system

Abstract

We provide the system. [Solution] A device that acquires speech as audio information, A device that instantly converts acquired audio information into text information, A device that analyzes converted text information to identify the background of the discussion and the intentions of the participants, A device that generates notifications to improve progress based on analysis results, A device that immediately provides the generated notification, A device that, after completion, compiles the analysis results and creates a record document and action plan, A system that includes a device for sharing instant notifications on information terminals.

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] In meetings and discussions, various opinions are flying around, and the content of the discussion often becomes complicated. As a result, problems such as the lack of enabling due to insufficient understanding among participants and the inability to obtain a clear action plan after the meeting have occurred. In addition, it is difficult to identify the speaker and the intention of the discussion from voice data, and there is often a problem that the minutes are inaccurate. It is necessary to solve these problems and improve the efficiency and effectiveness of the meeting.

Means for Solving the Problems

[0005] This invention provides a means for accurately and quickly recording the content of a meeting by acquiring speech during the meeting as audio data and converting it into text data in real time. By analyzing the acquired text data and identifying the context of the discussion and the intentions of the participants, feedback is generated to facilitate the smooth progress of the meeting. Furthermore, by identifying the speaker and classifying each statement, more detailed and accurate meeting minutes are created. The feedback and action plans generated from the analysis results are stored in cloud storage and made easily accessible to users, enabling efficient follow-up after the meeting.

[0006] "Audio data" refers to recordings of speeches made during meetings or discussions in digital format.

[0007] "Text data" refers to audio data converted into a string format that can be analyzed.

[0008] "Analysis" is the process of identifying the context of a discussion and the intentions of the participants based on text data.

[0009] "Feedback" refers to instructions and information generated based on analysis results to improve the progress of a meeting.

[0010] Meeting minutes are documents that summarize, organize, and record the content of discussions that took place during a meeting.

[0011] An "action plan" is a specific action plan set out based on the results of a meeting.

[0012] "Cloud storage" is a data storage system that saves analysis results, generated meeting minutes, and action plans, making them accessible via the internet. [Brief explanation of the drawing]

[0013] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] 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 Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Mode for Carrying Out the Invention

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

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

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

[0017] 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.

[0018] 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.

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

[0020] 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."

[0021] [First Embodiment]

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

[0023] 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.

[0024] 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).

[0025] 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.

[0026] 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.

[0027] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form 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.

[0028] 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.

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

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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".

[0034] This invention provides technology for conducting meetings and discussions more efficiently, based on an AI system for analyzing discussion intent. Specifically, it accurately grasps the content of meetings and generates feedback by converting audio during meetings into text in real time and analyzing the context of the discussion.

[0035] Acquisition and conversion of audio data

[0036] When a meeting begins, the terminal uses its built-in microphone to acquire audio. The acquired audio data is compressed using compression technology and sent to the server via the internet.

[0037] The server uses a speech recognition engine to convert speech data into text data. It also implements processes that enhance pronunciation clarity by utilizing noise reduction and speech enhancement technologies.

[0038] Text data analysis

[0039] The server analyzes text data using natural language processing techniques. A contextual analysis engine identifies the speaker's intent and the topic of discussion, and extracts important keywords.

[0040] To classify each utterance by speaker, speaker identification technology is applied. This is achieved using a model that learns unique linguistic patterns contained in the text.

[0041] Generating feedback and instructions

[0042] Based on the analysis results, the server generates feedback and specific instructions to improve the meeting's progress. This may include changing topics that have been discussed too much or identifying unresolved issues.

[0043] The generated feedback is immediately sent to the device and presented to the user visually or audibly.

[0044] Creating meeting minutes and action plans

[0045] After the meeting concludes, the server integrates the accumulated data to create reviewable meeting minutes and a clear action plan. This includes summaries of each topic discussed and conclusions, as well as specific actions recommended as next steps.

[0046] The completed meeting minutes and action plan are saved to cloud storage, making them accessible to users and shareable with other participants as needed.

[0047] Specific example

[0048] For example, in a product development meeting at a certain company, this system transcribes participants' comments in real time, identifying topics such as changes in product specifications, development schedules, and marketing strategies. If a comment such as "The release date of this product may be delayed" is made during the meeting, the system picks up on this comment and presents it as feedback suggesting a review of the milestones. After the meeting, meeting minutes accessible to all participants are provided on the cloud. This allows participants to review the tasks and actions they need to prepare for the next meeting.

[0049] The following describes the processing flow.

[0050] Step 1:

[0051] The device captures meeting audio in real time using its built-in microphone. The audio data is compressed and sent to the server over the network. A noise reduction algorithm is applied to optimize audio quality.

[0052] Step 2:

[0053] The server inputs the received audio data into a speech recognition engine and converts it into text data. Deep learning technology is used for speech recognition, and the speaker's accent and vocabulary are analyzed to improve accuracy.

[0054] Step 3:

[0055] The server uses a natural language processing engine to analyze the generated text data. This involves contextual analysis, keyword extraction, and sentiment analysis to identify the intent behind participants' statements and the topics of discussion.

[0056] Step 4:

[0057] Based on the results of the text analysis, the server generates feedback and instructions for the next steps to improve the progress of the meeting. The generated information is immediately sent to the terminal and presented to meeting participants visually or audibly.

[0058] Step 5:

[0059] The terminal displays feedback received from the server in the user interface. Users use this information to make decisions that will advance the discussion.

[0060] Step 6:

[0061] After the meeting concludes, the server integrates all analytical data to create detailed meeting minutes and an action plan for the next meeting. This includes a summary of the discussion, key conclusions, and recommended actions.

