system

Abstract

We provide the system. [Solution] Means of obtaining data related to laws and precedents from online resources, A means of analyzing acquired data using natural language processing technology and storing it in a database, A means for retrieving relevant information from a database and generating a set of questions based on information about the person being interrogated entered by the user, A means of receiving the suspect's response data and comparing it with legal and case law data to detect inconsistencies and inconsistencies, A system that includes means for generating reports based on analysis results and presenting them to the user.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In an interrogation, it is not easy for a police officer to analyze the suspect's statements in detail and detect contradictions or unnatural points. Therefore, in an actual interrogation, important information may be overlooked. Also, in order to improve the efficiency of interrogations, a system that can immediately refer to vast knowledge of laws and precedents is required. To solve these problems, a system that automatically acquires and analyzes knowledge of laws and precedents and detects contradictions in the suspect's statements and presents them to police officers is needed.

Means for Solving the Problems

[0005] The present invention provides a system that includes means for obtaining data related to laws and precedents from online resources, means for analyzing the obtained data using natural language processing technology and storing it in a database, means for obtaining relevant information from the database and generating a set of questions based on information about the person being interrogated entered by the user, means for receiving the suspect's response data and detecting inconsistencies and unnatural points by comparing it with legal and precedent data, and means for generating a report based on the analysis results and presenting it to the user. This system allows police officers to immediately check for inconsistencies in a suspect's statements during interrogation, thereby improving the efficiency and accuracy of interrogations.

[0006] "Data related to laws and precedents" refers to information including laws, ordinances, court judgments, and related interpretations and comments.

[0007] "Online resources" refer to databases, websites, APIs, or similar information sources that are accessible over the internet.

[0008] "Natural language processing technology" refers to techniques and methods that enable computers to understand, analyze, and generate human language, and includes technologies such as text mining, sentiment analysis, and topic modeling.

[0009] A "database" refers to an information system for storing structured information and for efficiently searching and managing it.

[0010] "User" refers to the human operators, such as police officers or interrogators, who operate this system.

[0011] "Persons under investigation" refers to suspects or witnesses who are being questioned.

[0012] A "question set" refers to a series of questions used in an interrogation, and is generated based on a specific purpose.

[0013] "Response data" refers to the oral or written answers provided by the person being interrogated in response to questions.

[0014] "Matching" refers to the process of comparing one piece of data with another to find matching or contradictory parts.

[0015] "Inconsistencies or unnatural points" refer to parts of the interrogated person's statements that do not match other data, or parts that are not logically consistent.

[0016] A "report" refers to a document or digital report summarizing the analysis results, and includes information useful for investigations. [Brief explanation of the drawing]

[0017] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Embodiment 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Embodiment 2 when the emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when the emotion engine is combined.

Mode for Carrying Out the Invention

[0018] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

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

[0020] In the following embodiments, a numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), etc.

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

[0022] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0023] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0024] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0025] [First Embodiment]

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

[0027] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0028] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0029] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0030] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0031] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0032] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

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

[0034] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0035] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0036] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0037] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0038] The system according to the present invention is for automatically acquiring, analyzing, and using data related to laws and precedents for investigation purposes. This system consists of a server, terminals, and users.

[0039] Knowledge acquisition and analysis

[0040] 1. Data Collection

[0041] The server accesses online resources and collects data related to laws and precedents. This includes information from government legal databases and court case databases.

[0042] 2. Data Analysis

[0043] The data collected by the server is analyzed using natural language processing techniques. Specifically, techniques such as text mining and topic modeling are applied to understand the content of the data.

[0044] 3. Storing in a database

[0045] The server analyzes the data and stores it in a database. This enables efficient searching and management.

[0046] Preparation for interrogation

[0047] 4. User login

[0048] The user logs into the system using their device. User authentication information is sent from the device to the server.

[0049] 5. Entering information about the person being interrogated.

[0050] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, and a summary of the case.

[0051] 6. Generating the Question Set

[0052] The server retrieves relevant information from the database and generates a set of questions. For example, in the case of an investigation into a traffic accident, questions based on traffic laws and relevant precedents will be generated.

[0053] Conducting an interrogation

[0054] 7. Commencement of Interrogation

[0055] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in either audio or text format.

[0056] 8. Data transmission and analysis

[0057] The terminal sends the recorded response data to the server. After receiving the response data, the server analyzes it by comparing it with legal and case law data that has been collected in advance.

[0058] 9. Detection of inconsistencies

[0059] The server uses natural language processing technology to analyze the content of the response and detect inconsistencies and unnatural points. For example, it might detect if a statement like "The car in front of me suddenly stopped at the intersection" contradicts the circumstantial evidence at the scene.

[0060] 10. Report generation

[0061] The server generates a report based on the analysis results. The report includes details about any inconsistencies or inconsistencies detected.

[0062] Example: Investigation of a car accident

[0063] As a concrete example, let's explain the investigation in a car accident.

[0064] 1. Data Collection

[0065] The server collects and analyzes case law data related to traffic laws and automobile accidents, and stores it in a database.

[0066] 2. User login

[0067] The user logs into the system using a terminal and enters the information of Mr. Tanaka, the person being investigated.

[0068] 3. Generating the Question Set

[0069] The server retrieves information about traffic accidents related to Mr. Tanaka and generates an appropriate set of questions.

[0070] 4. Conducting the interrogation

[0071] The user asks Mr. Tanaka, "Could you tell me the details of what happened in the accident?" Mr. Tanaka replies, "The car in front of me stopped suddenly at the intersection, so I also slammed on the brakes."

[0072] 5. Data transmission and analysis

[0073] The terminal records Mr. Tanaka's responses in audio or text format and sends them to the server. The server analyzes the data and detects any inconsistencies.

[0074] 6. Report generation

[0075] Based on the analysis results, the server generates a report pointing out inconsistencies between Mr. Tanaka's statements and the circumstantial evidence at the scene, and presents it to the user via the terminal.

[0076] As described above, the system according to the present invention improves the efficiency of police officers' work and enhances the accuracy of interrogations by acquiring, analyzing, and supporting legal and case law data.

[0077] The following describes the processing flow.

[0078] Step 1:

[0079] Data collection

[0080] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It can also collect data from publicly available websites using web scraping techniques.

[0081] Step 2:

[0082] Data Analysis

[0083] The server analyzes the collected data using natural language processing techniques. These include techniques such as text mining, sentiment analysis, and entity extraction. The analyzed data is categorized and organized by content.

[0084] Step 3:

[0085] Storage in database

[0086] The server stores the analyzed data in a database. Specifically, information related to laws is stored in the "Legal Matters" category, and information related to case law is stored in the "Case Law" category. Furthermore, indexes are created to streamline searching and query processing.

[0087] Step 4:

[0088] User login

[0089] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server performs authentication and grants permission to log in.

[0090] Step 5:

[0091] Entering information about the person being interrogated

[0092] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, address, and a summary of the case. The terminal then transmits this information to the server.

[0093] Step 6:

[0094] Question set generation

[0095] The server retrieves information related to the interrogation from the database and generates a set of questions. For example, if the person being interrogated is involved in a traffic accident, questions will be generated based on traffic laws and relevant past court precedents. The generated set of questions is then sent to the terminal.

[0096] Step 7:

[0097] Start of interrogation

[0098] The user uses a device to ask the person being interrogated system-generated questions. The person being interrogated's responses are recorded by the device in either audio or text format.

[0099] Step 8:

[0100] Sending response data

[0101] The device sends the recorded response data to the server. The data sent includes the interviewee's voice data, text data, or both.

[0102] Step 9:

[0103] Analysis and matching of response data

[0104] The server analyzes the received response data using natural language processing technology. The analyzed response data is then compared with legal and case law data stored in a database beforehand to detect inconsistencies and unnatural points.

[0105] Step 10:

[0106] Report generation

[0107] The server generates a report based on the analysis results. The report includes details about inconsistencies and unnatural points in the interrogated person's statements. The generated report is sent to the terminal and presented to the user.

[0108] Step 11:

[0109] Confirmation of results and further questioning

[0110] The user reviews the report using a terminal. If necessary, the user creates additional questions and conducts further investigations based on these questions. The additional questions are sent to the server for further analysis.

[0111] By repeating these steps, the accuracy and efficiency of interrogations can be improved.

[0112] (Example 1)

[0113] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0114] The current interrogation system requires a tremendous amount of time and effort because data collection and analysis based on laws and precedents are done manually. Furthermore, the quality and accuracy of the question sets generated during interrogations are low, making it difficult to efficiently detect inconsistencies or unnatural points in the interviewee's responses. Therefore, there is a need to improve the accuracy and efficiency of interrogations.

[0115] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0116] In this invention, the server includes means for acquiring data related to laws and precedents from online resources, means for analyzing the acquired data using natural language processing technology and storing it in a database, means for acquiring relevant information from the database based on information about the subject of investigation entered by the user and generating a set of questions using a generative AI model, means for receiving the subject's response data and detecting inconsistencies and unnatural points by comparing it with the data on laws and precedents, and means for generating a report based on the analysis results and presenting it to the user. This not only improves the quality of interrogations but also enables highly accurate and rapid interrogations.

[0117] "Online resources" refer to websites and databases that exist on the internet and contain data on laws and precedents provided by public institutions, companies, and other organizations.

[0118] "Legal and case law data" refers to digital data provided by government agencies and courts in various countries, including legal texts and the content of past court precedents.

[0119] "Natural language processing technology" refers to techniques that enable computers to understand and analyze human language, and includes methods such as text mining and topic modeling.

[0120] A "database" is a system for efficiently storing, managing, and retrieving large amounts of data, and includes relational databases and NoSQL databases.

[0121] A "subject of investigation" is a person who is the subject of questioning or investigation, and their basic information and information related to the case are entered into the system.

[0122] A "generative AI model" is a model that uses AI technology to automatically generate question sets and reports based on given input data. Examples include GPT-3 (registered trademark) and similar models.

[0123] A "question set" refers to a series of questions asked of the person being investigated, and is automatically generated by an AI model according to the purpose of the interrogation or investigation.

[0124] "Response data" refers to the data of responses provided by survey participants to questions, and is recorded in audio or text format.

[0125] "Inconsistencies or unnatural points" refer to parts of the survey respondents' responses that do not match or are logically inconsistent with the legal and case law data collected and analyzed in advance.

[0126] A "report" is a document that summarizes the analysis results and includes details about any inconsistencies or inconsistencies that were detected.

[0127] This invention relates to a system for automatically acquiring and analyzing data related to laws and precedents, and for efficiently conducting interrogations. This system consists of a server, terminals, and users.

[0128] Knowledge acquisition and analysis

[0129] 1. Data Collection

[0130] The server collects data related to laws and precedents from online resources on the internet. For example, it accesses government-provided legal databases and court case databases. Specifically, it uses Python crawling tools (e.g., Scrapy) to periodically crawl these databases and retrieve the necessary information.

[0131] 2. Data Analysis

[0132] The server analyzes the collected data using natural language processing (NLP) techniques. It applies text mining libraries (e.g., NLTK, SpaCy) and topic modeling techniques (e.g., LDA) to understand the data's content and extract relevant information.

[0133] 3. Storing in a database

[0134] The server stores the analyzed data in a database. Because fast and efficient data management is required here, relational databases (e.g., MySQL®, PostgreSQL) or NoSQL databases (e.g., MongoDB) are used.

[0135] Preparation for interrogation

[0136] 4. User login

[0137] Users log in to the system using their device. User authentication is performed by sending data from the device to the server, where security measures, including two-factor authentication, are implemented and the authentication process is completed.

[0138] 5. Entering information about the person being interrogated.

[0139] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, and a summary of the case. The entered data is stored in a database on the server.

[0140] 6. Generating the Question Set

[0141] The server retrieves relevant information from the database and generates a set of questions using a generative AI model (e.g., GPT-3). The generated set of questions includes the most appropriate questions for the research subject, based on the legal case.

[0142] Conducting an interrogation

[0143] 7. Commencement of Interrogation

[0144] The user asks the interviewee questions generated using the device. The interviewee's responses are recorded on the device in either audio or text format, and if audio data is used, it is converted to text using speech recognition technology (e.g., Google® Cloud Speech-to-Text).

[0145] 8. Data transmission and analysis

[0146] The terminal sends the recorded response data to the server. The server compares the received data with legal and case law data that has been collected and analyzed in advance, and performs analysis using NLP technology (e.g., BERT).

[0147] 9. Detection of inconsistencies

[0148] The server evaluates the responses based on the analysis results and detects inconsistencies and unnatural points. It also checks whether the statements are consistent with the facts by cross-referencing them with legal databases.

[0149] 10. Report generation

[0150] The server generates a report based on the analysis results and presents it to the user. The report includes details about any inconsistencies or unnatural points detected.

[0151] Examples of specific cases and prompt statements

[0152] As a concrete example, let's consider the investigation of a traffic accident.

[0153] 1. The server collects and analyzes case law data related to traffic laws and automobile accidents, and stores it in a database.

[0154] 2. The user logs into the system using a terminal and enters the information of Mr. Tanaka, the person being investigated.

[0155] 3. The server retrieves information about traffic accidents related to Mr. Tanaka and generates an appropriate set of questions.

[0156] Example of a prompt:

[0157] "Please generate a set of questions for an investigation into a traffic accident. The subject of the investigation is Mr. Tanaka, and we would like to ask about the detailed circumstances of the accident. Please include relevant traffic laws and precedents in the questions."

[0158] As described above, the system according to the present invention aims to improve the quality and efficiency of interrogations through the automatic acquisition, analysis, and examination of legal and case law data.

[0159] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0160] Step 1: Data Collection

[0161] The server collects data related to laws and precedents from online resources on the internet. Specifically, it uses Python crawling tools (e.g., Scrapy) to crawl government law databases and court precedent databases. The input is a list of URLs to be crawled, and the output is the retrieved HTML code and text data. The retrieved data is temporarily stored.

[0162] Specific operation: The server executes a Scrapy script, accesses the URL of the specified legal database, and collects the latest case law data.

[0163] Step 2: Data Analysis

[0164] The server analyzes the collected data using natural language processing (NLP) techniques. Specifically, it extracts keywords and topics from the collected text data using text mining libraries (e.g., NLTK, SpaCy). The input is the collected text data, and the output is the analyzed keyword and topic information.

[0165] Specific operation: The server uses the NLTK library to analyze legal terms contained in the collected data and extract important keywords and their relevance.

[0166] Step 3: Storing in the database

[0167] The server stores the analyzed data in a database. This can be a relational database (e.g., MySQL) or a NoSQL database (e.g., MongoDB). The input is the analyzed data, and the output is the completed storage of the data in the database.

[0168] Specific operation: The server stores the parsed data into MySQL using an INSERT statement.

[0169] Step 4: User Login

[0170] The user logs into the system using a terminal. The user ID and password are input data, which are sent to the server. The server compares this information with the authentication information in the database and sends the authentication result back to the terminal. The output indicates whether the login was successful or failed.

[0171] Specific operation: The user enters their ID and password on the login screen, and the server verifies this against the database information to perform authentication.

[0172] Step 5: Enter information about the person being investigated.

[0173] The user enters basic information about the person being interrogated into the terminal. This includes the person's name, age, and a summary of the case. The entered data is sent to the server, which stores it in a database. The output includes confirmation that the data has been successfully saved to the database.

[0174] Specific operation: The user enters Mr. Tanaka's information (name, age, details of the incident) into an input form, and the server receives this information and saves it to the database.

[0175] Step 6: Generate the question set

[0176] The server retrieves relevant information from the database and generates a set of questions using a generative AI model (e.g., GPT-3). Inputs include basic information about the person being interrogated and relevant legal and case law data, while output is the generated set of questions.

[0177] Specific operation: The server sends a prompt to the GPT-3 model saying "Generate a set of questions for a traffic accident investigation," retrieves the response, and constructs the question set.

[0178] Step 7: Start of interrogation

[0179] The user uses a device to ask questions generated by the device to the person being interviewed. The interviewee's responses are recorded on the device in either audio or text format. The input consists of the generated set of questions and the interviewee's responses, and the output is the recorded response data.

[0180] Specific operation: The user asks Mr. Tanaka, "Please tell me the details of what happened when the accident occurred," records Mr. Tanaka's voice response on the device, and converts the voice data into text using Google Cloud Speech-to-Text.

[0181] Step 8: Data transmission and analysis

[0182] The terminal sends the recorded response data to the server. The server compares the received data with pre-collected and analyzed legal and case law data and performs analysis using natural language processing techniques (e.g., BERT). The input consists of the response data and related legal and case law data, and the output is the analysis results.

[0183] Specific operation: The terminal sends the text-based response to the server, and the server performs contextual analysis of the response using a BERT model.

[0184] Step 9: Detecting inconsistencies

[0185] The server evaluates the answers based on the analysis results and detects inconsistencies and unnatural points. The input is the analysis results, and the output is the detected inconsistencies and unnatural points.

[0186] Specific operation: The server compares the case data for the relevant case with the response content and detects if the statement "the car in front suddenly stopped at the intersection" contradicts the situation at the accident scene.

[0187] Step 10: Generate the report

[0188] The server generates a report based on the analysis results and presents it to the user. The input consists of the analysis results and detected inconsistencies, while the output is the generated report.

[0189] Specific operation: The server compiles the analysis results, generates a report pointing out inconsistencies, and sends this report to the terminal in PDF format.

[0190] (Application Example 1)

[0191] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0192] In on-site interrogations and investigations, it is difficult for security personnel to ask efficient and accurate questions based on laws and precedents, and to immediately detect inconsistencies or suspicious points in answers. While rapid information acquisition and analysis on-site are required, conventional methods lack real-time capabilities, leading to decreased accuracy and efficiency in investigations. To address this challenge, there is a need for a system that effectively and quickly utilizes legal and precedent data to support on-site interrogations.

[0193] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0194] In this invention, the server includes means for acquiring data related to laws and precedents from online resources; means for analyzing the acquired data using natural language processing technology and storing it in a database; means for acquiring relevant information from the database based on information about the person being interrogated entered by the user and generating a set of questions; means for a security officer to display the questions using a head-mounted display and record the answers using speech recognition technology; means for receiving the recorded answer data and comparing it with the data on laws and precedents to detect inconsistencies and inconsistencies; and means for generating a report based on the analysis results and presenting it to the security officer. This enables rapid and accurate interrogations on-site.

[0195] "Online resources" is a general term for information sources such as databases and websites that exist on the internet.

[0196] A "law" is a set of rules and norms established by a state or local government that have the power to enforce order in society.

[0197] A "precedent" is a record of judgments and their reasons that have been handed down by courts in the past, and it serves as a reference for similar cases that may follow.

[0198] "Natural language processing technology" refers to techniques for understanding, analyzing, and generating human language using computers, and examples include text mining and topic modeling.

[0199] A "database" is a general term for a system that organizes and stores large amounts of data, and allows for efficient searching and updating.

[0200] A "person under investigation" refers to an individual who is investigated or questioned in order to provide information about a specific incident or situation.

[0201] A "question set" is a collection of related questions grouped together to achieve a specific objective.

[0202] A "security implementer" refers to an individual or organization that takes on the role of ensuring safety in a specific location or situation.

[0203] A "head-mounted display" is a type of display device worn on the head that displays information within the field of vision.

[0204] "Speech recognition technology" is a technology that analyzes speech data and converts it into text data.

[0205] A "contradiction" refers to a part of specific data or statements that is not logically consistent.

[0206] An "unnatural point" refers to a part of specific data or statements that is judged to deviate from what is normal or common sense.

[0207] A "report" is a document that summarizes analysis results or research findings, and is a type of report.

[0208] The system of this invention collects data related to laws and precedents from online resources, analyzes it using natural language processing technology, and generates question sets and analyzes answers for on-site interrogations and investigations. This system mainly consists of a server, a head-mounted display, and a user.

[0209] Collection of legal and case law data

[0210] The server automatically retrieves data related to laws and precedents from databases and websites on the internet. For example, government legal databases and court case databases are used.

[0211] Data analysis and database storage

[0212] The acquired data is analyzed on the server using natural language processing techniques. Specifically, techniques such as text mining and topic modeling are applied to extract important topics and case precedents. The analysis results are stored in a database, enabling efficient searching and management.

[0213] Question set generation

[0214] The user logs into the system via a head-mounted display and enters basic information about the person being interrogated. The server retrieves relevant information from the database and automatically generates an appropriate set of questions. This enables rapid and accurate interrogations on-site.

[0215] Speech recognition and analysis of responses

[0216] Security personnel use a head-mounted display to show generated questions and ask them to the person being interrogated. Responses are recorded in real time using speech recognition technology. The recorded data is sent to a server where it is compared with legal and case law data to detect inconsistencies and inconsistencies.

[0217] Report generation and presentation

[0218] The server generates a report based on the analysis results. The report includes details of any inconsistencies or anomalies detected. This report is presented to security personnel via a head-mounted display to assist with on-site investigations.

[0219] Examples of specific cases and prompt statements

[0220] As a concrete example of use, consider a scenario where a security officer is conducting an on-site investigation of a traffic accident. In this case, the officer inputs "legal data regarding traffic accidents" into the system via a head-mounted display and is instructed to generate "questions regarding procedures at the time of the accident."

[0221] Examples of prompt messages include the following:

[0222] "Please obtain legal data regarding traffic accidents and generate questions about procedures to be followed when an accident occurs."

[0223] In summary, the system of the present invention enables rapid and accurate on-site interrogations and supports security personnel through the efficient use of legal and case law data.

[0224] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0225] Step 1:

[0226] The server collects data related to laws and precedents from online resources.

[0227] Input: URLs of government legal databases, court case databases, etc.

[0228] Data processing: Web scraping is performed to extract necessary text information from HTML.

[0229] Output: Text data of collected laws and precedents.

[0230] Specific operation: The server uses a web scraping tool to retrieve HTML content from a specified URL and extracts text information using a library such as BeautifulSoup.

[0231] Step 2:

[0232] The server analyzes the acquired data using natural language processing technology and stores it in a database.

[0233] Input: Text data of collected laws and precedents.

[0234] Data processing: Extract important topics and case precedents through text mining and topic modeling.

[0235] Output: Analyzed data and its topic information.

