation is to classify the inquiry into predefined categories based on the subject.

[1552] Step 4: Retrieve related data

[1553] Server: Sends a query to a database (e.g. MySQL) based on the categorization. The input is the category information and the associated data request, and the output is the specific information retrieved from the database (e.g., the number of items in stock for product A). The specific operation is as follows:

[1554] Create a query: Create an SQL query based on the subject of your inquiry (e.g., "SELECT stock FROM products WHERE product_name="Product A"").

[1555] Data Acquisition: Pulling the required information from the database.

[1556] Step 5: Generate a response email

[1557] Server: Generates a reply email using the Jinja2 template engine based on the retrieved data. The input is the information retrieved from the database and a template, and the output is a reply email. The specific operations are as follows:

[1558] Embed data: Embed data in the template (e.g., "Your stock is {{ stock}}").

[1559] Email formatting: Format your emails in a natural style.

[1560] Step 6: Sending a response email

[1561] Server: The generated reply email is sent to the user (company representative) using the SMTP server (Postfix). The input is the reply email, and the output is a confirmation of successful sending. The specific operations are as follows:

[1562] Sending email: Sends email using the SMTP protocol.

[1563] Log: Logs successful transmission.

[1564] (Application example 1)

[1565] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1566] Modern electronic payment services are receiving an increasing number of inquiries from users, and they need to respond to these inquiries quickly and accurately. However, because the content of inquiries varies widely, manually responding to all inquiries is difficult, time-consuming, and costly. Furthermore, there is a lack of systems that can provide appropriate transaction information based on the content of inquiries and avoid inappropriate responses. Therefore, there is a need to develop a system that automates and streamlines inquiry responses in electronic payment services.

[1567] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1568] In this invention, the server includes means for receiving an inquiry message, means for analyzing the contents of the inquiry message, means for generating an appropriate response based on the analyzed content, means for transmitting the generated response, means for retrieving transaction information from a database based on the analyzed content, and means for customizing templates for the response, thereby enabling prompt and accurate responses to a variety of inquiries in electronic payment services.

[1569] An "inquiry message" is a message containing a question or confirmation that a user makes to a service provider.

[1570] "Analysis" refers to the methods and results used to examine and understand the content of given data or messages.

[1571] A "response" is a reply or answer provided to an inquiry or question.

[1572] "Sending" is the act of delivering data or messages to another device or network destination.

[1573] "Transaction information" refers to detailed data about a specific transaction, primarily in electronic payment services.

[1574] A "database" is an organized collection of data and a system that allows for efficient searching and retrieval.

[1575] A "template" is a document or file with a prescribed format or pattern, designed so that parts of its content can be easily changed.

[1576] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language.

[1577] A "generative AI model" is an artificial intelligence algorithm that generates optimal outputs for a task based on input data.

[1578] A "prompt sentence" is a sentence that contains instructions or a request to perform a specific task or action.

[1579] The present invention relates to a system that automates inquiries about electronic payment services and provides prompt and accurate responses. This system is comprised of multiple servers, terminals, and users, and is described in detail below.

[1580] System Configuration

[1581] The system includes the following components:

[1582] Server: The central component that receives query messages, analyzes their content, generates appropriate responses, and sends them. The server also hosts an AI engine and has the ability to analyze the content of messages using natural language processing techniques.

[1583] Terminal: A communication device such as a smartphone or PC used by a user, whose main role is to input inquiry messages and receive automatic responses from the server.

[1584] User: A person or company representative who uses electronic payment services and sends inquiry messages.

[1585] Technology and software used

[1586] The server performs natural language processing using a generative AI model based on Hugging Face's Transformers library, and uses an SQLite database to manage necessary data such as transaction and price information, and retrieves the appropriate data depending on the inquiry.