[0062] Step 7:

[0063] The final meeting minutes and action plan will be saved to cloud storage and made accessible to users. Users will use this information to prepare for and follow up on future meetings.

[0064] (Example 1)

[0065] 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."

[0066] Meetings and discussions often face challenges such as misunderstandings of what is said and the intentions behind it, leading to stalled progress. Furthermore, it is difficult to quickly and accurately create meeting minutes and action plans afterward. This can result in insufficient information sharing among participants and delays in decision-making.

[0067] 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.

[0068] In this invention, the server includes means for acquiring speech during a meeting as audio information, means for converting the acquired audio information into text information in real time, means for analyzing the converted text information to identify the context of the discussion and the intentions of the participants, means for applying a noise reduction filter and voice enhancement to the audio information, means for generating instructions to improve the progress of the meeting based on the analysis results, means for immediately providing the generated instructions to the participants, and means for compiling the analysis results after the meeting to create a report and an action plan. This improves the efficiency of meetings and enables rapid and accurate information sharing and decision-making.

[0069] "Audio information" refers to data recorded in digital format from speeches such as those heard in meetings or conversations.

[0070] "Acquiring" refers to the act of collecting audio information from a device.

[0071] "Textual information" refers to string data that has been converted from audio information.

[0072] "Real-time" means that processing and conversion occur simultaneously with the utterance.

[0073] "Analysis" refers to the process of processing information based on converted textual data to understand its context and intent.

[0074] "Context" refers to the framework that gives meaning and relevance within a statement or dialogue.

[0075] A "noise reduction filter" refers to a technology that reduces unwanted background noise from audio information.

[0076] "Voice enhancement" refers to the process of adjusting sound to make necessary sounds in audio information easier to hear.

[0077] "Instructions" refers to specific suggestions or commands generated based on the analysis results to support the progress of the meeting.

[0078] A "report" refers to a document that summarizes the content of discussions, conclusions, and action plans during a meeting.

[0079] An "action plan" refers to specific guidelines or procedures developed based on the matters discussed after the meeting.

[0080] "Participants" refers to those who speak or participate in a meeting or discussion.

[0081] A specific description will be given of embodiments for carrying out this invention.

[0082] This discussion intent analysis AI system aims to facilitate efficient meeting management and record-keeping. First, the terminal uses its built-in microphone to capture speech during the meeting as audio information. The audio information is converted into a digital format and processed efficiently through data compression technology. This data is then transmitted to a server via the internet.

[0083] The server uses an advanced speech recognition engine to convert speech information into text in real time. This conversion process is enhanced with noise reduction filters and voice enhancement techniques to improve accuracy. The converted text is time-stamped to maintain the order of speech.

[0084] Next, the server uses natural language processing (NLP) techniques to analyze the textual information. This analysis identifies the context of the discussion and the intentions of the speakers. Machine learning models are also used to identify speakers and classify each statement by speaker. This enables detailed tracking and analysis of the discussion.

[0085] Based on the analysis results, the server generates instructions to support the progress of the meeting. These instructions are presented to the user from the terminal as digital or audio output, providing real-time feedback.

[0086] After the meeting ends, the server automatically generates a report and action plan based on the data accumulated so far. The generated report is saved to cloud storage, making it easily accessible to the user and shareable with other participants.

[0087] As a concrete example, in a product development meeting, this system immediately picks up on a statement such as "The product launch date may be delayed" and instructs the system to re-evaluate milestones as appropriate feedback. An example of a prompt to be input into the generating AI model is, "Transcribe the statements made in this meeting in real time, analyze the speaker's intent, and generate feedback."

[0088] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0089] Step 1:

[0090] The device acquires audio during a meeting via its built-in microphone. This audio information is converted into a digital signal, and the data size is reduced using compression technology (e.g., MP3). The input is raw audio, and the output is compressed digital audio data.

[0091] Step 2:

[0092] Compressed audio data is transmitted to a server via the internet. The server utilizes a speech recognition engine to convert this data into text. The input is compressed digital audio data, and the output is text information with a timestamp. Noise reduction and speech enhancement algorithms are applied to optimize sound quality.

[0093] Step 3:

[0094] The server analyzes the converted text information using natural language processing (NLP) techniques. It performs contextual analysis and keyword extraction to identify participants' intentions and discussion topics. The input is text information, and the output is contextual data as a result of the analysis, extracted keywords, and speaker identification information.

[0095] Step 4:

[0096] Based on the analysis results, the server generates instructions to improve the meeting progress. This involves intent analysis using prompt sentences and utilizing a generative AI model to create specific feedback. The input is the analysis results from step 3, and the output is instructions and feedback presented in real time.

[0097] Step 5:

[0098] The generated instructions and feedback are provided to the user through the device. These are presented either on a screen or verbally using speech synthesis technology. The input is the instructions from step 4, and the output is the visual or auditory information received by the user.

[0099] Step 6:

[0100] After the meeting concludes, the server integrates all data to create a report and action plan. This process includes summarizing each generated instruction and statement. Input is data from all sessions, and output is the report and action plan as documents.

[0101] Step 7:

[0102] Finally, the generated report and execution plan are saved to cloud storage and configured to be accessible to the user. This process is for saving the output of step 6 to the cloud.

[0103] (Application Example 1)

[0104] 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."

[0105] A challenge arises when information sharing among staff at a store is not conducted quickly and accurately, as this can hinder customer service and operational efficiency. This problem stems particularly from the inability to respond quickly to real-time changes in customer requests and operational tasks, potentially resulting in a decline in service quality.

[0106] 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.

[0107] In this invention, the server includes a device for acquiring speech as audio information, a device for immediately converting the acquired audio information into text information, and a device for analyzing the converted text information to identify the background of the discussion and the intentions of the participants. This enables immediate information sharing among staff, real-time acquisition of information necessary for customer service and business operations, and efficient business operations.

