system

Abstract

We provide the system. [Solution] Means for acquiring audio data, A means of converting acquired audio data into text data in real time, A method for analyzing and extracting important agenda items from text data, A means of displaying the extracted important agenda items as a summary along with their supporting evidence, A system that includes means for providing users with summaries and corresponding audio data.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] [[ID=3!5]]In recent years, with the spread of teleconferences, recording of conversation contents has become more important. However, in the conventional method, the work of manually transcribing the conversation contents, extracting the key points therefrom, and creating minutes requires a great deal of time and labor. In addition, since it is performed manually, there may be a lack of accuracy and consistency. Furthermore, in many cases, means for confirming the voice data corresponding to the recorded text data are not well-prepared, and it is difficult to grasp the flow and nuances of the conversation. There is a need for a system that solves such problems and efficiently and accurately records, summarizes, and creates minutes of teleconference contents.

Means for Solving the Problems

[0005] The present invention solves the above problems by the following means: providing a system that includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing and extracting important agenda items from the text data, means for displaying the extracted important agenda items as a summary along with their supporting evidence, and means for providing the user with the summary and corresponding audio data. This enables not only the automatic and accurate recording of teleconference content and the efficient extraction and display of key points, but also allows the user to simultaneously review text and audio as needed. Furthermore, by using natural language processing technology to analyze and extract text data, the accuracy and consistency of the meeting minutes can be improved. In addition, by adding means for playing audio data related to the summary section specified by the user, it becomes easier to grasp the flow and nuances of the conversation.

[0006] "Audio data" refers to digital data of audio signals, including conversations and speeches during teleconferences.

[0007] "Real-time" refers to processing or responding to events almost simultaneously with their occurrence.

[0008] "Text data" refers to data in text format converted from audio data.

[0009] "Analysis" is the process of examining data and clarifying its structure and meaning.

[0010] "Important agenda items" refer to statements or topics that deserve particular attention during a teleconference.

[0011] "Extraction" is the process of taking a specific part from the whole.

[0012] "Evidence" refers to specific statements or pieces of evidence that support a particular conclusion or judgment.

[0013] A "summary" refers to a short, concise version of detailed information.

[0014] "Presentation" means visually providing text and graphics.

[0015] "Provision" refers to the distribution of information and services to users.

[0016] "Natural language processing technology" is a technology for a computer to understand, interpret, and generate human language.

Brief Description of Drawings

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

Mode for Carrying Out the Invention

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

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

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

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

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

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

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

[0025] [First Embodiment]

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

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

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

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

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

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

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

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

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

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

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

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

[0038] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. This invention acquires conversation content as audio data in real time, transcribes it, analyzes the text data to extract important agenda items, and generates a summary. Furthermore, it displays the supporting statements for each key point in an easy-to-understand manner, and allows users to simultaneously review the text and corresponding audio data. Specific embodiments of this system are described below.

[0039] Acquisition of audio data

[0040] The server activates the audio capture module as soon as the teleconference begins, and acquires audio data. The audio data is recorded in real time in an appropriate format (e.g., WAV or MP3). This recorded audio data is then used in subsequent processes.

[0041] Transcript

[0042] The server converts the acquired audio data into text data in real time via a speech recognition API (e.g., a general speech recognition service). This transcribed data is stored on the server and used later for summarization and analysis.

[0043] Extraction of important agenda items

[0044] The server ingests text data, analyzes the text using natural language processing (NLP) techniques, and extracts key agenda items. This involves techniques such as keyword extraction and text classification. The extracted agenda items are then organized into summaries.

[0045] Summary display and presentation of evidence

[0046] The server displays the extracted agenda items as summaries. It also displays the specific statements (evidence) corresponding to each summary. This allows users to easily understand the background and details of the key points.

[0047] Providing a user interface

[0048] The device provides a user-friendly interface. For example, a web portal displays the full text along with a summary, and when a user clicks on a specific meeting item, it can play audio data related to that point. It also generates PDF meeting minutes in a user-friendly format.

[0049] Specific example

[0050] 1. The user starts a teleconference.

[0051] "Teleconference started. Audio is being captured."

[0052] 2. The server captures the audio and converts it into text data in real time.

[0053] "We are converting audio data to text in real time."

[0054] 3. The server analyzes the text data using natural language processing technology and extracts important agenda items.

[0055] "Analysis in progress. Extracting key agenda items."

[0056] 4. The device provides a user interface, allowing the user to review the summary.

[0057] "You can review the key points. Related comments will also be displayed."

[0058] 5. The user reviews the summary and selects and plays audio data related to specific points.

[0059] "Let's review the statement in point 2."

[0060] "Playing the statement." (Audio data plays)

[0061] This system can improve work efficiency by efficiently and accurately recording the content of teleconferences and allowing users to review key points and their supporting evidence as needed.

[0062] The following describes the processing flow.

[0063] Step 1:

[0064] The user initiates a teleconference. Once the teleconference starts, audio data capture begins automatically.

[0065] Step 2:

[0066] The server activates the audio capture module and collects audio data from the teleconference in real time. The audio data is saved in an appropriate format.

[0067] Step 3:

[0068] The server sends the acquired audio data in real time to a speech recognition API (for example, a general speech recognition service), and converts the audio data into text data.

[0069] Step 4:

[0070] The server sequentially saves the text data received from the speech recognition API.

[0071] Step 5:

[0072] The server takes in text data and performs analysis using natural language processing (NLP) techniques. This extracts key agenda items from the text.

[0073] Step 6:

[0074] The server extracts the agenda items, organizes them into a summary, and identifies and saves the specific statements that support them.

[0075] Step 7:

[0076] The server prepares to display the generated summary and supporting evidence through the user interface.

[0077] Step 8:

[0078] The device displays an interface that provides the user with a summary and supporting evidence. The user can use this to confirm the key points.

[0079] Step 9:

[0080] The user reviews the displayed summary and supporting evidence. Clicking on a statement related to a specific point plays the corresponding audio data.

[0081] Step 10:

[0082] The device plays audio data related to the summary specified by the user. This allows the user to view the text and audio simultaneously, enabling them to grasp the flow and nuances of the conversation.

[0083] In this way, the system efficiently and accurately records the content of teleconferences through a series of processing steps and effectively provides the key points to the user.

[0084] (Example 1)

[0085] 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."

[0086] One challenge in conducting teleconferences is the difficulty in efficiently and accurately recording meeting content and easily reviewing summaries afterward. Traditional methods require each participant to review the entire meeting, which is time-consuming and laborious. Furthermore, there is a lack of effective systems for quickly extracting and displaying important agenda items.

[0087] 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.

[0088] In this invention, the server includes means for detecting the start of a teleconference, means for recording audio data in real time, means for converting the recorded audio data into text data in real time, means for analyzing and extracting important agenda items from the converted text data using natural language processing technology, means for displaying the extracted important agenda items as a summary along with their supporting evidence, means for providing the summary and corresponding audio data to the user, and means for providing a user interface. This makes it possible to efficiently and accurately record the content of a teleconference and to quickly extract and display important agenda items.

[0089] A "teleconference" is a communication method in which multiple participants in remote locations conduct a meeting via audio and video.

[0090] "Means for detecting the start of a meeting" refers to a system or method that automatically recognizes the start of a meeting based on the reservation information or schedule of the teleconference.

[0091] "Means for recording audio data in real time" refers to a system or device that captures the audio of a meeting in real time and saves it in digital format.

[0092] "Methods for converting to text data in real time" refer to technologies that instantly convert acquired audio data into text format, such as speech recognition services and software.

[0093] "Means of analysis and extraction using natural language processing technology" refers to algorithms and programs used to analyze text data and identify important words, phrases, and agenda items.

[0094] "Means of displaying as a summary" refers to methods or interfaces that visually organize the analyzed data and display it on the screen in a way that is easy for the user to understand.

[0095] "Means of providing audio data to users" refers to systems and methods for delivering recorded audio data to users in formats such as streaming or file download.

[0096] "Means of providing a user interface" refers to graphical screens and interaction methods that allow users to operate a system and access necessary information.

[0097] This invention relates to a system for efficiently recording the content of teleconferences, extracting important agenda items, generating summaries, and providing reproducible meeting minutes. The following describes in detail specific embodiments of this invention.

[0098] Acquisition of audio data

[0099] The server has a function to detect the start of a teleconference and obtains the meeting start time from calendar information or the reservation system. When the start time arrives, it automatically activates the audio capture module and records audio data in real time. The audio data is saved in WAV or MP3 format and used in the subsequent processing described below.

[0100] Transcript

[0101] The server sends the recorded audio data to a speech recognition service (e.g., a common speech recognition API, such as Google® Cloud Speech-to-Text or IBM Watson® Speech to Text) and converts it into text data in real time. This text data is stored on the server and used for later analysis and summary generation.

[0102] Extraction of important agenda items

[0103] The server retrieves stored text data and analyzes the text using natural language processing (NLP) techniques. This analysis employs keyword extraction and text classification technologies (e.g., NLTK, SpaCy). As a result of the analysis, important meeting items are extracted and organized into a summary.

[0104] Summary and presentation of evidence

[0105] The server displays key meeting minutes organized as summaries. Specific statements supporting each summary are also displayed. This allows users to easily understand the background and details of the key points.

[0106] Providing a user interface

[0107] The device provides a user-friendly interface. This interface is implemented as a web portal and mobile application, allowing for the display of summaries and full text, and playback of related audio data by clicking on the summary. It also generates PDF meeting minutes in a user-friendly format.

[0108] Specific example

[0109] 1. The user starts a teleconference.

[0110] "Teleconference started. Audio is being captured."

[0111] 2. The server captures the audio and converts it into text data in real time.

[0112] "We are converting audio data to text in real time."

[0113] 3. The server analyzes the text data using natural language processing technology and extracts important agenda items.

[0114] "Analysis in progress. Extracting key agenda items."

[0115] 4. The device provides a user interface, allowing the user to review the summary.

[0116] "You can review the key points. Related comments will also be displayed."

[0117] 5. The user reviews the summary and selects and plays audio data related to specific points.

[0118] "Let's review the statement in point 2."

[0119] "Playing the statement." (Audio data plays)

[0120] Examples of prompts for generative AI models

[0121] Example of a prompt:

[0122] Please transcribe the following teleconference audio and extract the key agenda items. Then, display the specific statements related to each agenda item and link to the audio data.

[0123] 1. The topic of the teleconference is "Project progress report."

[0124] 2. Record all statements made during the meeting and create a summary.

[0125] 3. Please also make it possible to play the audio data of the statements corresponding to each summary.

[0126] This system efficiently and accurately records the content of teleconferences and enables the rapid extraction and display of important agenda items. This leads to improved work efficiency and accurate information sharing.

[0127] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0128] Step 1:

[0129] The server detects the start of the teleconference.

[0130] Input: Schedule information and reservation system data.

[0131] Specific operation: The server refers to pre-configured calendar information to detect the teleconference start time. Based on the detected start time, it prepares to start the audio capture module.

[0132] Output: Preparing to start the audio capture module.

[0133] Step 2:

[0134] The server activates the audio capture module and records audio data in real time.

[0135] Input: Audio signal from a teleconference.

[0136] Specific operation: The audio capture module acquires audio directly from a microphone or audio interface and saves it in real time as digital data (WAV or MP3).

[0137] Output: Audio data file recorded in real time.

[0138] Step 3:

[0139] The server sends the recorded audio data to a speech recognition service, where it is converted into text data.

[0140] Input: Recorded audio data.

[0141] Specific operation: The server sends the acquired audio file to a speech recognition API (e.g., Google Cloud Speech-to-Text) and receives the resulting text data as a response.

[0142] Output: Real-time generated character data.

[0143] Step 4:

[0144] The server saves the text data to the database.

[0145] Input: Generated character data.

[0146] Specific operation: The server stores the generated character data in a database (e.g., MySQL®, PostgreSQL) as a timestamped record.

[0147] Output: Character data stored in the database.

[0148] Step 5:

[0149] The server analyzes the stored text data and extracts important agenda items.

[0150] Input: Character data stored in the database.

[0151] Specific operation: The server uses natural language processing (NLP) techniques to perform keyword extraction, text classification, and sentiment analysis. This identifies important meeting agenda items and organizes them into a summary.

[0152] Output: A list and summary of the key meeting agenda items extracted.

[0153] Step 6:

[0154] The server formats the data for displaying the summary and supporting evidence, and then sends it to the user interface.

[0155] Input: Key meeting agenda items and supporting statements extracted.

[0156] Specific actions: Format the data to display a summary and its supporting evidence, and prepare it for transmission to the endpoint.

[0157] Output: Formatted data.

[0158] Step 7:

[0159] The device provides a user-friendly interface.

[0160] Input: Formatted data.

[0161] Specific operation: Provide an interface for web portals and mobile applications that displays summaries, supporting evidence, and the full text. Implement a mechanism where, when a user clicks on a specific summary, the associated audio data is played.

[0162] Output: User interface.

[0163] Step 8:

[0164] The user interacts with the provided interface to obtain the necessary information.

[0165] Input: User interface screen.

[0166] Specific operation: Users view summaries through the interface and play relevant audio data by clicking on specific points. Furthermore, they can download meeting minutes in PDF format as needed.

[0167] Output: Playback of summary and related audio data, download of PDF meeting minutes.

[0168] (Application Example 1)

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

[0170] Traditional teleconference recording methods required participants to take notes themselves, which could lead to information overload, incompleteness, or misinterpretation. Furthermore, creating meeting minutes was time-consuming and labor-intensive, making it difficult even to grasp the key points. This was particularly problematic in factory meetings and conferences, negatively impacting work efficiency. A system capable of solving these problems was needed.

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

[0172] In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing and extracting important agenda items from the text data, means for displaying the extracted important agenda items as a summary along with their rationale, means for providing the user with the summary and corresponding audio data, means for recording the contents of meetings and conferences within the factory and automatically generating summaries, means for generating a meeting summary based on the extracted important agenda items, and means for displaying the summary on a user interface. This makes it possible to efficiently record the contents of meetings and conferences and automatically generate summaries, thereby significantly reducing the time and effort required for users to review meeting minutes.

[0173] "Audio data" refers to information recorded in digital format.

[0174] "Means of acquisition" refers to methods or devices for collecting specific data or information.

[0175] "Methods for converting to text data in real time" refer to methods or devices that instantly convert audio into text at the same time as it is heard.

[0176] "Important agenda items" are those that are particularly important topics or subjects of discussion in a meeting or discussion.

[0177] "Means of analysis and extraction" refers to methods or devices for analyzing data and extracting specific information from it.

[0178] A "summary" is a concise compilation of the most important parts of the original information.

[0179] A "user interface" refers to the screens or means by which a user operates a system or software.

[0180] "Meetings and discussions within the factory" refers to gatherings of managers and employees within the factory to discuss matters related to operations and production.

[0181] "Means for recording and automatically generating summaries" refers to methods or devices that save the content of meetings and discussions and automatically extract and summarize the key points based on that data.

[0182] "Extracted important agenda items" are the particularly important topics or items that have been extracted from the analyzed data.

[0183] "Means of providing information to the user" refers to methods or devices for presenting specific information to the user.

[0184] This invention relates to a system for efficiently recording the content of meetings and conferences within a factory and automatically generating summaries. This system performs a series of processes including audio data acquisition, transcription, text analysis, summary generation, and provision of a user interface. Specific embodiments are described below.

[0185] Acquisition of audio data

[0186] The server activates the audio capture module as soon as a meeting or discussion begins, and acquires audio data. The audio data is recorded in real time in an appropriate format (e.g., WAV or MP3). This recorded audio data is used in subsequent processes. Specifically, when a meeting starts, the server captures audio through the microphone and saves it as digital data.

[0187] Transcript

[0188] The server converts the acquired audio data into text data in real time using a speech recognition API. For example, the Python `speech_recognition` module can be used here. This transcribed data is stored on the server and used in the next analysis process.

[0189] Text analysis and extraction of key agenda items

[0190] The server analyzes the text data using natural language processing techniques to extract important agenda items. Specific techniques used for this process include natural language processing libraries such as spaCy and nltk. This automatically extracts particularly important discussions and decisions from the text data.

[0191] Summary generation and display

[0192] The server generates a summary of the extracted key agenda items. This summary should include specific statements (evidence) for each agenda item. The server displays the generated summary in the user interface (UI). The UI displays the full text along with the summary, and also includes a function that plays related audio data when the user clicks on a specific agenda item.

[0193] Providing a user interface

[0194] Terminals (e.g., robots and PCs in a factory) provide user-friendly interfaces. The web portal displays the full text along with a summary. Users can click on specific meeting items to play audio data related to those points. A PDF meeting transcript in an easy-to-read format is also generated.

[0195] Examples of specific cases and prompt statements

[0196] For example, suppose the following statement was made at a production meeting in a factory:

[0197] "We need to review next week's production schedule. Particular attention should be paid to the product quality on production line A."

[0198] A summary is generated based on this statement.

[0199] Examples of prompts to input into a generative AI model:

[0200] "Please analyze the following text and extract the key agenda items: We need to review next week's production schedule. Particular attention should be paid to the product quality of production line A."

[0201] As a result, a system can be realized that efficiently records the content of meetings and conferences within the factory and automatically generates summaries. By using this system, the time and effort required to create meeting minutes can be significantly reduced. In addition, important meeting items can be easily identified, improving meeting efficiency and accelerating business decision-making.

[0202] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0203] Step 1:

[0204] The server activates the audio capture module as soon as a meeting or discussion begins, and acquires audio data. This audio data is recorded in real time using a microphone in a digital format (e.g., WAV or MP3). The input is an audio signal, and the output is audio data in digital format.

[0205] Step 2:

[0206] The server converts acquired audio data into text data in real time. Specifically, it uses a speech recognition API (e.g., Python's speech_recognition module) to generate text from speech. The input is digital audio data, and the output is text data.

[0207] Step 3:

[0208] The server analyzes the generated text data using natural language processing (NLP) techniques to extract important agenda items. Specifically, it uses natural language processing libraries such as spaCy and nltk to identify specific keywords and phrases from the text. The input is text data, and the output is a list of important agenda items.

[0209] Step 4:

[0210] The server generates a summary of the extracted key agenda items. This summary also includes specific statements (evidence) for each agenda item. The input is a list of key agenda items, and the output is the summary text.

[0211] Step 5:

[0212] The terminal displays a summary provided by the server in its user interface. Specifically, it displays the summary and full text through a web portal or dedicated application, and when the user clicks on a specific meeting item, the related audio data is played. The input is the summary text and full text, and the output is the display in the user interface.