[0236] Specific operation: The server uses natural language processing libraries such as spaCy to parse text data and stores the data in a database such as SQLite.

[0237] Step 3:

[0238] The user logs into the system using a terminal and enters the information of the person being investigated.

[0239] Input: Basic information of the person being questioned (name, age, summary of the case).

[0240] Data processing: Perform user authentication and format checks on input information.

[0241] Output: Information about the person being interrogated is stored in the database.

[0242] Specific operation: The terminal sends the user's login information to the server, the server performs authentication, receives the information of the person being investigated, and stores it in the database.

[0243] Step 4:

[0244] The server retrieves relevant information from the database and generates a set of questions.

[0245] Input: Information on the person being interrogated and collected legal and case law data.

[0246] Data processing: Use a generative AI model to generate relevant questions.

[0247] Output: The generated set of questions.

[0248] Specific operation: The server retrieves relevant information from the database based on the query and uses a generative AI model (e.g., GPT-3) to generate the query.

[0249] Step 5:

[0250] Security personnel use a head-mounted display to show questions, and their answers are recorded using speech recognition technology.

[0251] Input: The generated set of questions.

[0252] Data processing: Speech recognition technology is used to convert the responses recorded via voice input into text.

[0253] Output: Text data of the recorded responses.

[0254] Specific operation: The head-mounted display presents a set of questions to the security officer and records the officer's voice input in real time, converting it to text.

[0255] Step 6:

[0256] The server receives the recorded response data and compares it with legal and case law data to detect inconsistencies and inconsistencies.

[0257] Input: Text data of the recorded response.

[0258] Data processing: Natural language processing techniques are used to compare and analyze response data with legal and case law data.

[0259] Output: Detection results for inconsistencies and unnatural points.

[0260] Specific operation: The server receives text data, performs matching and analysis using natural language processing techniques, and extracts the detection results.

[0261] Step 7:

[0262] The server generates a report based on the analysis results and presents it to the security implementer.

[0263] Input: Results of detecting inconsistencies and unnatural points.

[0264] Data processing: Based on the detection results, a report is generated, and the data is formatted for visual display.

[0265] Output: Final report.

[0266] Specific operation: The server generates a detailed report based on the detection results and presents it to the security officer via a head-mounted display.

[0267] The above outlines the specific processing flow of the system program that implements the application example.

[0268] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0269] The system according to this invention automatically acquires and analyzes data related to laws and precedents, and uses it to aid in interrogations. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the accuracy and effectiveness of interrogations are improved. The system consists of a server, terminals, and users, and its detailed operation is described below.

[0270] Knowledge acquisition and analysis

[0271] Data collection

[0272] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It is also possible to collect data from publicly available websites using web scraping techniques.

[0273] Data Analysis

[0274] The data collected by the server is analyzed using natural language processing techniques. Specifically, techniques such as text mining, sentiment analysis, and topic modeling are applied. The analyzed data is classified into categories according to its intended use.

[0275] Storage in database

[0276] The server stores the analyzed data in a database. Legal information is stored in the "Legal Matters" category, and information on case precedents is stored in the "Case Precedents" category. An index is created to streamline searching and query processing.

[0277] Preparation for interrogation

[0278] User login

[0279] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server performs authentication and grants permission to log in.

[0280] Input of information of the person under investigation

[0281] The user inputs the basic information of the person under investigation into the terminal. This information includes the name, age, address, summary of the case, etc. of the person under investigation. The terminal sends this information to the server.

[0282] Generation of question set

[0283] The server retrieves relevant information from the database and generates a question set. For example, in the case of a traffic accident, questions based on traffic laws and precedents are generated. The generated question set is sent to the terminal.

[0284] Use of emotion engine

[0285] Acquisition of emotion data

[0286] The terminal acquires the user's emotion data during the investigation. There are methods such as using voice analysis to identify emotions from the user's speech, or using the terminal's camera to analyze facial expressions and recognize emotions.

[0287] Analysis of emotion data

[0288] The server analyzes the emotion data obtained using the emotion engine. In this analysis, indicators for judging the user's stress level and the possibility of lying are generated.

[0289] Application of emotion data

[0290] The server adjusts the progress of the investigation based on the analyzed emotion data. For example, if the user is feeling high stress, the difficulty level of the questions is lowered or a break is proposed. The analysis result is presented to the user through the terminal.

[0291] Conduct of the investigation [[ID=四十八]]

[0292] [[ID=四十九]] [[ID=五十]]Start of the investigation[[ID=五十一]]

[0293] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in either audio or text format.

[0294] Sending and analyzing response data

[0295] The terminal sends the recorded response data to the server. The server analyzes the response data using natural language processing technology and compares it with legal and case law data to detect inconsistencies and inconsistencies.

[0296] Report generation

[0297] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on sentiment data. The generated report is sent to the terminal and presented to the user.

[0298] Examples

[0299] Example 1: Investigation of a car accident

[0300] The server collects and analyzes case data related to traffic laws and automobile accidents, and stores it in a database. The user logs into the system using a terminal and enters information about Mr. Tanaka, the person being investigated. Based on a set of questions generated by the server, the user asks Mr. Tanaka questions. Mr. Tanaka's answers are recorded, and the terminal sends the data to the server. The server generates a report, including any inconsistencies, based on the analysis results, and presents it to the user via the terminal.

[0301] Example 2: Use of an emotion engine

[0302] The user uses the emotion engine during the interrogation. The device analyzes Mr. Tanaka's facial expressions in real time and sends emotion data to the server. The server analyzes the emotion data and indicates that Mr. Tanaka is likely experiencing stress. The server adjusts the set of questions, and the device presents them to the user.

[0303] This system can significantly improve the accuracy and efficiency of investigations by combining legal and case data with sentiment analysis.

[0304] The following describes the processing flow.

[0305] Step 1:

[0306] Data collection

[0307] The server accesses online resources and collects data related to laws and cases. Specifically, it makes an API connection to the government's legal database and the court's case database to obtain the necessary information. It is also possible to collect data from publicly available websites using web scraping technology.

[0308] Step 2:

[0309] Data analysis

[0310] The server analyzes the collected data using natural language processing technology. Techniques such as text mining, sentiment analysis, and entity extraction are applied to understand the content of the data and classify it into categories according to its use.

[0311] Step 3:

[0312] Storage in the database

[0313] The server stores the analyzed data in the database. Information related to laws is stored in the regulation category, information related to cases is stored in the case category, and indexes for efficient search and query processing are also created.

[0314] Step 4:

[0315] User login

[0316] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server authenticates the user and grants permission to log in.

[0317] Step 5:

[0318] Entering information about the person being interrogated

[0319] The user uses a terminal to enter basic information about the person being interrogated. This information includes the person's name, age, address, and a summary of the case. The terminal then sends this information to the server.

[0320] Step 6:

[0321] Question set generation

[0322] The server retrieves information related to the person being interrogated from the database and generates a set of questions. For example, in the case of a traffic accident, specific questions based on traffic laws and relevant precedents are generated. The generated set of questions is then sent to the terminal.

[0323] Step 7:

[0324] Acquisition of emotional data

[0325] The device acquires emotional data from the user and the person being interrogated during questioning. Methods include identifying emotions from speech using voice analysis, and recognizing emotions by analyzing facial expressions using the device's camera.

[0326] Step 8:

[0327] Analysis of emotional data

[0328] The server analyzes emotional data acquired using an emotion engine. It generates indicators to determine the stress level of users and those being interrogated, as well as the likelihood of them lying.

[0329] Step 9:

[0330] Start of interrogation

[0331] The user uses a device to ask the person being interrogated the generated questions. The person being interrogated's responses are recorded on the device in either audio or text format.

[0332] Step 10:

[0333] Sending response data

[0334] The device sends the recorded response data to the server. The data sent includes the interviewee's voice data, text data, or both.

[0335] Step 11:

[0336] Analysis and matching of response data

[0337] The server analyzes the received response data using natural language processing technology. The analyzed response data is then compared with legal and case law data stored in a database to detect inconsistencies and inconsistencies.

[0338] Step 12:

[0339] Report generation

[0340] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The generated report is sent to the terminal and presented to the user.

[0341] Step 13:

[0342] Confirmation of results and further questioning

[0343] The user reviews the report using a terminal. If necessary, the user creates additional questions and conducts further investigations based on these questions. The additional questions are sent to the server for further analysis.

[0344] Through these specific processing steps, the accuracy and efficiency of interrogations can be significantly improved.

[0345] (Example 2)

[0346] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0347] While conventional interrogation systems have a certain degree of accuracy in generating questions based on laws and precedents and analyzing responses, a challenge lies in their lack of adjustments that take into account the emotions of the person being interrogated. Furthermore, there is a need for methods to achieve more accurate and fair interrogations by acquiring and analyzing the emotional data of the person being interrogated.

[0348] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0349] In this invention, the server includes means for acquiring data related to laws and precedents from online resources; means for analyzing the acquired data using natural language processing technology and storing it in a database; means for acquiring relevant information from the database and generating a set of questions based on information about the person being interrogated entered by the user; means for collecting emotional data of the person being interrogated using a terminal and analyzing it using emotion analysis technology; means for adjusting the progress of the interrogation in real time based on the analyzed emotional data; means for receiving the interrogation response data, analyzing it using natural language processing technology, and detecting inconsistencies and unnatural points by comparing it with legal and precedent data; and means for generating a report based on the analysis results and presenting it to the user. This makes it possible to achieve highly accurate and fair interrogations while taking into account the emotional data of the person being interrogated.

[0350] "Online resources" refer to information sources such as websites, databases, and APIs that are accessible via the internet.

[0351] A "database" is a system for storing structured data and for efficiently searching and managing it.

[0352] "Natural language processing technology" refers to technologies for understanding, analyzing, and generating human language using computers, and includes text mining, sentiment analysis, and topic modeling.

[0353] A "question set" is a collection of multiple questions asked for a specific purpose or target.

[0354] "Emotional analysis technology" is a technology that analyzes a person's emotional state from data such as audio and video, and is used to assess stress levels and truthfulness.

[0355] A "person under investigation" is a person who is being investigated in relation to laws or precedents.

[0356] A "terminal" refers to a device, such as a computer or smartphone, that a user uses to access a system.

[0357] "Response data" refers to data on the content of the answers given by the person being interrogated to the questions.

[0358] A "report" is a document that summarizes the results of an investigation, and includes analysis results, inconsistencies, and inconsistencies.

[0359] "Real-time" refers to processing occurring simultaneously with the event or with an extremely short delay.

[0360] The system according to this invention automatically acquires and analyzes data related to laws and precedents, and uses it to aid in interrogations. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the accuracy and effectiveness of interrogations are improved. This system consists of a server, terminals, and users.

[0361] Knowledge acquisition and analysis

[0362] Data collection

[0363] The server accesses online resources and collects data related to laws and precedents. Specifically, it obtains legal data using APIs provided by government agencies and collects precedent data from court websites using web scraping techniques. For example, legal data can be obtained using the "e-Stat" API, and precedent data can be collected using the "BeautifulSoup" library.

[0364] Data Analysis

[0365] The server analyzes the collected data using natural language processing techniques. Python's "NLTK" and "SpaCy" are used for the analysis. Specific processing includes text mining, sentiment analysis, and topic modeling. This extracts important parts of the collected text data and classifies them into categories according to their intended use.

[0366] Storage in database

[0367] The analyzed data is stored in a database by the server. For example, a MySQL database is used to categorize and store legal information in the "Legal Regulations" category and case law information in the "Case Law" category. Appropriate indexes are set up to streamline searching and querying.

[0368] Preparation for interrogation

[0369] User login

[0370] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server via HTTPS. The server verifies the user information against the database and performs authentication.

[0371] Entering information about the person being interrogated

[0372] The user enters the basic information of the person being interrogated into the terminal. This information includes name, age, address, and a summary of the case. This information is sent to the server in JSON format.

[0373] Question set generation

[0374] The server retrieves relevant information from the database and generates a set of questions using the Python "GPT-3" API. For example, in the case of a traffic accident, it forms questions based on the category "traffic laws" and sends the constructed questions to the terminal.

[0375] Using an Emotion Engine

[0376] Acquisition of emotional data

[0377] The device acquires user emotion data during questioning. It uses the "Google Cloud Speech-to-Text" API for voice analysis and "OpenCV" for facial expression analysis. The acquired emotion data is sent to the server in real time.

[0378] Analysis of emotional data

[0379] The server analyzes the emotional data it receives using "TENSORFLOW®". An algorithm is executed that evaluates the stress level and truthfulness of the interrogated person based on changes in voice tone and facial expressions.

[0380] Applications of emotional data

[0381] The server adjusts the progress of the interrogation based on the analysis results. If a high stress level is detected, the server may lower the difficulty of the question set it generates or send a message to the terminal suggesting a break.

[0382] Conducting an interrogation

[0383] Start of interrogation

[0384] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded using a microphone, and the device sends that data to a server.

[0385] Sending and analyzing response data

[0386] The terminal sends the recorded audio data to the server. The server uses "NLTK" or "SpaCy" to convert the audio data into text and compares it with legal and case law data to analyze for inconsistencies and inconsistencies.

[0387] Report generation

[0388] The server generates a report based on the final analysis results. The report includes a summary of all the interviewee's responses, inconsistencies, and a stress assessment based on sentiment analysis. The generated report is sent from the server to the terminal in PDF format and presented to the user.

[0389] Examples

[0390] Example 1: Investigation of a car accident

[0391] The server collects and analyzes case data related to traffic laws and automobile accidents, and stores it in a database. The user logs into the system using a terminal and enters information about the person being investigated. Based on a set of questions generated by the server, the user asks questions to the person being investigated. The person's answers are recorded, and the terminal sends the data to the server. The server generates a report, including any inconsistencies, based on the analysis results, and presents it to the user via the terminal.

[0392] Example 2: Use of an emotion engine

[0393] The user uses the emotion engine during the interrogation. The device analyzes the subject's facial expressions in real time and sends emotion data to the server. The server analyzes the emotion data and indicates that the subject is likely experiencing stress. The server adjusts the set of questions, which the device then presents to the user.

[0394] Examples of prompts for generative AI models

[0395] "Based on precedents related to traffic accidents, please automatically generate a set of questions for those being interrogated. During this process, use an emotion analysis engine to assess whether the subject is experiencing stress, and adjust the difficulty of the questions based on the results."

[0396] The above describes a specific implementation of the system program.

[0397] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0398] Step 1: Data Collection

[0399] The server accesses online resources and collects data related to laws and precedents. The server obtains legal data using APIs provided by government agencies and collects precedent data from court websites using web scraping techniques. Specifically, it obtains legal data in JSON format via the "e-Stat" API and uses the "BeautifulSoup" library to analyze precedent data from HTML pages and extract necessary information. Input is legal analysis information from online resources, and output is the acquired legal data and precedent data.

[0400] Step 2: Data Analysis

[0401] The server analyzes the collected data using natural language processing techniques. This analysis utilizes Python's "NLTK" and "SpaCy." The server processes the collected data through steps such as word tokenization, text mining, sentiment analysis, and topic modeling. The input is the collected raw data, and the output consists of analyzed key information and its results.

[0402] Step 3: Storing in the database

[0403] The server stores the parsed data in a database. Specifically, it uses a MySQL database, storing legal information in the "Legal Regulations" table and case law information in the "Case Law" table. It also creates indexes to improve search efficiency. The input is the parsed data, and the output is the data stored in the database and the indexes.

[0404] Step 4: User Login

[0405] A user logs into the system using a terminal. The user enters their ID and password, and this information is sent from the terminal to the server via HTTPS. The server compares this information with the user's information in the database and performs authentication. The input is the user's ID and password, and the output is the authentication result (success or failure) response.

[0406] Step 5: Enter information about the person being investigated.

[0407] The user enters basic information about the person being interrogated (name, age, address, case summary, etc.) into a terminal, and this information is sent from the terminal to the server in JSON format. The server stores the received information in a database. The input is the basic information of the person being interrogated, and the output is the information stored in the database.

[0408] Step 6: Generate the question set

[0409] The server retrieves relevant information from the database and generates a set of questions using the Python "GPT-3" API. Specifically, in the case of a traffic accident, it forms questions based on the category "traffic laws" and sends the constructed questions to the terminal. The input is relevant information from the database, and the output is the generated set of questions.

[0410] Step 7: Acquiring emotional data

[0411] The device acquires user emotion data during questioning. The "Google Cloud Speech-to-Text" API is used for speech analysis, and "OpenCV" is used for facial expression analysis. The emotion data acquired by the device is sent to the server in real time. The input is the user's emotion data, and the output is the analyzed data sent to the server.

[0412] Step 8: Analyzing emotional data

[0413] The server analyzes the emotional data it receives using TensorFlow. It then runs an algorithm that evaluates the stress level and truthfulness of the interrogated person based on changes in voice tone and facial expressions. The input is emotional data, and the output is the evaluation result of the stress level and truthfulness.

[0414] Step 9: Applying emotional data

[0415] The server adjusts the progress of the interrogation based on the analysis results. If a high stress level is detected, the server lowers the difficulty of the generated question set or generates a message suggesting a break and sends it to the terminal. The input is the evaluation result, and the output is the adjusted question set or the break suggestion message.

[0416] Step 10: Start of Interrogation

[0417] The user uses a device to generate questions and ask them to the person being interrogated. The person being interrogated's responses are recorded using a microphone, and the device sends this data to a server. The input is the person being interrogated's responses, and the output is the audio data sent to the server.

[0418] Step 11: Submitting and analyzing response data

[0419] The terminal sends the recorded audio data to the server. The server uses "NLTK" or "SpaCy" to convert the audio data into text and analyzes inconsistencies and inconsistencies by comparing it with legal and case law data. The input is the audio data, and the output is the converted text data and its analysis results.

[0420] Step 12: Generate the report

[0421] The server generates a report based on the final analysis results. The report includes a summary of all the interviewee's responses, inconsistencies, and a stress assessment based on sentiment analysis. The generated report is sent from the server to the terminal in PDF format and presented to the user. The input is the analysis results, and the output is a report in PDF format.

[0422] (Application Example 2)

[0423] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0424] In modern society, fraud and illegal activities are spreading with the increase in online transactions and communications. Traditional interrogation methods require manual collection and analysis of information related to laws and precedents, which is time-consuming, costly, and inconsistent in accuracy. Furthermore, it is difficult to properly assess the emotions and stress levels of those being interrogated, which can affect the progress of the interrogation. There is a need to solve these problems.

[0425] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for acquiring data related to laws and precedents from online resources, means for analyzing the acquired data using natural language processing technology and storing it in a database, means for acquiring relevant information from the database and generating a set of questions based on the information of the person being interrogated entered by the user, means for receiving the suspect's response data and detecting inconsistencies and unnatural points by comparing it with the data on laws and precedents, means for analyzing the user's emotional data using an emotion engine and adjusting the progress of the interrogation, means for generating a report based on the analysis results and presenting it to the user, means for analyzing the user's text and image data and detecting fraudulent activity, and means for evaluating and warning about the risk of fraud based on the analysis results. This enables the rapid detection and prevention of fraudulent activity and fraud in online transactions and communications, as well as improved accuracy and efficiency of interrogations.

[0426] "Data related to laws and precedents" refers to all information provided by the government and courts regarding laws and past court precedents.

[0427] "Online resources" refer to digital information sources such as websites, databases, and APIs that can be accessed via the internet.

[0428] "Means of acquisition" refers to the technologies and processes that a server uses to collect data from online resources.

[0429] "Natural language processing technology" refers to algorithms and models used to analyze text data and extract meaningful information.

[0430] "Means of storing data in a database" refers to systems and methods that efficiently store acquired and analyzed data, and enable searching and retrieval as needed.

[0431] "Information about the person being interrogated" refers to basic data about the person being interrogated, such as name, age, address, and a summary of the case.

[0432] "Methods for generating question sets" refers to technologies that automatically create a list of questions to conduct an interrogation based on data collected in advance.

[0433] A "suspect" refers to an individual who is under investigation and is suspected of having committed a specific act.

[0434] "Inconsistencies or unnatural points" refer to parts of the response data that do not match or are inconsistent with legal regulations or case law data.

[0435] An "emotion engine" refers to technology that analyzes a user's emotional state and takes appropriate measures based on that analysis.

[0436] "Means of adjusting the difficulty of questions" refers to systems or methods that dynamically change the content and complexity of questions during an interrogation, taking into account the user's emotional state.

[0437] "Methods for generating reports" refers to technology that automatically creates reports summarizing the results of interrogations and key points based on analysis results.

[0438] "Means of detecting fraudulent activity" refers to algorithms and technologies that analyze users' text and image data to identify fraud and other fraudulent activities at an early stage.

[0439] "Means of assessing and warning about the risk of fraud" refers to a system that, based on analysis results, notifies or warns users if it is determined that there is a high probability of fraud.

[0440] The system according to the present invention aims to rapidly detect and prevent fraud and misconduct in online transactions and communications. This system consists of multiple means for acquiring data related to laws and precedents, sentiment analysis using a sentiment engine, detection of misconduct, and risk assessment based on the results.

[0441] Knowledge acquisition and analysis

[0442] Data collection

[0443] The server accesses online resources via the internet to obtain data related to laws and precedents. Specifically, this involves collecting digital data provided by governments and courts via APIs, and collecting data from publicly available websites using web scraping techniques. This ensures that the latest legal and precedent data is always up-to-date.

[0444] Data Analysis

[0445] The server analyzes the acquired data using natural language processing techniques. Text mining, sentiment analysis, and topic modeling are applied, and the analyzed data is classified into categories. This allows for efficient searching of relevant information.

[0446] Storage in database

[0447] The analyzed data is stored in a database. Information related to laws is stored in the legal regulations category, and information related to case law is stored in the case law category, and indexes are created. This allows for quick searching and query processing.

[0448] Questionnaire generation and investigation

[0449] Entering information about the person being interrogated

[0450] The user uses a terminal to input basic information about the person being interrogated into the system. This information includes the person's name, age, address, and a summary of the case. The terminal sends this information to the server, which retrieves the relevant information from the database.

[0451] Question set generation

[0452] The server retrieves relevant information from the database and generates a set of questions. Based on the retrieved information, appropriate questions are automatically created. For example, if there is suspicion of fraud related to online transactions, questions based on applicable laws and precedents will be generated.

