System

Abstract

A system is provided.SOLUTION: A system including means for collecting documents related to past projects, means for converting the collected documents into a unified format, means for extracting important information by using a natural language processing technology, means for storing the extracted information in a database, means for analyzing a question from a user and retrieving related information from the database, means for summarizing the retrieved information and providing it to the user, and means for providing an interface through which the user inputs a question.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] When managing multiple projects in corporate activities, a large amount of documents related to past projects, such as meeting minutes, materials, and upper management reviews, accumulates. It can be difficult to quickly extract and reference the necessary information from this vast amount of information. Furthermore, the inability to efficiently reference past information can lead to repetitive problems when planning new projects or campaigns, reducing work efficiency. [Means for solving the problem]

[0005] This invention provides a means for collecting all documents related to past projects and converting them into a unified format. Furthermore, it builds a system that uses natural language processing technology to extract important information from the collected documents and store it in a database. This database is organized by project, linking related information together. It also provides an interface for users to input questions, analyzes the user's question, searches the database for related information, and provides a system that summarizes that information and presents it to the user. This system enables users to quickly and efficiently refer to past information, thereby improving the efficiency of new project planning and business operations.

[0006] "Documents" refers to all documents generated in corporate activities, including minutes, materials, senior management reviews, and other related information.

[0007] "Unified format" refers to a standard data structure for converting documents of different formats into one unified data format.

[0008] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate natural human language.

[0009] "Important information" refers to information directly related to decision-making or problem-solving, such as project progress, decisions, and issues.

[0010] A "database" refers to a collection of data that stores collected information in an organized manner and makes it possible to search and use it efficiently.

[0011] "Project unit" refers to a method of organizing a series of activities carried out based on a specific period or purpose as a single unit.

[0012] "Interface" refers to the user interface that allows the user to operate the system, and refers to the mechanism for inputting questions and displaying results.

[0013] "Question analysis" refers to the process of understanding the language entered by the user and interpreting its intent and meaning.

[0014] A "summary" is a short summary of a large amount of extracted information, organized into a concise and easy-to-understand format. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0016] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0017] First, the terms used in the following description will be explained.

[0018] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0019] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0020] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0021] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0022] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0026] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0027] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0028] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0029] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0030] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0033] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0034] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0035] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0036] ---

[0037] The present invention is a system that collects documents related to past projects, converts them into a unified format, extracts important information using natural language processing technology, and stores it in a database. It also provides an interface for users to input questions, analyzes the user's questions, searches the database for relevant information, and summarizes the information for the user.

[0038] The program of this system mainly performs the following processes.

[0039] Data collection and integration

[0040] The server collects past documents from the company's network drive or cloud storage. The documents collected include meeting minutes, materials, and upper management reviews. These documents exist in various formats (PDF, Word, Excel, etc.), so the server converts them into digital data using OCR technology and parsers and organizes them into a unified format.

[0041] Extracting information and storing it in a database

[0042] The server then uses natural language processing to analyze the documents converted into a unified format and extract important keywords and phrases, such as project status, decisions, and issues. The extracted information is then classified as metadata and stored in a relational or NoSQL database.

[0043] Building a question-answering system

[0044] The server provides an interface for users to enter questions, which runs on a terminal and allows users to enter questions in natural language.

[0045] When a user types a question, the device sends it to the server, which uses machine learning models and natural language processing algorithms to analyze the question and convert it into an internal query, which then searches for relevant information in a database and extracts the best answer.

[0046] Generate and provide answers

[0047] The server then summarizes the information extracted from the database and generates a specific answer for the user, which is then sent to the terminal and displayed to the user. This process allows users to quickly obtain important information related to past projects, greatly improving work efficiency.

[0048] Specific examples

[0049] Example 1: Finding new campaign ideas

[0050] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0051] 2. The device sends this question to the server.

[0052] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0053] 4. The server extracts and summarizes relevant information from the database (such as past campaign successes and failures, opinions from upper management, etc.) and provides appropriate advice to the user.

[0054] 5. The terminal displays the response from the server to the user.

[0055] Example 2: Checking the progress of a project

[0056] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0057] 2. The device sends this question to the server.

[0058] 3. The server parses the question and extracts the latest information related to Project A from the database.

[0059] 4. The server summarizes the extracted information and provides the user with the current status of the project, any issues, next steps, etc.

[0060] 5. The device displays the server's response to the user, showing specific progress, for example, "Project A is currently in stage 3, with major tasks 70% complete. The next step is X."

[0061] In this way, the system of the present invention makes it possible to quickly and efficiently refer to detailed data on past projects, which greatly contributes to the planning of new projects and the efficiency of business operations.

[0062] The processing flow will be explained below.

[0063] ---

[0064] Step 1:

[0065] The server accesses the company's network drive or cloud storage to collect documents related to past projects, such as meeting minutes, materials, and senior management reviews.

[0066] Step 2:

[0067] The server converts the collected documents into text data using OCR technology, or if the documents are electronic files, they are imported as is.

[0068] Step 3:

[0069] The server converts documents in different formats into a unified format (e.g., JSON), using an appropriate parser for each file format.

[0070] Step 4:

[0071] The server uses natural language processing techniques to extract key information and keywords from the converted documents, including project status, decisions, and issues.

[0072] Step 5:

[0073] The server classifies the extracted information as metadata and stores it in a relational or NoSQL database. The data is organized by project, and highly related information is linked together.

[0074] Step 6:

[0075] The server provides the terminal with an interface for the user to input a question, the interface being designed to allow the user to input a question in natural language.

[0076] Step 7:

[0077] The user inputs a question into the terminal. For example, "I'm having trouble with a new campaign. What should I do?"

[0078] Step 8:

[0079] The terminal sends the user's question to the server.

[0080] Step 9:

[0081] The server receives the question and analyzes it using natural language processing techniques, identifying important keywords and phrases from the question and converting them into an internal query.

[0082] Step 10:

[0083] The server uses the internal query to search the database and extract information relevant to the question.

[0084] Step 11:

[0085] The server generates summaries based on the extracted information, such as examples of successful past campaigns or summaries of senior management's opinions.

[0086] Step 12:

[0087] The server transmits the generated summary to the terminal.

[0088] Step 13:

[0089] The device displays the summary sent from the server to the user, providing specific advice such as, "Project X is an example of a past successful campaign, and method Y was effective in this project."

[0090] ---

[0091] These are the specific steps of the program's processing, which allow users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[0092] Example 1

[0093] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0094] Because information related to past business projects is scattered, it is difficult to quickly and efficiently obtain the necessary information. In addition, because this information is stored in various formats, it is difficult to manage and search it in a unified manner. This inefficient management and search of information is a factor that reduces business efficiency.

[0095] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0096] In this invention, the server includes a means for collecting materials related to past business projects, a means for converting the collected materials into a unified format, and a means for extracting important information using machine learning technology. This makes it possible to efficiently collect information related to past business projects, convert it into a unified format, and extract important information. It can also analyze questions from users, search for related information from an information storage device, and provide a summary of the searched information, allowing users to quickly obtain the information they need.

[0097] "Materials" refers to documents and data related to past business projects, including minutes, reports, plans, emails, etc.

[0098] A "unified format" is a data format for converting collected materials into a consistent format, which makes it possible to handle materials in different formats in a uniform manner.

[0099] "Machine learning technology" is a technology for analyzing large amounts of data and automatically learning and making predictions, and includes algorithm-based analysis methods and deep learning technology.

[0100] "Information storage device" refers to a storage system, such as a database or cloud storage, for storing extracted information, including relational databases and NoSQL databases.

[0101] "Character recognition technology" means technology for converting printed or handwritten characters into digital data, including optical character recognition (OCR) technology.

[0102] A "natural language processing model" is a model for understanding and analyzing human language, including distributed representation models (Word2Vec) and translation models (BERT, GPT-3 (registered trademark)).

[0103] A "high-speed search engine" is an engine for quickly searching for required information from large amounts of data, and includes full-text search engines (ElasticSearch (registered trademark), Solr).

[0104] "Internal query" refers to a database query that is generated by analyzing a user's natural language question and searching for relevant information.

[0105] "Input device" refers to a hardware or software interface through which a user can input queries or data, including keyboards, touchscreens, and voice input devices.

[0106] This system efficiently collects materials related to past business projects, extracts important information, stores it in a database, and promptly provides relevant information in response to user questions. This system is composed of the following entities: a server, a terminal, and a user.

[0107] Data collection and integration

[0108] The server collects business-related materials from the company's network drive or cloud storage (e.g., a general cloud storage service). This includes meeting minutes, reports, plans, emails, etc. The server accesses the cloud storage using an appropriate API and retrieves files from the specified target directory. The collected materials are broad and often relate to the entire company's activities.

[0109] Unified data format conversion

[0110] The server uses open-source character recognition tools (e.g., Tesseract) to convert collected materials into digital data using optical character recognition (OCR) and document parsers. For example, OCR is performed on PDF files to extract text, and common document parsing libraries (e.g., Apache POI) are used for Word and Excel files. The converted data is then formatted into a consistent format (e.g., JSON, XML).

[0111] Information Extraction

[0112] The server then analyzes the unified formatted material using natural language processing (NLP) techniques, such as open-source NLP libraries (e.g., SpaCy, NLTK) and advanced machine learning models (e.g., BERT, GPT-3), to extract important noun phrases and key phrases from the document and generate a summary of the entire document.

[0113] Storage in the database

[0114] The server stores the extracted information in an information storage device. This information storage device can be a relational database (e.g., a general relational database management system) or a NoSQL database (e.g., a general NoSQL database solution). A table containing progress and risk factors for each project is designed and the information is stored there. An index is also created using a full-text search engine (e.g., Elasticsearch) to enable fast searches.

[0115] Receiving and analyzing user questions

[0116] A user inputs a question through an interface provided on the device. For example, "What is the latest progress on project A?" The device then sends this question to the server. The server analyzes the received question using natural language processing techniques and machine learning models. For example, it uses sentiment analysis and intent classification to identify the intent of the question and generates a related query. The internal query is used to search for related information.

[0117] Database search and answer generation

[0118] The server uses the parsed query to search for information in the information storage device. For example, a query like "latest progress on Project A" might retrieve the latest updates from the database. Elasticsearch might also be used to quickly find relevant documents. The server then summarizes the information from the database and generates a specific answer for the user. For example, "Project A is currently in Stage 3, with 70% of key tasks completed. The next step is X." This answer is then sent to the device and displayed to the user.

[0119] Specific examples

[0120] Example 1: Finding new campaign ideas

[0121] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0122] 2. The device sends this question to the server.

[0123] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0124] 4. The server extracts and summarizes relevant information from the information storage device (such as examples of past campaign successes and failures, opinions of upper management, etc.) and provides appropriate advice to the user.

[0125] 5. The terminal displays the response from the server to the user.

[0126] Example 2: Checking the progress of a project

[0127] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0128] 2. The device sends this question to the server.

[0129] 3. The server analyzes the question and extracts the latest information related to Project A from the information repository.

[0130] 4. The server summarizes the extracted information and provides the user with the current status of the project, any issues, next steps, etc.

[0131] 5. The device displays the server's response to the user, showing specific progress, for example, "Project A is currently in stage 3, with major tasks 70% complete. The next step is X."

[0132] In this way, the system of the present invention allows for quick and efficient reference to detailed data related to past business projects, greatly contributing to the planning of new projects and the efficiency of business operations.

[0133] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0134] Step 1: Collect data

[0135] The server collects business-related materials from the company's network drive or cloud storage (e.g., a general cloud storage service). The server accesses the cloud storage using the appropriate API and retrieves files from the specified target directory. Inputs include the cloud storage API endpoint, authentication information, and the specified directory, and the collected raw material files are included as output. Specifically, the server retrieves a list of files and downloads each file in the list.

[0136] Step 2: Convert data into a unified format

[0137] The server converts the collected data into digital data using optical character recognition (OCR) technology and a document parser. The input is raw data files such as PDF, Word, and Excel, and the output is text data. For example, Tesseract OCR is used to extract text data from PDF files, and Apache POI is used to extract text data from Word and Excel files. The extracted text data is then formatted into a consistent, unified format (e.g., JSON or XML).

[0138] Step 3: Extracting information

[0139] The server analyzes the unified formatted data using natural language processing technology. The input is the unified formatted text data, and the output is information such as extracted important noun phrases and key phrases. Specifically, it uses SpaCy and NLTK to tokenize the document and extract important keywords and phrases. It also uses BERT and GPT-3 models to generate a summary of the entire document.

[0140] Step 4: Store in the database

[0141] The server stores the extracted information in an information storage device. The input is the extracted key information and summary data, and the output is the entries stored in a database. Specifically, it inserts the information into a relational database (e.g., a typical relational database management system) or a NoSQL database. In addition, it uses Elasticsearch to create an index, enabling fast searches.

[0142] Step 5: Receiving user questions

[0143] The user inputs a question through an interface provided on the terminal. The input is a natural language question typed by the user, and the output is the question sent to the server. For example, the user might input, "What is the latest progress on project A?" The terminal receives this question and sends it to the server.

[0144] Step 6: Parsing the Question

[0145] The server analyzes the received question using natural language processing technology and machine learning models. The input is the question submitted by the user, and the output is a related internal query. Specifically, it uses Sentiment Analysis and Intent Classification to identify the intent of the question, and keyphrase extraction to extract important keywords. The internal query is generated in SQL or Elasticsearch query language.

[0146] Step 7: Database Search

[0147] The server uses the parsed query to search for information in the information storage device. The input is the generated internal query, and the output is the search results. Specifically, it uses a relational database or Elasticsearch to retrieve related information, such as the latest update records and related documents.

[0148] Step 8: Generate and provide answers

[0149] The server summarizes the information retrieved from the database and generates a specific answer. The input is the search result data, and the output includes summary information provided to the user. For example, it generates content such as "Project A is currently in stage 3, with major tasks 70% complete. The next step is X." The server then sends the generated answer to the device and displays it to the user.

[0150] (Application example 1)

[0151] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0152] In recent years, efficient customer service and sales operations have become increasingly important in brick-and-mortar stores. However, store staff face challenges in quickly searching and retrieving vast amounts of past customer service history and sales data. Furthermore, they are often unable to provide prompt and appropriate responses to customer inquiries. This can lead to lower customer satisfaction and a deterioration in store operational efficiency. A system that can solve these problems is needed.

[0153] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0154] In this invention, the server includes a means for collecting documents related to past projects, a means for converting the collected documents into a unified format, and a means for extracting important information using natural language processing technology, which enables store staff to search and obtain past customer interaction history and sales data, and enables the server to quickly generate appropriate answers to store staff questions using a question-answering model.

[0155] "Documents related to past projects" refers to documents such as minutes, materials, and reviews related to projects that a company or organization has carried out in the past.

[0156] "Unified format" refers to converting documents that exist in various formats into a consistent format that makes them easier to parse and process.

[0157] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and respond to human language.

[0158] A "database" is a system that stores data in a structured way so that it can be efficiently searched and retrieved.

[0159] The "interface through which the user inputs questions" refers to the functions and screens that allow the user to input questions to the system in natural language.

[0160] "A means for store staff to search and obtain past customer interaction history and sales data" refers to a function that enables store staff to quickly find and use information related to past customer interactions and sales performance.

[0161] A "question answering model" refers to a machine learning model that generates appropriate answers to questions from users.

[0162] "Customer interaction history" is data that records past interactions with customers.

[0163] "Sales data" refers to data containing information about past sales activities, such as sales volume, timing, and sales amount.

[0164] MODE FOR CARRYING OUT THE INVENTION

[0165] This invention is a system that allows store staff in physical stores to quickly search and obtain past customer interaction history and sales data, and provides appropriate answers using a question-answering model.

[0166] 1. System Configuration

[0167] The system collects documents related to past projects, converts them into a unified format, extracts important information using natural language processing technology, and stores it in a database. It provides an interface for users to input questions, which then searches for and summarizes relevant information from the database and provides it to the user. A specific example is shown below.

[0168] Hardware and Software

[0169] The hardware and software used are as follows:

[0170] Flask: This Python web framework builds an API server to accept and process user queries.

[0171] transformers: A natural language processing library by Hugging Face. This library provides a question-answering model.

[0172] JSON: A data storage format that is easily used as a database.

[0173] 2. Program processing explanation

[0174] The server uses Flask to build an API server and accepts questions from users. Questions are sent in JSON format and processed within the server. The server uses natural language processing technology to analyze the user's question and search for information in a database. It uses the Hugging Face pipeline as a question-answering model to generate an appropriate answer using the question and related information in the database.

[0175] The server searches and retrieves information from the database, generates a response based on this information, and returns it to the user, allowing the user to quickly obtain the information they need.

[0176] Examples of concrete examples and prompts

[0177] For example, if a store employee asks, "What's the stock status of the latest smartphone accessories?", the question is sent to the server. The server searches past inventory and sales data and analyzes the information using natural language processing technology. Finally, the question-answering model generates an appropriate answer, which is displayed on the device.

[0178] The specific prompt is as follows:

[0179] Please let me know the latest smartphone accessories in stock.

[0180] In this way, the system of the present invention allows store staff to quickly and efficiently search and retrieve detailed customer interaction history and sales data, enabling them to provide appropriate responses to customers, thereby improving business efficiency and customer satisfaction.

[0181] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0182] Step 1:

[0183] Collect historical documents and convert them into a unified format.

[0184] The server collects past documents (meeting minutes, materials, upper management reviews, etc.) from the company's network drive or cloud storage. The collected documents (PDF, Word, Excel, etc.) are converted into digital data using OCR technology and parsers and organized into a unified format.

[0185] Input: Corporate historical documents in various formats

[0186] Output: Digital data converted into a unified format

[0187] Step 2:

[0188] Analyze the data and extract important information.

[0189] The server uses natural language processing technology to analyze the documents converted into a unified format and extract important keywords and phrases such as project progress, decisions, issues, etc. The extracted information is classified as metadata.

[0190] Input: Digital data converted into a unified format

[0191] Output: Metadata of extracted important information

[0192] Step 3:

[0193] Stores the data in the database.

[0194] The server then stores the extracted metadata of the key information in a relational or NoSQL database, which also contains data from other similar projects and organizes the information by relevance.

[0195] Input: Extracted metadata of important information

[0196] Output: Information organized in a database

[0197] Step 4:

[0198] Accepts and analyzes questions from users.

[0199] The user enters a question through the interface of their smartphone or smart glasses, which then sends the question to the server.

[0200] Input: A question typed by the user in natural language (e.g., "What is the stock status of the latest smartphone accessories?")

[0201] Output: Question data sent to the server

[0202] Step 5:

[0203] Analyze the question and perform a database search.

[0204] The server analyzes the user's question using a question-answering model, converts it into an internal query, and searches for related information in the database. Related information is extracted as search results.

[0205] Input: Question data submitted by the user

[0206] Output: Relevant information extracted from the database

[0207] Step 6:

[0208] Generate and provide answers.

[0209] The server generates and summarizes answers based on the extracted relevant information, and the generated answers are sent to the terminal and displayed to the user.

[0210] Input: Relevant information extracted from the database

[0211] Output: A summary answer provided to the user (e.g., "The latest smartphone accessories are in good stock.")

[0212] In this way, the entire system works together to improve the work efficiency of store staff in physical stores and increase customer satisfaction.

[0213] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0214] ---

[0215] This system collects documents related to past projects, converts them into a unified format, and uses natural language processing technology to extract important information and store it in a database. It also provides an interface for users to input questions, analyzes the user's question, searches the database for relevant information, summarizes the information, and provides it to the user. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, providing more appropriate answers and expressions based on the user's emotions.

[0216] The program of this system mainly performs the following processes.

[0217] Data collection and integration

[0218] The server collects past documents from the company's network drive or cloud storage. The documents collected include meeting minutes, materials, and upper management reviews. Because these documents exist in various formats (PDF, Word, Excel, etc.), the server converts them into digital data using OCR technology and parsers and organizes them into a unified format.