[0213] Step 6:

[0214] The user reviews the summary through the provided interface and plays audio data related to specific points. When the user clicks on a specific agenda item, the terminal plays the corresponding audio data. The input is the user's click operation, and the output is the playback of audio data.

[0215] 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.

[0216] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. In particular, by combining it with an emotion engine that recognizes user emotions, it becomes possible to create richer meeting minutes. This system has a series of functions including real-time acquisition of audio data, transcription, extraction of important agenda items, analysis of emotion data, summary generation, and playback of audio data. The following describes specific embodiments of this system.

[0217] Acquisition of audio data

[0218] The server activates the audio capture module as soon as the teleconference begins, collecting audio data in real time. The audio data is recorded in an appropriate format (e.g., WAV or MP3). This recorded audio data is later used for transcription and sentiment analysis.

[0219] Transcript

[0220] The server converts the acquired audio data into text data in real time via a speech recognition API. This transcribed data is stored on the server and used later for summarization and sentiment analysis.

[0221] Analysis of emotional data

[0222] The server uses an emotion engine to analyze the user's emotions in real time from voice and text data. This includes the ability to analyze voice tone and text content. For example, the user's emotions may be visualized as "tension" or "excitement."

[0223] Extraction of important agenda items

[0224] The server analyzes the text using natural language processing (NLP) techniques based on the analyzed character and sentiment data, and extracts key agenda items. This includes keyword extraction and text classification that take sentiment data into account.

[0225] Summary generation and display

[0226] The server integrates the extracted meeting minutes with sentiment data to generate a summary, and also identifies and saves the specific statements that support it. For example, parts of the sentiment data that show high levels of excitement are given particular attention.

[0227] Providing a user interface

[0228] The device provides a user-friendly interface. The web portal displays full text and sentiment data along with a summary, and users can click on specific meeting items to play audio data related to those points. It also generates PDF meeting minutes in a user-friendly format.

[0229] Specific example

[0230] 1. The user starts a teleconference.

[0231] "Teleconference started. Audio is being captured."

[0232] 2. The server captures the audio and converts it into text data in real time.

[0233] "We are converting audio data to text in real time."

[0234] 3. The server uses an emotion engine to analyze the user's emotions.

[0235] "Analyzing emotion data. Identifying user emotions."

[0236] 4. The server analyzes text data and sentiment data using natural language processing technology to extract important agenda items.

[0237] "Analysis in progress. Extracting key agenda items."

[0238] 5. The device provides a user interface, allowing users to view summaries and sentiment data.

[0239] "You can view key points and sentiment data. Related comments are also displayed."

[0240] 6. When a user reviews summaries and sentiment data and clicks on a statement related to a specific point, audio data related to that point is played.

[0241] "Let's review the statement in point 2."

[0242] "Playing the statement." (Audio data plays)

[0243] This system efficiently and accurately records the content of teleconferences and effectively provides users with key points, thereby improving work efficiency. Furthermore, by utilizing sentiment data, it makes it easier to understand the atmosphere and urgency of the meeting.

[0244] The following describes the processing flow.

[0245] Step 1:

[0246] The user initiates a teleconference. Once the teleconference starts, audio data capture begins automatically.

[0247] User: "Starting teleconference."

[0248] Step 2:

[0249] The server activates the audio capture module and collects audio data from the teleconference in real time. The audio data is saved in an appropriate format.

[0250] Server: "We are collecting audio data in real time."

[0251] Step 3:

[0252] The server collects audio data and sends it to a speech recognition API in real time, where it converts the audio into text data. This text data is then stored sequentially on the server.

[0253] Server: "Converting audio data to text data."

[0254] Step 4:

[0255] The server uses an emotion engine to analyze collected audio and text data in real time to recognize the user's emotions. The emotion engine analyzes the tone of voice and the content of the text.

[0256] Server: "Analyzing emotion data."

[0257] Step 5:

[0258] The server uses natural language processing (NLP) techniques to extract key agenda items based on the analyzed text and sentiment data.

[0259] Server: "We are analyzing text and sentiment data to extract key agenda items."

[0260] Step 6:

[0261] The server extracts the agenda items, organizes them into a summary, and saves them along with the corresponding specific statements. Sentiment data is also reflected in the summary.

[0262] Server: "We are generating summaries and tracking specific statements and sentiment data."

[0263] Step 7:

[0264] The device displays the generated summary and sentiment data through the user interface, allowing users to easily review key points.

[0265] Terminal: "Preparing the interface for summary and sentiment data."

[0266] Step 8:

[0267] Users can review the displayed summary and sentiment data. Selecting a specific agenda item will display comments and sentiment data related to that point.

[0268] User: "I'd like to confirm the specific points."

[0269] Terminal: "Displays selected points, related statements, and sentiment data."

[0270] Step 9:

[0271] When a user plays audio data related to a specific point, the relevant portion of the audio data will be played.

[0272] User: "Playing audio data related to the main points."

[0273] Terminal: "Related audio data will be played." (Audio data playback)

[0274] In this way, the system efficiently and accurately records the content of teleconferences through a series of processing steps, effectively providing users with key points and sentiment data. Users can simultaneously review text and audio, grasping the flow, nuances, and emotional changes of the conversation.

[0275] (Example 2)

[0276] 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".

[0277] Traditional teleconferencing systems struggled to efficiently record meeting content and generate key agenda items and summaries. Furthermore, they lacked sufficient means to analyze user sentiment data and understand the meeting atmosphere, making minute-taking cumbersome. Additionally, quickly accessing important statements during meetings was difficult.

[0278] The specific processing by the specific processing unit 290 of the data processing device 12 in Embodiment 2 is realized by the following means. In this invention, the server includes means for acquiring voice data, means for converting the acquired voice data into character data in real time, means for analyzing the user's emotion from the voice data and character data, means for extracting important agenda items based on the analyzed character data and emotion data, means for integrating the extracted agenda items and emotion data to generate a summary, and means for providing the generated summary and the corresponding voice data to the user. Thereby, it becomes possible to efficiently and accurately record the meeting content, extract important agenda items, and understand the atmosphere of the entire meeting.

[0279] "Voice data" refers to data obtained by digitizing voice signals such as those in a teleconference.

[0280] "Real time" refers to a state where data acquisition and processing are performed immediately without delay.

[0281] "Character data" refers to information in text format generated based on voice data.

[0282] "Emotion data" refers to data indicating the user's emotional state analyzed from voice or text.

[0283] "Agenda item" refers to the main topics or themes discussed in a teleconference.

[0284] "Natural language processing technology" is a technology for a computer to understand and analyze human language.

[0285] "Summary" is a text that concisely summarizes specific contents and key points.

[0286] "Server" is a computer system for processing data and providing various services to users.

[0287] A "user interface" is the environment that includes screens and input methods for a user to interact with a system.

[0288] A "voice capture module" is a device and software for collecting voice data and storing it as digital data.

[0289] A "database" is a system for systematically storing, searching, and editing information.

[0290] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. In particular, by combining it with an emotion engine that recognizes user emotions, it becomes possible to create richer meeting minutes. This system has a series of functions including real-time acquisition of audio data, transcription, extraction of important agenda items, analysis of emotion data, summary generation, and playback of audio data.

[0291] Hardware and software used

[0292] 1. Server

[0293] Voice capture module: Collects voice data in real time.

[0294] Speech Recognition API: Using "Google Speech-to-Text" as an example, we will convert speech data into text data in real time.

[0295] Emotion Engine: Using "Affectiva" as an example, we analyze the user's emotions from audio and text data.

[0296] Natural language processing technology: As an example, we will use "spaCy" to extract important agenda items based on text data and sentiment data.

[0297] Database: Stores analysis results and generated summaries.

[0298] 2. Terminal

[0299] User Interface: Displays summaries, sentiment data, and other information to the user using a web browser or dedicated application.

[0300] PDF generation tool: Used to output meeting minutes in PDF format.

[0301] Explanation of the program's processing flow

[0302] The server activates the audio capture module as soon as the teleconference begins, collecting audio data in real time. The collected audio data is recorded in WAV or MP3 format and used later for transcription and sentiment analysis.

[0303] Next, the server converts the audio data into text data in real time using a speech recognition API (Google Speech-to-Text). This text data is stored in a database. The text data generated through transcription is then used for subsequent summarization and sentiment analysis.

[0304] The server uses an emotion engine (Affectiva) to analyze the user's emotions in real time from voice and text data. Emotion analysis includes analyzing the tone of voice and the content of the text. For example, emotional data such as "tension" or "excitement" may be identified.

[0305] Furthermore, the server uses natural language processing technology (spaCy) to analyze text and sentiment data. It extracts important agenda items and performs keyword extraction and text classification that takes sentiment data into account. The extracted agenda items are used to generate summaries.

[0306] Subsequently, the server integrates the extracted meeting minutes and sentiment data to generate a summary. The generated summary also identifies and stores the specific statements that support it. In particular, sections where the sentiment data indicates high levels of excitement are given special attention in the summary.

[0307] Finally, the terminal provides a user interface that enables the user to view the summary, full text, and sentiment data. Clicking on a specific agenda item plays the audio data related to the key points. Additionally, PDF minutes in a reader-friendly format are also generated through the user interface.

[0308] Specific Example

[0309] 1. The user starts a teleconference.

[0310] Example: The user clicks the "Start Teleconference" button.

[0311] Result: The teleconference starts, and the server begins audio capture.

[0312] 2. The server captures the audio and converts it into text data in real time.

[0313] Example: The server batch processes the audio data every second and converts it into text data.

[0314] Result: Text data is generated as the teleconference progresses.

[0315] 3. The server analyzes the user's sentiment using a sentiment engine.

[0316] Example: The server saves analysis results such as "the user's voice tone has risen" during analysis.

[0317] Result: Sentiment data is generated, and timestamps corresponding to specific sentiments are recorded.

[0318] 4. The server analyzes the text data and sentiment data using natural language processing techniques to extract important agenda items.

[0319] Example: Keyphrase extraction and TextRank algorithms are used to extract important agenda items.

[0320] Result: Important meeting items are listed and saved in the database.

[0321] 5. The device provides a user interface, allowing users to view summaries and sentiment data.

[0322] Example: When a user accesses a web portal, they can click the "View Summary" button.

[0323] Results: A summary, key agenda items, sentiment data, and relevant comments are displayed.

[0324] Example of a prompt

[0325] 1. Record the start of the teleconference in real time.

[0326] 2. Convert the audio data to text and save it.

[0327] 3. Perform sentiment analysis from the text and audio data.

[0328] 4. Analyze the sentiment data and text data to extract the important agenda items.

[0329] 5. Generate a summary and display it in the user interface.

[0330] This invention makes it possible to efficiently record the content of teleconferences, extract important agenda items, and generate summaries. Furthermore, by analyzing sentiment data, it becomes easier to understand the atmosphere of the meeting and the emotions of the participants.

[0331] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0332] Step 1:

[0333] The server activates the audio capture module when the teleconference begins. The audio capture module collects audio data in real time and records it in WAV or MP3 format. The input is the audio signal from the teleconference, which is captured as digital audio data. The output is the recorded audio data. Specifically, the audio capture module is activated, the audio data is sent to the server in real time, and it is saved in the appropriate format.

[0334] Step 2:

[0335] The server converts acquired audio data into text data in real time. It analyzes the audio data using a speech recognition API (e.g., Google Speech-to-Text). The input is the audio data recorded in step 1, and the data processing involves converting the audio data into text data. The output is the converted text data. Specifically, the audio data is sent to the speech recognition API, and the text data returned by the API is saved to the server.

[0336] Step 3:

[0337] The server uses an emotion engine (e.g., Affectiva) to analyze the user's emotions in real time from audio and text data. The input consists of audio and text data, and the data processing involves speech tone analysis and text analysis. The output is the analyzed emotion data. Specifically, the emotion engine analyzes the tone of the audio and the content of the text, and creates a file that identifies emotions such as "tension" or "excitement."

[0338] Step 4:

[0339] The server extracts important agenda items based on text data and sentiment data using natural language processing techniques (e.g., spaCy). The input consists of text data and sentiment data, and data processing includes keyword extraction and text classification. The output is the extracted important agenda items. Specifically, it uses natural language processing techniques to analyze text data and list agenda items while considering sentiment data.

[0340] Step 5:

[0341] The server integrates extracted meeting minutes and sentiment data to generate a summary. The input consists of important meeting minutes and sentiment data, and the data processing involves calculations that integrate important statements and sentiment highlights. The output is the generated summary. Specifically, the extracted meeting minutes and sentiment data are saved to a database, and an algorithm runs to generate the summary.

[0342] Step 6:

[0343] The device provides a user interface, allowing users to view summaries, full text, and sentiment data. The input is the summary and corresponding audio data generated in step 5, and the output is a visually displayed summary and playable audio data. Specifically, when a user accesses the web portal and clicks the "View Summary" button, the summary, key minutes, sentiment data, and relevant comments are displayed.

[0344] Step 7:

[0345] When a user reviews summaries and sentiment data and clicks on a statement related to a specific point, the audio data related to that point is played. The input is the summary section and associated audio data specified by the user, and the output is the played audio data. Specifically, when a user clicks on a summary section in the web portal, the audio data corresponding to that statement is played, allowing them to listen to the details of the meeting content.

[0346] (Application Example 2)

[0347] 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".

[0348] Meetings and work instructions in factories require efficient and accurate recording. Traditional methods involved the cumbersome process of manually converting audio data into text and generating summaries. Furthermore, meeting minutes were created without incorporating emotional data from the meeting, making it difficult to issue work instructions that considered the atmosphere and urgency of the meeting. Therefore, a system was needed that provided a seamless process for real-time audio data acquisition, conversion to text, emotional data analysis, and automated work instruction generation.

[0349] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0350] In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing and extracting important agenda items from the text data, means for displaying the extracted important agenda items as a summary along with their rationale, means for providing the user with the summary and corresponding audio data and sentiment data, means for allowing the user to confirm important agenda items and summaries through a user interface, means for analyzing sentiment data to identify the user's emotions, and means for automatically generating work instructions based on the summary. This makes it possible to efficiently and accurately record the contents of meetings and automatically generate work instructions using sentiment data.

[0351] "Audio data" refers to sound information acquired from microphones and other audio input devices.

[0352] "Text data" refers to digital information obtained by converting audio data into text format.

[0353] "Important agenda items" refer to points or decisions that deserve particular attention during a meeting or discussion.

[0354] "Emotional data" refers to information that indicates the speaker's emotional state, and is analyzed from the tone of voice and the content of the text.

[0355] A "summary" is a text that concisely summarizes detailed information.

[0356] A "user interface" refers to the screens and means of operation that a user uses to interact with a system.

[0357] A "server" is a computer system that provides services over a network.

[0358] "Natural language processing technology" refers to the technology that enables computers to understand and generate human language.

[0359] A "work instruction sheet" is a document that contains instructions for performing a specific task.

[0360] This invention relates to a system for efficiently recording meetings and work instructions conducted within a factory, analyzing emotional data, and generating audio summaries and work instructions. Specifically, it performs real-time acquisition of audio data, conversion of text data, extraction of important agenda items, analysis of emotional data, summary generation, and automatic generation of work instructions.

[0361] The server uses a microphone to acquire audio data and activates an audio capture module to collect audio data in real time. The acquired audio data is recorded in an appropriate format (e.g., WAV or MP3). This recorded audio data is then used for transcription and sentiment analysis.

[0362] The server converts audio data acquired via a speech recognition API into text data in real time. This text data is stored on the server and used later for summarization and sentiment analysis.

[0363] The server uses an emotion engine to analyze the user's emotions in real time from voice and text data. This includes the ability to analyze voice tone and text content. For example, the user's emotions may be identified as "tension" or "excitement."

[0364] The server uses natural language processing (NLP) techniques to analyze the text based on the analyzed character and sentiment data, and extracts key agenda items. This process includes keyword extraction and text classification that take sentiment data into account.

[0365] The server integrates the extracted key meeting minutes with sentiment data to generate a summary, and also identifies and stores the specific statements that support it. For example, parts of the sentiment data showing high levels of excitement are given particular attention.

[0366] The device provides a user-friendly interface. The web portal displays full text and sentiment data along with a summary, and users can click on specific meeting items to play audio data related to those points. It also generates PDF meeting minutes in a user-friendly format.

[0367] Specific example:

[0368] If a user says, "Let's discuss the new production line," during a meeting, the system records the statement, analyzes whether the sentiment is positive, summarizes the key points, and incorporates them into the work instructions.

[0369] Example of a prompt:

[0370] Audio data: "We will discuss the new production line."

[0371] Emotional data: "Positive"

[0372] Summary generated: "Discussions were held regarding a new production line."

[0373] This invention significantly improves operational efficiency within a factory by efficiently and accurately recording meeting content and automatically generating work instructions using emotional data.

[0374] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0375] Step 1:

[0376] The server activates the means to acquire audio data and collects audio data in real time through the microphone. The audio data is recorded in an appropriate format such as WAV or MP3.

[0377] Input: Audio data acquired from the microphone

[0378] Output: Recorded audio file

[0379] Specific operation: The microphone picks up sound, the audio data is sent to the server in real time, and saved in the appropriate file format.

[0380] Step 2:

[0381] The server converts the recorded audio data into text data in real time using a speech recognition API. This converted text data is stored on the server.

[0382] Input: Audio file

[0383] Output: Character data (text format)

[0384] Specific operation: The speech recognition API analyzes the audio file, converts the audio to text, and the generated text data is saved to storage.

[0385] Step 3:

[0386] The server uses an emotion engine to analyze the user's emotions in real time from text and audio data. The analyzed emotion data is stored on the server.

[0387] Input: Text data, audio data

[0388] Output: Sentiment data

[0389] Specific operation: The emotion engine analyzes text data and voice tone, tags the detected emotions as "positive," "negative," "tense," etc., and stores them in a database.

[0390] Step 4:

[0391] The server uses natural language processing techniques to extract key agenda items from text and sentiment data. These extracted agenda items are stored along with the detected sentiment data.

[0392] Input: Text data, sentiment data

[0393] Output: Key Agenda Items

[0394] Specific operation: A natural language processing algorithm analyzes the text, extracts important keywords and phrases, and stores them in storage along with sentiment data.

[0395] Step 5:

[0396] The server generates a summary based on the extracted key meeting items and sentiment data. This summary also identifies and stores the specific statements that supported it.