[0108] "Speech" refers to linguistic expression used by people to convey intentions and information through sound.

[0109] "Audio information" refers to data obtained by acquiring and processing sound as a digital signal, making it analyzable.

[0110] "Textual information" refers to text data converted from audio information, and is used for analysis.

[0111] A "device" is a system of machinery or software designed to perform a specific function.

[0112] "The background of the discussion" refers to the context and situation in which participants speak, and it serves as the foundation for understanding their intentions.

[0113] "Instantly" means that processing or results can be obtained in near real-time with no delay.

[0114] "Analysis" is the process of organizing and breaking down information to understand its structure and meaning.

[0115] An "information terminal" is an electronic device used to display and manipulate data.

[0116] A "record document" is a formal written document in which information is organized and preserved for later reference.

[0117] An "action plan" is a detailed document outlining specific steps and methods for achieving a particular objective.

[0118] "Memory space" refers to digital space used to store data and information.

[0119] The system that implements this application consists of multiple components. The server uses a combination of appropriate hardware and software to instantly acquire and process audio information. Specifically, it uses a mobile device such as a smartphone or tablet to acquire audio, utilizing the microphone built into it. The device compresses the acquired audio information and transmits it to the server via the network.

[0120] The server uses an advanced speech recognition engine to convert audio information into text. This process leverages the Google® Cloud Speech-to-Text API, improving recognition accuracy through noise reduction and speech enhancement techniques. The server then analyzes the converted text and uses natural language processing techniques to identify the context of the discussion and the participants' intentions. This utilizes natural language processing libraries such as spaCy.

[0121] The analysis results are provided to the user as real-time feedback via an information terminal. Users can use this interface to immediately check the information and take necessary actions. After completion, the analysis data is compiled into a record document and action plan, and saved to cloud storage. Users can access and share this information at any time via their digital devices. Google Cloud Platform is used for the cloud service.

[0122] As a concrete example, consider information sharing regarding the arrival date of new products at a store. Using this system, a voice command such as "Check when the new products will arrive and inform everyone" would be instantly analyzed, and the results would be shared with the staff. This would allow each staff member to quickly prepare countermeasures and improve the quality of service provided to customers.

[0123] An example of a prompt for a generative AI model is: "Explain how to build a system that extracts important business tasks from store staff conversations and provides real-time feedback."

[0124] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0125] Step 1:

[0126] The device acquires conversational audio information using a microphone. The audio data, as input, is captured directly from the microphone as raw sound. The device then uses a compression algorithm to reduce the size of the audio data and convert it into a format suitable for transfer to the server. This process reduces the data size while maintaining the quality of the audio signal.

[0127] Step 2:

[0128] The server receives audio data sent from the terminal. The input is compressed audio data. The server uses the Google Cloud Speech-to-Text API to convert the audio data into text. During the conversion process, it uses techniques to suppress noise and emphasize important audio components. The output is text data that reflects the content of the discussion.

[0129] Step 3:

[0130] The server analyzes the text data. The input is the text data generated in step 2. It utilizes natural language processing techniques, specifically the spaCy library, to analyze the background of the discussion and the intentions of the participants. It extracts keywords and important contexts and structures the information based on them. The output is metadata of the analysis results, i.e., a set of identified intentions and keywords.

[0131] Step 4:

[0132] The server generates feedback based on the analysis results. The input is the metadata obtained in step 3. This generates effective feedback to improve the meeting's progress. Using a generation AI model, specific improvement suggestions and points are created, and the results are immediately transmitted to the information terminal. The output is the proposed feedback message and points.

[0133] Step 5:

[0134] Users receive this feedback through their information terminals and use it in real time to aid in decision-making. Users can modify their actions based on the feedback as needed, streamlining their work processes. Input is feedback messages, and output is improvement measures and corresponding actions.

[0135] Step 6:

[0136] The server generates a record document and an action plan based on the data analyzed after completion. The input is the data processed in all previous steps. The record document includes an overview of the meeting, a summary of the discussions, and decisions made, while the action plan outlines specific next steps and responsibilities. The output is the saved record document and action plan.

[0137] Step 7:

[0138] Users access record documents and action plans stored in cloud storage, and individually review and share them. Google Cloud Platform is used as the cloud service. Inputs are record documents and action plans, and outputs are the standardization of work through checking and sharing of verified information.

[0139] 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.

[0140] This invention provides a system characterized by its ability to analyze intentions and provide feedback during discussions, as well as to recognize participants' emotions. By incorporating an emotion engine, it enables real-time analysis of user emotions from speech during meetings and allows for the use of the results to improve the progress of the meeting.

[0141] Acquisition of voice data and sentiment analysis

[0142] The terminal begins capturing audio in real time as soon as the meeting starts and transmits it to the server. The audio data is then passed to the emotion engine simultaneously with speech recognition.

[0143] The server converts the received audio data into text using a speech recognition engine and analyzes the emotions contained in the utterances using an emotion engine. It identifies emotions such as joy, anger, and sadness from individual utterances and stores them as data.

[0144] Text data analysis and feedback generation

[0145] The server analyzes text data using natural language processing techniques to identify the context of the discussion and the intentions of the participants. This analysis is then combined with sentiment data to generate feedback that optimizes the meeting's progress.

[0146] The analyzed emotional data is also included as an important factor in creating meeting minutes and developing action plans. In particular, in situations where discussions are heated or there is no agreement, emotional data plays a role in regulating the progress of the meeting.