[0219] Extracting information and storing it in a database

[0220] The server then uses natural language processing to analyze the documents converted into a unified format and extract important keywords and phrases. This information includes project status, decisions, issues, etc. The extracted information is then classified as metadata and stored in a relational or NoSQL database.

[0221] Building a question-answering system

[0222] The server provides the terminal with an interface for the user to input a question, which runs on the terminal and is designed to allow the user to input a question in natural language.

[0223] When a user types a question, the device sends it to the server, which uses machine learning models and natural language processing algorithms to analyze the question and convert it into an internal query, which then searches for relevant information in a database and extracts the best answer.

[0224] Emotion engine recognizes emotions and adjusts responses

[0225] When analyzing the user's question, the server uses an emotion engine to recognize the user's emotion, which can identify the user's emotional state (e.g., stress, excitement, relief, etc.).

[0226] Based on the emotional state identified by the emotion engine, the server can adjust the content and expression of the response to suit the user. For example, if the user is under stress, the server can change the response to be more polite and reassuring.

[0227] Generate and provide answers

[0228] The server generates a summary based on the information extracted from the database and creates a specific answer for the user, using the answer adjusted by the emotion engine. The generated answer is sent to the terminal and displayed to the user.

[0229] Specific examples

[0230] Example 1: Finding new campaign ideas

[0231] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0232] 2. The device sends this question to the server.

[0233] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0234] 4. The server uses an emotion engine to recognize the user's emotional state, for example, identifying that the user is distressed and therefore stressed.

[0235] 5. The server extracts relevant information from the database (such as examples of past campaign successes and failures, opinions of senior management, etc.) and generates a summary that takes into account the user's emotional state.

[0236] 6. The device displays the response from the server to the user, for example, "Project X is an example of a successful campaign in the past, and method Y was effective in this project. Please stay calm and proceed."

[0237] Example 2: Checking the progress of a project

[0238] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0239] 2. The device sends this question to the server.

[0240] 3. The server parses the question and extracts the latest information related to Project A from the database.

[0241] 4. The server uses an emotion engine to recognize the user's emotional state, for example, if the user is in a hurry, it generates a quick and concise answer.

[0242] 5. The server summarizes the extracted information and generates a summary that takes into account the user's emotional state, for example, "Project A is currently in stage 3, with key tasks 70% complete. The next step is X."

[0243] 6. The terminal displays the response from the server to the user.

[0244] In this way, we can provide information tailored to the user's emotional state, resulting in a more appropriate and user-friendly system that allows users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[0245] The processing flow will be explained below.

[0246] ---

[0247] Step 1:

[0248] The server collects historical documents from the company's network drives and cloud storage, including meeting minutes, materials, and senior management reviews.

[0249] Step 2:

[0250] The server converts these documents into text data using OCR technology, or if they are electronic files, they are imported as is.

[0251] Step 3:

[0252] The server converts the documents into a unified format (e.g., JSON) using parsers for various file formats (e.g., PDF, Word, Excel, etc.).

[0253] Step 4:

[0254] The server uses natural language processing techniques to extract important information and keywords from the converted documents, such as project status, decisions, and issues.

[0255] Step 5:

[0256] The server categorizes the extracted information as metadata and stores it in a relational or NoSQL database, organized by project and linking related information together.

[0257] Step 6:

[0258] The server provides a terminal with an interface for users to input questions, and this interface is designed to allow users to input questions in natural language.

[0259] Step 7:

[0260] The user inputs a question into the terminal. For example, "I'm having trouble with a new campaign. What should I do?"

[0261] Step 8:

[0262] The terminal sends the user's question to the server.

[0263] Step 9:

[0264] Once the server receives the question, it uses natural language processing techniques to analyze it, identifying important keywords and phrases and converting them into an internal query.

[0265] Step 10:

[0266] The server uses an emotion engine to recognize emotions from the user's questions, for example, identifying a user as being in a stressful state if they are worried.

[0267] Step 11:

[0268] Based on the query and the emotion recognition results, the server searches the database and extracts relevant information, such as examples of past campaign successes and failures, and the opinions of senior management.

[0269] Step 12:

[0270] The server generates an appropriate summary based on the extracted information, taking into account emotion recognition. For example, it generates a summary that uses gentle language and provides reassurance to users who are feeling stressed.

[0271] Step 13:

[0272] The server transmits the generated summary to the terminal.

[0273] Step 14:

[0274] The device displays the summary sent from the server to the user, such as, "As a concrete example of a past successful campaign, method Y was extremely effective in project X, so please consider it. You can proceed calmly, so I'm sure it will be fine."

[0275] ---

[0276] These are the specific processing steps of a system that includes an emotion engine. This process allows users to receive appropriate information and advice according to their emotional state, enabling them to carry out their work more effectively.

[0277] Example 2

[0278] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0279] Conventional systems have had difficulty efficiently collecting and analyzing information related to past projects within a company, extracting important information, and providing appropriate answers to user questions. Furthermore, they lacked the ability to recognize the user's emotional state and adjust the content and expression of the answers based on that state, making it difficult to provide user-friendly information. The present invention aims to solve these problems and provide a system that quickly and appropriately provides the information users desire.

[0280] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0281] In this invention, the server includes means for collecting documents related to past projects, means for converting the collected documents into a unified format, means for extracting important information using natural language processing technology, means for storing the extracted information in a database, means for analyzing a question from a user and searching the database for related information, means for summarizing the searched information and providing it to the user, means for providing an interface for the user to input a question, and emotion recognition means for recognizing the user's emotional state in response to the input question and adjusting the content and expression of the answer, thereby making it possible to provide information according to the user's emotions.

[0282] "Documents related to past projects" refers to documents in various formats, such as minutes, materials, and reviews related to past projects of a company or organization.

[0283] "Unified format" refers to a format that converts documents of different formats (PDF, Word, Excel, etc.) into one common format.

[0284] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[0285] "Database" refers to a system for efficiently storing, retrieving, and managing digital data.

[0286] "Question interface" refers to an input means that allows a user to input a question to the system in natural language.

[0287] "Emotion recognition means" refers to technology that recognizes a user's emotional state (stress, anxiety, excitement, etc.) from the user's questions and input content.

[0288] "Generative AI model" refers to an algorithm or model designed to help artificial intelligence perform a specific task (e.g., natural language parsing or generation).

[0289] A "prompt" is an instruction or question that is input to an AI model, and refers to the input sentence that enables the model to generate an appropriate response.

[0290] This system collects documents related to past projects, converts them into a unified format, and uses natural language processing technology to extract important information and store it in a database. It also provides an interface for users to input questions, analyzes the user's question, searches the database for relevant information, summarizes the information, and provides it to the user. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, providing more appropriate answers and expressions based on the user's emotions.

[0291] Data collection and integration

[0292] The server collects historical documents from the company's network drive and cloud storage. These documents include meeting minutes, handouts, and senior management reviews, and they exist in various formats, such as PDF, Word, and Excel. This creates a foundation for unifying information from diverse data sources. Specifically, the server uses OCR technology (e.g., Tesseract OCR) and parser libraries (e.g., Python's docx library and openpyxl library) to convert these documents into digital data and organize them into a unified format.

[0293] Extracting information and storing it in a database

[0294] The server analyzes the documents converted into a unified format using natural language processing techniques (e.g., Stanford NLP or spaCy) to extract important keywords and phrases. This information includes project progress, decisions, and issues. The extracted information is classified as metadata and stored in a relational database (e.g., MySQL (registered trademark), PostgreSQL) or a NoSQL database (e.g., MongoDB). This creates a database that allows quick access to the information needed.

[0295] Building a question-answering system

[0296] The server provides the device with an interface through which the user can input a question. This interface can be web-based or app-style and is designed to allow the user to input a question in natural language. Once the user inputs a question, the device sends the question to the server. The server then uses a machine learning model (e.g., GPT-3) or a natural language processing algorithm to analyze the question and convert it into an internal query. This allows the optimal answer to the user's question to be quickly generated.

[0297] Emotion engine recognizes emotions and adjusts responses

[0298] During the question analysis process, the server uses an emotion engine (e.g., Microsoft® Azure® Cognitive Services emotion analysis API) to recognize the user's emotions. Emotion recognition can identify the user's emotional state, such as stress, anxiety, or excitement. Based on the results of emotion recognition, the server can adjust the content and expression of the answer. For example, if the user is stressed, the server can make the answer more polite and add expressions that give a sense of security.

[0299] Generate and provide answers

[0300] The server generates a summary based on the information extracted from the database and creates a specific answer, including content adjusted by the emotion engine. The generated answer is sent to the device and displayed to the user, allowing the user to make effective decisions based on the displayed information.

[0301] Specific examples

[0302] Example 1: Finding new campaign ideas

[0303] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0304] 2. The device sends this question to the server.

[0305] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0306] 4. The server uses an emotion engine to determine that the user is experiencing stress due to their concerns.

[0307] 5. The server extracts relevant information from the database (such as examples of past campaign successes and failures, opinions of senior management, etc.) and generates a summary that takes into account the user's emotional state.

[0308] 6. The device displays the response from the server to the user, for example, "Project X is an example of a successful campaign in the past, and method Y was effective in this project. Please stay calm and proceed."

[0309] Example 2: Checking the progress of a project

[0310] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0311] 2. The device sends this question to the server.

[0312] 3. The server parses the question and extracts the latest information related to Project A from the database.

[0313] 4. The server uses an emotion engine to generate quick and concise answers if the user is in a hurry.

[0314] 5. The server summarizes the extracted information and generates a summary that takes into account the user's emotional state, for example, "Project A is currently in stage 3, with key tasks 70% complete. The next step is X."

[0315] 6. The terminal displays the response from the server to the user.

[0316] In this way, a more appropriate and user-friendly system can be realized by providing information according to the user's emotional state, allowing users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[0317] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0318] Step 1:

[0319] A user logs into a terminal and begins the process of gathering documents related to past projects.

[0320] Input: User credentials

[0321] Output: System access permissions

[0322] What happens: The user enters their username and password, which the system authenticates.

[0323] Step 2:

[0324] The server collects past documents from the company's network drive or cloud storage.

[0325] Input: Path information for network drive or cloud storage

[0326] Output: A list of collected documents

[0327] Specific operation: The server scans the specified path for PDF, Word, Excel, and other files and saves them in temporary storage.

[0328] Step 3:

[0329] The server converts the collected documents into a unified format.

[0330] Input: Collected documents

[0331] Output: Data converted into a unified format

[0332] Specific operation: For PDF files, it uses Tesseract OCR to convert them to text data, and for Word and Excel files, it uses Python's docx and openpyxl libraries to extract the text data.

[0333] Step 4:

[0334] The server uses natural language processing techniques to extract key information from the documents that are converted into a unified format.

[0335] Input: Data converted into a unified format

[0336] Output: Extracted keywords and phrases

[0337] What it does: Tokenizes and tags parts of speech using Stanford NLP and spaCy to extract keywords such as project status and decisions.

[0338] Step 5:

[0339] The server stores the extracted information in a database.

[0340] Input: Extracted keywords or phrases

[0341] Output: Metadata stored in a database

[0342] Specific behavior: Stores information in a relational database (MySQL, PostgreSQL) or a NoSQL database (MongoDB). Fields include "Project name," "Progress," and "Decisions."

[0343] Step 6:

[0344] The user inputs a question through the terminal.

[0345] Input: User question

[0346] Output: Question data sent to the terminal

[0347] Specific operation: The user enters a question in natural language and clicks the "Send" button. The device sends this question to the server.

[0348] Step 7:

[0349] The server analyzes the user's question and searches the database for relevant information.

[0350] Input: User question

[0351] Output: Inner query and search results

[0352] What it does: It uses a machine learning model (e.g., GPT-3) to analyze the question and generate an internal query, which it then uses to search for relevant information in a database.

[0353] Step 8:

[0354] The server recognizes the user's emotional state based on the search results.

[0355] Input: User questions and search results

[0356] Output: Determining the user's emotional state

[0357] Specific behavior: Uses an emotion engine (e.g., Microsoft Azure Cognitive Services emotion analysis API) to recognize the user's emotional state from the questions.

[0358] Step 9:

[0359] The server adjusts the content and expression of the response based on the results of the emotion engine.

[0360] Input: User's emotional state and search results

[0361] Output: Adjusted answer

[0362] Specific behavior: Depending on the user's emotional state, the response may be more polite or specific. For example, a user in a stressful state may be given a reassuring response.

[0363] Step 10:

[0364] The server generates the final answer to provide to the user.

[0365] Input: Adjusted answer content

[0366] Output: Final answer

[0367] Specific actions: Integrate search results with tailored content to generate summaries, flesh out answers, and format them to appropriately respond to the user's question.

[0368] Step 11:

[0369] The terminal displays the final answer to the user.

[0370] Input: Final answer

[0371] Output: The answer displayed to the user

[0372] Specific operation: The final answer sent from the server is received and displayed on the user's device, allowing the user to view it and use it as a reference for making decisions.

[0373] (Application example 2)

[0374] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0375] Conventional systems have had difficulty efficiently collecting and analyzing documents related to past projects and security-related logs, and providing relevant information to users. Furthermore, they lacked the ability to recognize the user's emotional state and adjust the appropriate response content and expressions accordingly, making it difficult to improve user satisfaction. Therefore, there was a need for a system that could efficiently collect and analyze information and provide appropriate information based on the user's emotional state.

[0376] The identification process by the identification 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 documents related to past projects, means for converting the collected documents into a unified format, means for extracting important information using natural language processing technology, means for storing the extracted information in a database, means for analyzing questions from users and searching the database for related information, means for summarizing the searched information and providing it to the user, means for providing an interface for users to input questions, and means for combining an emotion engine that recognizes the user's emotions and adjusts the content and expression of the response. This makes it possible to efficiently collect and analyze information related to past projects and security, and to provide appropriate information based on the user's emotional state.

[0377] "Documents related to past projects" are documents such as minutes, reports, and materials related to projects that a company or organization has carried out in the past.

[0378] A "uniform format" is a standardized format that converts collected documents into the same format to facilitate analysis and search.

[0379] "Natural language processing technology" is a technology that enables computers to understand and analyze human language, and extract meaningful information from text data.

[0380] "Key information extraction means" refers to methods and algorithms that use natural language processing techniques to identify and extract key information, such as project status, decisions, and issues, from collected documents.

[0381] A "database" is a system for storing and managing extracted information, a collection of data structured to facilitate search and access.

[0382] "Means for analyzing a question from a user and searching for related information from a database" refers to a method or algorithm for analyzing a question entered by a user and searching for information related to that question from a database.

[0383] "Means for summarizing retrieved information and providing it to the user" refers to a method or algorithm that succinctly summarizes relevant information extracted from a database and provides it to the user in an easily understandable format.

[0384] The "interface through which the user enters a question" is the part of the software that provides a screen or input field into which the user can enter a question in text form.

[0385] An "emotion engine" is a technology that analyzes the user's emotional state from the text they input and adjusts the response content and expression appropriately based on that.

[0386] "Security-related logs and event reports" are documents or data files that record security incidents and other security-related information within a company or organization.

[0387] A "generative AI model" is an artificial intelligence model that uses pre-trained algorithms to generate appropriate responses or prompts to user questions.

[0388] A "prompt sentence" is a specific question or instruction sentence that is generated to be easy for the user to understand.

[0389] This system collects documents related to past projects and security-related logs and event reports, converts them into a unified format, analyzes them, and provides appropriate answers to user questions. It also has the ability to recognize the user's emotional state and adjust the content and expression of the answers.

[0390] Data collection and integration

[0391] The server collects documents related to past projects and security-related logs from the company's network drive or cloud storage. Because these documents exist in various formats, such as PDF, Word, and image files, the server converts them into digital data using OCR technology (e.g., the pytesseract library) and parsers, and organizes them into a unified format.

[0392] Extracting information and storing it in a database

[0393] The server then uses natural language processing techniques (e.g., the transformers library) to analyze the documents converted into a unified format and extract important keywords and phrases. This information includes project status, decisions, issues, the date and time of security incidents, and the scope of their impact. The extracted information is classified as metadata and stored in a relational or NoSQL database.

[0394] Building a question-answering system

[0395] The server provides an interface on the terminal for users to input questions. This interface runs on terminals such as smartphones and PCs and is designed to allow users to input questions in natural language. When a question is input by the user, the terminal sends the question to the server.

[0396] Emotion engine recognizes emotions and adjusts responses

[0397] When analyzing a user's question, the server uses an emotion engine to recognize the user's emotions. This emotion recognition allows the server to identify the user's emotional state. For example, if the user is stressed, the server changes the answer to be more polite and reassuring.

[0398] Generate and provide answers

[0399] The server generates a summary based on the information extracted from the database and creates a specific response for the user, using the response adjusted by the emotion engine. The generated response is sent to the terminal and displayed to the user.

[0400] Specific examples

[0401] Example 1: Asking about the last security incident

[0402] A user types "When was the last security incident?" into a smartphone app. The app sends the question to the server, which analyzes the question, extracts relevant information from a database, and generates an answer. The app then displays the answer from the server and the emotion recognition results. For example, it might say, "The last security incident occurred on December 1, 2022. If you would like to know more about the countermeasures, please continue asking questions."

[0403] Prompt Sentence Examples

[0404] "Please provide a comprehensive analysis of the security logs and provide a detailed summary of the last security incident."

[0405] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0406] Step 1:

[0407] The server collects past project-related documents and security logs from the company's network drive or cloud storage. To collect documents, the server accesses pre-configured paths or API endpoints and downloads data in various formats, such as PDF, Word, Excel, and image files. This creates a document collection. The input is the network drive or cloud storage, and the output is a list of collected document files.

[0408] Step 2:

[0409] The server converts the collected documents into a unified format using OCR technology and a parser. Specifically, it uses the pytesseract library to extract text data from image files, and converts PDF and Word documents into text data using an appropriate parser. This converts documents in various formats into a unified text format. The input is the collected document files, and the output is the converted text data.

[0410] Step 3:

[0411] The server uses natural language processing technology to analyze the text data converted into a unified format and extract important keywords and phrases. At this stage, the transformers library is used to perform text analysis and generate metadata such as project progress, decisions, issues, and the date and time of the security incident and its scope of impact. The input is the converted text data, and the output is the extracted metadata.

[0412] Step 4:

[0413] The server stores the extracted metadata in a relational or NoSQL database. Specifically, it organizes and stores the metadata in tables or collections for future searches and queries. The input is the extracted metadata, and the output is the results stored in the database.

[0414] Step 5:

[0415] The terminal provides an interface for users to input questions. Users input questions in natural language on the screen of their smartphone or PC, and the questions are sent to the server via the interface. The input is the user's text question, and the output is the result of sending the question data to the server.

[0416] Step 6:

[0417] The server analyzes the received user question using natural language processing technology and an emotion engine. Through the analysis, it identifies keywords in the question and the user's emotional state and converts them into an appropriate internal query. The input is the user's question data, and the output is the analysis result and the internal query.

[0418] Step 7:

[0419] The server uses the internal query to search for relevant information from a database. The search extracts appropriate information for the question and tailors the answer to the user's emotional state. The input is the internal query, and the output is the search results and tailored answer data.

[0420] Step 8:

[0421] The server uses a generative AI model to summarize and optimize answers to complete a specific answer for the user. It generates prompt sentences and provides answers that take the user's emotional state into account. The input is search results and adjusted answer data, and the output is a completed answer. Using a generative AI model improves the quality and appropriateness of answers.

[0422] Step 9:

[0423] The terminal displays the answers received from the server to the user. The user can check the detailed answers on the screen and use this interface to enter additional questions. The input is the answer data from the server, and the output is the answer display result to the user.

[0424] Example prompt sentence:

[0425] "Please provide a comprehensive analysis of the security logs and provide a detailed summary of the last security incident."

