system

Abstract

We provide the system. [Solution] To generate responses to inquiries from employees and partner companies, Means for collecting literature and files and storing them in a database, A method for training an artificial intelligence model based on a saved database, A method for analyzing user inquiries using natural language processing and generating the optimal response using a pre-trained artificial intelligence model, A means of providing the generated answer to the user, A means of automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models, Means to protect communications and data based on security policies, A system that includes this.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In enterprises, the number of daily inquiries from employees and external cooperation companies has increased, increasing the burden on the help desk. As a result, it has become difficult to quickly respond to user inquiries, and the quality of answers may decline. In addition, there is a problem that the man-hours for handling inquiries increase, hindering the efficient use of management resources. Furthermore, a great deal of maintenance man-hours are required for information update and management, and there are also security risks.

Means for Solving the Problems

[0005] In order to solve the above problems, the present invention provides the following means. That is,

[0006] The system provides means for collecting and storing documents and files in a database to generate answers to inquiries from employees and partner companies; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating the optimal answer using the trained artificial intelligence model; means for providing the generated answer to the user; means for automatically collecting new documents and files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on security policies.

[0007] "Documents and files" refers to documents and electronic files used both inside and outside the company, and includes operational manuals, user guides, and support documents.

[0008] A "database" is a collection of information that systematically stores information from various documents and files, making it easy to search and update.

[0009] An "artificial intelligence model" is an algorithm or set of programs that learns patterns and relationships based on input data and performs a specific task (for example, generating answers to inquiries).

[0010] "Natural language processing" is a technology that enables computers to understand, generate, and manipulate human language, and it includes techniques such as tokenizers and stemming.

[0011] An "inquiry" is a question or request made by a user seeking specific information or assistance.

[0012] An "optimal response" is a response that provides the most appropriate and accurate information in response to a user's inquiry.

[0013] "Automated collection" is the process by which a system self-containedly acquires necessary literature and files and adds them to a database without human intervention.

[0014] "Update" is a process of adding new information to existing data or models to maintain the latest state.

[0015] "Security policy" refers to guidelines and procedures for protecting data and communications from unauthorized access and tampering.

Brief Description of Drawings

[0016] [Figure 1] It is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] It is a conceptual diagram showing an example of the main functions of a data processing device and a smart device according to the first embodiment. [Figure 3] It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] [[ID=2�]]It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which multiple emotions are mapped. [Figure 10] It shows an emotion map to which multiple emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13]It is a sequence diagram showing the processing flow of the data processing system in Example 2 when combined with an emotion engine. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when combined with an emotion engine.

Embodiments for Carrying Out the Invention

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

[0018] First, the terms used in the following description will be explained.

[0019] In the following embodiments, the numbered processor (hereinafter simply referred to as "processor") may be one arithmetic unit or a combination of a plurality of arithmetic units. Also, the processor may be one type of arithmetic unit or a combination of a plurality of types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

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

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

[0022] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0024] [First Embodiment]

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

[0026] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0027] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0028] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0029] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0030] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

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

[0033] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0034] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0035] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0036] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0037] This invention is a system for streamlining the handling of inquiries both inside and outside a company. This system automatically collects documents and files, stores them in a database, trains an artificial intelligence model on them, and provides optimal answers to user inquiries. The specific program processing of this system is described below.

[0038] 1. Data Collection Phase

[0039] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in an integrated database.

[0040] 2. AI Model Training Phase

[0041] The server trains an artificial intelligence model based on literature and files stored in an integrated database. First, the data is preprocessed, and the text data is analyzed using natural language processing tools such as tokenizers and stemming. Then, a training dataset is created based on the preprocessed data, and the artificial intelligence model is trained using a machine learning algorithm (e.g., a Transformer-based model).

[0042] 3. Query Processing Phase

[0043] The user accesses the system using their device and enters an inquiry. For example: "Please tell me how to use the new office automation system." The device then sends this input to the server.

[0044] The server analyzes the received inquiry using natural language processing to understand its content. Then, based on the analyzed content, it uses a pre-trained artificial intelligence model to generate the optimal response. This generated response is then sent back to the user's device.

[0045] Specific example

[0046] The following is a concrete example illustrating how actual queries are processed.

[0047] User: "Please teach me how to use the new office automation system."

[0048] The device sends this query to the server.

[0049] The server analyzes the received text using a natural language processing tool (e.g., a tokenizer) and extracts "new OA system" and "how to use" as keywords.

[0050] The server uses a trained artificial intelligence model to extract information on "how to use the new office automation system" from relevant literature and generates the optimal response. This response is generated in a format such as, "The procedure for using the new office automation system is as follows..."

[0051] The device displays the generated answers to the user. The user can obtain specific and timely answers to their questions.

[0052] Maintenance Phase

[0053] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. Furthermore, it retrains its artificial intelligence model based on this new data to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy to ensure the security of the system.

[0054] In this way, the system can efficiently and securely handle inquiries from employees and partner companies. As a result, the workload of the help desk is reduced, and the quality of service to users is improved.

[0055] The following describes the processing flow.

[0056] Step 1:

[0057] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal and external repositories and online storage. This is done using scripts that automatically crawl the files at regular intervals.

[0058] Step 2:

[0059] The server analyzes the collected literature and files and generates metadata (e.g., file name, creation date, category). This analysis uses programs to analyze file formats and structures.

[0060] Step 3:

[0061] The server stores the contents of the documents and files, along with the generated metadata, in an integrated database. This storage process uses a storage management system to organize and store the files.

[0062] Step 4:

[0063] The server preprocesses the literature and files stored in the integrated database using natural language processing tools. Specifically, it uses a tokenizer to split the text and stemming to extract the core forms of words.

[0064] Step 5:

[0065] The server generates a training dataset based on preprocessed data and trains an artificial intelligence model using a machine learning algorithm (e.g., a Transformer-based model). During this process, the model's performance is evaluated by dividing it into training and test data.

[0066] Step 6:

[0067] The user enters their inquiry using their device. For example, they might type, "Please tell me how to use the new office automation system." The device then sends this input to the server.

[0068] Step 7:

[0069] The server analyzes the received query using natural language processing tools to extract key keywords and their relationships. This analysis utilizes tools for noun extraction and dependency analysis.

[0070] Step 8:

[0071] The server uses a trained artificial intelligence model based on the analysis results to generate the optimal response. For example, it might generate a specific response such as, "The procedure for using the new office automation system is as follows..."

[0072] Step 9:

[0073] The server sends the generated response back to the user's device. The device then displays this response to the user.

[0074] Step 10:

[0075] The server periodically recollects new literature and updated files, updating the database and artificial intelligence models. This includes the process of analyzing and retraining on new data.

[0076] Step 11:

[0077] The server encrypts all data communications and implements access control based on security policies. This ensures the security of the entire system.

[0078] (Example 1)

[0079] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0080] Responding to inquiries both inside and outside the company requires considerable effort and time, making efficient work difficult. Furthermore, there is a lack of systems to properly manage information provided by various departments and partner companies and to provide prompt and accurate responses. Keeping up-to-date with daily updates is also a major challenge. This invention aims to solve these problems and provide a system that streamlines inquiry handling.

[0081] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0082] In this invention, the server includes means for collecting and storing documents and electronic files in a database for generating answers to inquiries from information providers; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating the optimal answer using the trained artificial intelligence model; means for providing the generated answer to the user; means for automatically collecting new documents and electronic files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on security policies. This makes it possible to respond to inquiries efficiently and accurately based on the latest information.

[0083] "Information provider" refers to the department or partner company that provides documents or electronic files.

[0084] An "inquiry" refers to a question or request that a user enters into the system.

[0085] "Documents" refer to text data such as business manuals and user guides.

[0086] An "electronic file" is data stored in a digital format, including PDFs and document files.

[0087] A "database" refers to a system for managing and storing collected documents and electronic files.

[0088] An "artificial intelligence model" refers to a program trained using machine learning algorithms that generates responses based on input data.

[0089] "Natural language processing" refers to the technology of using computers to understand, analyze, and generate human language.

[0090] A "tokenizer" refers to a software tool that divides text into words or phrases.

[0091] "Stemming" refers to a natural language processing technique that extracts the root portion of a word.

[0092] A "security policy" refers to the rules and procedures for ensuring the security of data and communications.

[0093] "Users" refer to individuals or organizations that submit inquiries to the system and receive responses.

[0094] "Automated data collection" refers to a process in which the system automatically collects data without requiring manual operation.

[0095] "Communication" refers to the sending and receiving of data and information between systems.

[0096] "Answer" refers to the response that the system generates in response to a user's inquiry.

[0097] "Updating" refers to the process of adding new information to databases and artificial intelligence models to keep them up-to-date.

[0098] This invention is a system for streamlining the handling of inquiries both inside and outside a company. This system automatically collects documents and electronic files, stores them in a database, trains an artificial intelligence model based on the stored data, and provides optimal answers to user inquiries. Specific embodiments of this system are described below.

[0099] The server automatically collects documents and electronic files provided by various departments and partner companies within the company at regular intervals. These documents and electronic files include instruction manuals and guidelines. The server uses data collection scripts and APIs to access the internet and internal networks and store this data in an integrated database. Hardware such as dedicated servers and cloud storage is used in this data collection process.

[0100] The server trains an artificial intelligence model based on documents and electronic files stored in an integrated database. First, the server preprocesses the data and analyzes the text data using natural language processing tools such as tokenizers and stemming (e.g., NLTK, spaCy). Then, it creates a training dataset based on the preprocessed data and trains a Transformer-based model (e.g., BERT, GPT) using a machine learning framework (e.g., TENSORFLOW®, PyTorch).

[0101] Users access the system using their devices and enter inquiries. For example, they might ask, "How do I use the new office automation system?" The device sends this input to the server. The server analyzes the received inquiry using natural language processing tools and understands its content. Based on the analysis, it then generates the optimal answer using a trained artificial intelligence model. This generated answer is then sent back to the user's device.

[0102] The following is a concrete example illustrating how actual queries are processed.

[0103] "Please teach me how to use the new office automation system."

[0104] "What are the latest guidelines regarding employee onboarding and offboarding procedures?"

[0105] "I want to know how to set up project management software."

[0106] The server automatically collects new documents and updated electronic files from repositories and online storage on a regular basis, updating the database. Furthermore, it retrains its artificial intelligence model based on this new data, enabling it to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy and ensures system security using the SSL / TLS protocol.

[0107] In this way, the system can efficiently and securely handle inquiries from employees and partner companies. As a result, the workload of the help desk is reduced, and the quality of service to users is improved.

[0108] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0109] Step 1: Prepare for data collection

[0110] The server configures connections to collect documents and electronic files from various departments and partner companies. This includes configuring API and FTP connections.

[0111] Input: Connection information for API endpoints and FTP servers of each department and partner company.

[0112] Output: A list of endpoints that have completed the necessary connection settings for data collection and are therefore available for connection.

[0113] Step 2: Perform data collection

[0114] The server automatically starts collecting data at specified time intervals. For example, it can trigger a collection task every Monday at 9:00 AM.

[0115] Input: Regular schedule information.

[0116] Output: A list of documents and electronic files downloaded from each endpoint.

[0117] Specific operation: The server makes HTTP requests or FTP connections to API endpoints or FTP servers of each department or partner company to download documents and files.

[0118] Step 3: Save the data to the database

[0119] The server saves downloaded documents and electronic files to a temporary directory, and then moves this data to the integrated database.

[0120] Input: Downloaded documents and electronic files.

[0121] Output: Entries for documents and electronic files stored in the integrated database.

[0122] Specific actions: Use scripts and ETL tools to perform data transformation and cleaning for database integration.

[0123] Step 4: Data Preprocessing

[0124] The server extracts data from the integrated database and performs preprocessing. Specifically, it tokenizes the text data using a tokenizer and performs stemming and stop word removal.

[0125] Input: Documents and electronic files extracted from an integrated database.

[0126] Output: Preprocessed text data.

[0127] Specific operation: Preprocessing is performed using natural language processing tools (e.g., NLTK, spaCy).

[0128] Step 5: Training the artificial intelligence model

[0129] The server creates a training dataset based on preprocessed data and trains the model using a machine learning framework (e.g., TensorFlow, PyTorch).

[0130] Input: Preprocessed text data.

[0131] Output: A trained artificial intelligence model.

[0132] Specific operation: Train the model using a Transformer-based model (e.g., BERT, GPT).

[0133] Step 6: Enter your inquiry

[0134] The user accesses the system using their own device and enters their inquiry. For example, they might enter, "Please tell me how to use the new office automation system."

[0135] Input: User inquiry details.

[0136] Output: The content of the query sent to the server.

[0137] Specific operation: The terminal sends the user's input query to the server as an HTTP request.

[0138] Step 7: Analyze the inquiry

[0139] The server analyzes the received query using natural language processing tools. Specifically, it uses a tokenizer to split the text and extract keywords.

[0140] Input: User inquiry details.

[0141] Output: Extracted keywords and query structure.

[0142] Specific operation: Performs text analysis using a tokenizer and stemming.

[0143] Step 8: Generating the answer

[0144] The server uses a trained artificial intelligence model to generate the best possible answer to the user's inquiry.

[0145] Input: Extracted keywords or query structure.

[0146] Output: The generated answer text.

[0147] Specific operation: Generate an answer using a pre-trained model and send that answer to the terminal as an HTTP response.

[0148] Step 9: Display the answer

[0149] The terminal displays the response sent back from the server to the user.

[0150] Input: Response text from the server.

[0151] Output: The answer displayed to the user.

[0152] Specific operation: Render the answer text on the screen and display it visually to the user.

[0153] Step 10: Database Update

[0154] The server automatically collects new documents and updated files from repositories and online storage on a regular basis and updates the database.

[0155] Input: A new document or electronic file.

[0156] Output: Updated integrated database.

[0157] Specific operation: The automated collection task is executed periodically, adding new data to the database.

[0158] Step 11: Retrain the model

[0159] The server retrains the artificial intelligence model using the new data.

[0160] Input: Updated text data.

[0161] Output: A retrained artificial intelligence model.

[0162] Specific operation: The machine learning algorithm is run again using new data to train the model with the latest information.

[0163] Step 12: Security Protection

[0164] The server uses the SSL / TLS protocol to encrypt all data communications.

[0165] Input: Communication data.

[0166] Output: Encrypted communication data.

[0167] Specific operation: Implements the SSL / TLS protocol to encrypt and protect communication data.

[0168] (Application Example 1)

[0169] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0170] Conventional systems make it difficult to improve operational efficiency and provide prompt responses to inquiries in logistics centers. In particular, the lack of a system that allows workers to quickly obtain necessary information makes delays and errors more likely. The present invention aims to solve these problems and provide a system that improves the operational efficiency of logistics centers.

[0171] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0172] In this invention, the server includes means for collecting documents and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating optimal answers using the trained artificial intelligence model; means for responding to inquiries from smart devices to provide logistics center work guidelines and procedures; means for providing the generated answers to users; means for automatically collecting new documents and files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on security policies. This enables improved work efficiency and quick and accurate response to inquiries in the logistics center.

[0173] "Documents and files" refer to documents and electronic files related to business operations both inside and outside the company, such as operational manuals, user guides, and support documents.

[0174] A "database" is a system that centrally manages collected literature and files, making them searchable and accessible.

[0175] An "artificial intelligence model" is an AI system trained using machine learning algorithms that generates the optimal answer to a specific task.

[0176] "Natural language processing" is a language processing technique that analyzes text-based inquiries from users and understands their meaning.

[0177] A "logistics center" is a warehouse-type facility that handles operations such as receiving, storing, and shipping goods and materials.

[0178] "Smart devices" are advanced electronic devices with communication capabilities, such as smartphones, smart glasses, and head-mounted displays.

[0179] A "security policy" is a set of rules and procedures aimed at protecting data and communications.

[0180] "Work guidelines" are guidelines that describe the procedures and methods for various tasks in a logistics center.

[0181] An "inquiry" is a question or request that a user enters into the system to seek information.

[0182] "Answer" refers to the information provided in response to a user's inquiry.

[0183] This invention is a system aimed at improving the efficiency of operations and speeding up inquiry response in logistics centers. This system automatically collects documents and files, stores them in a database, trains an artificial intelligence model on them, and provides optimal answers to user inquiries. The specific program processing of this system is described below.

[0184] 1. Data Collection Phase

[0185] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in a database.

[0186] 2. AI Model Training Phase

[0187] The server trains an artificial intelligence model based on literature and files stored in the database. First, the data is preprocessed, and the text data is analyzed using natural language processing tools such as tokenizers and stemming. This process uses libraries such as spaCy and Huggingface Transformers. After that, a training dataset is created based on the preprocessed data, and the artificial intelligence model is trained using machine learning algorithms such as the BERT model.

[0188] 3. Query Processing Phase

[0189] The user accesses the system using their smart device (e.g., smartphone or smart glasses) and enters an inquiry. For example: "Please tell me how to package the new product." The device sends this input to the server. The server analyzes the received inquiry using natural language processing tools and understands its content. Based on the analysis, it generates the optimal answer using a trained artificial intelligence model. This generated answer is then sent back to the user's device.

[0190] Specific example

[0191] Consider a scenario where a user asks via smartphone, "How should I pack the product?" This inquiry is analyzed by the server, and the optimal packing procedure is displayed based on relevant manuals.

[0192] 4. Maintenance Phase

[0193] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. Furthermore, it retrains its artificial intelligence model based on this new data to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy to ensure the security of the system.

[0194] Example of a prompt

[0195] As a concrete example, the following prompt statements are possible:

[0196] User: "Please tell me how to pack the product."

[0197] Prompt: "Analyze the information on product packaging methods in the logistics center's work procedures and manuals, and indicate the optimal procedure."

[0198] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0199] Step 1: Data Collection

[0200] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in a database.

[0201] Input: Literature and files

[0202] Output: Data stored in the database

[0203] Specific operation: Retrieve data from file folders or online storage and upload it to the database system.

[0204] Step 2: Data Preprocessing

[0205] The server preprocesses the stored literature and files. It analyzes the text data using natural language processing tools such as tokenizers and stemming tools (e.g., spaCy).

[0206] Input: Literature and files stored in the database

[0207] Output: Preprocessed text data

[0208] Specific operation: Tokenize the text, perform stemming, and save the analysis results as structured data.

[0209] Step 3: AI Model Training

[0210] The server trains an artificial intelligence model using machine learning algorithms such as the BERT model, based on preprocessed data. This data is processed using the Huggingface Transformers library.

[0211] Input: Preprocessed text data

[0212] Output: Trained artificial intelligence model

[0213] Specific operation: Preprocessed text data is created as a training dataset, and training is performed using the BERT model.

[0214] Step 4: Enter your inquiry

[0215] Users access the system using their smart devices and enter their inquiries. For example, they might submit an inquiry such as, "Please tell me how to pack the product."

[0216] Input: User inquiry text

[0217] Output: Sending a query to the server

[0218] Specific operation: The user enters text through the interface of a smart device and sends that inquiry to the server.

[0219] Step 5: Inquiry Analysis

[0220] The server analyzes the received query using a natural language processing tool (e.g., spaCy) to understand its content.