[0147] Real-time emotion change notifications and meeting minute creation

[0148] The terminal displays feedback and sentiment analysis results sent from the server to the user in real time. This allows the user to make decisions based not only on the context of the discussion but also on the flow of emotions.

[0149] After the meeting concludes, the server integrates all data to create meeting minutes and an action plan. Sentimental data is used as an indicator to assess the impact of each statement, contributing to an objective meeting record.

[0150] Specific example

[0151] For example, in a meeting about the launch of a new product, if a participant expresses strong concerns about market reaction, this system identifies the emotions such as "anxiety" and "concern" contained in their statement. The analysis results in feedback such as "The team needs to discuss this concern further," which is immediately made public to other participants. Subsequently, at the end of the meeting, detailed minutes are created, including how the concerns were discussed and what was decided as the next steps. Sentimental data is an important indicator of which topics generated more emotional discussion.

[0152] The following describes the processing flow.

[0153] Step 1:

[0154] The terminal uses a microphone to capture audio in real time at the start of the meeting. The captured audio data is compressed and then sent to the server. Noise cancellation technology is applied to ensure audio quality.

[0155] Step 2:

[0156] The server converts the received audio data into text data using a speech recognition engine. The speech recognition uses a highly accurate deep learning model to handle a wide variety of speech patterns.

[0157] Step 3:

[0158] The server passes the audio data to the emotion engine, which analyzes the user's emotions from the conversation. It identifies the emotional tone in each utterance, such as joy, anger, or sadness, and stores this data along with other analysis data.

[0159] Step 4:

[0160] The server analyzes text data using natural language processing algorithms to identify the context of the discussion and the intentions of the participants. Sentiment data plays a crucial role in this analysis, deepening the understanding of how the discussion is progressing.

[0161] Step 5:

[0162] The server generates feedback based on the analysis results to improve the members' progress. This includes summaries of key points from the discussion and instructions on the direction of the meeting based on sentiment analysis.

[0163] Step 6:

[0164] The device displays generated feedback and real-time analyzed sentiment data to the user. This allows the user to understand the emotional state of the participants while facilitating the discussion.

[0165] Step 7:

[0166] After the meeting concludes, the server integrates all data to create detailed minutes and action plans. Sentimental data is used to assess the impact of the dialogue and show how the discussion evolved.

[0167] Step 8:

[0168] The completed meeting minutes and action plans are saved to cloud storage and made accessible to users as needed. This allows meeting participants to clearly define their next steps and follow up efficiently.

[0169] (Example 2)

[0170] 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".

[0171] Traditional meeting systems made it difficult to provide feedback that took participants' emotions into account, potentially influencing the progress of discussions and decision-making processes. Furthermore, it was challenging to quickly and accurately analyze individual statements and their intentions, and to effectively utilize emotional data. This resulted in decreased meeting efficiency and productivity.

[0172] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0173] In this invention, the server includes means for acquiring audio information during a meeting as data, means for converting the acquired audio information into text information in real time, means for analyzing the converted text information to identify the context of the discussion and the intentions of the participants, means for recognizing the emotions contained in the analyzed text information and storing them as data, means for generating feedback to optimize the progress of the meeting based on the analysis results, and means for compiling the emotion data and analysis results after the meeting to create a meeting record and an activity plan. This enables the provision of feedback that takes into account the emotions of the participants in real time, and makes it possible to grasp intentions and optimize the progress of the meeting using emotion data.

[0174] "Audio information" refers to a series of sound data collected during a meeting, used to record the progress of the meeting and the statements made.

[0175] "Textual information" refers to text data converted in real time based on audio information, and is used to visually recognize and analyze the content of meetings.

[0176] "Context" refers to the linguistic background information of statements made during a meeting, as well as the circumstances and flow in which those statements were made, and is of great importance to the progress of the discussion.

[0177] "Intention" refers to the purpose or aim that a speaker is trying to convey through their words, and it is a factor that influences the direction of the discussion.

[0178] "Emotions" are psychological states underpinned by words and actions, such as expressing feelings like joy, anger, and sadness.

[0179] "Feedback" refers to opinions and suggestions provided to participants based on analyzed data to optimize the flow of a meeting, and is important information for indicating areas for improvement.

[0180] "Meeting minutes" are documents that summarize the discussions, statements, and decisions made during a meeting, and are used as reference material for future reference.

[0181] An "activity plan" is a document that outlines specific future action plans and schedules based on the results of a meeting, and it supports the organization's strategic execution.

[0182] "Data" refers to facts and information that have been measured, collected, and analyzed, and includes numerical values, text, and other information that are processed and stored within a system.

[0183] This system optimizes meeting progress by analyzing audio during meetings in real time, interpreting participants' emotions and intentions from their statements, and providing feedback.

[0184] The terminal uses a built-in high-performance microphone to capture meeting audio information in real time. A dedicated application is installed that has the ability to continuously monitor audio input and acquire data. Audio data is transmitted to the server immediately using a low-latency protocol (e.g., WebSocket).

[0185] The server converts the received audio information into text using a speech recognition engine (e.g., a speech recognition API from a major cloud service provider). Furthermore, the server uses an emotion analysis engine and natural language processing technology (e.g., a natural language model API from a commercial AI provider) to identify emotions from the spoken content and to deeply analyze the context of the discussion and the intentions of the participants.

[0186] The analyzed information is used to generate feedback that takes into account the emotional data contained in each statement and the flow of the discussion. This feedback is provided to participants in real time and serves as an important indicator for the progress of the meeting. After the meeting, meeting records and activity plans are automatically generated based on the accumulated data and saved to shared network storage.

[0187] As a concrete example, in a meeting about the market strategy for a new product, if a participant expresses concerns about the product's design, the system analyzes the emotion of "anxiety" contained in their statement and provides feedback such as "further design evaluation is needed." This feedback is instantly displayed to the participants, allowing everyone to share their concerns.