[0453] Using emotion engines for fraud detection

[0454] Acquisition and analysis of emotional data

[0455] The terminal acquires emotional data from the user and the subject during questioning. Using voice analysis and facial recognition technology, it identifies emotions from speech and analyzes facial expressions using a camera. The server analyzes this data using an emotion engine and evaluates the emotional state.

[0456] Applications of emotional data

[0457] The server adjusts the interrogation process based on the analyzed emotional data. For example, if the person is experiencing high stress levels, it may lower the difficulty of the questions or suggest a break. It also displays a warning to the user if it determines there is a high risk of fraud.

[0458] Report generation and output

[0459] Sending and analyzing answer data

[0460] The user uses a device to ask questions of the person being interrogated and records the answer data. This data is recorded on the device in text or audio format and sent to a server. The server uses natural language processing technology to analyze the answer data and compare it with legal and case law data to detect inconsistencies and inconsistencies.

[0461] Report generation and presentation

[0462] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The generated report is presented to the user via the terminal.

[0463] Specific example

[0464] For example, if a suspicious transaction occurs during an online transaction, this system analyzes the chat content in real time and uses an emotion engine to detect the risk of fraud. If it is determined to be "high risk," it displays a warning to the user and generates a detailed report.

[0465] Example prompts for using generative AI models

[0466] "Analyze the transaction details and assess the risk of fraud. Transaction details: [Specific transaction details] User sentiment: [Details of sentiment state] Present the analysis results in a list format."

[0467] This system will be a powerful tool for early detection and prevention of fraudulent activity in online transactions and communications.

[0468] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0469] Step 1:

[0470] The server retrieves data related to laws and precedents from online resources. Specifically, it collects digital data provided by governments and courts via API connections. It may also collect data from publicly available websites using web scraping techniques. The server retrieves this data and stores it in a database. The input is data from online resources, and the output is analytical data stored in the database.

[0471] Step 2:

[0472] The server analyzes the acquired data using natural language processing techniques. Specifically, it applies text mining, sentiment analysis, and topic modeling. As a result of the analysis, the data is classified into categories and stored again in the database. The input is raw data acquired from online resources, and the output is analyzed data classified into categories.

[0473] Step 3:

[0474] The user uses a terminal to input basic information about the person being interrogated (name, age, address, case summary, etc.). The entered information is sent from the terminal to the server. The input is the basic information of the person being interrogated, and the output is the user's input information sent to the server.

[0475] Step 4:

[0476] The server retrieves relevant information from the database based on the information of the person being interrogated and generates a set of questions. The set of questions is automatically generated to include appropriate questions for the person being interrogated. The input is the basic information of the person being interrogated, and the output is the generated set of questions.

[0477] Step 5:

[0478] The device acquires emotional data from the user or the person being interrogated. This emotional data is obtained using voice analysis and facial recognition technology. Specifically, voice analysis is performed to identify emotions from the user's statements, and facial expressions are analyzed using a camera. The inputs are audio and camera footage, and the output is the analyzed emotional data.

[0479] Step 6:

[0480] The server analyzes the acquired emotion data. Using an emotion engine, it evaluates the user's emotional state (stress level, likelihood of lying, etc.) and generates indicators. The input is emotion data, and the output is the emotion analysis result.

[0481] Step 7:

[0482] The server adjusts the interview process based on the analyzed sentiment data. For example, if the user is experiencing high levels of stress, it may lower the difficulty of the questions or suggest a break. The input is the sentiment analysis results, and the output is an adjusted set of questions or a break suggestion.

[0483] Step 8:

[0484] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in audio or text format and sent from the device to the server. The input is the person's responses, and the output is the response data sent to the server.

[0485] Step 9:

[0486] The server analyzes the response data using natural language processing techniques and compares it with legal and case law data to detect inconsistencies and inconsistencies. The input is the response data, and the output is the analysis results.

[0487] Step 10:

[0488] The server generates a report based on the analysis results and presents it to the user via the terminal. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The input is the analysis results, and the output is the generated report.

[0489] This allows for a clear explanation of the overall system processing flow and the specific actions performed at each step.

[0490] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0491] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0492] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0493] [Second Embodiment]

[0494] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0495] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0496] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0497] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0498] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0499] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0500] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0501] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0502] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0503] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0504] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0505] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0506] The system according to the present invention is designed to automatically acquire and analyze data related to laws and precedents, and to use this data for investigations. This system consists of a server, terminals, and users.

[0507] Knowledge acquisition and analysis

[0508] 1. Data Collection

[0509] The server accesses online resources and collects data related to laws and precedents. This includes information from government legal databases and court case databases.

[0510] 2. Data Analysis

[0511] The data collected by the server is analyzed using natural language processing techniques. Specifically, techniques such as text mining and topic modeling are applied to understand the content of the data.

[0512] 3. Storing in a database

[0513] The server analyzes the data and stores it in a database. This enables efficient searching and management.

[0514] Preparation for interrogation

[0515] 4. User login

[0516] The user logs into the system using their device. User authentication information is sent from the device to the server.

[0517] 5. Entering information about the person being interrogated.

[0518] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, and a summary of the case.

[0519] 6. Generating the Question Set

[0520] The server retrieves relevant information from the database and generates a set of questions. For example, in the case of an investigation into a traffic accident, questions based on traffic laws and relevant precedents will be generated.

[0521] Conducting an interrogation

[0522] 7. Commencement of Interrogation

[0523] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in either audio or text format.

[0524] 8. Data transmission and analysis

[0525] The terminal sends the recorded response data to the server. After receiving the response data, the server analyzes it by comparing it with legal and case law data that has been collected in advance.

[0526] 9. Detection of inconsistencies

[0527] The server uses natural language processing technology to analyze the content of the response and detect inconsistencies and unnatural points. For example, it might detect if a statement like "The car in front of me suddenly stopped at the intersection" contradicts the circumstantial evidence at the scene.

[0528] 10. Report generation

[0529] The server generates a report based on the analysis results. The report includes details about any inconsistencies or inconsistencies detected.

[0530] Example: Investigation of a car accident

[0531] As a concrete example, let's explain the investigation in a car accident.

[0532] 1. Data Collection

[0533] The server collects and analyzes case law data related to traffic laws and automobile accidents, and stores it in a database.

[0534] 2. User login

[0535] The user logs into the system using a terminal and enters the information of Mr. Tanaka, the person being investigated.

[0536] 3. Generating the Question Set

[0537] The server retrieves information about traffic accidents related to Mr. Tanaka and generates an appropriate set of questions.

[0538] 4. Conducting the interrogation

[0539] The user asks Mr. Tanaka, "Could you tell me the details of what happened in the accident?" Mr. Tanaka replies, "The car in front of me stopped suddenly at the intersection, so I also slammed on the brakes."

[0540] 5. Data transmission and analysis

[0541] The terminal records Mr. Tanaka's responses in audio or text format and sends them to the server. The server analyzes the data and detects any inconsistencies.

[0542] 6. Report generation

[0543] Based on the analysis results, the server generates a report pointing out inconsistencies between Mr. Tanaka's statements and the circumstantial evidence at the scene, and presents it to the user via the terminal.

[0544] As described above, the system according to the present invention improves the efficiency of police officers' work and enhances the accuracy of interrogations by acquiring, analyzing, and supporting legal and case law data.

[0545] The following describes the processing flow.

[0546] Step 1:

[0547] Data collection

[0548] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It can also collect data from publicly available websites using web scraping techniques.

[0549] Step 2:

[0550] Data Analysis

[0551] The server analyzes the collected data using natural language processing techniques. These include techniques such as text mining, sentiment analysis, and entity extraction. The analyzed data is categorized and organized by content.

[0552] Step 3:

[0553] Storage in database

[0554] The server stores the analyzed data in a database. Specifically, information related to laws is stored in the "Legal Matters" category, and information related to case law is stored in the "Case Law" category. Furthermore, indexes are created to streamline searching and query processing.

[0555] Step 4:

[0556] User login

[0557] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server performs authentication and grants permission to log in.

[0558] Step 5:

[0559] Entering information about the person being interrogated

[0560] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, address, and a summary of the case. The terminal then transmits this information to the server.

[0561] Step 6:

[0562] Question set generation

[0563] The server retrieves information related to the interrogation from the database and generates a set of questions. For example, if the person being interrogated is involved in a traffic accident, questions will be generated based on traffic laws and relevant past court precedents. The generated set of questions is then sent to the terminal.

[0564] Step 7:

[0565] Start of interrogation

[0566] The user uses a device to ask the person being interrogated system-generated questions. The person being interrogated's responses are recorded by the device in either audio or text format.

[0567] Step 8:

[0568] Sending response data

[0569] The device sends the recorded response data to the server. The data sent includes the interviewee's voice data, text data, or both.

[0570] Step 9:

[0571] Analysis and matching of response data

[0572] The server analyzes the received response data using natural language processing technology. The analyzed response data is then compared with legal and case law data stored in a database beforehand to detect inconsistencies and unnatural points.

[0573] Step 10:

[0574] Report generation

[0575] The server generates a report based on the analysis results. The report includes details about inconsistencies and unnatural points in the interrogated person's statements. The generated report is sent to the terminal and presented to the user.

[0576] Step 11:

[0577] Confirmation of results and further questioning

[0578] The user reviews the report using a terminal. If necessary, the user creates additional questions and conducts further investigations based on these questions. The additional questions are sent to the server for further analysis.

[0579] By repeating these steps, the accuracy and efficiency of interrogations can be improved.

[0580] (Example 1)

[0581] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0582] The current interrogation system requires a tremendous amount of time and effort because data collection and analysis based on laws and precedents are done manually. Furthermore, the quality and accuracy of the question sets generated during interrogations are low, making it difficult to efficiently detect inconsistencies or unnatural points in the interviewee's responses. Therefore, there is a need to improve the accuracy and efficiency of interrogations.

[0583] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0584] In this invention, the server includes means for acquiring data related to laws and precedents from online resources, means for analyzing the acquired data using natural language processing technology and storing it in a database, means for acquiring relevant information from the database based on information about the subject of investigation entered by the user and generating a set of questions using a generative AI model, means for receiving the subject's response data and detecting inconsistencies and unnatural points by comparing it with the data on laws and precedents, and means for generating a report based on the analysis results and presenting it to the user. This not only improves the quality of interrogations but also enables highly accurate and rapid interrogations.

[0585] "Online resources" refer to websites and databases that exist on the internet and contain data on laws and precedents provided by public institutions, companies, and other organizations.

[0586] "Legal and case law data" refers to digital data provided by government agencies and courts in various countries, including legal texts and the content of past court precedents.

[0587] "Natural language processing technology" refers to techniques that enable computers to understand and analyze human language, and includes methods such as text mining and topic modeling.

[0588] A "database" is a system for efficiently storing, managing, and retrieving large amounts of data, and includes relational databases and NoSQL databases.

[0589] A "subject of investigation" is a person who is the subject of questioning or investigation, and their basic information and information related to the case are entered into the system.

[0590] A "generative AI model" is a model that uses AI technology to automatically generate question sets and reports based on given input data. Examples include GPT-3 and similar models.

[0591] A "question set" refers to a series of questions asked of the person being investigated, and is automatically generated by an AI model according to the purpose of the interrogation or investigation.

[0592] "Response data" refers to the data of responses provided by survey participants to questions, and is recorded in audio or text format.

[0593] "Inconsistencies or unnatural points" refer to parts of the survey respondents' responses that do not match or are logically inconsistent with the legal and case law data collected and analyzed in advance.

[0594] A "report" is a document that summarizes the analysis results and includes details about any inconsistencies or inconsistencies that were detected.

[0595] This invention relates to a system for automatically acquiring and analyzing data related to laws and precedents, and for efficiently conducting interrogations. This system consists of a server, terminals, and users.

[0596] Knowledge acquisition and analysis

[0597] 1. Data Collection

[0598] The server collects data related to laws and precedents from online resources on the internet. For example, it accesses government-provided legal databases and court case databases. Specifically, it uses Python crawling tools (e.g., Scrapy) to periodically crawl these databases and retrieve the necessary information.

[0599] 2. Data Analysis

[0600] The server analyzes the collected data using natural language processing (NLP) techniques. It applies text mining libraries (e.g., NLTK, SpaCy) and topic modeling techniques (e.g., LDA) to understand the data's content and extract relevant information.

[0601] 3. Storing in a database

[0602] The server stores the analyzed data in a database. Because fast and efficient data management is required here, relational databases (e.g., MySQL, PostgreSQL) or NoSQL databases (e.g., MongoDB) are used.

[0603] Preparation for interrogation

[0604] 4. User login

[0605] Users log in to the system using their device. User authentication is performed by sending data from the device to the server, where security measures, including two-factor authentication, are implemented and the authentication process is completed.

[0606] 5. Entering information about the person being interrogated.

[0607] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, and a summary of the case. The entered data is stored in a database on the server.

[0608] 6. Generating the Question Set

[0609] The server retrieves relevant information from the database and generates a set of questions using a generative AI model (e.g., GPT-3). The generated set of questions includes the most appropriate questions for the research subject, based on the legal case.

[0610] Conducting an interrogation

[0611] 7. Commencement of Interrogation

[0612] The user asks the interviewee questions generated using the device. The interviewee's responses are recorded on the device in either audio or text format, and if audio data is used, it is converted to text using speech recognition technology (e.g., Google Cloud Speech-to-Text).

[0613] 8. Data transmission and analysis

[0614] The terminal sends the recorded response data to the server. The server compares the received data with legal and case law data that has been collected and analyzed in advance, and performs analysis using NLP technology (e.g., BERT).

[0615] 9. Detection of inconsistencies

[0616] The server evaluates the responses based on the analysis results and detects inconsistencies and unnatural points. It also checks whether the statements are consistent with the facts by cross-referencing them with legal databases.

[0617] 10. Report generation

[0618] The server generates a report based on the analysis results and presents it to the user. The report includes details about any inconsistencies or unnatural points detected.

[0619] Examples of specific cases and prompt statements

[0620] As a concrete example, let's consider the investigation of a traffic accident.

[0621] 1. The server collects and analyzes case law data related to traffic laws and automobile accidents, and stores it in a database.

[0622] 2. The user logs into the system using a terminal and enters the information of Mr. Tanaka, the person being investigated.

[0623] 3. The server retrieves information about traffic accidents related to Mr. Tanaka and generates an appropriate set of questions.

[0624] Example of a prompt:

[0625] "Please generate a set of questions for an investigation into a traffic accident. The subject of the investigation is Mr. Tanaka, and we would like to ask about the detailed circumstances of the accident. Please include relevant traffic laws and precedents in the questions."

[0626] As described above, the system according to the present invention aims to improve the quality and efficiency of interrogations through the automatic acquisition, analysis, and examination of legal and case law data.

[0627] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0628] Step 1: Data Collection

[0629] The server collects data related to laws and precedents from online resources on the internet. Specifically, it uses Python crawling tools (e.g., Scrapy) to crawl government law databases and court precedent databases. The input is a list of URLs to be crawled, and the output is the retrieved HTML code and text data. The retrieved data is temporarily stored.

[0630] Specific operation: The server executes a Scrapy script, accesses the URL of the specified legal database, and collects the latest case law data.

[0631] Step 2: Data Analysis

[0632] The server analyzes the collected data using natural language processing (NLP) techniques. Specifically, it extracts keywords and topics from the collected text data using text mining libraries (e.g., NLTK, SpaCy). The input is the collected text data, and the output is the analyzed keyword and topic information.

[0633] Specific operation: The server uses the NLTK library to analyze legal terms contained in the collected data and extract important keywords and their relevance.

[0634] Step 3: Storing in the database

[0635] The server stores the analyzed data in a database. This can be a relational database (e.g., MySQL) or a NoSQL database (e.g., MongoDB). The input is the analyzed data, and the output is the completed storage of the data in the database.

[0636] Specific operation: The server stores the parsed data into MySQL using an INSERT statement.

[0637] Step 4: User Login

[0638] The user logs into the system using a terminal. The user ID and password are input data, which are sent to the server. The server compares this information with the authentication information in the database and sends the authentication result back to the terminal. The output indicates whether the login was successful or failed.

[0639] Specific operation: The user enters their ID and password on the login screen, and the server verifies this against the database information to perform authentication.

[0640] Step 5: Enter information about the person being investigated.

[0641] The user enters basic information about the person being interrogated into the terminal. This includes the person's name, age, and a summary of the case. The entered data is sent to the server, which stores it in a database. The output includes confirmation that the data has been successfully saved to the database.

[0642] Specific operation: The user enters Mr. Tanaka's information (name, age, details of the incident) into an input form, and the server receives this information and saves it to the database.

[0643] Step 6: Generate the question set

[0644] The server retrieves relevant information from the database and generates a set of questions using a generative AI model (e.g., GPT-3). Inputs include basic information about the person being interrogated and relevant legal and case law data, while output is the generated set of questions.

[0645] Specific operation: The server sends a prompt to the GPT-3 model saying "Generate a set of questions for a traffic accident investigation," retrieves the response, and constructs the question set.

[0646] Step 7: Start of interrogation

[0647] The user uses a device to ask questions generated by the device to the person being interviewed. The interviewee's responses are recorded on the device in either audio or text format. The input consists of the generated set of questions and the interviewee's responses, and the output is the recorded response data.

[0648] Specific operation: The user asks Mr. Tanaka, "Please tell me the details of what happened when the accident occurred," records Mr. Tanaka's voice response on the device, and converts the voice data into text using Google Cloud Speech-to-Text.

[0649] Step 8: Data transmission and analysis

[0650] The terminal sends the recorded response data to the server. The server compares the received data with pre-collected and analyzed legal and case law data and performs analysis using natural language processing techniques (e.g., BERT). The input consists of the response data and related legal and case law data, and the output is the analysis results.

[0651] Specific operation: The terminal sends the text-based response to the server, and the server performs contextual analysis of the response using a BERT model.

[0652] Step 9: Detecting inconsistencies

[0653] The server evaluates the answers based on the analysis results and detects inconsistencies and unnatural points. The input is the analysis results, and the output is the detected inconsistencies and unnatural points.

[0654] Specific operation: The server compares the case data for the relevant case with the response content and detects if the statement "the car in front suddenly stopped at the intersection" contradicts the situation at the accident scene.

[0655] Step 10: Generate the report

[0656] The server generates a report based on the analysis results and presents it to the user. The input consists of the analysis results and detected inconsistencies, while the output is the generated report.

[0657] Specific operation: The server compiles the analysis results, generates a report pointing out inconsistencies, and sends this report to the terminal in PDF format.

[0658] (Application Example 1)

[0659] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0660] In on-site interrogations and investigations, it is difficult for security personnel to ask efficient and accurate questions based on laws and precedents, and to immediately detect inconsistencies or suspicious points in answers. While rapid information acquisition and analysis on-site are required, conventional methods lack real-time capabilities, leading to decreased accuracy and efficiency in investigations. To address this challenge, there is a need for a system that effectively and quickly utilizes legal and precedent data to support on-site interrogations.

[0661] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0662] In this invention, the server includes means for acquiring data related to laws and precedents from online resources; means for analyzing the acquired data using natural language processing technology and storing it in a database; means for acquiring relevant information from the database based on information about the person being interrogated entered by the user and generating a set of questions; means for a security officer to display the questions using a head-mounted display and record the answers using speech recognition technology; means for receiving the recorded answer data and comparing it with the data on laws and precedents to detect inconsistencies and inconsistencies; and means for generating a report based on the analysis results and presenting it to the security officer. This enables rapid and accurate interrogations on-site.

[0663] "Online resources" is a general term for information sources such as databases and websites that exist on the internet.

[0664] A "law" is a set of rules and norms established by a state or local government that have the power to enforce order in society.

[0665] A "precedent" is a record of judgments and their reasons that have been handed down by courts in the past, and it serves as a reference for similar cases that may follow.

[0666] "Natural language processing technology" refers to techniques for understanding, analyzing, and generating human language using computers, and examples include text mining and topic modeling.

[0667] A "database" is a general term for a system that organizes and stores large amounts of data, and allows for efficient searching and updating.

[0668] A "person under investigation" refers to an individual who is investigated or questioned in order to provide information about a specific incident or situation.

[0669] A "question set" is a collection of related questions grouped together to achieve a specific objective.

[0670] A "security implementer" refers to an individual or organization that takes on the role of ensuring safety in a specific location or situation.

[0671] A "head-mounted display" is a type of display device worn on the head that displays information within the field of vision.

[0672] "Speech recognition technology" is a technology that analyzes speech data and converts it into text data.

[0673] A "contradiction" refers to a part of specific data or statements that is not logically consistent.

[0674] An "unnatural point" refers to a part of specific data or statements that is judged to deviate from what is normal or common sense.

[0675] A "report" is a document that summarizes analysis results or research findings, and is a type of report.

[0676] The system of this invention collects data related to laws and precedents from online resources, analyzes it using natural language processing technology, and generates question sets and analyzes answers for on-site interrogations and investigations. This system mainly consists of a server, a head-mounted display, and a user.

[0677] Collection of legal and case law data

[0678] The server automatically retrieves data related to laws and precedents from databases and websites on the internet. For example, government legal databases and court case databases are used.

[0679] Data analysis and database storage

[0680] The acquired data is analyzed on the server using natural language processing techniques. Specifically, techniques such as text mining and topic modeling are applied to extract important topics and case precedents. The analysis results are stored in a database, enabling efficient searching and management.

[0681] Question set generation

[0682] The user logs into the system via a head-mounted display and enters basic information about the person being interrogated. The server retrieves relevant information from the database and automatically generates an appropriate set of questions. This enables rapid and accurate interrogations on-site.

[0683] Speech recognition and analysis of responses

[0684] Security personnel use a head-mounted display to show generated questions and ask them to the person being interrogated. Responses are recorded in real time using speech recognition technology. The recorded data is sent to a server where it is compared with legal and case law data to detect inconsistencies and inconsistencies.

[0685] Report generation and presentation

[0686] The server generates a report based on the analysis results. The report includes details of any inconsistencies or anomalies detected. This report is presented to security personnel via a head-mounted display to assist with on-site investigations.