[0221] Input: User inquiry text

[0222] Output: Analyzed keywords and queries

[0223] Specific operation: The received inquiry text is tokenized, keywords are extracted, and queries are generated.

[0224] Step 6: Generate Answer

[0225] The server generates the optimal response using a pre-trained artificial intelligence model based on the analyzed data. This generated response includes information extracted from relevant literature.

[0226] Input: Analyzed keywords and queries

[0227] Output: The best answer for the user

[0228] Specific operation: Using a pre-trained artificial intelligence model, extract information based on keywords and queries, and generate answers in natural language.

[0229] Step 7: Provide your answer

[0230] The server sends the generated answer back to the user's device and displays it to the user. The user can obtain specific and timely answers to their questions.

[0231] Input: Response text from the server

[0232] Output: The answer displayed on the user's smart device.

[0233] Specific operation: The response text sent from the server is displayed on the user's device.

[0234] Step 8: Maintenance

[0235] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. It also retrains artificial intelligence models based on this new data. Furthermore, all data communications are encrypted according to security policies to ensure system security.

[0236] Input: New documents and files, security policy

[0237] Output: Updated database and retrained artificial intelligence model

[0238] Specific actions: Regularly collect and store data, retrain artificial intelligence models, and implement security measures.

[0239] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0240] This invention provides a system that streamlines the handling of inquiries from employees and partner companies, and recognizes user emotions to provide appropriate responses. This system automatically collects documents and files, stores them in a database, and trains an artificial intelligence (AI) model. Furthermore, by combining this with an emotion engine, it recognizes user emotions and provides appropriate responses tailored to those emotions.

[0241] Data collection phase

[0242] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal repositories and online storage. The collected documentation and files are stored in an integrated database at regular intervals.

[0243] AI model training phase

[0244] The server preprocesses literature and files stored in the integrated database using natural language processing (NLP) tools. Specifically, it uses a tokenizer to segment the text and stemming to extract the core forms of words. Then, it generates a training dataset based on the preprocessed data and trains an AI model using a machine learning algorithm (e.g., a transformer model). This model is used to generate the best possible answers to user inquiries.

[0245] Query processing phase

[0246] The user enters their inquiry into the system from their terminal. For example, they might enter, "Please tell me how to use the new office automation system." This input is then sent from the terminal to the server.

[0247] The server analyzes the received inquiry using NLP tools to extract key keywords and their relationships. It then uses a pre-trained AI model to generate the optimal response.

[0248] Emotion recognition phase

[0249] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis. The emotion engine analyzes the user's input text and adjusts the tone and content of the AI ​​model's response accordingly. For example, if the user indicates that they are "distressed," the response will be adjusted to a gentle and polite tone.

[0250] Answer generation phase

[0251] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The device then displays this response to the user. This allows the user to receive specific and timely responses that are tailored to their emotions.

[0252] Specific example

[0253] The following is an example of how to process a specific query using the emotion engine.

[0254] User: "I'm having trouble figuring out how to use the new office automation system. Please help me."

[0255] The device sends this query to the server.

[0256] The server analyzes the received text using an NLP tool and extracts keywords such as "OA system," "how to use," and "having trouble," along with the associated emotions.

[0257] The server activates an emotion engine based on the extracted emotion "distressed," and uses an AI model to generate a gentle and polite response, such as "Don't worry, here's how to use the new OA system..."

[0258] The device displays the generated response to the user. By receiving a response that is tailored to their own emotions, the user can resolve the problem with confidence.

[0259] Maintenance Phase

[0260] The server periodically recollects new literature and updated files, updating the database and AI models. This includes the process of analyzing new data and retraining the AI ​​models. Furthermore, the server ensures system security by encrypting all data communications and implementing access control based on security policies.

[0261] In this way, this system, equipped with an emotion engine, enables efficient and emotionally sensitive inquiry handling, contributing to improved user experience and reduced workload for the help desk.

[0262] The following describes the processing flow.

[0263] Step 1:

[0264] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal and external repositories and online storage. This collection is performed at specific time intervals or triggered by specific events.

[0265] Step 2:

[0266] The server analyzes the collected literature and files and generates file metadata (e.g., file name, creation date, category). This metadata generation uses tools for file format detection and text extraction.

[0267] Step 3:

[0268] The server stores the contents of the documents and files, along with the generated metadata, in an integrated database. This storage process uses a database management system to normalize the data and make it searchable efficiently.

[0269] Step 4:

[0270] The server preprocesses the literature and files stored in the integrated database using natural language processing tools. Specifically, this preprocessing involves splitting the text using a tokenizer and converting words into their stem forms using stemming.

[0271] Step 5:

[0272] The server generates a training dataset based on preprocessed data and trains an artificial intelligence model using machine learning algorithms. This training includes cross-validation between the training dataset and the test dataset.

[0273] Step 6:

[0274] The user accesses the system using a terminal and enters an inquiry. For example, enter "I'm really struggling because I don't understand how to use the new OA system at all. Please help." The terminal sends this input to the server.

[0275] Step 7:

[0276] The server analyzes the received inquiry content using a natural language processing tool and extracts the main keywords and their relationships. For example, keywords such as "OA system", "usage method", and "being troubled" are extracted.

[0277] Step 8:

[0278] The server activates the sentiment engine based on the extracted keywords. The sentiment engine analyzes the sentiment such as "being troubled" from the user's text and adjusts the tone and content of the answer based on that sentiment.

[0279] Step 9:

[0280] The server generates an optimal answer using a trained artificial intelligence model based on the results of the sentiment engine. For example, it generates an answer in a gentle tone such as "Please don't worry. The usage procedure of the new OA system is as follows..." [[ID=​​​​​​​​​​​​​​​​ Step 12:

[0286] The server encrypts all data communications and implements access control based on the security policy, thereby ensuring the security of the entire system.

[0287] (Example 2)

[0288] Next, Example 2 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart device 14 is referred to as the "terminal".

[0289] Conventional inquiry response systems ignored the user's emotions and mechanically provided answers, making it difficult to improve user satisfaction. Also, sufficient measures are required for model updates and data security to provide optimal answers according to the inquiry content. The present invention addresses such problems and aims to achieve efficient and emotion - considerate inquiry response.

[0290] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0291] In this invention, the server includes means for collecting documents and files and storing them in a database, means for training an artificial intelligence model based on the stored database, means for analyzing inquiries from users by natural language processing and generating an optimal answer using the trained artificial intelligence model, means for providing the generated answer to the user, means for automatically collecting new documents and files periodically and updating the database and the artificial intelligence model, means for protecting communications and data based on the security policy, and means for performing sentiment analysis based on the content of the user's inquiry and adjusting the answer in an appropriate tone and content according to the analysis result. Thereby, efficient and emotion - considerate inquiry response becomes possible.

[0292] "Documents and files" refers to a collection of data containing information necessary for responding to inquiries, such as operational manuals, user guides, and support documents.

[0293] A "database" is a system for efficiently storing, managing, and searching for documents and files.

[0294] An "artificial intelligence model" is a model trained to generate the best possible answer to a query using machine learning algorithms.

[0295] "Natural language processing" is a technique that uses tools such as tokenizers and stemming to analyze user inquiries.

[0296] "Sentiment analysis" is the process of identifying emotions from the content of a user's inquiry and adjusting the response to match those emotions in terms of tone and content.

[0297] "Means of protecting communications and data" refer to technologies that encrypt data communications using SSL / TLS, etc., and restrict access to system resources using access control lists (ACLs).

[0298] "Means for automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models" refers to the process of continuously collecting new information, updating databases based on that information, and retraining AI models to keep them up-to-date.

[0299] "Methods for analyzing inquiries and generating answers" refers to a function that analyzes user inquiries using NLP tools and generates the optimal answer using a pre-trained artificial intelligence model.

[0300] This invention is a system that streamlines the handling of inquiries from employees and partner companies, and recognizes user emotions to provide appropriate responses. This system automatically collects documents and files, stores them in a database, and trains an artificial intelligence model. Furthermore, by combining this with an emotion engine, it recognizes user emotions and provides appropriate responses tailored to those emotions.

[0301] Data collection phase

[0302] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal repositories and online storage. The collected documentation and files are periodically saved to an integrated database. This process is performed using a crawler and a database management system (e.g., MySQL®).

[0303] AI model training phase

[0304] The server preprocesses literature and files stored in the integrated database using natural language processing (NLP) tools (e.g., NLTK, Spacy). It splits the text using a tokenizer and extracts word stems using stemming. Then, it generates a training dataset based on the preprocessed data and trains an AI model using machine learning algorithms (e.g., BERT, GPT-3®). This AI model is used to generate the best possible answers to user inquiries.

[0305] Query processing phase

[0306] The user enters their inquiry into the system from their terminal. For example, they might enter, "Please tell me how to use the new office automation system." This input is then sent from the terminal to the server.

[0307] The server analyzes the received inquiry using NLP tools to extract key keywords and their relationships. It then uses a pre-trained AI model to generate the optimal response.

[0308] Emotional recognition phase

[0309] The server activates the emotion engine based on the content of the user's inquiry and performs emotion analysis. The emotion engine analyzes the emotion from the user's input text and adjusts the tone and content of the answer of the AI model according to the emotion. For example, when the user shows an emotion of "being troubled", the answer is adjusted to a gentle and polite tone.

[0310] Answer generation phase

[0311] The server generates an answer adjusted by the emotion engine and returns it to the user's terminal. The terminal displays this answer to the user. As a result, the user can obtain a specific and prompt answer adapted to their emotion.

[0312] Maintenance phase

[0313] The server periodically re-collects new documents and updated files, and updates the database and the AI model. This process includes the work of analyzing new data and re-training the AI model. In addition, the server encrypts all data communications and implements access control based on the security policy to ensure the security of the system.

[0314] Specific example

[0315] The following shows a specific example of processing an inquiry using the emotion engine.

[0316] User: "I'm completely lost about how to use the new OA system and I'm in trouble. Please help me."

[0317] The terminal sends this inquiry to the server.

[0318] The server analyzes the received text with an NLP tool and extracts the keywords and emotions of "OA system", "usage method", and "being troubled".

[0319] The server activates an emotion engine based on the extracted emotion "distressed," and uses an AI model to generate a gentle and polite response, such as "Don't worry, here's how to use the new OA system..."

[0320] The device displays the generated response to the user. By receiving a response that is tailored to their own emotions, the user can resolve the problem with confidence.

[0321] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0322] Step 1:

[0323] The server executes data collection scripts to download operational manuals, user guides, and support documents from internal repositories and online storage.

[0324] Input: URLs of internal repositories and online storage, API keys, etc.

[0325] Output: Downloaded literature and files

[0326] Specific operation: The server uses a crawler to download literature and files from the specified URL via FTP or API.

[0327] Step 2:

[0328] The server saves downloaded documents and files to an integrated database.

[0329] Input: Downloaded literature and files

[0330] Output: Data stored in the integrated database

[0331] Specific operation: The server inserts the literature and files into the database management system (e.g., MySQL).

[0332] Step 3:

[0333] The server preprocesses the literature and files stored in the integrated database using NLP tools.

[0334] Input: Literature and files stored in the integrated database

[0335] Output: Preprocessed text data

[0336] Specific operation: The server uses NLP tools (NLTK, Spacy) to split the text with a tokenizer and perform stemming.

[0337] Step 4:

[0338] The server generates a training dataset based on pre-processed data.

[0339] Input: Preprocessed text data

[0340] Output: Training dataset

[0341] Specific operation: The server converts the preprocessed data into a CSV file, and then uses TensorFlow or PyTorch to generate a training dataset based on that file.

[0342] Step 5:

[0343] The server uses machine learning algorithms to train the AI ​​model.

[0344] Input: Training dataset

[0345] Output: Trained AI model file

[0346] Specific operation: The server trains an AI model using a Transformer model (e.g., the Hugging Face Transformers library) and saves it to "transformer_model.bin".

[0347] Step 6:

[0348] The user enters an inquiry from their device, and the device sends that information to the server.

[0349] Input: Inquiry details (Example: "Please tell me how to use the new office automation system.")

[0350] Output: Query data sent to the server

[0351] Specific operation: The user enters a query into an input form on their device, and the device uses a REST API to send this data to the server.

[0352] Step 7:

[0353] The server analyzes the received query content using an NLP tool.

[0354] Input: Inquiry details

[0355] Output: Key keywords and their relationships

[0356] Specific operation: The server uses Spacy to extract key keywords such as "OA system" and "how to use".

[0357] Step 8:

[0358] The server uses a trained AI model to generate the optimal answer.

[0359] Input: Analyzed query content

[0360] Output: Generated answer

[0361] Specific operation: The server loads "transformer_model.bin" and generates a response based on the query.

[0362] Step 9:

[0363] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis.

[0364] Input: Inquiry details

[0365] Output: Emotion analysis results

[0366] Specific operation: The server calls an emotion engine (e.g., Google® Cloud Natural Language API) to analyze emotions and identify the emotion "distressed".

[0367] Step 10:

[0368] The server adjusts the tone and content of the AI ​​model's responses based on the sentiment analysis results.

[0369] Input: Sentiment analysis results, generated responses

[0370] Output: Adjusted answer

[0371] Specific action: Based on the emotion "distressed," the server changes the tone of its response to a gentler and more polite one.

[0372] Step 11:

[0373] The server sends the adjusted response back to the terminal, which then displays it to the user.

[0374] Input: Adjusted answer

[0375] Output: Answer displayed to the user

[0376] Specific operation: The server sends the prepared response back to the terminal in JSON format as a REST API response, and the terminal displays it on the screen.

[0377] Step 12:

[0378] The server periodically recollects new literature and updated files, updating the database and AI models.

[0379] Input: New literature or updated files

[0380] Output: Updated database and AI model

[0381] Specific operation: The server uses scheduled jobs to collect new data, insert it into the database, and retrain the AI ​​model based on the new data.

[0382] Step 13:

[0383] The server encrypts all data communications and implements access control.

[0384] Input: Communication data, access request

[0385] Output: Encrypted data, secure access

[0386] Specific operation: The server encrypts data communication using SSL / TLS and ensures security using Access Control Lists (ACLs).

[0387] (Application Example 2)

[0388] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0389] Previously, responses to inquiries from employees and partner companies were inefficient, and responses that considered user emotions were not provided. As a result, improvements in user experience and the efficiency of inquiry handling were not fully achieved. Furthermore, in security services, there is a need for responses that give users peace of mind in times of anxiety or emergency.

[0390] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting literature and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for recognizing the user's emotions, analyzing them using natural language processing, and generating an optimal response using the trained artificial intelligence model; means for providing the generated response to the user; means for adjusting the tone of the response based on the user's emotions using an emotion recognition engine; means for automatically collecting new literature and files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on a security policy. This makes it possible to provide prompt and appropriate inquiry response and security services that take into consideration the user's emotions.

[0391] "Documents and files" refer to materials containing information necessary to respond to user inquiries, including operational manuals, user guides, and security-related guidelines.

[0392] A "database" is an information system that stores collected literature and files, and allows for the efficient management and retrieval of information necessary to respond to inquiries.

[0393] An "artificial intelligence model" is an algorithm that learns from collected literature and files and is used to automatically generate the best possible answers to user inquiries.

[0394] "Natural language processing" is a technology used to analyze user inquiries and extract key keywords and their relationships.

[0395] An "emotion recognition engine" is a system that analyzes emotions from the user's input text and adjusts the tone and content of the response accordingly.

[0396] A "security policy" is a set of rules and guidelines designed to ensure data communication and information protection within a system.

[0397] "Response tone" refers to the tone and attitude of the response, which are adjusted with consideration for the user's feelings.

[0398] "New documents and files" refer to materials that are added regularly, including the latest operational manuals, user guides, and security-related guidelines.

[0399] The system for carrying out this invention is a response system that integrates multiple means to provide prompt and appropriate answers to inquiries while taking into consideration the user's feelings. This system mainly uses the following hardware and software.

[0400] First, the server automatically collects documents and files from internal repositories and online storage and stores them in an integrated database. These documents and files include operational manuals, user guides, and security-related guidelines. The server uses the requests library to retrieve these documents.

[0401] Next, the server trains an artificial intelligence model based on the stored literature and files. This process utilizes natural language processing (NLP) tools. Specifically, it uses a tokenizer to segment the text and stemming to extract the core forms of words. Then, it generates a training dataset based on the preprocessed data and trains the AI ​​model using a machine learning algorithm (e.g., a transformer model). The pipeline function of the transformers library is used in this training phase.

[0402] Users enter inquiries into the system from their devices. For example, if a user makes an inquiry such as, "I'm scared because I heard sounds like someone broke into my house tonight," this input is sent from the device to the server.

[0403] The server analyzes the received inquiry using natural language processing tools to extract key keywords and their relationships. Simultaneously, it analyzes the user's emotions using an emotion recognition engine and adjusts the tone and content of the AI ​​model's response according to the emotions the user expresses.

[0404] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The user's device then displays this response to the user. In this way, the user can obtain a specific and timely response that is tailored to their emotions.

[0405] For example, if a user enters a question saying, "I'm scared because I heard sounds like someone broke into my house tonight," it will be processed as follows:

[0406] 1. Receive a user inquiry.

[0407] 2. Natural language processing tools were used to extract the keywords and emotions "intrusion" and "scary".

[0408] 3. The emotion analysis engine determined it to be "fear".

[0409] 4. Based on appropriate advice regarding "intrusion," adjust the tone of your response according to your emotions.

[0410] 5. Generate and provide the user with a response such as, "Don't worry. Contact the police immediately and evacuate to a safe place."

[0411] Examples of prompts to input into a generative AI model are as follows:

[0412] Q: I'm scared because I heard sounds tonight that sound like someone broke into my house.

[0413] A: Don't worry. Contact the police immediately and evacuate to a safe place.

[0414] In this way, this system, equipped with an emotion engine, provides a concrete means to achieve efficient inquiry handling that takes emotions into consideration.

[0415] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0416] Step 1:

[0417] The server automatically collects documents and files from internal repositories and online storage. It retrieves documents such as operational manuals, user guides, and security-related guidelines using the requests library and stores their contents in an integrated database. Input consists of documents and files in internal repositories and online storage, while output consists of documents and files stored in the integrated database.

[0418] Step 2:

[0419] The server trains an artificial intelligence model based on literature and files stored in an integrated database. Natural language processing (NLP) tools split the text (tokenizer) and extract the core form of words using stemming. Then, a training dataset is generated using the preprocessed data, and a transformer model is trained using the pipeline function of the transformers library. The input is literature and files stored in the integrated database, and the output is the trained artificial intelligence model.

[0420] Step 3:

[0421] The user enters the inquiry using their device. For example, the text might be, "I'm scared because I heard sounds like someone broke into my house tonight." The input is the user's inquiry text, and the output is the inquiry sent from the device to the server.

[0422] Step 4:

[0423] The server analyzes user inquiries using natural language processing tools to extract key keywords and their relationships. This analysis utilizes tokenizers and stemming, and an emotion recognition engine is used to analyze the user's emotions. The input is the user's inquiry text, and the output is the extracted keywords and emotion data.

[0424] Step 5:

[0425] The server generates the optimal response using a trained artificial intelligence model based on extracted keywords and user sentiment. This process involves generating prompts, inputting them into the AI ​​model, and obtaining the response. The input consists of extracted keywords and sentiment data, while the output is the generated response text.