[1587] Program processing

[1588] When the server receives a query message, it converts it into text format and sends it to the AI ​​engine. The AI ​​engine analyzes the content of the query message and performs appropriate tokenization, part-of-speech tagging, and topic extraction. Based on the analysis results, it then classifies the query content into an appropriate category and retrieves related data (e.g., transaction information and price information) from the database.

[1589] The server then combines the retrieved data with the template to generate a natural-sounding response, possibly using a template engine (e.g., Jinja2), which is then automatically sent to the user's device.

[1590] Examples and prompts

[1591] As a concrete example, if a user makes a query such as "What is the transaction status of product A?", the server will act as follows:

[1592] 1. User: Enter "Please tell me the transaction status of product A" on your smartphone and send it.

[1593] 2. Server: Receives the message and analyzes its contents. The AI ​​engine determines that the subject is "transaction status."

[1594] 3. Server: Retrieve the transaction status related to product A from the database and generate a natural response based on it.

[1595] 4. Server: Send the generated response "The latest transaction status for product A is 'Completed'" to the user's terminal.

[1596] Prompt Sentence Examples

[1597] (Prompt example 1)

[1598] User: Please tell me the price of product A.

[1599] System: The price of product A is 5,000 yen. Please check the details.

[1600] (Prompt example 2)

[1601] User: What is the transaction status for Item B?

[1602] System: The latest transaction status for Item B is 'Pending'. We will notify you as soon as there is an update.

[1603] In this way, the system can provide quick and accurate responses to various inquiries about electronic payment services.

[1604] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1605] Step 1:

[1606] Users input and send inquiry messages from their smartphones, PCs, or other devices. The input messages are then sent from the user's device to the server.

[1607] Step 2:

[1608] The server receives the query message and converts the data sent in the form of an email message or chat message into a text format, resulting in text data that can be parsed.

[1609] Step 3:

[1610] The server sends the received text data to the AI ​​engine (generative AI model) and analyzes the content. In this step, the following processes are performed using natural language processing technology:

[1611] Tokenization: Breaking the message content down into words.

[1612] Part-of-speech tagging: tag each word with its part of speech.

[1613] Thematic extraction: Identifying the main topics and keywords of a message.

[1614] The analysis results in themes and keywords related to the content of the query.

[1615] Step 4:

[1616] Based on the analysis, the server classifies the query into an appropriate category, such as "price inquiry," "transaction status inquiry," etc. This classification serves as a preliminary step in generating an appropriate response.

[1617] Step 5:

[1618] The server retrieves the necessary information from the database based on the category. For example, if the query is "price inquiry," it retrieves the relevant product price information from the database. Specifically, it executes an SQL query to extract the necessary data, such as price information and transaction status.

[1619] Step 6:

[1620] The server uses the retrieved data and the template to generate a response in a natural style. Using a template engine (such as Jinja2), the server embeds the data in the template to create a response. This response is an appropriate response to the query.

[1621] Step 7:

[1622] The server then sends the generated response to the user's device. The response message is delivered to the user using the SMTP protocol or the chat application's API. This process allows the user to receive a prompt and appropriate response.

[1623] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1624] The present invention provides a system that responds quickly and appropriately to inquiry emails from companies. This system automatically performs a series of processes: receiving an inquiry email, analyzing its content, generating an appropriate response, and sending it as a reply email. Furthermore, by combining this system with an emotion engine that recognizes the user's emotions, the system can adjust the tone and content of the response to match the user's emotions. The system's program processing is explained below in natural language.

[1625] System Configuration

[1626] The system includes the following components:

[1627] Server: Hosts the mail server and AI engine, receives and analyzes inquiry emails, generates and sends responses, and is also equipped with an emotion engine to analyze user emotions.

[1628] Device: A PC or mobile device used by a user (company representative). Sends an inquiry email and receives an automatic reply email from the server.

[1629] User: Company representative. Responsible for sending inquiry emails to the system.

[1630] Program processing

[1631] 1. Server: Receives inquiry emails. Receives emails sent by users (company personnel) and converts the email text into text format.