[0687] Examples of specific cases and prompt statements

[0688] As a concrete example of use, consider a scenario where a security officer is conducting an on-site investigation of a traffic accident. In this case, the officer inputs "legal data regarding traffic accidents" into the system via a head-mounted display and is instructed to generate "questions regarding procedures at the time of the accident."

[0689] Examples of prompt messages include the following:

[0690] "Please obtain legal data regarding traffic accidents and generate questions about procedures to be followed when an accident occurs."

[0691] In summary, the system of the present invention enables rapid and accurate on-site interrogations and supports security personnel through the efficient use of legal and case law data.

[0692] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0693] Step 1:

[0694] The server collects data related to laws and precedents from online resources.

[0695] Input: URLs of government legal databases, court case databases, etc.

[0696] Data processing: Web scraping is performed to extract necessary text information from HTML.

[0697] Output: Text data of collected laws and precedents.

[0698] Specific operation: The server uses a web scraping tool to retrieve HTML content from a specified URL and extracts text information using a library such as BeautifulSoup.

[0699] Step 2:

[0700] The server analyzes the acquired data using natural language processing technology and stores it in a database.

[0701] Input: Text data of collected laws and precedents.

[0702] Data processing: Extract important topics and case precedents through text mining and topic modeling.

[0703] Output: Analyzed data and its topic information.

[0704] Specific operation: The server uses natural language processing libraries such as spaCy to parse text data and stores the data in a database such as SQLite.

[0705] Step 3:

[0706] The user logs into the system using a terminal and enters the information of the person being investigated.

[0707] Input: Basic information of the person being questioned (name, age, summary of the case).

[0708] Data processing: Perform user authentication and format checks on input information.

[0709] Output: Information about the person being interrogated is stored in the database.

[0710] Specific operation: The terminal sends the user's login information to the server, the server performs authentication, receives the information of the person being investigated, and stores it in the database.

[0711] Step 4:

[0712] The server retrieves relevant information from the database and generates a set of questions.

[0713] Input: Information on the person being interrogated and collected legal and case law data.

[0714] Data processing: Use a generative AI model to generate relevant questions.

[0715] Output: The generated set of questions.

[0716] Specific operation: The server retrieves relevant information from the database based on the query and uses a generative AI model (e.g., GPT-3) to generate the query.

[0717] Step 5:

[0718] Security personnel use a head-mounted display to show questions, and their answers are recorded using speech recognition technology.

[0719] Input: The generated set of questions.

[0720] Data processing: Speech recognition technology is used to convert the responses recorded via voice input into text.

[0721] Output: Text data of the recorded responses.

[0722] Specific operation: The head-mounted display presents a set of questions to the security officer and records the officer's voice input in real time, converting it to text.

[0723] Step 6:

[0724] The server receives the recorded response data and compares it with legal and case law data to detect inconsistencies and inconsistencies.

[0725] Input: Text data of the recorded response.

[0726] Data processing: Natural language processing techniques are used to compare and analyze response data with legal and case law data.

[0727] Output: Detection results for inconsistencies and unnatural points.

[0728] Specific operation: The server receives text data, performs matching and analysis using natural language processing techniques, and extracts the detection results.

[0729] Step 7:

[0730] The server generates a report based on the analysis results and presents it to the security implementer.

[0731] Input: Results of detecting inconsistencies and unnatural points.

[0732] Data processing: Based on the detection results, a report is generated, and the data is formatted for visual display.

[0733] Output: Final report.

[0734] Specific operation: The server generates a detailed report based on the detection results and presents it to the security officer via a head-mounted display.

[0735] The above outlines the specific processing flow of the system program that implements the application example.

[0736] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0737] The system according to this invention automatically acquires and analyzes data related to laws and precedents, and uses it to aid in interrogations. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the accuracy and effectiveness of interrogations are improved. The system consists of a server, terminals, and users, and its detailed operation is described below.

[0738] Knowledge acquisition and analysis

[0739] Data collection

[0740] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It is also possible to collect data from publicly available websites using web scraping techniques.

[0741] Data Analysis

[0742] The data collected by the server is analyzed using natural language processing techniques. Specifically, techniques such as text mining, sentiment analysis, and topic modeling are applied. The analyzed data is classified into categories according to its intended use.

[0743] Storage in database

[0744] The server stores the analyzed data in a database. Legal information is stored in the "Legal Matters" category, and information on case precedents is stored in the "Case Precedents" category. An index is created to streamline searching and query processing.

[0745] Preparation for interrogation

[0746] User login

[0747] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server performs authentication and grants permission to log in.

[0748] Entering information about the person being interrogated

[0749] The user enters the basic information of the person being interrogated into the terminal. This information includes the person's name, age, address, and a summary of the case. The terminal then transmits this information to the server.

[0750] Question set generation

[0751] The server retrieves relevant information from the database and generates a set of questions. For example, in the case of a traffic accident, questions based on traffic laws and precedents are generated. The generated set of questions is then sent to the terminal.

[0752] Using an Emotion Engine

[0753] Acquisition of emotional data

[0754] The device acquires user emotion data during questioning. This can be done by using voice analysis to identify emotions from the user's statements, or by using the device's camera to analyze facial expressions and recognize emotions.

[0755] Analysis of emotional data

[0756] The server analyzes emotional data acquired using an emotion engine. This analysis generates indicators to determine the user's stress level, the likelihood of them lying, and other factors.

[0757] Applications of emotional data

[0758] The server adjusts the interrogation process based on the analyzed emotional data. For example, if the user is experiencing high levels of stress, it may lower the difficulty of the questions or suggest a break. The analysis results are presented to the user via their device.

[0759] Conducting an interrogation

[0760] Start of interrogation

[0761] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in either audio or text format.

[0762] Sending and analyzing response data

[0763] The terminal sends the recorded response data to the server. The server analyzes the response data using natural language processing technology and compares it with legal and case law data to detect inconsistencies and inconsistencies.

[0764] Report generation

[0765] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on sentiment data. The generated report is sent to the terminal and presented to the user.

[0766] Examples

[0767] Example 1: Investigation of a car accident

[0768] The server collects and analyzes case data related to traffic laws and automobile accidents, and stores it in a database. The user logs into the system using a terminal and enters information about Mr. Tanaka, the person being investigated. Based on a set of questions generated by the server, the user asks Mr. Tanaka questions. Mr. Tanaka's answers are recorded, and the terminal sends the data to the server. The server generates a report, including any inconsistencies, based on the analysis results, and presents it to the user via the terminal.

[0769] Example 2: Use of an emotion engine

[0770] The user uses the emotion engine during the interrogation. The device analyzes Mr. Tanaka's facial expressions in real time and sends emotion data to the server. The server analyzes the emotion data and indicates that Mr. Tanaka is likely experiencing stress. The server adjusts the set of questions, and the device presents them to the user.

[0771] This system can significantly improve the accuracy and efficiency of interrogations by combining legal and case law data with sentiment analysis.

[0772] The following describes the processing flow.

[0773] Step 1:

[0774] Data collection

[0775] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It can also collect data from publicly available websites using web scraping techniques.

[0776] Step 2:

[0777] Data Analysis

[0778] The server analyzes the collected data using natural language processing techniques. Techniques such as text mining, sentiment analysis, and entity extraction are applied to understand the data's content and classify it into categories appropriate for its intended use.

[0779] Step 3:

[0780] Storage in database

[0781] The server stores the analyzed data in a database. Legal information is stored in the "Legal Matters" category, and information on case precedents is stored in the "Case Precedents" category. Indexes are also created for efficient searching and query processing.

[0782] Step 4:

[0783] User login

[0784] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server authenticates the user and grants permission to log in.

[0785] Step 5:

[0786] Entering information about the person being interrogated

[0787] The user uses a terminal to enter basic information about the person being interrogated. This information includes the person's name, age, address, and a summary of the case. The terminal then sends this information to the server.

[0788] Step 6:

[0789] Question set generation

[0790] The server retrieves information related to the person being interrogated from the database and generates a set of questions. For example, in the case of a traffic accident, specific questions based on traffic laws and relevant precedents are generated. The generated set of questions is then sent to the terminal.

[0791] Step 7:

[0792] Acquisition of emotional data

[0793] The device acquires emotional data from the user and the person being interrogated during questioning. Methods include identifying emotions from speech using voice analysis, and recognizing emotions by analyzing facial expressions using the device's camera.

[0794] Step 8:

[0795] Analysis of emotional data

[0796] The server analyzes emotional data acquired using an emotion engine. It generates indicators to determine the stress level of users and those being interrogated, as well as the likelihood of them lying.

[0797] Step 9:

[0798] Start of interrogation

[0799] The user uses a device to ask the person being interrogated the generated questions. The person being interrogated's responses are recorded on the device in either audio or text format.

[0800] Step 10:

[0801] Sending response data

[0802] The device sends the recorded response data to the server. The data sent includes the interviewee's voice data, text data, or both.

[0803] Step 11:

[0804] Analysis and matching of response data

[0805] The server analyzes the received response data using natural language processing technology. The analyzed response data is then compared with legal and case law data stored in a database to detect inconsistencies and inconsistencies.

[0806] Step 12:

[0807] Report generation

[0808] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The generated report is sent to the terminal and presented to the user.

[0809] Step 13:

[0810] Confirmation of results and further questioning

[0811] The user reviews the report using a terminal. If necessary, the user creates additional questions and conducts further investigations based on these questions. The additional questions are sent to the server for further analysis.

[0812] Through these specific processing steps, the accuracy and efficiency of interrogations can be significantly improved.

[0813] (Example 2)

[0814] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0815] While conventional interrogation systems have a certain degree of accuracy in generating questions based on laws and precedents and analyzing responses, a challenge lies in their lack of adjustments that take into account the emotions of the person being interrogated. Furthermore, there is a need for methods to achieve more accurate and fair interrogations by acquiring and analyzing the emotional data of the person being interrogated.

[0816] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0817] In this invention, the server includes means for acquiring data related to laws and precedents from online resources; means for analyzing the acquired data using natural language processing technology and storing it in a database; means for acquiring relevant information from the database and generating a set of questions based on information about the person being interrogated entered by the user; means for collecting emotional data of the person being interrogated using a terminal and analyzing it using emotion analysis technology; means for adjusting the progress of the interrogation in real time based on the analyzed emotional data; means for receiving the interrogation response data, analyzing it using natural language processing technology, and detecting inconsistencies and unnatural points by comparing it with legal and precedent data; and means for generating a report based on the analysis results and presenting it to the user. This makes it possible to achieve highly accurate and fair interrogations while taking into account the emotional data of the person being interrogated.

[0818] "Online resources" refer to information sources such as websites, databases, and APIs that are accessible via the internet.

[0819] A "database" is a system for storing structured data and for efficiently searching and managing it.

[0820] "Natural language processing technology" refers to technologies for understanding, analyzing, and generating human language using computers, and includes text mining, sentiment analysis, and topic modeling.

[0821] A "question set" is a collection of multiple questions asked for a specific purpose or target.

[0822] "Emotional analysis technology" is a technology that analyzes a person's emotional state from data such as audio and video, and is used to assess stress levels and truthfulness.

[0823] A "person under investigation" is a person who is being investigated in relation to laws or precedents.

[0824] A "terminal" refers to a device, such as a computer or smartphone, that a user uses to access a system.

[0825] "Response data" refers to data on the content of the answers given by the person being interrogated to the questions.

[0826] A "report" is a document that summarizes the results of an investigation, and includes analysis results, inconsistencies, and inconsistencies.

[0827] "Real-time" refers to processing occurring simultaneously with the event or with an extremely short delay.

[0828] The system according to this invention automatically acquires and analyzes data related to laws and precedents, and uses it to aid in interrogations. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the accuracy and effectiveness of interrogations are improved. This system consists of a server, terminals, and users.

[0829] Knowledge acquisition and analysis

[0830] Data collection

[0831] The server accesses online resources and collects data related to laws and precedents. Specifically, it obtains legal data using APIs provided by government agencies and collects precedent data from court websites using web scraping techniques. For example, legal data can be obtained using the "e-Stat" API, and precedent data can be collected using the "BeautifulSoup" library.

[0832] Data Analysis

[0833] The server analyzes the collected data using natural language processing techniques. Python's "NLTK" and "SpaCy" are used for the analysis. Specific processing includes text mining, sentiment analysis, and topic modeling. This extracts important parts of the collected text data and classifies them into categories according to their intended use.

[0834] Storage in database

[0835] The analyzed data is stored in a database by the server. For example, a MySQL database is used to categorize and store legal information in the "Legal Regulations" category and case law information in the "Case Law" category. Appropriate indexes are set up to streamline searching and querying.

[0836] Preparation for interrogation

[0837] User login

[0838] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server via HTTPS. The server verifies the user information against the database and performs authentication.

[0839] Entering information about the person being interrogated

[0840] The user enters the basic information of the person being interrogated into the terminal. This information includes name, age, address, and a summary of the case. This information is sent to the server in JSON format.

[0841] Question set generation

[0842] The server retrieves relevant information from the database and generates a set of questions using the Python "GPT-3" API. For example, in the case of a traffic accident, it forms questions based on the category "traffic laws" and sends the constructed questions to the terminal.

[0843] Using an Emotion Engine

[0844] Acquisition of emotional data

[0845] The device acquires user emotion data during questioning. It uses the "Google Cloud Speech-to-Text" API for voice analysis and "OpenCV" for facial expression analysis. The acquired emotion data is sent to the server in real time.

[0846] Analysis of emotional data

[0847] The server analyzes the emotional data it receives using TensorFlow. An algorithm is then executed to evaluate the stress level and veracity of the interrogated person based on changes in voice tone and facial expressions.

[0848] Applications of emotional data

[0849] The server adjusts the progress of the interrogation based on the analysis results. If a high stress level is detected, the server may lower the difficulty of the question set it generates or send a message to the terminal suggesting a break.

[0850] Conducting an interrogation

[0851] Start of interrogation

[0852] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded using a microphone, and the device sends that data to a server.

[0853] Sending and analyzing response data

[0854] The terminal sends the recorded audio data to the server. The server uses "NLTK" or "SpaCy" to convert the audio data into text and compares it with legal and case law data to analyze for inconsistencies and inconsistencies.

[0855] Report generation

[0856] The server generates a report based on the final analysis results. The report includes a summary of all the interviewee's responses, inconsistencies, and a stress assessment based on sentiment analysis. The generated report is sent from the server to the terminal in PDF format and presented to the user.

[0857] Examples

[0858] Example 1: Investigation of a car accident

[0859] The server collects and analyzes case data related to traffic laws and automobile accidents, and stores it in a database. The user logs into the system using a terminal and enters information about the person being investigated. Based on a set of questions generated by the server, the user asks questions to the person being investigated. The person's answers are recorded, and the terminal sends the data to the server. The server generates a report, including any inconsistencies, based on the analysis results, and presents it to the user via the terminal.

[0860] Example 2: Use of an emotion engine

[0861] The user uses the emotion engine during the interrogation. The device analyzes the subject's facial expressions in real time and sends emotion data to the server. The server analyzes the emotion data and indicates that the subject is likely experiencing stress. The server adjusts the set of questions, which the device then presents to the user.

[0862] Examples of prompts for generative AI models

[0863] "Based on precedents related to traffic accidents, please automatically generate a set of questions for those being interrogated. During this process, use an emotion analysis engine to assess whether the subject is experiencing stress, and adjust the difficulty of the questions based on the results."

[0864] The above describes a specific implementation of the system program.

[0865] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0866] Step 1: Data Collection

[0867] The server accesses online resources and collects data related to laws and precedents. The server obtains legal data using APIs provided by government agencies and collects precedent data from court websites using web scraping techniques. Specifically, it obtains legal data in JSON format via the "e-Stat" API and uses the "BeautifulSoup" library to analyze precedent data from HTML pages and extract necessary information. Input is legal analysis information from online resources, and output is the acquired legal data and precedent data.

[0868] Step 2: Data Analysis

[0869] The server analyzes the collected data using natural language processing techniques. This analysis utilizes Python's "NLTK" and "SpaCy." The server processes the collected data through steps such as word tokenization, text mining, sentiment analysis, and topic modeling. The input is the collected raw data, and the output consists of analyzed key information and its results.

[0870] Step 3: Storing in the database

[0871] The server stores the parsed data in a database. Specifically, it uses a MySQL database, storing legal information in the "Legal Regulations" table and case law information in the "Case Law" table. It also creates indexes to improve search efficiency. The input is the parsed data, and the output is the data stored in the database and the indexes.

[0872] Step 4: User Login

[0873] A user logs into the system using a terminal. The user enters their ID and password, and this information is sent from the terminal to the server via HTTPS. The server compares this information with the user's information in the database and performs authentication. The input is the user's ID and password, and the output is the authentication result (success or failure) response.

[0874] Step 5: Enter information about the person being investigated.

[0875] The user enters basic information about the person being interrogated (name, age, address, case summary, etc.) into a terminal, and this information is sent from the terminal to the server in JSON format. The server stores the received information in a database. The input is the basic information of the person being interrogated, and the output is the information stored in the database.

[0876] Step 6: Generate the question set

[0877] The server retrieves relevant information from the database and generates a set of questions using the Python "GPT-3" API. Specifically, in the case of a traffic accident, it forms questions based on the category "traffic laws" and sends the constructed questions to the terminal. The input is relevant information from the database, and the output is the generated set of questions.

[0878] Step 7: Acquiring emotional data

[0879] The device acquires user emotion data during questioning. The "Google Cloud Speech-to-Text" API is used for speech analysis, and "OpenCV" is used for facial expression analysis. The emotion data acquired by the device is sent to the server in real time. The input is the user's emotion data, and the output is the analyzed data sent to the server.

[0880] Step 8: Analyzing emotional data

[0881] The server analyzes the emotional data it receives using TensorFlow. It then runs an algorithm that evaluates the stress level and truthfulness of the interrogated person based on changes in voice tone and facial expressions. The input is emotional data, and the output is the evaluation result of the stress level and truthfulness.

[0882] Step 9: Applying emotional data

[0883] The server adjusts the progress of the interrogation based on the analysis results. If a high stress level is detected, the server lowers the difficulty of the generated question set or generates a message suggesting a break and sends it to the terminal. The input is the evaluation result, and the output is the adjusted question set or the break suggestion message.

[0884] Step 10: Start of Interrogation

[0885] The user uses a device to generate questions and ask them to the person being interrogated. The person being interrogated's responses are recorded using a microphone, and the device sends this data to a server. The input is the person being interrogated's responses, and the output is the audio data sent to the server.

[0886] Step 11: Submitting and analyzing response data

[0887] The terminal sends the recorded audio data to the server. The server uses "NLTK" or "SpaCy" to convert the audio data into text and analyzes inconsistencies and inconsistencies by comparing it with legal and case law data. The input is the audio data, and the output is the converted text data and its analysis results.

[0888] Step 12: Generate the report

[0889] The server generates a report based on the final analysis results. The report includes a summary of all the interviewee's responses, inconsistencies, and a stress assessment based on sentiment analysis. The generated report is sent from the server to the terminal in PDF format and presented to the user. The input is the analysis results, and the output is a report in PDF format.

[0890] (Application Example 2)

[0891] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0892] In modern society, fraud and illegal activities are spreading with the increase in online transactions and communications. Traditional interrogation methods require manual collection and analysis of information related to laws and precedents, which is time-consuming, costly, and inconsistent in accuracy. Furthermore, it is difficult to properly assess the emotions and stress levels of those being interrogated, which can affect the progress of the interrogation. There is a need to solve these problems.

[0893] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for acquiring data related to laws and precedents from online resources, means for analyzing the acquired data using natural language processing technology and storing it in a database, means for acquiring relevant information from the database and generating a set of questions based on the information of the person being interrogated entered by the user, means for receiving the suspect's response data and detecting inconsistencies and unnatural points by comparing it with the data on laws and precedents, means for analyzing the user's emotional data using an emotion engine and adjusting the progress of the interrogation, means for generating a report based on the analysis results and presenting it to the user, means for analyzing the user's text and image data and detecting fraudulent activity, and means for evaluating and warning about the risk of fraud based on the analysis results. This enables the rapid detection and prevention of fraudulent activity and fraud in online transactions and communications, as well as improved accuracy and efficiency of interrogations.

[0894] "Data related to laws and precedents" refers to all information provided by the government and courts regarding laws and past court precedents.

[0895] "Online resources" refer to digital information sources such as websites, databases, and APIs that can be accessed via the internet.

[0896] "Means of acquisition" refers to the technologies and processes that a server uses to collect data from online resources.

[0897] "Natural language processing technology" refers to algorithms and models used to analyze text data and extract meaningful information.

[0898] "Means of storing data in a database" refers to systems and methods that efficiently store acquired and analyzed data, and enable searching and retrieval as needed.

[0899] "Information about the person being interrogated" refers to basic data about the person being interrogated, such as name, age, address, and a summary of the case.

[0900] "Methods for generating question sets" refers to technologies that automatically create a list of questions to conduct an interrogation based on data collected in advance.

[0901] A "suspect" refers to an individual who is under investigation and is suspected of having committed a specific act.

[0902] "Inconsistencies or unnatural points" refer to parts of the response data that do not match or are inconsistent with legal regulations or case law data.

[0903] An "emotion engine" refers to technology that analyzes a user's emotional state and takes appropriate measures based on that analysis.

[0904] "Means of adjusting the difficulty of questions" refers to systems or methods that dynamically change the content and complexity of questions during an interrogation, taking into account the user's emotional state.

[0905] "Methods for generating reports" refers to technology that automatically creates reports summarizing the results of interrogations and key points based on analysis results.

[0906] "Means of detecting fraudulent activity" refers to algorithms and technologies that analyze users' text and image data to identify fraud and other fraudulent activities at an early stage.

[0907] "Means of assessing and warning about the risk of fraud" refers to a system that, based on analysis results, notifies or warns users if it is determined that there is a high probability of fraud.

[0908] The system according to the present invention aims to rapidly detect and prevent fraud and misconduct in online transactions and communications. This system consists of multiple means for acquiring data related to laws and precedents, sentiment analysis using a sentiment engine, detection of misconduct, and risk assessment based on the results.

[0909] Knowledge acquisition and analysis

[0910] Data collection

[0911] The server accesses online resources via the internet to obtain data related to laws and precedents. Specifically, this involves collecting digital data provided by governments and courts via APIs, and collecting data from publicly available websites using web scraping techniques. This ensures that the latest legal and precedent data is always up-to-date.