[0426] Step 6:

[0427] The server sends a response, refined by the emotion recognition engine, to the user's device, which then displays this response to the user. The input is the generated response text, and the output is the response displayed on the user's device.

[0428] Step 7:

[0429] The server automatically collects new literature and files on a regular basis, updating the integrated database and artificial intelligence model. This process analyzes the new data and retrains the AI ​​model. The input is the newly collected literature and files, and the output is the updated database and retrained AI model.

[0430] The above processing steps enable prompt and appropriate responses to inquiries while taking user emotions into consideration.

[0431] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0432] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include those described above. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions shown by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0433] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0434] [Second Embodiment]

[0435] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0436] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0437] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0438] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0439] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0440] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0441] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0442] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0443] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0444] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0445] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0446] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0447] This invention is a system for streamlining the handling of inquiries both inside and outside a company. This system automatically collects documents and files, stores them in a database, trains an artificial intelligence model on them, and provides optimal answers to user inquiries. The specific program processing of this system is described below.

[0448] 1. Data Collection Phase

[0449] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in an integrated database.

[0450] 2. AI Model Training Phase

[0451] The server trains an artificial intelligence model based on literature and files stored in an integrated database. First, the data is preprocessed, and the text data is analyzed using natural language processing tools such as tokenizers and stemming. Then, a training dataset is created based on the preprocessed data, and the artificial intelligence model is trained using a machine learning algorithm (e.g., a Transformer-based model).

[0452] 3. Query Processing Phase

[0453] The user accesses the system using their device and enters an inquiry. For example: "Please tell me how to use the new office automation system." The device then sends this input to the server.

[0454] The server analyzes the received inquiry using natural language processing to understand its content. Then, based on the analyzed content, it uses a pre-trained artificial intelligence model to generate the optimal response. This generated response is then sent back to the user's device.

[0455] Specific example

[0456] The following is a concrete example illustrating how actual queries are processed.

[0457] User: "Please teach me how to use the new office automation system."

[0458] The device sends this query to the server.

[0459] The server analyzes the received text using a natural language processing tool (e.g., a tokenizer) and extracts "new OA system" and "how to use" as keywords.

[0460] The server uses a trained artificial intelligence model to extract information on "how to use the new office automation system" from relevant literature and generates the optimal response. This response is generated in a format such as, "The procedure for using the new office automation system is as follows..."

[0461] The device displays the generated answers to the user. The user can obtain specific and timely answers to their questions.

[0462] Maintenance Phase

[0463] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. Furthermore, it retrains its artificial intelligence model based on this new data to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy to ensure the security of the system.

[0464] In this way, the system can efficiently and securely handle inquiries from employees and partner companies. As a result, the workload of the help desk is reduced, and the quality of service to users is improved.

[0465] The following describes the processing flow.

[0466] Step 1:

[0467] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal and external repositories and online storage. This is done using scripts that automatically crawl the files at regular intervals.

[0468] Step 2:

[0469] The server analyzes the collected literature and files and generates metadata (e.g., file name, creation date, category). This analysis uses programs to analyze file formats and structures.

[0470] Step 3:

[0471] The server stores the contents of the documents and files, along with the generated metadata, in an integrated database. This storage process uses a storage management system to organize and store the files.

[0472] Step 4:

[0473] The server preprocesses the literature and files stored in the integrated database using natural language processing tools. Specifically, it uses a tokenizer to split the text and stemming to extract the core forms of words.

[0474] Step 5:

[0475] The server generates a training dataset based on preprocessed data and trains an artificial intelligence model using a machine learning algorithm (e.g., a Transformer-based model). During this process, the model's performance is evaluated by dividing it into training and test data.

[0476] Step 6:

[0477] The user enters their inquiry using their device. For example, they might type, "Please tell me how to use the new office automation system." The device then sends this input to the server.

[0478] Step 7:

[0479] The server analyzes the received query using natural language processing tools to extract key keywords and their relationships. This analysis utilizes tools for noun extraction and dependency analysis.

[0480] Step 8:

[0481] The server uses a trained artificial intelligence model based on the analysis results to generate the optimal response. For example, it might generate a specific response such as, "The procedure for using the new office automation system is as follows..."

[0482] Step 9:

[0483] The server sends the generated response back to the user's device. The device then displays this response to the user.

[0484] Step 10:

[0485] The server periodically recollects new literature and updated files, updating the database and artificial intelligence models. This includes the process of analyzing and retraining on new data.

[0486] Step 11:

[0487] The server encrypts all data communications and implements access control based on security policies. This ensures the security of the entire system.

[0488] (Example 1)

[0489] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0490] Responding to inquiries both inside and outside the company requires considerable effort and time, making efficient work difficult. Furthermore, there is a lack of systems to properly manage information provided by various departments and partner companies and to provide prompt and accurate responses. Keeping up-to-date with daily updates is also a major challenge. This invention aims to solve these problems and provide a system that streamlines inquiry handling.

[0491] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0492] In this invention, the server includes means for collecting and storing documents and electronic files in a database for generating answers to inquiries from information providers; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating the optimal answer using the trained artificial intelligence model; means for providing the generated answer to the user; means for automatically collecting new documents and electronic files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on security policies. This makes it possible to respond to inquiries efficiently and accurately based on the latest information.

[0493] "Information provider" refers to the department or partner company that provides documents or electronic files.

[0494] An "inquiry" refers to a question or request that a user enters into the system.

[0495] "Documents" refer to text data such as business manuals and user guides.

[0496] An "electronic file" is data stored in a digital format, including PDFs and document files.

[0497] A "database" refers to a system for managing and storing collected documents and electronic files.

[0498] An "artificial intelligence model" refers to a program trained using machine learning algorithms that generates responses based on input data.

[0499] "Natural language processing" refers to the technology of using computers to understand, analyze, and generate human language.

[0500] A "tokenizer" refers to a software tool that divides text into words or phrases.

[0501] "Stemming" refers to a natural language processing technique that extracts the root portion of a word.

[0502] A "security policy" refers to the rules and procedures for ensuring the security of data and communications.

[0503] "Users" refer to individuals or organizations that submit inquiries to the system and receive responses.

[0504] "Automated data collection" refers to a process in which the system automatically collects data without requiring manual operation.

[0505] "Communication" refers to the sending and receiving of data and information between systems.

[0506] "Answer" refers to the response that the system generates in response to a user's inquiry.

[0507] "Updating" refers to the process of adding new information to databases and artificial intelligence models to keep them up-to-date.

[0508] This invention is a system for streamlining the handling of inquiries both inside and outside a company. This system automatically collects documents and electronic files, stores them in a database, trains an artificial intelligence model based on the stored data, and provides optimal answers to user inquiries. Specific embodiments of this system are described below.

[0509] The server automatically collects documents and electronic files provided by various departments and partner companies within the company at regular intervals. These documents and electronic files include instruction manuals and guidelines. The server uses data collection scripts and APIs to access the internet and internal networks and store this data in an integrated database. Hardware such as dedicated servers and cloud storage is used in this data collection process.

[0510] The server trains an artificial intelligence model based on documents and electronic files stored in an integrated database. First, the server preprocesses the data and analyzes the text data using natural language processing tools such as tokenizers and stemming (e.g., NLTK, spaCy). Then, it creates a training dataset based on the preprocessed data and trains a Transformer-based model (e.g., BERT, GPT) using a machine learning framework (e.g., TensorFlow, PyTorch).

[0511] Users access the system using their devices and enter inquiries. For example, they might ask, "How do I use the new office automation system?" The device sends this input to the server. The server analyzes the received inquiry using natural language processing tools and understands its content. Based on the analysis, it then generates the optimal answer using a trained artificial intelligence model. This generated answer is then sent back to the user's device.

[0512] The following is a concrete example illustrating how actual queries are processed.

[0513] "Please teach me how to use the new office automation system."

[0514] "What are the latest guidelines regarding employee onboarding and offboarding procedures?"

[0515] "I want to know how to set up project management software."

[0516] The server automatically collects new documents and updated electronic files from repositories and online storage on a regular basis, updating the database. Furthermore, it retrains its artificial intelligence model based on this new data, enabling it to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy and ensures system security using the SSL / TLS protocol.

[0517] In this way, the system can efficiently and securely handle inquiries from employees and partner companies. As a result, the workload of the help desk is reduced, and the quality of service to users is improved.

[0518] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0519] Step 1: Prepare for data collection

[0520] The server configures connections to collect documents and electronic files from various departments and partner companies. This includes configuring API and FTP connections.

[0521] Input: Connection information for API endpoints and FTP servers of each department and partner company.

[0522] Output: A list of endpoints that have completed the necessary connection settings for data collection and are therefore available for connection.

[0523] Step 2: Perform data collection

[0524] The server automatically starts collecting data at specified time intervals. For example, it can trigger a collection task every Monday at 9:00 AM.

[0525] Input: Regular schedule information.

[0526] Output: A list of documents and electronic files downloaded from each endpoint.

[0527] Specific operation: The server makes HTTP requests or FTP connections to API endpoints or FTP servers of each department or partner company to download documents and files.

[0528] Step 3: Save the data to the database

[0529] The server saves downloaded documents and electronic files to a temporary directory, and then moves this data to the integrated database.

[0530] Input: Downloaded documents and electronic files.

[0531] Output: Entries for documents and electronic files stored in the integrated database.

[0532] Specific actions: Use scripts and ETL tools to perform data transformation and cleaning for database integration.

[0533] Step 4: Data Preprocessing

[0534] The server extracts data from the integrated database and performs preprocessing. Specifically, it tokenizes the text data using a tokenizer and performs stemming and stop word removal.

[0535] Input: Documents and electronic files extracted from an integrated database.

[0536] Output: Preprocessed text data.

[0537] Specific operation: Preprocessing is performed using natural language processing tools (e.g., NLTK, spaCy).

[0538] Step 5: Training the artificial intelligence model

[0539] The server creates a training dataset based on preprocessed data and trains the model using a machine learning framework (e.g., TensorFlow, PyTorch).

[0540] Input: Preprocessed text data.

[0541] Output: A trained artificial intelligence model.

[0542] Specific operation: Train the model using a Transformer-based model (e.g., BERT, GPT).

[0543] Step 6: Enter your inquiry

[0544] The user accesses the system using their own device and enters their inquiry. For example, they might enter, "Please tell me how to use the new office automation system."

[0545] Input: User inquiry details.

[0546] Output: The content of the query sent to the server.

[0547] Specific operation: The terminal sends the user's input query to the server as an HTTP request.

[0548] Step 7: Analyze the inquiry

[0549] The server analyzes the received query using natural language processing tools. Specifically, it uses a tokenizer to split the text and extract keywords.

[0550] Input: User inquiry details.

[0551] Output: Extracted keywords and query structure.

[0552] Specific operation: Performs text analysis using a tokenizer and stemming.

[0553] Step 8: Generating the answer

[0554] The server uses a trained artificial intelligence model to generate the best possible answer to the user's inquiry.

[0555] Input: Extracted keywords or query structure.

[0556] Output: The generated answer text.

[0557] Specific operation: Generate an answer using a pre-trained model and send that answer to the terminal as an HTTP response.

[0558] Step 9: Display the answer

[0559] The terminal displays the response sent back from the server to the user.

[0560] Input: Response text from the server.

[0561] Output: The answer displayed to the user.

[0562] Specific operation: Render the answer text on the screen and display it visually to the user.

[0563] Step 10: Database Update

[0564] The server automatically collects new documents and updated files from repositories and online storage on a regular basis and updates the database.

[0565] Input: A new document or electronic file.

[0566] Output: Updated integrated database.

[0567] Specific operation: The automated collection task is executed periodically, adding new data to the database.

[0568] Step 11: Retrain the model

[0569] The server retrains the artificial intelligence model using the new data.

[0570] Input: Updated text data.

[0571] Output: A retrained artificial intelligence model.

[0572] Specific operation: The machine learning algorithm is run again using new data to train the model with the latest information.

[0573] Step 12: Security Protection

[0574] The server uses the SSL / TLS protocol to encrypt all data communications.

[0575] Input: Communication data.

[0576] Output: Encrypted communication data.

[0577] Specific operation: Implements the SSL / TLS protocol to encrypt and protect communication data.

[0578] (Application Example 1)

[0579] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0580] Conventional systems make it difficult to improve operational efficiency and provide prompt responses to inquiries in logistics centers. In particular, the lack of a system that allows workers to quickly obtain necessary information makes delays and errors more likely. The present invention aims to solve these problems and provide a system that improves the operational efficiency of logistics centers.

[0581] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0582] In this invention, the server includes means for collecting documents and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating optimal answers using the trained artificial intelligence model; means for responding to inquiries from smart devices to provide logistics center work guidelines and procedures; means for providing the generated answers to users; means for automatically collecting new documents and files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on security policies. This enables improved work efficiency and quick and accurate response to inquiries in the logistics center.

[0583] "Documents and files" refer to documents and electronic files related to business operations both inside and outside the company, such as operational manuals, user guides, and support documents.

[0584] A "database" is a system that centrally manages collected literature and files, making them searchable and accessible.

[0585] An "artificial intelligence model" is an AI system trained using machine learning algorithms that generates the optimal answer to a specific task.

[0586] "Natural language processing" is a language processing technique that analyzes text-based inquiries from users and understands their meaning.

[0587] A "logistics center" is a warehouse-type facility that handles operations such as receiving, storing, and shipping goods and materials.

[0588] "Smart devices" are advanced electronic devices with communication capabilities, such as smartphones, smart glasses, and head-mounted displays.

[0589] A "security policy" is a set of rules and procedures aimed at protecting data and communications.

[0590] "Work guidelines" are guidelines that describe the procedures and methods for various tasks in a logistics center.

[0591] An "inquiry" is a question or request that a user enters into the system to seek information.

[0592] "Answer" refers to the information provided in response to a user's inquiry.

[0593] This invention is a system aimed at improving the efficiency of operations and speeding up inquiry response in logistics centers. This system automatically collects documents and files, stores them in a database, trains an artificial intelligence model on them, and provides optimal answers to user inquiries. The specific program processing of this system is described below.

[0594] 1. Data Collection Phase

[0595] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in a database.

[0596] 2. AI Model Training Phase

[0597] The server trains an artificial intelligence model based on literature and files stored in the database. First, the data is preprocessed, and the text data is analyzed using natural language processing tools such as tokenizers and stemming. This process uses libraries such as spaCy and Huggingface Transformers. After that, a training dataset is created based on the preprocessed data, and the artificial intelligence model is trained using machine learning algorithms such as the BERT model.

[0598] 3. Query Processing Phase

[0599] The user accesses the system using their smart device (e.g., smartphone or smart glasses) and enters an inquiry. For example: "Please tell me how to package the new product." The device sends this input to the server. The server analyzes the received inquiry using natural language processing tools and understands its content. Based on the analysis, it generates the optimal answer using a trained artificial intelligence model. This generated answer is then sent back to the user's device.

[0600] Specific example

[0601] Consider a scenario where a user asks via smartphone, "How should I pack the product?" This inquiry is analyzed by the server, and the optimal packing procedure is displayed based on relevant manuals.

[0602] 4. Maintenance Phase

[0603] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. Furthermore, it retrains its artificial intelligence model based on this new data to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy to ensure the security of the system.

[0604] Example of a prompt

[0605] As a concrete example, the following prompt statements are possible:

[0606] User: "Please tell me how to pack the product."

[0607] Prompt: "Analyze the information on product packaging methods in the logistics center's work procedures and manuals, and indicate the optimal procedure."

[0608] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0609] Step 1: Data Collection

[0610] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in a database.

[0611] Input: Literature and files

[0612] Output: Data stored in the database

[0613] Specific operation: Retrieve data from file folders or online storage and upload it to the database system.

[0614] Step 2: Data Preprocessing

[0615] The server preprocesses the stored literature and files. It analyzes the text data using natural language processing tools such as tokenizers and stemming tools (e.g., spaCy).

[0616] Input: Literature and files stored in the database

[0617] Output: Preprocessed text data

[0618] Specific operation: Tokenize the text, perform stemming, and save the analysis results as structured data.

[0619] Step 3: AI Model Training

[0620] The server trains an artificial intelligence model using machine learning algorithms such as the BERT model, based on preprocessed data. This data is processed using the Huggingface Transformers library.

[0621] Input: Preprocessed text data

[0622] Output: Trained artificial intelligence model

[0623] Specific operation: Preprocessed text data is created as a training dataset, and training is performed using the BERT model.

[0624] Step 4: Enter your inquiry

[0625] Users access the system using their smart devices and enter their inquiries. For example, they might submit an inquiry such as, "Please tell me how to pack the product."

[0626] Input: User inquiry text

[0627] Output: Sending a query to the server

[0628] Specific operation: The user enters text through the interface of a smart device and sends that inquiry to the server.

[0629] Step 5: Inquiry Analysis

[0630] The server analyzes the received query using a natural language processing tool (e.g., spaCy) to understand its content.

[0631] Input: User inquiry text

[0632] Output: Analyzed keywords and queries

[0633] Specific operation: The received inquiry text is tokenized, keywords are extracted, and queries are generated.

[0634] Step 6: Generate Answer

[0635] The server generates the optimal response using a pre-trained artificial intelligence model based on the analyzed data. This generated response includes information extracted from relevant literature.

[0636] Input: Analyzed keywords and queries

[0637] Output: The best answer for the user

[0638] Specific operation: Using a pre-trained artificial intelligence model, extract information based on keywords and queries, and generate answers in natural language.

[0639] Step 7: Provide your answer

[0640] The server sends the generated answer back to the user's device and displays it to the user. The user can obtain specific and timely answers to their questions.

[0641] Input: Response text from the server

[0642] Output: The answer displayed on the user's smart device.

[0643] Specific operation: The response text sent from the server is displayed on the user's device.

[0644] Step 8: Maintenance

[0645] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. It also retrains artificial intelligence models based on this new data. Furthermore, all data communications are encrypted according to security policies to ensure system security.

[0646] Input: New documents and files, security policy

[0647] Output: Updated database and retrained artificial intelligence model

[0648] Specific actions: Regularly collect and store data, retrain artificial intelligence models, and implement security measures.

[0649] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0650] This invention provides a system that streamlines the handling of inquiries from employees and partner companies, and recognizes user emotions to provide appropriate responses. This system automatically collects documents and files, stores them in a database, and trains an artificial intelligence (AI) model. Furthermore, by combining this with an emotion engine, it recognizes user emotions and provides appropriate responses tailored to those emotions.

[0651] Data collection phase

[0652] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal repositories and online storage. The collected documentation and files are stored in an integrated database at regular intervals.

[0653] AI model training phase

[0654] The server preprocesses literature and files stored in the integrated database using natural language processing (NLP) tools. Specifically, it uses a tokenizer to segment the text and stemming to extract the core forms of words. Then, it generates a training dataset based on the preprocessed data and trains an AI model using a machine learning algorithm (e.g., a transformer model). This model is used to generate the best possible answers to user inquiries.

[0655] Query processing phase

[0656] The user enters their inquiry into the system from their terminal. For example, they might enter, "Please tell me how to use the new office automation system." This input is then sent from the terminal to the server.