[1632] 2. Server: Sends the email text to the AI ​​engine and analyzes the email content using natural language processing. Specifically, it performs the following processes:

[1633] Tokenization: Breaking down email content into words.

[1634] Part-of-speech tagging: tag each word with its part of speech.

[1635] Topic extraction: Extracting keywords and phrases to identify the subject of a query.

[1636] 3. Server: Analyzes the user's emotions from the email content using an emotion engine. Specifically, it uses emotion analysis technology to determine the emotional state, such as positive, negative, or neutral.

[1637] 4. Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry. Based on this categorization, the AI ​​engine generates the optimal answer corresponding to the inquiry.

[1638] 5. Server: Based on the analyzed content, the server retrieves relevant data from the database, such as inventory information, price information, etc., depending on the subject of the inquiry.

[1639] 6. Server: Generates a response email by embedding the acquired data using a template. Template engines such as Jinja2 can be used. Furthermore, the tone and content of the response email are adjusted based on the analysis results of the emotion engine.

[1640] 7. Server: Sends the generated response email to the user. An automatic reply email is sent to the user's email address using an SMTP server.

[1641] Specific examples of processing

[1642] Example 1: Product inventory inquiry

[1643] User: A company representative sends an email inquiring about the stock status of Product A.

[1644] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[1645] AI engine: Analyzes the email content and determines that the main question is about stock status. The emotion engine recognizes the user's emotion as negative. It retrieves stock information for product A from the database and generates an answer based on that information.

[1646] Generated answer:

[1647] "Product A's current stock status is as follows:

[1648] Number in stock: 150

[1649] Next arrival: October 15, 2023

[1650] We apologize for keeping you waiting and appreciate your understanding."

[1651] This email will include proposal materials such as an operation manual for the inventory management system and a detailed catalog for Product A.

[1652] Server: The generated email is automatically sent to the company representative.

[1653] Example 2: Price enquiry

[1654] User: A company representative sends an email inquiring about the price of Product B.

[1655] Server: Receives inquiry emails and sends the email text to the AI ​​engine.

[1656] AI engine: Analyzes the email content and determines that the main topic of the inquiry is "price." The emotion engine recognizes the user's emotion as "positive." It retrieves the latest price information for Product B from the price information database and generates an answer based on that information.

[1657] Generated answer:

[1658] "Product B's latest price information is as follows:

[1659] Unit price: 5,000 yen

[1660] Discount information: 10% off this month only

[1661] Please refer to the price list for details. We look forward to your continued support.

[1662] This email will include proposal materials such as a price list and details of discount campaigns.

[1663] Server: The generated email is automatically sent to the company representative.

[1664] This allows the system of the present invention to respond to inquiries from companies in a prompt and appropriate manner, taking into consideration the feelings of the user.

[1665] The processing flow will be explained below.

[1666] Step 1:

[1667] User: A company representative creates and sends an inquiry email.

[1668] Step 2:

[1669] Server: The mail server receives the inquiry email.

[1670] The SMTP server listens for emails on port 25, stores them in the inbox, and converts the email content into text format.

[1671] Step 3:

[1672] Server: Sends the email converted to text format to the AI ​​engine.

[1673] The email text is sent as a REST API request to the AI ​​engine for analysis.

[1674] Step 4:

[1675] AI engine: Analyzes email content using natural language processing technology.

[1676] Tokenization: Breaking down email content into words.

[1677] Part-of-speech tagging: tag each word with its part of speech.

[1678] Topic extraction: Extracting keywords and phrases to identify the subject of a query, using techniques such as morphological analyzers, POS taggers, TF-IDF, and BERT.

[1679] Step 5:

[1680] Server: Receives the analysis results from the AI ​​engine and categorizes the inquiry.

[1681] Based on the analysis results, the products are divided into categories such as "inventory," "price," and "specifications."