[0912] Data Analysis

[0913] The server analyzes the acquired data using natural language processing techniques. Text mining, sentiment analysis, and topic modeling are applied, and the analyzed data is classified into categories. This allows for efficient searching of relevant information.

[0914] Storage in database

[0915] The analyzed data is stored in a database. Information related to laws is stored in the legal regulations category, and information related to case law is stored in the case law category, and indexes are created. This allows for quick searching and query processing.

[0916] Questionnaire generation and investigation

[0917] Entering information about the person being interrogated

[0918] The user uses a terminal to input basic information about the person being interrogated into the system. This information includes the person's name, age, address, and a summary of the case. The terminal sends this information to the server, which retrieves the relevant information from the database.

[0919] Question set generation

[0920] The server retrieves relevant information from the database and generates a set of questions. Based on the retrieved information, appropriate questions are automatically created. For example, if there is suspicion of fraud related to online transactions, questions based on applicable laws and precedents will be generated.

[0921] Using emotion engines for fraud detection

[0922] Acquisition and analysis of emotional data

[0923] The terminal acquires emotional data from the user and the subject during questioning. Using voice analysis and facial recognition technology, it identifies emotions from speech and analyzes facial expressions using a camera. The server analyzes this data using an emotion engine and evaluates the emotional state.

[0924] Applications of emotional data

[0925] The server adjusts the interrogation process based on the analyzed emotional data. For example, if the person is experiencing high stress levels, it may lower the difficulty of the questions or suggest a break. It also displays a warning to the user if it determines there is a high risk of fraud.

[0926] Report generation and output

[0927] Sending and analyzing answer data

[0928] The user uses a device to ask questions of the person being interrogated and records the answer data. This data is recorded on the device in text or audio format and sent to a server. The server uses natural language processing technology to analyze the answer data and compare it with legal and case law data to detect inconsistencies and inconsistencies.

[0929] Report generation and presentation

[0930] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The generated report is presented to the user via the terminal.

[0931] Specific example

[0932] For example, if a suspicious transaction occurs during an online transaction, this system analyzes the chat content in real time and uses an emotion engine to detect the risk of fraud. If it is determined to be "high risk," it displays a warning to the user and generates a detailed report.

[0933] Example prompts for using generative AI models

[0934] "Analyze the transaction details and assess the risk of fraud. Transaction details: [Specific transaction details] User sentiment: [Details of sentiment state] Present the analysis results in a list format."

[0935] This system will be a powerful tool for early detection and prevention of fraudulent activity in online transactions and communications.

[0936] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0937] Step 1:

[0938] The server retrieves data related to laws and precedents from online resources. Specifically, it collects digital data provided by governments and courts via API connections. It may also collect data from publicly available websites using web scraping techniques. The server retrieves this data and stores it in a database. The input is data from online resources, and the output is analytical data stored in the database.

[0939] Step 2:

[0940] The server analyzes the acquired data using natural language processing techniques. Specifically, it applies text mining, sentiment analysis, and topic modeling. As a result of the analysis, the data is classified into categories and stored again in the database. The input is raw data acquired from online resources, and the output is analyzed data classified into categories.

[0941] Step 3:

[0942] The user uses a terminal to input basic information about the person being interrogated (name, age, address, case summary, etc.). The entered information is sent from the terminal to the server. The input is the basic information of the person being interrogated, and the output is the user's input information sent to the server.

[0943] Step 4:

[0944] The server retrieves relevant information from the database based on the information of the person being interrogated and generates a set of questions. The set of questions is automatically generated to include appropriate questions for the person being interrogated. The input is the basic information of the person being interrogated, and the output is the generated set of questions.

[0945] Step 5:

[0946] The device acquires emotional data from the user or the person being interrogated. This emotional data is obtained using voice analysis and facial recognition technology. Specifically, voice analysis is performed to identify emotions from the user's statements, and facial expressions are analyzed using a camera. The inputs are audio and camera footage, and the output is the analyzed emotional data.

[0947] Step 6:

[0948] The server analyzes the acquired emotion data. Using an emotion engine, it evaluates the user's emotional state (stress level, likelihood of lying, etc.) and generates indicators. The input is emotion data, and the output is the emotion analysis result.

[0949] Step 7:

[0950] The server adjusts the interview process based on the analyzed sentiment data. For example, if the user is experiencing high levels of stress, it may lower the difficulty of the questions or suggest a break. The input is the sentiment analysis results, and the output is an adjusted set of questions or a break suggestion.

[0951] Step 8:

[0952] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in audio or text format and sent from the device to the server. The input is the person's responses, and the output is the response data sent to the server.

[0953] Step 9:

[0954] The server analyzes the response data using natural language processing techniques and compares it with legal and case law data to detect inconsistencies and inconsistencies. The input is the response data, and the output is the analysis results.

[0955] Step 10:

[0956] The server generates a report based on the analysis results and presents it to the user via the terminal. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The input is the analysis results, and the output is the generated report.

[0957] This allows for a clear explanation of the overall system processing flow and the specific actions performed at each step.

[0958] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0959] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0960] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0961] [Third Embodiment]

[0962] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0963] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0964] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0965] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0966] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0967] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0968] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0969] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0970] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0971] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0972] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0973] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0974] The system according to the present invention is for automatically acquiring, analyzing, and using data related to laws and precedents for investigation purposes. This system consists of a server, terminals, and users.

[0975] Knowledge acquisition and analysis

[0976] 1. Data Collection

[0977] The server accesses online resources and collects data related to laws and precedents. This includes information from government legal databases and court case databases.

[0978] 2. Data Analysis

[0979] The data collected by the server is analyzed using natural language processing techniques. Specifically, techniques such as text mining and topic modeling are applied to understand the content of the data.

[0980] 3. Storing in a database

[0981] The server analyzes the data and stores it in a database. This enables efficient searching and management.

[0982] Preparation for interrogation

[0983] 4. User login

[0984] The user logs into the system using their device. User authentication information is sent from the device to the server.

[0985] 5. Entering information about the person being interrogated.

[0986] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, and a summary of the case.

[0987] 6. Generating the Question Set

[0988] The server retrieves relevant information from the database and generates a set of questions. For example, in the case of an investigation into a traffic accident, questions based on traffic laws and relevant precedents will be generated.

[0989] Conducting an interrogation

[0990] 7. Commencement of Interrogation

[0991] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in either audio or text format.

[0992] 8. Data transmission and analysis

[0993] The terminal sends the recorded response data to the server. After receiving the response data, the server analyzes it by comparing it with legal and case law data that has been collected in advance.

[0994] 9. Detection of inconsistencies

[0995] The server uses natural language processing technology to analyze the content of the response and detect inconsistencies and unnatural points. For example, it might detect if a statement like "The car in front of me suddenly stopped at the intersection" contradicts the circumstantial evidence at the scene.

[0996] 10. Report generation

[0997] The server generates a report based on the analysis results. The report includes details about any inconsistencies or inconsistencies detected.

[0998] Example: Investigation of a car accident

[0999] As a concrete example, let's explain the investigation in a car accident.

[1000] 1. Data Collection

[1001] The server collects and analyzes case law data related to traffic laws and automobile accidents, and stores it in a database.

[1002] 2. User login

[1003] The user logs into the system using a terminal and enters the information of Mr. Tanaka, the person being investigated.

[1004] 3. Generating the Question Set

[1005] The server retrieves information about traffic accidents related to Mr. Tanaka and generates an appropriate set of questions.

[1006] 4. Conducting the interrogation

[1007] The user asks Mr. Tanaka, "Could you tell me the details of what happened in the accident?" Mr. Tanaka replies, "The car in front of me stopped suddenly at the intersection, so I also slammed on the brakes."

[1008] 5. Data transmission and analysis

[1009] The terminal records Mr. Tanaka's responses in audio or text format and sends them to the server. The server analyzes the data and detects any inconsistencies.

[1010] 6. Report generation

[1011] Based on the analysis results, the server generates a report pointing out inconsistencies between Mr. Tanaka's statements and the circumstantial evidence at the scene, and presents it to the user via the terminal.

[1012] As described above, the system according to the present invention improves the efficiency of police officers' work and enhances the accuracy of interrogations by acquiring, analyzing, and supporting legal and case law data.

[1013] The following describes the processing flow.

[1014] Step 1:

[1015] Data collection

[1016] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It can also collect data from publicly available websites using web scraping techniques.

[1017] Step 2:

[1018] Data Analysis

[1019] The server analyzes the collected data using natural language processing techniques. These include techniques such as text mining, sentiment analysis, and entity extraction. The analyzed data is categorized and organized by content.

[1020] Step 3:

[1021] Storage in database

[1022] The server stores the analyzed data in a database. Specifically, information related to laws is stored in the "Legal Matters" category, and information related to case law is stored in the "Case Law" category. Furthermore, indexes are created to streamline searching and query processing.

[1023] Step 4:

[1024] User login

[1025] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server performs authentication and grants permission to log in.

[1026] Step 5:

[1027] Entering information about the person being interrogated

[1028] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, address, and a summary of the case. The terminal then transmits this information to the server.

[1029] Step 6:

[1030] Question set generation

[1031] The server retrieves information related to the interrogation from the database and generates a set of questions. For example, if the person being interrogated is involved in a traffic accident, questions will be generated based on traffic laws and relevant past court precedents. The generated set of questions is then sent to the terminal.

[1032] Step 7:

[1033] Start of interrogation

[1034] The user uses a device to ask the person being interrogated system-generated questions. The person being interrogated's responses are recorded by the device in either audio or text format.

[1035] Step 8:

[1036] Sending response data

[1037] The device sends the recorded response data to the server. The data sent includes the interviewee's voice data, text data, or both.

[1038] Step 9:

[1039] Analysis and matching of response data

[1040] The server analyzes the received response data using natural language processing technology. The analyzed response data is then compared with legal and case law data stored in a database beforehand to detect inconsistencies and unnatural points.

[1041] Step 10:

[1042] Report generation

[1043] The server generates a report based on the analysis results. The report includes details about inconsistencies and unnatural points in the interrogated person's statements. The generated report is sent to the terminal and presented to the user.

[1044] Step 11:

[1045] Confirmation of results and further questioning

[1046] The user reviews the report using a terminal. If necessary, the user creates additional questions and conducts further investigations based on these questions. The additional questions are sent to the server for further analysis.

[1047] By repeating these steps, the accuracy and efficiency of interrogations can be improved.

[1048] (Example 1)

[1049] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1050] The current interrogation system requires a tremendous amount of time and effort because data collection and analysis based on laws and precedents are done manually. Furthermore, the quality and accuracy of the question sets generated during interrogations are low, making it difficult to efficiently detect inconsistencies or unnatural points in the interviewee's responses. Therefore, there is a need to improve the accuracy and efficiency of interrogations.

[1051] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1052] In this invention, the server includes means for acquiring data related to laws and precedents from online resources, means for analyzing the acquired data using natural language processing technology and storing it in a database, means for acquiring relevant information from the database based on information about the subject of investigation entered by the user and generating a set of questions using a generative AI model, means for receiving the subject's response data and detecting inconsistencies and unnatural points by comparing it with the data on laws and precedents, and means for generating a report based on the analysis results and presenting it to the user. This not only improves the quality of interrogations but also enables highly accurate and rapid interrogations.

[1053] "Online resources" refer to websites and databases that exist on the internet and contain data on laws and precedents provided by public institutions, companies, and other organizations.

[1054] "Legal and case law data" refers to digital data provided by government agencies and courts in various countries, including legal texts and the content of past court precedents.

[1055] "Natural language processing technology" refers to techniques that enable computers to understand and analyze human language, and includes methods such as text mining and topic modeling.

[1056] A "database" is a system for efficiently storing, managing, and retrieving large amounts of data, and includes relational databases and NoSQL databases.

[1057] A "subject of investigation" is a person who is the subject of questioning or investigation, and their basic information and information related to the case are entered into the system.

[1058] A "generative AI model" is a model that uses AI technology to automatically generate question sets and reports based on given input data. Examples include GPT-3 and similar models.

[1059] A "question set" refers to a series of questions asked of the person being investigated, and is automatically generated by an AI model according to the purpose of the interrogation or investigation.

[1060] "Response data" refers to the data of responses provided by survey participants to questions, and is recorded in audio or text format.

[1061] "Inconsistencies or unnatural points" refer to parts of the survey respondents' responses that do not match or are logically inconsistent with the legal and case law data collected and analyzed in advance.

[1062] A "report" is a document that summarizes the analysis results and includes details about any inconsistencies or inconsistencies that were detected.

[1063] This invention relates to a system for automatically acquiring and analyzing data related to laws and precedents, and for efficiently conducting interrogations. This system consists of a server, terminals, and users.

[1064] Knowledge acquisition and analysis

[1065] 1. Data Collection

[1066] The server collects data related to laws and precedents from online resources on the internet. For example, it accesses government-provided legal databases and court case databases. Specifically, it uses Python crawling tools (e.g., Scrapy) to periodically crawl these databases and retrieve the necessary information.

[1067] 2. Data Analysis

[1068] The server analyzes the collected data using natural language processing (NLP) techniques. It applies text mining libraries (e.g., NLTK, SpaCy) and topic modeling techniques (e.g., LDA) to understand the data's content and extract relevant information.

[1069] 3. Storing in a database

[1070] The server stores the analyzed data in a database. Because fast and efficient data management is required here, relational databases (e.g., MySQL, PostgreSQL) or NoSQL databases (e.g., MongoDB) are used.

[1071] Preparation for interrogation

[1072] 4. User login

[1073] Users log in to the system using their device. User authentication is performed by sending data from the device to the server, where security measures, including two-factor authentication, are implemented and the authentication process is completed.

[1074] 5. Entering information about the person being interrogated.

[1075] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, and a summary of the case. The entered data is stored in a database on the server.

[1076] 6. Generating the Question Set

[1077] The server retrieves relevant information from the database and generates a set of questions using a generative AI model (e.g., GPT-3). The generated set of questions includes the most appropriate questions for the research subject, based on the legal case.

[1078] Conducting an interrogation

[1079] 7. Commencement of Interrogation

[1080] The user asks the interviewee questions generated using the device. The interviewee's responses are recorded on the device in either audio or text format, and if audio data is used, it is converted to text using speech recognition technology (e.g., Google Cloud Speech-to-Text).

[1081] 8. Data transmission and analysis

[1082] The terminal sends the recorded response data to the server. The server compares the received data with legal and case law data that has been collected and analyzed in advance, and performs analysis using NLP technology (e.g., BERT).

[1083] 9. Detection of inconsistencies

[1084] The server evaluates the responses based on the analysis results and detects inconsistencies and unnatural points. It also checks whether the statements are consistent with the facts by cross-referencing them with legal databases.

[1085] 10. Report generation

[1086] The server generates a report based on the analysis results and presents it to the user. The report includes details about any inconsistencies or unnatural points detected.

[1087] Examples of specific cases and prompt statements

[1088] As a concrete example, let's consider the investigation of a traffic accident.

[1089] 1. The server collects and analyzes case law data related to traffic laws and automobile accidents, and stores it in a database.

[1090] 2. The user logs into the system using a terminal and enters the information of Mr. Tanaka, the person being investigated.

[1091] 3. The server retrieves information about traffic accidents related to Mr. Tanaka and generates an appropriate set of questions.

[1092] Example of a prompt:

[1093] "Please generate a set of questions for an investigation into a traffic accident. The subject of the investigation is Mr. Tanaka, and we would like to ask about the detailed circumstances of the accident. Please include relevant traffic laws and precedents in the questions."

[1094] As described above, the system according to the present invention aims to improve the quality and efficiency of interrogations through the automatic acquisition, analysis, and examination of legal and case law data.

[1095] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1096] Step 1: Data Collection

[1097] The server collects data related to laws and precedents from online resources on the internet. Specifically, it uses Python crawling tools (e.g., Scrapy) to crawl government law databases and court precedent databases. The input is a list of URLs to be crawled, and the output is the retrieved HTML code and text data. The retrieved data is temporarily stored.

[1098] Specific operation: The server executes a Scrapy script, accesses the URL of the specified legal database, and collects the latest case law data.

[1099] Step 2: Data Analysis

[1100] The server analyzes the collected data using natural language processing (NLP) techniques. Specifically, it extracts keywords and topics from the collected text data using text mining libraries (e.g., NLTK, SpaCy). The input is the collected text data, and the output is the analyzed keyword and topic information.

[1101] Specific operation: The server uses the NLTK library to analyze legal terms contained in the collected data and extract important keywords and their relevance.

[1102] Step 3: Storing in the database

[1103] The server stores the analyzed data in a database. This can be a relational database (e.g., MySQL) or a NoSQL database (e.g., MongoDB). The input is the analyzed data, and the output is the completed storage of the data in the database.

[1104] Specific operation: The server stores the parsed data into MySQL using an INSERT statement.

[1105] Step 4: User Login

[1106] The user logs into the system using a terminal. The user ID and password are input data, which are sent to the server. The server compares this information with the authentication information in the database and sends the authentication result back to the terminal. The output indicates whether the login was successful or failed.

[1107] Specific operation: The user enters their ID and password on the login screen, and the server verifies this against the database information to perform authentication.

[1108] Step 5: Enter information about the person being investigated.

[1109] The user enters basic information about the person being interrogated into the terminal. This includes the person's name, age, and a summary of the case. The entered data is sent to the server, which stores it in a database. The output includes confirmation that the data has been successfully saved to the database.

[1110] Specific operation: The user enters Mr. Tanaka's information (name, age, details of the incident) into an input form, and the server receives this information and saves it to the database.

[1111] Step 6: Generate the question set

[1112] The server retrieves relevant information from the database and generates a set of questions using a generative AI model (e.g., GPT-3). Inputs include basic information about the person being interrogated and relevant legal and case law data, while output is the generated set of questions.

[1113] Specific operation: The server sends a prompt to the GPT-3 model saying "Generate a set of questions for a traffic accident investigation," retrieves the response, and constructs the question set.

[1114] Step 7: Start of interrogation

[1115] The user uses a device to ask questions generated by the device to the person being interviewed. The interviewee's responses are recorded on the device in either audio or text format. The input consists of the generated set of questions and the interviewee's responses, and the output is the recorded response data.

[1116] Specific operation: The user asks Mr. Tanaka, "Please tell me the details of what happened when the accident occurred," records Mr. Tanaka's voice response on the device, and converts the voice data into text using Google Cloud Speech-to-Text.

[1117] Step 8: Data transmission and analysis

[1118] The terminal sends the recorded response data to the server. The server compares the received data with pre-collected and analyzed legal and case law data and performs analysis using natural language processing techniques (e.g., BERT). The input consists of the response data and related legal and case law data, and the output is the analysis results.

[1119] Specific operation: The terminal sends the text-based response to the server, and the server performs contextual analysis of the response using a BERT model.

[1120] Step 9: Detecting inconsistencies

[1121] The server evaluates the answers based on the analysis results and detects inconsistencies and unnatural points. The input is the analysis results, and the output is the detected inconsistencies and unnatural points.

[1122] Specific operation: The server compares the case data for the relevant case with the response content and detects if the statement "the car in front suddenly stopped at the intersection" contradicts the situation at the accident scene.

[1123] Step 10: Generate the report

[1124] The server generates a report based on the analysis results and presents it to the user. The input consists of the analysis results and detected inconsistencies, while the output is the generated report.

[1125] Specific operation: The server compiles the analysis results, generates a report pointing out inconsistencies, and sends this report to the terminal in PDF format.

[1126] (Application Example 1)

[1127] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1128] In on-site interrogations and investigations, it is difficult for security personnel to ask efficient and accurate questions based on laws and precedents, and to immediately detect inconsistencies or suspicious points in answers. While rapid information acquisition and analysis on-site are required, conventional methods lack real-time capabilities, leading to decreased accuracy and efficiency in investigations. To address this challenge, there is a need for a system that effectively and quickly utilizes legal and precedent data to support on-site interrogations.

[1129] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1130] In this invention, the server includes means for acquiring data related to laws and precedents from online resources; means for analyzing the acquired data using natural language processing technology and storing it in a database; means for acquiring relevant information from the database based on information about the person being interrogated entered by the user and generating a set of questions; means for a security officer to display the questions using a head-mounted display and record the answers using speech recognition technology; means for receiving the recorded answer data and comparing it with the data on laws and precedents to detect inconsistencies and inconsistencies; and means for generating a report based on the analysis results and presenting it to the security officer. This enables rapid and accurate interrogations on-site.

[1131] "Online resources" is a general term for information sources such as databases and websites that exist on the internet.

[1132] A "law" is a set of rules and norms established by a state or local government that have the power to enforce order in society.

[1133] A "precedent" is a record of judgments and their reasons that have been handed down by courts in the past, and it serves as a reference for similar cases that may follow.

[1134] "Natural language processing technology" refers to techniques for understanding, analyzing, and generating human language using computers, and examples include text mining and topic modeling.

[1135] A "database" is a general term for a system that organizes and stores large amounts of data, and allows for efficient searching and updating.

[1136] A "person under investigation" refers to an individual who is investigated or questioned in order to provide information about a specific incident or situation.

[1137] A "question set" is a collection of related questions grouped together to achieve a specific objective.

[1138] A "security implementer" refers to an individual or organization that takes on the role of ensuring safety in a specific location or situation.

[1139] A "head-mounted display" is a type of display device worn on the head that displays information within the field of vision.

[1140] "Speech recognition technology" is a technology that analyzes speech data and converts it into text data.

[1141] A "contradiction" refers to a part of specific data or statements that is not logically consistent.

[1142] An "unnatural point" refers to a part of specific data or statements that is judged to deviate from what is normal or common sense.

[1143] A "report" is a document that summarizes analysis results or research findings, and is a type of report.

[1144] The present invention's system collects data related to laws and precedents from online resources, analyzes it using natural language processing technology, and generates question sets and analyzes answers for on-site interrogations and investigations. This system mainly consists of a server, a head-mounted display, and a user.

[1145] Collection of legal and case law data

[1146] The server automatically retrieves data related to laws and precedents from databases and websites on the internet. For example, government legal databases and court case databases are used.