[0657] The server analyzes the received inquiry using NLP tools to extract key keywords and their relationships. It then uses a pre-trained AI model to generate the optimal response.

[0658] Emotion recognition phase

[0659] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis. The emotion engine analyzes the user's input text and adjusts the tone and content of the AI ​​model's response accordingly. For example, if the user indicates that they are "distressed," the response will be adjusted to a gentle and polite tone.

[0660] Answer generation phase

[0661] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The device then displays this response to the user. This allows the user to receive specific and timely responses that are tailored to their emotions.

[0662] Specific example

[0663] The following is an example of how to process a specific query using the emotion engine.

[0664] User: "I'm having trouble figuring out how to use the new office automation system. Please help me."

[0665] The device sends this query to the server.

[0666] The server analyzes the received text using an NLP tool and extracts keywords such as "OA system," "how to use," and "having trouble," along with the associated emotions.

[0667] The server activates an emotion engine based on the extracted emotion "distressed," and uses an AI model to generate a gentle and polite response, such as "Don't worry, here's how to use the new OA system..."

[0668] The device displays the generated response to the user. By receiving a response that is tailored to their own emotions, the user can resolve the problem with confidence.

[0669] Maintenance Phase

[0670] The server periodically recollects new literature and updated files, updating the database and AI models. This includes the process of analyzing new data and retraining the AI ​​models. Furthermore, the server ensures system security by encrypting all data communications and implementing access control based on security policies.

[0671] In this way, this system, equipped with an emotion engine, enables efficient and emotionally sensitive inquiry handling, contributing to improved user experience and reduced workload for the help desk.

[0672] The following describes the processing flow.

[0673] Step 1:

[0674] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal and external repositories and online storage. This collection is performed at specific time intervals or triggered by specific events.

[0675] Step 2:

[0676] The server analyzes the collected literature and files and generates file metadata (e.g., file name, creation date, category). This metadata generation uses tools for file format detection and text extraction.

[0677] Step 3:

[0678] The server stores the contents of the documents and files, along with the generated metadata, in an integrated database. This storage process uses a database management system to normalize the data and make it searchable efficiently.

[0679] Step 4:

[0680] The server preprocesses the literature and files stored in the integrated database using natural language processing tools. Specifically, this preprocessing involves splitting the text using a tokenizer and converting words into their stem forms using stemming.

[0681] Step 5:

[0682] The server generates a training dataset based on preprocessed data and trains an artificial intelligence model using machine learning algorithms. This training includes cross-validation between the training dataset and the test dataset.

[0683] Step 6:

[0684] A user accesses the system using a terminal and enters an inquiry. For example, they might enter, "I'm having trouble figuring out how to use the new office automation system; please help me." The terminal then sends this input to the server.

[0685] Step 7:

[0686] The server analyzes the received inquiry using natural language processing tools and extracts key keywords and their relationships. For example, keywords such as "OA system," "how to use," and "I'm having trouble" might be extracted.

[0687] Step 8:

[0688] The server activates the emotion engine based on the extracted keywords. The emotion engine analyzes the user's text to determine emotions such as "distressed," and adjusts the tone and content of the response based on those emotions.

[0689] Step 9:

[0690] Based on the results of the emotion engine, the server uses a trained artificial intelligence model to generate the optimal response. For example, it might generate a response in a gentle tone such as, "Don't worry, the instructions for using the new OA system are as follows..."

[0691] Step 10:

[0692] The server sends the generated response back to the user's device. The device displays this response to the user. The user can receive specific and timely responses that are tailored to their emotions.

[0693] Step 11:

[0694] The server periodically recollects new literature and updated files, updating the database and artificial intelligence models. This includes the process of analyzing new data and retraining the AI ​​models.

[0695] Step 12:

[0696] The server encrypts all data communications and implements access control based on security policies. This ensures the security of the entire system.

[0697] (Example 2)

[0698] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0699] Conventional inquiry handling systems often disregard user emotions and provide mechanical answers, making it difficult to increase user satisfaction. Furthermore, adequate support is needed for model updates to provide optimal answers tailored to the content of inquiries, as well as for data security. This invention aims to address these problems and achieve efficient and emotionally sensitive inquiry handling.

[0700] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0701] In this invention, the server includes means for collecting literature and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating optimal answers using the trained artificial intelligence model; means for providing the generated answers to the user; means for automatically collecting new literature and files periodically and updating the database and artificial intelligence model; means for protecting communications and data based on security policies; and means for performing sentiment analysis based on the content of user inquiries and adjusting the answers with appropriate tone and content according to the analysis results. This enables efficient and emotionally sensitive inquiry handling.

[0702] "Documents and files" refers to a collection of data containing information necessary for responding to inquiries, such as operational manuals, user guides, and support documents.

[0703] A "database" is a system for efficiently storing, managing, and searching for documents and files.

[0704] An "artificial intelligence model" is a model trained to generate the best possible answer to a query using machine learning algorithms.

[0705] "Natural language processing" is a technique that uses tools such as tokenizers and stemming to analyze user inquiries.

[0706] "Sentiment analysis" is the process of identifying emotions from the content of a user's inquiry and adjusting the response to match those emotions in terms of tone and content.

[0707] "Means of protecting communications and data" refer to technologies that encrypt data communications using SSL / TLS, etc., and restrict access to system resources using access control lists (ACLs).

[0708] "Means for automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models" refers to the process of continuously collecting new information, updating databases based on that information, and retraining AI models to keep them up-to-date.

[0709] "Methods for analyzing inquiries and generating answers" refers to a function that analyzes user inquiries using NLP tools and generates the optimal answer using a pre-trained artificial intelligence model.

[0710] This invention is a system that streamlines the handling of inquiries from employees and partner companies, and recognizes user emotions to provide appropriate responses. This system automatically collects documents and files, stores them in a database, and trains an artificial intelligence model. Furthermore, by combining this with an emotion engine, it recognizes user emotions and provides appropriate responses tailored to those emotions.

[0711] Data collection phase

[0712] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal repositories and online storage. The collected documentation and files are periodically saved to an integrated database. This process is performed using a crawler and a database management system (e.g., MySQL).

[0713] AI model training phase

[0714] The server preprocesses literature and files stored in the integrated database using natural language processing (NLP) tools (e.g., NLTK, Spacy). It tokenizes the text and extracts word stems using stemming. Then, it generates a training dataset based on the preprocessed data and trains an AI model using machine learning algorithms (e.g., BERT, GPT-3). This AI model is used to generate the best possible answers to user inquiries.

[0715] Query processing phase

[0716] The user enters their inquiry into the system from their terminal. For example, they might enter, "Please tell me how to use the new office automation system." This input is then sent from the terminal to the server.

[0717] The server analyzes the received inquiry using NLP tools to extract key keywords and their relationships. It then uses a pre-trained AI model to generate the optimal response.

[0718] Emotion recognition phase

[0719] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis. The emotion engine analyzes the user's input text and adjusts the tone and content of the AI ​​model's response accordingly. For example, if the user indicates that they are "distressed," the response will be adjusted to a gentle and polite tone.

[0720] Answer generation phase

[0721] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The device then displays this response to the user. This allows the user to receive specific and timely responses that are tailored to their emotions.

[0722] Maintenance Phase

[0723] The server periodically recollects new literature and updated files, updating the database and AI models. This process includes analyzing new data and retraining the AI ​​models. Furthermore, the server ensures system security by encrypting all data communications and implementing access controls based on security policies.

[0724] Specific example

[0725] The following is an example of how to process a specific query using the emotion engine.

[0726] User: "I'm having trouble figuring out how to use the new office automation system. Please help me."

[0727] The device sends this query to the server.

[0728] The server analyzes the received text using an NLP tool and extracts keywords such as "OA system," "how to use," and "having trouble," along with the associated emotions.

[0729] The server activates an emotion engine based on the extracted emotion "distressed," and uses an AI model to generate a gentle and polite response, such as "Don't worry, here's how to use the new OA system..."

[0730] The device displays the generated response to the user. By receiving a response that is tailored to their own emotions, the user can resolve the problem with confidence.

[0731] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0732] Step 1:

[0733] The server executes data collection scripts to download operational manuals, user guides, and support documents from internal repositories and online storage.

[0734] Input: URLs of internal repositories and online storage, API keys, etc.

[0735] Output: Downloaded literature and files

[0736] Specific operation: The server uses a crawler to download literature and files from the specified URL via FTP or API.

[0737] Step 2:

[0738] The server saves downloaded documents and files to an integrated database.

[0739] Input: Downloaded literature and files

[0740] Output: Data stored in the integrated database

[0741] Specific operation: The server inserts the literature and files into the database management system (e.g., MySQL).

[0742] Step 3:

[0743] The server preprocesses the literature and files stored in the integrated database using NLP tools.

[0744] Input: Literature and files stored in the integrated database

[0745] Output: Preprocessed text data

[0746] Specific operation: The server uses NLP tools (NLTK, Spacy) to split the text with a tokenizer and perform stemming.

[0747] Step 4:

[0748] The server generates a training dataset based on pre-processed data.

[0749] Input: Preprocessed text data

[0750] Output: Training dataset

[0751] Specific operation: The server converts the preprocessed data into a CSV file, and then uses TensorFlow or PyTorch to generate a training dataset based on that file.

[0752] Step 5:

[0753] The server uses machine learning algorithms to train the AI ​​model.

[0754] Input: Training dataset

[0755] Output: Trained AI model file

[0756] Specific operation: The server trains an AI model using a Transformer model (e.g., the Hugging Face Transformers library) and saves it to "transformer_model.bin".

[0757] Step 6:

[0758] The user enters an inquiry from their device, and the device sends that information to the server.

[0759] Input: Inquiry details (Example: "Please tell me how to use the new office automation system.")

[0760] Output: Query data sent to the server

[0761] Specific operation: The user enters a query into an input form on their device, and the device uses a REST API to send this data to the server.

[0762] Step 7:

[0763] The server analyzes the received query content using an NLP tool.

[0764] Input: Inquiry details

[0765] Output: Key keywords and their relationships

[0766] Specific operation: The server uses Spacy to extract key keywords such as "OA system" and "how to use".

[0767] Step 8:

[0768] The server uses a trained AI model to generate the optimal answer.

[0769] Input: Analyzed query content

[0770] Output: Generated answer

[0771] Specific operation: The server loads "transformer_model.bin" and generates a response based on the query.

[0772] Step 9:

[0773] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis.

[0774] Input: Inquiry details

[0775] Output: Emotion analysis results

[0776] Specific operation: The server calls an emotion engine (e.g., Google Cloud Natural Language API) to analyze emotions and identify the emotion "distressed".

[0777] Step 10:

[0778] The server adjusts the tone and content of the AI ​​model's responses based on the sentiment analysis results.

[0779] Input: Sentiment analysis results, generated responses

[0780] Output: Adjusted answer

[0781] Specific action: Based on the emotion "distressed," the server changes the tone of its response to a gentler and more polite one.

[0782] Step 11:

[0783] The server sends the adjusted response back to the terminal, which then displays it to the user.

[0784] Input: Adjusted answer

[0785] Output: Answer displayed to the user

[0786] Specific operation: The server sends the prepared response back to the terminal in JSON format as a REST API response, and the terminal displays it on the screen.

[0787] Step 12:

[0788] The server periodically recollects new literature and updated files, updating the database and AI models.

[0789] Input: New literature or updated files

[0790] Output: Updated database and AI model

[0791] Specific operation: The server uses scheduled jobs to collect new data, insert it into the database, and retrain the AI ​​model based on the new data.

[0792] Step 13:

[0793] The server encrypts all data communications and implements access control.

[0794] Input: Communication data, access request

[0795] Output: Encrypted data, secure access

[0796] Specific operation: The server encrypts data communication using SSL / TLS and ensures security using Access Control Lists (ACLs).

[0797] (Application Example 2)

[0798] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0799] Previously, responses to inquiries from employees and partner companies were inefficient, and responses that considered user emotions were not provided. As a result, improvements in user experience and the efficiency of inquiry handling were not fully achieved. Furthermore, in security services, there is a need for responses that give users peace of mind in times of anxiety or emergency.

[0800] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting literature and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for recognizing the user's emotions, analyzing them using natural language processing, and generating an optimal response using the trained artificial intelligence model; means for providing the generated response to the user; means for adjusting the tone of the response based on the user's emotions using an emotion recognition engine; means for automatically collecting new literature and files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on a security policy. This makes it possible to provide prompt and appropriate inquiry response and security services that take into consideration the user's emotions.

[0801] "Documents and files" refer to materials containing information necessary to respond to user inquiries, including operational manuals, user guides, and security-related guidelines.

[0802] A "database" is an information system that stores collected literature and files, and allows for the efficient management and retrieval of information necessary to respond to inquiries.

[0803] An "artificial intelligence model" is an algorithm that learns from collected literature and files and is used to automatically generate the best possible answers to user inquiries.

[0804] "Natural language processing" is a technology used to analyze user inquiries and extract key keywords and their relationships.

[0805] An "emotion recognition engine" is a system that analyzes emotions from the user's input text and adjusts the tone and content of the response accordingly.

[0806] A "security policy" is a set of rules and guidelines designed to ensure data communication and information protection within a system.

[0807] "Response tone" refers to the tone and attitude of the response, which are adjusted with consideration for the user's feelings.

[0808] "New documents and files" refer to materials that are added regularly, including the latest operational manuals, user guides, and security-related guidelines.

[0809] The system for carrying out this invention is a response system that integrates multiple means to provide prompt and appropriate answers to inquiries while taking into consideration the user's feelings. This system mainly uses the following hardware and software.

[0810] First, the server automatically collects documents and files from internal repositories and online storage and stores them in an integrated database. These documents and files include operational manuals, user guides, and security-related guidelines. The server uses the requests library to retrieve these documents.

[0811] Next, the server trains an artificial intelligence model based on the stored literature and files. This process utilizes natural language processing (NLP) tools. Specifically, it uses a tokenizer to segment the text and stemming to extract the core forms of words. Then, it generates a training dataset based on the preprocessed data and trains the AI ​​model using a machine learning algorithm (e.g., a transformer model). The pipeline function of the transformers library is used in this training phase.

[0812] Users enter inquiries into the system from their devices. For example, if a user makes an inquiry such as, "I'm scared because I heard sounds like someone broke into my house tonight," this input is sent from the device to the server.

[0813] The server analyzes the received inquiry using natural language processing tools to extract key keywords and their relationships. Simultaneously, it analyzes the user's emotions using an emotion recognition engine and adjusts the tone and content of the AI ​​model's response according to the emotions the user expresses.

[0814] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The user's device then displays this response to the user. In this way, the user can obtain a specific and timely response that is tailored to their emotions.

[0815] For example, if a user enters a question saying, "I'm scared because I heard sounds like someone broke into my house tonight," it will be processed as follows:

[0816] 1. Receive a user inquiry.

[0817] 2. Natural language processing tools were used to extract the keywords and emotions "intrusion" and "scary".

[0818] 3. The emotion analysis engine determined it to be "fear".

[0819] 4. Based on appropriate advice regarding "intrusion," adjust the tone of your response according to your emotions.

[0820] 5. Generate and provide the user with a response such as, "Don't worry. Contact the police immediately and evacuate to a safe place."

[0821] Examples of prompts to input into a generative AI model are as follows:

[0822] Q: I'm scared because I heard sounds tonight that sound like someone broke into my house.

[0823] A: Don't worry. Contact the police immediately and evacuate to a safe place.

[0824] In this way, this system, equipped with an emotion engine, provides a concrete means to achieve efficient inquiry handling that takes emotions into consideration.

[0825] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0826] Step 1:

[0827] The server automatically collects documents and files from internal repositories and online storage. It retrieves documents such as operational manuals, user guides, and security-related guidelines using the requests library and stores their contents in an integrated database. Input consists of documents and files in internal repositories and online storage, while output consists of documents and files stored in the integrated database.

[0828] Step 2:

[0829] The server trains an artificial intelligence model based on literature and files stored in an integrated database. Natural language processing (NLP) tools split the text (tokenizer) and extract the core form of words using stemming. Then, a training dataset is generated using the preprocessed data, and a transformer model is trained using the pipeline function of the transformers library. The input is literature and files stored in the integrated database, and the output is the trained artificial intelligence model.

[0830] Step 3:

[0831] The user enters the inquiry using their device. For example, the text might be, "I'm scared because I heard sounds like someone broke into my house tonight." The input is the user's inquiry text, and the output is the inquiry sent from the device to the server.

[0832] Step 4:

[0833] The server analyzes user inquiries using natural language processing tools to extract key keywords and their relationships. This analysis utilizes tokenizers and stemming, and an emotion recognition engine is used to analyze the user's emotions. The input is the user's inquiry text, and the output is the extracted keywords and emotion data.

[0834] Step 5:

[0835] The server generates the optimal response using a trained artificial intelligence model based on extracted keywords and user sentiment. This process involves generating prompts, inputting them into the AI ​​model, and obtaining the response. The input consists of extracted keywords and sentiment data, while the output is the generated response text.

[0836] Step 6:

[0837] The server sends a response, refined by the emotion recognition engine, to the user's device, which then displays this response to the user. The input is the generated response text, and the output is the response displayed on the user's device.

[0838] Step 7:

[0839] The server automatically collects new literature and files on a regular basis, updating the integrated database and artificial intelligence model. This process analyzes the new data and retrains the AI ​​model. The input is the newly collected literature and files, and the output is the updated database and retrained AI model.

[0840] The above processing steps enable prompt and appropriate responses to inquiries while taking user emotions into consideration.

[0841] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0842] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0843] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0844] [Third Embodiment]

[0845] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0846] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0847] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0848] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0849] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0850] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0851] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0852] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0853] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0854] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0855] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0856] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0857] This invention is a system for streamlining the handling of inquiries both inside and outside a company. This system automatically collects documents and files, stores them in a database, trains an artificial intelligence model on them, and provides optimal answers to user inquiries. The specific program processing of this system is described below.

[0858] 1. Data Collection Phase

[0859] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in an integrated database.

[0860] 2. AI Model Training Phase

[0861] The server trains an artificial intelligence model based on literature and files stored in an integrated database. First, the data is preprocessed, and the text data is analyzed using natural language processing tools such as tokenizers and stemming. Then, a training dataset is created based on the preprocessed data, and the artificial intelligence model is trained using a machine learning algorithm (e.g., a Transformer-based model).

[0862] 3. Query Processing Phase

[0863] The user accesses the system using their device and enters an inquiry. For example: "Please tell me how to use the new office automation system." The device then sends this input to the server.

[0864] The server analyzes the received inquiry using natural language processing to understand its content. Then, based on the analyzed content, it uses a pre-trained artificial intelligence model to generate the optimal response. This generated response is then sent back to the user's device.

[0865] Specific example

[0866] The following is a concrete example illustrating how actual queries are processed.

[0867] User: "Please tell me how to use the new office automation system."

[0868] The device sends this query to the server.

[0869] The server analyzes the received text using a natural language processing tool (e.g., a tokenizer) and extracts "new OA system" and "how to use" as keywords.

[0870] The server uses a trained artificial intelligence model to extract information on "how to use the new office automation system" from relevant literature and generates the optimal response. This response is generated in a format such as, "The procedure for using the new office automation system is as follows..."