[1682] Step 6:

[1683] Server: Sends text formatted email to the emotion engine and analyzes the user's emotions.

[1684] Sentiment analysis techniques are used to determine emotional states such as positive, negative, or neutral.

[1685] Step 7:

[1686] Server: Select the appropriate template according to the category.

[1687] For example, an "inventory" inquiry uses an availability response template, and a "price" inquiry uses a price information response template.

[1688] Step 8:

[1689] Server: Retrieves the required data from the database.

[1690] Use Python scripts and SQL queries to pull relevant data from the database, such as inventory information, pricing information, etc.

[1691] Step 9:

[1692] Server: The acquired data is embedded in the template and a reply email is generated.

[1693] Use a template engine such as Jinja2 to create emails in a natural style. Adjust the tone and content of the response email based on the analysis results of the emotion engine.

[1694] Step 10:

[1695] Server: The generated response email is sent to the user.

[1696] Use an email sending module (such as smtplib) to send an automatic reply email to the user's email address via the SMTP server.

[1697] Step 11:

[1698] User: The company representative receives the automatic reply email and checks the contents.

[1699] Example 2

[1700] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1701] In conventional inquiry email response systems, analyzing email content and generating appropriate responses is often done manually, making it difficult to respond quickly and accurately.In addition, it is not possible to adjust the tone or content of responses according to the user's emotions, making it difficult to improve user satisfaction.

[1702] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1703] In this invention, the server includes means for receiving an inquiry email, means for analyzing the content of the inquiry email using natural language processing technology, means for recognizing the emotional state from the email content, means for retrieving necessary data from a database based on the analyzed content, means for generating an appropriate response based on the analyzed content and the retrieved data, and means for transmitting the generated response. This enables quick and accurate response to inquiries and can generate responses that correspond to the user's emotions, which is expected to improve user satisfaction.

[1704] An "inquiry email" is an email sent by a user (company representative) to inquire about information from the system.

[1705] "Natural language processing technology" refers to a set of techniques and algorithms that enable computers to understand and process human language.

[1706] The "emotional state" refers to the emotional state that the user is in when sending the inquiry email, and includes positive, negative, neutral, and the like.

[1707] A "database" is a structured collection of data that allows a system to store and retrieve required information.

[1708] An "appropriate response" is an email that is automatically generated based on the contents of the inquiry email and provides answers and information to the user's questions.

[1709] "Adjusting tone and expression" means changing the content of the reply email depending on the user's emotional state and providing information in an appropriate manner.

[1710] "Emotion analysis technology" is a technology for identifying and classifying user emotions from text data.

[1711] This invention is a system for responding promptly and appropriately to inquiries from companies via email. This system fully automates the process of receiving, analyzing, generating responses, and replying to emails. It also has the ability to recognize the user's emotional state and adjust the tone and content of the response based on that state.

[1712] System Configuration

[1713] The system includes the following components:

[1714] Server: Hosts the mail server and AI engine, receives and analyzes inquiry emails, generates and sends responses, and is also equipped with a sentiment analysis engine to analyze user sentiment.

[1715] Terminal: A personal computer or mobile terminal used by a user (company representative). Sends inquiry emails and receives automatic reply emails from the server.

[1716] User: A company representative who is responsible for sending inquiry emails to the system.

[1717] Program processing

[1718] The system processes an inquiry email as follows:

[1719] Hardware and Software

[1720] Server: An IMAP server is used to receive inquiry emails. Natural language processing technology (e.g., spaCy) is used to analyze email content, and a sentiment analysis engine (e.g., VADER) is used for sentiment analysis. A database management system such as MySQL is used for the database, and a template engine (e.g., Jinja2) is used to generate response emails. An SMTP server is used to send emails.

[1721] Specific examples

[1722] Example 1: Product inventory inquiry

[1723] User: A company representative sends an email inquiring about the stock status of Product A.