[0397] Input: Key meeting agenda items, sentiment data

[0398] Output: Summary text

[0399] Specific operation: A natural language generation model takes important meeting items and sentiment data as input, generates a concise summary based on that, and saves specific parts of the statements.

[0400] Step 6:

[0401] The device displays summaries, full text, and sentiment data through a user interface. When the user clicks on a specific meeting item, it plays audio data related to that point. It also generates meeting minutes in PDF format.

[0402] Input: Summary text, full text, sentiment data, audio data

[0403] Output: Displayed summary, full text, sentiment data, and PDF meeting minutes.

[0404] Specific operation: The user interface displays a summary text and related data, and clicking a link plays the audio data. Additionally, a PDF generator produces meeting minutes.

[0405] 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.

[0406] Data generation model 58 is a type of 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.

[0407] 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.

[0408] [Second Embodiment]

[0409] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0410] 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.

[0411] 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).

[0412] 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.

[0413] 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.

[0414] 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).

[0415] 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.

[0416] 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.

[0417] 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.

[0418] 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.

[0419] 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.

[0420] 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".

[0421] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. This invention acquires conversation content as audio data in real time, transcribes it, analyzes the text data to extract important agenda items, and generates a summary. Furthermore, it displays the supporting statements for each key point in an easy-to-understand manner, and allows users to simultaneously review the text and corresponding audio data. Specific embodiments of this system are described below.

[0422] Acquisition of audio data

[0423] The server activates the audio capture module as soon as the teleconference begins, and acquires audio data. The audio data is recorded in real time in an appropriate format (e.g., WAV or MP3). This recorded audio data is then used in subsequent processes.

[0424] Transcript

[0425] The server converts the acquired audio data into text data in real time via a speech recognition API (e.g., a general speech recognition service). This transcribed data is stored on the server and used later for summarization and analysis.

[0426] Extraction of important agenda items

[0427] The server ingests text data, analyzes the text using natural language processing (NLP) techniques, and extracts key agenda items. This involves techniques such as keyword extraction and text classification. The extracted agenda items are then organized into summaries.

[0428] Summary display and presentation of evidence

[0429] The server displays the extracted agenda items as summaries. It also displays the specific statements (evidence) corresponding to each summary. This allows users to easily understand the background and details of the key points.

[0430] Providing a user interface

[0431] The device provides a user-friendly interface. For example, a web portal displays the full text along with a summary, and when a user clicks on a specific meeting item, it can play audio data related to that point. It also generates PDF meeting minutes in a user-friendly format.

[0432] Specific example

[0433] 1. The user starts a teleconference.

[0434] "Teleconference started. Audio is being captured."

[0435] 2. The server captures the audio and converts it into text data in real time.

[0436] "We are converting audio data to text in real time."

[0437] 3. The server analyzes the text data using natural language processing technology and extracts important agenda items.

[0438] "Analysis in progress. Extracting key agenda items."

[0439] 4. The device provides a user interface, allowing the user to review the summary.

[0440] "You can review the key points. Related comments will also be displayed."

[0441] 5. The user reviews the summary and selects and plays audio data related to specific points.

[0442] "Let's review the statement in point 2."

[0443] "Playing the statement." (Audio data plays)

[0444] This system can improve work efficiency by efficiently and accurately recording the content of teleconferences and allowing users to review key points and their supporting evidence as needed.

[0445] The following describes the processing flow.

[0446] Step 1:

[0447] The user initiates a teleconference. Once the teleconference starts, audio data capture begins automatically.

[0448] Step 2:

[0449] The server activates the audio capture module and collects audio data from the teleconference in real time. The audio data is saved in an appropriate format.

[0450] Step 3:

[0451] The server sends the acquired audio data in real time to a speech recognition API (for example, a general speech recognition service), and converts the audio data into text data.

[0452] Step 4:

[0453] The server sequentially saves the text data received from the speech recognition API.

[0454] Step 5:

[0455] The server takes in text data and performs analysis using natural language processing (NLP) techniques. This extracts key agenda items from the text.

[0456] Step 6:

[0457] The server extracts the agenda items, organizes them into a summary, and identifies and saves the specific statements that support them.

[0458] Step 7:

[0459] The server prepares to display the generated summary and supporting evidence through the user interface.

[0460] Step 8:

[0461] The device displays an interface that provides the user with a summary and supporting evidence. The user can use this to confirm the key points.

[0462] Step 9:

[0463] The user reviews the displayed summary and supporting evidence. Clicking on a statement related to a specific point plays the corresponding audio data.

[0464] Step 10:

[0465] The device plays audio data related to the summary specified by the user. This allows the user to view the text and audio simultaneously, enabling them to grasp the flow and nuances of the conversation.

[0466] In this way, the system efficiently and accurately records the content of teleconferences through a series of processing steps and effectively provides the user with the key points.

[0467] (Example 1)

[0468] 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."

[0469] One challenge in conducting teleconferences is the difficulty in efficiently and accurately recording meeting content and easily reviewing summaries afterward. Traditional methods require each participant to review the entire meeting, which is time-consuming and laborious. Furthermore, there is a lack of effective systems for quickly extracting and displaying important agenda items.

[0470] 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.

[0471] In this invention, the server includes means for detecting the start of a teleconference, means for recording audio data in real time, means for converting the recorded audio data into text data in real time, means for analyzing and extracting important agenda items from the converted text data using natural language processing technology, means for displaying the extracted important agenda items as a summary along with their supporting evidence, means for providing the summary and corresponding audio data to the user, and means for providing a user interface. This makes it possible to efficiently and accurately record the content of a teleconference and to quickly extract and display important agenda items.

[0472] A "teleconference" is a communication method in which multiple participants in remote locations conduct a meeting via audio and video.

[0473] "Means for detecting the start of a meeting" refers to a system or method that automatically recognizes the start of a meeting based on the reservation information or schedule of the teleconference.

[0474] "Means for recording audio data in real time" refers to a system or device that captures the audio of a meeting in real time and saves it in digital format.

[0475] "Methods for converting to text data in real time" refer to technologies that instantly convert acquired audio data into text format, such as speech recognition services and software.

[0476] "Means of analysis and extraction using natural language processing technology" refers to algorithms and programs used to analyze text data and identify important words, phrases, and agenda items.

[0477] "Means of displaying as a summary" refers to methods or interfaces that visually organize the analyzed data and display it on the screen in a way that is easy for the user to understand.

[0478] "Means of providing audio data to users" refers to systems and methods for delivering recorded audio data to users in formats such as streaming or file download.

[0479] "Means of providing a user interface" refers to graphical screens and interaction methods that allow users to operate a system and access necessary information.

[0480] This invention relates to a system for efficiently recording the content of teleconferences, extracting important agenda items, generating summaries, and providing reproducible meeting minutes. The following describes in detail specific embodiments of this invention.

[0481] Acquisition of audio data

[0482] The server has a function to detect the start of a teleconference and obtains the meeting start time from calendar information or the reservation system. When the start time arrives, it automatically activates the audio capture module and records audio data in real time. The audio data is saved in WAV or MP3 format and used in the subsequent processing described below.

[0483] Transcript

[0484] The server sends the recorded audio data to a speech recognition service (e.g., common speech recognition APIs, Google Cloud Speech-to-Text, or IBM Watson Speech to Text) and converts it into text data in real time. This text data is stored on the server and used for later analysis and summary generation.

[0485] Extraction of important agenda items

[0486] The server retrieves stored text data and analyzes the text using natural language processing (NLP) techniques. This analysis employs keyword extraction and text classification technologies (e.g., NLTK, SpaCy). As a result of the analysis, important meeting items are extracted and organized into a summary.

[0487] Summary and presentation of evidence

[0488] The server displays key meeting minutes organized as summaries. Specific statements supporting each summary are also displayed. This allows users to easily understand the background and details of the key points.

[0489] Providing a user interface

[0490] The device provides a user-friendly interface. This interface is implemented as a web portal and mobile application, allowing for the display of summaries and full text, and playback of related audio data by clicking on the summary. It also generates PDF meeting minutes in a user-friendly format.

[0491] Specific example

[0492] 1. The user starts a teleconference.

[0493] "Teleconference started. Audio is being captured."

[0494] 2. The server captures the audio and converts it into text data in real time.

[0495] "We are converting audio data to text in real time."

[0496] 3. The server analyzes the text data using natural language processing technology and extracts important agenda items.

[0497] "Analysis in progress. Extracting key agenda items."

[0498] 4. The device provides a user interface, allowing the user to review the summary.

[0499] "You can review the key points. Related comments will also be displayed."

[0500] 5. The user reviews the summary and selects and plays audio data related to specific points.

[0501] "Let's review the statement in point 2."

[0502] "Playing the statement." (Audio data plays)

[0503] Examples of prompts for generative AI models

[0504] Example of a prompt:

[0505] Please transcribe the following teleconference audio and extract the key agenda items. Then, display the specific statements related to each agenda item and link to the audio data.

[0506] 1. The topic of the teleconference is "Project progress report."

[0507] 2. Record all statements made during the meeting and create a summary.

[0508] 3. Please also make it possible to play the audio data of the statements corresponding to each summary.

[0509] This system efficiently and accurately records the content of teleconferences and enables the rapid extraction and display of important agenda items. This leads to improved work efficiency and accurate information sharing.

[0510] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0511] Step 1:

[0512] The server detects the start of the teleconference.

[0513] Input: Schedule information and reservation system data.

[0514] Specific operation: The server refers to pre-configured calendar information to detect the teleconference start time. Based on the detected start time, it prepares to start the audio capture module.

[0515] Output: Preparing to start the audio capture module.

[0516] Step 2:

[0517] The server activates the audio capture module and records audio data in real time.

[0518] Input: Audio signal from a teleconference.

[0519] Specific operation: The audio capture module acquires audio directly from a microphone or audio interface and saves it in real time as digital data (WAV or MP3).

[0520] Output: Audio data file recorded in real time.

[0521] Step 3:

[0522] The server sends the recorded audio data to a speech recognition service, where it is converted into text data.

[0523] Input: Recorded audio data.

[0524] Specific operation: The server sends the acquired audio file to a speech recognition API (e.g., Google Cloud Speech-to-Text) and receives the resulting text data as a response.

[0525] Output: Real-time generated character data.

[0526] Step 4:

[0527] The server saves the text data to the database.

[0528] Input: Generated character data.

[0529] Specific operation: The server stores the generated character data in a database (e.g., MySQL, PostgreSQL) as a timestamped record.

[0530] Output: Character data stored in the database.

[0531] Step 5:

[0532] The server analyzes the stored text data and extracts important agenda items.

[0533] Input: Character data stored in the database.

[0534] Specific operation: The server uses natural language processing (NLP) techniques to perform keyword extraction, text classification, and sentiment analysis. This identifies important meeting agenda items and organizes them into a summary.

[0535] Output: A list and summary of the key meeting agenda items extracted.

[0536] Step 6:

[0537] The server formats the data for displaying the summary and supporting evidence, and then sends it to the user interface.

[0538] Input: Key meeting agenda items and supporting statements extracted.

[0539] Specific actions: Format the data to display a summary and its supporting evidence, and prepare it for transmission to the endpoint.

[0540] Output: Formatted data.

[0541] Step 7:

[0542] The device provides a user-friendly interface.

[0543] Input: Formatted data.

[0544] Specific operation: Provide an interface for web portals and mobile applications that displays summaries, supporting evidence, and the full text. Implement a mechanism where, when a user clicks on a specific summary, the associated audio data is played.

[0545] Output: User interface.

[0546] Step 8:

[0547] The user interacts with the provided interface to obtain the necessary information.

[0548] Input: User interface screen.

[0549] Specific operation: Users view summaries through the interface and play relevant audio data by clicking on specific points. Furthermore, they can download meeting minutes in PDF format as needed.

[0550] Output: Playback of summary and related audio data, download of PDF meeting minutes.

[0551] (Application Example 1)

[0552] 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."

[0553] Traditional teleconference recording methods required participants to take notes themselves, which could lead to information overload, incompleteness, or misinterpretation. Furthermore, creating meeting minutes was time-consuming and labor-intensive, making it difficult even to grasp the key points. This was particularly problematic in factory meetings and conferences, negatively impacting work efficiency. A system capable of solving these problems was needed.

[0554] 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.

[0555] In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing and extracting important agenda items from the text data, means for displaying the extracted important agenda items as a summary along with their rationale, means for providing the user with the summary and corresponding audio data, means for recording the contents of meetings and conferences within the factory and automatically generating summaries, means for generating a meeting summary based on the extracted important agenda items, and means for displaying the summary on a user interface. This makes it possible to efficiently record the contents of meetings and conferences and automatically generate summaries, thereby significantly reducing the time and effort required for users to review meeting minutes.

[0556] "Audio data" refers to information recorded in digital format.

[0557] "Means of acquisition" refers to methods or devices for collecting specific data or information.

[0558] "Methods for converting to text data in real time" refers to methods or devices that instantly convert audio into text at the same time as it is heard.

[0559] "Important agenda items" are those that are particularly important topics or subjects of discussion in a meeting or discussion.

[0560] "Means of analysis and extraction" refers to methods or devices for analyzing data and extracting specific information from it.

[0561] A "summary" is a concise compilation of the most important parts of the original information.

[0562] A "user interface" refers to the screens or means by which a user operates a system or software.

[0563] "Meetings and discussions within the factory" refers to gatherings of managers and employees within the factory to discuss matters related to operations and production.

[0564] "Means for recording and automatically generating summaries" refers to methods or devices that save the content of meetings and discussions and automatically extract and summarize the key points based on that data.

[0565] "Extracted key agenda items" are particularly important topics or items that have been extracted from the analyzed data.

[0566] "Means of providing information to the user" refers to methods or devices for presenting specific information to the user.

[0567] This invention relates to a system for efficiently recording the content of meetings and conferences within a factory and automatically generating summaries. This system performs a series of processes including audio data acquisition, transcription, text analysis, summary generation, and provision of a user interface. Specific embodiments are described below.

[0568] Acquisition of audio data

[0569] The server activates the audio capture module as soon as a meeting or discussion begins, and acquires audio data. The audio data is recorded in real time in an appropriate format (e.g., WAV or MP3). This recorded audio data is used in subsequent processes. Specifically, when a meeting starts, the server captures audio through the microphone and saves it as digital data.

[0570] Transcript

[0571] The server converts the acquired audio data into text data in real time using a speech recognition API. For example, the Python `speech_recognition` module can be used here. This transcribed data is stored on the server and used in the next analysis process.

[0572] Text analysis and extraction of key agenda items

[0573] The server analyzes the text data using natural language processing techniques to extract important agenda items. Specific techniques used for this process include natural language processing libraries such as spaCy and nltk. This automatically extracts particularly important discussions and decisions from the text data.

[0574] Summary generation and display

[0575] The server generates a summary of the extracted key agenda items. This summary should include specific statements (evidence) for each agenda item. The server displays the generated summary in the user interface (UI). The UI displays the full text along with the summary, and also includes a function that plays related audio data when the user clicks on a specific agenda item.

[0576] Providing a user interface

[0577] Terminals (e.g., robots and PCs in a factory) provide user-friendly interfaces. The web portal displays the full text along with a summary. Users can click on specific meeting items to play audio data related to those points. A PDF meeting transcript in an easy-to-read format is also generated.

[0578] Examples of specific cases and prompt statements

[0579] For example, suppose the following statement was made at a production meeting in a factory:

[0580] "We need to review next week's production schedule. Particular attention should be paid to the product quality on production line A."

[0581] A summary is generated based on this statement.

[0582] Examples of prompts to input into a generative AI model:

[0583] "Please analyze the following text and extract the key agenda items: We need to review next week's production schedule. Particular attention should be paid to the product quality of production line A."

[0584] As a result, a system can be realized that efficiently records the content of meetings and conferences within the factory and automatically generates summaries. By using this system, the time and effort required to create meeting minutes can be significantly reduced. In addition, important meeting items can be easily identified, improving meeting efficiency and accelerating business decision-making.

[0585] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0586] Step 1:

[0587] The server activates the audio capture module as soon as a meeting or conference begins, and acquires audio data. This audio data is recorded in real time using a microphone in a digital format (e.g., WAV or MP3). The input is an audio signal, and the output is audio data in digital format.

[0588] Step 2:

[0589] The server converts acquired audio data into text data in real time. Specifically, it uses a speech recognition API (e.g., Python's speech_recognition module) to generate text from speech. The input is digital audio data, and the output is text data.

[0590] Step 3:

[0591] The server analyzes the generated text data using natural language processing (NLP) techniques to extract important agenda items. Specifically, it uses natural language processing libraries such as spaCy and nltk to identify specific keywords and phrases from the text. The input is text data, and the output is a list of important agenda items.

[0592] Step 4:

[0593] The server generates a summary of the extracted key agenda items. This summary includes specific statements (evidence) for each agenda item. The input is a list of key agenda items, and the output is the summary text.

[0594] Step 5:

[0595] The terminal displays a summary provided by the server in its user interface. Specifically, it displays the summary and full text through a web portal or dedicated application, and when the user clicks on a specific meeting item, the related audio data is played. The input is the summary text and full text, and the output is the display in the user interface.

[0596] Step 6:

[0597] The user reviews the summary through the provided interface and plays audio data related to specific points. When the user clicks on a specific agenda item, the terminal plays the corresponding audio data. The input is the user's click operation, and the output is the playback of audio data.

[0598] 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.

[0599] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. In particular, by combining it with an emotion engine that recognizes user emotions, it becomes possible to create richer meeting minutes. This system has a series of functions including real-time acquisition of audio data, transcription, extraction of important agenda items, analysis of emotion data, summary generation, and playback of audio data. The following describes specific embodiments of this system.

[0600] Acquisition of audio data

[0601] The server activates the audio capture module as soon as the teleconference begins, collecting audio data in real time. The audio data is recorded in an appropriate format (e.g., WAV or MP3). This recorded audio data is later used for transcription and sentiment analysis.

[0602] Transcript

[0603] The server converts the acquired audio data into text data in real time via a speech recognition API. This transcribed data is stored on the server and used later for summarization and sentiment analysis.

[0604] Analysis of emotional data

[0605] The server uses an emotion engine to analyze the user's emotions in real time from voice and text data. This includes the ability to analyze voice tone and text content. For example, the user's emotions may be visualized as "tension" or "excitement."

[0606] Extraction of important agenda items

[0607] The server analyzes the text using natural language processing (NLP) techniques based on the analyzed character and sentiment data, extracting key agenda items. This includes keyword extraction and text classification that take sentiment data into account.