[0871] The device displays the generated answers to the user. The user can obtain specific and timely answers to their questions.

[0872] Maintenance Phase

[0873] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. Furthermore, it retrains its artificial intelligence model based on this new data to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy to ensure the security of the system.

[0874] In this way, the system can efficiently and securely handle inquiries from employees and partner companies. As a result, the workload of the help desk is reduced, and the quality of service to users is improved.

[0875] The following describes the processing flow.

[0876] Step 1:

[0877] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal and external repositories and online storage. This is done using scripts that automatically crawl the files at regular intervals.

[0878] Step 2:

[0879] The server analyzes the collected literature and files and generates metadata (e.g., file name, creation date, category). This analysis uses programs to analyze file formats and structures.

[0880] Step 3:

[0881] The server stores the contents of the documents and files, along with the generated metadata, in an integrated database. This storage process uses a storage management system to organize and store the files.

[0882] Step 4:

[0883] The server preprocesses the literature and files stored in the integrated database using natural language processing tools. Specifically, it uses a tokenizer to split the text and stemming to extract the core forms of words.

[0884] Step 5:

[0885] The server generates a training dataset based on preprocessed data and trains an artificial intelligence model using a machine learning algorithm (e.g., a Transformer-based model). During this process, the model's performance is evaluated by dividing it into training and test data.

[0886] Step 6:

[0887] The user enters their inquiry using their device. For example, they might type, "Please tell me how to use the new office automation system." The device then sends this input to the server.

[0888] Step 7:

[0889] The server analyzes the received query using natural language processing tools to extract key keywords and their relationships. This analysis utilizes tools for noun extraction and dependency analysis.

[0890] Step 8:

[0891] The server uses a trained artificial intelligence model based on the analysis results to generate the optimal response. For example, it might generate a specific response such as, "The procedure for using the new office automation system is as follows..."

[0892] Step 9:

[0893] The server sends the generated response back to the user's device. The device then displays this response to the user.

[0894] Step 10:

[0895] The server periodically recollects new literature and updated files, updating the database and artificial intelligence models. This includes the process of analyzing and retraining on new data.

[0896] Step 11:

[0897] The server encrypts all data communications and implements access control based on security policies. This ensures the security of the entire system.

[0898] (Example 1)

[0899] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0900] Responding to inquiries both inside and outside the company requires considerable effort and time, making efficient work difficult. Furthermore, there is a lack of systems to properly manage information provided by various departments and partner companies and to provide prompt and accurate responses. Keeping up-to-date with daily updates is also a major challenge. This invention aims to solve these problems and provide a system that streamlines inquiry handling.

[0901] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0902] In this invention, the server includes means for collecting and storing documents and electronic files in a database for generating answers to inquiries from information providers; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating the optimal answer using the trained artificial intelligence model; means for providing the generated answer to the user; means for automatically collecting new documents and electronic files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on security policies. This makes it possible to respond to inquiries efficiently and accurately based on the latest information.

[0903] "Information provider" refers to the department or partner company that provides documents or electronic files.

[0904] An "inquiry" refers to a question or request that a user enters into the system.

[0905] "Documents" refer to text data such as business manuals and user guides.

[0906] An "electronic file" is data stored in a digital format, including PDFs and document files.

[0907] A "database" refers to a system for managing and storing collected documents and electronic files.

[0908] An "artificial intelligence model" refers to a program trained using machine learning algorithms that generates responses based on input data.

[0909] "Natural language processing" refers to the technology of using computers to understand, analyze, and generate human language.

[0910] A "tokenizer" refers to a software tool that divides text into words or phrases.

[0911] "Stemming" refers to a natural language processing technique that extracts the root portion of a word.

[0912] A "security policy" refers to the rules and procedures for ensuring the security of data and communications.

[0913] "Users" refer to individuals or organizations that submit inquiries to the system and receive responses.

[0914] "Automated data collection" refers to a process in which the system automatically collects data without requiring manual operation.

[0915] "Communication" refers to the sending and receiving of data and information between systems.

[0916] "Answer" refers to the response that the system generates in response to a user's inquiry.

[0917] "Updating" refers to the process of adding new information to databases and artificial intelligence models to keep them up-to-date.

[0918] This invention is a system for streamlining the handling of inquiries both inside and outside a company. This system automatically collects documents and electronic files, stores them in a database, trains an artificial intelligence model based on the stored data, and provides optimal answers to user inquiries. Specific embodiments of this system are described below.

[0919] The server automatically collects documents and electronic files provided by various departments and partner companies within the company at regular intervals. These documents and electronic files include instruction manuals and guidelines. The server uses data collection scripts and APIs to access the internet and internal networks and store this data in an integrated database. Hardware such as dedicated servers and cloud storage is used in this data collection process.

[0920] The server trains an artificial intelligence model based on documents and electronic files stored in an integrated database. First, the server preprocesses the data and analyzes the text data using natural language processing tools such as tokenizers and stemming (e.g., NLTK, spaCy). Then, it creates a training dataset based on the preprocessed data and trains a Transformer-based model (e.g., BERT, GPT) using a machine learning framework (e.g., TensorFlow, PyTorch).

[0921] Users access the system using their devices and enter inquiries. For example, they might ask, "How do I use the new office automation system?" The device sends this input to the server. The server analyzes the received inquiry using natural language processing tools and understands its content. Based on the analysis, it then generates the optimal answer using a trained artificial intelligence model. This generated answer is then sent back to the user's device.

[0922] The following is a concrete example illustrating how actual queries are processed.

[0923] "Please teach me how to use the new office automation system."

[0924] "What are the latest guidelines regarding employee onboarding and offboarding procedures?"

[0925] "I want to know how to set up project management software."

[0926] The server automatically collects new documents and updated electronic files from repositories and online storage periodically to update its database. Furthermore, it retrains its artificial intelligence model based on this new data, enabling it to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy and ensures system security using the SSL / TLS protocol.

[0927] In this way, the system can efficiently and securely handle inquiries from employees and partner companies. As a result, the workload of the help desk is reduced, and the quality of service to users is improved.

[0928] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0929] Step 1: Preparing for data collection

[0930] The server configures connections to collect documents and electronic files from various departments and partner companies. This includes configuring API and FTP connections.

[0931] Input: Connection information for API endpoints and FTP servers of each department and partner company.

[0932] Output: A list of endpoints that have completed the necessary connection settings for data collection and are therefore connectable.

[0933] Step 2: Perform data collection

[0934] The server automatically starts collecting data at specified time intervals. For example, it can trigger a collection task every Monday at 9:00 AM.

[0935] Input: Regular schedule information.

[0936] Output: A list of documents and electronic files downloaded from each endpoint.

[0937] Specific operation: The server makes HTTP requests or FTP connections to API endpoints or FTP servers of each department or partner company to download documents and files.

[0938] Step 3: Save the data to the database

[0939] The server saves downloaded documents and electronic files to a temporary directory, and then moves this data to the integrated database.

[0940] Input: Downloaded documents and electronic files.

[0941] Output: Entries for documents and electronic files stored in the integrated database.

[0942] Specific actions: Use scripts and ETL tools to perform data transformation and cleaning for database integration.

[0943] Step 4: Data Preprocessing

[0944] The server extracts data from the integrated database and performs preprocessing. Specifically, it tokenizes the text data using a tokenizer and performs stemming and stop word removal.

[0945] Input: Documents and electronic files extracted from an integrated database.

[0946] Output: Preprocessed text data.

[0947] Specific operation: Preprocessing is performed using natural language processing tools (e.g., NLTK, spaCy).

[0948] Step 5: Training the artificial intelligence model

[0949] The server creates a training dataset based on preprocessed data and trains the model using a machine learning framework (e.g., TensorFlow, PyTorch).

[0950] Input: Preprocessed text data.

[0951] Output: A trained artificial intelligence model.

[0952] Specific operation: Train the model using a Transformer-based model (e.g., BERT, GPT).

[0953] Step 6: Enter your inquiry

[0954] The user accesses the system using their own device and enters their inquiry. For example, they might enter, "Please tell me how to use the new office automation system."

[0955] Input: User inquiry details.

[0956] Output: The content of the query sent to the server.

[0957] Specific operation: The terminal sends the user's input query to the server as an HTTP request.

[0958] Step 7: Analyze the inquiry

[0959] The server analyzes the received query using natural language processing tools. Specifically, it uses a tokenizer to split the text and extract keywords.

[0960] Input: User inquiry details.

[0961] Output: Extracted keywords and query structure.

[0962] Specific operation: Performs text analysis using a tokenizer and stemming.

[0963] Step 8: Generating the answer

[0964] The server uses a trained artificial intelligence model to generate the best possible answer to the user's inquiry.

[0965] Input: Extracted keywords or query structure.

[0966] Output: The generated answer text.

[0967] Specific operation: Generate an answer using a pre-trained model and send that answer to the terminal as an HTTP response.

[0968] Step 9: Display the answer

[0969] The terminal displays the response sent back from the server to the user.

[0970] Input: Response text from the server.

[0971] Output: The answer displayed to the user.

[0972] Specific operation: Render the answer text on the screen and display it visually to the user.

[0973] Step 10: Database Update

[0974] The server automatically collects new documents and updated files from repositories and online storage on a regular basis and updates the database.

[0975] Input: A new document or electronic file.

[0976] Output: Updated integrated database.

[0977] Specific operation: The automated collection task is executed periodically, adding new data to the database.

[0978] Step 11: Retrain the model

[0979] The server retrains the artificial intelligence model using the new data.

[0980] Input: Updated text data.

[0981] Output: A retrained artificial intelligence model.

[0982] Specific operation: The machine learning algorithm is run again using new data to train the model with the latest information.

[0983] Step 12: Security Protection

[0984] The server uses the SSL / TLS protocol to encrypt all data communications.

[0985] Input: Communication data.

[0986] Output: Encrypted communication data.

[0987] Specific operation: Implements the SSL / TLS protocol to encrypt and protect communication data.

[0988] (Application Example 1)

[0989] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0990] Conventional systems make it difficult to improve operational efficiency and provide prompt responses to inquiries in logistics centers. In particular, the lack of a system that allows workers to quickly obtain necessary information makes delays and errors more likely. The present invention aims to solve these problems and provide a system that improves the operational efficiency of logistics centers.

[0991] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0992] In this invention, the server includes means for collecting documents and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating optimal answers using the trained artificial intelligence model; means for responding to inquiries from smart devices to provide logistics center work guidelines and procedures; means for providing the generated answers to users; means for automatically collecting new documents and files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on security policies. This enables improved work efficiency and quick and accurate response to inquiries in the logistics center.

[0993] "Documents and files" refer to documents and electronic files related to business operations both inside and outside the company, such as operational manuals, user guides, and support documents.

[0994] A "database" is a system that centrally manages collected literature and files, making them searchable and accessible.

[0995] An "artificial intelligence model" is an AI system trained using machine learning algorithms that generates the optimal answer to a specific task.

[0996] "Natural language processing" is a language processing technique that analyzes text-based inquiries from users and understands their meaning.

[0997] A "logistics center" is a warehouse-type facility that handles operations such as receiving, storing, and shipping goods and materials.

[0998] "Smart devices" are advanced electronic devices with communication capabilities, such as smartphones, smart glasses, and head-mounted displays.

[0999] A "security policy" is a set of rules and procedures aimed at protecting data and communications.

[1000] "Work guidelines" are guidelines that describe the procedures and methods for various tasks in a logistics center.

[1001] An "inquiry" is a question or request that a user enters into the system to seek information.

[1002] "Answer" refers to the information provided in response to a user's inquiry.

[1003] This invention is a system aimed at improving the efficiency of operations and speeding up inquiry response in logistics centers. This system automatically collects documents and files, stores them in a database, trains an artificial intelligence model on them, and provides optimal answers to user inquiries. The specific program processing of this system is described below.

[1004] 1. Data Collection Phase

[1005] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in a database.

[1006] 2. AI Model Training Phase

[1007] The server trains an artificial intelligence model based on literature and files stored in the database. First, the data is preprocessed, and the text data is analyzed using natural language processing tools such as tokenizers and stemming. This process uses libraries such as spaCy and Huggingface Transformers. After that, a training dataset is created based on the preprocessed data, and the artificial intelligence model is trained using machine learning algorithms such as the BERT model.

[1008] 3. Query Processing Phase

[1009] The user accesses the system using their smart device (e.g., smartphone or smart glasses) and enters an inquiry. For example: "Please tell me how to package the new product." The device sends this input to the server. The server analyzes the received inquiry using natural language processing tools and understands its content. Based on the analysis, it generates the optimal answer using a trained artificial intelligence model. This generated answer is then sent back to the user's device.

[1010] Specific example

[1011] Consider a scenario where a user asks via smartphone, "How should I pack the product?" This inquiry is analyzed by the server, and the optimal packing procedure is displayed based on relevant manuals.

[1012] 4. Maintenance Phase

[1013] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. Furthermore, it retrains its artificial intelligence model based on this new data to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy to ensure the security of the system.

[1014] Example of a prompt

[1015] As a concrete example, the following prompt statements are possible:

[1016] User: "Please tell me how to pack the product."

[1017] Prompt: "Analyze the information on product packaging methods in the logistics center's work procedures and manuals, and indicate the optimal procedure."

[1018] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1019] Step 1: Data Collection

[1020] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in a database.

[1021] Input: Literature and files

[1022] Output: Data stored in the database

[1023] Specific operation: Retrieve data from file folders or online storage and upload it to the database system.

[1024] Step 2: Data Preprocessing

[1025] The server preprocesses the stored literature and files. It analyzes the text data using natural language processing tools such as tokenizers and stemming tools (e.g., spaCy).

[1026] Input: Literature and files stored in the database

[1027] Output: Preprocessed text data

[1028] Specific operation: Tokenize the text, perform stemming, and save the analysis results as structured data.

[1029] Step 3: AI Model Training

[1030] The server trains an artificial intelligence model using machine learning algorithms such as the BERT model, based on preprocessed data. This data is processed using the Huggingface Transformers library.

[1031] Input: Preprocessed text data

[1032] Output: Trained artificial intelligence model

[1033] Specific operation: Preprocessed text data is created as a training dataset, and training is performed using the BERT model.

[1034] Step 4: Enter your inquiry

[1035] Users access the system using their smart devices and enter their inquiries. For example, they might submit an inquiry such as, "Please tell me how to pack the product."

[1036] Input: User inquiry text

[1037] Output: Sending a query to the server

[1038] Specific operation: The user enters text through the interface of a smart device and sends that inquiry to the server.

[1039] Step 5: Inquiry Analysis

[1040] The server analyzes the received query using a natural language processing tool (e.g., spaCy) to understand its content.

[1041] Input: User inquiry text

[1042] Output: Analyzed keywords and queries

[1043] Specific operation: The received inquiry text is tokenized, keywords are extracted, and queries are generated.

[1044] Step 6: Generate Answer

[1045] The server generates the optimal response using a pre-trained artificial intelligence model based on the analyzed data. This generated response includes information extracted from relevant literature.

[1046] Input: Analyzed keywords and queries

[1047] Output: The best answer for the user

[1048] Specific operation: Using a pre-trained artificial intelligence model, extract information based on keywords and queries, and generate answers in natural language.

[1049] Step 7: Provide your answer

[1050] The server sends the generated answer back to the user's device and displays it to the user. The user can obtain specific and timely answers to their questions.

[1051] Input: Response text from the server

[1052] Output: The answer displayed on the user's smart device.

[1053] Specific operation: The response text sent from the server is displayed on the user's device.

[1054] Step 8: Maintenance

[1055] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. It also retrains artificial intelligence models based on this new data. Furthermore, all data communications are encrypted according to security policies to ensure system security.

[1056] Input: New documents and files, security policy

[1057] Output: Updated database and retrained artificial intelligence model

[1058] Specific actions: Regularly collect and store data, retrain artificial intelligence models, and implement security measures.

[1059] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1060] This invention provides a system that streamlines the handling of inquiries from employees and partner companies, and recognizes user emotions to provide appropriate responses. This system automatically collects documents and files, stores them in a database, and trains an artificial intelligence (AI) model. Furthermore, by combining this with an emotion engine, it recognizes user emotions and provides appropriate responses tailored to those emotions.

[1061] Data collection phase

[1062] The server automatically collects documents and files such as operational manuals, user guides, and support documents from internal repositories and online storage. The collected documents and files are stored in an integrated database at regular intervals.

[1063] AI model training phase

[1064] The server preprocesses literature and files stored in the integrated database using natural language processing (NLP) tools. Specifically, it uses a tokenizer to segment the text and stemming to extract the core forms of words. Then, it generates a training dataset based on the preprocessed data and trains an AI model using a machine learning algorithm (e.g., a transformer model). This model is used to generate the best possible answers to user inquiries.

[1065] Query processing phase

[1066] The user enters their inquiry into the system from their terminal. For example, they might enter, "Please tell me how to use the new OA system." This input is then sent from the terminal to the server.

[1067] The server analyzes the received inquiry using NLP tools to extract key keywords and their relationships. It then uses a pre-trained AI model to generate the optimal response.

[1068] Emotion recognition phase

[1069] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis. The emotion engine analyzes the user's input text and adjusts the tone and content of the AI ​​model's response accordingly. For example, if the user indicates that they are "distressed," the response will be adjusted to a gentle and polite tone.

[1070] Answer generation phase

[1071] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The device then displays this response to the user. This allows the user to receive specific and timely responses that are tailored to their emotions.

[1072] Specific example

[1073] The following is an example of how to process a specific query using the emotion engine.

[1074] User: "I'm having trouble figuring out how to use the new office automation system. Please help me."

[1075] The device sends this query to the server.

[1076] The server analyzes the received text using an NLP tool and extracts keywords such as "OA system," "how to use," and "having trouble," along with the associated emotions.

[1077] The server activates an emotion engine based on the extracted emotion "distressed," and uses an AI model to generate a gentle and polite response, such as "Don't worry, here's how to use the new OA system..."

[1078] The device displays the generated response to the user. By receiving a response that is tailored to their own emotions, the user can resolve the problem with confidence.

[1079] Maintenance Phase

[1080] The server periodically recollects new literature and updated files, updating the database and AI models. This includes the process of analyzing new data and retraining the AI ​​models. Furthermore, the server ensures system security by encrypting all data communications and implementing access controls based on security policies.

[1081] In this way, this system, equipped with an emotion engine, enables efficient and emotionally sensitive inquiry handling, contributing to improved user experience and reduced workload for the help desk.

[1082] The following describes the processing flow.

[1083] Step 1:

[1084] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal and external repositories and online storage. This collection is performed at specific time intervals or triggered by specific events.

[1085] Step 2:

[1086] The server analyzes the collected literature and files and generates file metadata (e.g., file name, creation date, category). This metadata generation uses tools for file format detection and text extraction.

[1087] Step 3:

[1088] The server stores the contents of the documents and files, along with the generated metadata, in an integrated database. This storage process uses a database management system to normalize the data and make it searchable efficiently.

[1089] Step 4:

[1090] The server preprocesses the literature and files stored in the integrated database using natural language processing tools. Specifically, this preprocessing involves splitting the text using a tokenizer and converting words into their stem forms using stemming.