[1724] Server: Receives the inquiry email and sends the email text to the natural language processing engine.

[1725] Natural language processing engine: Analyzes the email content and determines that the main question is about stock status. The sentiment analysis engine recognizes the user's sentiment as negative. The engine retrieves stock information for product A from the database and generates a response based on that information.

[1726] Generated answer:

[1727] text

[1728] The current stock status of product A is as follows:

[1729] Quantity in stock: 150

[1730] Next expected arrival date: October 15, 2023

[1731] We apologize for the wait and appreciate your understanding.

[1732] This email will include proposal materials such as an inventory management system operation manual and a detailed catalog for product A.

[1733] Server: The generated email is automatically sent to the company representative.

[1734] Example 2: Price enquiry

[1735] User: A company representative sends an email inquiring about the price of Product B.

[1736] Server: Receives the inquiry email and sends the email text to the natural language processing engine.

[1737] Natural language processing engine: Analyzes the email content and determines that "price" is the main topic of the inquiry. The sentiment analysis engine recognizes the user's sentiment as "positive." The latest price information for Product B is retrieved from the price information database and an answer is generated based on that information.

[1738] Generated answer:

[1739] text

[1740] The latest price information for product B is as follows:

[1741] Unit price: 5,000 yen

[1742] Discount information: 10% off this month only

[1743] Please refer to the price list for details. We look forward to your continued support.

[1744] This email will include a price list and details of discount campaigns as proposal materials.

[1745] Server: The generated email is automatically sent to the company representative.

[1746] Prompt Sentence Examples

[1747] Below are some example prompts to input to a generative AI model:

[1748] text

[1749] Design a system that analyzes user inquiry emails, recognizes their emotional state, and generates the best possible response. However, use a template engine to generate the response email and adjust the tone depending on the emotional state.

[1750] This allows the system of the present invention to respond to inquiry emails from companies quickly and accurately while taking into consideration the user's feelings.

[1751] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1752] Step 1: Receiving an inquiry email

[1753] Server: Receives inquiry emails sent by users (company personnel) via the IMAP server. The emails are retrieved from a specific mailbox and the character code is converted to UTF-8.

[1754] Input: New emails from the IMAP mail server

[1755] Output: UTF-8 formatted email text

[1756] Specific operation: Connect to the IMAP server, check for new emails, and retrieve them. Convert the character code to UTF-8.

[1757] Step 2: Analyzing the email content

[1758] Server: The received email text is sent to a natural language processing engine (e.g., spaCy) to analyze the query. The analysis involves tokenization, part-of-speech tagging, and topic extraction.

[1759] Input: UTF-8 formatted email text

[1760] Output: Parsed email content (tokens, part-of-speech tags, subject)

[1761] What it does: Tokenizes email text, tags each word with a part-of-speech tag, and extracts key keywords and phrases.

[1762] Step 3: Sentiment Analysis

[1763] Server: The analyzed email content is sent to an emotion analysis engine (e.g., VADER) to analyze the user's emotional state.

[1764] Input: Parsed email content (tokens, part-of-speech tags, subject)

[1765] Output: Emotional state (positive, negative, neutral)

[1766] Specific operation: The analyzed email content is input into an emotion analysis engine to determine the emotional state.

[1767] Step 4: Categorize the inquiry

[1768] Server: Based on the analysis results of the natural language processing engine, the inquiry content is categorized (e.g., "stock status," "price," etc.).

[1769] Input: Parsed email content, emotional state

[1770] Output: Category information (availability, price, etc.)

[1771] What it does: Maps email content to specific categories and selects the appropriate response template.

[1772] Step 5: Retrieve related data

[1773] Server: Based on the categorized inquiry, retrieves the necessary information from the database, such as stock information and price information.

[1774] Input: Category information

[1775] Output: Retrieved data (stock information, price information, etc.)

[1776] Specific operation: Executes a database query, retrieves the corresponding data, and stores the retrieved data in internal memory.