[0426] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0427] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0428] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0429] [Second embodiment]

[0430] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0431] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0432] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0433] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0434] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0435] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0436] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0437] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0438] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0439] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0440] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0441] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0442] ---

[0443] The present invention is a system that collects documents related to past projects, converts them into a unified format, extracts important information using natural language processing technology, and stores it in a database. It also provides an interface for users to input questions, analyzes the user's questions, searches the database for relevant information, and summarizes the information for the user.

[0444] The program of this system mainly performs the following processes.

[0445] Data collection and integration

[0446] The server collects past documents from the company's network drive or cloud storage. The documents collected include meeting minutes, materials, and upper management reviews. These documents exist in various formats (PDF, Word, Excel, etc.), so the server converts them into digital data using OCR technology and parsers and organizes them into a unified format.

[0447] Extracting information and storing it in a database

[0448] The server then uses natural language processing to analyze the documents converted into a unified format and extract important keywords and phrases, such as project status, decisions, and issues. The extracted information is then classified as metadata and stored in a relational or NoSQL database.

[0449] Building a question-answering system

[0450] The server provides an interface for users to enter questions, which runs on a terminal and allows users to enter questions in natural language.

[0451] When a user types a question, the device sends it to the server, which uses machine learning models and natural language processing algorithms to analyze the question and convert it into an internal query, which then searches for relevant information in a database and extracts the best answer.

[0452] Generate and provide answers

[0453] The server then summarizes the information extracted from the database and generates a specific answer for the user, which is then sent to the terminal and displayed to the user. This process allows users to quickly obtain important information related to past projects, greatly improving work efficiency.

[0454] Specific examples

[0455] Example 1: Finding new campaign ideas

[0456] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0457] 2. The device sends this question to the server.

[0458] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0459] 4. The server extracts and summarizes relevant information from the database (such as past campaign successes and failures, opinions from upper management, etc.) and provides appropriate advice to the user.

[0460] 5. The terminal displays the response from the server to the user.

[0461] Example 2: Checking the progress of a project

[0462] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0463] 2. The device sends this question to the server.

[0464] 3. The server parses the question and extracts the latest information related to Project A from the database.

[0465] 4. The server summarizes the extracted information and provides the user with the current status of the project, any issues, next steps, etc.

[0466] 5. The device displays the server's response to the user, showing specific progress, for example, "Project A is currently in stage 3, with major tasks 70% complete. The next step is X."

[0467] In this way, the system of the present invention makes it possible to quickly and efficiently refer to detailed data on past projects, which greatly contributes to the planning of new projects and the efficiency of business operations.

[0468] The processing flow will be explained below.

[0469] ---

[0470] Step 1:

[0471] The server accesses the company's network drive or cloud storage to collect documents related to past projects, such as meeting minutes, materials, and senior management reviews.

[0472] Step 2:

[0473] The server converts the collected documents into text data using OCR technology, or if the documents are electronic files, they are imported as is.

[0474] Step 3:

[0475] The server converts documents in different formats into a unified format (e.g., JSON), using an appropriate parser for each file format.

[0476] Step 4:

[0477] The server uses natural language processing techniques to extract key information and keywords from the converted documents, including project status, decisions, and issues.

[0478] Step 5:

[0479] The server classifies the extracted information as metadata and stores it in a relational or NoSQL database. The data is organized by project, and highly related information is linked together.

[0480] Step 6:

[0481] The server provides the terminal with an interface for the user to input a question, the interface being designed to allow the user to input a question in natural language.

[0482] Step 7:

[0483] The user inputs a question into the terminal. For example, "I'm having trouble with a new campaign. What should I do?"

[0484] Step 8:

[0485] The terminal sends the user's question to the server.

[0486] Step 9:

[0487] The server receives the question and analyzes it using natural language processing techniques, identifying important keywords and phrases from the question and converting them into an internal query.

[0488] Step 10:

[0489] The server uses the internal query to search the database and extract information relevant to the question.

[0490] Step 11:

[0491] The server generates summaries based on the extracted information, such as examples of successful past campaigns or summaries of senior management's opinions.

[0492] Step 12:

[0493] The server transmits the generated summary to the terminal.

[0494] Step 13:

[0495] The device displays the summary sent from the server to the user, providing specific advice such as, "Project X is an example of a past successful campaign, and method Y was effective in this project."

[0496] ---

[0497] These are the specific steps of the program's processing, which allow users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[0498] Example 1

[0499] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0500] Because information related to past business projects is scattered, it is difficult to quickly and efficiently obtain the necessary information. In addition, because this information is stored in various formats, it is difficult to manage and search it in a unified manner. This inefficient management and search of information is a factor that reduces business efficiency.

[0501] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0502] In this invention, the server includes a means for collecting materials related to past business projects, a means for converting the collected materials into a unified format, and a means for extracting important information using machine learning technology. This makes it possible to efficiently collect information related to past business projects, convert it into a unified format, and extract important information. It can also analyze questions from users, search for related information from an information storage device, and provide a summary of the searched information, allowing users to quickly obtain the information they need.

[0503] "Materials" refers to documents and data related to past business projects, including minutes, reports, plans, emails, etc.

[0504] A "unified format" is a data format for converting collected materials into a consistent format, which makes it possible to handle materials in different formats in a uniform manner.

[0505] "Machine learning technology" is a technology for analyzing large amounts of data and automatically learning and making predictions, and includes algorithm-based analysis methods and deep learning technology.

[0506] "Information storage device" refers to a storage system, such as a database or cloud storage, for storing extracted information, including relational databases and NoSQL databases.

[0507] "Character recognition technology" means technology for converting printed or handwritten characters into digital data, including optical character recognition (OCR) technology.

[0508] A "natural language processing model" is a model for understanding and analyzing human language, and includes distributed representation models (Word2Vec) and translation models (BERT, GPT-3).

[0509] A "high-speed search engine" is an engine for quickly searching for necessary information from large amounts of data, and includes full-text search engines (Elasticsearch, Solr).

[0510] "Internal query" refers to a database query that is generated by analyzing a user's natural language question and searching for relevant information.

[0511] "Input device" refers to a hardware or software interface through which a user can input queries or data, including keyboards, touchscreens, and voice input devices.

[0512] This system efficiently collects materials related to past business projects, extracts important information, stores it in a database, and promptly provides relevant information in response to user questions. This system is composed of the following entities: a server, a terminal, and a user.

[0513] Data collection and integration

[0514] The server collects business-related materials from the company's network drive or cloud storage (e.g., a general cloud storage service). This includes meeting minutes, reports, plans, emails, etc. The server accesses the cloud storage using an appropriate API and retrieves files from the specified target directory. The collected materials are broad and often relate to the entire company's activities.

[0515] Unified data format conversion

[0516] The server uses open-source character recognition tools (e.g., Tesseract) to convert collected materials into digital data using optical character recognition (OCR) and document parsers. For example, OCR is performed on PDF files to extract text, and common document parsing libraries (e.g., Apache POI) are used for Word and Excel files. The converted data is then formatted into a consistent format (e.g., JSON, XML).

[0517] Information Extraction

[0518] The server then analyzes the unified formatted material using natural language processing (NLP) techniques, such as open-source NLP libraries (e.g., SpaCy, NLTK) and advanced machine learning models (e.g., BERT, GPT-3), to extract important noun phrases and key phrases from the document and generate a summary of the entire document.

[0519] Storage in the database

[0520] The server stores the extracted information in an information storage device. This information storage device can be a relational database (e.g., a general relational database management system) or a NoSQL database (e.g., a general NoSQL database solution). A table containing progress and risk factors for each project is designed and the information is stored there. An index is also created using a full-text search engine (e.g., Elasticsearch) to enable fast searches.

[0521] Receiving and analyzing user questions

[0522] A user inputs a question through an interface provided on the device. For example, "What is the latest progress on project A?" The device then sends this question to the server. The server analyzes the received question using natural language processing techniques and machine learning models. For example, it uses sentiment analysis and intent classification to identify the intent of the question and generates a related query. The internal query is used to search for related information.

[0523] Database search and answer generation

[0524] The server uses the parsed query to search for information in the information storage device. For example, a query like "latest progress on Project A" might retrieve the latest updates from the database. Elasticsearch might also be used to quickly find relevant documents. The server then summarizes the information from the database and generates a specific answer for the user. For example, "Project A is currently in Stage 3, with 70% of key tasks completed. The next step is X." This answer is then sent to the device and displayed to the user.

[0525] Specific examples

[0526] Example 1: Finding new campaign ideas

[0527] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0528] 2. The device sends this question to the server.

[0529] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0530] 4. The server extracts and summarizes relevant information from the information storage device (such as examples of past campaign successes and failures, opinions of upper management, etc.) and provides appropriate advice to the user.

[0531] 5. The terminal displays the response from the server to the user.

[0532] Example 2: Checking the progress of a project

[0533] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0534] 2. The device sends this question to the server.

[0535] 3. The server analyzes the question and extracts the latest information related to Project A from the information repository.

[0536] 4. The server summarizes the extracted information and provides the user with the current status of the project, any issues, next steps, etc.

[0537] 5. The device displays the server's response to the user, showing specific progress, for example, "Project A is currently in stage 3, with major tasks 70% complete. The next step is X."

[0538] In this way, the system of the present invention allows for quick and efficient reference to detailed data related to past business projects, greatly contributing to the planning of new projects and the efficiency of business operations.

[0539] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0540] Step 1: Collect data

[0541] The server collects business-related materials from the company's network drive or cloud storage (e.g., a general cloud storage service). The server accesses the cloud storage using the appropriate API and retrieves files from the specified target directory. Inputs include the cloud storage API endpoint, authentication information, and the specified directory, and the collected raw material files are included as output. Specifically, the server retrieves a list of files and downloads each file in the list.

[0542] Step 2: Convert data into a unified format

[0543] The server converts the collected data into digital data using optical character recognition (OCR) technology and a document parser. The input is raw data files such as PDF, Word, and Excel, and the output is text data. For example, Tesseract OCR is used to extract text data from PDF files, and Apache POI is used to extract text data from Word and Excel files. The extracted text data is then formatted into a consistent, unified format (e.g., JSON or XML).

[0544] Step 3: Extracting information

[0545] The server analyzes the unified formatted data using natural language processing technology. The input is the unified formatted text data, and the output is information such as extracted important noun phrases and key phrases. Specifically, it uses SpaCy and NLTK to tokenize the document and extract important keywords and phrases. It also uses BERT and GPT-3 models to generate a summary of the entire document.

[0546] Step 4: Store in the database

[0547] The server stores the extracted information in an information storage device. The input is the extracted key information and summary data, and the output is the entries stored in a database. Specifically, it inserts the information into a relational database (e.g., a typical relational database management system) or a NoSQL database. In addition, it uses Elasticsearch to create an index, enabling fast searches.

[0548] Step 5: Receiving user questions

[0549] The user inputs a question through an interface provided on the terminal. The input is a natural language question typed by the user, and the output is the question sent to the server. For example, the user might input, "What is the latest progress on project A?" The terminal receives this question and sends it to the server.

[0550] Step 6: Parsing the Question

[0551] The server analyzes the received question using natural language processing technology and machine learning models. The input is the question submitted by the user, and the output is a related internal query. Specifically, it uses Sentiment Analysis and Intent Classification to identify the intent of the question, and keyphrase extraction to extract important keywords. The internal query is generated in SQL or Elasticsearch query language.

[0552] Step 7: Database Search

[0553] The server uses the parsed query to search for information in the information storage device. The input is the generated internal query, and the output is the search results. Specifically, it uses a relational database or Elasticsearch to retrieve related information, such as the latest update records and related documents.

[0554] Step 8: Generate and provide answers

[0555] The server summarizes the information retrieved from the database and generates a specific answer. The input is the search result data, and the output includes summary information provided to the user. For example, it generates content such as "Project A is currently in stage 3, with major tasks 70% complete. The next step is X." The server then sends the generated answer to the device and displays it to the user.

[0556] (Application example 1)

[0557] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0558] In recent years, efficient customer service and sales operations have become increasingly important in brick-and-mortar stores. However, store staff face challenges in quickly searching and retrieving vast amounts of past customer service history and sales data. Furthermore, they are often unable to provide prompt and appropriate responses to customer inquiries. This can lead to lower customer satisfaction and a deterioration in store operational efficiency. A system that can solve these problems is needed.

[0559] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0560] In this invention, the server includes a means for collecting documents related to past projects, a means for converting the collected documents into a unified format, and a means for extracting important information using natural language processing technology, which enables store staff to search and obtain past customer interaction history and sales data, and enables the server to quickly generate appropriate answers to store staff questions using a question-answering model.

[0561] "Documents related to past projects" refers to documents such as minutes, materials, and reviews related to projects that a company or organization has carried out in the past.

[0562] "Unified format" refers to converting documents that exist in various formats into a consistent format that makes them easier to parse and process.

[0563] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and respond to human language.

[0564] A "database" is a system that stores data in a structured way so that it can be efficiently searched and retrieved.

[0565] The "interface through which the user inputs questions" refers to the functions and screens that allow the user to input questions to the system in natural language.

[0566] "A means for store staff to search and obtain past customer interaction history and sales data" refers to a function that enables store staff to quickly find and use information related to past customer interactions and sales performance.

[0567] A "question answering model" refers to a machine learning model that generates appropriate answers to questions from users.

[0568] "Customer interaction history" is data that records past interactions with customers.

[0569] "Sales data" refers to data containing information about past sales activities, such as sales volume, timing, and sales amount.

[0570] MODE FOR CARRYING OUT THE INVENTION

[0571] This invention is a system that allows store staff in physical stores to quickly search and obtain past customer interaction history and sales data, and provides appropriate answers using a question-answering model.

[0572] 1. System Configuration

[0573] The system collects documents related to past projects, converts them into a unified format, extracts important information using natural language processing technology, and stores it in a database. It provides an interface for users to input questions, which then searches for and summarizes relevant information from the database and provides it to the user. A specific example is shown below.

[0574] Hardware and Software

[0575] The hardware and software used are as follows:

[0576] Flask: This Python web framework builds an API server to accept and process user queries.

[0577] transformers: A natural language processing library by Hugging Face. This library provides a question-answering model.

[0578] JSON: A data storage format that is easily used as a database.

[0579] 2. Program processing explanation

[0580] The server uses Flask to build an API server and accepts questions from users. Questions are sent in JSON format and processed within the server. The server uses natural language processing technology to analyze the user's question and search for information in a database. It uses the Hugging Face pipeline as a question-answering model to generate an appropriate answer using the question and related information in the database.

[0581] The server searches and retrieves information from the database, generates a response based on this information, and returns it to the user, allowing the user to quickly obtain the information they need.

[0582] Examples of concrete examples and prompts

[0583] For example, if a store employee asks, "What's the stock status of the latest smartphone accessories?", the question is sent to the server. The server searches past inventory and sales data and analyzes the information using natural language processing technology. Finally, the question-answering model generates an appropriate answer, which is displayed on the device.

[0584] The specific prompt is as follows:

[0585] Please let me know the latest smartphone accessories in stock.

[0586] In this way, the system of the present invention allows store staff to quickly and efficiently search and retrieve detailed customer interaction history and sales data, enabling them to provide appropriate responses to customers, thereby improving business efficiency and customer satisfaction.

[0587] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0588] Step 1:

[0589] Collect historical documents and convert them into a unified format.

[0590] The server collects past documents (meeting minutes, materials, upper management reviews, etc.) from the company's network drive or cloud storage. The collected documents (PDF, Word, Excel, etc.) are converted into digital data using OCR technology and parsers and organized into a unified format.

[0591] Input: Corporate historical documents in various formats

[0592] Output: Digital data converted into a unified format

[0593] Step 2:

[0594] Analyze the data and extract important information.

[0595] The server uses natural language processing technology to analyze the documents converted into a unified format and extract important keywords and phrases such as project progress, decisions, issues, etc. The extracted information is classified as metadata.

[0596] Input: Digital data converted into a unified format

[0597] Output: Metadata of extracted important information

[0598] Step 3:

[0599] Stores the data in the database.

[0600] The server then stores the extracted metadata of the key information in a relational or NoSQL database, which also contains data from other similar projects and organizes the information by relevance.

[0601] Input: Extracted metadata of important information

[0602] Output: Information organized in a database

[0603] Step 4:

[0604] Accepts and analyzes questions from users.

[0605] The user enters a question through the interface of their smartphone or smart glasses, which then sends the question to the server.

[0606] Input: A question typed by the user in natural language (e.g., "What is the stock status of the latest smartphone accessories?")

[0607] Output: Question data sent to the server

[0608] Step 5:

[0609] Analyze the question and perform a database search.

[0610] The server analyzes the user's question using a question-answering model, converts it into an internal query, and searches for related information in the database. Related information is extracted as search results.

[0611] Input: Question data submitted by the user

[0612] Output: Relevant information extracted from the database

[0613] Step 6:

[0614] Generate and provide answers.

[0615] The server generates and summarizes answers based on the extracted relevant information, and the generated answers are sent to the terminal and displayed to the user.

[0616] Input: Relevant information extracted from the database

[0617] Output: A summary answer provided to the user (e.g., "The latest smartphone accessories are in good stock.")

[0618] In this way, the entire system works together to improve the work efficiency of store staff in physical stores and increase customer satisfaction.

[0619] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0620] ---

[0621] This system collects documents related to past projects, converts them into a unified format, and uses natural language processing technology to extract important information and store it in a database. It also provides an interface for users to input questions, analyzes the user's question, searches the database for relevant information, summarizes the information, and provides it to the user. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, providing more appropriate answers and expressions based on the user's emotions.

[0622] The program of this system mainly performs the following processes.

[0623] Data collection and integration

[0624] The server collects past documents from the company's network drive or cloud storage. The documents collected include meeting minutes, materials, and upper management reviews. Because these documents exist in various formats (PDF, Word, Excel, etc.), the server converts them into digital data using OCR technology and parsers and organizes them into a unified format.

[0625] Extracting information and storing it in a database

[0626] The server then uses natural language processing to analyze the documents converted into a unified format and extract important keywords and phrases. This information includes project status, decisions, issues, etc. The extracted information is then classified as metadata and stored in a relational or NoSQL database.

[0627] Building a question-answering system

[0628] The server provides the terminal with an interface for the user to input a question, which runs on the terminal and is designed to allow the user to input a question in natural language.

[0629] When a user types a question, the device sends it to the server, which uses machine learning models and natural language processing algorithms to analyze the question and convert it into an internal query, which then searches for relevant information in a database and extracts the best answer.

[0630] Emotion engine recognizes emotions and adjusts responses

[0631] When analyzing the user's question, the server uses an emotion engine to recognize the user's emotion, which can identify the user's emotional state (e.g., stress, excitement, relief, etc.).

[0632] Based on the emotional state identified by the emotion engine, the server can adjust the content and expression of the response to suit the user. For example, if the user is under stress, the server can change the response to be more polite and reassuring.

[0633] Generate and provide answers

[0634] The server generates a summary based on the information extracted from the database and creates a specific answer for the user, using the answer adjusted by the emotion engine. The generated answer is sent to the terminal and displayed to the user.

[0635] Specific examples

[0636] Example 1: Finding new campaign ideas

[0637] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0638] 2. The device sends this question to the server.

[0639] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0640] 4. The server uses an emotion engine to recognize the user's emotional state, for example, identifying that the user is distressed and therefore stressed.

[0641] 5. The server extracts relevant information from the database (such as examples of past campaign successes and failures, opinions of senior management, etc.) and generates a summary that takes into account the user's emotional state.

[0642] 6. The device displays the response from the server to the user, for example, "Project X is an example of a successful campaign in the past, and method Y was effective in this project. Please stay calm and proceed."

[0643] Example 2: Checking the progress of a project

[0644] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0645] 2. The device sends this question to the server.

[0646] 3. The server parses the question and extracts the latest information related to Project A from the database.

[0647] 4. The server uses an emotion engine to recognize the user's emotional state, for example, if the user is in a hurry, it generates a quick and concise answer.

[0648] 5. The server summarizes the extracted information and generates a summary that takes into account the user's emotional state, for example, "Project A is currently in stage 3, with key tasks 70% complete. The next step is X."