[1091] Step 5:

[1092] The server generates a training dataset based on preprocessed data and trains an artificial intelligence model using machine learning algorithms. This training includes cross-validation between the training dataset and the test dataset.

[1093] Step 6:

[1094] The user accesses the system using a terminal and enters an inquiry. For example, they might enter, "I'm having trouble figuring out how to use the new office automation system; please help me." The terminal then sends this input to the server.

[1095] Step 7:

[1096] The server analyzes the received inquiry using natural language processing tools and extracts key keywords and their relationships. For example, keywords such as "OA system," "how to use," and "I'm having trouble" might be extracted.

[1097] Step 8:

[1098] The server activates the emotion engine based on the extracted keywords. The emotion engine analyzes the user's text to determine emotions such as "distress," and adjusts the tone and content of the response based on those emotions.

[1099] Step 9:

[1100] Based on the results of the emotion engine, the server uses a trained artificial intelligence model to generate the optimal response. For example, it might generate a response in a gentle tone such as, "Don't worry, the instructions for using the new OA system are as follows..."

[1101] Step 10:

[1102] The server sends the generated response back to the user's device. The device displays this response to the user. The user can receive specific and timely responses that are tailored to their emotions.

[1103] Step 11:

[1104] The server periodically recollects new literature and updated files, updating the database and artificial intelligence models. This includes the process of analyzing new data and retraining the AI ​​models.

[1105] Step 12:

[1106] The server encrypts all data communications and implements access control based on security policies. This ensures the security of the entire system.

[1107] (Example 2)

[1108] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1109] Conventional inquiry handling systems often disregard user emotions and provide mechanical answers, making it difficult to increase user satisfaction. Furthermore, adequate support is needed for model updates to provide optimal answers tailored to the content of inquiries, as well as for data security. This invention aims to address these problems and achieve efficient and emotionally sensitive inquiry handling.

[1110] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1111] In this invention, the server includes means for collecting literature and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating optimal answers using the trained artificial intelligence model; means for providing the generated answers to the user; means for automatically collecting new literature and files periodically and updating the database and artificial intelligence model; means for protecting communications and data based on security policies; and means for performing sentiment analysis based on the content of user inquiries and adjusting the answers with appropriate tone and content according to the analysis results. This enables efficient and emotionally sensitive inquiry handling.

[1112] "Documents and files" refers to a collection of data containing information necessary for responding to inquiries, such as operational manuals, user guides, and support documents.

[1113] A "database" is a system for efficiently storing, managing, and searching for documents and files.

[1114] An "artificial intelligence model" is a model trained to generate the best possible answer to a query using machine learning algorithms.

[1115] "Natural language processing" is a technique that uses tools such as tokenizers and stemming to analyze user inquiries.

[1116] "Sentiment analysis" is the process of identifying emotions from the content of a user's inquiry and adjusting the response to match those emotions in terms of tone and content.

[1117] "Means of protecting communications and data" refer to technologies that encrypt data communications using SSL / TLS, etc., and restrict access to system resources using access control lists (ACLs).

[1118] "Means for automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models" refers to the process of continuously collecting new information, updating databases based on that information, and retraining AI models to keep them up-to-date.

[1119] "Methods for analyzing inquiries and generating answers" refers to a function that analyzes user inquiries using NLP tools and generates the optimal answer using a pre-trained artificial intelligence model.

[1120] This invention is a system that streamlines the handling of inquiries from employees and partner companies, and recognizes user emotions to provide appropriate responses. This system automatically collects documents and files, stores them in a database, and trains an artificial intelligence model. Furthermore, by combining this with an emotion engine, it recognizes user emotions and provides appropriate responses tailored to those emotions.

[1121] Data collection phase

[1122] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal repositories and online storage. The collected documentation and files are periodically saved to an integrated database. This process is performed using a crawler and a database management system (e.g., MySQL).

[1123] AI model training phase

[1124] The server preprocesses literature and files stored in the integrated database using natural language processing (NLP) tools (e.g., NLTK, Spacy). It tokenizes the text and extracts word stems using stemming. Then, it generates a training dataset based on the preprocessed data and trains an AI model using machine learning algorithms (e.g., BERT, GPT-3). This AI model is used to generate the best possible answers to user inquiries.

[1125] Query processing phase

[1126] The user enters their inquiry into the system from their terminal. For example, they might enter, "Please tell me how to use the new OA system." This input is then sent from the terminal to the server.

[1127] The server analyzes the received inquiry using NLP tools to extract key keywords and their relationships. It then uses a pre-trained AI model to generate the optimal response.

[1128] Emotion recognition phase

[1129] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis. The emotion engine analyzes the user's input text and adjusts the tone and content of the AI ​​model's response accordingly. For example, if the user indicates that they are "distressed," the response will be adjusted to a gentle and polite tone.

[1130] Answer generation phase

[1131] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The device then displays this response to the user. This allows the user to receive specific and timely responses that are tailored to their emotions.

[1132] Maintenance Phase

[1133] The server periodically recollects new literature and updated files, updating the database and AI models. This process includes analyzing new data and retraining the AI ​​models. Furthermore, the server ensures system security by encrypting all data communications and implementing access controls based on security policies.

[1134] Specific example

[1135] The following is an example of how to process a specific query using the emotion engine.

[1136] User: "I'm having trouble figuring out how to use the new office automation system. Please help me."

[1137] The device sends this query to the server.

[1138] The server analyzes the received text using an NLP tool and extracts keywords such as "OA system," "how to use," and "having trouble," along with the associated emotions.

[1139] The server activates an emotion engine based on the extracted emotion "distressed," and uses an AI model to generate a gentle and polite response, such as "Don't worry, here's how to use the new OA system..."

[1140] The device displays the generated response to the user. By receiving a response that is tailored to their own emotions, the user can resolve the problem with confidence.

[1141] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1142] Step 1:

[1143] The server executes data collection scripts to download operational manuals, user guides, and support documents from internal repositories and online storage.

[1144] Input: URLs of internal repositories and online storage, API keys, etc.

[1145] Output: Downloaded literature and files

[1146] Specific operation: The server uses a crawler to download literature and files from the specified URL via FTP or API.

[1147] Step 2:

[1148] The server saves downloaded documents and files to an integrated database.

[1149] Input: Downloaded literature and files

[1150] Output: Data stored in the integrated database

[1151] Specific operation: The server inserts the literature and files into the database management system (e.g., MySQL).

[1152] Step 3:

[1153] The server preprocesses the literature and files stored in the integrated database using NLP tools.

[1154] Input: Literature and files stored in the integrated database

[1155] Output: Preprocessed text data

[1156] Specific operation: The server uses NLP tools (NLTK, Spacy) to split the text with a tokenizer and perform stemming.

[1157] Step 4:

[1158] The server generates a training dataset based on pre-processed data.

[1159] Input: Preprocessed text data

[1160] Output: Training dataset

[1161] Specific operation: The server converts the preprocessed data into a CSV file, and then uses TensorFlow or PyTorch to generate a training dataset based on that file.

[1162] Step 5:

[1163] The server uses machine learning algorithms to train the AI ​​model.

[1164] Input: Training dataset

[1165] Output: Trained AI model file

[1166] Specific operation: The server trains an AI model using a Transformer model (e.g., the Hugging Face Transformers library) and saves it to "transformer_model.bin".

[1167] Step 6:

[1168] The user enters an inquiry from their device, and the device sends that information to the server.

[1169] Input: Inquiry details (Example: "Please tell me how to use the new office automation system.")

[1170] Output: Query data sent to the server

[1171] Specific operation: The user enters a query into an input form on their device, and the device uses a REST API to send this data to the server.

[1172] Step 7:

[1173] The server analyzes the received query content using an NLP tool.

[1174] Input: Inquiry details

[1175] Output: Key keywords and their relationships

[1176] Specific operation: The server uses Spacy to extract key keywords such as "OA system" and "how to use".

[1177] Step 8:

[1178] The server uses a trained AI model to generate the optimal answer.

[1179] Input: Analyzed query content

[1180] Output: Generated answer

[1181] Specific operation: The server loads "transformer_model.bin" and generates a response based on the query.

[1182] Step 9:

[1183] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis.

[1184] Input: Inquiry details

[1185] Output: Emotion analysis results

[1186] Specific operation: The server calls an emotion engine (e.g., Google Cloud Natural Language API) to analyze emotions and identify the emotion "distressed".

[1187] Step 10:

[1188] The server adjusts the tone and content of the AI ​​model's responses based on the sentiment analysis results.

[1189] Input: Sentiment analysis results, generated responses

[1190] Output: Adjusted answer

[1191] Specific action: Based on the emotion "distressed," the server changes the tone of its response to a gentler and more polite one.

[1192] Step 11:

[1193] The server sends the adjusted response back to the terminal, which then displays it to the user.

[1194] Input: Adjusted answer

[1195] Output: Answer displayed to the user

[1196] Specific operation: The server sends the prepared response back to the terminal in JSON format as a REST API response, and the terminal displays it on the screen.

[1197] Step 12:

[1198] The server periodically recollects new literature and updated files, updating the database and AI models.

[1199] Input: New literature or updated files

[1200] Output: Updated database and AI model

[1201] Specific operation: The server uses scheduled jobs to collect new data, insert it into the database, and retrain the AI ​​model based on the new data.

[1202] Step 13:

[1203] The server encrypts all data communications and implements access control.

[1204] Input: Communication data, access request

[1205] Output: Encrypted data, secure access

[1206] Specific operation: The server encrypts data communication using SSL / TLS and ensures security using Access Control Lists (ACLs).

[1207] (Application Example 2)

[1208] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1209] Previously, responses to inquiries from employees and partner companies were inefficient, and responses that considered user emotions were not provided. As a result, improvements in user experience and the efficiency of inquiry handling were not fully achieved. Furthermore, in security services, there is a need for responses that give users peace of mind in times of anxiety or emergency.

[1210] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting literature and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for recognizing the user's emotions, analyzing them using natural language processing, and generating an optimal response using the trained artificial intelligence model; means for providing the generated response to the user; means for adjusting the tone of the response based on the user's emotions using an emotion recognition engine; means for automatically collecting new literature and files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on a security policy. This makes it possible to provide prompt and appropriate inquiry response and security services that take into consideration the user's emotions.

[1211] "Documents and files" refer to materials containing information necessary to respond to user inquiries, including operational manuals, user guides, and security-related guidelines.

[1212] A "database" is an information system that stores collected literature and files, and allows for the efficient management and retrieval of information necessary to respond to inquiries.

[1213] An "artificial intelligence model" is an algorithm that learns from collected literature and files and is used to automatically generate the best possible answers to user inquiries.

[1214] "Natural language processing" is a technology used to analyze user inquiries and extract key keywords and their relationships.

[1215] An "emotion recognition engine" is a system that analyzes emotions from the user's input text and adjusts the tone and content of the response accordingly.

[1216] A "security policy" is a set of rules and guidelines designed to ensure data communication and information protection within a system.

[1217] "Response tone" refers to the tone and attitude of the response, which are adjusted with consideration for the user's feelings.

[1218] "New documents and files" refer to materials that are added regularly, including the latest operational manuals, user guides, and security-related guidelines.

[1219] The system for carrying out this invention is a response system that integrates multiple means to provide prompt and appropriate answers to inquiries while taking into consideration the user's feelings. This system mainly uses the following hardware and software.

[1220] First, the server automatically collects documents and files from internal repositories and online storage and stores them in an integrated database. These documents and files include operational manuals, user guides, and security-related guidelines. The server uses the requests library to retrieve these documents.

[1221] Next, the server trains an artificial intelligence model based on the stored literature and files. This process utilizes natural language processing (NLP) tools. Specifically, it uses a tokenizer to segment the text and stemming to extract the core forms of words. Then, it generates a training dataset based on the preprocessed data and trains the AI ​​model using a machine learning algorithm (e.g., a transformer model). The pipeline function of the transformers library is used in this training phase.

[1222] Users enter inquiries into the system from their devices. For example, if a user makes an inquiry such as, "I'm scared because I heard sounds like someone broke into my house tonight," this input is sent from the device to the server.

[1223] The server analyzes the received inquiry using natural language processing tools to extract key keywords and their relationships. Simultaneously, it analyzes the user's emotions using an emotion recognition engine and adjusts the tone and content of the AI ​​model's response according to the emotions the user expresses.

[1224] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The user's device then displays this response to the user. In this way, the user can obtain a specific and timely response that is tailored to their emotions.

[1225] For example, if a user enters a question saying, "I'm scared because I heard sounds like someone broke into my house tonight," it will be processed as follows:

[1226] 1. Receive a user inquiry.

[1227] 2. Natural language processing tools were used to extract the keywords and emotions "intrusion" and "scary".

[1228] 3. The emotion analysis engine determined it to be "fear".

[1229] 4. Based on appropriate advice regarding "intrusion," adjust the tone of your response according to your emotions.

[1230] 5. Generate and provide the user with a response such as, "Don't worry. Contact the police immediately and evacuate to a safe place."

[1231] Examples of prompts to input into a generative AI model are as follows:

[1232] Q: I'm scared because I heard sounds tonight that sound like someone broke into my house.

[1233] A: Don't worry. Contact the police immediately and evacuate to a safe place.

[1234] In this way, this system, equipped with an emotion engine, provides a concrete means to achieve efficient inquiry handling that takes emotions into consideration.

[1235] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1236] Step 1:

[1237] The server automatically collects documents and files from internal repositories and online storage. It retrieves documents such as operational manuals, user guides, and security-related guidelines using the requests library and stores their contents in an integrated database. Input consists of documents and files in internal repositories and online storage, while output consists of documents and files stored in the integrated database.

[1238] Step 2:

[1239] The server trains an artificial intelligence model based on literature and files stored in an integrated database. Natural language processing (NLP) tools split the text (tokenizer) and extract the core form of words using stemming. Then, a training dataset is generated using the preprocessed data, and a transformer model is trained using the pipeline function of the transformers library. The input is literature and files stored in the integrated database, and the output is the trained artificial intelligence model.

[1240] Step 3:

[1241] The user enters the inquiry using their device. For example, the text might be, "I'm scared because I heard sounds like someone broke into my house tonight." The input is the user's inquiry text, and the output is the inquiry sent from the device to the server.

[1242] Step 4:

[1243] The server analyzes user inquiries using natural language processing tools to extract key keywords and their relationships. This analysis utilizes tokenizers and stemming, and an emotion recognition engine is used to analyze the user's emotions. The input is the user's inquiry text, and the output is the extracted keywords and emotion data.

[1244] Step 5:

[1245] The server generates the optimal response using a trained artificial intelligence model based on extracted keywords and user sentiment. This process involves generating prompts, inputting them into the AI ​​model, and obtaining the response. The input consists of extracted keywords and sentiment data, while the output is the generated response text.

[1246] Step 6:

[1247] The server sends a response, refined by the emotion recognition engine, to the user's device, which then displays this response to the user. The input is the generated response text, and the output is the response displayed on the user's device.

[1248] Step 7:

[1249] The server automatically collects new literature and files on a regular basis, updating the integrated database and artificial intelligence model. This process analyzes the new data and retrains the AI ​​model. The input is the newly collected literature and files, and the output is the updated database and retrained AI model.

[1250] The above processing steps enable prompt and appropriate responses to inquiries while taking user emotions into consideration.

[1251] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1252] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1253] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1254] [Fourth Embodiment]

[1255] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1256] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1257] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1258] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1259] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1260] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1261] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1262] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1263] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1264] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1265] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1266] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1267] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1268] This invention is a system for streamlining the handling of inquiries both inside and outside a company. This system automatically collects documents and files, stores them in a database, trains an artificial intelligence model on them, and provides optimal answers to user inquiries. The specific program processing of this system is described below.

[1269] 1. Data Collection Phase

[1270] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in an integrated database.

[1271] 2. AI Model Training Phase

[1272] The server trains an artificial intelligence model based on literature and files stored in an integrated database. First, the data is preprocessed, and the text data is analyzed using natural language processing tools such as tokenizers and stemming. Then, a training dataset is created based on the preprocessed data, and the artificial intelligence model is trained using a machine learning algorithm (e.g., a Transformer-based model).

[1273] 3. Query Processing Phase

[1274] The user accesses the system using their device and enters an inquiry. For example: "Please tell me how to use the new office automation system." The device then sends this input to the server.

[1275] The server analyzes the received inquiry using natural language processing to understand its content. Then, based on the analyzed content, it uses a pre-trained artificial intelligence model to generate the optimal response. This generated response is then sent back to the user's device.

[1276] Specific example

[1277] The following is a concrete example illustrating how actual queries are processed.

[1278] User: "Please tell me how to use the new office automation system."

[1279] The device sends this query to the server.

[1280] The server analyzes the received text using a natural language processing tool (e.g., a tokenizer) and extracts "new OA system" and "how to use" as keywords.

[1281] The server uses a trained artificial intelligence model to extract information on "how to use the new office automation system" from relevant literature and generates the optimal response. This response is generated in a format such as, "The procedure for using the new office automation system is as follows..."

[1282] The device displays the generated answers to the user. The user can obtain specific and timely answers to their questions.

[1283] Maintenance Phase

[1284] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. Furthermore, it retrains its artificial intelligence model based on this new data to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy to ensure the security of the system.

[1285] In this way, the system can efficiently and securely handle inquiries from employees and partner companies. As a result, the workload of the help desk is reduced, and the quality of service to users is improved.

[1286] The following describes the processing flow.

[1287] Step 1:

[1288] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal and external repositories and online storage. This is done using scripts that automatically crawl the files at regular intervals.

[1289] Step 2:

[1290] The server analyzes the collected literature and files and generates metadata (e.g., file name, creation date, category). This analysis uses programs to analyze file formats and structures.

[1291] Step 3:

[1292] The server stores the contents of the documents and files, along with the generated metadata, in an integrated database. This storage process uses a storage management system to organize and store the files.

[1293] Step 4:

[1294] The server preprocesses the literature and files stored in the integrated database using natural language processing tools. Specifically, it uses a tokenizer to split the text and stemming to extract the core forms of words.

[1295] Step 5:

[1296] The server generates a training dataset based on preprocessed data and trains an artificial intelligence model using a machine learning algorithm (e.g., a Transformer-based model). During this process, the model's performance is evaluated by dividing it into training and test data.

[1297] Step 6:

[1298] The user enters their inquiry using their device. For example, they might type, "Please tell me how to use the new office automation system." The device then sends this input to the server.

[1299] Step 7:

[1300] The server analyzes the received query using natural language processing tools to extract key keywords and their relationships. This analysis utilizes tools for noun extraction and dependency analysis.

[1301] Step 8:

[1302] The server uses a trained artificial intelligence model based on the analysis results to generate the optimal response. For example, it might generate a specific response such as, "The procedure for using the new office automation system is as follows..."

[1303] Step 9:

[1304] The server sends the generated response back to the user's device. The device then displays this response to the user.

[1305] Step 10:

[1306] The server periodically recollects new literature and updated files, updating the database and artificial intelligence models. This includes the process of analyzing and retraining on new data.

[1307] Step 11:

[1308] The server encrypts all data communications and implements access control based on security policies. This ensures the security of the entire system.