[1777] Step 6: Generate a response email

[1778] Server: Use a template engine (e.g., Jinja2) to embed the acquired data into the response email. Adjust the tone and content of the response email based on the results of sentiment analysis.

[1779] Input: Template, captured data, emotional state

[1780] Output: Generated response email

[1781] What it does: Embed data in templates to tailor the wording and tone of emails based on emotional state.

[1782] Step 7: Sending an email

[1783] Server: The generated response email is sent to the user via the SMTP server.

[1784] Input: Generated response email

[1785] Output: Email sent

[1786] Specific operation: Connect to the SMTP server, add the generated email to the sending queue, and send it to the user's email address.

[1787] (Application example 2)

[1788] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1789] Current inquiry response systems are required to respond quickly and appropriately to inquiry emails from users. However, previous systems provided mechanical responses without taking emotions into account, which often resulted in a poor user experience. Particularly in the field of e-commerce, flexible responses that match the user's emotions are important, and this aspect needs to be improved.

[1790] The specific processing 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 receiving an inquiry email, means for analyzing the content of the inquiry email, means for generating an appropriate response, means for sending the generated response, means for analyzing the user's emotions, and means for adjusting the tone and content of the response based on the analyzed user's emotions. This enables a quick and appropriate response that takes the user's emotions into consideration.

[1791] An "inquiry email" is an electronic message sent by a user to a business or organization seeking information or support.

[1792] "Means of analysis" refers to the processes and techniques used to understand the content of received inquiry emails and identify their meaning and intent.

[1793] "Natural language processing technology" refers to a series of technologies and methods that allow computers to understand, analyze, and generate human language.

[1794] "Means for generating an appropriate answer" refers to a process for automatically creating the most appropriate answer to a user's inquiry based on the analysis results.

[1795] "Means for analyzing user emotions" refers to a technology that determines the user's emotional state from the content of the inquiry email and classifies it as a numerical value or category.

[1796] "Methods for adjusting the tone and content of responses" refers to techniques for changing the wording and expression of responses to match the analyzed emotional state.

[1797] A "database" is an information system used to manage information related to inquiries (e.g., product inventory, pricing information, etc.).

[1798] "Mobile devices such as smartphones" refers to devices that are portable and can connect to the Internet and use applications.

[1799] MODE FOR CARRYING OUT THE INVENTION

[1800] This invention is an automatic response system for responding quickly and accurately to inquiry emails from companies. This system automatically performs a series of processes, from receiving the inquiry email, analyzing its content, generating and sending a response, and even adjusting the tone of the response to take the user's emotions into consideration.

[1801] System Configuration

[1802] The system includes the following components:

[1803] 1. Server: Equipped with an emotion engine that receives and analyzes inquiry emails, generates and sends responses, and analyzes user emotions.

[1804] 2. Terminal: A mobile terminal such as a PC or smartphone used by a user (company representative). It sends an inquiry email and receives an automatic reply email from the server.

[1805] 3. User: Company representative. Responsible for sending inquiry emails to the system.

[1806] Hardware and software used

[1807] Mail server: Gmail API

[1808] AI engine: Google Cloud Natural Language API

[1809] Sentiment analysis engine: IBM Watson Natural Language Understanding API

[1810] Template engine: Jinja2

[1811] Database: PostgreSQL

[1812] System Operation

[1813] The server first receives the inquiry email. At this time, it retrieves the email using the Gmail API, etc. The received email text is analyzed using the Google Cloud Natural Language API, and tokenization, part-of-speech tagging, and topic extraction are performed.

[1814] The emotional state of the received email is then analyzed using the IBM Watson Natural Language Understanding API, which can determine whether the email content carries a positive, negative, or neutral sentiment.