[1147] Data analysis and database storage

[1148] The acquired data is analyzed on the server using natural language processing techniques. Specifically, techniques such as text mining and topic modeling are applied to extract important topics and case precedents. The analysis results are stored in a database, enabling efficient searching and management.

[1149] Question set generation

[1150] The user logs into the system via a head-mounted display and enters basic information about the person being interrogated. The server retrieves relevant information from the database and automatically generates an appropriate set of questions. This enables rapid and accurate interrogations on-site.

[1151] Speech recognition and analysis of responses

[1152] Security personnel use a head-mounted display to show generated questions and ask them to the person being interrogated. Responses are recorded in real time using speech recognition technology. The recorded data is sent to a server where it is compared with legal and case law data to detect inconsistencies and inconsistencies.

[1153] Report generation and presentation

[1154] The server generates a report based on the analysis results. The report includes details of any inconsistencies or anomalies detected. This report is presented to security personnel via a head-mounted display to assist with on-site investigations.

[1155] Examples of specific cases and prompt statements

[1156] As a concrete example of use, consider a scenario where a security officer is conducting an on-site investigation of a traffic accident. In this case, the officer inputs "legal data regarding traffic accidents" into the system via a head-mounted display and is instructed to generate "questions regarding procedures at the time of the accident."

[1157] Examples of prompt messages include the following:

[1158] "Please obtain legal data regarding traffic accidents and generate questions about procedures to be followed when an accident occurs."

[1159] In summary, the system of the present invention enables rapid and accurate on-site interrogations and supports security personnel through the efficient use of legal and case law data.

[1160] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1161] Step 1:

[1162] The server collects data related to laws and precedents from online resources.

[1163] Input: URLs of government legal databases, court case databases, etc.

[1164] Data processing: Web scraping is performed to extract necessary text information from HTML.

[1165] Output: Text data of collected laws and precedents.

[1166] Specific operation: The server uses a web scraping tool to retrieve HTML content from a specified URL and extracts text information using a library such as BeautifulSoup.

[1167] Step 2:

[1168] The server analyzes the acquired data using natural language processing technology and stores it in a database.

[1169] Input: Text data of collected laws and precedents.

[1170] Data processing: Extract important topics and case precedents through text mining and topic modeling.

[1171] Output: Analyzed data and its topic information.

[1172] Specific operation: The server uses natural language processing libraries such as spaCy to parse text data and stores the data in a database such as SQLite.

[1173] Step 3:

[1174] The user logs into the system using a terminal and enters the information of the person being investigated.

[1175] Input: Basic information of the person being questioned (name, age, summary of the case).

[1176] Data processing: Perform user authentication and format checks on input information.

[1177] Output: Information about the person being interrogated is stored in the database.

[1178] Specific operation: The terminal sends the user's login information to the server, the server performs authentication, receives the information of the person being investigated, and stores it in the database.

[1179] Step 4:

[1180] The server retrieves relevant information from the database and generates a set of questions.

[1181] Input: Information on the person being interrogated and collected legal and case law data.

[1182] Data processing: Use a generative AI model to generate relevant questions.

[1183] Output: The generated set of questions.

[1184] Specific operation: The server retrieves relevant information from the database based on the query and uses a generative AI model (e.g., GPT-3) to generate the query.

[1185] Step 5:

[1186] Security personnel use a head-mounted display to show questions, and their answers are recorded using speech recognition technology.

[1187] Input: The generated set of questions.

[1188] Data processing: Speech recognition technology is used to convert the responses recorded via voice input into text.

[1189] Output: Text data of the recorded responses.

[1190] Specific operation: The head-mounted display presents a set of questions to the security officer and records the officer's voice input in real time, converting it to text.

[1191] Step 6:

[1192] The server receives the recorded response data and compares it with legal and case law data to detect inconsistencies and inconsistencies.

[1193] Input: Text data of the recorded response.

[1194] Data processing: Natural language processing techniques are used to compare and analyze response data with legal and case law data.

[1195] Output: Detection results for inconsistencies and unnatural points.

[1196] Specific operation: The server receives text data, performs matching and analysis using natural language processing techniques, and extracts the detection results.

[1197] Step 7:

[1198] The server generates a report based on the analysis results and presents it to the security implementer.

[1199] Input: Results of detecting inconsistencies and unnatural points.

[1200] Data processing: Based on the detection results, a report is generated, and the data is formatted for visual display.

[1201] Output: Final report.

[1202] Specific operation: The server generates a detailed report based on the detection results and presents it to the security officer via a head-mounted display.

[1203] The above outlines the specific processing flow of the system program that implements the application example.

[1204] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1205] The system according to this invention automatically acquires and analyzes data related to laws and precedents, and uses it to aid in interrogations. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the accuracy and effectiveness of interrogations are improved. The system consists of a server, terminals, and users, and its detailed operation is described below.

[1206] Knowledge acquisition and analysis

[1207] Data collection

[1208] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It is also possible to collect data from publicly available websites using web scraping techniques.

[1209] Data Analysis

[1210] The data collected by the server is analyzed using natural language processing techniques. Specifically, techniques such as text mining, sentiment analysis, and topic modeling are applied. The analyzed data is classified into categories according to its intended use.

[1211] Storage in database

[1212] The server stores the analyzed data in a database. Legal information is stored in the "Legal Matters" category, and information on case precedents is stored in the "Case Precedents" category. An index is created to streamline searching and query processing.

[1213] Preparation for interrogation

[1214] User login

[1215] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server performs authentication and grants permission to log in.

[1216] Entering information about the person being interrogated

[1217] The user enters the basic information of the person being interrogated into the terminal. This information includes the person's name, age, address, and a summary of the case. The terminal then transmits this information to the server.

[1218] Question set generation

[1219] The server retrieves relevant information from the database and generates a set of questions. For example, in the case of a traffic accident, questions based on traffic laws and precedents are generated. The generated set of questions is then sent to the terminal.

[1220] Using an Emotion Engine

[1221] Acquisition of emotional data

[1222] The device acquires user emotion data during questioning. This can be done by using voice analysis to identify emotions from the user's statements, or by using the device's camera to analyze facial expressions and recognize emotions.

[1223] Analysis of emotional data

[1224] The server analyzes emotional data acquired using an emotion engine. This analysis generates indicators to determine the user's stress level, the likelihood of them lying, and other factors.

[1225] Applications of emotional data

[1226] The server adjusts the interrogation process based on the analyzed emotional data. For example, if the user is experiencing high levels of stress, it may lower the difficulty of the questions or suggest a break. The analysis results are presented to the user via their device.

[1227] Conducting an interrogation

[1228] Start of interrogation

[1229] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in either audio or text format.

[1230] Sending and analyzing response data

[1231] The terminal sends the recorded response data to the server. The server analyzes the response data using natural language processing technology and compares it with legal and case law data to detect inconsistencies and inconsistencies.

[1232] Report generation

[1233] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on sentiment data. The generated report is sent to the terminal and presented to the user.

[1234] Examples

[1235] Example 1: Investigation of a car accident

[1236] The server collects and analyzes case data related to traffic laws and automobile accidents, and stores it in a database. The user logs into the system using a terminal and enters information about Mr. Tanaka, the person being investigated. Based on a set of questions generated by the server, the user asks Mr. Tanaka questions. Mr. Tanaka's answers are recorded, and the terminal sends the data to the server. The server generates a report, including any inconsistencies, based on the analysis results, and presents it to the user via the terminal.

[1237] Example 2: Use of an emotion engine

[1238] The user uses the emotion engine during the interrogation. The device analyzes Mr. Tanaka's facial expressions in real time and sends emotion data to the server. The server analyzes the emotion data and indicates that Mr. Tanaka is likely experiencing stress. The server adjusts the set of questions, and the device presents them to the user.

[1239] This system can significantly improve the accuracy and efficiency of interrogations by combining legal and case law data with sentiment analysis.

[1240] The following describes the processing flow.

[1241] Step 1:

[1242] Data collection

[1243] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It can also collect data from publicly available websites using web scraping techniques.

[1244] Step 2:

[1245] Data Analysis

[1246] The server analyzes the collected data using natural language processing techniques. Techniques such as text mining, sentiment analysis, and entity extraction are applied to understand the data's content and classify it into categories appropriate for its intended use.

[1247] Step 3:

[1248] Storage in database

[1249] The server stores the analyzed data in a database. Legal information is stored in the "Legal Matters" category, and information on case law is stored in the "Case Law" category. Indexes are also created for efficient searching and query processing.

[1250] Step 4:

[1251] User login

[1252] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server authenticates the user and grants permission to log in.

[1253] Step 5:

[1254] Entering information about the person being interrogated

[1255] The user uses a terminal to enter basic information about the person being interrogated. This information includes the person's name, age, address, and a summary of the case. The terminal then sends this information to the server.

[1256] Step 6:

[1257] Question set generation

[1258] The server retrieves information related to the person being interrogated from the database and generates a set of questions. For example, in the case of a traffic accident, specific questions based on traffic laws and relevant precedents are generated. The generated set of questions is then sent to the terminal.

[1259] Step 7:

[1260] Acquisition of emotional data

[1261] The device acquires emotional data from the user and the person being interrogated during questioning. Methods include identifying emotions from speech using voice analysis, and recognizing emotions by analyzing facial expressions using the device's camera.

[1262] Step 8:

[1263] Analysis of emotional data

[1264] The server analyzes emotional data acquired using an emotion engine. It generates indicators to determine the stress level of users and those being interrogated, as well as the likelihood of them lying.

[1265] Step 9:

[1266] Start of interrogation

[1267] The user uses a device to ask the person being interrogated the generated questions. The person being interrogated's responses are recorded on the device in either audio or text format.

[1268] Step 10:

[1269] Sending response data

[1270] The device sends the recorded response data to the server. The data sent includes the interviewee's voice data, text data, or both.

[1271] Step 11:

[1272] Analysis and matching of response data

[1273] The server analyzes the received response data using natural language processing technology. The analyzed response data is then compared with legal and case law data stored in a database to detect inconsistencies and inconsistencies.

[1274] Step 12:

[1275] Report generation

[1276] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The generated report is sent to the terminal and presented to the user.

[1277] Step 13:

[1278] Confirmation of results and further questioning

[1279] The user reviews the report using a terminal. If necessary, the user creates additional questions and conducts further investigations based on these questions. The additional questions are sent to the server for further analysis.

[1280] Through these specific processing steps, the accuracy and efficiency of interrogations can be significantly improved.

[1281] (Example 2)

[1282] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1283] While conventional interrogation systems have a certain degree of accuracy in generating questions based on laws and precedents and analyzing responses, a challenge lies in their lack of adjustments that take into account the emotions of the person being interrogated. Furthermore, there is a need for methods to achieve more accurate and fair interrogations by acquiring and analyzing the emotional data of the person being interrogated.

[1284] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1285] In this invention, the server includes means for acquiring data related to laws and precedents from online resources; means for analyzing the acquired data using natural language processing technology and storing it in a database; means for acquiring relevant information from the database and generating a set of questions based on information about the person being interrogated entered by the user; means for collecting emotional data of the person being interrogated using a terminal and analyzing it using emotion analysis technology; means for adjusting the progress of the interrogation in real time based on the analyzed emotional data; means for receiving the interrogation response data, analyzing it using natural language processing technology, and detecting inconsistencies and unnatural points by comparing it with legal and precedent data; and means for generating a report based on the analysis results and presenting it to the user. This makes it possible to achieve highly accurate and fair interrogations while taking into account the emotional data of the person being interrogated.

[1286] "Online resources" refer to information sources such as websites, databases, and APIs that are accessible via the internet.

[1287] A "database" is a system for storing structured data and for efficiently searching and managing it.

[1288] "Natural language processing technology" refers to technologies for understanding, analyzing, and generating human language using computers, and includes text mining, sentiment analysis, and topic modeling.

[1289] A "question set" is a collection of multiple questions asked for a specific purpose or target.

[1290] "Emotional analysis technology" is a technology that analyzes a person's emotional state from data such as audio and video, and is used to assess stress levels and truthfulness.

[1291] A "person under investigation" is a person who is being investigated in relation to laws or precedents.

[1292] A "terminal" refers to a device, such as a computer or smartphone, that a user uses to access a system.

[1293] "Response data" refers to data on the content of the answers given by the person being interrogated to the questions.

[1294] A "report" is a document that summarizes the results of an investigation, and includes analysis results, inconsistencies, and inconsistencies.

[1295] "Real-time" refers to processing occurring simultaneously with the event or with an extremely short delay.

[1296] The system according to this invention automatically acquires and analyzes data related to laws and precedents, and uses it to aid in interrogations. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the accuracy and effectiveness of interrogations are improved. This system consists of a server, terminals, and users.

[1297] Knowledge acquisition and analysis

[1298] Data collection

[1299] The server accesses online resources and collects data related to laws and precedents. Specifically, it obtains legal data using APIs provided by government agencies and collects precedent data from court websites using web scraping techniques. For example, legal data can be obtained using the "e-Stat" API, and precedent data can be collected using the "BeautifulSoup" library.

[1300] Data Analysis

[1301] The server analyzes the collected data using natural language processing techniques. Python's "NLTK" and "SpaCy" are used for the analysis. Specific processing includes text mining, sentiment analysis, and topic modeling. This extracts important parts of the collected text data and classifies them into categories according to their intended use.

[1302] Storage in database

[1303] The analyzed data is stored in a database by the server. For example, a MySQL database is used to categorize and store legal information in the "Legal Regulations" category and case law information in the "Case Law" category. Appropriate indexes are set up to streamline searching and querying.

[1304] Preparation for interrogation

[1305] User login

[1306] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server via HTTPS. The server verifies the user information against the database and performs authentication.

[1307] Entering information about the person being interrogated

[1308] The user enters the basic information of the person being interrogated into the terminal. This information includes name, age, address, and a summary of the case. This information is sent to the server in JSON format.

[1309] Question set generation

[1310] The server retrieves relevant information from the database and generates a set of questions using the Python "GPT-3" API. For example, in the case of a traffic accident, it forms questions based on the category "traffic laws" and sends the constructed questions to the terminal.

[1311] Using an Emotion Engine

[1312] Acquisition of emotional data

[1313] The device acquires user emotion data during questioning. It uses the "Google Cloud Speech-to-Text" API for voice analysis and "OpenCV" for facial expression analysis. The acquired emotion data is sent to the server in real time.

[1314] Analysis of emotional data

[1315] The server analyzes the emotional data it receives using TensorFlow. An algorithm is then executed to evaluate the stress level and veracity of the interrogated person based on changes in voice tone and facial expressions.

[1316] Applications of emotional data

[1317] The server adjusts the progress of the interrogation based on the analysis results. If a high stress level is detected, the server may lower the difficulty of the question set it generates or send a message to the terminal suggesting a break.

[1318] Conducting an interrogation

[1319] Start of interrogation

[1320] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded using a microphone, and the device sends that data to a server.

[1321] Sending and analyzing response data

[1322] The terminal sends the recorded audio data to the server. The server uses "NLTK" or "SpaCy" to convert the audio data into text and compares it with legal and case law data to analyze for inconsistencies and inconsistencies.

[1323] Report generation

[1324] The server generates a report based on the final analysis results. The report includes a summary of all the interviewee's responses, inconsistencies, and a stress assessment based on sentiment analysis. The generated report is sent from the server to the terminal in PDF format and presented to the user.

[1325] Examples

[1326] Example 1: Investigation of a car accident

[1327] The server collects and analyzes case data related to traffic laws and automobile accidents, and stores it in a database. The user logs into the system using a terminal and enters information about the person being investigated. Based on a set of questions generated by the server, the user asks questions to the person being investigated. The person's answers are recorded, and the terminal sends the data to the server. The server generates a report, including any inconsistencies, based on the analysis results, and presents it to the user via the terminal.

[1328] Example 2: Use of an emotion engine

[1329] The user uses the emotion engine during the interrogation. The device analyzes the subject's facial expressions in real time and sends emotion data to the server. The server analyzes the emotion data and indicates that the subject is likely experiencing stress. The server adjusts the set of questions, which the device then presents to the user.

[1330] Examples of prompts for generative AI models

[1331] "Based on precedents related to traffic accidents, please automatically generate a set of questions for those being interrogated. During this process, use an emotion analysis engine to assess whether the subject is experiencing stress, and adjust the difficulty of the questions based on the results."

[1332] The above describes a specific implementation of the system program.

[1333] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1334] Step 1: Data Collection

[1335] The server accesses online resources and collects data related to laws and precedents. The server obtains legal data using APIs provided by government agencies and collects precedent data from court websites using web scraping techniques. Specifically, it obtains legal data in JSON format via the "e-Stat" API and uses the "BeautifulSoup" library to analyze precedent data from HTML pages and extract necessary information. Input is legal analysis information from online resources, and output is the acquired legal data and precedent data.

[1336] Step 2: Data Analysis

[1337] The server analyzes the collected data using natural language processing techniques. This analysis utilizes Python's "NLTK" and "SpaCy." The server processes the collected data through steps such as word tokenization, text mining, sentiment analysis, and topic modeling. The input is the collected raw data, and the output consists of analyzed key information and its results.

[1338] Step 3: Storing in the database

[1339] The server stores the parsed data in a database. Specifically, it uses a MySQL database, storing legal information in the "Legal Regulations" table and case law information in the "Case Law" table. It also creates indexes to improve search efficiency. The input is the parsed data, and the output is the data stored in the database and the indexes.

[1340] Step 4: User Login

[1341] A user logs into the system using a terminal. The user enters their ID and password, and this information is sent from the terminal to the server via HTTPS. The server compares this information with the user's information in the database and performs authentication. The input is the user's ID and password, and the output is the authentication result (success or failure) response.

[1342] Step 5: Enter information about the person being investigated.

[1343] The user enters basic information about the person being interrogated (name, age, address, case summary, etc.) into a terminal, and this information is sent from the terminal to the server in JSON format. The server stores the received information in a database. The input is the basic information of the person being interrogated, and the output is the information stored in the database.

[1344] Step 6: Generate the question set

[1345] The server retrieves relevant information from the database and generates a set of questions using the Python "GPT-3" API. Specifically, in the case of a traffic accident, it forms questions based on the category "traffic laws" and sends the constructed questions to the terminal. The input is relevant information from the database, and the output is the generated set of questions.

[1346] Step 7: Acquiring emotional data

[1347] The device acquires user emotion data during questioning. The "Google Cloud Speech-to-Text" API is used for speech analysis, and "OpenCV" is used for facial expression analysis. The emotion data acquired by the device is sent to the server in real time. The input is the user's emotion data, and the output is the analyzed data sent to the server.

[1348] Step 8: Analyzing emotional data

[1349] The server analyzes the emotional data it receives using TensorFlow. It then runs an algorithm that evaluates the stress level and truthfulness of the interrogated person based on changes in voice tone and facial expressions. The input is emotional data, and the output is the evaluation result of the stress level and truthfulness.

[1350] Step 9: Applying emotional data

[1351] The server adjusts the progress of the interrogation based on the analysis results. If a high stress level is detected, the server lowers the difficulty of the generated question set or generates a message suggesting a break and sends it to the terminal. The input is the evaluation result, and the output is the adjusted question set or the break suggestion message.

[1352] Step 10: Start of Interrogation

[1353] The user uses a device to generate questions and ask them to the person being interrogated. The person being interrogated's responses are recorded using a microphone, and the device sends this data to a server. The input is the person being interrogated's responses, and the output is the audio data sent to the server.

[1354] Step 11: Submitting and analyzing response data

[1355] The terminal sends the recorded audio data to the server. The server uses "NLTK" or "SpaCy" to convert the audio data into text and analyzes inconsistencies and inconsistencies by comparing it with legal and case law data. The input is the audio data, and the output is the converted text data and its analysis results.

[1356] Step 12: Generate the report

[1357] The server generates a report based on the final analysis results. The report includes a summary of all the interviewee's responses, inconsistencies, and a stress assessment based on sentiment analysis. The generated report is sent from the server to the terminal in PDF format and presented to the user. The input is the analysis results, and the output is a report in PDF format.

[1358] (Application Example 2)

[1359] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1360] In modern society, fraud and illegal activities are spreading with the increase in online transactions and communications. Traditional interrogation methods require manual collection and analysis of information related to laws and precedents, which is time-consuming, costly, and inconsistent in accuracy. Furthermore, it is difficult to properly assess the emotions and stress levels of those being interrogated, which can affect the progress of the interrogation. There is a need to solve these problems.

[1361] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for acquiring data related to laws and precedents from online resources, means for analyzing the acquired data using natural language processing technology and storing it in a database, means for acquiring relevant information from the database and generating a set of questions based on the information of the person being interrogated entered by the user, means for receiving the suspect's response data and detecting inconsistencies and unnatural points by comparing it with the data on laws and precedents, means for analyzing the user's emotional data using an emotion engine and adjusting the progress of the interrogation, means for generating a report based on the analysis results and presenting it to the user, means for analyzing the user's text and image data and detecting fraudulent activity, and means for evaluating and warning about the risk of fraud based on the analysis results. This enables the rapid detection and prevention of fraudulent activity and fraud in online transactions and communications, as well as improved accuracy and efficiency of interrogations.

[1362] "Data related to laws and precedents" refers to all information provided by the government and courts regarding laws and past court precedents.

[1363] "Online resources" refer to digital information sources such as websites, databases, and APIs that can be accessed via the internet.

[1364] "Means of acquisition" refers to the technologies and processes that a server uses to collect data from online resources.

[1365] "Natural language processing technology" refers to algorithms and models used to analyze text data and extract meaningful information.

[1366] "Means of storing data in a database" refers to systems and methods that efficiently store acquired and analyzed data, and enable searching and retrieval as needed.

[1367] "Information about the person being interrogated" refers to basic data about the person being interrogated, such as name, age, address, and a summary of the case.

[1368] "Methods for generating question sets" refers to technologies that automatically create a list of questions to conduct an interrogation based on data collected in advance.

[1369] A "suspect" refers to an individual who is under investigation and is suspected of having committed a specific act.

[1370] "Inconsistencies or unnatural points" refer to parts of the response data that do not match or are inconsistent with legal regulations or case law data.

[1371] An "emotion engine" refers to technology that analyzes a user's emotional state and takes appropriate measures based on that analysis.