[1309] (Example 1)

[1310] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1311] Responding to inquiries both inside and outside the company requires considerable effort and time, making efficient work difficult. Furthermore, there is a lack of systems to properly manage information provided by various departments and partner companies and to provide prompt and accurate responses. Keeping up-to-date with daily updates is also a major challenge. This invention aims to solve these problems and provide a system that streamlines inquiry handling.

[1312] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1313] In this invention, the server includes means for collecting and storing documents and electronic files in a database for generating answers to inquiries from information providers; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating the optimal answer using the trained artificial intelligence model; means for providing the generated answer to the user; means for automatically collecting new documents and electronic files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on security policies. This makes it possible to respond to inquiries efficiently and accurately based on the latest information.

[1314] "Information provider" refers to the department or partner company that provides documents or electronic files.

[1315] An "inquiry" refers to a question or request that a user enters into the system.

[1316] "Documents" refer to text data such as business manuals and user guides.

[1317] An "electronic file" is data stored in a digital format, including PDFs and document files.

[1318] A "database" refers to a system for managing and storing collected documents and electronic files.

[1319] An "artificial intelligence model" refers to a program trained using machine learning algorithms that generates responses based on input data.

[1320] "Natural language processing" refers to the technology of using computers to understand, analyze, and generate human language.

[1321] A "tokenizer" refers to a software tool that divides text into words or phrases.

[1322] "Stemming" refers to a natural language processing technique that extracts the root portion of a word.

[1323] A "security policy" refers to the rules and procedures for ensuring the security of data and communications.

[1324] "Users" refer to individuals or organizations that submit inquiries to the system and receive responses.

[1325] "Automated data collection" refers to a process in which the system automatically collects data without requiring manual operation.

[1326] "Communication" refers to the sending and receiving of data and information between systems.

[1327] "Answer" refers to the response that the system generates in response to a user's inquiry.

[1328] "Updating" refers to the process of adding new information to databases and artificial intelligence models to keep them up-to-date.

[1329] This invention is a system for streamlining the handling of inquiries both inside and outside a company. This system automatically collects documents and electronic files, stores them in a database, trains an artificial intelligence model based on the stored data, and provides optimal answers to user inquiries. Specific embodiments of this system are described below.

[1330] The server automatically collects documents and electronic files provided by various departments and partner companies within the company at regular intervals. These documents and electronic files include instruction manuals and guidelines. The server uses data collection scripts and APIs to access the internet and internal networks and store this data in an integrated database. Hardware such as dedicated servers and cloud storage is used in this data collection process.

[1331] The server trains an artificial intelligence model based on documents and electronic files stored in an integrated database. First, the server preprocesses the data and analyzes the text data using natural language processing tools such as tokenizers and stemming (e.g., NLTK, spaCy). Then, it creates a training dataset based on the preprocessed data and trains a Transformer-based model (e.g., BERT, GPT) using a machine learning framework (e.g., TensorFlow, PyTorch).

[1332] Users access the system using their devices and enter inquiries. For example, they might ask, "How do I use the new office automation system?" The device sends this input to the server. The server analyzes the received inquiry using natural language processing tools and understands its content. Based on the analysis, it then generates the optimal answer using a trained artificial intelligence model. This generated answer is then sent back to the user's device.

[1333] The following is a concrete example illustrating how actual queries are processed.

[1334] "Please teach me how to use the new office automation system."

[1335] "What are the latest guidelines regarding employee onboarding and offboarding procedures?"

[1336] "I want to know how to set up project management software."

[1337] The server automatically collects new documents and updated electronic files from repositories and online storage periodically to update its database. Furthermore, it retrains its artificial intelligence model based on this new data, enabling it to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy and ensures system security using the SSL / TLS protocol.

[1338] In this way, the system can efficiently and securely handle inquiries from employees and partner companies. As a result, the workload of the help desk is reduced, and the quality of service to users is improved.

[1339] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1340] Step 1: Preparing for data collection

[1341] The server configures connections to collect documents and electronic files from various departments and partner companies. This includes configuring API and FTP connections.

[1342] Input: Connection information for API endpoints and FTP servers of each department and partner company.

[1343] Output: A list of endpoints that have completed the necessary connection settings for data collection and are therefore connectable.

[1344] Step 2: Perform data collection

[1345] The server automatically starts collecting data at specified time intervals. For example, it can trigger a collection task every Monday at 9:00 AM.

[1346] Input: Regular schedule information.

[1347] Output: A list of documents and electronic files downloaded from each endpoint.

[1348] Specific operation: The server makes HTTP requests or FTP connections to API endpoints or FTP servers of each department or partner company to download documents and files.

[1349] Step 3: Save the data to the database

[1350] The server saves downloaded documents and electronic files to a temporary directory, and then moves this data to the integrated database.

[1351] Input: Downloaded documents and electronic files.

[1352] Output: Entries for documents and electronic files stored in the integrated database.

[1353] Specific actions: Use scripts and ETL tools to perform data transformation and cleaning for database integration.

[1354] Step 4: Data Preprocessing

[1355] The server extracts data from the integrated database and performs preprocessing. Specifically, it tokenizes the text data using a tokenizer and performs stemming and stop word removal.

[1356] Input: Documents and electronic files extracted from an integrated database.

[1357] Output: Preprocessed text data.

[1358] Specific operation: Preprocessing is performed using natural language processing tools (e.g., NLTK, spaCy).

[1359] Step 5: Training the artificial intelligence model

[1360] The server creates a training dataset based on preprocessed data and trains the model using a machine learning framework (e.g., TensorFlow, PyTorch).

[1361] Input: Preprocessed text data.

[1362] Output: A trained artificial intelligence model.

[1363] Specific operation: Train the model using a Transformer-based model (e.g., BERT, GPT).

[1364] Step 6: Enter your inquiry

[1365] The user accesses the system using their own device and enters their inquiry. For example, they might enter, "Please tell me how to use the new office automation system."

[1366] Input: User inquiry details.

[1367] Output: The content of the query sent to the server.

[1368] Specific operation: The terminal sends the user's input query to the server as an HTTP request.

[1369] Step 7: Analyze the inquiry

[1370] The server analyzes the received query using natural language processing tools. Specifically, it uses a tokenizer to split the text and extract keywords.

[1371] Input: User inquiry details.

[1372] Output: Extracted keywords and query structure.

[1373] Specific operation: Performs text analysis using a tokenizer and stemming.

[1374] Step 8: Generating the answer

[1375] The server uses a trained artificial intelligence model to generate the best possible answer to the user's inquiry.

[1376] Input: Extracted keywords or query structure.

[1377] Output: The generated answer text.

[1378] Specific operation: Generate an answer using a pre-trained model and send that answer to the terminal as an HTTP response.

[1379] Step 9: Display the answer

[1380] The terminal displays the response sent back from the server to the user.

[1381] Input: Response text from the server.

[1382] Output: The answer displayed to the user.

[1383] Specific operation: Render the answer text on the screen and display it visually to the user.

[1384] Step 10: Database Update

[1385] The server automatically collects new documents and updated files from repositories and online storage on a regular basis and updates the database.

[1386] Input: A new document or electronic file.

[1387] Output: Updated integrated database.

[1388] Specific operation: The automated collection task is executed periodically, adding new data to the database.

[1389] Step 11: Retrain the model

[1390] The server retrains the artificial intelligence model using the new data.

[1391] Input: Updated text data.

[1392] Output: A retrained artificial intelligence model.

[1393] Specific operation: The machine learning algorithm is run again using new data to train the model with the latest information.

[1394] Step 12: Security Protection

[1395] The server uses the SSL / TLS protocol to encrypt all data communications.

[1396] Input: Communication data.

[1397] Output: Encrypted communication data.

[1398] Specific operation: Implements the SSL / TLS protocol to encrypt and protect communication data.

[1399] (Application Example 1)

[1400] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1401] Conventional systems make it difficult to improve operational efficiency and provide prompt responses to inquiries in logistics centers. In particular, the lack of a system that allows workers to quickly obtain necessary information makes delays and errors more likely. The present invention aims to solve these problems and provide a system that improves the operational efficiency of logistics centers.

[1402] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1403] In this invention, the server includes means for collecting documents and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating optimal answers using the trained artificial intelligence model; means for responding to inquiries from smart devices to provide logistics center work guidelines and procedures; means for providing the generated answers to users; means for automatically collecting new documents and files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on security policies. This enables improved work efficiency and quick and accurate response to inquiries in the logistics center.

[1404] "Documents and files" refer to documents and electronic files related to business operations both inside and outside the company, such as operational manuals, user guides, and support documents.

[1405] A "database" is a system that centrally manages collected literature and files, making them searchable and accessible.

[1406] An "artificial intelligence model" is an AI system trained using machine learning algorithms that generates the optimal answer to a specific task.

[1407] "Natural language processing" is a language processing technique that analyzes text-based inquiries from users and understands their meaning.

[1408] A "logistics center" is a warehouse-type facility that handles operations such as receiving, storing, and shipping goods and materials.

[1409] "Smart devices" are advanced electronic devices with communication capabilities, such as smartphones, smart glasses, and head-mounted displays.

[1410] A "security policy" is a set of rules and procedures aimed at protecting data and communications.

[1411] "Work guidelines" are guidelines that describe the procedures and methods for various tasks in a logistics center.

[1412] An "inquiry" is a question or request that a user enters into the system to seek information.

[1413] "Answer" refers to the information provided in response to a user's inquiry.

[1414] This invention is a system aimed at improving the efficiency of operations and speeding up inquiry response in logistics centers. This system automatically collects documents and files, stores them in a database, trains an artificial intelligence model on them, and provides optimal answers to user inquiries. The specific program processing of this system is described below.

[1415] 1. Data Collection Phase

[1416] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in a database.

[1417] 2. AI Model Training Phase

[1418] The server trains an artificial intelligence model based on literature and files stored in the database. First, the data is preprocessed, and the text data is analyzed using natural language processing tools such as tokenizers and stemming. This process uses libraries such as spaCy and Huggingface Transformers. After that, a training dataset is created based on the preprocessed data, and the artificial intelligence model is trained using machine learning algorithms such as the BERT model.

[1419] 3. Query Processing Phase

[1420] The user accesses the system using their smart device (e.g., smartphone or smart glasses) and enters an inquiry. For example: "Please tell me how to package the new product." The device sends this input to the server. The server analyzes the received inquiry using natural language processing tools and understands its content. Based on the analysis, it generates the optimal answer using a trained artificial intelligence model. This generated answer is then sent back to the user's device.

[1421] Specific example

[1422] Consider a scenario where a user asks via smartphone, "How should I pack the product?" This inquiry is analyzed by the server, and the optimal packing procedure is displayed based on relevant manuals.

[1423] 4. Maintenance Phase

[1424] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. Furthermore, it retrains its artificial intelligence model based on this new data to respond to user inquiries using the latest information. In addition, the server encrypts all data communications in accordance with its security policy to ensure the security of the system.

[1425] Example of a prompt

[1426] As a concrete example, the following prompt statements are possible:

[1427] User: "Please tell me how to pack the product."

[1428] Prompt: "Analyze the information on product packaging methods in the logistics center's work procedures and manuals, and indicate the optimal procedure."

[1429] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1430] Step 1: Data Collection

[1431] The server automatically collects documents and files provided by each department and partner company at regular intervals. These documents and files include operational manuals, user guides, and support documents. The collected data is stored in a database.

[1432] Input: Literature and files

[1433] Output: Data stored in the database

[1434] Specific operation: Retrieve data from file folders or online storage and upload it to the database system.

[1435] Step 2: Data Preprocessing

[1436] The server preprocesses the stored literature and files. It analyzes the text data using natural language processing tools such as tokenizers and stemming tools (e.g., spaCy).

[1437] Input: Literature and files stored in the database

[1438] Output: Preprocessed text data

[1439] Specific operation: Tokenize the text, perform stemming, and save the analysis results as structured data.

[1440] Step 3: AI Model Training

[1441] The server trains an artificial intelligence model using machine learning algorithms such as the BERT model, based on preprocessed data. This data is processed using the Huggingface Transformers library.

[1442] Input: Preprocessed text data

[1443] Output: Trained artificial intelligence model

[1444] Specific operation: Preprocessed text data is created as a training dataset, and training is performed using the BERT model.

[1445] Step 4: Enter your inquiry

[1446] Users access the system using their smart devices and enter their inquiries. For example, they might submit an inquiry such as, "Please tell me how to pack the product."

[1447] Input: User inquiry text

[1448] Output: Sending a query to the server

[1449] Specific operation: The user enters text through the interface of a smart device and sends that inquiry to the server.

[1450] Step 5: Inquiry Analysis

[1451] The server analyzes the received query using a natural language processing tool (e.g., spaCy) to understand its content.

[1452] Input: User inquiry text

[1453] Output: Analyzed keywords and queries

[1454] Specific operation: The received inquiry text is tokenized, keywords are extracted, and queries are generated.

[1455] Step 6: Generate Answer

[1456] The server generates the optimal response using a pre-trained artificial intelligence model based on the analyzed data. This generated response includes information extracted from relevant literature.

[1457] Input: Analyzed keywords and queries

[1458] Output: The best answer for the user

[1459] Specific operation: Using a pre-trained artificial intelligence model, extract information based on keywords and queries, and generate answers in natural language.

[1460] Step 7: Provide your answer

[1461] The server sends the generated answer back to the user's device and displays it to the user. The user can obtain specific and timely answers to their questions.

[1462] Input: Response text from the server

[1463] Output: The answer displayed on the user's smart device.

[1464] Specific operation: The response text sent from the server is displayed on the user's device.

[1465] Step 8: Maintenance

[1466] The server automatically collects new literature and updated files from repositories and online storage on a regular basis to update the database. It also retrains artificial intelligence models based on this new data. Furthermore, all data communications are encrypted according to security policies to ensure system security.

[1467] Input: New documents and files, security policy

[1468] Output: Updated database and retrained artificial intelligence model

[1469] Specific actions: Regularly collect and store data, retrain artificial intelligence models, and implement security measures.

[1470] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1471] This invention provides a system that streamlines the handling of inquiries from employees and partner companies, and recognizes user emotions to provide appropriate responses. This system automatically collects documents and files, stores them in a database, and trains an artificial intelligence (AI) model. Furthermore, by combining this with an emotion engine, it recognizes user emotions and provides appropriate responses tailored to those emotions.

[1472] Data collection phase

[1473] The server automatically collects documents and files such as operational manuals, user guides, and support documents from internal repositories and online storage. The collected documents and files are stored in an integrated database at regular intervals.

[1474] AI model training phase

[1475] The server preprocesses literature and files stored in the integrated database using natural language processing (NLP) tools. Specifically, it uses a tokenizer to segment the text and stemming to extract the core forms of words. Then, it generates a training dataset based on the preprocessed data and trains an AI model using a machine learning algorithm (e.g., a transformer model). This model is used to generate the best possible answers to user inquiries.

[1476] Query processing phase

[1477] The user enters their inquiry into the system from their terminal. For example, they might enter, "Please tell me how to use the new OA system." This input is then sent from the terminal to the server.

[1478] The server analyzes the received inquiry using NLP tools to extract key keywords and their relationships. It then uses a pre-trained AI model to generate the optimal response.

[1479] Emotion recognition phase

[1480] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis. The emotion engine analyzes the user's input text and adjusts the tone and content of the AI ​​model's response accordingly. For example, if the user indicates that they are "distressed," the response will be adjusted to a gentle and polite tone.

[1481] Answer generation phase

[1482] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The device then displays this response to the user. This allows the user to receive specific and timely responses that are tailored to their emotions.

[1483] Specific example

[1484] The following is an example of how to process a specific query using the emotion engine.

[1485] User: "I'm having trouble figuring out how to use the new office automation system. Please help me."

[1486] The device sends this query to the server.

[1487] The server analyzes the received text using an NLP tool and extracts keywords such as "OA system," "how to use," and "having trouble," along with the associated emotions.

[1488] The server activates an emotion engine based on the extracted emotion "distressed," and uses an AI model to generate a gentle and polite response, such as "Don't worry, here's how to use the new OA system..."

[1489] The device displays the generated response to the user. By receiving a response that is tailored to their own emotions, the user can resolve the problem with confidence.

[1490] Maintenance Phase

[1491] The server periodically recollects new literature and updated files, updating the database and AI models. This includes the process of analyzing new data and retraining the AI ​​models. Furthermore, the server ensures system security by encrypting all data communications and implementing access controls based on security policies.

[1492] In this way, this system, equipped with an emotion engine, enables efficient and emotionally sensitive inquiry handling, contributing to improved user experience and reduced workload for the help desk.

[1493] The following describes the processing flow.

[1494] Step 1:

[1495] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal and external repositories and online storage. This collection is performed at specific time intervals or triggered by specific events.

[1496] Step 2:

[1497] The server analyzes the collected literature and files and generates file metadata (e.g., file name, creation date, category). This metadata generation uses tools for file format detection and text extraction.

[1498] Step 3:

[1499] The server stores the contents of the documents and files, along with the generated metadata, in an integrated database. This storage process uses a database management system to normalize the data and make it searchable efficiently.

[1500] Step 4:

[1501] The server preprocesses the literature and files stored in the integrated database using natural language processing tools. Specifically, this preprocessing involves splitting the text using a tokenizer and converting words into their stem forms using stemming.

[1502] Step 5:

[1503] The server generates a training dataset based on preprocessed data and trains an artificial intelligence model using machine learning algorithms. This training includes cross-validation between the training dataset and the test dataset.

[1504] Step 6:

[1505] The user accesses the system using a terminal and enters an inquiry. For example, they might enter, "I'm having trouble figuring out how to use the new office automation system; please help me." The terminal then sends this input to the server.

[1506] Step 7:

[1507] The server analyzes the received inquiry using natural language processing tools and extracts key keywords and their relationships. For example, keywords such as "OA system," "how to use," and "I'm having trouble" might be extracted.

[1508] Step 8:

[1509] The server activates the emotion engine based on the extracted keywords. The emotion engine analyzes the user's text to determine emotions such as "distress," and adjusts the tone and content of the response based on those emotions.

[1510] Step 9:

[1511] Based on the results of the emotion engine, the server uses a trained artificial intelligence model to generate the optimal response. For example, it might generate a response in a gentle tone such as, "Don't worry, the instructions for using the new OA system are as follows..."

[1512] Step 10:

[1513] The server sends the generated response back to the user's device. The device displays this response to the user. The user can receive specific and timely responses that are tailored to their emotions.

[1514] Step 11:

[1515] The server periodically recollects new literature and updated files, updating the database and artificial intelligence models. This includes the process of analyzing new data and retraining the AI ​​models.

[1516] Step 12:

[1517] The server encrypts all data communications and implements access control based on security policies. This ensures the security of the entire system.

[1518] (Example 2)

[1519] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1520] Conventional inquiry handling systems often disregard user emotions and provide mechanical answers, making it difficult to increase user satisfaction. Furthermore, adequate support is needed for model updates to provide optimal answers tailored to the content of inquiries, as well as for data security. This invention aims to address these problems and achieve efficient and emotionally sensitive inquiry handling.