[0649] 6. The terminal displays the response from the server to the user.

[0650] In this way, we can provide information tailored to the user's emotional state, resulting in a more appropriate and user-friendly system that allows users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[0651] The processing flow will be explained below.

[0652] ---

[0653] Step 1:

[0654] The server collects historical documents from the company's network drives and cloud storage, including meeting minutes, materials, and senior management reviews.

[0655] Step 2:

[0656] The server converts these documents into text data using OCR technology, or if they are electronic files, they are imported as is.

[0657] Step 3:

[0658] The server converts the documents into a unified format (e.g., JSON) using parsers for various file formats (e.g., PDF, Word, Excel, etc.).

[0659] Step 4:

[0660] The server uses natural language processing techniques to extract important information and keywords from the converted documents, such as project status, decisions, and issues.

[0661] Step 5:

[0662] The server categorizes the extracted information as metadata and stores it in a relational or NoSQL database, organized by project and linking related information together.

[0663] Step 6:

[0664] The server provides a terminal with an interface for users to input questions, and this interface is designed to allow users to input questions in natural language.

[0665] Step 7:

[0666] The user inputs a question into the terminal. For example, "I'm having trouble with a new campaign. What should I do?"

[0667] Step 8:

[0668] The terminal sends the user's question to the server.

[0669] Step 9:

[0670] Once the server receives the question, it uses natural language processing techniques to analyze it, identifying important keywords and phrases and converting them into an internal query.

[0671] Step 10:

[0672] The server uses an emotion engine to recognize emotions from the user's questions, for example, identifying a user as being in a stressful state if they are worried.

[0673] Step 11:

[0674] Based on the query and the emotion recognition results, the server searches the database and extracts relevant information, such as examples of past campaign successes and failures, and the opinions of senior management.

[0675] Step 12:

[0676] The server generates an appropriate summary based on the extracted information, taking into account emotion recognition. For example, it generates a summary that uses gentle language and provides reassurance to users who are feeling stressed.

[0677] Step 13:

[0678] The server transmits the generated summary to the terminal.

[0679] Step 14:

[0680] The device displays the summary sent from the server to the user, such as, "As a concrete example of a past successful campaign, method Y was extremely effective in project X, so please consider it. You can proceed calmly, so I'm sure it will be fine."

[0681] ---

[0682] These are the specific processing steps of a system that includes an emotion engine. This process allows users to receive appropriate information and advice according to their emotional state, enabling them to carry out their work more effectively.

[0683] Example 2

[0684] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0685] Conventional systems have had difficulty efficiently collecting and analyzing information related to past projects within a company, extracting important information, and providing appropriate answers to user questions. Furthermore, they lacked the ability to recognize the user's emotional state and adjust the content and expression of the answers based on that state, making it difficult to provide user-friendly information. The present invention aims to solve these problems and provide a system that quickly and appropriately provides the information users desire.

[0686] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0687] In this invention, the server includes means for collecting documents related to past projects, means for converting the collected documents into a unified format, means for extracting important information using natural language processing technology, means for storing the extracted information in a database, means for analyzing a question from a user and searching the database for related information, means for summarizing the searched information and providing it to the user, means for providing an interface for the user to input a question, and emotion recognition means for recognizing the user's emotional state in response to the input question and adjusting the content and expression of the answer, thereby making it possible to provide information according to the user's emotions.

[0688] "Documents related to past projects" refers to documents in various formats, such as minutes, materials, and reviews related to past projects of a company or organization.

[0689] "Unified format" refers to a format that converts documents of different formats (PDF, Word, Excel, etc.) into one common format.

[0690] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[0691] "Database" refers to a system for efficiently storing, retrieving, and managing digital data.

[0692] "Question interface" refers to an input means that allows a user to input a question to the system in natural language.

[0693] "Emotion recognition means" refers to technology that recognizes a user's emotional state (stress, anxiety, excitement, etc.) from the user's questions and input content.

[0694] "Generative AI model" refers to an algorithm or model designed to help artificial intelligence perform a specific task (e.g., natural language parsing or generation).

[0695] A "prompt" is an instruction or question that is input to an AI model, and refers to the input sentence that enables the model to generate an appropriate response.

[0696] This system collects documents related to past projects, converts them into a unified format, and uses natural language processing technology to extract important information and store it in a database. It also provides an interface for users to input questions, analyzes the user's question, searches the database for relevant information, summarizes the information, and provides it to the user. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, providing more appropriate answers and expressions based on the user's emotions.

[0697] Data collection and integration

[0698] The server collects historical documents from the company's network drive and cloud storage. These documents include meeting minutes, handouts, and senior management reviews, and they exist in various formats, such as PDF, Word, and Excel. This creates a foundation for unifying information from diverse data sources. Specifically, the server uses OCR technology (e.g., Tesseract OCR) and parser libraries (e.g., Python's docx library and openpyxl library) to convert these documents into digital data and organize them into a unified format.

[0699] Extracting information and storing it in a database

[0700] The server analyzes the documents converted into a unified format using natural language processing techniques (e.g., Stanford NLP or spaCy) to extract important keywords and phrases. This information includes project status, decisions, and issues. The extracted information is classified as metadata and stored in a relational database (e.g., MySQL, PostgreSQL) or a NoSQL database (e.g., MongoDB). This creates a database that allows quick access to the information needed.

[0701] Building a question-answering system

[0702] The server provides the device with an interface through which the user can input a question. This interface can be web-based or app-style and is designed to allow the user to input a question in natural language. Once the user inputs a question, the device sends the question to the server. The server then uses a machine learning model (e.g., GPT-3) or a natural language processing algorithm to analyze the question and convert it into an internal query. This allows the optimal answer to the user's question to be quickly generated.

[0703] Emotion engine recognizes emotions and adjusts responses

[0704] During the question analysis process, the server uses an emotion engine (e.g., Microsoft Azure Cognitive Services' Emotion Analysis API) to recognize the user's emotions. Emotion recognition can identify the user's emotional state, such as stress, anxiety, or excitement. Based on the results of emotion recognition, the server can adjust the content and expression of the answer. For example, if the user is stressed, the server can make the answer more polite and add expressions that give a sense of security.

[0705] Generate and provide answers

[0706] The server generates a summary based on the information extracted from the database and creates a specific answer, including content adjusted by the emotion engine. The generated answer is sent to the device and displayed to the user, allowing the user to make effective decisions based on the displayed information.

[0707] Specific examples

[0708] Example 1: Finding new campaign ideas

[0709] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0710] 2. The device sends this question to the server.

[0711] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0712] 4. The server uses an emotion engine to determine that the user is experiencing stress due to their concerns.

[0713] 5. The server extracts relevant information from the database (such as examples of past campaign successes and failures, opinions of senior management, etc.) and generates a summary that takes into account the user's emotional state.

[0714] 6. The device displays the response from the server to the user, for example, "Project X is an example of a successful campaign in the past, and method Y was effective in this project. Please stay calm and proceed."

[0715] Example 2: Checking the progress of a project

[0716] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0717] 2. The device sends this question to the server.

[0718] 3. The server parses the question and extracts the latest information related to Project A from the database.

[0719] 4. The server uses an emotion engine to generate quick and concise answers if the user is in a hurry.

[0720] 5. The server summarizes the extracted information and generates a summary that takes into account the user's emotional state, for example, "Project A is currently in stage 3, with key tasks 70% complete. The next step is X."

[0721] 6. The terminal displays the response from the server to the user.

[0722] In this way, a more appropriate and user-friendly system can be realized by providing information according to the user's emotional state, allowing users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[0723] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0724] Step 1:

[0725] A user logs into a terminal and begins the process of gathering documents related to past projects.

[0726] Input: User credentials

[0727] Output: System access permissions

[0728] What happens: The user enters their username and password, which the system authenticates.

[0729] Step 2:

[0730] The server collects past documents from the company's network drive or cloud storage.

[0731] Input: Path information for network drive or cloud storage

[0732] Output: A list of collected documents

[0733] Specific operation: The server scans the specified path for PDF, Word, Excel, and other files and saves them in temporary storage.

[0734] Step 3:

[0735] The server converts the collected documents into a unified format.

[0736] Input: Collected documents

[0737] Output: Data converted into a unified format

[0738] Specific operation: For PDF files, it uses Tesseract OCR to convert them to text data, and for Word and Excel files, it uses Python's docx and openpyxl libraries to extract the text data.

[0739] Step 4:

[0740] The server uses natural language processing techniques to extract key information from the documents that are converted into a unified format.

[0741] Input: Data converted into a unified format

[0742] Output: Extracted keywords and phrases

[0743] What it does: Tokenizes and tags parts of speech using Stanford NLP and spaCy to extract keywords such as project status and decisions.

[0744] Step 5:

[0745] The server stores the extracted information in a database.

[0746] Input: Extracted keywords or phrases

[0747] Output: Metadata stored in a database

[0748] Specific behavior: Stores information in a relational database (MySQL, PostgreSQL) or a NoSQL database (MongoDB). Fields include "Project name," "Progress," and "Decisions."

[0749] Step 6:

[0750] The user inputs a question through the terminal.

[0751] Input: User question

[0752] Output: Question data sent to the terminal

[0753] Specific operation: The user enters a question in natural language and clicks the "Send" button. The device sends this question to the server.

[0754] Step 7:

[0755] The server analyzes the user's question and searches the database for relevant information.

[0756] Input: User question

[0757] Output: Inner query and search results

[0758] What it does: It uses a machine learning model (e.g., GPT-3) to analyze the question and generate an internal query, which it then uses to search for relevant information in a database.

[0759] Step 8:

[0760] The server recognizes the user's emotional state based on the search results.

[0761] Input: User questions and search results

[0762] Output: Determining the user's emotional state

[0763] Specific behavior: Uses an emotion engine (e.g., Microsoft Azure Cognitive Services emotion analysis API) to recognize the user's emotional state from the questions.

[0764] Step 9:

[0765] The server adjusts the content and expression of the response based on the results of the emotion engine.

[0766] Input: User's emotional state and search results

[0767] Output: Adjusted answer

[0768] Specific behavior: Depending on the user's emotional state, the response may be more polite or specific. For example, a user in a stressful state may be given a reassuring response.

[0769] Step 10:

[0770] The server generates the final answer to provide to the user.

[0771] Input: Adjusted answer content

[0772] Output: Final answer

[0773] Specific actions: Integrate search results with tailored content to generate summaries, flesh out answers, and format them to appropriately respond to the user's question.

[0774] Step 11:

[0775] The terminal displays the final answer to the user.

[0776] Input: Final answer

[0777] Output: The answer displayed to the user

[0778] Specific operation: The final answer sent from the server is received and displayed on the user's device, allowing the user to view it and use it as a reference for making decisions.

[0779] (Application example 2)

[0780] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0781] Conventional systems have had difficulty efficiently collecting and analyzing documents related to past projects and security-related logs, and providing relevant information to users. Furthermore, they lacked the ability to recognize the user's emotional state and adjust the appropriate response content and expressions accordingly, making it difficult to improve user satisfaction. Therefore, there was a need for a system that could efficiently collect and analyze information and provide appropriate information based on the user's emotional state.

[0782] The identification process by the identification 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 documents related to past projects, means for converting the collected documents into a unified format, means for extracting important information using natural language processing technology, means for storing the extracted information in a database, means for analyzing questions from users and searching the database for related information, means for summarizing the searched information and providing it to the user, means for providing an interface for users to input questions, and means for combining an emotion engine that recognizes the user's emotions and adjusts the content and expression of the response. This makes it possible to efficiently collect and analyze information related to past projects and security, and to provide appropriate information based on the user's emotional state.

[0783] "Documents related to past projects" are documents such as minutes, reports, and materials related to projects that a company or organization has carried out in the past.

[0784] A "uniform format" is a standardized format that converts collected documents into the same format to facilitate analysis and search.

[0785] "Natural language processing technology" is a technology that enables computers to understand and analyze human language, and extract meaningful information from text data.

[0786] "Key information extraction means" refers to methods and algorithms that use natural language processing techniques to identify and extract key information, such as project status, decisions, and issues, from collected documents.

[0787] A "database" is a system for storing and managing extracted information, a collection of data structured to facilitate search and access.

[0788] "Means for analyzing a question from a user and searching for related information from a database" refers to a method or algorithm for analyzing a question entered by a user and searching for information related to that question from a database.

[0789] "Means for summarizing retrieved information and providing it to the user" refers to a method or algorithm that succinctly summarizes relevant information extracted from a database and provides it to the user in an easily understandable format.

[0790] The "interface through which the user enters a question" is the part of the software that provides a screen or input field into which the user can enter a question in text form.

[0791] An "emotion engine" is a technology that analyzes the user's emotional state from the text they input and adjusts the response content and expression appropriately based on that.

[0792] "Security-related logs and event reports" are documents or data files that record security incidents and other security-related information within a company or organization.

[0793] A "generative AI model" is an artificial intelligence model that uses pre-trained algorithms to generate appropriate responses or prompts to user questions.

[0794] A "prompt sentence" is a specific question or instruction sentence that is generated to be easy for the user to understand.

[0795] This system collects documents related to past projects and security-related logs and event reports, converts them into a unified format, analyzes them, and provides appropriate answers to user questions. It also has the ability to recognize the user's emotional state and adjust the content and expression of the answers.

[0796] Data collection and integration

[0797] The server collects documents related to past projects and security-related logs from the company's network drive or cloud storage. Because these documents exist in various formats, such as PDF, Word, and image files, the server converts them into digital data using OCR technology (e.g., the pytesseract library) and parsers, and organizes them into a unified format.

[0798] Extracting information and storing it in a database

[0799] The server then uses natural language processing techniques (e.g., the transformers library) to analyze the documents converted into a unified format and extract important keywords and phrases. This information includes project status, decisions, issues, the date and time of security incidents, and the scope of their impact. The extracted information is classified as metadata and stored in a relational or NoSQL database.

[0800] Building a question-answering system

[0801] The server provides an interface on the terminal for users to input questions. This interface runs on terminals such as smartphones and PCs and is designed to allow users to input questions in natural language. When a question is input by the user, the terminal sends the question to the server.

[0802] Emotion engine recognizes emotions and adjusts responses

[0803] When analyzing a user's question, the server uses an emotion engine to recognize the user's emotions. This emotion recognition allows the server to identify the user's emotional state. For example, if the user is stressed, the server changes the answer to be more polite and reassuring.

[0804] Generate and provide answers

[0805] The server generates a summary based on the information extracted from the database and creates a specific response for the user, using the response adjusted by the emotion engine. The generated response is sent to the terminal and displayed to the user.

[0806] Specific examples

[0807] Example 1: Asking about the last security incident

[0808] A user types "When was the last security incident?" into a smartphone app. The app sends the question to the server, which analyzes the question, extracts relevant information from a database, and generates an answer. The app then displays the answer from the server and the emotion recognition results. For example, it might say, "The last security incident occurred on December 1, 2022. If you would like to know more about the countermeasures, please continue asking questions."

[0809] Prompt Sentence Examples

[0810] "Please provide a comprehensive analysis of the security logs and provide a detailed summary of the last security incident."

[0811] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0812] Step 1:

[0813] The server collects past project-related documents and security logs from the company's network drive or cloud storage. To collect documents, the server accesses pre-configured paths or API endpoints and downloads data in various formats, such as PDF, Word, Excel, and image files. This creates a document collection. The input is the network drive or cloud storage, and the output is a list of collected document files.

[0814] Step 2:

[0815] The server converts the collected documents into a unified format using OCR technology and a parser. Specifically, it uses the pytesseract library to extract text data from image files, and converts PDF and Word documents into text data using an appropriate parser. This converts documents in various formats into a unified text format. The input is the collected document files, and the output is the converted text data.

[0816] Step 3:

[0817] The server uses natural language processing technology to analyze the text data converted into a unified format and extract important keywords and phrases. At this stage, the transformers library is used to perform text analysis and generate metadata such as project progress, decisions, issues, and the date and time of the security incident and its scope of impact. The input is the converted text data, and the output is the extracted metadata.

[0818] Step 4:

[0819] The server stores the extracted metadata in a relational or NoSQL database. Specifically, it organizes and stores the metadata in tables or collections for future searches and queries. The input is the extracted metadata, and the output is the results stored in the database.

[0820] Step 5:

[0821] The terminal provides an interface for users to input questions. Users input questions in natural language on the screen of their smartphone or PC, and the questions are sent to the server via the interface. The input is the user's text question, and the output is the result of sending the question data to the server.

[0822] Step 6:

[0823] The server analyzes the received user question using natural language processing technology and an emotion engine. Through the analysis, it identifies keywords in the question and the user's emotional state and converts them into an appropriate internal query. The input is the user's question data, and the output is the analysis result and the internal query.

[0824] Step 7:

[0825] The server uses the internal query to search for relevant information from a database. The search extracts appropriate information for the question and tailors the answer to the user's emotional state. The input is the internal query, and the output is the search results and tailored answer data.

[0826] Step 8:

[0827] The server uses a generative AI model to summarize and optimize answers to complete a specific answer for the user. It generates prompt sentences and provides answers that take the user's emotional state into account. The input is search results and adjusted answer data, and the output is a completed answer. Using a generative AI model improves the quality and appropriateness of answers.

[0828] Step 9:

[0829] The terminal displays the answers received from the server to the user. The user can check the detailed answers on the screen and use this interface to enter additional questions. The input is the answer data from the server, and the output is the answer display result to the user.

[0830] Example prompt sentence:

[0831] "Please provide a comprehensive analysis of the security logs and provide a detailed summary of the last security incident."

[0832] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0833] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0834] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0835] [Third embodiment]

[0836] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0837] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0838] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0839] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0840] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0841] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0842] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0843] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0844] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0845] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0846] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0847] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0848] ---

[0849] The present invention is a system that collects documents related to past projects, converts them into a unified format, extracts important information using natural language processing technology, and stores it in a database. It also provides an interface for users to input questions, analyzes the user's questions, searches the database for relevant information, and summarizes the information for the user.

[0850] The program of this system mainly performs the following processes.

[0851] Data collection and integration

[0852] The server collects past documents from the company's network drive or cloud storage. The documents collected include meeting minutes, materials, and upper management reviews. These documents exist in various formats (PDF, Word, Excel, etc.), so the server converts them into digital data using OCR technology and parsers and organizes them into a unified format.

[0853] Extracting information and storing it in a database

[0854] The server then uses natural language processing to analyze the documents converted into a unified format and extract important keywords and phrases, such as project status, decisions, and issues. The extracted information is then classified as metadata and stored in a relational or NoSQL database.

[0855] Building a question-answering system

[0856] The server provides an interface for users to enter questions, which runs on a terminal and allows users to enter questions in natural language.

[0857] When a user types a question, the device sends it to the server, which uses machine learning models and natural language processing algorithms to analyze the question and convert it into an internal query, which then searches for relevant information in a database and extracts the best answer.

[0858] Generate and provide answers

[0859] The server then summarizes the information extracted from the database and generates a specific answer for the user, which is then sent to the terminal and displayed to the user. This process allows users to quickly obtain important information related to past projects, greatly improving work efficiency.

[0860] Specific examples

[0861] Example 1: Finding new campaign ideas

[0862] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0863] 2. The device sends this question to the server.

[0864] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0865] 4. The server extracts and summarizes relevant information from the database (such as past campaign successes and failures, opinions from upper management, etc.) and provides appropriate advice to the user.

[0866] 5. The terminal displays the response from the server to the user.

[0867] Example 2: Checking the progress of a project

[0868] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0869] 2. The device sends this question to the server.

[0870] 3. The server parses the question and extracts the latest information related to Project A from the database.

[0871] 4. The server summarizes the extracted information and provides the user with the current status of the project, any issues, next steps, etc.

[0872] 5. The device displays the server's response to the user, showing specific progress, for example, "Project A is currently in stage 3, with major tasks 70% complete. The next step is X."