[1372] "Means of adjusting the difficulty of questions" refers to systems or methods that dynamically change the content and complexity of questions during an interrogation, taking into account the user's emotional state.

[1373] "Methods for generating reports" refers to technology that automatically creates reports summarizing the results of interrogations and key points based on analysis results.

[1374] "Means of detecting fraudulent activity" refers to algorithms and technologies that analyze users' text and image data to identify fraud and other fraudulent activities at an early stage.

[1375] "Means of assessing and warning about the risk of fraud" refers to a system that, based on analysis results, notifies or warns users if it is determined that there is a high probability of fraud.

[1376] The system according to the present invention aims to rapidly detect and prevent fraud and misconduct in online transactions and communications. This system consists of multiple means for acquiring data related to laws and precedents, sentiment analysis using a sentiment engine, detection of misconduct, and risk assessment based on the results.

[1377] Knowledge acquisition and analysis

[1378] Data collection

[1379] The server accesses online resources via the internet to obtain data related to laws and precedents. Specifically, this involves collecting digital data provided by governments and courts via APIs, and collecting data from publicly available websites using web scraping techniques. This ensures that the latest legal and precedent data is always up-to-date.

[1380] Data Analysis

[1381] The server analyzes the acquired data using natural language processing techniques. Text mining, sentiment analysis, and topic modeling are applied, and the analyzed data is classified into categories. This allows for efficient searching of relevant information.

[1382] Storage in database

[1383] The analyzed data is stored in a database. Information related to laws is stored in the legal regulations category, and information related to case law is stored in the case law category, and indexes are created. This allows for quick searching and query processing.

[1384] Questionnaire generation and investigation

[1385] Entering information about the person being interrogated

[1386] The user uses a terminal to input basic information about the person being interrogated into the system. This information includes the person's name, age, address, and a summary of the case. The terminal sends this information to the server, which retrieves the relevant information from the database.

[1387] Question set generation

[1388] The server retrieves relevant information from the database and generates a set of questions. Based on the retrieved information, appropriate questions are automatically created. For example, if there is suspicion of fraud related to online transactions, questions based on applicable laws and precedents will be generated.

[1389] Using emotion engines for fraud detection

[1390] Acquisition and analysis of emotional data

[1391] The terminal acquires emotional data from the user and the subject during questioning. Using voice analysis and facial recognition technology, it identifies emotions from speech and analyzes facial expressions using a camera. The server analyzes this data using an emotion engine and evaluates the emotional state.

[1392] Applications of emotional data

[1393] The server adjusts the interrogation process based on the analyzed emotional data. For example, if the person is experiencing high stress levels, it may lower the difficulty of the questions or suggest a break. It also displays a warning to the user if it determines there is a high risk of fraud.

[1394] Report generation and output

[1395] Sending and analyzing answer data

[1396] The user uses a device to ask questions of the person being interrogated and records the answer data. This data is recorded on the device in text or audio format and sent to a server. The server uses natural language processing technology to analyze the answer data and compare it with legal and case law data to detect inconsistencies and inconsistencies.

[1397] Report generation and presentation

[1398] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The generated report is presented to the user via the terminal.

[1399] Specific example

[1400] For example, if a suspicious transaction occurs during an online transaction, this system analyzes the chat content in real time and uses an emotion engine to detect the risk of fraud. If it is determined to be "high risk," it displays a warning to the user and generates a detailed report.

[1401] Example prompts for using generative AI models

[1402] "Analyze the transaction details and assess the risk of fraud. Transaction details: [Specific transaction details] User sentiment: [Details of sentiment state] Present the analysis results in a list format."

[1403] This system will be a powerful tool for early detection and prevention of fraudulent activity in online transactions and communications.

[1404] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1405] Step 1:

[1406] The server retrieves data related to laws and precedents from online resources. Specifically, it collects digital data provided by governments and courts via API connections. It may also collect data from publicly available websites using web scraping techniques. The server retrieves this data and stores it in a database. The input is data from online resources, and the output is analytical data stored in the database.

[1407] Step 2:

[1408] The server analyzes the acquired data using natural language processing techniques. Specifically, it applies text mining, sentiment analysis, and topic modeling. As a result of the analysis, the data is classified into categories and stored again in the database. The input is raw data acquired from online resources, and the output is analyzed data classified into categories.

[1409] Step 3:

[1410] The user uses a terminal to input basic information about the person being interrogated (name, age, address, case summary, etc.). The entered information is sent from the terminal to the server. The input is the basic information of the person being interrogated, and the output is the user's input information sent to the server.

[1411] Step 4:

[1412] The server retrieves relevant information from the database based on the information of the person being interrogated and generates a set of questions. The set of questions is automatically generated to include appropriate questions for the person being interrogated. The input is the basic information of the person being interrogated, and the output is the generated set of questions.

[1413] Step 5:

[1414] The device acquires emotional data from the user or the person being interrogated. This emotional data is obtained using voice analysis and facial recognition technology. Specifically, voice analysis is performed to identify emotions from the user's statements, and facial expressions are analyzed using a camera. The inputs are audio and camera footage, and the output is the analyzed emotional data.

[1415] Step 6:

[1416] The server analyzes the acquired emotional data. Using an emotion engine, it evaluates the user's emotional state (stress level, likelihood of lying, etc.) and generates indicators. The input is emotional data, and the output is the result of the emotion analysis.

[1417] Step 7:

[1418] The server adjusts the interview process based on the analyzed sentiment data. For example, if the user is experiencing high levels of stress, it may lower the difficulty of the questions or suggest a break. The input is the sentiment analysis results, and the output is an adjusted set of questions or a break suggestion.

[1419] Step 8:

[1420] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in audio or text format and sent from the device to the server. The input is the person's responses, and the output is the response data sent to the server.

[1421] Step 9:

[1422] The server analyzes the response data using natural language processing techniques and compares it with legal and case law data to detect inconsistencies and inconsistencies. The input is the response data, and the output is the analysis results.

[1423] Step 10:

[1424] The server generates a report based on the analysis results and presents it to the user via the terminal. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The input is the analysis results, and the output is the generated report.

[1425] This allows for a clear explanation of the overall system processing flow and the specific actions performed at each step.

[1426] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1427] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1428] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1429] [Fourth Embodiment]

[1430] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1431] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1432] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1433] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1434] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1435] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1436] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1437] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1438] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1439] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1440] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1441] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1442] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1443] The system according to the present invention is for automatically acquiring, analyzing, and using data related to laws and precedents for investigation purposes. This system consists of a server, terminals, and users.

[1444] Knowledge acquisition and analysis

[1445] 1. Data Collection

[1446] The server accesses online resources and collects data related to laws and precedents. This includes information from government legal databases and court case databases.

[1447] 2. Data Analysis

[1448] The data collected by the server is analyzed using natural language processing techniques. Specifically, techniques such as text mining and topic modeling are applied to understand the content of the data.

[1449] 3. Storing in a database

[1450] The server analyzes the data and stores it in a database. This enables efficient searching and management.

[1451] Preparation for interrogation

[1452] 4. User login

[1453] The user logs into the system using their device. User authentication information is sent from the device to the server.

[1454] 5. Entering information about the person being interrogated.

[1455] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, and a summary of the case.

[1456] 6. Generating the Question Set

[1457] The server retrieves relevant information from the database and generates a set of questions. For example, in the case of an investigation into a traffic accident, questions based on traffic laws and relevant precedents will be generated.

[1458] Conducting an interrogation

[1459] 7. Commencement of Interrogation

[1460] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in either audio or text format.

[1461] 8. Data transmission and analysis

[1462] The terminal sends the recorded response data to the server. After receiving the response data, the server analyzes it by comparing it with legal and case law data that has been collected in advance.

[1463] 9. Detection of inconsistencies

[1464] The server uses natural language processing technology to analyze the content of the response and detect inconsistencies and unnatural points. For example, it might detect if a statement like "The car in front of me suddenly stopped at the intersection" contradicts the circumstantial evidence at the scene.

[1465] 10. Report generation

[1466] The server generates a report based on the analysis results. The report includes details about any inconsistencies or inconsistencies detected.

[1467] Example: Investigation of a car accident

[1468] As a concrete example, let's explain the investigation in a car accident.

[1469] 1. Data Collection

[1470] The server collects and analyzes case law data related to traffic laws and automobile accidents, and stores it in a database.

[1471] 2. User login

[1472] The user logs into the system using a terminal and enters the information of Mr. Tanaka, the person being investigated.

[1473] 3. Generating the Question Set

[1474] The server retrieves information about traffic accidents related to Mr. Tanaka and generates an appropriate set of questions.

[1475] 4. Conducting the interrogation

[1476] The user asks Mr. Tanaka, "Could you tell me the details of what happened in the accident?" Mr. Tanaka replies, "The car in front of me stopped suddenly at the intersection, so I also slammed on the brakes."

[1477] 5. Data transmission and analysis

[1478] The terminal records Mr. Tanaka's responses in audio or text format and sends them to the server. The server analyzes the data and detects any inconsistencies.

[1479] 6. Report generation

[1480] Based on the analysis results, the server generates a report pointing out inconsistencies between Mr. Tanaka's statements and the circumstantial evidence at the scene, and presents it to the user via the terminal.

[1481] As described above, the system according to the present invention improves the efficiency of police officers' work and enhances the accuracy of interrogations by acquiring, analyzing, and supporting legal and case law data.

[1482] The following describes the processing flow.

[1483] Step 1:

[1484] Data collection

[1485] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It can also collect data from publicly available websites using web scraping techniques.

[1486] Step 2:

[1487] Data Analysis

[1488] The server analyzes the collected data using natural language processing techniques. These include techniques such as text mining, sentiment analysis, and entity extraction. The analyzed data is categorized and organized by content.

[1489] Step 3:

[1490] Storage in database

[1491] The server stores the analyzed data in a database. Specifically, information related to laws is stored in the "Legal Matters" category, and information related to case law is stored in the "Case Law" category. Furthermore, indexes are created to streamline searching and query processing.

[1492] Step 4:

[1493] User login

[1494] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server performs authentication and grants permission to log in.

[1495] Step 5:

[1496] Entering information about the person being interrogated

[1497] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, address, and a summary of the case. The terminal then transmits this information to the server.

[1498] Step 6:

[1499] Question set generation

[1500] The server retrieves information related to the interrogation from the database and generates a set of questions. For example, if the person being interrogated is involved in a traffic accident, questions will be generated based on traffic laws and relevant past court precedents. The generated set of questions is then sent to the terminal.

[1501] Step 7:

[1502] Start of interrogation

[1503] The user uses a device to ask the person being interrogated system-generated questions. The person being interrogated's responses are recorded by the device in either audio or text format.

[1504] Step 8:

[1505] Sending response data

[1506] The device sends the recorded response data to the server. The data sent includes the interviewee's voice data, text data, or both.

[1507] Step 9:

[1508] Analysis and matching of response data

[1509] The server analyzes the received response data using natural language processing technology. The analyzed response data is then compared with legal and case law data stored in a database beforehand to detect inconsistencies and unnatural points.

[1510] Step 10:

[1511] Report generation

[1512] The server generates a report based on the analysis results. The report includes details about inconsistencies and unnatural points in the interrogated person's statements. The generated report is sent to the terminal and presented to the user.

[1513] Step 11:

[1514] Confirmation of results and further questioning

[1515] The user reviews the report using a terminal. If necessary, the user creates additional questions and conducts further investigations based on these questions. The additional questions are sent to the server for further analysis.

[1516] By repeating these steps, the accuracy and efficiency of interrogations can be improved.

[1517] (Example 1)

[1518] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1519] The current interrogation system requires a tremendous amount of time and effort because data collection and analysis based on laws and precedents are done manually. Furthermore, the quality and accuracy of the question sets generated during interrogations are low, making it difficult to efficiently detect inconsistencies or unnatural points in the interviewee's responses. Therefore, there is a need to improve the accuracy and efficiency of interrogations.

[1520] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1521] In this invention, the server includes means for acquiring data related to laws and precedents from online resources, means for analyzing the acquired data using natural language processing technology and storing it in a database, means for acquiring relevant information from the database based on information about the subject of investigation entered by the user and generating a set of questions using a generative AI model, means for receiving the subject's response data and detecting inconsistencies and unnatural points by comparing it with the data on laws and precedents, and means for generating a report based on the analysis results and presenting it to the user. This not only improves the quality of interrogations but also enables highly accurate and rapid interrogations.

[1522] "Online resources" refer to websites and databases that exist on the internet and contain data on laws and precedents provided by public institutions, companies, and other organizations.

[1523] "Legal and case law data" refers to digital data provided by government agencies and courts in various countries, including legal texts and the content of past court precedents.

[1524] "Natural language processing technology" refers to techniques that enable computers to understand and analyze human language, and includes methods such as text mining and topic modeling.

[1525] A "database" is a system for efficiently storing, managing, and retrieving large amounts of data, and includes relational databases and NoSQL databases.

[1526] A "subject of investigation" is a person who is the subject of questioning or investigation, and their basic information and information related to the case are entered into the system.

[1527] A "generative AI model" is a model that uses AI technology to automatically generate question sets and reports based on given input data. Examples include GPT-3 and similar models.

[1528] A "question set" refers to a series of questions asked of the person being investigated, and is automatically generated by an AI model according to the purpose of the interrogation or investigation.

[1529] "Response data" refers to the data of responses provided by survey participants to questions, and is recorded in audio or text format.

[1530] "Inconsistencies or unnatural points" refer to parts of the survey respondents' responses that do not match or are logically inconsistent with the legal and case law data collected and analyzed in advance.

[1531] A "report" is a document that summarizes the analysis results and includes details about any inconsistencies or inconsistencies that were detected.

[1532] This invention relates to a system for automatically acquiring and analyzing data related to laws and precedents, and for efficiently conducting interrogations. This system consists of a server, terminals, and users.

[1533] Knowledge acquisition and analysis

[1534] 1. Data Collection

[1535] The server collects data related to laws and precedents from online resources on the internet. For example, it accesses government-provided legal databases and court case databases. Specifically, it uses Python crawling tools (e.g., Scrapy) to periodically crawl these databases and retrieve the necessary information.

[1536] 2. Data Analysis

[1537] The server analyzes the collected data using natural language processing (NLP) techniques. It applies text mining libraries (e.g., NLTK, SpaCy) and topic modeling techniques (e.g., LDA) to understand the data's content and extract relevant information.

[1538] 3. Storing in a database

[1539] The server stores the analyzed data in a database. Because fast and efficient data management is required here, relational databases (e.g., MySQL, PostgreSQL) or NoSQL databases (e.g., MongoDB) are used.

[1540] Preparation for interrogation

[1541] 4. User login

[1542] Users log in to the system using their device. User authentication is performed by sending data from the device to the server, where security measures, including two-factor authentication, are implemented and the authentication process is completed.

[1543] 5. Entering information about the person being interrogated.

[1544] The user enters basic information about the person being interrogated into the terminal. This information includes the person's name, age, and a summary of the case. The entered data is stored in a database on the server.

[1545] 6. Generating the Question Set

[1546] The server retrieves relevant information from the database and generates a set of questions using a generative AI model (e.g., GPT-3). The generated set of questions includes the most appropriate questions for the research subject, based on the legal case.

[1547] Conducting an interrogation

[1548] 7. Commencement of Interrogation

[1549] The user asks the interviewee questions generated using the device. The interviewee's responses are recorded on the device in either audio or text format, and if audio data is used, it is converted to text using speech recognition technology (e.g., Google Cloud Speech-to-Text).

[1550] 8. Data transmission and analysis

[1551] The terminal sends the recorded response data to the server. The server compares the received data with legal and case law data that has been collected and analyzed in advance, and performs analysis using NLP technology (e.g., BERT).

[1552] 9. Detection of inconsistencies

[1553] The server evaluates the responses based on the analysis results and detects inconsistencies and unnatural points. It also checks whether the statements are consistent with the facts by cross-referencing them with legal databases.

[1554] 10. Report generation

[1555] The server generates a report based on the analysis results and presents it to the user. The report includes details about any inconsistencies or unnatural points detected.

[1556] Examples of specific cases and prompt statements

[1557] As a concrete example, let's consider the investigation of a traffic accident.

[1558] 1. The server collects and analyzes case law data related to traffic laws and automobile accidents, and stores it in a database.

[1559] 2. The user logs into the system using a terminal and enters the information of Mr. Tanaka, the person being investigated.

[1560] 3. The server retrieves information about traffic accidents related to Mr. Tanaka and generates an appropriate set of questions.

[1561] Example of a prompt:

[1562] "Please generate a set of questions for an investigation into a traffic accident. The subject of the investigation is Mr. Tanaka, and we would like to ask about the detailed circumstances of the accident. Please include relevant traffic laws and precedents in the questions."

[1563] As described above, the system according to the present invention aims to improve the quality and efficiency of interrogations through the automatic acquisition, analysis, and examination of legal and case law data.

[1564] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1565] Step 1: Data Collection

[1566] The server collects data related to laws and precedents from online resources on the internet. Specifically, it uses Python crawling tools (e.g., Scrapy) to crawl government law databases and court precedent databases. The input is a list of URLs to be crawled, and the output is the retrieved HTML code and text data. The retrieved data is temporarily stored.

[1567] Specific operation: The server executes a Scrapy script, accesses the URL of the specified legal database, and collects the latest case law data.

[1568] Step 2: Data Analysis

[1569] The server analyzes the collected data using natural language processing (NLP) techniques. Specifically, it extracts keywords and topics from the collected text data using text mining libraries (e.g., NLTK, SpaCy). The input is the collected text data, and the output is the analyzed keyword and topic information.

[1570] Specific operation: The server uses the NLTK library to analyze legal terms contained in the collected data and extract important keywords and their relevance.

[1571] Step 3: Storing in the database

[1572] The server stores the analyzed data in a database. This can be a relational database (e.g., MySQL) or a NoSQL database (e.g., MongoDB). The input is the analyzed data, and the output is the completed storage of the data in the database.

[1573] Specific operation: The server stores the parsed data into MySQL using an INSERT statement.

[1574] Step 4: User Login

[1575] The user logs into the system using a terminal. The user ID and password are input data, which are sent to the server. The server compares this information with the authentication information in the database and sends the authentication result back to the terminal. The output indicates whether the login was successful or failed.

[1576] Specific operation: The user enters their ID and password on the login screen, and the server verifies this against the database information to perform authentication.

[1577] Step 5: Enter information about the person being investigated.

[1578] The user enters basic information about the person being interrogated into the terminal. This includes the person's name, age, and a summary of the case. The entered data is sent to the server, which stores it in a database. The output includes confirmation that the data has been successfully saved to the database.

[1579] Specific operation: The user enters Mr. Tanaka's information (name, age, details of the incident) into an input form, and the server receives this information and saves it to the database.

[1580] Step 6: Generate the question set

[1581] The server retrieves relevant information from the database and generates a set of questions using a generative AI model (e.g., GPT-3). Inputs include basic information about the person being interrogated and relevant legal and case law data, while output is the generated set of questions.

[1582] Specific operation: The server sends a prompt to the GPT-3 model saying "Generate a set of questions for a traffic accident investigation," retrieves the response, and constructs the question set.

[1583] Step 7: Start of interrogation

[1584] The user uses a device to ask questions generated by the device to the person being interviewed. The interviewee's responses are recorded on the device in either audio or text format. The input consists of the generated set of questions and the interviewee's responses, and the output is the recorded response data.

[1585] Specific operation: The user asks Mr. Tanaka, "Please tell me the details of what happened when the accident occurred," records Mr. Tanaka's voice response on the device, and converts the voice data into text using Google Cloud Speech-to-Text.

[1586] Step 8: Data transmission and analysis

[1587] The terminal sends the recorded response data to the server. The server compares the received data with pre-collected and analyzed legal and case law data and performs analysis using natural language processing techniques (e.g., BERT). The input consists of the response data and related legal and case law data, and the output is the analysis results.

[1588] Specific operation: The terminal sends the text-based response to the server, and the server performs contextual analysis of the response using a BERT model.

[1589] Step 9: Detecting inconsistencies

[1590] The server evaluates the answers based on the analysis results and detects inconsistencies and unnatural points. The input is the analysis results, and the output is the detected inconsistencies and unnatural points.

[1591] Specific operation: The server compares the case data for the relevant case with the response content and detects if the statement "the car in front suddenly stopped at the intersection" contradicts the situation at the accident scene.

[1592] Step 10: Generate the report

[1593] The server generates a report based on the analysis results and presents it to the user. The input consists of the analysis results and detected inconsistencies, while the output is the generated report.

[1594] Specific operation: The server compiles the analysis results, generates a report pointing out inconsistencies, and sends this report to the terminal in PDF format.

[1595] (Application Example 1)

[1596] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1597] In on-site interrogations and investigations, it is difficult for security personnel to ask efficient and accurate questions based on laws and precedents, and to immediately detect inconsistencies or suspicious points in answers. While rapid information acquisition and analysis on-site are required, conventional methods lack real-time capabilities, leading to decreased accuracy and efficiency in investigations. To address this challenge, there is a need for a system that effectively and quickly utilizes legal and precedent data to support on-site interrogations.

[1598] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1599] In this invention, the server includes means for acquiring data related to laws and precedents from online resources; means for analyzing the acquired data using natural language processing technology and storing it in a database; means for acquiring relevant information from the database based on information about the person being interrogated entered by the user and generating a set of questions; means for a security officer to display the questions using a head-mounted display and record the answers using speech recognition technology; means for receiving the recorded answer data and comparing it with the data on laws and precedents to detect inconsistencies and inconsistencies; and means for generating a report based on the analysis results and presenting it to the security officer. This enables rapid and accurate interrogations on-site.

[1600] "Online resources" is a general term for information sources such as databases and websites that exist on the internet.

[1601] A "law" is a set of rules and norms established by a state or local government that have the power to enforce order in society.

[1602] A "precedent" is a record of judgments and their reasons that have been handed down by courts in the past, and it serves as a reference for similar cases that may follow.

[1603] "Natural language processing technology" refers to techniques for understanding, analyzing, and generating human language using computers, and examples include text mining and topic modeling.

[1604] A "database" is a general term for a system that organizes and stores large amounts of data, and allows for efficient searching and updating.