[1815] Based on the analysis results, an appropriate answer is generated. The answer is generated using the Jinja2 template engine, retrieving the necessary data from the database (PostgreSQL) according to the inquiry content. The tone of the generated answer is adjusted based on the results of sentiment analysis, so that the content takes the user's emotions into consideration.

[1816] Finally, the generated response is automatically sent to the user using an SMTP server, allowing the user to receive a prompt and appropriate response.

[1817] Specific examples

[1818] Example 1: Product inventory inquiry

[1819] A user sends an email inquiry about the stock status of product A. The server receives the inquiry email and sends the email text to the AI ​​engine for analysis. The AI ​​engine determines that "stock status" is the main content of the inquiry, and the emotion engine recognizes the user's emotion as "negative." The server retrieves stock information for product A from the database and generates an answer based on that information. The generated answer includes the following:

[1820] "Product A's current stock status is as follows:

[1821] Number in stock: 150

[1822] Next arrival: October 15, 2023

[1823] We apologize for keeping you waiting and appreciate your understanding."

[1824] Example 2: Price enquiry

[1825] A user sends an email inquiry about the price of product B. The server receives the inquiry email and sends the email text to the AI ​​engine for analysis. The AI ​​engine determines that "price" is the main topic of the inquiry, and the emotion engine recognizes the user's emotion as "positive." The latest price information for product B is retrieved from the price information database, and an answer is generated based on that information. The generated answer includes the following:

[1826] "Product B's latest price information is as follows:

[1827] Unit price: 5,000 yen

[1828] Discount information: 10% off this month only

[1829] Please refer to the price list for details. We look forward to your continued support.

[1830] Prompt Sentence Examples

[1831] "Generate an auto-reply message based on the email sent by the user. The content of the email should be, 'Please let me know if the product is in stock.'"

[1832] This allows the system of the present invention to respond quickly and appropriately while taking into consideration the user's feelings.

[1833] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1834] Step 1:

[1835] The server receives an inquiry email from a user using the Gmail API. At this time, it obtains the email content in text format. The input is the inquiry email sent by the user, and the output is the email content in text format.

[1836] Step 2:

[1837] The server sends the received email text to the Google Cloud Natural Language API for natural language processing. Specific operations include tokenizing the text, tagging parts of speech, and extracting themes to analyze the meaning and intent of the email. The input is the email content in text format, and the output is the analysis results: keywords and themes.

[1838] Step 3:

[1839] The server uses the IBM Watson Natural Language Understanding API to analyze the user's emotions from the email content. Specifically, it uses sentiment analysis technology to determine emotional states such as positive, negative, and neutral as numerical values ​​or categories. The input is the email content in text format, and the output is the result of the sentiment analysis.

[1840] Step 4:

[1841] The server categorizes the query based on the analysis results and retrieves the necessary data from the database. For example, it retrieves inventory and price information from a PostgreSQL database. The input is the subject and keywords resulting from the analysis, and the output is the retrieved related information.

[1842] Step 5:

[1843] The server uses the Jinja2 template engine to embed the retrieved data and generate an appropriate reply email. Furthermore, it adjusts the tone and content of the reply email based on the analysis results of the emotion engine. Specific operations include modifying the email text based on the emotion analysis. The input is the retrieved related information and the emotion analysis results, and the output is the generated reply email.

[1844] Step 6:

[1845] The server automatically sends the generated reply email to the user using the SMTP server. Specifically, it sets the email address, subject, and body, and performs the sending process via the SMTP server. The input is the generated reply email, and the output is the email sent to the user.

[1846] This allows the system to respond to inquiries promptly and appropriately, taking into consideration the user's feelings.

[1847] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1848] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1849] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1850] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1851] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1852] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1853] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1854] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1855] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1856] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1857] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1858] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1859] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1860] 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.

[1861] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1862] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1863] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1864] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1865] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1866] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1867] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1868] The following is further disclosed regarding the above embodiment.