[0873] In this way, the system of the present invention makes it possible to quickly and efficiently refer to detailed data on past projects, which greatly contributes to the planning of new projects and the efficiency of business operations.

[0874] The processing flow will be explained below.

[0875] ---

[0876] Step 1:

[0877] The server accesses the company's network drive or cloud storage to collect documents related to past projects, such as meeting minutes, materials, and senior management reviews.

[0878] Step 2:

[0879] The server converts the collected documents into text data using OCR technology, or if the documents are electronic files, they are imported as is.

[0880] Step 3:

[0881] The server converts documents in different formats into a unified format (e.g., JSON), using an appropriate parser for each file format.

[0882] Step 4:

[0883] The server uses natural language processing techniques to extract key information and keywords from the converted documents, including project status, decisions, and issues.

[0884] Step 5:

[0885] The server classifies the extracted information as metadata and stores it in a relational or NoSQL database. The data is organized by project, and highly related information is linked together.

[0886] Step 6:

[0887] The server provides the terminal with an interface for the user to input a question, the interface being designed to allow the user to input a question in natural language.

[0888] Step 7:

[0889] The user inputs a question into the terminal. For example, "I'm having trouble with a new campaign. What should I do?"

[0890] Step 8:

[0891] The terminal sends the user's question to the server.

[0892] Step 9:

[0893] The server receives the question and analyzes it using natural language processing techniques, identifying important keywords and phrases from the question and converting them into an internal query.

[0894] Step 10:

[0895] The server uses the internal query to search the database and extract information relevant to the question.

[0896] Step 11:

[0897] The server generates summaries based on the extracted information, such as examples of successful past campaigns or summaries of senior management's opinions.

[0898] Step 12:

[0899] The server transmits the generated summary to the terminal.

[0900] Step 13:

[0901] The device displays the summary sent from the server to the user, providing specific advice such as, "Project X is an example of a past successful campaign, and method Y was effective in this project."

[0902] ---

[0903] These are the specific steps of the program's processing, which allow users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[0904] Example 1

[0905] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0906] Because information related to past business projects is scattered, it is difficult to quickly and efficiently obtain the necessary information. In addition, because this information is stored in various formats, it is difficult to manage and search it in a unified manner. This inefficient management and search of information is a factor that reduces business efficiency.

[0907] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0908] In this invention, the server includes a means for collecting materials related to past business projects, a means for converting the collected materials into a unified format, and a means for extracting important information using machine learning technology. This makes it possible to efficiently collect information related to past business projects, convert it into a unified format, and extract important information. It can also analyze questions from users, search for related information from an information storage device, and provide a summary of the searched information, allowing users to quickly obtain the information they need.

[0909] "Materials" refers to documents and data related to past business projects, including minutes, reports, plans, emails, etc.

[0910] A "unified format" is a data format for converting collected materials into a consistent format, which makes it possible to handle materials in different formats in a uniform manner.

[0911] "Machine learning technology" is a technology for analyzing large amounts of data and automatically learning and making predictions, and includes algorithm-based analysis methods and deep learning technology.

[0912] "Information storage device" refers to a storage system, such as a database or cloud storage, for storing extracted information, including relational databases and NoSQL databases.

[0913] "Character recognition technology" means technology for converting printed or handwritten characters into digital data, including optical character recognition (OCR) technology.

[0914] A "natural language processing model" is a model for understanding and analyzing human language, and includes distributed representation models (Word2Vec) and translation models (BERT, GPT-3).

[0915] A "high-speed search engine" is an engine for quickly searching for necessary information from large amounts of data, and includes full-text search engines (Elasticsearch, Solr).

[0916] "Internal query" refers to a database query that is generated by analyzing a user's natural language question and searching for relevant information.

[0917] "Input device" refers to a hardware or software interface through which a user can input queries or data, including keyboards, touchscreens, and voice input devices.

[0918] This system efficiently collects materials related to past business projects, extracts important information, stores it in a database, and promptly provides relevant information in response to user questions. This system is composed of the following entities: a server, a terminal, and a user.

[0919] Data collection and integration

[0920] The server collects business-related materials from the company's network drive or cloud storage (e.g., a general cloud storage service). This includes meeting minutes, reports, plans, emails, etc. The server accesses the cloud storage using an appropriate API and retrieves files from the specified target directory. The collected materials are broad and often relate to the entire company's activities.

[0921] Unified data format conversion

[0922] The server uses open-source character recognition tools (e.g., Tesseract) to convert collected materials into digital data using optical character recognition (OCR) and document parsers. For example, OCR is performed on PDF files to extract text, and common document parsing libraries (e.g., Apache POI) are used for Word and Excel files. The converted data is then formatted into a consistent format (e.g., JSON, XML).

[0923] Information Extraction

[0924] The server then analyzes the unified formatted material using natural language processing (NLP) techniques, such as open-source NLP libraries (e.g., SpaCy, NLTK) and advanced machine learning models (e.g., BERT, GPT-3), to extract important noun phrases and key phrases from the document and generate a summary of the entire document.

[0925] Storage in the database

[0926] The server stores the extracted information in an information storage device. This information storage device can be a relational database (e.g., a general relational database management system) or a NoSQL database (e.g., a general NoSQL database solution). A table containing progress and risk factors for each project is designed and the information is stored there. An index is also created using a full-text search engine (e.g., Elasticsearch) to enable fast searches.

[0927] Receiving and analyzing user questions

[0928] A user inputs a question through an interface provided on the device. For example, "What is the latest progress on project A?" The device then sends this question to the server. The server analyzes the received question using natural language processing techniques and machine learning models. For example, it uses sentiment analysis and intent classification to identify the intent of the question and generates a related query. The internal query is used to search for related information.

[0929] Database search and answer generation

[0930] The server uses the parsed query to search for information in the information storage device. For example, a query like "latest progress on Project A" might retrieve the latest updates from the database. Elasticsearch might also be used to quickly find relevant documents. The server then summarizes the information from the database and generates a specific answer for the user. For example, "Project A is currently in Stage 3, with 70% of key tasks completed. The next step is X." This answer is then sent to the device and displayed to the user.

[0931] Specific examples

[0932] Example 1: Finding new campaign ideas

[0933] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[0934] 2. The device sends this question to the server.

[0935] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[0936] 4. The server extracts and summarizes relevant information from the information storage device (such as examples of past campaign successes and failures, opinions of upper management, etc.) and provides appropriate advice to the user.

[0937] 5. The terminal displays the response from the server to the user.

[0938] Example 2: Checking the progress of a project

[0939] 1. The user types into the terminal, "Tell me about the progress of Project A."

[0940] 2. The device sends this question to the server.

[0941] 3. The server analyzes the question and extracts the latest information related to Project A from the information repository.

[0942] 4. The server summarizes the extracted information and provides the user with the current status of the project, any issues, next steps, etc.

[0943] 5. The device displays the server's response to the user, showing specific progress, for example, "Project A is currently in stage 3, with major tasks 70% complete. The next step is X."

[0944] In this way, the system of the present invention allows for quick and efficient reference to detailed data related to past business projects, greatly contributing to the planning of new projects and the efficiency of business operations.

[0945] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0946] Step 1: Collect data

[0947] The server collects business-related materials from the company's network drive or cloud storage (e.g., a general cloud storage service). The server accesses the cloud storage using the appropriate API and retrieves files from the specified target directory. Inputs include the cloud storage API endpoint, authentication information, and the specified directory, and the collected raw material files are included as output. Specifically, the server retrieves a list of files and downloads each file in the list.

[0948] Step 2: Convert data into a unified format

[0949] The server converts the collected data into digital data using optical character recognition (OCR) technology and a document parser. The input is raw data files such as PDF, Word, and Excel, and the output is text data. For example, Tesseract OCR is used to extract text data from PDF files, and Apache POI is used to extract text data from Word and Excel files. The extracted text data is then formatted into a consistent, unified format (e.g., JSON or XML).

[0950] Step 3: Extracting information

[0951] The server analyzes the unified formatted data using natural language processing technology. The input is the unified formatted text data, and the output is information such as extracted important noun phrases and key phrases. Specifically, it uses SpaCy and NLTK to tokenize the document and extract important keywords and phrases. It also uses BERT and GPT-3 models to generate a summary of the entire document.

[0952] Step 4: Store in the database

[0953] The server stores the extracted information in an information storage device. The input is the extracted key information and summary data, and the output is the entries stored in a database. Specifically, it inserts the information into a relational database (e.g., a typical relational database management system) or a NoSQL database. In addition, it uses Elasticsearch to create an index, enabling fast searches.

[0954] Step 5: Receiving user questions

[0955] The user inputs a question through an interface provided on the terminal. The input is a natural language question typed by the user, and the output is the question sent to the server. For example, the user might input, "What is the latest progress on project A?" The terminal receives this question and sends it to the server.

[0956] Step 6: Parsing the Question

[0957] The server analyzes the received question using natural language processing technology and machine learning models. The input is the question submitted by the user, and the output is a related internal query. Specifically, it uses Sentiment Analysis and Intent Classification to identify the intent of the question, and keyphrase extraction to extract important keywords. The internal query is generated in SQL or Elasticsearch query language.

[0958] Step 7: Database Search

[0959] The server uses the parsed query to search for information in the information storage device. The input is the generated internal query, and the output is the search results. Specifically, it uses a relational database or Elasticsearch to retrieve related information, such as the latest update records and related documents.

[0960] Step 8: Generate and provide answers

[0961] The server summarizes the information retrieved from the database and generates a specific answer. The input is the search result data, and the output includes summary information provided to the user. For example, it generates content such as "Project A is currently in stage 3, with major tasks 70% complete. The next step is X." The server then sends the generated answer to the device and displays it to the user.

[0962] (Application example 1)

[0963] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0964] In recent years, efficient customer service and sales operations have become increasingly important in brick-and-mortar stores. However, store staff face challenges in quickly searching and retrieving vast amounts of past customer service history and sales data. Furthermore, they are often unable to provide prompt and appropriate responses to customer inquiries. This can lead to lower customer satisfaction and a deterioration in store operational efficiency. A system that can solve these problems is needed.

[0965] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0966] In this invention, the server includes a means for collecting documents related to past projects, a means for converting the collected documents into a unified format, and a means for extracting important information using natural language processing technology, which enables store staff to search and obtain past customer interaction history and sales data, and enables the server to quickly generate appropriate answers to store staff questions using a question-answering model.

[0967] "Documents related to past projects" refers to documents such as minutes, materials, and reviews related to projects that a company or organization has carried out in the past.

[0968] "Unified format" refers to converting documents that exist in various formats into a consistent format that makes them easier to parse and process.

[0969] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and respond to human language.

[0970] A "database" is a system that stores data in a structured way so that it can be efficiently searched and retrieved.

[0971] The "interface through which the user inputs questions" refers to the functions and screens that allow the user to input questions to the system in natural language.

[0972] "A means for store staff to search and obtain past customer interaction history and sales data" refers to a function that enables store staff to quickly find and use information related to past customer interactions and sales performance.

[0973] A "question answering model" refers to a machine learning model that generates appropriate answers to questions from users.

[0974] "Customer interaction history" is data that records past interactions with customers.

[0975] "Sales data" refers to data containing information about past sales activities, such as sales volume, timing, and sales amount.

[0976] MODE FOR CARRYING OUT THE INVENTION

[0977] This invention is a system that allows store staff in physical stores to quickly search and obtain past customer interaction history and sales data, and provides appropriate answers using a question-answering model.

[0978] 1. System Configuration

[0979] The system collects documents related to past projects, converts them into a unified format, extracts important information using natural language processing technology, and stores it in a database. It provides an interface for users to input questions, which then searches for and summarizes relevant information from the database and provides it to the user. A specific example is shown below.

[0980] Hardware and Software

[0981] The hardware and software used are as follows:

[0982] Flask: This Python web framework builds an API server to accept and process user queries.

[0983] transformers: A natural language processing library by Hugging Face. This library provides a question-answering model.

[0984] JSON: A data storage format that is easily used as a database.

[0985] 2. Program processing explanation

[0986] The server uses Flask to build an API server and accepts questions from users. Questions are sent in JSON format and processed within the server. The server uses natural language processing technology to analyze the user's question and search for information in a database. It uses the Hugging Face pipeline as a question-answering model to generate an appropriate answer using the question and related information in the database.

[0987] The server searches and retrieves information from the database, generates a response based on this information, and returns it to the user, allowing the user to quickly obtain the information they need.

[0988] Examples of concrete examples and prompts

[0989] For example, if a store employee asks, "What's the stock status of the latest smartphone accessories?", the question is sent to the server. The server searches past inventory and sales data and analyzes the information using natural language processing technology. Finally, the question-answering model generates an appropriate answer, which is displayed on the device.

[0990] The specific prompt is as follows:

[0991] Please let me know the latest smartphone accessories in stock.

[0992] In this way, the system of the present invention allows store staff to quickly and efficiently search and retrieve detailed customer interaction history and sales data, enabling them to provide appropriate responses to customers, thereby improving business efficiency and customer satisfaction.

[0993] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0994] Step 1:

[0995] Collect historical documents and convert them into a unified format.

[0996] The server collects past documents (meeting minutes, materials, upper management reviews, etc.) from the company's network drive or cloud storage. The collected documents (PDF, Word, Excel, etc.) are converted into digital data using OCR technology and parsers and organized into a unified format.

[0997] Input: Corporate historical documents in various formats

[0998] Output: Digital data converted into a unified format

[0999] Step 2:

[1000] Analyze the data and extract important information.

[1001] The server uses natural language processing technology to analyze the documents converted into a unified format and extract important keywords and phrases such as project progress, decisions, issues, etc. The extracted information is classified as metadata.

[1002] Input: Digital data converted into a unified format

[1003] Output: Metadata of extracted important information

[1004] Step 3:

[1005] Stores the data in the database.

[1006] The server then stores the extracted metadata of the key information in a relational or NoSQL database, which also contains data from other similar projects and organizes the information by relevance.

[1007] Input: Extracted metadata of important information

[1008] Output: Information organized in a database

[1009] Step 4:

[1010] Accepts and analyzes questions from users.

[1011] The user enters a question through the interface of their smartphone or smart glasses, which then sends the question to the server.

[1012] Input: A question typed by the user in natural language (e.g., "What is the stock status of the latest smartphone accessories?")

[1013] Output: Question data sent to the server

[1014] Step 5:

[1015] Analyze the question and perform a database search.

[1016] The server analyzes the user's question using a question-answering model, converts it into an internal query, and searches for related information in the database. Related information is extracted as search results.

[1017] Input: Question data submitted by the user

[1018] Output: Relevant information extracted from the database

[1019] Step 6:

[1020] Generate and provide answers.

[1021] The server generates and summarizes answers based on the extracted relevant information, and the generated answers are sent to the terminal and displayed to the user.

[1022] Input: Relevant information extracted from the database

[1023] Output: A summary answer provided to the user (e.g., "The latest smartphone accessories are in good stock.")

[1024] In this way, the entire system works together to improve the work efficiency of store staff in physical stores and increase customer satisfaction.

[1025] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1026] ---

[1027] This system collects documents related to past projects, converts them into a unified format, and uses natural language processing technology to extract important information and store it in a database. It also provides an interface for users to input questions, analyzes the user's question, searches the database for relevant information, summarizes the information, and provides it to the user. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, providing more appropriate answers and expressions based on the user's emotions.

[1028] The program of this system mainly performs the following processes.

[1029] Data collection and integration

[1030] The server collects past documents from the company's network drive or cloud storage. The documents collected include meeting minutes, materials, and upper management reviews. Because these documents exist in various formats (PDF, Word, Excel, etc.), the server converts them into digital data using OCR technology and parsers and organizes them into a unified format.

[1031] Extracting information and storing it in a database

[1032] The server then uses natural language processing to analyze the documents converted into a unified format and extract important keywords and phrases. This information includes project status, decisions, issues, etc. The extracted information is then classified as metadata and stored in a relational or NoSQL database.

[1033] Building a question-answering system

[1034] The server provides the terminal with an interface for the user to input a question, which runs on the terminal and is designed to allow the user to input a question in natural language.

[1035] When a user types a question, the device sends it to the server, which uses machine learning models and natural language processing algorithms to analyze the question and convert it into an internal query, which then searches for relevant information in a database and extracts the best answer.

[1036] Emotion engine recognizes emotions and adjusts responses

[1037] When analyzing the user's question, the server uses an emotion engine to recognize the user's emotion, which can identify the user's emotional state (e.g., stress, excitement, relief, etc.).

[1038] Based on the emotional state identified by the emotion engine, the server can adjust the content and expression of the response to suit the user. For example, if the user is under stress, the server can change the response to be more polite and reassuring.

[1039] Generate and provide answers

[1040] The server generates a summary based on the information extracted from the database and creates a specific answer for the user, using the answer adjusted by the emotion engine. The generated answer is sent to the terminal and displayed to the user.

[1041] Specific examples

[1042] Example 1: Finding new campaign ideas

[1043] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[1044] 2. The device sends this question to the server.

[1045] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[1046] 4. The server uses an emotion engine to recognize the user's emotional state, for example, identifying that the user is distressed and therefore stressed.

[1047] 5. The server extracts relevant information from the database (such as examples of past campaign successes and failures, opinions of senior management, etc.) and generates a summary that takes into account the user's emotional state.

[1048] 6. The device displays the response from the server to the user, for example, "Project X is an example of a successful campaign in the past, and method Y was effective in this project. Please stay calm and proceed."

[1049] Example 2: Checking the progress of a project

[1050] 1. The user types into the terminal, "Tell me about the progress of Project A."

[1051] 2. The device sends this question to the server.

[1052] 3. The server parses the question and extracts the latest information related to Project A from the database.

[1053] 4. The server uses an emotion engine to recognize the user's emotional state, for example, if the user is in a hurry, it generates a quick and concise answer.

[1054] 5. The server summarizes the extracted information and generates a summary that takes into account the user's emotional state, for example, "Project A is currently in stage 3, with key tasks 70% complete. The next step is X."

[1055] 6. The terminal displays the response from the server to the user.

[1056] In this way, we can provide information tailored to the user's emotional state, resulting in a more appropriate and user-friendly system that allows users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[1057] The processing flow will be explained below.

[1058] ---

[1059] Step 1:

[1060] The server collects historical documents from the company's network drives and cloud storage, including meeting minutes, materials, and senior management reviews.

[1061] Step 2:

[1062] The server converts these documents into text data using OCR technology, or if they are electronic files, they are imported as is.

[1063] Step 3:

[1064] The server converts the documents into a unified format (e.g., JSON) using parsers for various file formats (e.g., PDF, Word, Excel, etc.).

[1065] Step 4:

[1066] The server uses natural language processing techniques to extract important information and keywords from the converted documents, such as project status, decisions, and issues.

[1067] Step 5:

[1068] The server categorizes the extracted information as metadata and stores it in a relational or NoSQL database, organized by project and linking related information together.

[1069] Step 6:

[1070] The server provides a terminal with an interface for users to input questions, and this interface is designed to allow users to input questions in natural language.

[1071] Step 7:

[1072] The user inputs a question into the terminal. For example, "I'm having trouble with a new campaign. What should I do?"

[1073] Step 8:

[1074] The terminal sends the user's question to the server.

[1075] Step 9:

[1076] Once the server receives the question, it uses natural language processing techniques to analyze it, identifying important keywords and phrases and converting them into an internal query.

[1077] Step 10:

[1078] The server uses an emotion engine to recognize emotions from the user's questions, for example, identifying a user as being in a stressful state if they are worried.

[1079] Step 11:

[1080] Based on the query and the emotion recognition results, the server searches the database and extracts relevant information, such as examples of past campaign successes and failures, and the opinions of senior management.

[1081] Step 12:

[1082] The server generates an appropriate summary based on the extracted information, taking into account emotion recognition. For example, it generates a summary that uses gentle language and provides reassurance to users who are feeling stressed.

[1083] Step 13:

[1084] The server transmits the generated summary to the terminal.