[0608] Summary generation and display

[0609] The server integrates the extracted meeting minutes with sentiment data to generate a summary, and also identifies and stores the specific statements that support it. For example, parts of the sentiment data that show high levels of excitement are given particular attention.

[0610] Providing a user interface

[0611] The device provides a user-friendly interface. The web portal displays full text and sentiment data along with a summary, and users can click on specific meeting items to play audio data related to those points. It also generates PDF meeting minutes in a user-friendly format.

[0612] Specific example

[0613] 1. The user starts a teleconference.

[0614] "Teleconference started. Audio is being captured."

[0615] 2. The server captures the audio and converts it into text data in real time.

[0616] "We are converting audio data to text in real time."

[0617] 3. The server uses an emotion engine to analyze the user's emotions.

[0618] "Analyzing emotion data. Identifying user emotions."

[0619] 4. The server analyzes text data and sentiment data using natural language processing technology to extract important agenda items.

[0620] "Analysis in progress. Extracting key agenda items."

[0621] 5. The device provides a user interface, allowing users to view summaries and sentiment data.

[0622] "You can view key points and sentiment data. Related comments are also displayed."

[0623] 6. When a user reviews summaries and sentiment data and clicks on a statement related to a specific point, audio data related to that point is played.

[0624] "Let's review the statement in point 2."

[0625] "Playing the statement." (Audio data plays)

[0626] This system efficiently and accurately records the content of teleconferences and effectively provides users with key points, thereby improving work efficiency. Furthermore, by utilizing sentiment data, it makes it easier to understand the atmosphere and urgency of the meeting.

[0627] The following describes the processing flow.

[0628] Step 1:

[0629] The user initiates a teleconference. Once the teleconference starts, audio data capture begins automatically.

[0630] User: "Starting teleconference."

[0631] Step 2:

[0632] The server activates the audio capture module and collects audio data from the teleconference in real time. The audio data is saved in an appropriate format.

[0633] Server: "We are collecting audio data in real time."

[0634] Step 3:

[0635] The server sends the collected audio data to a speech recognition API in real time, converting the audio into text data. This text data is then stored sequentially on the server.

[0636] Server: "Converting audio data to text data."

[0637] Step 4:

[0638] The server uses an emotion engine to analyze collected audio and text data in real time to recognize the user's emotions. The emotion engine analyzes the tone of voice and the content of the text.

[0639] Server: "Analyzing emotion data."

[0640] Step 5:

[0641] The server uses natural language processing (NLP) techniques to extract key agenda items based on the analyzed text and sentiment data.

[0642] Server: "We are analyzing text and sentiment data to extract key agenda items."

[0643] Step 6:

[0644] The server extracts the agenda items, organizes them into a summary, and saves them along with the corresponding specific statements. Sentiment data is also reflected in the summary.

[0645] Server: "We are generating summaries and tracking specific statements and sentiment data."

[0646] Step 7:

[0647] The device displays the generated summary and sentiment data through the user interface, allowing users to easily review key points.

[0648] Terminal: "Preparing the interface for summary and sentiment data."

[0649] Step 8:

[0650] Users can review the displayed summary and sentiment data. Selecting a specific agenda item will display comments and sentiment data related to that point.

[0651] User: "I'd like to confirm the specific points."

[0652] Terminal: "Displays selected points, related statements, and sentiment data."

[0653] Step 9:

[0654] When a user plays audio data related to a specific point, the relevant portion of the audio data will be played.

[0655] User: "Playing audio data related to the main points."

[0656] Terminal: "Related audio data will be played." (Audio data playback)

[0657] In this way, the system efficiently and accurately records the content of teleconferences through a series of processing steps, effectively providing users with key points and sentiment data. Users can review text and audio simultaneously, grasping the flow, nuances, and emotional changes of the conversation.

[0658] (Example 2)

[0659] 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".

[0660] Traditional teleconferencing systems struggled to efficiently record meeting content and generate key agenda items and summaries. Furthermore, they lacked sufficient means to analyze user sentiment data and understand the meeting atmosphere, making minute-taking cumbersome. Additionally, quickly accessing important statements during meetings was difficult.

[0661] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing the user's emotions from the audio data and text data, means for extracting important agenda items based on the analyzed text data and emotion data, means for integrating the extracted agenda items and emotion data to generate a summary, and means for providing the user with the generated summary and corresponding audio data. This enables efficient and accurate recording of meeting content, extraction of important agenda items, and understanding of the overall atmosphere of the meeting.

[0662] "Audio data" refers to data obtained by digitizing audio signals from teleconferences, etc.

[0663] "Real-time" refers to a state where data acquisition and processing are performed instantly and without delay.

[0664] "Text data" refers to information in text format generated from audio data.

[0665] "Emotional data" refers to data that indicates a user's emotional state, analyzed from voice and text.

[0666] "Meeting items" refer to the main topics or themes discussed during the teleconference.

[0667] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[0668] A "summary" is a text that concisely summarizes the specific content and key points.

[0669] A "server" is a computer system that processes data and provides various services to users.

[0670] A "user interface" is the environment that includes screens and input methods for a user to interact with a system.

[0671] A "voice capture module" is a device and software for collecting voice data and storing it as digital data.

[0672] A "database" is a system for systematically storing, searching, and editing information.

[0673] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. In particular, by combining it with an emotion engine that recognizes user emotions, it becomes possible to create richer meeting minutes. This system has a series of functions including real-time acquisition of audio data, transcription, extraction of important agenda items, analysis of emotion data, summary generation, and playback of audio data.

[0674] Hardware and software used

[0675] 1. Server

[0676] Voice capture module: Collects voice data in real time.

[0677] Speech Recognition API: Using "Google Speech-to-Text" as an example, we will convert speech data into text data in real time.

[0678] Emotion Engine: Using "Affectiva" as an example, we analyze the user's emotions from audio and text data.

[0679] Natural language processing technology: As an example, we will use "spaCy" to extract important agenda items based on text data and sentiment data.

[0680] Database: Stores analysis results and generated summaries.

[0681] 2. Terminal

[0682] User Interface: Displays summaries, sentiment data, and other information to the user using a web browser or dedicated application.

[0683] PDF generation tool: Used to output meeting minutes in PDF format.

[0684] Explanation of the program's processing flow

[0685] The server activates the audio capture module as soon as the teleconference begins, collecting audio data in real time. The collected audio data is recorded in WAV or MP3 format and used later for transcription and sentiment analysis.

[0686] Next, the server converts the audio data into text data in real time using a speech recognition API (Google Speech-to-Text). This text data is stored in a database. The text data generated through transcription is then used for subsequent summarization and sentiment analysis.

[0687] The server uses an emotion engine (Affectiva) to analyze the user's emotions in real time from voice and text data. Emotion analysis includes analyzing the tone of voice and the content of the text. For example, emotional data such as "tension" or "excitement" may be identified.

[0688] Furthermore, the server uses natural language processing technology (spaCy) to analyze text and sentiment data. It extracts important agenda items and performs keyword extraction and text classification that takes sentiment data into account. The extracted agenda items are used to generate summaries.

[0689] Subsequently, the server integrates the extracted meeting minutes and sentiment data to generate a summary. The generated summary also identifies and stores the specific statements that support it. In particular, sections where the sentiment data indicates high levels of excitement are given special attention in the summary.

[0690] Finally, the device provides a user interface, allowing users to view summaries, full text, and sentiment data. Clicking on a specific meeting item plays audio data related to that point. The user interface also generates easy-to-read PDF meeting minutes.

[0691] Specific example

[0692] 1. The user starts a teleconference.

[0693] Example: The user clicks the "Start Teleconference" button.

[0694] Result: The teleconference begins, and the server starts audio capture.

[0695] 2. The server captures the audio and converts it into text data in real time.

[0696] Example: The server batches audio data every second and converts it into text data.

[0697] Result: Text data is generated as the teleconference progresses.

[0698] 3. The server uses an emotion engine to analyze the user's emotions.

[0699] Example: The server saves analysis results such as "the user's voice tone is rising" during the analysis.

[0700] Result: Emotional data is generated, and a timestamp corresponding to a specific emotion is recorded.

[0701] 4. The server analyzes text data and sentiment data using natural language processing technology to extract important agenda items.

[0702] Example: Keyphrase extraction and TextRank algorithms are used to extract important agenda items.

[0703] Result: Important meeting items are listed and saved in the database.

[0704] 5. The device provides a user interface, allowing users to view summaries and sentiment data.

[0705] Example: When a user accesses a web portal, they can click the "View Summary" button.

[0706] Results: A summary, key agenda items, sentiment data, and relevant comments are displayed.

[0707] Example of a prompt

[0708] 1. Record the start of the teleconference in real time.

[0709] 2. Convert the audio data to text and save it.

[0710] 3. Perform sentiment analysis from the text and audio data.

[0711] 4. Analyze the sentiment data and text data to extract the important agenda items.

[0712] 5. Generate a summary and display it in the user interface.

[0713] This invention makes it possible to efficiently record the content of teleconferences, extract important agenda items, and generate summaries. Furthermore, by analyzing sentiment data, it becomes easier to understand the atmosphere of the meeting and the emotions of the participants.

[0714] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0715] Step 1:

[0716] The server activates the audio capture module when the teleconference begins. The audio capture module collects audio data in real time and records it in WAV or MP3 format. The input is the audio signal from the teleconference, which is captured as digital audio data. The output is the recorded audio data. Specifically, the audio capture module is activated, the audio data is sent to the server in real time, and it is saved in the appropriate format.

[0717] Step 2:

[0718] The server converts acquired audio data into text data in real time. It analyzes the audio data using a speech recognition API (e.g., Google Speech-to-Text). The input is the audio data recorded in step 1, and the data processing involves converting the audio data into text data. The output is the converted text data. Specifically, the audio data is sent to the speech recognition API, and the text data returned by the API is saved to the server.

[0719] Step 3:

[0720] The server uses an emotion engine (e.g., Affectiva) to analyze the user's emotions in real time from audio and text data. The input consists of audio and text data, and the data processing involves speech tone analysis and text analysis. The output is the analyzed emotion data. Specifically, the emotion engine analyzes the tone of the audio and the content of the text, and creates a file that identifies emotions such as "tension" or "excitement."

[0721] Step 4:

[0722] The server extracts important agenda items based on text data and sentiment data using natural language processing techniques (e.g., spaCy). The input consists of text data and sentiment data, and data processing includes keyword extraction and text classification. The output is the extracted important agenda items. Specifically, it uses natural language processing techniques to analyze text data and list agenda items while considering sentiment data.

[0723] Step 5:

[0724] The server integrates extracted meeting minutes and sentiment data to generate a summary. The input consists of important meeting minutes and sentiment data, and the data processing involves calculations that integrate important statements and sentiment highlights. The output is the generated summary. Specifically, the extracted meeting minutes and sentiment data are saved to a database, and an algorithm runs to generate the summary.

[0725] Step 6:

[0726] The device provides a user interface, allowing users to view summaries, full text, and sentiment data. The input is the summary and corresponding audio data generated in step 5, and the output is a visually displayed summary and playable audio data. Specifically, when a user accesses the web portal and clicks the "View Summary" button, the summary, key minutes, sentiment data, and relevant comments are displayed.

[0727] Step 7:

[0728] When a user reviews summaries and sentiment data and clicks on a statement related to a specific point, the audio data related to that point is played. The input is the summary section and associated audio data specified by the user, and the output is the played audio data. Specifically, when a user clicks on a summary section in the web portal, the audio data corresponding to that statement is played, allowing them to listen to the details of the meeting content.

[0729] (Application Example 2)

[0730] 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."

[0731] Meetings and work instructions in factories require efficient and accurate recording. Traditional methods involved the cumbersome process of manually converting audio data into text and generating summaries. Furthermore, meeting minutes were created without incorporating emotional data from the meeting, making it difficult to issue work instructions that considered the atmosphere and urgency of the meeting. Therefore, a system was needed that provided a seamless process for real-time audio data acquisition, conversion to text, emotional data analysis, and automated work instruction generation.

[0732] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0733] In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing and extracting important agenda items from the text data, means for displaying the extracted important agenda items as a summary along with their rationale, means for providing the user with the summary and corresponding audio data and sentiment data, means for allowing the user to confirm important agenda items and summaries through a user interface, means for analyzing sentiment data to identify the user's emotions, and means for automatically generating work instructions based on the summary. This makes it possible to efficiently and accurately record the contents of meetings and automatically generate work instructions using sentiment data.

[0734] "Audio data" refers to sound information acquired from microphones and other audio input devices.

[0735] "Text data" refers to digital information obtained by converting audio data into text format.

[0736] "Important agenda items" refer to points or decisions that deserve particular attention during a meeting or discussion.

[0737] "Emotional data" refers to information that indicates the speaker's emotional state, and is analyzed from the tone of voice and the content of the text.

[0738] A "summary" is a text that concisely summarizes detailed information.

[0739] A "user interface" refers to the screens and means of operation that a user uses to interact with a system.

[0740] A "server" is a computer system that provides services over a network.

[0741] "Natural language processing technology" refers to the technology that enables computers to understand and generate human language.

[0742] A "work instruction sheet" is a document that contains instructions for performing a specific task.

[0743] This invention relates to a system for efficiently recording meetings and work instructions conducted within a factory, analyzing emotional data, and generating audio summaries and work instructions. Specifically, it performs real-time acquisition of audio data, conversion of text data, extraction of important agenda items, analysis of emotional data, summary generation, and automatic generation of work instructions.

[0744] The server uses a microphone to acquire audio data and activates an audio capture module to collect audio data in real time. The acquired audio data is recorded in an appropriate format (e.g., WAV or MP3). This recorded audio data is then used for transcription and sentiment analysis.

[0745] The server converts audio data acquired via a speech recognition API into text data in real time. This text data is stored on the server and used later for summarization and sentiment analysis.

[0746] The server uses an emotion engine to analyze the user's emotions in real time from voice and text data. This includes the ability to analyze voice tone and text content. For example, the user's emotions may be identified as "tension" or "excitement."

[0747] The server analyzes the text using natural language processing (NLP) techniques based on the analyzed character and sentiment data, and extracts key agenda items. This process includes keyword extraction and text classification that take sentiment data into account.

[0748] The server integrates the extracted key meeting minutes with sentiment data to generate a summary, and also identifies and stores the specific statements that support it. For example, parts of the sentiment data showing high levels of excitement are given particular attention.

[0749] The device provides a user-friendly interface. The web portal displays full text and sentiment data along with a summary, and users can click on specific meeting items to play audio data related to those points. It also generates PDF meeting minutes in a user-friendly format.

[0750] Specific example:

[0751] If a user says, "Let's discuss the new production line," during a meeting, the system records the statement, analyzes whether the sentiment is positive, summarizes the key points, and incorporates them into the work instructions.

[0752] Example of a prompt:

[0753] Audio data: "We will discuss the new production line."

[0754] Emotional data: "Positive"

[0755] Summary generated: "Discussions were held regarding a new production line."

[0756] This invention significantly improves operational efficiency within a factory by efficiently and accurately recording meeting content and automatically generating work instructions using emotional data.

[0757] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0758] Step 1:

[0759] The server activates the means to acquire audio data and collects audio data in real time through the microphone. The audio data is recorded in an appropriate format such as WAV or MP3.

[0760] Input: Audio data acquired from the microphone

[0761] Output: Recorded audio file

[0762] Specific operation: The microphone picks up sound, the audio data is sent to the server in real time, and saved in the appropriate file format.

[0763] Step 2:

[0764] The server converts the recorded audio data into text data in real time using a speech recognition API. This converted text data is stored on the server.

[0765] Input: Audio file

[0766] Output: Character data (text format)

[0767] Specific operation: The speech recognition API analyzes the audio file, converts the audio to text, and the generated text data is saved to storage.

[0768] Step 3:

[0769] The server uses an emotion engine to analyze the user's emotions in real time from text and audio data. The analyzed emotion data is stored on the server.

[0770] Input: Text data, audio data

[0771] Output: Sentiment data

[0772] Specific operation: The emotion engine analyzes text data and voice tone, tags the detected emotions as "positive," "negative," "tense," etc., and stores them in a database.

[0773] Step 4:

[0774] The server uses natural language processing techniques to extract key agenda items from text and sentiment data. These extracted agenda items are stored along with the detected sentiment data.

[0775] Input: Text data, sentiment data

[0776] Output: Key Agenda Items

[0777] Specific operation: A natural language processing algorithm analyzes the text, extracts important keywords and phrases, and stores them in storage along with sentiment data.

[0778] Step 5:

[0779] The server generates a summary based on the extracted key meeting items and sentiment data. This summary also identifies and stores the specific statements that supported it.

[0780] Input: Key meeting items, sentiment data

[0781] Output: Summary text

[0782] Specific operation: A natural language generation model takes important meeting items and sentiment data as input, generates a concise summary based on that, and saves specific parts of the statements.

[0783] Step 6:

[0784] The device displays summaries, full text, and sentiment data through a user interface. When the user clicks on a specific meeting item, it plays audio data related to that point. It also generates meeting minutes in PDF format.

[0785] Input: Summary text, full text, sentiment data, audio data

[0786] Output: Displayed summary, full text, sentiment data, and PDF meeting minutes.

[0787] Specific operation: The user interface displays a summary text and related data, and clicking a link plays the audio data. Additionally, a PDF generator produces meeting minutes.

[0788] 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.

[0789] 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.

[0790] 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.

[0791] [Third Embodiment]

[0792] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0793] 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.

[0794] 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).

[0795] 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.

[0796] 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.

[0797] 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).

[0798] 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.

[0799] 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.

[0800] 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.

[0801] 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.

[0802] 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.

[0803] 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".

[0804] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. This invention acquires conversation content as audio data in real time, transcribes it, analyzes the text data to extract important agenda items, and generates a summary. Furthermore, it displays the supporting statements for each key point in an easy-to-understand manner, and allows users to simultaneously review the text and corresponding audio data. Specific embodiments of this system are described below.

[0805] Acquisition of audio data

[0806] The server activates the audio capture module as soon as the teleconference begins, and acquires audio data. The audio data is recorded in real time in an appropriate format (e.g., WAV or MP3). This recorded audio data is then used in subsequent processes.

[0807] Transcript

[0808] The server converts the acquired audio data into text data in real time via a speech recognition API (e.g., a general speech recognition service). This transcribed data is stored on the server and used later for summarization and analysis.