[1869] (Claim 1)

[1870] A means for receiving inquiry emails;

[1871] means for analyzing the content of the inquiry email;

[1872] means for generating an appropriate response based on the analyzed content;

[1873] means for transmitting the generated response;

[1874] A system including:

[1875] (Claim 2)

[1876] 10. The system of claim 1, wherein the analyzing means uses natural language processing techniques.

[1877] (Claim 3)

[1878] 2. The system according to claim 1, further comprising means for acquiring necessary data from a database based on the analyzed content.

[1879] "Example 1"

[1880] (Claim 1)

[1881] A means for receiving inquiry emails;

[1882] means for analyzing the content of the inquiry email using natural language processing technology;

[1883] means for categorizing the inquiry content based on the analysis result;

[1884] means for acquiring related data from a database according to the categorized content;

[1885] a means for generating an appropriate response using the acquired data;

[1886] means for transmitting the generated response;

[1887] A system including:

[1888] (Claim 2)

[1889] The system of claim 1, wherein the analyzing means uses a generative AI model and obtains the analysis result using a prompt sentence.

[1890] (Claim 3)

[1891] 2. The system according to claim 1, wherein the answer generating means uses a template engine to embed the acquired data and generate an answer in a natural style.

[1892] "Application Example 1"

[1893] (Claim 1)

[1894] means for receiving an inquiry message;

[1895] means for analyzing the content of the inquiry message;

[1896] means for generating an appropriate response based on the analyzed content;

[1897] means for transmitting the generated response;

[1898] means for acquiring transaction information from a database based on the analyzed content;

[1899] a means for customizing the response template;

[1900] A system including:

[1901] (Claim 2)

[1902] 10. The system of claim 1, wherein the analyzing means uses natural language processing techniques.

[1903] (Claim 3)

[1904] 10. The system of claim 1, further comprising means for generating an optimal response based on a prompt sentence, the system using a generative AI model to generate the response.

[1905] "Example 2: Combining Emotion Engines"

[1906] (Claim 1)

[1907] A means for receiving inquiry emails;

[1908] means for analyzing the content of the inquiry email using natural language processing technology;

[1909] means for recognizing an emotional state from the content of the email;

[1910] means for acquiring necessary data from a database based on the analyzed content;

[1911] means for generating an appropriate response based on the analyzed content and the acquired data;

[1912] means for transmitting the generated response;

[1913] A system including:

[1914] (Claim 2)

[1915] 2. The system according to claim 1, further comprising means for adjusting the tone and expression of the reply email based on the analyzed content.

[1916] (Claim 3)

[1917] 10. The system of claim 1, further comprising means for using sentiment analysis techniques to determine the emotional state of the inquiry email.

[1918] "Application example 2 when combining emotion engines"

[1919] (Claim 1)

[1920] A means for receiving inquiry emails;

[1921] means for analyzing the content of the inquiry email;

[1922] means for generating an appropriate response based on the analyzed content;

[1923] means for transmitting the generated response;

[1924] means for analyzing user emotions;

[1925] means for adjusting the tone and content of responses based on the analyzed user emotions;

[1926] A system including:

[1927] (Claim 2)

[1928] 10. The system of claim 1, wherein the analyzing means uses natural language processing techniques.

[1929] (Claim 3)

[1930] 2. The system according to claim 1, further comprising means for acquiring necessary data from a database based on the analyzed content.

[1931] (Claim 4)

[1932] The system of claim 1, having the function of receiving an inquiry email from a mobile terminal such as a smartphone, analyzing it, generating and sending a response, and having means including a tone adjustment function based on the user's emotions. [Explanation of symbols]

[1933] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for receiving inquiry emails; means for analyzing the content of the inquiry email; means for generating an appropriate response based on the analyzed content; means for transmitting the generated response; A system including:

2. The system of claim 1 , wherein the analyzing means uses natural language processing techniques.

3. 2. The system according to claim 1, further comprising means for acquiring necessary data from a database based on the analyzed content.

Images