[1085] Step 14:

[1086] The device displays the summary sent from the server to the user, such as, "As a concrete example of a past successful campaign, method Y was extremely effective in project X, so please consider it. You can proceed calmly, so I'm sure it will be fine."

[1087] ---

[1088] These are the specific processing steps of a system that includes an emotion engine. This process allows users to receive appropriate information and advice according to their emotional state, enabling them to carry out their work more effectively.

[1089] Example 2

[1090] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1091] Conventional systems have had difficulty efficiently collecting and analyzing information related to past projects within a company, extracting important information, and providing appropriate answers to user questions. Furthermore, they lacked the ability to recognize the user's emotional state and adjust the content and expression of the answers based on that state, making it difficult to provide user-friendly information. The present invention aims to solve these problems and provide a system that quickly and appropriately provides the information users desire.

[1092] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1093] In this invention, the server includes means for collecting documents related to past projects, means for converting the collected documents into a unified format, means for extracting important information using natural language processing technology, means for storing the extracted information in a database, means for analyzing a question from a user and searching the database for related information, means for summarizing the searched information and providing it to the user, means for providing an interface for the user to input a question, and emotion recognition means for recognizing the user's emotional state in response to the input question and adjusting the content and expression of the answer, thereby making it possible to provide information according to the user's emotions.

[1094] "Documents related to past projects" refers to documents in various formats, such as minutes, materials, and reviews related to past projects of a company or organization.

[1095] "Unified format" refers to a format that converts documents of different formats (PDF, Word, Excel, etc.) into one common format.

[1096] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[1097] "Database" refers to a system for efficiently storing, retrieving, and managing digital data.

[1098] "Question interface" refers to an input means that allows a user to input a question to the system in natural language.

[1099] "Emotion recognition means" refers to technology that recognizes a user's emotional state (stress, anxiety, excitement, etc.) from the user's questions and input content.

[1100] "Generative AI model" refers to an algorithm or model designed to help artificial intelligence perform a specific task (e.g., natural language parsing or generation).

[1101] A "prompt" is an instruction or question that is input to an AI model, and refers to the input sentence that enables the model to generate an appropriate response.

[1102] This system collects documents related to past projects, converts them into a unified format, and uses natural language processing technology to extract important information and store it in a database. It also provides an interface for users to input questions, analyzes the user's question, searches the database for relevant information, summarizes the information, and provides it to the user. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, providing more appropriate answers and expressions based on the user's emotions.

[1103] Data collection and integration

[1104] The server collects historical documents from the company's network drive and cloud storage. These documents include meeting minutes, handouts, and senior management reviews, and they exist in various formats, such as PDF, Word, and Excel. This creates a foundation for unifying information from diverse data sources. Specifically, the server uses OCR technology (e.g., Tesseract OCR) and parser libraries (e.g., Python's docx library and openpyxl library) to convert these documents into digital data and organize them into a unified format.

[1105] Extracting information and storing it in a database

[1106] The server analyzes the documents converted into a unified format using natural language processing techniques (e.g., Stanford NLP or spaCy) to extract important keywords and phrases. This information includes project status, decisions, and issues. The extracted information is classified as metadata and stored in a relational database (e.g., MySQL, PostgreSQL) or a NoSQL database (e.g., MongoDB). This creates a database that allows quick access to the information needed.

[1107] Building a question-answering system

[1108] The server provides the device with an interface through which the user can input a question. This interface can be web-based or app-style and is designed to allow the user to input a question in natural language. Once the user inputs a question, the device sends the question to the server. The server then uses a machine learning model (e.g., GPT-3) or a natural language processing algorithm to analyze the question and convert it into an internal query. This allows the optimal answer to the user's question to be quickly generated.

[1109] Emotion engine recognizes emotions and adjusts responses

[1110] During the question analysis process, the server uses an emotion engine (e.g., Microsoft Azure Cognitive Services' Emotion Analysis API) to recognize the user's emotions. Emotion recognition can identify the user's emotional state, such as stress, anxiety, or excitement. Based on the results of emotion recognition, the server can adjust the content and expression of the answer. For example, if the user is stressed, the server can make the answer more polite and add expressions that give a sense of security.

[1111] Generate and provide answers

[1112] The server generates a summary based on the information extracted from the database and creates a specific answer, including content adjusted by the emotion engine. The generated answer is sent to the device and displayed to the user, allowing the user to make effective decisions based on the displayed information.

[1113] Specific examples

[1114] Example 1: Finding new campaign ideas

[1115] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[1116] 2. The device sends this question to the server.

[1117] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[1118] 4. The server uses an emotion engine to determine that the user is experiencing stress due to their concerns.

[1119] 5. The server extracts relevant information from the database (such as examples of past campaign successes and failures, opinions of senior management, etc.) and generates a summary that takes into account the user's emotional state.

[1120] 6. The device displays the response from the server to the user, for example, "Project X is an example of a successful campaign in the past, and method Y was effective in this project. Please stay calm and proceed."

[1121] Example 2: Checking the progress of a project

[1122] 1. The user types into the terminal, "Tell me about the progress of Project A."

[1123] 2. The device sends this question to the server.

[1124] 3. The server parses the question and extracts the latest information related to Project A from the database.

[1125] 4. The server uses an emotion engine to generate quick and concise answers if the user is in a hurry.

[1126] 5. The server summarizes the extracted information and generates a summary that takes into account the user's emotional state, for example, "Project A is currently in stage 3, with key tasks 70% complete. The next step is X."

[1127] 6. The terminal displays the response from the server to the user.

[1128] In this way, a more appropriate and user-friendly system can be realized by providing information according to the user's emotional state, allowing users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[1129] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1130] Step 1:

[1131] A user logs into a terminal and begins the process of gathering documents related to past projects.

[1132] Input: User credentials

[1133] Output: System access permissions

[1134] What happens: The user enters their username and password, which the system authenticates.

[1135] Step 2:

[1136] The server collects past documents from the company's network drive or cloud storage.

[1137] Input: Path information for network drive or cloud storage

[1138] Output: A list of collected documents

[1139] Specific operation: The server scans the specified path for PDF, Word, Excel, and other files and saves them in temporary storage.

[1140] Step 3:

[1141] The server converts the collected documents into a unified format.

[1142] Input: Collected documents

[1143] Output: Data converted into a unified format

[1144] Specific operation: For PDF files, it uses Tesseract OCR to convert them to text data, and for Word and Excel files, it uses Python's docx and openpyxl libraries to extract the text data.

[1145] Step 4:

[1146] The server uses natural language processing techniques to extract key information from the documents that are converted into a unified format.

[1147] Input: Data converted into a unified format

[1148] Output: Extracted keywords and phrases

[1149] What it does: Tokenizes and tags parts of speech using Stanford NLP and spaCy to extract keywords such as project status and decisions.

[1150] Step 5:

[1151] The server stores the extracted information in a database.

[1152] Input: Extracted keywords or phrases

[1153] Output: Metadata stored in a database

[1154] Specific behavior: Stores information in a relational database (MySQL, PostgreSQL) or a NoSQL database (MongoDB). Fields include "Project name," "Progress," and "Decisions."

[1155] Step 6:

[1156] The user inputs a question through the terminal.

[1157] Input: User question

[1158] Output: Question data sent to the terminal

[1159] Specific operation: The user enters a question in natural language and clicks the "Send" button. The device sends this question to the server.

[1160] Step 7:

[1161] The server analyzes the user's question and searches the database for relevant information.

[1162] Input: User question

[1163] Output: Inner query and search results

[1164] What it does: It uses a machine learning model (e.g., GPT-3) to analyze the question and generate an internal query, which it then uses to search for relevant information in a database.

[1165] Step 8:

[1166] The server recognizes the user's emotional state based on the search results.

[1167] Input: User questions and search results

[1168] Output: Determining the user's emotional state

[1169] Specific behavior: Uses an emotion engine (e.g., Microsoft Azure Cognitive Services emotion analysis API) to recognize the user's emotional state from the questions.

[1170] Step 9:

[1171] The server adjusts the content and expression of the response based on the results of the emotion engine.

[1172] Input: User's emotional state and search results

[1173] Output: Adjusted answer

[1174] Specific behavior: Depending on the user's emotional state, the response may be more polite or specific. For example, a user in a stressful state may be given a reassuring response.

[1175] Step 10:

[1176] The server generates the final answer to provide to the user.

[1177] Input: Adjusted answer content

[1178] Output: Final answer

[1179] Specific actions: Integrate search results with tailored content to generate summaries, flesh out answers, and format them to appropriately respond to the user's question.

[1180] Step 11:

[1181] The terminal displays the final answer to the user.

[1182] Input: Final answer

[1183] Output: The answer displayed to the user

[1184] Specific operation: The final answer sent from the server is received and displayed on the user's device, allowing the user to view it and use it as a reference for making decisions.

[1185] (Application example 2)

[1186] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1187] Conventional systems have had difficulty efficiently collecting and analyzing documents related to past projects and security-related logs, and providing relevant information to users. Furthermore, they lacked the ability to recognize the user's emotional state and adjust the appropriate response content and expressions accordingly, making it difficult to improve user satisfaction. Therefore, there was a need for a system that could efficiently collect and analyze information and provide appropriate information based on the user's emotional state.

[1188] The identification process by the identification 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 documents related to past projects, means for converting the collected documents into a unified format, means for extracting important information using natural language processing technology, means for storing the extracted information in a database, means for analyzing questions from users and searching the database for related information, means for summarizing the searched information and providing it to the user, means for providing an interface for users to input questions, and means for combining an emotion engine that recognizes the user's emotions and adjusts the content and expression of the response. This makes it possible to efficiently collect and analyze information related to past projects and security, and to provide appropriate information based on the user's emotional state.

[1189] "Documents related to past projects" are documents such as minutes, reports, and materials related to projects that a company or organization has carried out in the past.

[1190] A "uniform format" is a standardized format that converts collected documents into the same format to facilitate analysis and search.

[1191] "Natural language processing technology" is a technology that enables computers to understand and analyze human language, and extract meaningful information from text data.

[1192] "Key information extraction means" refers to methods and algorithms that use natural language processing techniques to identify and extract key information, such as project status, decisions, and issues, from collected documents.

[1193] A "database" is a system for storing and managing extracted information, a collection of data structured to facilitate search and access.

[1194] "Means for analyzing a question from a user and searching for related information from a database" refers to a method or algorithm for analyzing a question entered by a user and searching for information related to that question from a database.

[1195] "Means for summarizing retrieved information and providing it to the user" refers to a method or algorithm that succinctly summarizes relevant information extracted from a database and provides it to the user in an easily understandable format.

[1196] The "interface through which the user enters a question" is the part of the software that provides a screen or input field into which the user can enter a question in text form.

[1197] An "emotion engine" is a technology that analyzes the user's emotional state from the text they input and adjusts the response content and expression appropriately based on that.

[1198] "Security-related logs and event reports" are documents or data files that record security incidents and other security-related information within a company or organization.

[1199] A "generative AI model" is an artificial intelligence model that uses pre-trained algorithms to generate appropriate responses or prompts to user questions.

[1200] A "prompt sentence" is a specific question or instruction sentence that is generated to be easy for the user to understand.

[1201] This system collects documents related to past projects and security-related logs and event reports, converts them into a unified format, analyzes them, and provides appropriate answers to user questions. It also has the ability to recognize the user's emotional state and adjust the content and expression of the answers.

[1202] Data collection and integration

[1203] The server collects documents related to past projects and security-related logs from the company's network drive or cloud storage. Because these documents exist in various formats, such as PDF, Word, and image files, the server converts them into digital data using OCR technology (e.g., the pytesseract library) and parsers, and organizes them into a unified format.

[1204] Extracting information and storing it in a database

[1205] The server then uses natural language processing techniques (e.g., the transformers library) to analyze the documents converted into a unified format and extract important keywords and phrases. This information includes project status, decisions, issues, the date and time of security incidents, and the scope of their impact. The extracted information is classified as metadata and stored in a relational or NoSQL database.

[1206] Building a question-answering system

[1207] The server provides an interface on the terminal for users to input questions. This interface runs on terminals such as smartphones and PCs and is designed to allow users to input questions in natural language. When a question is input by the user, the terminal sends the question to the server.

[1208] Emotion engine recognizes emotions and adjusts responses

[1209] When analyzing a user's question, the server uses an emotion engine to recognize the user's emotions. This emotion recognition allows the server to identify the user's emotional state. For example, if the user is stressed, the server changes the answer to be more polite and reassuring.

[1210] Generate and provide answers

[1211] The server generates a summary based on the information extracted from the database and creates a specific response for the user, using the response adjusted by the emotion engine. The generated response is sent to the terminal and displayed to the user.

[1212] Specific examples

[1213] Example 1: Asking about the last security incident

[1214] A user types "When was the last security incident?" into a smartphone app. The app sends the question to the server, which analyzes the question, extracts relevant information from a database, and generates an answer. The app then displays the answer from the server and the emotion recognition results. For example, it might say, "The last security incident occurred on December 1, 2022. If you would like to know more about the countermeasures, please continue asking questions."

[1215] Prompt Sentence Examples

[1216] "Please provide a comprehensive analysis of the security logs and provide a detailed summary of the last security incident."

[1217] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1218] Step 1:

[1219] The server collects past project-related documents and security logs from the company's network drive or cloud storage. To collect documents, the server accesses pre-configured paths or API endpoints and downloads data in various formats, such as PDF, Word, Excel, and image files. This creates a document collection. The input is the network drive or cloud storage, and the output is a list of collected document files.

[1220] Step 2:

[1221] The server converts the collected documents into a unified format using OCR technology and a parser. Specifically, it uses the pytesseract library to extract text data from image files, and converts PDF and Word documents into text data using an appropriate parser. This converts documents in various formats into a unified text format. The input is the collected document files, and the output is the converted text data.

[1222] Step 3:

[1223] The server uses natural language processing technology to analyze the text data converted into a unified format and extract important keywords and phrases. At this stage, the transformers library is used to perform text analysis and generate metadata such as project progress, decisions, issues, and the date and time of the security incident and its scope of impact. The input is the converted text data, and the output is the extracted metadata.

[1224] Step 4:

[1225] The server stores the extracted metadata in a relational or NoSQL database. Specifically, it organizes and stores the metadata in tables or collections for future searches and queries. The input is the extracted metadata, and the output is the results stored in the database.

[1226] Step 5:

[1227] The terminal provides an interface for users to input questions. Users input questions in natural language on the screen of their smartphone or PC, and the questions are sent to the server via the interface. The input is the user's text question, and the output is the result of sending the question data to the server.

[1228] Step 6:

[1229] The server analyzes the received user question using natural language processing technology and an emotion engine. Through the analysis, it identifies keywords in the question and the user's emotional state and converts them into an appropriate internal query. The input is the user's question data, and the output is the analysis result and the internal query.

[1230] Step 7:

[1231] The server uses the internal query to search for relevant information from a database. The search extracts appropriate information for the question and tailors the answer to the user's emotional state. The input is the internal query, and the output is the search results and tailored answer data.

[1232] Step 8:

[1233] The server uses a generative AI model to summarize and optimize answers to complete a specific answer for the user. It generates prompt sentences and provides answers that take the user's emotional state into account. The input is search results and adjusted answer data, and the output is a completed answer. Using a generative AI model improves the quality and appropriateness of answers.

[1234] Step 9:

[1235] The terminal displays the answers received from the server to the user. The user can check the detailed answers on the screen and use this interface to enter additional questions. The input is the answer data from the server, and the output is the answer display result to the user.

[1236] Example prompt sentence:

[1237] "Please provide a comprehensive analysis of the security logs and provide a detailed summary of the last security incident."

[1238] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1239] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1240] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1241] [Fourth embodiment]

[1242] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1243] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1244] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1245] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1246] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1247] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1248] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1249] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1250] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1251] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1252] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1253] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1254] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1255] ---

[1256] The present invention is a system that collects documents related to past projects, converts them into a unified format, extracts important information using natural language processing technology, and stores it in a database. It also provides an interface for users to input questions, analyzes the user's questions, searches the database for relevant information, and summarizes the information for the user.

[1257] The program of this system mainly performs the following processes.

[1258] Data collection and integration

[1259] The server collects past documents from the company's network drive or cloud storage. The documents collected include meeting minutes, materials, and upper management reviews. These documents exist in various formats (PDF, Word, Excel, etc.), so the server converts them into digital data using OCR technology and parsers and organizes them into a unified format.

[1260] Extracting information and storing it in a database

[1261] The server then uses natural language processing to analyze the documents converted into a unified format and extract important keywords and phrases, such as project status, decisions, and issues. The extracted information is then classified as metadata and stored in a relational or NoSQL database.

[1262] Building a question-answering system

[1263] The server provides an interface for users to enter questions, which runs on a terminal and allows users to enter questions in natural language.

[1264] When a user types a question, the device sends it to the server, which uses machine learning models and natural language processing algorithms to analyze the question and convert it into an internal query, which then searches for relevant information in a database and extracts the best answer.

[1265] Generate and provide answers

[1266] The server then summarizes the information extracted from the database and generates a specific answer for the user, which is then sent to the terminal and displayed to the user. This process allows users to quickly obtain important information related to past projects, greatly improving work efficiency.

[1267] Specific examples

[1268] Example 1: Finding new campaign ideas

[1269] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[1270] 2. The device sends this question to the server.

[1271] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[1272] 4. The server extracts and summarizes relevant information from the database (such as past campaign successes and failures, opinions from upper management, etc.) and provides appropriate advice to the user.

[1273] 5. The terminal displays the response from the server to the user.

[1274] Example 2: Checking the progress of a project

[1275] 1. The user types into the terminal, "Tell me about the progress of Project A."

[1276] 2. The device sends this question to the server.

[1277] 3. The server parses the question and extracts the latest information related to Project A from the database.

[1278] 4. The server summarizes the extracted information and provides the user with the current status of the project, any issues, next steps, etc.

[1279] 5. The device displays the server's response to the user, showing specific progress, for example, "Project A is currently in stage 3, with major tasks 70% complete. The next step is X."

[1280] In this way, the system of the present invention makes it possible to quickly and efficiently refer to detailed data on past projects, which greatly contributes to the planning of new projects and the efficiency of business operations.

[1281] The processing flow will be explained below.

[1282] ---

[1283] Step 1:

[1284] The server accesses the company's network drive or cloud storage to collect documents related to past projects, such as meeting minutes, materials, and senior management reviews.

[1285] Step 2:

[1286] The server converts the collected documents into text data using OCR technology, or if the documents are electronic files, they are imported as is.

[1287] Step 3:

[1288] The server converts documents in different formats into a unified format (e.g., JSON), using an appropriate parser for each file format.

[1289] Step 4:

[1290] The server uses natural language processing techniques to extract key information and keywords from the converted documents, including project status, decisions, and issues.

[1291] Step 5:

[1292] The server classifies the extracted information as metadata and stores it in a relational or NoSQL database. The data is organized by project, and highly related information is linked together.

[1293] Step 6:

[1294] The server provides the terminal with an interface for the user to input a question, the interface being designed to allow the user to input a question in natural language.

[1295] Step 7:

[1296] The user inputs a question into the terminal. For example, "I'm having trouble with a new campaign. What should I do?"

[1297] Step 8:

[1298] The terminal sends the user's question to the server.

[1299] Step 9:

[1300] The server receives the question and analyzes it using natural language processing techniques, identifying important keywords and phrases from the question and converting them into an internal query.

[1301] Step 10:

[1302] The server uses the internal query to search the database and extract information relevant to the question.

[1303] Step 11:

[1304] The server generates summaries based on the extracted information, such as examples of successful past campaigns or summaries of senior management's opinions.

[1305] Step 12:

[1306] The server transmits the generated summary to the terminal.

[1307] Step 13:

[1308] The device displays the summary sent from the server to the user, providing specific advice such as, "Project X is an example of a past successful campaign, and method Y was effective in this project."