[0809] Extraction of important agenda items

[0810] The server ingests text data, analyzes the text using natural language processing (NLP) techniques, and extracts key agenda items. This involves techniques such as keyword extraction and text classification. The extracted agenda items are then organized into summaries.

[0811] Summary display and presentation of evidence

[0812] The server displays the extracted agenda items as summaries. It also displays the specific statements (evidence) corresponding to each summary. This allows users to easily understand the background and details of the key points.

[0813] Providing a user interface

[0814] The device provides a user-friendly interface. For example, a web portal displays the full text along with a summary, and when a user clicks on a specific meeting item, it can play audio data related to that point. It also generates PDF meeting minutes in a user-friendly format.

[0815] Specific example

[0816] 1. The user starts a teleconference.

[0817] "Teleconference started. Audio is being captured."

[0818] 2. The server captures the audio and converts it into text data in real time.

[0819] "We are converting audio data to text in real time."

[0820] 3. The server analyzes the text data using natural language processing technology and extracts important agenda items.

[0821] "Analysis in progress. Extracting key agenda items."

[0822] 4. The device provides a user interface, allowing the user to review the summary.

[0823] "You can review the key points. Related comments will also be displayed."

[0824] 5. The user reviews the summary and selects and plays audio data related to specific points.

[0825] "Let's review the statement in point 2."

[0826] "Playing the statement." (Audio data plays)

[0827] This system can improve work efficiency by efficiently and accurately recording the content of teleconferences and allowing users to review key points and their supporting evidence as needed.

[0828] The following describes the processing flow.

[0829] Step 1:

[0830] The user initiates a teleconference. Once the teleconference starts, audio data capture begins automatically.

[0831] Step 2:

[0832] The server activates the audio capture module and collects audio data from the teleconference in real time. The audio data is saved in an appropriate format.

[0833] Step 3:

[0834] The server sends the acquired audio data in real time to a speech recognition API (for example, a general speech recognition service), and converts the audio data into text data.

[0835] Step 4:

[0836] The server sequentially saves the text data received from the speech recognition API.

[0837] Step 5:

[0838] The server takes in text data and performs analysis using natural language processing (NLP) techniques. This extracts key agenda items from the text.

[0839] Step 6:

[0840] The server extracts the agenda items, organizes them into a summary, and identifies and saves the specific statements that support them.

[0841] Step 7:

[0842] The server prepares to display the generated summary and supporting evidence through the user interface.

[0843] Step 8:

[0844] The device displays an interface that provides the user with a summary and supporting evidence. The user can use this to confirm the key points.

[0845] Step 9:

[0846] The user reviews the displayed summary and supporting evidence. Clicking on a statement related to a specific point plays the corresponding audio data.

[0847] Step 10:

[0848] The device plays audio data related to the summary specified by the user. This allows the user to view the text and audio simultaneously, enabling them to grasp the flow and nuances of the conversation.

[0849] In this way, the system efficiently and accurately records the content of teleconferences through a series of processing steps and effectively provides the user with the key points.

[0850] (Example 1)

[0851] 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."

[0852] One challenge in conducting teleconferences is the difficulty in efficiently and accurately recording meeting content and easily reviewing summaries afterward. Traditional methods require each participant to review the entire meeting, which is time-consuming and laborious. Furthermore, there is a lack of effective systems for quickly extracting and displaying important agenda items.

[0853] 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.

[0854] In this invention, the server includes means for detecting the start of a teleconference, means for recording audio data in real time, means for converting the recorded audio data into text data in real time, means for analyzing and extracting important agenda items from the converted text data using natural language processing technology, means for displaying the extracted important agenda items as a summary along with their supporting evidence, means for providing the summary and corresponding audio data to the user, and means for providing a user interface. This makes it possible to efficiently and accurately record the content of a teleconference and to quickly extract and display important agenda items.

[0855] A "teleconference" is a communication method in which multiple participants in remote locations conduct a meeting via audio and video.

[0856] "Means for detecting the start of a meeting" refers to a system or method that automatically recognizes the start of a meeting based on the reservation information or schedule of the teleconference.

[0857] "Means for recording audio data in real time" refers to a system or device that captures the audio of a meeting in real time and saves it in digital format.

[0858] "Methods for converting to text data in real time" refer to technologies that instantly convert acquired audio data into text format, such as speech recognition services and software.

[0859] "Means of analysis and extraction using natural language processing technology" refers to algorithms and programs used to analyze text data and identify important words, phrases, and agenda items.

[0860] "Means of displaying as a summary" refers to methods or interfaces that visually organize the analyzed data and display it on the screen in a way that is easy for the user to understand.

[0861] "Means of providing audio data to users" refers to systems and methods for delivering recorded audio data to users in formats such as streaming or file download.

[0862] "Means of providing a user interface" refers to graphical screens and interaction methods that allow users to operate a system and access necessary information.

[0863] This invention relates to a system for efficiently recording the content of teleconferences, extracting important agenda items, generating summaries, and providing reproducible meeting minutes. The following describes in detail specific embodiments of this invention.

[0864] Acquisition of audio data

[0865] The server has a function to detect the start of a teleconference and obtains the meeting start time from calendar information or the reservation system. When the start time arrives, it automatically activates the audio capture module and records audio data in real time. The audio data is saved in WAV or MP3 format and used in the subsequent processing described below.

[0866] Transcript

[0867] The server sends the recorded audio data to a speech recognition service (e.g., common speech recognition APIs, Google Cloud Speech-to-Text, or IBM Watson Speech to Text) and converts it into text data in real time. This text data is stored on the server and used for later analysis and summary generation.

[0868] Extraction of important agenda items

[0869] The server retrieves stored text data and analyzes the text using natural language processing (NLP) techniques. This analysis employs keyword extraction and text classification technologies (e.g., NLTK, SpaCy). As a result of the analysis, important meeting items are extracted and organized into a summary.

[0870] Summary and presentation of evidence

[0871] The server displays key meeting minutes organized as summaries. Specific statements supporting each summary are also displayed. This allows users to easily understand the background and details of the key points.

[0872] Providing a user interface

[0873] The device provides a user-friendly interface. This interface is implemented as a web portal and mobile application, allowing for the display of summaries and full text, and playback of related audio data by clicking on the summary. It also generates PDF meeting minutes in a user-friendly format.

[0874] Specific example

[0875] 1. The user starts a teleconference.

[0876] "Teleconference started. Audio is being captured."

[0877] 2. The server captures the audio and converts it into text data in real time.

[0878] "We are converting audio data to text in real time."

[0879] 3. The server analyzes the text data using natural language processing technology and extracts important agenda items.

[0880] "Analysis in progress. Extracting key agenda items."

[0881] 4. The device provides a user interface, allowing the user to review the summary.

[0882] "You can review the key points. Related comments will also be displayed."

[0883] 5. The user reviews the summary and selects and plays audio data related to specific points.

[0884] "Let's review the statement in point 2."

[0885] "Playing the statement." (Audio data plays)

[0886] Examples of prompts for generative AI models

[0887] Example of a prompt:

[0888] Please transcribe the following teleconference audio and extract the key agenda items. Then, display the specific statements related to each agenda item and link to the audio data.

[0889] 1. The topic of the teleconference is "Project progress report."

[0890] 2. Record all statements made during the meeting and create a summary.

[0891] 3. Please also make it possible to play the audio data of the statements corresponding to each summary.

[0892] This system efficiently and accurately records the content of teleconferences and enables the rapid extraction and display of important agenda items. This leads to improved work efficiency and accurate information sharing.

[0893] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0894] Step 1:

[0895] The server detects the start of the teleconference.

[0896] Input: Schedule information and reservation system data.

[0897] Specific operation: The server refers to pre-configured calendar information to detect the teleconference start time. Based on the detected start time, it prepares to start the audio capture module.

[0898] Output: Preparing to start the audio capture module.

[0899] Step 2:

[0900] The server activates the audio capture module and records audio data in real time.

[0901] Input: Audio signal from a teleconference.

[0902] Specific operation: The audio capture module acquires audio directly from a microphone or audio interface and saves it in real time as digital data (WAV or MP3).

[0903] Output: Audio data file recorded in real time.

[0904] Step 3:

[0905] The server sends the recorded audio data to a speech recognition service, where it is converted into text data.

[0906] Input: Recorded audio data.

[0907] Specific operation: The server sends the acquired audio file to a speech recognition API (e.g., Google Cloud Speech-to-Text) and receives the resulting text data as a response.

[0908] Output: Real-time generated character data.

[0909] Step 4:

[0910] The server saves the text data to the database.

[0911] Input: Generated character data.

[0912] Specific operation: The server stores the generated character data in a database (e.g., MySQL, PostgreSQL) as a timestamped record.

[0913] Output: Character data stored in the database.

[0914] Step 5:

[0915] The server analyzes the stored text data and extracts important agenda items.

[0916] Input: Character data stored in the database.

[0917] Specific operation: The server uses natural language processing (NLP) techniques to perform keyword extraction, text classification, and sentiment analysis. This identifies important meeting agenda items and organizes them into a summary.

[0918] Output: A list and summary of the key meeting agenda items extracted.

[0919] Step 6:

[0920] The server formats the data for displaying the summary and supporting evidence, and then sends it to the user interface.

[0921] Input: Key meeting agenda items and supporting statements extracted.

[0922] Specific actions: Format the data to display a summary and its supporting evidence, and prepare it for transmission to the endpoint.

[0923] Output: Formatted data.

[0924] Step 7:

[0925] The device provides a user-friendly interface.

[0926] Input: Formatted data.

[0927] Specific operation: Provide an interface for web portals and mobile applications that displays summaries, supporting evidence, and the full text. Implement a mechanism where, when a user clicks on a specific summary, the associated audio data is played.

[0928] Output: User interface.

[0929] Step 8:

[0930] The user interacts with the provided interface to obtain the necessary information.

[0931] Input: User interface screen.

[0932] Specific operation: Users view summaries through the interface and play relevant audio data by clicking on specific points. Furthermore, they can download meeting minutes in PDF format as needed.

[0933] Output: Playback of summary and related audio data, download of PDF meeting minutes.

[0934] (Application Example 1)

[0935] 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."

[0936] Traditional teleconference recording methods required participants to take notes themselves, which could lead to information overload, incompleteness, or misinterpretation. Furthermore, creating meeting minutes was time-consuming and labor-intensive, making it difficult even to grasp the key points. This was particularly problematic in factory meetings and conferences, negatively impacting work efficiency. A system capable of solving these problems was needed.

[0937] 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.

[0938] In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing and extracting important agenda items from the text data, means for displaying the extracted important agenda items as a summary along with their rationale, means for providing the user with the summary and corresponding audio data, means for recording the contents of meetings and conferences within the factory and automatically generating summaries, means for generating a meeting summary based on the extracted important agenda items, and means for displaying the summary on a user interface. This makes it possible to efficiently record the contents of meetings and conferences and automatically generate summaries, thereby significantly reducing the time and effort required for users to review meeting minutes.

[0939] "Audio data" refers to information recorded in digital format.

[0940] "Means of acquisition" refers to methods or devices for collecting specific data or information.

[0941] "Methods for converting to text data in real time" refers to methods or devices that instantly convert audio into text at the same time as it is heard.

[0942] "Important agenda items" are those that are particularly important topics or subjects of discussion in a meeting or discussion.

[0943] "Means of analysis and extraction" refers to methods or devices for analyzing data and extracting specific information from it.

[0944] A "summary" is a concise compilation of the most important parts of the original information.

[0945] A "user interface" refers to the screens or means by which a user operates a system or software.

[0946] "Meetings and discussions within the factory" refers to gatherings of managers and employees within the factory to discuss matters related to operations and production.

[0947] "Means for recording and automatically generating summaries" refers to methods or devices that save the content of meetings and discussions and automatically extract and summarize the key points based on that data.

[0948] "Extracted important agenda items" are the particularly important topics or items that have been extracted from the analyzed data.

[0949] "Means of providing information to the user" refers to methods or devices for presenting specific information to the user.

[0950] This invention relates to a system for efficiently recording the content of meetings and conferences within a factory and automatically generating summaries. This system performs a series of processes including audio data acquisition, transcription, text analysis, summary generation, and provision of a user interface. Specific embodiments are described below.

[0951] Acquisition of audio data

[0952] The server activates the audio capture module as soon as a meeting or discussion begins, and acquires audio data. The audio data is recorded in real time in an appropriate format (e.g., WAV or MP3). This recorded audio data is used in subsequent processes. Specifically, when a meeting starts, the server captures audio through the microphone and saves it as digital data.

[0953] Transcript

[0954] The server converts the acquired audio data into text data in real time using a speech recognition API. For example, the Python `speech_recognition` module can be used here. This transcribed data is stored on the server and used in the next analysis process.

[0955] Text analysis and extraction of key agenda items

[0956] The server analyzes the text data using natural language processing techniques to extract important agenda items. Specific techniques used for this process include natural language processing libraries such as spaCy and nltk. This automatically extracts particularly important discussions and decisions from the text data.

[0957] Summary generation and display

[0958] The server generates a summary of the extracted key agenda items. This summary should include specific statements (evidence) for each agenda item. The server displays the generated summary in the user interface (UI). The UI displays the full text along with the summary, and also includes a function that plays related audio data when the user clicks on a specific agenda item.

[0959] Providing a user interface

[0960] Terminals (e.g., robots and PCs in a factory) provide user-friendly interfaces. The web portal displays the full text along with a summary. Users can click on specific meeting items to play audio data related to those points. A PDF meeting transcript in an easy-to-read format is also generated.

[0961] Examples of specific cases and prompt statements

[0962] For example, suppose the following statement was made at a production meeting in a factory:

[0963] "We need to review next week's production schedule. Particular attention should be paid to the product quality on production line A."

[0964] A summary is generated based on this statement.

[0965] Examples of prompts to input into a generative AI model:

[0966] "Please analyze the following text and extract the key agenda items: We need to review next week's production schedule. Particular attention should be paid to the product quality of production line A."

[0967] As a result, a system can be realized that efficiently records the content of meetings and conferences within the factory and automatically generates summaries. By using this system, the time and effort required to create meeting minutes can be significantly reduced. In addition, important meeting items can be easily identified, improving meeting efficiency and accelerating business decision-making.

[0968] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0969] Step 1:

[0970] The server activates the audio capture module as soon as a meeting or discussion begins, and acquires audio data. This audio data is recorded in real time using a microphone in a digital format (e.g., WAV or MP3). The input is an audio signal, and the output is audio data in digital format.

[0971] Step 2:

[0972] The server converts acquired audio data into text data in real time. Specifically, it uses a speech recognition API (e.g., Python's speech_recognition module) to generate text from speech. The input is digital audio data, and the output is text data.

[0973] Step 3:

[0974] The server analyzes the generated text data using natural language processing (NLP) techniques to extract important agenda items. Specifically, it uses natural language processing libraries such as spaCy and nltk to identify specific keywords and phrases from the text. The input is text data, and the output is a list of important agenda items.

[0975] Step 4:

[0976] The server generates a summary of the extracted key agenda items. This summary also includes specific statements (evidence) for each agenda item. The input is a list of key agenda items, and the output is the summary text.

[0977] Step 5:

[0978] The terminal displays a summary provided by the server in its user interface. Specifically, it displays the summary and full text through a web portal or dedicated application, and when the user clicks on a specific meeting item, the related audio data is played. The input is the summary text and full text, and the output is the display in the user interface.

[0979] Step 6:

[0980] The user reviews the summary through the provided interface and plays audio data related to specific points. When the user clicks on a specific agenda item, the terminal plays the corresponding audio data. The input is the user's click operation, and the output is the playback of audio data.

[0981] 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.

[0982] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. In particular, by combining it with an emotion engine that recognizes user emotions, it becomes possible to create richer meeting minutes. This system has a series of functions including real-time acquisition of audio data, transcription, extraction of important agenda items, analysis of emotion data, summary generation, and playback of audio data. The following describes specific embodiments of this system.

[0983] Acquisition of audio data

[0984] The server activates the audio capture module as soon as the teleconference begins, collecting audio data in real time. The audio data is recorded in an appropriate format (e.g., WAV or MP3). This recorded audio data is later used for transcription and sentiment analysis.

[0985] Transcript

[0986] The server converts the acquired audio data into text data in real time via a speech recognition API. This transcribed data is stored on the server and used later for summarization and sentiment analysis.

[0987] Analysis of emotional data

[0988] The server uses an emotion engine to analyze the user's emotions in real time from voice and text data. This includes the ability to analyze voice tone and text content. For example, the user's emotions may be visualized as "tension" or "excitement."

[0989] Extraction of important agenda items

[0990] The server analyzes the text using natural language processing (NLP) techniques based on the analyzed character and sentiment data, and extracts key agenda items. This includes keyword extraction and text classification that take sentiment data into account.

[0991] Summary generation and display

[0992] The server integrates the extracted meeting minutes with sentiment data to generate a summary, and also identifies and saves the specific statements that support it. For example, parts of the sentiment data that show high levels of excitement are given particular attention.

[0993] Providing a user interface

[0994] The device provides a user-friendly interface. The web portal displays full text and sentiment data along with a summary, and users can click on specific meeting items to play audio data related to those points. It also generates PDF meeting minutes in a user-friendly format.

[0995] Specific example

[0996] 1. The user starts a teleconference.

[0997] "Teleconference started. Audio is being captured."

[0998] 2. The server captures the audio and converts it into text data in real time.

[0999] "We are converting audio data to text in real time."

[1000] 3. The server uses an emotion engine to analyze the user's emotions.

[1001] "Analyzing emotion data. Identifying user emotions."

[1002] 4. The server analyzes text data and sentiment data using natural language processing technology to extract important agenda items.

[1003] "Analysis in progress. Extracting key agenda items."

[1004] 5. The device provides a user interface, allowing users to view summaries and sentiment data.

[1005] "You can view key points and sentiment data. Related comments are also displayed."

[1006] 6. When a user reviews summaries and sentiment data and clicks on a statement related to a specific point, audio data related to that point is played.

[1007] "Let's review the statement in point 2."

[1008] "Playing the statement." (Audio data plays)

[1009] This system efficiently and accurately records the content of teleconferences and effectively provides users with key points, thereby improving work efficiency. Furthermore, by utilizing sentiment data, it makes it easier to understand the atmosphere and urgency of the meeting.

[1010] The following describes the processing flow.

[1011] Step 1:

[1012] The user initiates a teleconference. Once the teleconference starts, audio data capture begins automatically.

[1013] User: "Starting teleconference."