[0188] An example of a prompt could be, "In a meeting about a new product launch, please tell me how to identify participants' anxieties and concerns and generate appropriate feedback." This system makes it easier for users to understand the flow of emotions and obtain information to make better decisions.

[0189] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0190] Step 1:

[0191] The terminal captures audio in real time using its built-in microphone as soon as the meeting starts. The input is ambient sound, and this audio is sent to the server as data. Low-latency protocols are used to efficiently capture audio.

[0192] Step 2:

[0193] The server receives audio data sent from the terminal. The input is audio data. This data is input to the speech recognition engine and processed to convert it into text information. The output is text information in which the spoken content has been converted into text format.

[0194] Step 3:

[0195] The server inputs the obtained text information into the sentiment analysis engine. The input is text information. Here, the engine identifies emotions and generates sentiment data such as "joy," "anger," and "sadness." The output is the sentiment data corresponding to each statement.

[0196] Step 4:

[0197] The server analyzes the context of a discussion and the intentions of the participants using textual and sentiment data. The input consists of textual and sentiment data. It utilizes a generative AI model to clarify participants' intentions and identify which statements are important. The output is the analysis results, including the discussion context and participants' intentions.

[0198] Step 5:

[0199] The server generates real-time feedback based on the analysis results. The input is the context and sentiment data of the discussion. Based on this, it extracts suggestions and points to note for improving the meeting's progress and creates a feedback message as output.

[0200] Step 6:

[0201] The terminal displays the generated feedback to participants in real time. The input is feedback messages from the server. In operation, it visually displays feedback and changes in emotion on the display to help participants respond immediately.

[0202] Step 7:

[0203] The server integrates all data after the meeting and creates meeting minutes and an action plan. Inputs include all spoken statements and associated sentiment data. It integrates and summarizes the data, generating meeting minutes and an action plan outlining the next steps as output.

[0204] (Application Example 2)

[0205] 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".

[0206] Conventional conversation analysis systems only identify context and understand intent based on the content of the conversation, and do not adequately provide real-time analysis and feedback on the emotional state of participants. Therefore, there is a need for a mechanism that facilitates appropriate responses in situations where emotions such as tension and stress are abnormally heightened. In particular, in workplace environments, there is a need to detect sudden changes in emotions early and respond appropriately.

[0207] 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.

[0208] In this invention, the server includes means for acquiring audio data of a conversation, means for converting the acquired audio data into text data in real time, means for analyzing the converted text data to identify the context of the discussion and the intentions of the participants, means for generating feedback based on the analysis results and participant sentiment analysis to improve the progress of the conversation, means for providing the generated feedback to the participants in real time, means for creating a record and plan after the conversation has ended based on the analysis results and sentiment data, and means for detecting abnormal sentiment levels and issuing an alarm based on sentiment data. This makes it possible to quickly detect changes in sentiment and take immediate action as needed.

[0209] "Conversation audio data" refers to data that electronically records participants' speech in its original form.

[0210] "Real-time conversion to text data" refers to the process of instantly converting acquired audio data into text information.

[0211] "The context of the discussion" refers to information that indicates the topic and progress of a conversation.

[0212] "Participant's intent" refers to the speaker's purpose or information that expresses their objective in a conversation.

[0213] "Generating feedback" means creating information that suggests areas for improvement and action guidelines based on the analysis results.

[0214] "Providing feedback to participants in real time" means notifying participants of the generated feedback immediately.

[0215] "Creating records and plans" means organizing the data obtained after the conversation and developing documents and guidelines for future actions that can be referenced later.

[0216] "Emotional data" refers to information that indicates the emotional state of a person, extracted from the content of their speech.

[0217] "Detecting abnormal emotional levels" means identifying a state in which a participant's emotions deviate from the normal range.

[0218] "Issuing an alarm" means notifying relevant parties to take precautions when an anomaly is detected.

[0219] The system that implements this application includes a program that acquires and analyzes conversational audio data in real time. This system mainly consists of three elements: a server, a terminal, and a user.

[0220] The server's role is to receive audio data transmitted from terminals and convert it into text data using speech recognition software. Specifically, the audio data is transcribed using Microsoft® Azure® speech recognition APIs, and then analyzed as emotion data using Amazon AWS® emotion analysis APIs. This data analysis clarifies the emotional state and intentions of the conversation participants and generates real-time feedback based on this. Furthermore, if an abnormally high emotional level is detected, a mechanism is in place to issue an alert to designated parties.

[0221] The terminal is a smart device worn by the user (e.g., smart glasses), which is used to capture audio data. This audio data is immediately transferred to a server for analysis. The user receives feedback that is displayed in real time based on the analysis results, providing information that is useful for facilitating meetings and discussions.

[0222] For example, if an employee's stress level becomes abnormally high during a workplace meeting, the system can detect this and notify the mental health officer, enabling early follow-up. This system is expected to improve the workplace environment and maintain employees' mental health.

[0223] Examples of prompt messages include the following:

[0224] "I want to develop a system that analyzes emotions from the following audio data and immediately notifies users if their stress levels are abnormal. Please provide specific program code and examples of API usage."

[0225] This makes it possible to appropriately manage the emotional state of participants in a conversational environment and promote good communication.

[0226] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0227] Step 1:

[0228] The device captures audio data of the conversation. During this process, the smart device records the surrounding conversation, and the acquired audio is prepared as digital data in its original format. The input is ambient sound, and the output is digital audio data.

[0229] Step 2:

[0230] The terminal transfers the captured audio data to the server. The audio data is sent to and received by the server in real time using a secure protocol. The input is digital audio data, and the output is the audio data sent to the server.