[1309] ---

[1310] These are the specific steps of the program's processing, which allow users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[1311] Example 1

[1312] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1313] Because information related to past business projects is scattered, it is difficult to quickly and efficiently obtain the necessary information. In addition, because this information is stored in various formats, it is difficult to manage and search it in a unified manner. This inefficient management and search of information is a factor that reduces business efficiency.

[1314] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1315] In this invention, the server includes a means for collecting materials related to past business projects, a means for converting the collected materials into a unified format, and a means for extracting important information using machine learning technology. This makes it possible to efficiently collect information related to past business projects, convert it into a unified format, and extract important information. It can also analyze questions from users, search for related information from an information storage device, and provide a summary of the searched information, allowing users to quickly obtain the information they need.

[1316] "Materials" refers to documents and data related to past business projects, including minutes, reports, plans, emails, etc.

[1317] A "unified format" is a data format for converting collected materials into a consistent format, which makes it possible to handle materials in different formats in a uniform manner.

[1318] "Machine learning technology" is a technology for analyzing large amounts of data and automatically learning and making predictions, and includes algorithm-based analysis methods and deep learning technology.

[1319] "Information storage device" refers to a storage system, such as a database or cloud storage, for storing extracted information, including relational databases and NoSQL databases.

[1320] "Character recognition technology" means technology for converting printed or handwritten characters into digital data, including optical character recognition (OCR) technology.

[1321] A "natural language processing model" is a model for understanding and analyzing human language, and includes distributed representation models (Word2Vec) and translation models (BERT, GPT-3).

[1322] A "high-speed search engine" is an engine for quickly searching for necessary information from large amounts of data, and includes full-text search engines (Elasticsearch, Solr).

[1323] "Internal query" refers to a database query that is generated by analyzing a user's natural language question and searching for relevant information.

[1324] "Input device" refers to a hardware or software interface through which a user can input queries or data, including keyboards, touchscreens, and voice input devices.

[1325] This system efficiently collects materials related to past business projects, extracts important information, stores it in a database, and promptly provides relevant information in response to user questions. This system is composed of the following entities: a server, a terminal, and a user.

[1326] Data collection and integration

[1327] The server collects business-related materials from the company's network drive or cloud storage (e.g., a general cloud storage service). This includes meeting minutes, reports, plans, emails, etc. The server accesses the cloud storage using an appropriate API and retrieves files from the specified target directory. The collected materials are broad and often relate to the entire company's activities.

[1328] Unified data format conversion

[1329] The server uses open-source character recognition tools (e.g., Tesseract) to convert collected materials into digital data using optical character recognition (OCR) and document parsers. For example, OCR is performed on PDF files to extract text, and common document parsing libraries (e.g., Apache POI) are used for Word and Excel files. The converted data is then formatted into a consistent format (e.g., JSON, XML).

[1330] Information Extraction

[1331] The server then analyzes the unified formatted material using natural language processing (NLP) techniques, such as open-source NLP libraries (e.g., SpaCy, NLTK) and advanced machine learning models (e.g., BERT, GPT-3), to extract important noun phrases and key phrases from the document and generate a summary of the entire document.

[1332] Storage in the database

[1333] The server stores the extracted information in an information storage device. This information storage device can be a relational database (e.g., a general relational database management system) or a NoSQL database (e.g., a general NoSQL database solution). A table containing progress and risk factors for each project is designed and the information is stored there. An index is also created using a full-text search engine (e.g., Elasticsearch) to enable fast searches.

[1334] Receiving and analyzing user questions

[1335] A user inputs a question through an interface provided on the device. For example, "What is the latest progress on project A?" The device then sends this question to the server. The server analyzes the received question using natural language processing techniques and machine learning models. For example, it uses sentiment analysis and intent classification to identify the intent of the question and generates a related query. The internal query is used to search for related information.

[1336] Database search and answer generation

[1337] The server uses the parsed query to search for information in the information storage device. For example, a query like "latest progress on Project A" might retrieve the latest updates from the database. Elasticsearch might also be used to quickly find relevant documents. The server then summarizes the information from the database and generates a specific answer for the user. For example, "Project A is currently in Stage 3, with 70% of key tasks completed. The next step is X." This answer is then sent to the device and displayed to the user.

[1338] Specific examples

[1339] Example 1: Finding new campaign ideas

[1340] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[1341] 2. The device sends this question to the server.

[1342] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[1343] 4. The server extracts and summarizes relevant information from the information storage device (such as examples of past campaign successes and failures, opinions of upper management, etc.) and provides appropriate advice to the user.

[1344] 5. The terminal displays the response from the server to the user.

[1345] Example 2: Checking the progress of a project

[1346] 1. The user types into the terminal, "Tell me about the progress of Project A."

[1347] 2. The device sends this question to the server.

[1348] 3. The server analyzes the question and extracts the latest information related to Project A from the information repository.

[1349] 4. The server summarizes the extracted information and provides the user with the current status of the project, any issues, next steps, etc.

[1350] 5. The device displays the server's response to the user, showing specific progress, for example, "Project A is currently in stage 3, with major tasks 70% complete. The next step is X."

[1351] In this way, the system of the present invention allows for quick and efficient reference to detailed data related to past business projects, greatly contributing to the planning of new projects and the efficiency of business operations.

[1352] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1353] Step 1: Collect data

[1354] The server collects business-related materials from the company's network drive or cloud storage (e.g., a general cloud storage service). The server accesses the cloud storage using the appropriate API and retrieves files from the specified target directory. Inputs include the cloud storage API endpoint, authentication information, and the specified directory, and the collected raw material files are included as output. Specifically, the server retrieves a list of files and downloads each file in the list.

[1355] Step 2: Convert data into a unified format

[1356] The server converts the collected data into digital data using optical character recognition (OCR) technology and a document parser. The input is raw data files such as PDF, Word, and Excel, and the output is text data. For example, Tesseract OCR is used to extract text data from PDF files, and Apache POI is used to extract text data from Word and Excel files. The extracted text data is then formatted into a consistent, unified format (e.g., JSON or XML).

[1357] Step 3: Extracting information

[1358] The server analyzes the unified formatted data using natural language processing technology. The input is the unified formatted text data, and the output is information such as extracted important noun phrases and key phrases. Specifically, it uses SpaCy and NLTK to tokenize the document and extract important keywords and phrases. It also uses BERT and GPT-3 models to generate a summary of the entire document.

[1359] Step 4: Store in the database

[1360] The server stores the extracted information in an information storage device. The input is the extracted key information and summary data, and the output is the entries stored in a database. Specifically, it inserts the information into a relational database (e.g., a typical relational database management system) or a NoSQL database. In addition, it uses Elasticsearch to create an index, enabling fast searches.

[1361] Step 5: Receiving user questions

[1362] The user inputs a question through an interface provided on the terminal. The input is a natural language question typed by the user, and the output is the question sent to the server. For example, the user might input, "What is the latest progress on project A?" The terminal receives this question and sends it to the server.

[1363] Step 6: Parsing the Question

[1364] The server analyzes the received question using natural language processing technology and machine learning models. The input is the question submitted by the user, and the output is a related internal query. Specifically, it uses Sentiment Analysis and Intent Classification to identify the intent of the question, and keyphrase extraction to extract important keywords. The internal query is generated in SQL or Elasticsearch query language.

[1365] Step 7: Database Search

[1366] The server uses the parsed query to search for information in the information storage device. The input is the generated internal query, and the output is the search results. Specifically, it uses a relational database or Elasticsearch to retrieve related information, such as the latest update records and related documents.

[1367] Step 8: Generate and provide answers

[1368] The server summarizes the information retrieved from the database and generates a specific answer. The input is the search result data, and the output includes summary information provided to the user. For example, it generates content such as "Project A is currently in stage 3, with major tasks 70% complete. The next step is X." The server then sends the generated answer to the device and displays it to the user.

[1369] (Application example 1)

[1370] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1371] In recent years, efficient customer service and sales operations have become increasingly important in brick-and-mortar stores. However, store staff face challenges in quickly searching and retrieving vast amounts of past customer service history and sales data. Furthermore, they are often unable to provide prompt and appropriate responses to customer inquiries. This can lead to lower customer satisfaction and a deterioration in store operational efficiency. A system that can solve these problems is needed.

[1372] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1373] In this invention, the server includes a means for collecting documents related to past projects, a means for converting the collected documents into a unified format, and a means for extracting important information using natural language processing technology, which enables store staff to search and obtain past customer interaction history and sales data, and enables the server to quickly generate appropriate answers to store staff questions using a question-answering model.

[1374] "Documents related to past projects" refers to documents such as minutes, materials, and reviews related to projects that a company or organization has carried out in the past.

[1375] "Unified format" refers to converting documents that exist in various formats into a consistent format that makes them easier to parse and process.

[1376] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and respond to human language.

[1377] A "database" is a system that stores data in a structured way so that it can be efficiently searched and retrieved.

[1378] The "interface through which the user inputs questions" refers to the functions and screens that allow the user to input questions to the system in natural language.

[1379] "A means for store staff to search and obtain past customer interaction history and sales data" refers to a function that enables store staff to quickly find and use information related to past customer interactions and sales performance.

[1380] A "question answering model" refers to a machine learning model that generates appropriate answers to questions from users.

[1381] "Customer interaction history" is data that records past interactions with customers.

[1382] "Sales data" refers to data containing information about past sales activities, such as sales volume, timing, and sales amount.

[1383] MODE FOR CARRYING OUT THE INVENTION

[1384] This invention is a system that allows store staff in physical stores to quickly search and obtain past customer interaction history and sales data, and provides appropriate answers using a question-answering model.

[1385] 1. System Configuration

[1386] The system collects documents related to past projects, converts them into a unified format, extracts important information using natural language processing technology, and stores it in a database. It provides an interface for users to input questions, which then searches for and summarizes relevant information from the database and provides it to the user. A specific example is shown below.

[1387] Hardware and Software

[1388] The hardware and software used are as follows:

[1389] Flask: This Python web framework builds an API server to accept and process user queries.

[1390] transformers: A natural language processing library by Hugging Face. This library provides a question-answering model.

[1391] JSON: A data storage format that is easily used as a database.

[1392] 2. Program processing explanation

[1393] The server uses Flask to build an API server and accepts questions from users. Questions are sent in JSON format and processed within the server. The server uses natural language processing technology to analyze the user's question and search for information in a database. It uses the Hugging Face pipeline as a question-answering model to generate an appropriate answer using the question and related information in the database.

[1394] The server searches and retrieves information from the database, generates a response based on this information, and returns it to the user, allowing the user to quickly obtain the information they need.

[1395] Examples of concrete examples and prompts

[1396] For example, if a store employee asks, "What's the stock status of the latest smartphone accessories?", the question is sent to the server. The server searches past inventory and sales data and analyzes the information using natural language processing technology. Finally, the question-answering model generates an appropriate answer, which is displayed on the device.

[1397] The specific prompt is as follows:

[1398] Please let me know the latest smartphone accessories in stock.

[1399] In this way, the system of the present invention allows store staff to quickly and efficiently search and retrieve detailed customer interaction history and sales data, enabling them to provide appropriate responses to customers, thereby improving business efficiency and customer satisfaction.

[1400] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1401] Step 1:

[1402] Collect historical documents and convert them into a unified format.

[1403] The server collects past documents (meeting minutes, materials, upper management reviews, etc.) from the company's network drive or cloud storage. The collected documents (PDF, Word, Excel, etc.) are converted into digital data using OCR technology and parsers and organized into a unified format.

[1404] Input: Corporate historical documents in various formats

[1405] Output: Digital data converted into a unified format

[1406] Step 2:

[1407] Analyze the data and extract important information.

[1408] The server uses natural language processing technology to analyze the documents converted into a unified format and extract important keywords and phrases such as project progress, decisions, issues, etc. The extracted information is classified as metadata.

[1409] Input: Digital data converted into a unified format

[1410] Output: Metadata of extracted important information

[1411] Step 3:

[1412] Stores the data in the database.

[1413] The server then stores the extracted metadata of the key information in a relational or NoSQL database, which also contains data from other similar projects and organizes the information by relevance.

[1414] Input: Extracted metadata of important information

[1415] Output: Information organized in a database

[1416] Step 4:

[1417] Accepts and analyzes questions from users.

[1418] The user enters a question through the interface of their smartphone or smart glasses, which then sends the question to the server.

[1419] Input: A question typed by the user in natural language (e.g., "What is the stock status of the latest smartphone accessories?")

[1420] Output: Question data sent to the server

[1421] Step 5:

[1422] Analyze the question and perform a database search.

[1423] The server analyzes the user's question using a question-answering model, converts it into an internal query, and searches for related information in the database. Related information is extracted as search results.

[1424] Input: Question data submitted by the user

[1425] Output: Relevant information extracted from the database

[1426] Step 6:

[1427] Generate and provide answers.

[1428] The server generates and summarizes answers based on the extracted relevant information, and the generated answers are sent to the terminal and displayed to the user.

[1429] Input: Relevant information extracted from the database

[1430] Output: A summary answer provided to the user (e.g., "The latest smartphone accessories are in good stock.")

[1431] In this way, the entire system works together to improve the work efficiency of store staff in physical stores and increase customer satisfaction.

[1432] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1433] ---

[1434] This system collects documents related to past projects, converts them into a unified format, and uses natural language processing technology to extract important information and store it in a database. It also provides an interface for users to input questions, analyzes the user's question, searches the database for relevant information, summarizes the information, and provides it to the user. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, providing more appropriate answers and expressions based on the user's emotions.

[1435] The program of this system mainly performs the following processes.

[1436] Data collection and integration

[1437] The server collects past documents from the company's network drive or cloud storage. The documents collected include meeting minutes, materials, and upper management reviews. Because these documents exist in various formats (PDF, Word, Excel, etc.), the server converts them into digital data using OCR technology and parsers and organizes them into a unified format.

[1438] Extracting information and storing it in a database

[1439] The server then uses natural language processing to analyze the documents converted into a unified format and extract important keywords and phrases. This information includes project status, decisions, issues, etc. The extracted information is then classified as metadata and stored in a relational or NoSQL database.

[1440] Building a question-answering system

[1441] The server provides the terminal with an interface for the user to input a question, which runs on the terminal and is designed to allow the user to input a question in natural language.

[1442] When a user types a question, the device sends it to the server, which uses machine learning models and natural language processing algorithms to analyze the question and convert it into an internal query, which then searches for relevant information in a database and extracts the best answer.

[1443] Emotion engine recognizes emotions and adjusts responses

[1444] When analyzing the user's question, the server uses an emotion engine to recognize the user's emotion, which can identify the user's emotional state (e.g., stress, excitement, relief, etc.).

[1445] Based on the emotional state identified by the emotion engine, the server can adjust the content and expression of the response to suit the user. For example, if the user is under stress, the server can change the response to be more polite and reassuring.

[1446] Generate and provide answers

[1447] The server generates a summary based on the information extracted from the database and creates a specific answer for the user, using the answer adjusted by the emotion engine. The generated answer is sent to the terminal and displayed to the user.

[1448] Specific examples

[1449] Example 1: Finding new campaign ideas

[1450] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[1451] 2. The device sends this question to the server.

[1452] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[1453] 4. The server uses an emotion engine to recognize the user's emotional state, for example, identifying that the user is distressed and therefore stressed.

[1454] 5. The server extracts relevant information from the database (such as examples of past campaign successes and failures, opinions of senior management, etc.) and generates a summary that takes into account the user's emotional state.

[1455] 6. The device displays the response from the server to the user, for example, "Project X is an example of a successful campaign in the past, and method Y was effective in this project. Please stay calm and proceed."

[1456] Example 2: Checking the progress of a project

[1457] 1. The user types into the terminal, "Tell me about the progress of Project A."

[1458] 2. The device sends this question to the server.

[1459] 3. The server parses the question and extracts the latest information related to Project A from the database.

[1460] 4. The server uses an emotion engine to recognize the user's emotional state, for example, if the user is in a hurry, it generates a quick and concise answer.

[1461] 5. The server summarizes the extracted information and generates a summary that takes into account the user's emotional state, for example, "Project A is currently in stage 3, with key tasks 70% complete. The next step is X."

[1462] 6. The terminal displays the response from the server to the user.

[1463] In this way, we can provide information tailored to the user's emotional state, resulting in a more appropriate and user-friendly system that allows users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[1464] The processing flow will be explained below.

[1465] ---

[1466] Step 1:

[1467] The server collects historical documents from the company's network drives and cloud storage, including meeting minutes, materials, and senior management reviews.

[1468] Step 2:

[1469] The server converts these documents into text data using OCR technology, or if they are electronic files, they are imported as is.

[1470] Step 3:

[1471] The server converts the documents into a unified format (e.g., JSON) using parsers for various file formats (e.g., PDF, Word, Excel, etc.).

[1472] Step 4:

[1473] The server uses natural language processing techniques to extract important information and keywords from the converted documents, such as project status, decisions, and issues.

[1474] Step 5:

[1475] The server categorizes the extracted information as metadata and stores it in a relational or NoSQL database, organized by project and linking related information together.

[1476] Step 6:

[1477] The server provides a terminal with an interface for users to input questions, and this interface is designed to allow users to input questions in natural language.

[1478] Step 7:

[1479] The user inputs a question into the terminal. For example, "I'm having trouble with a new campaign. What should I do?"

[1480] Step 8:

[1481] The terminal sends the user's question to the server.

[1482] Step 9:

[1483] Once the server receives the question, it uses natural language processing techniques to analyze it, identifying important keywords and phrases and converting them into an internal query.

[1484] Step 10:

[1485] The server uses an emotion engine to recognize emotions from the user's questions, for example, identifying a user as being in a stressful state if they are worried.

[1486] Step 11:

[1487] Based on the query and the emotion recognition results, the server searches the database and extracts relevant information, such as examples of past campaign successes and failures, and the opinions of senior management.

[1488] Step 12:

[1489] The server generates an appropriate summary based on the extracted information, taking into account emotion recognition. For example, it generates a summary that uses gentle language and provides reassurance to users who are feeling stressed.

[1490] Step 13:

[1491] The server transmits the generated summary to the terminal.

[1492] Step 14:

[1493] The device displays the summary sent from the server to the user, such as, "As a concrete example of a past successful campaign, method Y was extremely effective in project X, so please consider it. You can proceed calmly, so I'm sure it will be fine."

[1494] ---

[1495] These are the specific processing steps of a system that includes an emotion engine. This process allows users to receive appropriate information and advice according to their emotional state, enabling them to carry out their work more effectively.

[1496] Example 2

[1497] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1498] Conventional systems have had difficulty efficiently collecting and analyzing information related to past projects within a company, extracting important information, and providing appropriate answers to user questions. Furthermore, they lacked the ability to recognize the user's emotional state and adjust the content and expression of the answers based on that state, making it difficult to provide user-friendly information. The present invention aims to solve these problems and provide a system that quickly and appropriately provides the information users desire.

[1499] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1500] In this invention, the server includes means for collecting documents related to past projects, means for converting the collected documents into a unified format, means for extracting important information using natural language processing technology, means for storing the extracted information in a database, means for analyzing a question from a user and searching the database for related information, means for summarizing the searched information and providing it to the user, means for providing an interface for the user to input a question, and emotion recognition means for recognizing the user's emotional state in response to the input question and adjusting the content and expression of the answer, thereby making it possible to provide information according to the user's emotions.

[1501] "Documents related to past projects" refers to documents in various formats, such as minutes, materials, and reviews related to past projects of a company or organization.

[1502] "Unified format" refers to a format that converts documents of different formats (PDF, Word, Excel, etc.) into one common format.

[1503] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[1504] "Database" refers to a system for efficiently storing, retrieving, and managing digital data.

[1505] "Question interface" refers to an input means that allows a user to input a question to the system in natural language.

[1506] "Emotion recognition means" refers to technology that recognizes a user's emotional state (stress, anxiety, excitement, etc.) from the user's questions and input content.

[1507] "Generative AI model" refers to an algorithm or model designed to help artificial intelligence perform a specific task (e.g., natural language parsing or generation).