[1014] Step 2:

[1015] The server activates the audio capture module and collects audio data from the teleconference in real time. The audio data is saved in an appropriate format.

[1016] Server: "We are collecting audio data in real time."

[1017] Step 3:

[1018] The server collects audio data and sends it to a speech recognition API in real time, where it converts the audio into text data. This text data is then stored sequentially on the server.

[1019] Server: "Converting audio data to text data."

[1020] Step 4:

[1021] The server uses an emotion engine to analyze collected audio and text data in real time to recognize the user's emotions. The emotion engine analyzes the tone of voice and the content of the text.

[1022] Server: "Analyzing emotion data."

[1023] Step 5:

[1024] The server uses natural language processing (NLP) techniques to extract key agenda items based on the analyzed text and sentiment data.

[1025] Server: "We are analyzing text and sentiment data to extract key agenda items."

[1026] Step 6:

[1027] The server extracts the agenda items, organizes them into a summary, and saves them along with the corresponding specific statements. Sentiment data is also reflected in the summary.

[1028] Server: "We are generating summaries and tracking specific statements and sentiment data."

[1029] Step 7:

[1030] The device displays the generated summary and sentiment data through the user interface, allowing users to easily review key points.

[1031] Terminal: "Preparing the interface for summary and sentiment data."

[1032] Step 8:

[1033] Users can review the displayed summary and sentiment data. Selecting a specific agenda item will display comments and sentiment data related to that point.

[1034] User: "I'd like to confirm the specific points."

[1035] Terminal: "Displays selected points, related statements, and sentiment data."

[1036] Step 9:

[1037] When a user plays audio data related to a specific point, the relevant portion of the audio data will be played.

[1038] User: "Playing audio data related to the main points."

[1039] Terminal: "Related audio data will be played." (Audio data playback)

[1040] In this way, the system efficiently and accurately records the content of teleconferences through a series of processing steps, effectively providing users with key points and sentiment data. Users can review text and audio simultaneously, grasping the flow, nuances, and emotional changes of the conversation.

[1041] (Example 2)

[1042] 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."

[1043] Traditional teleconferencing systems struggled to efficiently record meeting content and generate key agenda items and summaries. Furthermore, they lacked sufficient means to analyze user sentiment data and understand the meeting atmosphere, making minute-taking cumbersome. Additionally, quickly accessing important statements during meetings was difficult.

[1044] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing the user's emotions from the audio data and text data, means for extracting important agenda items based on the analyzed text data and emotion data, means for integrating the extracted agenda items and emotion data to generate a summary, and means for providing the user with the generated summary and corresponding audio data. This enables efficient and accurate recording of meeting content, extraction of important agenda items, and understanding of the overall atmosphere of the meeting.

[1045] "Audio data" refers to data obtained by digitizing audio signals from teleconferences, etc.

[1046] "Real-time" refers to a state where data acquisition and processing are performed instantly and without delay.

[1047] "Text data" refers to information in text format generated from audio data.

[1048] "Emotional data" refers to data that indicates a user's emotional state, analyzed from voice and text.

[1049] "Meeting items" refer to the main topics or themes discussed during the teleconference.

[1050] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[1051] A "summary" is a text that concisely summarizes the specific content and key points.

[1052] A "server" is a computer system that processes data and provides various services to users.

[1053] A "user interface" is the environment that includes screens and input methods for a user to interact with a system.

[1054] A "voice capture module" is a device and software for collecting voice data and storing it as digital data.

[1055] A "database" is a system for systematically storing, searching, and editing information.

[1056] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. In particular, by combining it with an emotion engine that recognizes user emotions, it becomes possible to create richer meeting minutes. This system has a series of functions including real-time acquisition of audio data, transcription, extraction of important agenda items, analysis of emotion data, summary generation, and playback of audio data.

[1057] Hardware and software used

[1058] 1. Server

[1059] Voice capture module: Collects voice data in real time.

[1060] Speech Recognition API: Using "Google Speech-to-Text" as an example, we will convert speech data into text data in real time.

[1061] Emotion Engine: Using "Affectiva" as an example, we analyze the user's emotions from audio and text data.

[1062] Natural language processing technology: As an example, we will use "spaCy" to extract important agenda items based on text data and sentiment data.

[1063] Database: Stores analysis results and generated summaries.

[1064] 2. Terminal

[1065] User Interface: Displays summaries, sentiment data, and other information to the user using a web browser or dedicated application.

[1066] PDF generation tool: Used to output meeting minutes in PDF format.

[1067] Explanation of the program's processing flow

[1068] The server activates the audio capture module as soon as the teleconference begins, collecting audio data in real time. The collected audio data is recorded in WAV or MP3 format and used later for transcription and sentiment analysis.

[1069] Next, the server converts the audio data into text data in real time using a speech recognition API (Google Speech-to-Text). This text data is stored in a database. The text data generated through transcription is then used for subsequent summarization and sentiment analysis.

[1070] The server uses an emotion engine (Affectiva) to analyze the user's emotions in real time from voice and text data. Emotion analysis includes analyzing the tone of voice and the content of the text. For example, emotional data such as "tension" or "excitement" may be identified.

[1071] Furthermore, the server uses natural language processing technology (spaCy) to analyze text and sentiment data. It extracts important agenda items and performs keyword extraction and text classification that takes sentiment data into account. The extracted agenda items are used to generate summaries.

[1072] Subsequently, the server integrates the extracted meeting minutes and sentiment data to generate a summary. The generated summary also identifies and stores the specific statements that support it. In particular, sections where the sentiment data indicates high levels of excitement are given special attention in the summary.

[1073] Finally, the device provides a user interface, allowing users to view summaries, full text, and sentiment data. Clicking on a specific meeting item plays audio data related to that point. The user interface also generates easy-to-read PDF meeting minutes.

[1074] Specific example

[1075] 1. The user starts a teleconference.

[1076] Example: The user clicks the "Start Teleconference" button.

[1077] Result: The teleconference begins, and the server starts audio capture.

[1078] 2. The server captures the audio and converts it into text data in real time.

[1079] Example: The server batches audio data every second and converts it into text data.

[1080] Result: Text data is generated as the teleconference progresses.

[1081] 3. The server uses an emotion engine to analyze the user's emotions.

[1082] Example: The server saves analysis results such as "the user's voice tone is rising" during the analysis.

[1083] Result: Emotional data is generated, and a timestamp corresponding to a specific emotion is recorded.

[1084] 4. The server analyzes text data and sentiment data using natural language processing technology to extract important agenda items.

[1085] Example: Keyphrase extraction and TextRank algorithms are used to extract important agenda items.

[1086] Result: Important meeting items are listed and saved in the database.

[1087] 5. The device provides a user interface, allowing users to view summaries and sentiment data.

[1088] Example: When a user accesses a web portal, they can click the "View Summary" button.

[1089] Results: A summary, key agenda items, sentiment data, and relevant comments are displayed.

[1090] Example of a prompt

[1091] 1. Record the start of the teleconference in real time.

[1092] 2. Convert the audio data to text and save it.

[1093] 3. Perform sentiment analysis from the text and audio data.

[1094] 4. Analyze the sentiment data and text data to extract the important agenda items.

[1095] 5. Generate a summary and display it in the user interface.

[1096] This invention makes it possible to efficiently record the content of teleconferences, extract important agenda items, and generate summaries. Furthermore, by analyzing sentiment data, it becomes easier to understand the atmosphere of the meeting and the emotions of the participants.

[1097] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1098] Step 1:

[1099] The server activates the audio capture module when the teleconference begins. The audio capture module collects audio data in real time and records it in WAV or MP3 format. The input is the audio signal from the teleconference, which is captured as digital audio data. The output is the recorded audio data. Specifically, the audio capture module is activated, the audio data is sent to the server in real time, and it is saved in the appropriate format.

[1100] Step 2:

[1101] The server converts acquired audio data into text data in real time. It analyzes the audio data using a speech recognition API (e.g., Google Speech-to-Text). The input is the audio data recorded in step 1, and the data processing involves converting the audio data into text data. The output is the converted text data. Specifically, the audio data is sent to the speech recognition API, and the text data returned by the API is saved to the server.

[1102] Step 3:

[1103] The server uses an emotion engine (e.g., Affectiva) to analyze the user's emotions in real time from audio and text data. The input consists of audio and text data, and the data processing involves speech tone analysis and text analysis. The output is the analyzed emotion data. Specifically, the emotion engine analyzes the tone of the audio and the content of the text, and creates a file that identifies emotions such as "tension" or "excitement."

[1104] Step 4:

[1105] The server extracts important agenda items based on text data and sentiment data using natural language processing techniques (e.g., spaCy). The input consists of text data and sentiment data, and data processing includes keyword extraction and text classification. The output is the extracted important agenda items. Specifically, it uses natural language processing techniques to analyze text data and list agenda items while considering sentiment data.

[1106] Step 5:

[1107] The server integrates extracted meeting minutes and sentiment data to generate a summary. The input consists of important meeting minutes and sentiment data, and the data processing involves calculations that integrate important statements and sentiment highlights. The output is the generated summary. Specifically, the extracted meeting minutes and sentiment data are saved to a database, and an algorithm runs to generate the summary.

[1108] Step 6:

[1109] The device provides a user interface, allowing users to view summaries, full text, and sentiment data. The input is the summary and corresponding audio data generated in step 5, and the output is a visually displayed summary and playable audio data. Specifically, when a user accesses the web portal and clicks the "View Summary" button, the summary, key minutes, sentiment data, and relevant comments are displayed.

[1110] Step 7:

[1111] When a user reviews summaries and sentiment data and clicks on a statement related to a specific point, the audio data related to that point is played. The input is the summary section and associated audio data specified by the user, and the output is the played audio data. Specifically, when a user clicks on a summary section in the web portal, the audio data corresponding to that statement is played, allowing them to listen to the details of the meeting content.

[1112] (Application Example 2)

[1113] 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."

[1114] Meetings and work instructions in factories require efficient and accurate recording. Traditional methods involved the cumbersome process of manually converting audio data into text and generating summaries. Furthermore, meeting minutes were created without incorporating emotional data from the meeting, making it difficult to issue work instructions that considered the atmosphere and urgency of the meeting. Therefore, a system was needed that provided a seamless process for real-time audio data acquisition, conversion to text, emotional data analysis, and automated work instruction generation.

[1115] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1116] In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing and extracting important agenda items from the text data, means for displaying the extracted important agenda items as a summary along with their rationale, means for providing the user with the summary and corresponding audio data and sentiment data, means for allowing the user to confirm important agenda items and summaries through a user interface, means for analyzing sentiment data to identify the user's emotions, and means for automatically generating work instructions based on the summary. This makes it possible to efficiently and accurately record the contents of meetings and automatically generate work instructions using sentiment data.

[1117] "Audio data" refers to sound information acquired from microphones and other audio input devices.

[1118] "Text data" refers to digital information obtained by converting audio data into text format.

[1119] "Important agenda items" refer to points or decisions that deserve particular attention during a meeting or discussion.

[1120] "Emotional data" refers to information that indicates the speaker's emotional state, and is analyzed from the tone of voice and the content of the text.

[1121] A "summary" is a text that concisely summarizes detailed information.

[1122] A "user interface" refers to the screens and means of operation that allow a user to interact with a system.

[1123] A "server" is a computer system that provides services over a network.

[1124] "Natural language processing technology" refers to the technology that enables computers to understand and generate human language.

[1125] A "work instruction sheet" is a document that contains instructions for performing a specific task.

[1126] This invention relates to a system for efficiently recording meetings and work instructions conducted within a factory, analyzing emotional data, and generating audio summaries and work instructions. Specifically, it performs real-time acquisition of audio data, conversion of text data, extraction of important agenda items, analysis of emotional data, summary generation, and automatic generation of work instructions.

[1127] The server uses a microphone to acquire audio data and activates an audio capture module to collect audio data in real time. The acquired audio data is recorded in an appropriate format (e.g., WAV or MP3). This recorded audio data is then used for transcription and sentiment analysis.

[1128] The server converts audio data acquired via a speech recognition API into text data in real time. This text data is stored on the server and used later for summarization and sentiment analysis.

[1129] The server uses an emotion engine to analyze the user's emotions in real time from voice and text data. This includes the ability to analyze voice tone and text content. For example, the user's emotions may be identified as "tension" or "excitement."

[1130] The server analyzes the text using natural language processing (NLP) techniques based on the analyzed character and sentiment data, and extracts key agenda items. This process includes keyword extraction and text classification that take sentiment data into account.

[1131] The server integrates the extracted key meeting minutes with sentiment data to generate a summary, and also identifies and stores the specific statements that support it. For example, parts of the sentiment data showing high levels of excitement are given particular attention.

[1132] The device provides a user-friendly interface. The web portal displays full text and sentiment data along with a summary, and users can click on specific meeting items to play audio data related to those points. It also generates PDF meeting minutes in a user-friendly format.

[1133] Specific example:

[1134] If a user says, "Let's discuss the new production line," during a meeting, the system records the statement, analyzes whether the sentiment is positive, summarizes the key points, and incorporates them into the work instructions.

[1135] Example of a prompt:

[1136] Audio data: "We will discuss the new production line."

[1137] Emotional data: "Positive"

[1138] Summary generated: "Discussions were held regarding a new production line."

[1139] This invention significantly improves operational efficiency within a factory by efficiently and accurately recording meeting content and automatically generating work instructions using emotional data.

[1140] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1141] Step 1:

[1142] The server activates the means to acquire audio data and collects audio data in real time through the microphone. The audio data is recorded in an appropriate format such as WAV or MP3.

[1143] Input: Audio data acquired from the microphone

[1144] Output: Recorded audio file

[1145] Specific operation: The microphone picks up sound, the audio data is sent to the server in real time, and saved in the appropriate file format.

[1146] Step 2:

[1147] The server converts the recorded audio data into text data in real time using a speech recognition API. This converted text data is stored on the server.

[1148] Input: Audio file

[1149] Output: Character data (text format)

[1150] Specific operation: The speech recognition API analyzes the audio file, converts the audio to text, and the generated text data is saved to storage.

[1151] Step 3:

[1152] The server uses an emotion engine to analyze the user's emotions in real time from text and audio data. The analyzed emotion data is stored on the server.

[1153] Input: Text data, audio data

[1154] Output: Sentiment data

[1155] Specific operation: The emotion engine analyzes text data and voice tone, tags the detected emotions as "positive," "negative," "tense," etc., and stores them in a database.

[1156] Step 4:

[1157] The server uses natural language processing techniques to extract key agenda items from text and sentiment data. These extracted agenda items are stored along with the detected sentiment data.

[1158] Input: Text data, sentiment data

[1159] Output: Key Agenda Items

[1160] Specific operation: A natural language processing algorithm analyzes the text, extracts important keywords and phrases, and stores them in storage along with sentiment data.

[1161] Step 5:

[1162] The server generates a summary based on the extracted key meeting items and sentiment data. This summary also identifies and stores the specific statements that supported it.

[1163] Input: Key meeting items, sentiment data

[1164] Output: Summary text

[1165] Specific operation: A natural language generation model takes important meeting items and sentiment data as input, generates a concise summary based on that, and saves specific parts of the statements.

[1166] Step 6:

[1167] The device displays summaries, full text, and sentiment data through a user interface. When the user clicks on a specific meeting item, it plays audio data related to that point. It also generates meeting minutes in PDF format.

[1168] Input: Summary text, full text, sentiment data, audio data

[1169] Output: Displayed summary, full text, sentiment data, and PDF meeting minutes.

[1170] Specific operation: The user interface displays a summary text and related data, and clicking a link plays the audio data. Additionally, a PDF generator produces meeting minutes.

[1171] 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.

[1172] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). An 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.

[1173] 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.

[1174] [Fourth Embodiment]

[1175] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1176] 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.

[1177] 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).

[1178] 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.

[1179] 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.

[1180] 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).

[1181] 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.

[1182] 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.

[1183] 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.

[1184] 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.

[1185] 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.

[1186] 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.

[1187] 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".

[1188] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. This invention acquires conversation content as audio data in real time, transcribes it, analyzes the text data to extract important agenda items, and generates a summary. Furthermore, it displays the supporting statements for each key point in an easy-to-understand manner, and allows users to simultaneously review the text and corresponding audio data. Specific embodiments of this system are described below.

[1189] Acquisition of audio data

[1190] The server activates the audio capture module as soon as the teleconference begins, and acquires audio data. The audio data is recorded in real time in an appropriate format (e.g., WAV or MP3). This recorded audio data is then used in subsequent processes.

[1191] Transcript

[1192] The server converts the acquired audio data into text data in real time via a speech recognition API (e.g., a general speech recognition service). This transcribed data is stored on the server and used later for summarization and analysis.

[1193] Extraction of important agenda items

[1194] The server ingests text data, analyzes the text using natural language processing (NLP) techniques, and extracts key agenda items. This involves techniques such as keyword extraction and text classification. The extracted agenda items are then organized into summaries.

[1195] Summary display and presentation of evidence

[1196] The server displays the extracted agenda items as summaries. It also displays the specific statements (evidence) corresponding to each summary. This allows users to easily understand the background and details of the key points.

[1197] Providing a user interface

[1198] The device provides a user-friendly interface. For example, a web portal displays the full text along with a summary, and when a user clicks on a specific meeting item, it can play audio data related to that point. It also generates PDF meeting minutes in a user-friendly format.

[1199] Specific example

[1200] 1. The user starts a teleconference.

[1201] "Teleconference started. Audio is being captured."

[1202] 2. The server captures the audio and converts it into text data in real time.

[1203] "We are converting audio data to text in real time."

[1204] 3. The server analyzes the text data using natural language processing technology and extracts important agenda items.

[1205] "Analysis in progress. Extracting key agenda items."

[1206] 4. The device provides a user interface, allowing the user to review the summary.

[1207] "You can review the key points. Related comments will also be displayed."

[1208] 5. The user reviews the summary and selects and plays audio data related to specific points.

[1209] "Let's review the statement in point 2."

[1210] "Playing the statement." (Audio data plays)

[1211] This system can improve work efficiency by efficiently and accurately recording the content of teleconferences and allowing users to review key points and their supporting evidence as needed.

[1212] The following describes the processing flow.

[1213] Step 1:

[1214] The user initiates a teleconference. Once the teleconference starts, audio data capture begins automatically.

[1215] Step 2:

[1216] The server activates the audio capture module and collects audio data from the teleconference in real time. The audio data is saved in an appropriate format.