[0231] Step 3:

[0232] The server converts the received audio data into text data using speech recognition software. Here, Microsoft Azure's speech recognition API is used to process the audio into text information. The input is audio data, and the output is text data.

[0233] Step 4:

[0234] The server analyzes the received text data using Amazon AWS sentiment analysis API and generates sentiment data for each statement. The input is text data, and the output after data analysis is sentiment data. In this process, the user's emotional state is classified into multiple categories such as "surprise," "joy," and "fear."

[0235] Step 5:

[0236] The server uses the analysis results to construct feedback generated from the conversation context and sentiment data, and sends it to the terminal. It compiles information useful for facilitating the process for each participant. The input is the analyzed text and sentiment data, and the output is feedback data.

[0237] Step 6:

[0238] The device displays feedback received by the user in real time. Based on this information, the user can understand the progress of the conversation and their own or others' emotional states. The input is feedback data, and the output is the information displayed on the user's screen.

[0239] Step 7:

[0240] At the end of each session, the server automatically creates a record and plan based on the retained data. This record integrates sentiment data and feedback, and includes an action plan for the next conversation. Inputs are the text and sentiment data recorded during the session, and outputs are the recorded document and the plan for the next steps.

[0241] 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.

[0242] 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.

[0243] 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.

[0244] [Second Embodiment]

[0245] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0246] 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.

[0247] 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).

[0248] 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.

[0249] 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.

[0250] 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).

[0251] 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.

[0252] 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.

[0253] 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.

[0254] 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.

[0255] 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.

[0256] 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".

[0257] This invention provides technology for conducting meetings and discussions more efficiently, based on an AI system for analyzing discussion intent. Specifically, it accurately grasps the content of meetings and generates feedback by converting audio during meetings into text in real time and analyzing the context of the discussion.

[0258] Acquisition and conversion of audio data

[0259] When a meeting begins, the terminal uses its built-in microphone to acquire audio. The acquired audio data is compressed using compression technology and sent to the server via the internet.

[0260] The server uses a speech recognition engine to convert speech data into text data. It also implements processes that enhance pronunciation clarity by utilizing noise reduction and speech enhancement technologies.

[0261] Text data analysis

[0262] The server analyzes text data using natural language processing techniques. A contextual analysis engine identifies the speaker's intent and the topic of discussion, and extracts important keywords.

[0263] To classify each utterance by speaker, speaker identification technology is applied. This is achieved using a model that learns unique linguistic patterns contained in the text.

[0264] Generating feedback and instructions

[0265] Based on the analysis results, the server generates feedback and specific instructions to improve the meeting's progress. This may include changing topics that have been discussed too much or identifying unresolved issues.

[0266] The generated feedback is immediately sent to the device and presented to the user visually or audibly.

[0267] Creating meeting minutes and action plans

[0268] After the meeting concludes, the server integrates the accumulated data to create reviewable meeting minutes and a clear action plan. This includes summaries of each topic discussed and conclusions, as well as specific actions recommended as next steps.

[0269] The completed meeting minutes and action plan are saved to cloud storage, making them accessible to users and shareable with other participants as needed.

[0270] Specific example

[0271] For example, in a product development meeting at a certain company, this system transcribes participants' comments in real time, identifying topics such as changes in product specifications, development schedules, and marketing strategies. If a comment such as "The release date of this product may be delayed" is made during the meeting, the system picks up on this comment and presents it as feedback suggesting a review of the milestones. After the meeting, meeting minutes accessible to all participants are provided on the cloud. This allows participants to review the tasks and actions they need to prepare for the next meeting.

[0272] The following describes the processing flow.

[0273] Step 1:

[0274] The device captures meeting audio in real time using its built-in microphone. The audio data is compressed and sent to the server over the network. A noise reduction algorithm is applied to optimize audio quality.

[0275] Step 2:

[0276] The server inputs the received audio data into a speech recognition engine and converts it into text data. Deep learning technology is used for speech recognition, and the speaker's accent and vocabulary are analyzed to improve accuracy.

[0277] Step 3:

[0278] The server uses a natural language processing engine to analyze the generated text data. This involves contextual analysis, keyword extraction, and sentiment analysis to identify the intent behind participants' statements and the topics of discussion.

[0279] Step 4:

[0280] Based on the results of text analysis, the server generates feedback for improving the progress of the meeting and instructions regarding the next actions. The generated information is immediately sent to the terminal and presented visually or audibly to the meeting participants.

[0281] Step 5:

[0282] The terminal displays the feedback received from the server on the user interface. The user makes decisions for advancing the discussion based on this information.

[0283] Step 6:

[0284] After the meeting ends, the server integrates all the analysis data and creates detailed meeting minutes and an action plan for the next meeting. This includes summaries of statements, important conclusions, and recommended actions.

[0285] Step 7:

[0286] The final meeting minutes and action plan are saved in cloud storage and made accessible to the users. The users use this information to prepare for and follow up on the next meeting.

[0287] (Example 1)

[0288] Next, Example 1 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart glasses 214 are referred to as the "terminal".

[0289] Meetings and discussions often have the problem that the content and intent of statements are not correctly understood, causing delays in the progress of the discussion. Also, it is difficult to create meeting minutes and action plans quickly and accurately after the meeting. As a result, information sharing among participants is insufficient, and decision-making may be delayed.

[0290] 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.