[1508] A "prompt" is an instruction or question that is input to an AI model, and refers to the input sentence that enables the model to generate an appropriate response.

[1509] This system collects documents related to past projects, converts them into a unified format, and uses natural language processing technology to extract important information and store it in a database. It also provides an interface for users to input questions, analyzes the user's question, searches the database for relevant information, summarizes the information, and provides it to the user. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, providing more appropriate answers and expressions based on the user's emotions.

[1510] Data collection and integration

[1511] The server collects historical documents from the company's network drive and cloud storage. These documents include meeting minutes, handouts, and senior management reviews, and they exist in various formats, such as PDF, Word, and Excel. This creates a foundation for unifying information from diverse data sources. Specifically, the server uses OCR technology (e.g., Tesseract OCR) and parser libraries (e.g., Python's docx library and openpyxl library) to convert these documents into digital data and organize them into a unified format.

[1512] Extracting information and storing it in a database

[1513] The server analyzes the documents converted into a unified format using natural language processing techniques (e.g., Stanford NLP or spaCy) to extract important keywords and phrases. This information includes project status, decisions, and issues. The extracted information is classified as metadata and stored in a relational database (e.g., MySQL, PostgreSQL) or a NoSQL database (e.g., MongoDB). This creates a database that allows quick access to the information needed.

[1514] Building a question-answering system

[1515] The server provides the device with an interface through which the user can input a question. This interface can be web-based or app-style and is designed to allow the user to input a question in natural language. Once the user inputs a question, the device sends the question to the server. The server then uses a machine learning model (e.g., GPT-3) or a natural language processing algorithm to analyze the question and convert it into an internal query. This allows the optimal answer to the user's question to be quickly generated.

[1516] Emotion engine recognizes emotions and adjusts responses

[1517] During the question analysis process, the server uses an emotion engine (e.g., Microsoft Azure Cognitive Services' Emotion Analysis API) to recognize the user's emotions. Emotion recognition can identify the user's emotional state, such as stress, anxiety, or excitement. Based on the results of emotion recognition, the server can adjust the content and expression of the answer. For example, if the user is stressed, the server can make the answer more polite and add expressions that give a sense of security.

[1518] Generate and provide answers

[1519] The server generates a summary based on the information extracted from the database and creates a specific answer, including content adjusted by the emotion engine. The generated answer is sent to the device and displayed to the user, allowing the user to make effective decisions based on the displayed information.

[1520] Specific examples

[1521] Example 1: Finding new campaign ideas

[1522] 1. The user types into their device, "I'm having trouble with a new campaign. What should I do?"

[1523] 2. The device sends this question to the server.

[1524] 3. The server uses natural language processing to analyze the question and identify keywords such as "new campaign," "concerns," and "past success stories."

[1525] 4. The server uses an emotion engine to determine that the user is experiencing stress due to their concerns.

[1526] 5. The server extracts relevant information from the database (such as examples of past campaign successes and failures, opinions of senior management, etc.) and generates a summary that takes into account the user's emotional state.

[1527] 6. The device displays the response from the server to the user, for example, "Project X is an example of a successful campaign in the past, and method Y was effective in this project. Please stay calm and proceed."

[1528] Example 2: Checking the progress of a project

[1529] 1. The user types into the terminal, "Tell me about the progress of Project A."

[1530] 2. The device sends this question to the server.

[1531] 3. The server parses the question and extracts the latest information related to Project A from the database.

[1532] 4. The server uses an emotion engine to generate quick and concise answers if the user is in a hurry.

[1533] 5. The server summarizes the extracted information and generates a summary that takes into account the user's emotional state, for example, "Project A is currently in stage 3, with key tasks 70% complete. The next step is X."

[1534] 6. The terminal displays the response from the server to the user.

[1535] In this way, a more appropriate and user-friendly system can be realized by providing information according to the user's emotional state, allowing users to quickly and efficiently retrieve past information and use it to plan new projects and campaigns.

[1536] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1537] Step 1:

[1538] A user logs into a terminal and begins the process of gathering documents related to past projects.

[1539] Input: User credentials

[1540] Output: System access permissions

[1541] What happens: The user enters their username and password, which the system authenticates.

[1542] Step 2:

[1543] The server collects past documents from the company's network drive or cloud storage.

[1544] Input: Path information for network drive or cloud storage

[1545] Output: A list of collected documents

[1546] Specific operation: The server scans the specified path for PDF, Word, Excel, and other files and saves them in temporary storage.

[1547] Step 3:

[1548] The server converts the collected documents into a unified format.

[1549] Input: Collected documents

[1550] Output: Data converted into a unified format

[1551] Specific operation: For PDF files, it uses Tesseract OCR to convert them to text data, and for Word and Excel files, it uses Python's docx and openpyxl libraries to extract the text data.

[1552] Step 4:

[1553] The server uses natural language processing techniques to extract key information from the documents that are converted into a unified format.

[1554] Input: Data converted into a unified format

[1555] Output: Extracted keywords and phrases

[1556] What it does: Tokenizes and tags parts of speech using Stanford NLP and spaCy to extract keywords such as project status and decisions.

[1557] Step 5:

[1558] The server stores the extracted information in a database.

[1559] Input: Extracted keywords or phrases

[1560] Output: Metadata stored in a database

[1561] Specific behavior: Stores information in a relational database (MySQL, PostgreSQL) or a NoSQL database (MongoDB). Fields include "Project name," "Progress," and "Decisions."

[1562] Step 6:

[1563] The user inputs a question through the terminal.

[1564] Input: User question

[1565] Output: Question data sent to the terminal

[1566] Specific operation: The user enters a question in natural language and clicks the "Send" button. The device sends this question to the server.

[1567] Step 7:

[1568] The server analyzes the user's question and searches the database for relevant information.

[1569] Input: User question

[1570] Output: Inner query and search results

[1571] What it does: It uses a machine learning model (e.g., GPT-3) to analyze the question and generate an internal query, which it then uses to search for relevant information in a database.

[1572] Step 8:

[1573] The server recognizes the user's emotional state based on the search results.

[1574] Input: User questions and search results

[1575] Output: Determining the user's emotional state

[1576] Specific behavior: Uses an emotion engine (e.g., Microsoft Azure Cognitive Services emotion analysis API) to recognize the user's emotional state from the questions.

[1577] Step 9:

[1578] The server adjusts the content and expression of the response based on the results of the emotion engine.

[1579] Input: User's emotional state and search results

[1580] Output: Adjusted answer

[1581] Specific behavior: Depending on the user's emotional state, the response may be more polite or specific. For example, a user in a stressful state may be given a reassuring response.

[1582] Step 10:

[1583] The server generates the final answer to provide to the user.

[1584] Input: Adjusted answer content

[1585] Output: Final answer

[1586] Specific actions: Integrate search results with tailored content to generate summaries, flesh out answers, and format them to appropriately respond to the user's question.

[1587] Step 11:

[1588] The terminal displays the final answer to the user.

[1589] Input: Final answer

[1590] Output: The answer displayed to the user

[1591] Specific operation: The final answer sent from the server is received and displayed on the user's device, allowing the user to view it and use it as a reference for making decisions.

[1592] (Application example 2)

[1593] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1594] Conventional systems have had difficulty efficiently collecting and analyzing documents related to past projects and security-related logs, and providing relevant information to users. Furthermore, they lacked the ability to recognize the user's emotional state and adjust the appropriate response content and expressions accordingly, making it difficult to improve user satisfaction. Therefore, there was a need for a system that could efficiently collect and analyze information and provide appropriate information based on the user's emotional state.

[1595] The identification process by the identification 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 documents related to past projects, means for converting the collected documents into a unified format, means for extracting important information using natural language processing technology, means for storing the extracted information in a database, means for analyzing questions from users and searching the database for related information, means for summarizing the searched information and providing it to the user, means for providing an interface for users to input questions, and means for combining an emotion engine that recognizes the user's emotions and adjusts the content and expression of the response. This makes it possible to efficiently collect and analyze information related to past projects and security, and to provide appropriate information based on the user's emotional state.

[1596] "Documents related to past projects" are documents such as minutes, reports, and materials related to projects that a company or organization has carried out in the past.

[1597] A "uniform format" is a standardized format that converts collected documents into the same format to facilitate analysis and search.

[1598] "Natural language processing technology" is a technology that enables computers to understand and analyze human language, and extract meaningful information from text data.

[1599] "Key information extraction means" refers to methods and algorithms that use natural language processing techniques to identify and extract key information, such as project status, decisions, and issues, from collected documents.

[1600] A "database" is a system for storing and managing extracted information, a collection of data structured to facilitate search and access.

[1601] "Means for analyzing a question from a user and searching for related information from a database" refers to a method or algorithm for analyzing a question entered by a user and searching for information related to that question from a database.

[1602] "Means for summarizing retrieved information and providing it to the user" refers to a method or algorithm that succinctly summarizes relevant information extracted from a database and provides it to the user in an easily understandable format.

[1603] The "interface through which the user enters a question" is the part of the software that provides a screen or input field into which the user can enter a question in text form.

[1604] An "emotion engine" is a technology that analyzes the user's emotional state from the text they input and adjusts the response content and expression appropriately based on that.

[1605] "Security-related logs and event reports" are documents or data files that record security incidents and other security-related information within a company or organization.

[1606] A "generative AI model" is an artificial intelligence model that uses pre-trained algorithms to generate appropriate responses or prompts to user questions.

[1607] A "prompt sentence" is a specific question or instruction sentence that is generated to be easy for the user to understand.

[1608] This system collects documents related to past projects and security-related logs and event reports, converts them into a unified format, analyzes them, and provides appropriate answers to user questions. It also has the ability to recognize the user's emotional state and adjust the content and expression of the answers.

[1609] Data collection and integration

[1610] The server collects documents related to past projects and security-related logs from the company's network drive or cloud storage. Because these documents exist in various formats, such as PDF, Word, and image files, the server converts them into digital data using OCR technology (e.g., the pytesseract library) and parsers, and organizes them into a unified format.

[1611] Extracting information and storing it in a database

[1612] The server then uses natural language processing techniques (e.g., the transformers library) to analyze the documents converted into a unified format and extract important keywords and phrases. This information includes project status, decisions, issues, the date and time of security incidents, and the scope of their impact. The extracted information is classified as metadata and stored in a relational or NoSQL database.

[1613] Building a question-answering system

[1614] The server provides an interface on the terminal for users to input questions. This interface runs on terminals such as smartphones and PCs and is designed to allow users to input questions in natural language. When a question is input by the user, the terminal sends the question to the server.

[1615] Emotion engine recognizes emotions and adjusts responses

[1616] When analyzing a user's question, the server uses an emotion engine to recognize the user's emotions. This emotion recognition allows the server to identify the user's emotional state. For example, if the user is stressed, the server changes the answer to be more polite and reassuring.

[1617] Generate and provide answers

[1618] The server generates a summary based on the information extracted from the database and creates a specific response for the user, using the response adjusted by the emotion engine. The generated response is sent to the terminal and displayed to the user.

[1619] Specific examples

[1620] Example 1: Asking about the last security incident

[1621] A user types "When was the last security incident?" into a smartphone app. The app sends the question to the server, which analyzes the question, extracts relevant information from a database, and generates an answer. The app then displays the answer from the server and the emotion recognition results. For example, it might say, "The last security incident occurred on December 1, 2022. If you would like to know more about the countermeasures, please continue asking questions."

[1622] Prompt Sentence Examples

[1623] "Please provide a comprehensive analysis of the security logs and provide a detailed summary of the last security incident."

[1624] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1625] Step 1:

[1626] The server collects past project-related documents and security logs from the company's network drive or cloud storage. To collect documents, the server accesses pre-configured paths or API endpoints and downloads data in various formats, such as PDF, Word, Excel, and image files. This creates a document collection. The input is the network drive or cloud storage, and the output is a list of collected document files.

[1627] Step 2:

[1628] The server converts the collected documents into a unified format using OCR technology and a parser. Specifically, it uses the pytesseract library to extract text data from image files, and converts PDF and Word documents into text data using an appropriate parser. This converts documents in various formats into a unified text format. The input is the collected document files, and the output is the converted text data.

[1629] Step 3:

[1630] The server uses natural language processing technology to analyze the text data converted into a unified format and extract important keywords and phrases. At this stage, the transformers library is used to perform text analysis and generate metadata such as project progress, decisions, issues, and the date and time of the security incident and its scope of impact. The input is the converted text data, and the output is the extracted metadata.

[1631] Step 4:

[1632] The server stores the extracted metadata in a relational or NoSQL database. Specifically, it organizes and stores the metadata in tables or collections for future searches and queries. The input is the extracted metadata, and the output is the results stored in the database.

[1633] Step 5:

[1634] The terminal provides an interface for users to input questions. Users input questions in natural language on the screen of their smartphone or PC, and the questions are sent to the server via the interface. The input is the user's text question, and the output is the result of sending the question data to the server.

[1635] Step 6:

[1636] The server analyzes the received user question using natural language processing technology and an emotion engine. Through the analysis, it identifies keywords in the question and the user's emotional state and converts them into an appropriate internal query. The input is the user's question data, and the output is the analysis result and the internal query.

[1637] Step 7:

[1638] The server uses the internal query to search for relevant information from a database. The search extracts appropriate information for the question and tailors the answer to the user's emotional state. The input is the internal query, and the output is the search results and tailored answer data.

[1639] Step 8:

[1640] The server uses a generative AI model to summarize and optimize answers to complete a specific answer for the user. It generates prompt sentences and provides answers that take the user's emotional state into account. The input is search results and adjusted answer data, and the output is a completed answer. Using a generative AI model improves the quality and appropriateness of answers.

[1641] Step 9:

[1642] The terminal displays the answers received from the server to the user. The user can check the detailed answers on the screen and use this interface to enter additional questions. The input is the answer data from the server, and the output is the answer display result to the user.

[1643] Example prompt sentence:

[1644] "Please provide a comprehensive analysis of the security logs and provide a detailed summary of the last security incident."

[1645] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1646] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1647] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1648] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1649] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1650] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1651] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1652] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1653] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1654] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1655] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1656] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1657] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1658] 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.

[1659] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1660] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1661] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1662] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1663] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1664] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1665] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1666] The following is further disclosed regarding the above embodiment.

[1667] (Claim 1)

[1668] A means of collecting documentation related to past projects;

[1669] A means of converting collected documents into a unified format;

[1670] a means of extracting key information using natural language processing techniques;

[1671] means for storing the extracted information in a database;

[1672] A means for analyzing a user's question and searching a database for relevant information;

[1673] means for summarizing the retrieved information and providing it to the user;

[1674] means for providing an interface for a user to input a question;

[1675] A system including:

[1676] (Claim 2)

[1677] 10. The system of claim 1, wherein the collected documents are converted into digital data using OCR technology.

[1678] (Claim 3)

[1679] 2. The system according to claim 1, which organizes data related to past projects by project and links related information together.

[1680] "Example 1"

[1681] (Claim 1)

[1682] A means of collecting materials related to past business projects;

[1683] A means of converting collected materials into a unified format;

[1684] a means of extracting key information using machine learning techniques;

[1685] means for storing the extracted information in an information storage device;

[1686] means for analyzing a question from a user and retrieving related information from an information storage device;

[1687] means for summarizing the retrieved information and providing it to the user;

[1688] means for providing an input device for a user to input a question;

[1689] A system including:

[1690] (Claim 2)

[1691] 10. The system of claim 1, wherein the collected materials are converted into digital data using character recognition technology.

[1692] (Claim 3)

[1693] 2. The system according to claim 1, which organizes data related to past business projects by project and links related information together.

[1694] (Claim 4)

[1695] 10. The system of claim 1, wherein a natural language processing model is used to parse a user's question and convert it into an internal query.

[1696] (Claim 5)

[1697] 10. The system of claim 1, further comprising: searching for relevant information using a high-speed search engine based on the analyzed query.

[1698] (Claim 6)

[1699] 10. The system of claim 1, wherein the system uses data stored in the information storage device to generate a summary corresponding to a question entered by a user.

[1700] (Claim 7)

[1701] 10. The system of claim 1, further comprising a user interface that operates on the device to allow a user to quickly obtain information.

[1702] "Application Example 1"

[1703] (Claim 1)

[1704] A means of collecting documentation related to past projects;

[1705] A means of converting collected documents into a unified format;

[1706] a means of extracting key information using natural language processing techniques;

[1707] means for storing the extracted information in a database;

[1708] A means for analyzing a user's question and searching a database for relevant information;

[1709] means for summarizing the retrieved information and providing it to the user;

[1710] means for providing an interface for a user to input a question;

[1711] A means for store staff to search and obtain past customer service history and sales data,

[1712] A means for generating appropriate answers to questions from store staff using a question-answering model;

[1713] A system including:

[1714] (Claim 2)

[1715] 10. The system of claim 1, wherein the collected documents are converted into digital data using OCR technology.

[1716] (Claim 3)

[1717] 2. The system according to claim 1, which organizes data related to past projects by project and links related information together.

[1718] "Example 2: Combining Emotion Engines"

[1719] (Claim 1)

[1720] A means of collecting documentation related to past projects;

[1721] A means of converting collected documents into a unified format;

[1722] a means of extracting key information using natural language processing techniques;

[1723] means for storing the extracted information in a database;

[1724] A means for analyzing a user's question and searching a database for relevant information;

[1725] means for summarizing the retrieved information and providing it to the user;

[1726] means for providing an interface for a user to input a question;

[1727] an emotion recognition means for recognizing an emotional state of a user in response to a question input by the user and adjusting the content and expression of the answer;

[1728] A system including:

[1729] (Claim 2)

[1730] 10. The system of claim 1, wherein the collected documents are converted into digital data using OCR technology.

[1731] (Claim 3)

[1732] 2. The system according to claim 1, which organizes data related to past projects by project and links related information together.

[1733] (Claim 4)

[1734] 10. The system of claim 1, wherein a generative AI model is used to analyze a user's input question and convert it into an appropriate internal query.

[1735] (Claim 5)

[1736] 2. The system of claim 1, wherein the content and expression of the response are adjusted based on the user's emotional state.

[1737] "Application example 2 when combining emotion engines"

[1738] (Claim 1)

[1739] A means of collecting documentation related to past projects;

[1740] A means of converting collected documents into a unified format;

[1741] a means of extracting key information using natural language processing techniques;

[1742] means for storing the extracted information in a database;

[1743] A means for analyzing a user's question and searching a database for relevant information;

[1744] means for summarizing the retrieved information and providing it to the user;

[1745] means for providing an interface for a user to input a question;

[1746] A method that combines an emotion engine that recognizes the user's emotions and adjusts the content and expression of the answers;

[1747] A system including:

[1748] (Claim 2)

[1749] 10. The system of claim 1, wherein the collected documents are converted into digital data using OCR technology.

[1750] (Claim 3)

[1751] 2. The system according to claim 1, which organizes data related to past projects by project and links related information together.

[1752] (Claim 4)

[1753] 10. The system of claim 1, wherein security-related logs and event reports are collected, converted into a unified format, and analyzed.

[1754] (Claim 5)

[1755] The system according to claim 1 uses a generative AI model that analyzes emotions and generates appropriate answers when a user inputs a question.

[1756] (Claim 6)

[1757] 6. The system according to claim 5, wherein when a security administrator inputs a question about a past incident, the system generates a prompt sentence that extracts relevant information from logs and reports, summarizes it, and provides it. [Explanation of symbols]

[1758] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of collecting documentation related to past projects; A means of converting collected documents into a unified format; a means of extracting key information using natural language processing techniques; means for storing the extracted information in a database; A means for analyzing a user's question and searching a database for relevant information; means for summarizing the retrieved information and providing it to the user; means for providing an interface for a user to input a question; A system including:

2. 10. The system of claim 1, wherein the collected documents are converted into digital data using OCR technology.

3. 2. The system according to claim 1, wherein data relating to past projects is organized by project unit, and related information is linked together.

Images