[1521] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1522] In this invention, the server includes means for collecting literature and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for analyzing user inquiries using natural language processing and generating optimal answers using the trained artificial intelligence model; means for providing the generated answers to the user; means for automatically collecting new literature and files periodically and updating the database and artificial intelligence model; means for protecting communications and data based on security policies; and means for performing sentiment analysis based on the content of user inquiries and adjusting the answers with appropriate tone and content according to the analysis results. This enables efficient and emotionally sensitive inquiry handling.

[1523] "Documents and files" refers to a collection of data containing information necessary for responding to inquiries, such as operational manuals, user guides, and support documents.

[1524] A "database" is a system for efficiently storing, managing, and searching for documents and files.

[1525] An "artificial intelligence model" is a model trained to generate the best possible answer to a query using machine learning algorithms.

[1526] "Natural language processing" is a technique that uses tools such as tokenizers and stemming to analyze user inquiries.

[1527] "Sentiment analysis" is the process of identifying emotions from the content of a user's inquiry and adjusting the response to match those emotions in terms of tone and content.

[1528] "Means of protecting communications and data" refer to technologies that encrypt data communications using SSL / TLS, etc., and restrict access to system resources using access control lists (ACLs).

[1529] "Means for automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models" refers to the process of continuously collecting new information, updating databases based on that information, and retraining AI models to keep them up-to-date.

[1530] "Methods for analyzing inquiries and generating answers" refers to a function that analyzes user inquiries using NLP tools and generates the optimal answer using a pre-trained artificial intelligence model.

[1531] This invention is a system that streamlines the handling of inquiries from employees and partner companies, and recognizes user emotions to provide appropriate responses. This system automatically collects documents and files, stores them in a database, and trains an artificial intelligence model. Furthermore, by combining this with an emotion engine, it recognizes user emotions and provides appropriate responses tailored to those emotions.

[1532] Data collection phase

[1533] The server automatically collects documentation and files, such as operational manuals, user guides, and support documents, from internal repositories and online storage. The collected documentation and files are periodically saved to an integrated database. This process is performed using a crawler and a database management system (e.g., MySQL).

[1534] AI model training phase

[1535] The server preprocesses literature and files stored in the integrated database using natural language processing (NLP) tools (e.g., NLTK, Spacy). It tokenizes the text and extracts word stems using stemming. Then, it generates a training dataset based on the preprocessed data and trains an AI model using machine learning algorithms (e.g., BERT, GPT-3). This AI model is used to generate the best possible answers to user inquiries.

[1536] Query processing phase

[1537] The user enters their inquiry into the system from their terminal. For example, they might enter, "Please tell me how to use the new OA system." This input is then sent from the terminal to the server.

[1538] The server analyzes the received inquiry using NLP tools to extract key keywords and their relationships. It then uses a pre-trained AI model to generate the optimal response.

[1539] Emotion recognition phase

[1540] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis. The emotion engine analyzes the user's input text and adjusts the tone and content of the AI ​​model's response accordingly. For example, if the user indicates that they are "distressed," the response will be adjusted to a gentle and polite tone.

[1541] Answer generation phase

[1542] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The device then displays this response to the user. This allows the user to receive specific and timely responses that are tailored to their emotions.

[1543] Maintenance Phase

[1544] The server periodically recollects new literature and updated files, updating the database and AI models. This process includes analyzing new data and retraining the AI ​​models. Furthermore, the server ensures system security by encrypting all data communications and implementing access controls based on security policies.

[1545] Specific example

[1546] The following is an example of how to process a specific query using the emotion engine.

[1547] User: "I'm having trouble figuring out how to use the new office automation system. Please help me."

[1548] The device sends this query to the server.

[1549] The server analyzes the received text using an NLP tool and extracts keywords such as "OA system," "how to use," and "having trouble," along with the associated emotions.

[1550] The server activates an emotion engine based on the extracted emotion "distressed," and uses an AI model to generate a gentle and polite response, such as "Don't worry, here's how to use the new OA system..."

[1551] The device displays the generated response to the user. By receiving a response that is tailored to their own emotions, the user can resolve the problem with confidence.

[1552] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1553] Step 1:

[1554] The server executes data collection scripts to download operational manuals, user guides, and support documents from internal repositories and online storage.

[1555] Input: URLs of internal repositories and online storage, API keys, etc.

[1556] Output: Downloaded literature and files

[1557] Specific operation: The server uses a crawler to download literature and files from the specified URL via FTP or API.

[1558] Step 2:

[1559] The server saves downloaded documents and files to an integrated database.

[1560] Input: Downloaded literature and files

[1561] Output: Data stored in the integrated database

[1562] Specific operation: The server inserts the literature and files into the database management system (e.g., MySQL).

[1563] Step 3:

[1564] The server preprocesses the literature and files stored in the integrated database using NLP tools.

[1565] Input: Literature and files stored in the integrated database

[1566] Output: Preprocessed text data

[1567] Specific operation: The server uses NLP tools (NLTK, Spacy) to split the text with a tokenizer and perform stemming.

[1568] Step 4:

[1569] The server generates a training dataset based on pre-processed data.

[1570] Input: Preprocessed text data

[1571] Output: Training dataset

[1572] Specific operation: The server converts the preprocessed data into a CSV file, and then uses TensorFlow or PyTorch to generate a training dataset based on that file.

[1573] Step 5:

[1574] The server uses machine learning algorithms to train the AI ​​model.

[1575] Input: Training dataset

[1576] Output: Trained AI model file

[1577] Specific operation: The server trains an AI model using a Transformer model (e.g., the Hugging Face Transformers library) and saves it to "transformer_model.bin".

[1578] Step 6:

[1579] The user enters an inquiry from their device, and the device sends that information to the server.

[1580] Input: Inquiry details (Example: "Please tell me how to use the new office automation system.")

[1581] Output: Query data sent to the server

[1582] Specific operation: The user enters a query into an input form on their device, and the device uses a REST API to send this data to the server.

[1583] Step 7:

[1584] The server analyzes the received query content using an NLP tool.

[1585] Input: Inquiry details

[1586] Output: Key keywords and their relationships

[1587] Specific operation: The server uses Spacy to extract key keywords such as "OA system" and "how to use".

[1588] Step 8:

[1589] The server uses a trained AI model to generate the optimal answer.

[1590] Input: Analyzed query content

[1591] Output: Generated answer

[1592] Specific operation: The server loads "transformer_model.bin" and generates a response based on the query.

[1593] Step 9:

[1594] The server activates an emotion engine based on the user's inquiry and performs sentiment analysis.

[1595] Input: Inquiry details

[1596] Output: Emotion analysis results

[1597] Specific operation: The server calls an emotion engine (e.g., Google Cloud Natural Language API) to analyze emotions and identify the emotion "distressed".

[1598] Step 10:

[1599] The server adjusts the tone and content of the AI ​​model's responses based on the sentiment analysis results.

[1600] Input: Sentiment analysis results, generated responses

[1601] Output: Adjusted answer

[1602] Specific action: Based on the emotion "distressed," the server changes the tone of its response to a gentler and more polite one.

[1603] Step 11:

[1604] The server sends the adjusted response back to the terminal, which then displays it to the user.

[1605] Input: Adjusted answer

[1606] Output: Answer displayed to the user

[1607] Specific operation: The server sends the prepared response back to the terminal in JSON format as a REST API response, and the terminal displays it on the screen.

[1608] Step 12:

[1609] The server periodically recollects new literature and updated files, updating the database and AI models.

[1610] Input: New literature or updated files

[1611] Output: Updated database and AI model

[1612] Specific operation: The server uses scheduled jobs to collect new data, insert it into the database, and retrain the AI ​​model based on the new data.

[1613] Step 13:

[1614] The server encrypts all data communications and implements access control.

[1615] Input: Communication data, access request

[1616] Output: Encrypted data, secure access

[1617] Specific operation: The server encrypts data communication using SSL / TLS and ensures security using Access Control Lists (ACLs).

[1618] (Application Example 2)

[1619] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1620] Previously, responses to inquiries from employees and partner companies were inefficient, and responses that considered user emotions were not provided. As a result, improvements in user experience and the efficiency of inquiry handling were not fully achieved. Furthermore, in security services, there is a need for responses that give users peace of mind in times of anxiety or emergency.

[1621] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting literature and files and storing them in a database; means for training an artificial intelligence model based on the stored database; means for recognizing the user's emotions, analyzing them using natural language processing, and generating an optimal response using the trained artificial intelligence model; means for providing the generated response to the user; means for adjusting the tone of the response based on the user's emotions using an emotion recognition engine; means for automatically collecting new literature and files periodically and updating the database and artificial intelligence model; and means for protecting communications and data based on a security policy. This makes it possible to provide prompt and appropriate inquiry response and security services that take into consideration the user's emotions.

[1622] "Documents and files" refer to materials containing information necessary to respond to user inquiries, including operational manuals, user guides, and security-related guidelines.

[1623] A "database" is an information system that stores collected literature and files, and allows for the efficient management and retrieval of information necessary to respond to inquiries.

[1624] An "artificial intelligence model" is an algorithm that learns from collected literature and files and is used to automatically generate the best possible answers to user inquiries.

[1625] "Natural language processing" is a technology used to analyze user inquiries and extract key keywords and their relationships.

[1626] An "emotion recognition engine" is a system that analyzes emotions from the user's input text and adjusts the tone and content of the response accordingly.

[1627] A "security policy" is a set of rules and guidelines designed to ensure data communication and information protection within a system.

[1628] "Response tone" refers to the tone and attitude of the response, which are adjusted with consideration for the user's feelings.

[1629] "New documents and files" refer to materials that are added regularly, including the latest operational manuals, user guides, and security-related guidelines.

[1630] The system for carrying out this invention is a response system that integrates multiple means to provide prompt and appropriate answers to inquiries while taking into consideration the user's feelings. This system mainly uses the following hardware and software.

[1631] First, the server automatically collects documents and files from internal repositories and online storage and stores them in an integrated database. These documents and files include operational manuals, user guides, and security-related guidelines. The server uses the requests library to retrieve these documents.

[1632] Next, the server trains an artificial intelligence model based on the stored literature and files. This process utilizes natural language processing (NLP) tools. Specifically, it uses a tokenizer to segment the text and stemming to extract the core forms of words. Then, it generates a training dataset based on the preprocessed data and trains the AI ​​model using a machine learning algorithm (e.g., a transformer model). The pipeline function of the transformers library is used in this training phase.

[1633] Users enter inquiries into the system from their devices. For example, if a user makes an inquiry such as, "I'm scared because I heard sounds like someone broke into my house tonight," this input is sent from the device to the server.

[1634] The server analyzes the received inquiry using natural language processing tools to extract key keywords and their relationships. Simultaneously, it analyzes the user's emotions using an emotion recognition engine and adjusts the tone and content of the AI ​​model's response according to the emotions the user expresses.

[1635] The server generates a response adjusted by the emotion engine and sends it back to the user's device. The user's device then displays this response to the user. In this way, the user can obtain a specific and timely response that is tailored to their emotions.

[1636] For example, if a user enters a question saying, "I'm scared because I heard sounds like someone broke into my house tonight," it will be processed as follows:

[1637] 1. Receive a user inquiry.

[1638] 2. Natural language processing tools were used to extract the keywords and emotions "intrusion" and "scary".

[1639] 3. The emotion analysis engine determined it to be "fear".

[1640] 4. Based on appropriate advice regarding "intrusion," adjust the tone of your response according to your emotions.

[1641] 5. Generate and provide the user with a response such as, "Don't worry. Contact the police immediately and evacuate to a safe place."

[1642] Examples of prompts to input into a generative AI model are as follows:

[1643] Q: I'm scared because I heard sounds tonight that sound like someone broke into my house.

[1644] A: Don't worry. Contact the police immediately and evacuate to a safe place.

[1645] In this way, this system, equipped with an emotion engine, provides a concrete means to achieve efficient inquiry handling that takes emotions into consideration.

[1646] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1647] Step 1:

[1648] The server automatically collects documents and files from internal repositories and online storage. It retrieves documents such as operational manuals, user guides, and security-related guidelines using the requests library and stores their contents in an integrated database. Input consists of documents and files in internal repositories and online storage, while output consists of documents and files stored in the integrated database.

[1649] Step 2:

[1650] The server trains an artificial intelligence model based on literature and files stored in an integrated database. Natural language processing (NLP) tools split the text (tokenizer) and extract the core form of words using stemming. Then, a training dataset is generated using the preprocessed data, and a transformer model is trained using the pipeline function of the transformers library. The input is literature and files stored in the integrated database, and the output is the trained artificial intelligence model.

[1651] Step 3:

[1652] The user enters the inquiry using their device. For example, text such as, "I'm scared because I heard sounds like someone broke into my house tonight." The input is the user's inquiry text, and the output is the inquiry sent from the device to the server.

[1653] Step 4:

[1654] The server analyzes user inquiries using natural language processing tools to extract key keywords and their relationships. This analysis utilizes tokenizers and stemming, and an emotion recognition engine is used to analyze the user's emotions. The input is the user's inquiry text, and the output is the extracted keywords and emotion data.

[1655] Step 5:

[1656] The server generates the optimal response using a trained artificial intelligence model based on extracted keywords and user sentiment. This process involves generating prompts, inputting them into the AI ​​model, and obtaining the response. The input consists of extracted keywords and sentiment data, while the output is the generated response text.

[1657] Step 6:

[1658] The server sends a response, refined by the emotion recognition engine, to the user's device, which then displays this response to the user. The input is the generated response text, and the output is the response displayed on the user's device.

[1659] Step 7:

[1660] The server automatically collects new literature and files on a regular basis, updating the integrated database and artificial intelligence model. This process analyzes the new data and retrains the AI ​​model. The input is the newly collected literature and files, and the output is the updated database and retrained AI model.

[1661] The above processing steps enable prompt and appropriate responses to inquiries while taking user emotions into consideration.

[1662] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1663] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1664] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1665] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1666] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1667] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1668] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1669] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1670] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1671] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1672] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1673] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1674] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1675] 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.

[1676] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1677] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1678] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1679] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1680] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1681] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1682] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[1683] The following is further disclosed regarding the embodiments described above.

[1684] (Claim 1)

[1685] To generate responses to inquiries from employees and partner companies,

[1686] Means for collecting literature and files and storing them in a database,

[1687] A method for training an artificial intelligence model based on a saved database,

[1688] A method for analyzing user inquiries using natural language processing and generating the optimal response using a pre-trained artificial intelligence model,

[1689] A means of providing the generated answer to the user,

[1690] A means of automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models,

[1691] Means to protect communications and data based on security policies,

[1692] A system that includes this.

[1693] (Claim 2)

[1694] The system according to claim 1, characterized in that the means for generating the optimal response to a user inquiry is to use a natural language processing tool including a tokenizer and stemming.

[1695] (Claim 3)

[1696] The system according to claim 1, characterized in that the aforementioned documents and files include an operations manual and a user guide.

[1697] "Example 1"

[1698] (Claim 1)

[1699] To generate responses to inquiries from information providers,

[1700] Means for collecting documents and electronic files and storing them in a database,

[1701] A method for training an artificial intelligence model based on a saved database,

[1702] A method for analyzing user inquiries using natural language processing and generating optimal answers using a pre-trained artificial intelligence model,

[1703] A means of providing the generated answers to the user,

[1704] A means of automatically collecting new documents and electronic files on a regular basis and updating databases and artificial intelligence models,

[1705] Means to protect communications and data based on security policies,

[1706] A system that includes this.

[1707] (Claim 2)

[1708] The system according to claim 1, characterized in that the means for generating the optimal response to a user inquiry is to use a natural language processing tool including a tokenizer and stemming.

[1709] (Claim 3)

[1710] The system according to claim 1, characterized in that the aforementioned documents and electronic files include instruction manuals and guidelines.

[1711] "Application Example 1"

[1712] (Claim 1)

[1713] To generate responses to inquiries from employees and partner companies,

[1714] Means for collecting literature and files and storing them in a database,

[1715] A method for training an artificial intelligence model based on a saved database,

[1716] A method for analyzing user inquiries using natural language processing and generating the optimal response using a pre-trained artificial intelligence model,

[1717] To provide operational guidelines and procedures for the logistics center, we need a means to respond to inquiries from smart devices,

[1718] A means of providing the generated answer to the user,

[1719] A means of automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models,

[1720] Means to protect communications and data based on security policies,

[1721] A system that includes this.

[1722] (Claim 2)

[1723] The system according to claim 1, characterized in that the means for generating the optimal response to a user inquiry is to use a natural language processing tool including a tokenizer and stemming.

[1724] (Claim 3)

[1725] The system according to claim 1, characterized in that the aforementioned documents and files include an operations manual and a user guide.

[1726] "Example 2 of combining an emotion engine"

[1727] (Claim 1)

[1728] To generate responses to inquiries from employees and partner companies,

[1729] Means for collecting literature and files and storing them in a database,

[1730] A method for training an artificial intelligence model based on a saved database,

[1731] A method for analyzing user inquiries using natural language processing and generating the optimal response using a pre-trained artificial intelligence model,

[1732] A means of providing the generated answer to the user,

[1733] A means of automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models,

[1734] Means to protect communications and data based on security policies,

[1735] A means of performing sentiment analysis based on the user's inquiry and adjusting the response to an appropriate tone and content according to the analysis results,

[1736] A system that includes this.

[1737] (Claim 2)

[1738] The system according to claim 1, characterized in that the means for generating the optimal response to a user inquiry is to use a natural language processing tool including a tokenizer and stemming.

[1739] (Claim 3)

[1740] The system according to claim 1, characterized in that the aforementioned documents and files include an operations manual and a user guide.

[1741] "Application example 2 when combining with an emotional engine"

[1742] (Claim 1)

[1743] To generate responses to inquiries from employees and partner companies,

[1744] Means for collecting literature and files and storing them in a database,

[1745] A method for training an artificial intelligence model based on a saved database,

[1746] A means for recognizing user emotions, analyzing them using natural language processing, and generating the optimal response using a pre-trained artificial intelligence model,

[1747] A means of providing the generated answer to the user,

[1748] A means of adjusting the tone of responses based on the user's emotions using an emotion recognition engine,

[1749] A means of automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models,

[1750] Means to protect communications and data based on security policies,

[1751] A system that includes this.

[1752] (Claim 2)

[1753] The system according to claim 1, characterized in that the means for generating the optimal response to a user inquiry is to use a natural language processing tool including a tokenizer and stemming.

[1754] (Claim 3)

[1755] The system according to claim 1, characterized in that the aforementioned documents and files include security-related guidelines in addition to the operational manual and user guide. [Explanation of Symbols]

[1756] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. To generate responses to inquiries from employees and partner companies, Means for collecting literature and files and storing them in a database, A method for training an artificial intelligence model based on a saved database, A method for analyzing user inquiries using natural language processing and generating the optimal response using a pre-trained artificial intelligence model, A means of providing the generated answer to the user, A means of automatically collecting new literature and files on a regular basis and updating databases and artificial intelligence models, Means to protect communications and data based on security policies, A system that includes this.

2. The system according to claim 1, characterized in that the means for generating the optimal response to a user inquiry is to use a natural language processing tool including a tokenizer and stemming.

3. The system according to claim 1, characterized in that the aforementioned documents and files include an operations manual and a user guide.

Images