[0291] In this invention, the server includes means for acquiring speech during a meeting as audio information, means for converting the acquired audio information into text information in real time, means for analyzing the converted text information to identify the context of the discussion and the intentions of the participants, means for applying a noise reduction filter and voice enhancement to the audio information, means for generating instructions to improve the progress of the meeting based on the analysis results, means for immediately providing the generated instructions to the participants, and means for compiling the analysis results after the meeting to create a report and an action plan. This improves the efficiency of meetings and enables rapid and accurate information sharing and decision-making.

[0292] "Audio information" refers to data recorded in digital format from speeches such as those heard in meetings or conversations.

[0293] "Acquiring" refers to the act of collecting audio information from a device.

[0294] "Textual information" refers to string data that has been converted from audio information.

[0295] "Real-time" means that processing and conversion occur simultaneously with the utterance.

[0296] "Analysis" refers to the process of processing information based on converted textual data to understand its context and intent.

[0297] "Context" refers to the framework that gives meaning and relevance within a statement or dialogue.

[0298] A "noise reduction filter" refers to a technology that reduces unwanted background noise from audio information.

[0299] "Voice enhancement" refers to the process of adjusting sound to make necessary sounds in audio information easier to hear.

[0300] "Instruction" means specific proposals or commands generated based on the analysis results to support the progress of the meeting.

[0301] "Report" refers to a document that summarizes the content, conclusions, and implementation plan of the discussions during the meeting.

[0302] "Implementation plan" represents specific action guidelines or procedure manuals established based on the matters considered after the meeting.

[0303] "Participant" refers to the speaker or those involved in the meeting or discussion.

[0304] The embodiments for implementing this invention will be specifically described.

[0305] This discussion intention analysis AI system aims at efficient meeting progress and record creation. First, the terminal uses the built-in microphone to acquire the utterances during the meeting as audio information. The audio information is converted into digital format and efficiently processed through data compression technology. This data is transmitted to the server via the Internet.

[0306] On the server, by using an advanced speech recognition engine, the audio information is converted into character information in real time. This conversion process is enhanced in accuracy by using noise removal filters and voice enhancement techniques. The converted character information is assigned a timestamp to maintain the order of the utterances.

[0307] Next, the server analyzes the character information using natural language processing (NLP) technology. Through this analysis, the context of the discussion and the intention of the speaker are identified. Also, a machine learning model is used to identify the speaker and classify each utterance by speaker. This enables detailed tracking and analysis of the discussion.

[0308] Based on the analysis results, the server generates instructions to support the progress of the meeting. These instructions are presented to the user from the terminal as digital or voice output, providing real-time feedback.

[0309] After the meeting ends, the server automatically generates a report and action plan based on the data accumulated so far. The generated report is saved to cloud storage, making it easily accessible to the user and shareable with other participants.

[0310] As a concrete example, in a product development meeting, this system immediately picks up on a statement such as "The product launch date may be delayed" and instructs the system to re-evaluate milestones as appropriate feedback. An example of a prompt to be input into the generating AI model is, "Transcribe the statements made in this meeting in real time, analyze the speaker's intent, and generate feedback."

[0311] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0312] Step 1:

[0313] The device acquires audio during a meeting via its built-in microphone. This audio information is converted into a digital signal, and the data size is reduced using compression technology (e.g., MP3). The input is raw audio, and the output is compressed digital audio data.

[0314] Step 2:

[0315] Compressed audio data is transmitted to a server via the internet. The server utilizes a speech recognition engine to convert this data into text. The input is compressed digital audio data, and the output is text information with a timestamp. Noise reduction and speech enhancement algorithms are applied to optimize sound quality.

[0316] Step 3:

[0317] The server analyzes the converted text information using natural language processing (NLP) techniques. It performs contextual analysis and keyword extraction to identify participants' intentions and discussion topics. The input is text information, and the output is contextual data as a result of the analysis, extracted keywords, and speaker identification information.

[0318] Step 4:

[0319] Based on the analysis results, the server generates instructions to improve the meeting progress. This involves intent analysis using prompt sentences and utilizing a generative AI model to create specific feedback. The input is the analysis results from step 3, and the output is instructions and feedback presented in real time.

[0320] Step 5:

[0321] The generated instructions and feedback are provided to the user through the device. These are presented either on a screen or verbally using speech synthesis technology. The input is the instructions from step 4, and the output is the visual or auditory information received by the user.

[0322] Step 6:

[0323] After the meeting concludes, the server integrates all data to create a report and action plan. This process includes summarizing each generated instruction and statement. Input is data from all sessions, and output is the report and action plan as documents.

[0324] Step 7:

[0325] Finally, the generated report and execution plan are saved to cloud storage and configured to be accessible to the user. This process is for saving the output of step 6 to the cloud.

[0326] (Application Example 1)

[0327] 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."

[0328] A challenge arises when information sharing among staff at a store is not conducted quickly and accurately, as this can hinder customer service and operational efficiency. This problem stems particularly from the inability to respond quickly to real-time changes in customer requests and operational tasks, potentially resulting in a decline in service quality.

[0329] 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.

[0330] In this invention, the server includes a device for acquiring speech as audio information, a device for immediately converting the acquired audio information into text information, and a device for analyzing the converted text information to identify the background of the discussion and the intentions of the participants. This enables immediate information sharing among staff, real-time acquisition of information necessary for customer service and business operations, and efficient business operations.

[0331] "Speech" refers to linguistic expression used by people to convey intentions and information through sound.

[0332] "Audio information" refers to data obtained by acquiring and processing sound as a digital signal, making it analyzable.

[0333] "Textual information" refers to text data converted from audio information, and is used for analysis.

[0334] A "device" is a system of machinery or software designed to perform a specific function.

[0335] "The background of the discussion" refers to the context and situation in which participants speak, and it serves as the foundation for understanding their intentions.

[0336] "Instantly" means that processing or results can be obtained in near real-time with no delay.