[1217] Step 3:

[1218] The server sends the acquired audio data in real time to a speech recognition API (for example, a general speech recognition service), and converts the audio data into text data.

[1219] Step 4:

[1220] The server sequentially saves the text data received from the speech recognition API.

[1221] Step 5:

[1222] The server takes in text data and performs analysis using natural language processing (NLP) techniques. This extracts key agenda items from the text.

[1223] Step 6:

[1224] The server extracts the agenda items, organizes them into a summary, and identifies and saves the specific statements that support them.

[1225] Step 7:

[1226] The server prepares to display the generated summary and supporting evidence through the user interface.

[1227] Step 8:

[1228] The device displays an interface that provides the user with a summary and supporting evidence. The user can use this to confirm the key points.

[1229] Step 9:

[1230] The user reviews the displayed summary and supporting evidence. Clicking on a statement related to a specific point plays the corresponding audio data.

[1231] Step 10:

[1232] The device plays audio data related to the summary specified by the user. This allows the user to view the text and audio simultaneously, enabling them to grasp the flow and nuances of the conversation.

[1233] In this way, the system efficiently and accurately records the content of teleconferences through a series of processing steps and effectively provides the key points to the user.

[1234] (Example 1)

[1235] 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".

[1236] One challenge in conducting teleconferences is the difficulty in efficiently and accurately recording meeting content and easily reviewing summaries afterward. Traditional methods require each participant to review the entire meeting, which is time-consuming and laborious. Furthermore, there is a lack of effective systems for quickly extracting and displaying important agenda items.

[1237] 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.

[1238] In this invention, the server includes means for detecting the start of a teleconference, means for recording audio data in real time, means for converting the recorded audio data into text data in real time, means for analyzing and extracting important agenda items from the converted text data using natural language processing technology, means for displaying the extracted important agenda items as a summary along with their supporting evidence, means for providing the summary and corresponding audio data to the user, and means for providing a user interface. This makes it possible to efficiently and accurately record the content of a teleconference and to quickly extract and display important agenda items.

[1239] A "teleconference" is a communication method in which multiple participants in remote locations conduct a meeting via audio and video.

[1240] "Means for detecting the start of a meeting" refers to a system or method that automatically recognizes the start of a meeting based on the reservation information or schedule of the teleconference.

[1241] "Means for recording audio data in real time" refers to a system or device that captures the audio of a meeting in real time and saves it in digital format.

[1242] "Methods for converting to text data in real time" refer to technologies that instantly convert acquired audio data into text format, such as speech recognition services and software.

[1243] "Means of analysis and extraction using natural language processing technology" refers to algorithms and programs used to analyze text data and identify important words, phrases, and agenda items.

[1244] "Means of displaying as a summary" refers to methods or interfaces that visually organize the analyzed data and display it on the screen in a way that is easy for the user to understand.

[1245] "Means of providing audio data to users" refers to systems and methods for delivering recorded audio data to users in formats such as streaming or file download.

[1246] "Means of providing a user interface" refers to graphical screens and interaction methods that allow users to operate a system and access necessary information.

[1247] This invention relates to a system for efficiently recording the content of teleconferences, extracting important agenda items, generating summaries, and providing reproducible meeting minutes. The following describes in detail specific embodiments of this invention.

[1248] Acquisition of audio data

[1249] The server has a function to detect the start of a teleconference and obtains the meeting start time from calendar information or the reservation system. When the start time arrives, it automatically activates the audio capture module and records audio data in real time. The audio data is saved in WAV or MP3 format and used in the subsequent processing described below.

[1250] Transcript

[1251] The server sends the recorded audio data to a speech recognition service (e.g., common speech recognition APIs, Google Cloud Speech-to-Text, or IBM Watson Speech to Text) and converts it into text data in real time. This text data is stored on the server and used for later analysis and summary generation.

[1252] Extraction of important agenda items

[1253] The server retrieves stored text data and analyzes the text using natural language processing (NLP) techniques. This analysis employs keyword extraction and text classification technologies (e.g., NLTK, SpaCy). As a result of the analysis, important meeting items are extracted and organized into a summary.

[1254] Summary and presentation of evidence

[1255] The server displays key meeting minutes organized as summaries. Specific statements supporting each summary are also displayed. This allows users to easily understand the background and details of the key points.

[1256] Providing a user interface

[1257] The device provides a user-friendly interface. This interface is implemented as a web portal and mobile application, allowing for the display of summaries and full text, and playback of related audio data by clicking on the summary. It also generates PDF meeting minutes in a user-friendly format.

[1258] Specific example

[1259] 1. The user starts a teleconference.

[1260] "Teleconference started. Audio is being captured."

[1261] 2. The server captures the audio and converts it into text data in real time.

[1262] "We are converting audio data to text in real time."

[1263] 3. The server analyzes the text data using natural language processing technology and extracts important agenda items.

[1264] "Analysis in progress. Extracting key agenda items."

[1265] 4. The device provides a user interface, allowing the user to review the summary.

[1266] "You can review the key points. Related comments will also be displayed."

[1267] 5. The user reviews the summary and selects and plays audio data related to specific points.

[1268] "Let's review the statement in point 2."

[1269] "Playing the statement." (Audio data plays)

[1270] Examples of prompts for generative AI models

[1271] Example of a prompt:

[1272] Please transcribe the following teleconference audio and extract the key agenda items. Then, display the specific statements related to each agenda item and link to the audio data.

[1273] 1. The topic of the teleconference is "Project progress report."

[1274] 2. Record all statements made during the meeting and create a summary.

[1275] 3. Please also make it possible to play the audio data of the statements corresponding to each summary.

[1276] This system efficiently and accurately records the content of teleconferences and enables the rapid extraction and display of important agenda items. This leads to improved work efficiency and accurate information sharing.

[1277] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1278] Step 1:

[1279] The server detects the start of the teleconference.

[1280] Input: Schedule information and reservation system data.

[1281] Specific operation: The server refers to pre-configured calendar information to detect the teleconference start time. Based on the detected start time, it prepares to start the audio capture module.

[1282] Output: Preparing to start the audio capture module.

[1283] Step 2:

[1284] The server activates the audio capture module and records audio data in real time.

[1285] Input: Audio signal from a teleconference.

[1286] Specific operation: The audio capture module acquires audio directly from a microphone or audio interface and saves it in real time as digital data (WAV or MP3).

[1287] Output: Audio data file recorded in real time.

[1288] Step 3:

[1289] The server sends the recorded audio data to a speech recognition service, where it is converted into text data.

[1290] Input: Recorded audio data.

[1291] Specific operation: The server sends the acquired audio file to a speech recognition API (e.g., Google Cloud Speech-to-Text) and receives the resulting text data as a response.

[1292] Output: Real-time generated character data.

[1293] Step 4:

[1294] The server saves the text data to the database.

[1295] Input: Generated character data.

[1296] Specific operation: The server stores the generated character data in a database (e.g., MySQL, PostgreSQL) as a timestamped record.

[1297] Output: Character data stored in the database.

[1298] Step 5:

[1299] The server analyzes the stored text data and extracts important agenda items.

[1300] Input: Character data stored in the database.

[1301] Specific operation: The server uses natural language processing (NLP) techniques to perform keyword extraction, text classification, and sentiment analysis. This identifies important meeting agenda items and organizes them into a summary.

[1302] Output: A list and summary of the key meeting agenda items extracted.

[1303] Step 6:

[1304] The server formats the data for displaying the summary and supporting evidence, and then sends it to the user interface.

[1305] Input: Key meeting agenda items and supporting statements extracted.

[1306] Specific actions: Format the data to display a summary and its supporting evidence, and prepare it for transmission to the endpoint.

[1307] Output: Formatted data.

[1308] Step 7:

[1309] The device provides a user-friendly interface.

[1310] Input: Formatted data.

[1311] Specific operation: Provide an interface for web portals and mobile applications that displays summaries, supporting evidence, and the full text. Implement a mechanism where, when a user clicks on a specific summary, the associated audio data is played.

[1312] Output: User interface.

[1313] Step 8:

[1314] The user interacts with the provided interface to obtain the necessary information.

[1315] Input: User interface screen.

[1316] Specific operation: Users view summaries through the interface and play relevant audio data by clicking on specific points. Furthermore, they can download meeting minutes in PDF format as needed.

[1317] Output: Playback of summary and related audio data, download of PDF meeting minutes.

[1318] (Application Example 1)

[1319] 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".

[1320] Traditional teleconference recording methods required participants to take notes themselves, which could lead to information overload, incompleteness, or misinterpretation. Furthermore, creating meeting minutes was time-consuming and labor-intensive, making it difficult even to grasp the key points. This was particularly problematic in factory meetings and conferences, negatively impacting work efficiency. A system capable of solving these problems was needed.

[1321] 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.

[1322] In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing and extracting important agenda items from the text data, means for displaying the extracted important agenda items as a summary along with their rationale, means for providing the user with the summary and corresponding audio data, means for recording the contents of meetings and conferences within the factory and automatically generating summaries, means for generating a meeting summary based on the extracted important agenda items, and means for displaying the summary on a user interface. This makes it possible to efficiently record the contents of meetings and conferences and automatically generate summaries, thereby significantly reducing the time and effort required for users to review meeting minutes.

[1323] "Audio data" refers to information recorded in digital format.

[1324] "Means of acquisition" refers to methods or devices for collecting specific data or information.

[1325] "Methods for converting to text data in real time" refer to methods or devices that instantly convert audio into text at the same time as it is heard.

[1326] "Important agenda items" are those that are particularly important topics or subjects of discussion in a meeting or discussion.

[1327] "Means of analysis and extraction" refers to methods or devices for analyzing data and extracting specific information from it.

[1328] A "summary" is a concise compilation of the most important parts of the original information.

[1329] A "user interface" refers to the screens or means by which a user operates a system or software.

[1330] "Meetings and discussions within the factory" refers to gatherings of managers and employees within the factory to discuss matters related to operations and production.

[1331] "Means for recording and automatically generating summaries" refers to methods or devices that save the content of meetings and discussions and automatically extract and summarize the key points based on that data.

[1332] "Extracted important agenda items" are the particularly important topics or items that have been extracted from the analyzed data.

[1333] "Means of providing information to the user" refers to methods or devices for presenting specific information to the user.

[1334] This invention relates to a system for efficiently recording the content of meetings and conferences within a factory and automatically generating summaries. This system performs a series of processes including audio data acquisition, transcription, text analysis, summary generation, and provision of a user interface. Specific embodiments are described below.

[1335] Acquisition of audio data

[1336] The server activates the audio capture module as soon as a meeting or discussion begins, and acquires audio data. The audio data is recorded in real time in an appropriate format (e.g., WAV or MP3). This recorded audio data is used in subsequent processes. Specifically, when a meeting starts, the server captures audio through the microphone and saves it as digital data.

[1337] Transcript

[1338] The server converts the acquired audio data into text data in real time using a speech recognition API. For example, the Python `speech_recognition` module can be used here. This transcribed data is stored on the server and used in the next analysis process.

[1339] Text analysis and extraction of key agenda items

[1340] The server analyzes the text data using natural language processing techniques to extract important agenda items. Specific techniques used for this process include natural language processing libraries such as spaCy and nltk. This automatically extracts particularly important discussions and decisions from the text data.

[1341] Summary generation and display

[1342] The server generates a summary of the extracted key agenda items. This summary should include specific statements (evidence) for each agenda item. The server displays the generated summary in the user interface (UI). The UI displays the full text along with the summary, and also includes a function that plays related audio data when the user clicks on a specific agenda item.

[1343] Providing a user interface

[1344] Terminals (e.g., robots and PCs in a factory) provide user-friendly interfaces. The web portal displays the full text along with a summary. Users can click on specific meeting items to play audio data related to those points. A PDF meeting transcript in an easy-to-read format is also generated.

[1345] Examples of specific cases and prompt statements

[1346] For example, suppose the following statement was made at a production meeting in a factory:

[1347] "We need to review next week's production schedule. Particular attention should be paid to the product quality on production line A."

[1348] A summary is generated based on this statement.

[1349] Examples of prompts to input into a generative AI model:

[1350] "Please analyze the following text and extract the key agenda items: We need to review next week's production schedule. Particular attention should be paid to the product quality of production line A."

[1351] As a result, a system can be realized that efficiently records the content of meetings and conferences within the factory and automatically generates summaries. By using this system, the time and effort required to create meeting minutes can be significantly reduced. In addition, important meeting items can be easily identified, improving meeting efficiency and accelerating business decision-making.

[1352] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1353] Step 1:

[1354] The server activates the audio capture module as soon as a meeting or discussion begins, and acquires audio data. This audio data is recorded in real time using a microphone in a digital format (e.g., WAV or MP3). The input is an audio signal, and the output is audio data in digital format.

[1355] Step 2:

[1356] The server converts acquired audio data into text data in real time. Specifically, it uses a speech recognition API (e.g., Python's speech_recognition module) to generate text from speech. The input is digital audio data, and the output is text data.

[1357] Step 3:

[1358] The server analyzes the generated text data using natural language processing (NLP) techniques to extract important agenda items. Specifically, it uses natural language processing libraries such as spaCy and nltk to identify specific keywords and phrases from the text. The input is text data, and the output is a list of important agenda items.

[1359] Step 4:

[1360] The server generates a summary of the extracted key agenda items. This summary also includes specific statements (evidence) for each agenda item. The input is a list of key agenda items, and the output is the summary text.

[1361] Step 5:

[1362] The terminal displays a summary provided by the server in its user interface. Specifically, it displays the summary and full text through a web portal or dedicated application, and when the user clicks on a specific meeting item, the related audio data is played. The input is the summary text and full text, and the output is the display in the user interface.

[1363] Step 6:

[1364] The user reviews the summary through the provided interface and plays audio data related to specific points. When the user clicks on a specific agenda item, the terminal plays the corresponding audio data. The input is the user's click operation, and the output is the playback of audio data.

[1365] 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.

[1366] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. In particular, by combining it with an emotion engine that recognizes user emotions, it becomes possible to create richer meeting minutes. This system has a series of functions including real-time acquisition of audio data, transcription, extraction of important agenda items, analysis of emotion data, summary generation, and playback of audio data. The following describes specific embodiments of this system.

[1367] Acquisition of audio data

[1368] The server activates the audio capture module as soon as the teleconference begins, collecting audio data in real time. The audio data is recorded in an appropriate format (e.g., WAV or MP3). This recorded audio data is later used for transcription and sentiment analysis.

[1369] Transcript

[1370] The server converts the acquired audio data into text data in real time via a speech recognition API. This transcribed data is stored on the server and used later for summarization and sentiment analysis.

[1371] Analysis of emotional data

[1372] The server uses an emotion engine to analyze the user's emotions in real time from voice and text data. This includes the ability to analyze voice tone and text content. For example, the user's emotions may be visualized as "tension" or "excitement."

[1373] Extraction of important agenda items

[1374] The server analyzes the text using natural language processing (NLP) techniques based on the analyzed character and sentiment data, and extracts key agenda items. This includes keyword extraction and text classification that take sentiment data into account.

[1375] Summary generation and display

[1376] The server integrates the extracted meeting minutes with sentiment data to generate a summary, and also identifies and saves the specific statements that support it. For example, parts of the sentiment data that show high levels of excitement are given particular attention.

[1377] Providing a user interface

[1378] The device provides a user-friendly interface. The web portal displays full text and sentiment data along with a summary, and users can click on specific meeting items to play audio data related to those points. It also generates PDF meeting minutes in a user-friendly format.

[1379] Specific example

[1380] 1. The user starts a teleconference.

[1381] "Teleconference started. Audio is being captured."

[1382] 2. The server captures the audio and converts it into text data in real time.

[1383] "We are converting audio data to text in real time."

[1384] 3. The server uses an emotion engine to analyze the user's emotions.

[1385] "Analyzing emotion data. Identifying user emotions."

[1386] 4. The server analyzes text data and sentiment data using natural language processing technology to extract important agenda items.

[1387] "Analysis in progress. Extracting key agenda items."

[1388] 5. The device provides a user interface, allowing users to view summaries and sentiment data.

[1389] "You can view key points and sentiment data. Related comments are also displayed."

[1390] 6. When a user reviews summaries and sentiment data and clicks on a statement related to a specific point, audio data related to that point is played.

[1391] "Let's review the statement in point 2."

[1392] "Playing the statement." (Audio data plays)

[1393] This system efficiently and accurately records the content of teleconferences and effectively provides users with key points, thereby improving work efficiency. Furthermore, by utilizing sentiment data, it makes it easier to understand the atmosphere and urgency of the meeting.

[1394] The following describes the processing flow.

[1395] Step 1:

[1396] The user initiates a teleconference. Once the teleconference starts, audio data capture begins automatically.

[1397] User: "Starting teleconference."

[1398] Step 2:

[1399] The server activates the audio capture module and collects audio data from the teleconference in real time. The audio data is saved in an appropriate format.

[1400] Server: "We are collecting audio data in real time."

[1401] Step 3:

[1402] The server collects audio data and sends it to a speech recognition API in real time, where it converts the audio into text data. This text data is then stored sequentially on the server.

[1403] Server: "Converting audio data to text data."

[1404] Step 4:

[1405] The server uses an emotion engine to analyze collected audio and text data in real time to recognize the user's emotions. The emotion engine analyzes the tone of voice and the content of the text.

[1406] Server: "Analyzing emotion data."

[1407] Step 5:

[1408] The server uses natural language processing (NLP) techniques to extract key agenda items based on the analyzed text and sentiment data.

[1409] Server: "We are analyzing text and sentiment data to extract key agenda items."

[1410] Step 6:

[1411] The server extracts the agenda items, organizes them into a summary, and saves them along with the corresponding specific statements. Sentiment data is also reflected in the summary.

[1412] Server: "We are generating summaries and tracking specific statements and sentiment data."

[1413] Step 7:

[1414] The device displays the generated summary and sentiment data through the user interface, allowing users to easily review key points.

[1415] Terminal: "Preparing the interface for summary and sentiment data."

[1416] Step 8:

[1417] Users can review the displayed summary and sentiment data. Selecting a specific agenda item will display comments and sentiment data related to that point.

[1418] User: "I'd like to confirm the specific points."

[1419] Terminal: "Displays selected points, related statements, and sentiment data."

[1420] Step 9:

[1421] When a user plays audio data related to a specific point, the relevant portion of the audio data will be played.

[1422] User: "Playing audio data related to the main points."