[1605] A "person under investigation" refers to an individual who is investigated or questioned in order to provide information about a specific incident or situation.

[1606] A "question set" is a collection of related questions grouped together to achieve a specific objective.

[1607] A "security implementer" refers to an individual or organization that takes on the role of ensuring safety in a specific location or situation.

[1608] A "head-mounted display" is a type of display device worn on the head that displays information within the field of vision.

[1609] "Speech recognition technology" is a technology that analyzes speech data and converts it into text data.

[1610] A "contradiction" refers to a part of specific data or statements that is not logically consistent.

[1611] An "unnatural point" refers to a part of specific data or statements that is judged to deviate from what is normal or common sense.

[1612] A "report" is a document that summarizes analysis results or research findings, and is a type of report.

[1613] The system of this invention collects data related to laws and precedents from online resources, analyzes it using natural language processing technology, and generates question sets and analyzes answers for on-site interrogations and investigations. This system mainly consists of a server, a head-mounted display, and a user.

[1614] Collection of legal and case law data

[1615] The server automatically retrieves data related to laws and precedents from databases and websites on the internet. For example, government legal databases and court case databases are used.

[1616] Data analysis and database storage

[1617] The acquired data is analyzed on the server using natural language processing techniques. Specifically, techniques such as text mining and topic modeling are applied to extract important topics and case precedents. The analysis results are stored in a database, enabling efficient searching and management.

[1618] Question set generation

[1619] The user logs into the system via a head-mounted display and enters basic information about the person being interrogated. The server retrieves relevant information from the database and automatically generates an appropriate set of questions. This enables rapid and accurate interrogations on-site.

[1620] Speech recognition and analysis of responses

[1621] Security personnel use a head-mounted display to show generated questions and ask them to the person being interrogated. Responses are recorded in real time using speech recognition technology. The recorded data is sent to a server where it is compared with legal and case law data to detect inconsistencies and inconsistencies.

[1622] Report generation and presentation

[1623] The server generates a report based on the analysis results. The report includes details of any inconsistencies or anomalies detected. This report is presented to security personnel via a head-mounted display to assist with on-site investigations.

[1624] Examples of specific cases and prompt statements

[1625] As a concrete example of use, consider a scenario where a security officer is conducting an on-site investigation of a traffic accident. In this case, the officer inputs "legal data regarding traffic accidents" into the system via a head-mounted display and is instructed to generate "questions regarding procedures at the time of the accident."

[1626] Examples of prompt messages include the following:

[1627] "Please obtain legal data regarding traffic accidents and generate questions about procedures to be followed when an accident occurs."

[1628] In summary, the system of the present invention enables rapid and accurate on-site interrogations and supports security personnel through the efficient use of legal and case law data.

[1629] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1630] Step 1:

[1631] The server collects data related to laws and precedents from online resources.

[1632] Input: URLs of government legal databases, court case databases, etc.

[1633] Data processing: Web scraping is performed to extract necessary text information from HTML.

[1634] Output: Text data of collected laws and precedents.

[1635] Specific operation: The server uses a web scraping tool to retrieve HTML content from a specified URL and extracts text information using a library such as BeautifulSoup.

[1636] Step 2:

[1637] The server analyzes the acquired data using natural language processing technology and stores it in a database.

[1638] Input: Text data of collected laws and precedents.

[1639] Data processing: Extract important topics and case precedents through text mining and topic modeling.

[1640] Output: Analyzed data and its topic information.

[1641] Specific operation: The server uses natural language processing libraries such as spaCy to parse text data and stores the data in a database such as SQLite.

[1642] Step 3:

[1643] The user logs into the system using a terminal and enters the information of the person being investigated.

[1644] Input: Basic information of the person being questioned (name, age, summary of the case).

[1645] Data processing: Perform user authentication and format checks on input information.

[1646] Output: Information about the person being interrogated is stored in the database.

[1647] Specific operation: The terminal sends the user's login information to the server, the server performs authentication, receives the information of the person being investigated, and stores it in the database.

[1648] Step 4:

[1649] The server retrieves relevant information from the database and generates a set of questions.

[1650] Input: Information on the person being interrogated and collected legal and case law data.

[1651] Data processing: Use a generative AI model to generate relevant questions.

[1652] Output: The generated set of questions.

[1653] Specific operation: The server retrieves relevant information from the database based on the query and uses a generative AI model (e.g., GPT-3) to generate the query.

[1654] Step 5:

[1655] Security personnel use a head-mounted display to show questions, and their answers are recorded using speech recognition technology.

[1656] Input: The generated set of questions.

[1657] Data processing: Speech recognition technology is used to convert the responses recorded via voice input into text.

[1658] Output: Text data of the recorded responses.

[1659] Specific operation: The head-mounted display presents a set of questions to the security officer and records the officer's voice input in real time, converting it to text.

[1660] Step 6:

[1661] The server receives the recorded response data and compares it with legal and case law data to detect inconsistencies and inconsistencies.

[1662] Input: Text data of the recorded response.

[1663] Data processing: Natural language processing techniques are used to compare and analyze response data with legal and case law data.

[1664] Output: Detection results for inconsistencies and unnatural points.

[1665] Specific operation: The server receives text data, performs matching and analysis using natural language processing techniques, and extracts the detection results.

[1666] Step 7:

[1667] The server generates a report based on the analysis results and presents it to the security implementer.

[1668] Input: Results of detecting inconsistencies and unnatural points.

[1669] Data processing: Based on the detection results, a report is generated, and the data is formatted for visual display.

[1670] Output: Final report.

[1671] Specific operation: The server generates a detailed report based on the detection results and presents it to the security officer via a head-mounted display.

[1672] The above outlines the specific processing flow of the system program that implements the application example.

[1673] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1674] The system according to this invention automatically acquires and analyzes data related to laws and precedents, and uses it to aid in interrogations. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the accuracy and effectiveness of interrogations are improved. The system consists of a server, terminals, and users, and its detailed operation is described below.

[1675] Knowledge acquisition and analysis

[1676] Data collection

[1677] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It is also possible to collect data from publicly available websites using web scraping techniques.

[1678] Data Analysis

[1679] The data collected by the server is analyzed using natural language processing techniques. Specifically, techniques such as text mining, sentiment analysis, and topic modeling are applied. The analyzed data is classified into categories according to its intended use.

[1680] Storage in database

[1681] The server stores the analyzed data in a database. Legal information is stored in the "Legal Matters" category, and information on case precedents is stored in the "Case Precedents" category. An index is created to streamline searching and query processing.

[1682] Preparation for interrogation

[1683] User login

[1684] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server performs authentication and grants permission to log in.

[1685] Entering information about the person being interrogated

[1686] The user enters the basic information of the person being interrogated into the terminal. This information includes the person's name, age, address, and a summary of the case. The terminal then transmits this information to the server.

[1687] Question set generation

[1688] The server retrieves relevant information from the database and generates a set of questions. For example, in the case of a traffic accident, questions based on traffic laws and precedents are generated. The generated set of questions is then sent to the terminal.

[1689] Using an Emotion Engine

[1690] Acquisition of emotional data

[1691] The device acquires user emotion data during questioning. This can be done by using voice analysis to identify emotions from the user's statements, or by using the device's camera to analyze facial expressions and recognize emotions.

[1692] Analysis of emotional data

[1693] The server analyzes emotional data acquired using an emotion engine. This analysis generates indicators to determine the user's stress level, the likelihood of them lying, and other factors.

[1694] Applications of emotional data

[1695] The server adjusts the interrogation process based on the analyzed emotional data. For example, if the user is experiencing high levels of stress, it may lower the difficulty of the questions or suggest a break. The analysis results are presented to the user via their device.

[1696] Conducting an interrogation

[1697] Start of interrogation

[1698] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded on the device in either audio or text format.

[1699] Sending and analyzing response data

[1700] The terminal sends the recorded response data to the server. The server analyzes the response data using natural language processing technology and compares it with legal and case law data to detect inconsistencies and inconsistencies.

[1701] Report generation

[1702] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on sentiment data. The generated report is sent to the terminal and presented to the user.

[1703] Examples

[1704] Example 1: Investigation of a car accident

[1705] The server collects and analyzes case data related to traffic laws and automobile accidents, and stores it in a database. The user logs into the system using a terminal and enters information about Mr. Tanaka, the person being investigated. Based on a set of questions generated by the server, the user asks Mr. Tanaka questions. Mr. Tanaka's answers are recorded, and the terminal sends the data to the server. The server generates a report, including any inconsistencies, based on the analysis results, and presents it to the user via the terminal.

[1706] Example 2: Use of an emotion engine

[1707] The user uses the emotion engine during the interrogation. The device analyzes Mr. Tanaka's facial expressions in real time and sends emotion data to the server. The server analyzes the emotion data and indicates that Mr. Tanaka is likely experiencing stress. The server adjusts the set of questions, and the device presents them to the user.

[1708] This system can significantly improve the accuracy and efficiency of interrogations by combining legal and case law data with sentiment analysis.

[1709] The following describes the processing flow.

[1710] Step 1:

[1711] Data collection

[1712] The server accesses online resources and collects data related to laws and precedents. Specifically, it connects to government legal databases and court case databases via API to obtain the necessary information. It can also collect data from publicly available websites using web scraping techniques.

[1713] Step 2:

[1714] Data Analysis

[1715] The server analyzes the collected data using natural language processing techniques. Techniques such as text mining, sentiment analysis, and entity extraction are applied to understand the data's content and classify it into categories appropriate for its intended use.

[1716] Step 3:

[1717] Storage in database

[1718] The server stores the analyzed data in a database. Legal information is stored in the "Legal Matters" category, and information on case precedents is stored in the "Case Precedents" category. Indexes are also created for efficient searching and query processing.

[1719] Step 4:

[1720] User login

[1721] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server. The server authenticates the user and grants permission to log in.

[1722] Step 5:

[1723] Entering information about the person being interrogated

[1724] The user uses a terminal to enter basic information about the person being interrogated. This information includes the person's name, age, address, and a summary of the case. The terminal then sends this information to the server.

[1725] Step 6:

[1726] Question set generation

[1727] The server retrieves information related to the person being interrogated from the database and generates a set of questions. For example, in the case of a traffic accident, specific questions based on traffic laws and relevant precedents are generated. The generated set of questions is then sent to the terminal.

[1728] Step 7:

[1729] Acquisition of emotional data

[1730] The device acquires emotional data from the user and the person being interrogated during questioning. Methods include identifying emotions from speech using voice analysis, and recognizing emotions by analyzing facial expressions using the device's camera.

[1731] Step 8:

[1732] Analysis of emotional data

[1733] The server analyzes emotional data acquired using an emotion engine. It generates indicators to determine the stress level of users and those being interrogated, as well as the likelihood of them lying.

[1734] Step 9:

[1735] Start of interrogation

[1736] The user uses a device to ask the person being interrogated the generated questions. The person being interrogated's responses are recorded on the device in either audio or text format.

[1737] Step 10:

[1738] Sending response data

[1739] The device sends the recorded response data to the server. The data sent includes the interviewee's voice data, text data, or both.

[1740] Step 11:

[1741] Analysis and matching of response data

[1742] The server analyzes the received response data using natural language processing technology. The analyzed response data is then compared with legal and case law data stored in a database to detect inconsistencies and inconsistencies.

[1743] Step 12:

[1744] Report generation

[1745] The server generates a report based on the analysis results. The report includes inconsistencies and unnatural points in the interrogated person's statements, as well as analysis results based on emotional data. The generated report is sent to the terminal and presented to the user.

[1746] Step 13:

[1747] Confirmation of results and further questioning

[1748] The user reviews the report using a terminal. If necessary, the user creates additional questions and conducts further investigations based on these questions. The additional questions are sent to the server for further analysis.

[1749] Through these specific processing steps, the accuracy and efficiency of interrogations can be significantly improved.

[1750] (Example 2)

[1751] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1752] While conventional interrogation systems have a certain degree of accuracy in generating questions based on laws and precedents and analyzing responses, a challenge lies in their lack of adjustments that take into account the emotions of the person being interrogated. Furthermore, there is a need for methods to achieve more accurate and fair interrogations by acquiring and analyzing the emotional data of the person being interrogated.

[1753] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1754] In this invention, the server includes means for acquiring data related to laws and precedents from online resources; means for analyzing the acquired data using natural language processing technology and storing it in a database; means for acquiring relevant information from the database and generating a set of questions based on information about the person being interrogated entered by the user; means for collecting emotional data of the person being interrogated using a terminal and analyzing it using emotion analysis technology; means for adjusting the progress of the interrogation in real time based on the analyzed emotional data; means for receiving the interrogation response data, analyzing it using natural language processing technology, and detecting inconsistencies and unnatural points by comparing it with legal and precedent data; and means for generating a report based on the analysis results and presenting it to the user. This makes it possible to achieve highly accurate and fair interrogations while taking into account the emotional data of the person being interrogated.

[1755] "Online resources" refer to information sources such as websites, databases, and APIs that are accessible via the internet.

[1756] A "database" is a system for storing structured data and for efficiently searching and managing it.

[1757] "Natural language processing technology" refers to technologies for understanding, analyzing, and generating human language using computers, and includes text mining, sentiment analysis, and topic modeling.

[1758] A "question set" is a collection of multiple questions asked for a specific purpose or target.

[1759] "Emotional analysis technology" is a technology that analyzes a person's emotional state from data such as audio and video, and is used to assess stress levels and truthfulness.

[1760] A "person under investigation" is a person who is being investigated in relation to laws or precedents.

[1761] A "terminal" refers to a device, such as a computer or smartphone, that a user uses to access a system.

[1762] "Response data" refers to data on the content of the answers given by the person being interrogated to the questions.

[1763] A "report" is a document that summarizes the results of an investigation, and includes analysis results, inconsistencies, and inconsistencies.

[1764] "Real-time" refers to processing occurring simultaneously with the event or with an extremely short delay.

[1765] The system according to this invention automatically acquires and analyzes data related to laws and precedents, and uses it to aid in interrogations. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, the accuracy and effectiveness of interrogations are improved. This system consists of a server, terminals, and users.

[1766] Knowledge acquisition and analysis

[1767] Data collection

[1768] The server accesses online resources and collects data related to laws and precedents. Specifically, it obtains legal data using APIs provided by government agencies and collects precedent data from court websites using web scraping techniques. For example, legal data can be obtained using the "e-Stat" API, and precedent data can be collected using the "BeautifulSoup" library.

[1769] Data Analysis

[1770] The server analyzes the collected data using natural language processing techniques. Python's "NLTK" and "SpaCy" are used for the analysis. Specific processing includes text mining, sentiment analysis, and topic modeling. This extracts important parts of the collected text data and classifies them into categories according to their intended use.

[1771] Storage in database

[1772] The analyzed data is stored in a database by the server. For example, a MySQL database is used to categorize and store legal information in the "Legal Regulations" category and case law information in the "Case Law" category. Appropriate indexes are set up to streamline searching and querying.

[1773] Preparation for interrogation

[1774] User login

[1775] The user logs into the system using their device. The user enters their ID and password, and the device sends this information to the server via HTTPS. The server verifies the user information against the database and performs authentication.

[1776] Entering information about the person being interrogated

[1777] The user enters the basic information of the person being interrogated into the terminal. This information includes name, age, address, and a summary of the case. This information is sent to the server in JSON format.

[1778] Question set generation

[1779] The server retrieves relevant information from the database and generates a set of questions using the Python "GPT-3" API. For example, in the case of a traffic accident, it forms questions based on the category "traffic laws" and sends the constructed questions to the terminal.

[1780] Using an Emotion Engine

[1781] Acquisition of emotional data

[1782] The device acquires user emotion data during questioning. It uses the "Google Cloud Speech-to-Text" API for voice analysis and "OpenCV" for facial expression analysis. The acquired emotion data is sent to the server in real time.

[1783] Analysis of emotional data

[1784] The server analyzes the emotional data it receives using TensorFlow. An algorithm is then executed to evaluate the stress level and veracity of the interrogated person based on changes in voice tone and facial expressions.

[1785] Applications of emotional data

[1786] The server adjusts the progress of the interrogation based on the analysis results. If a high stress level is detected, the server may lower the difficulty of the question set it generates or send a message to the terminal suggesting a break.

[1787] Conducting an interrogation

[1788] Start of interrogation

[1789] The user uses a device to generate questions and ask them to the person being interrogated. The person's responses are recorded using a microphone, and the device sends that data to a server.

[1790] Sending and analyzing response data

[1791] The terminal sends the recorded audio data to the server. The server uses "NLTK" or "SpaCy" to convert the audio data into text and compares it with legal and case law data to analyze for inconsistencies and inconsistencies.

[1792] Report generation

[1793] The server generates a report based on the final analysis results. The report includes a summary of all the interviewee's responses, inconsistencies, and a stress assessment based on sentiment analysis. The generated report is sent from the server to the terminal in PDF format and presented to the user.

[1794] Examples

[1795] Example 1: Investigation of a car accident

[1796] The server collects and analyzes case data related to traffic laws and automobile accidents, and stores it in a database. The user logs into the system using a terminal and enters information about the person being investigated. Based on a set of questions generated by the server, the user asks questions to the person being investigated. The person's answers are recorded, and the terminal sends the data to the server. The server generates a report, including any inconsistencies, based on the analysis results, and presents it to the user via the terminal.

[1797] Example 2: Use of an emotion engine

[1798] The user uses the emotion engine during the interrogation. The device analyzes the subject's facial expressions in real time and sends emotion data to the server. The server analyzes the emotion data and indicates that the subject is likely experiencing stress. The server adjusts the set of questions, which the device then presents to the user.

[1799] Examples of prompts for generative AI models

[1800] "Based on precedents related to traffic accidents, please automatically generate a set of questions for those being interrogated. During this process, use an emotion analysis engine to assess whether the subject is experiencing stress, and adjust the difficulty of the questions based on the results."

[1801] The above describes a specific implementation of the system program.

[1802] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1803] Step 1: Data Collection

[1804] The server accesses online resources and collects data related to laws and precedents. The server obtains legal data using APIs provided by government agencies and collects precedent data from court websites using web scraping techniques. Specifically, it obtains legal data in JSON format via the "e-Stat" API and uses the "BeautifulSoup" library to analyze precedent data from HTML pages and extract necessary information. Input is legal analysis information from online resources, and output is the acquired legal data and precedent data.

[1805] Step 2: Data Analysis

[1806] The server analyzes the collected data using natural language processing techniques. This analysis utilizes Python's "NLTK" and "SpaCy." The server processes the collected data through steps such as word tokenization, text mining, sentiment analysis, and topic modeling. The input is the collected raw data, and the output consists of analyzed key information and its results.

[1807] Step 3: Storing in the database

[1808] The server stores the parsed data in a database. Specifically, it uses a MySQL database, storing legal information in the "Legal Regulations" table and case law information in the "Case Law" table. It also creates indexes to improve search efficiency. The input is the parsed data, and the output is the data stored in the database and the indexes.

[1809] Step 4: User Login

[1810] A user logs into the system using a terminal. The user enters their ID and password, and this information is sent from the terminal to the server via HTTPS. The server compares this information with the user's information in the database and performs authentication. The input is the user's ID and password, and the output is the authentication result (success or failure) response.

[1811] Step 5: Enter information about the person being investigated.

[1812] The user enters basic information about the person being interrogated (name, age, address, case summary, etc.) into a terminal, and this information is sent from the terminal to the server in JSON format. The server stores the received information in a database. The input is the basic information of the person being interrogated, and the output is the information stored in the database.

[1813] Step 6: Generate the question set

[1814] The server retrieves relevant information from the database and generates a set of questions using the Python "GPT-3" API. Specifically, in the case of a traffic accident, it forms questions based on the category "traffic laws" and sends the constructed questions to the terminal. The input is relevant information from the database, and the output is the generated set of questions.

[1815] Step 7: Acquiring emotional data

[1816] The device acquires user emotion data during questioning. The "Google Cloud Speech-to-Text" API is used for speech analysis, and "OpenCV" is used for facial expression analysis. The emotion data acquired by the device is sent to the server in real time. The input is the user's emotion data, and the output is the analyzed data sent to the server.

[1817] Step 8: Analyzing emotional data

[1818] The server analyzes the emotional data it receives using TensorFlow. It then runs an algorithm that evaluates the stress level and truthfulness of the interrogated person based on changes in voice tone and facial expressions. The input is emotional data, and the output is the evaluation result of the stress level and truthfulness.

[1819] Step 9: Applying emotional data

[1820] The server adjusts the progress of the interrogation based on the analysis results. If a high stress level is detected, the server lowers the difficulty of the generated question set or generates a message suggesting a break and sends it to the terminal. The input is the evaluation result, and the output is the adjusted question set or the break suggestion message.

[1821] Step 10: Start of Interrogation

[1822] The user uses a device to generate questions and ask them to the person being interrogated. The person being interrogated's responses are recorded using a microphone, and the device sends this data to a server. The input is the person being interrogated's responses, and the output is the audio data sent to the server.

[1823] Step 11: Submitting and analyzing response data

[1824] The terminal sends the recorded audio data to the server. The server uses "NLTK" or "SpaCy" to convert the audio data into text and analyzes inconsistencies and inconsistencies by comparing it with legal and case law data. The input is the audio data, and the output is the converted text data and its analysis results.

[1825] Step 12: Generate the report

[1826] The server generates a report based on the final analysis results. The report includes a summary of all the interviewee's responses, inconsistencies, and a stress assessment based on sentiment analysis. The generated report is sent from the server to the terminal in PDF format and presented to the user. The input is the analysis results, and the output is a report in PDF format.

[1827] (Application Example 2)

[1828] Next, we will explain application example 2. In the following explanation, the data processing device 1...

Claims

1. Means of obtaining data related to laws and precedents from online resources, A means of analyzing acquired data using natural language processing technology and storing it in a database, A means for retrieving relevant information from a database and generating a set of questions based on information about the person being interrogated entered by the user, A means of receiving the suspect's response data and comparing it with legal and case law data to detect inconsistencies and inconsistencies, A system that includes means for generating reports based on analysis results and presenting them to the user.

2. The system according to claim 1, which records and analyzes the suspect's responses via voice input.

3. The system according to claim 1, wherein a user creates additional questions, and the system updates the content of the interrogation based on those questions.

Images