[1423] Terminal: "Related audio data will be played." (Audio data playback)

[1424] In this way, the system efficiently and accurately records the content of teleconferences through a series of processing steps, effectively providing users with key points and sentiment data. Users can review text and audio simultaneously, grasping the flow, nuances, and emotional changes of the conversation.

[1425] (Example 2)

[1426] 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".

[1427] Traditional teleconferencing systems struggled to efficiently record meeting content and generate key agenda items and summaries. Furthermore, they lacked sufficient means to analyze user sentiment data and understand the meeting atmosphere, making minute-taking cumbersome. Additionally, quickly accessing important statements during meetings was difficult.

[1428] The identification processing performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing the user's emotions from the audio data and text data, means for extracting important agenda items based on the analyzed text data and emotion data, means for integrating the extracted agenda items and emotion data to generate a summary, and means for providing the user with the generated summary and corresponding audio data. This enables efficient and accurate recording of meeting content, extraction of important agenda items, and understanding of the overall atmosphere of the meeting.

[1429] "Audio data" refers to data obtained by digitizing audio signals from teleconferences, etc.

[1430] "Real-time" refers to a state where data acquisition and processing are performed instantly and without delay.

[1431] "Text data" refers to information in text format generated from audio data.

[1432] "Emotional data" refers to data that indicates a user's emotional state, analyzed from voice and text.

[1433] "Meeting items" refer to the main topics or themes discussed during the teleconference.

[1434] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[1435] A "summary" is a text that concisely summarizes the specific content and key points.

[1436] A "server" is a computer system that processes data and provides various services to users.

[1437] A "user interface" is the environment that includes screens and input methods for a user to interact with a system.

[1438] A "voice capture module" is a device and software for collecting voice data and storing it as digital data.

[1439] A "database" is a system for systematically storing, searching, and editing information.

[1440] This invention relates to a system for efficiently recording the content of teleconferences, generating summaries, and creating meeting minutes. In particular, by combining it with an emotion engine that recognizes user emotions, it becomes possible to create richer meeting minutes. This system has a series of functions including real-time acquisition of audio data, transcription, extraction of important agenda items, analysis of emotion data, summary generation, and playback of audio data.

[1441] Hardware and software used

[1442] 1. Server

[1443] Voice capture module: Collects voice data in real time.

[1444] Speech Recognition API: Using "Google Speech-to-Text" as an example, we will convert speech data into text data in real time.

[1445] Emotion Engine: Using "Affectiva" as an example, we analyze the user's emotions from audio and text data.

[1446] Natural language processing technology: As an example, we will use "spaCy" to extract important agenda items based on text data and sentiment data.

[1447] Database: Stores analysis results and generated summaries.

[1448] 2. Terminal

[1449] User Interface: Displays summaries, sentiment data, and other information to the user using a web browser or dedicated application.

[1450] PDF generation tool: Used to output meeting minutes in PDF format.

[1451] Explanation of the program's processing flow

[1452] The server activates the audio capture module as soon as the teleconference begins, collecting audio data in real time. The collected audio data is recorded in WAV or MP3 format and used later for transcription and sentiment analysis.

[1453] Next, the server converts the audio data into text data in real time using a speech recognition API (Google Speech-to-Text). This text data is stored in a database. The text data generated through transcription is then used for subsequent summarization and sentiment analysis.

[1454] The server uses an emotion engine (Affectiva) to analyze the user's emotions in real time from voice and text data. Emotion analysis includes analyzing the tone of voice and the content of the text. For example, emotional data such as "tension" or "excitement" may be identified.

[1455] Furthermore, the server uses natural language processing technology (spaCy) to analyze text and sentiment data. It extracts important agenda items and performs keyword extraction and text classification that takes sentiment data into account. The extracted agenda items are used to generate summaries.

[1456] Subsequently, the server integrates the extracted meeting minutes and sentiment data to generate a summary. The generated summary also identifies and stores the specific statements that support it. In particular, sections where the sentiment data indicates high levels of excitement are given special attention in the summary.

[1457] Finally, the device provides a user interface, allowing users to view summaries, full text, and sentiment data. Clicking on a specific meeting item plays audio data related to that point. The user interface also generates easy-to-read PDF meeting minutes.

[1458] Specific example

[1459] 1. The user starts a teleconference.

[1460] Example: The user clicks the "Start Teleconference" button.

[1461] Result: The teleconference begins, and the server starts audio capture.

[1462] 2. The server captures the audio and converts it into text data in real time.

[1463] Example: The server batches audio data every second and converts it into text data.

[1464] Result: Text data is generated as the teleconference progresses.

[1465] 3. The server uses an emotion engine to analyze the user's emotions.

[1466] Example: The server saves analysis results such as "the user's voice tone is rising" during the analysis.

[1467] Result: Emotional data is generated, and a timestamp corresponding to a specific emotion is recorded.

[1468] 4. The server analyzes text data and sentiment data using natural language processing technology to extract important agenda items.

[1469] Example: Keyphrase extraction and TextRank algorithms are used to extract important agenda items.

[1470] Result: Important meeting items are listed and saved in the database.

[1471] 5. The device provides a user interface, allowing users to view summaries and sentiment data.

[1472] Example: When a user accesses a web portal, they can click the "View Summary" button.

[1473] Results: A summary, key agenda items, sentiment data, and relevant comments are displayed.

[1474] Example of a prompt

[1475] 1. Record the start of the teleconference in real time.

[1476] 2. Convert the audio data to text and save it.

[1477] 3. Perform sentiment analysis from the text and audio data.

[1478] 4. Analyze the sentiment data and text data to extract the important agenda items.

[1479] 5. Generate a summary and display it in the user interface.

[1480] This invention makes it possible to efficiently record the content of teleconferences, extract important agenda items, and generate summaries. Furthermore, by analyzing sentiment data, it becomes easier to understand the atmosphere of the meeting and the emotions of the participants.

[1481] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1482] Step 1:

[1483] The server activates the audio capture module when the teleconference begins. The audio capture module collects audio data in real time and records it in WAV or MP3 format. The input is the audio signal from the teleconference, which is captured as digital audio data. The output is the recorded audio data. Specifically, the audio capture module is activated, the audio data is sent to the server in real time, and it is saved in the appropriate format.

[1484] Step 2:

[1485] The server converts acquired audio data into text data in real time. It analyzes the audio data using a speech recognition API (e.g., Google Speech-to-Text). The input is the audio data recorded in step 1, and the data processing involves converting the audio data into text data. The output is the converted text data. Specifically, the audio data is sent to the speech recognition API, and the text data returned by the API is saved to the server.

[1486] Step 3:

[1487] The server uses an emotion engine (e.g., Affectiva) to analyze the user's emotions in real time from audio and text data. The input consists of audio and text data, and the data processing involves speech tone analysis and text analysis. The output is the analyzed emotion data. Specifically, the emotion engine analyzes the tone of the audio and the content of the text, and creates a file that identifies emotions such as "tension" or "excitement."

[1488] Step 4:

[1489] The server extracts important agenda items based on text data and sentiment data using natural language processing techniques (e.g., spaCy). The input consists of text data and sentiment data, and data processing includes keyword extraction and text classification. The output is the extracted important agenda items. Specifically, it uses natural language processing techniques to analyze text data and list agenda items while considering sentiment data.

[1490] Step 5:

[1491] The server integrates extracted meeting minutes and sentiment data to generate a summary. The input consists of important meeting minutes and sentiment data, and the data processing involves calculations that integrate important statements and sentiment highlights. The output is the generated summary. Specifically, the extracted meeting minutes and sentiment data are saved to a database, and an algorithm runs to generate the summary.

[1492] Step 6:

[1493] The device provides a user interface, allowing users to view summaries, full text, and sentiment data. The input is the summary and corresponding audio data generated in step 5, and the output is a visually displayed summary and playable audio data. Specifically, when a user accesses the web portal and clicks the "View Summary" button, the summary, key minutes, sentiment data, and relevant comments are displayed.

[1494] Step 7:

[1495] When a user reviews summaries and sentiment data and clicks on a statement related to a specific point, the audio data related to that point is played. The input is the summary section and associated audio data specified by the user, and the output is the played audio data. Specifically, when a user clicks on a summary section in the web portal, the audio data corresponding to that statement is played, allowing them to listen to the details of the meeting content.

[1496] (Application Example 2)

[1497] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1498] Meetings and work instructions in factories require efficient and accurate recording. Traditional methods involved the cumbersome process of manually converting audio data into text and generating summaries. Furthermore, meeting minutes were created without incorporating emotional data from the meeting, making it difficult to issue work instructions that considered the atmosphere and urgency of the meeting. Therefore, a system was needed that provided a seamless process for real-time audio data acquisition, conversion to text, emotional data analysis, and automated work instruction generation.

[1499] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1500] In this invention, the server includes means for acquiring audio data, means for converting the acquired audio data into text data in real time, means for analyzing and extracting important agenda items from the text data, means for displaying the extracted important agenda items as a summary along with their rationale, means for providing the user with the summary and corresponding audio data and sentiment data, means for allowing the user to confirm important agenda items and summaries through a user interface, means for analyzing sentiment data to identify the user's emotions, and means for automatically generating work instructions based on the summary. This makes it possible to efficiently and accurately record the contents of meetings and automatically generate work instructions using sentiment data.

[1501] "Audio data" refers to sound information acquired from microphones and other audio input devices.

[1502] "Text data" refers to digital information obtained by converting audio data into text format.

[1503] "Important agenda items" refer to points or decisions that deserve particular attention during a meeting or discussion.

[1504] "Emotional data" refers to information that indicates the speaker's emotional state, and is analyzed from the tone of voice and the content of the text.

[1505] A "summary" is a text that concisely summarizes detailed information.

[1506] A "user interface" refers to the screens and means of operation that a user uses to interact with a system.

[1507] A "server" is a computer system that provides services over a network.

[1508] "Natural language processing technology" refers to the technology that enables computers to understand and generate human language.

[1509] A "work instruction sheet" is a document that contains instructions for performing a specific task.

[1510] This invention relates to a system for efficiently recording meetings and work instructions conducted within a factory, analyzing emotional data, and generating audio summaries and work instructions. Specifically, it performs real-time acquisition of audio data, conversion of text data, extraction of important agenda items, analysis of emotional data, summary generation, and automatic generation of work instructions.

[1511] The server uses a microphone to acquire audio data and activates an audio capture module to collect audio data in real time. The acquired audio data is recorded in an appropriate format (e.g., WAV or MP3). This recorded audio data is then used for transcription and sentiment analysis.

[1512] The server converts audio data acquired via a speech recognition API into text data in real time. This text data is stored on the server and used later for summarization and sentiment analysis.

[1513] The server uses an emotion engine to analyze the user's emotions in real time from voice and text data. This includes the ability to analyze voice tone and text content. For example, the user's emotions may be identified as "tension" or "excitement."

[1514] The server analyzes the text using natural language processing (NLP) techniques based on the analyzed character and sentiment data, and extracts key agenda items. This process includes keyword extraction and text classification that take sentiment data into account.

[1515] The server integrates the extracted key meeting minutes with sentiment data to generate a summary, and also identifies and stores the specific statements that support it. For example, parts of the sentiment data showing high levels of excitement are given particular attention.

[1516] The device provides a user-friendly interface. The web portal displays full text and sentiment data along with a summary, and users can click on specific meeting items to play audio data related to those points. It also generates PDF meeting minutes in a user-friendly format.

[1517] Specific example:

[1518] If a user says, "Let's discuss the new production line," during a meeting, the system records the statement, analyzes whether the sentiment is positive, summarizes the key points, and incorporates them into the work instructions.

[1519] Example of a prompt:

[1520] Audio data: "We will discuss the new production line."

[1521] Emotional data: "Positive"

[1522] Summary generated: "Discussions were held regarding a new production line."

[1523] This invention significantly improves operational efficiency within a factory by efficiently and accurately recording meeting content and automatically generating work instructions using emotional data.

[1524] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1525] Step 1:

[1526] The server activates the means to acquire audio data and collects audio data in real time through the microphone. The audio data is recorded in an appropriate format such as WAV or MP3.

[1527] Input: Audio data acquired from the microphone

[1528] Output: Recorded audio file

[1529] Specific operation: The microphone picks up sound, the audio data is sent to the server in real time, and saved in the appropriate file format.

[1530] Step 2:

[1531] The server converts the recorded audio data into text data in real time using a speech recognition API. This converted text data is stored on the server.

[1532] Input: Audio file

[1533] Output: Character data (text format)

[1534] Specific operation: The speech recognition API analyzes the audio file, converts the audio to text, and the generated text data is saved to storage.

[1535] Step 3:

[1536] The server uses an emotion engine to analyze the user's emotions in real time from text and audio data. The analyzed emotion data is stored on the server.

[1537] Input: Text data, audio data

[1538] Output: Sentiment data

[1539] Specific operation: The emotion engine analyzes text data and voice tone, tags the detected emotions as "positive," "negative," "tense," etc., and stores them in a database.

[1540] Step 4:

[1541] The server uses natural language processing techniques to extract key agenda items from text and sentiment data. These extracted agenda items are stored along with the detected sentiment data.

[1542] Input: Text data, sentiment data

[1543] Output: Key Agenda Items

[1544] Specific operation: A natural language processing algorithm analyzes the text, extracts important keywords and phrases, and stores them in storage along with sentiment data.

[1545] Step 5:

[1546] The server generates a summary based on the extracted key meeting items and sentiment data. This summary also identifies and stores the specific statements that supported it.

[1547] Input: Key meeting items, sentiment data

[1548] Output: Summary text

[1549] Specific operation: A natural language generation model takes important meeting items and sentiment data as input, generates a concise summary based on that, and saves specific parts of the statements.

[1550] Step 6:

[1551] The device displays summaries, full text, and sentiment data through a user interface. When the user clicks on a specific meeting item, it plays audio data related to that point. It also generates meeting minutes in PDF format.

[1552] Input: Summary text, full text, sentiment data, audio data

[1553] Output: Displayed summary, full text, sentiment data, and PDF meeting minutes.

[1554] Specific operation: The user interface displays a summary text and related data, and clicking a link plays the audio data. Additionally, a PDF generator produces meeting minutes.

[1555] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1556] 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.

[1557] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1558] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1559] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1560] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1561] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1562] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1563] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1564] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1565] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1566] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1567] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1568] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1569] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1570] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1571] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1572] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1573] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1574] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1575] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[1576] The following is further disclosed regarding the embodiments described above.

[1577] (Claim 1)

[1578] Means for acquiring audio data,

[1579] A means of converting acquired audio data into text data in real time,

[1580] A method for analyzing and extracting important agenda items from text data,

[1581] A means of displaying the extracted important agenda items as a summary along with their supporting evidence,

[1582] A system that includes means for providing users with summaries and corresponding audio data.

[1583] (Claim 2)

[1584] The system according to claim 1, which uses natural language processing technology for the analysis and extraction of character data.

[1585] (Claim 3)

[1586] The system according to claim 1, further comprising means for playing audio data related to a summary portion specified by the user.

[1587] "Example 1"

[1588] (Claim 1)

[1589] A means of detecting the start of a teleconference,

[1590] A means of recording audio data in real time,

[1591] A means of converting recorded audio data into text data in real time,

[1592] A means of analyzing and extracting important agenda items from converted text data using natural language processing technology,

[1593] A means of displaying the extracted important agenda items as a summary along with their supporting evidence,

[1594] A means of providing the user with a summary and corresponding audio data,

[1595] Means for providing a user interface

[1596] A system that includes this.

[1597] (Claim 2)

[1598] The system according to claim 1, further comprising means for activating an audio capture module and acquiring audio data.

[1599] (Claim 3)

[1600] The system according to claim 1, further comprising means for playing audio data related to a summary portion specified by the user.

[1601] "Application Example 1"

[1602] (Claim 1)

[1603] Means for acquiring audio data,

[1604] A means of converting acquired audio data into text data in real time,

[1605] A method for analyzing and extracting important agenda items from text data,

[1606] A means of displaying the extracted important agenda items as a summary along with their supporting evidence,

[1607] A means of providing the user with a summary and corresponding audio data,

[1608] A means of recording the contents of meetings and conferences within the factory and automatically generating summaries,

[1609] A means for generating a meeting summary based on the extracted key agenda items,

[1610] A system that includes means for displaying a summary on a user interface.

[1611] (Claim 2)

[1612] The system according to claim 1, which uses natural language processing technology for the analysis and extraction of character data.

[1613] (Claim 3)

[1614] The system according to claim 1, further comprising means for playing audio data related to a summary portion specified by the user.

[1615] "Example 2 of combining an emotion engine"

[1616] (Claim 1)

[1617] Means for acquiring audio data,

[1618] A means of converting acquired audio data into text data in real time,

[1619] A means for analyzing user emotions from audio and text data,

[1620] A means of extracting important agenda items based on analyzed text data and sentiment data,

[1621] A means of integrating extracted meeting minutes and sentiment data to generate a summary,

[1622] A system including means for providing the user with a generated summary and corresponding audio data.

[1623] (Claim 2)

[1624] The system according to claim 1, which uses natural language processing technology for analyzing text data and audio data.

[1625] (Claim 3)

[1626] The system according to claim 1, further comprising means for playing audio data related to a summary portion specified by the user.

[1627] "Application example 2 when combining with an emotional engine"

[1628] (Claim 1)

[1629] Means for acquiring audio data,

[1630] A means of converting acquired audio data into text data in real time,

[1631] A method for analyzing and extracting important agenda items from text data,

[1632] A means of displaying the extracted important agenda items as a summary along with their supporting evidence,

[1633] A means of providing the user with a summary and corresponding audio and sentiment data,

[1634] A means to review important meeting items and summaries through the user interface,

[1635] A means of identifying a user's emotions by analyzing emotional data,

[1636] A system that includes means for automatically generating work instructions based on a summary.

[1637] (Claim 2)

[1638] The system according to claim 1, which uses natural language processing technology for the analysis and extraction of text data and sentiment data.

[1639] (Claim 3)

[1640] The system according to claim 1, further comprising means for playing audio data related to a summary portion specified by the user. [Explanation of Symbols]

[1641] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. Means for acquiring audio data, A means of converting acquired audio data into text data in real time, A method for analyzing and extracting important agenda items from text data, A means of displaying the extracted important agenda items as a summary along with their supporting evidence, A system that includes means for providing users with summaries and corresponding audio data.

2. The system according to claim 1, which uses natural language processing technology for the analysis and extraction of character data.

3. The system according to claim 1, further comprising means for playing audio data related to a summary portion specified by the user.

Images