[0337] "Analysis" is the process of organizing and breaking down information to understand its structure and meaning.

[0338] An "information terminal" is an electronic device used to display and manipulate data.

[0339] A "record document" is a formal written document in which information is organized and preserved for later reference.

[0340] An "action plan" is a detailed document outlining specific steps and methods for achieving a particular objective.

[0341] "Memory space" refers to digital space used to store data and information.

[0342] The system that implements this application consists of multiple components. The server uses a combination of appropriate hardware and software to instantly acquire and process audio information. Specifically, it uses a mobile device such as a smartphone or tablet to acquire audio, utilizing the microphone built into it. The device compresses the acquired audio information and transmits it to the server via the network.

[0343] The server uses an advanced speech recognition engine to convert audio information into text. This process leverages the Google Cloud Speech-to-Text API, improving recognition accuracy through noise reduction and speech enhancement techniques. The server then analyzes the converted text, using natural language processing techniques to identify the context of the discussion and the participants' intentions. This utilizes natural language processing libraries such as spaCy.

[0344] The analysis results are provided to the user as real-time feedback via an information terminal. Users can use this interface to immediately check the information and take necessary actions. After completion, the analysis data is compiled into a record document and action plan, and saved to cloud storage. Users can access and share this information at any time via their digital devices. Google Cloud Platform is used for the cloud service.

[0345] As a concrete example, consider information sharing regarding the arrival date of new products at a store. Using this system, a voice command such as "Check when the new products will arrive and inform everyone" would be instantly analyzed, and the results would be shared with the staff. This would allow each staff member to quickly prepare countermeasures and improve the quality of service provided to customers.

[0346] An example of a prompt for a generative AI model is: "Explain how to build a system that extracts important business tasks from store staff conversations and provides real-time feedback."

[0347] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0348] Step 1:

[0349] The device acquires conversational audio information using a microphone. The audio data, as input, is captured directly from the microphone as raw sound. The device then uses a compression algorithm to reduce the size of the audio data and convert it into a format suitable for transfer to the server. This process reduces the data size while maintaining the quality of the audio signal.

[0350] Step 2:

[0351] The server receives audio data sent from the terminal. The input is compressed audio data. The server uses the Google Cloud Speech-to-Text API to convert the audio data into text. During the conversion process, it uses techniques to suppress noise and emphasize important audio components. The output is text data that reflects the content of the discussion.

[0352] Step 3:

[0353] The server analyzes the text data. The input is the text data generated in step 2. It utilizes natural language processing techniques, specifically the spaCy library, to analyze the background of the discussion and the intentions of the participants. It extracts keywords and important contexts and structures the information based on them. The output is metadata of the analysis results, i.e., a set of identified intentions and keywords.

[0354] Step 4:

[0355] The server generates feedback based on the analysis results. The input is the metadata obtained in step 3. This generates effective feedback to improve the meeting's progress. Using a generation AI model, specific improvement suggestions and points are created, and the results are immediately transmitted to the information terminal. The output is the proposed feedback message and points.

[0356] Step 5:

[0357] Users receive this feedback through their information terminals and use it in real time to aid in decision-making. Users can modify their actions based on the feedback as needed, streamlining their work processes. Input is feedback messages, and output is improvement measures and corresponding actions.

[0358] Step 6:

[0359] The server generates a record document and an action plan based on the data analyzed after completion. The input is the data processed in all previous steps. The record document includes an overview of the meeting, a summary of the discussions, and decisions made, while the action plan outlines specific next steps and responsibilities. The output is the saved record document and action plan.

[0360] Step 7:

[0361] Users access record documents and action plans stored in cloud storage, and individually review and share them. Google Cloud Platform is used as the cloud service. Inputs are record documents and action plans, and outputs are the standardization of work through checking and sharing of verified information.

[0362] 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.

[0363] This invention provides a system characterized by its ability to analyze intentions and provide feedback during discussions, as well as to recognize participants' emotions. By incorporating an emotion engine, it enables real-time analysis of user emotions from speech during meetings and allows for the use of the results to improve the progress of the meeting.

[0364] Acquisition of voice data and sentiment analysis

[0365] The terminal begins capturing audio in real time as soon as the meeting starts and transmits it to the server. The audio data is then passed to the emotion engine simultaneously with speech recognition.

[0366] The server converts the received audio data into text using a speech recognition engine and analyzes the emotions contained in the utterances using an emotion engine. It identifies emotions such as joy, anger, and sadness from individual utterances and stores them as data.

[0367] Text data analysis and feedback generation

[0368] The server analyzes text data using natural language processing techniques to identify the context of the discussion and the intentions of the participants. This analysis is then combined with sentiment data to generate feedback that optimizes the meeting's progress.

[0369] The analyzed emotional data is also included as an important factor in creating meeting minutes and developing action plans. In particular, in situations where discussions are heated or there is no agreement, emotional data plays a role in regulating the progress of the meeting.

[0370] Real-time emotion change notifications and meeting minute creation

[0371] The terminal displays feedback and sentiment analysis results sent from the server to the user in real time. This allows the user to make decisions based not only on the context of the discussion but also on the flow of emotions.

[0372] After the meeting concludes, the server integrates all data to create meeting minutes and an action plan. Sentimental data is used as an indicator to assess the impact of each statement, contributing to an objective meeting record.

[0373] Specific example

[0374] For example, in a meeting about the launch of a new product, if a participant expresses strong concerns about market reaction, this system identifies the emotions such as "anxiety" and "concern" contained in their statement. The analysis results in feedback such as "The team needs to discuss this concern further," which is immediately made public to other participants. Subsequently, at the end of the meeting, detailed minutes are created, including how the concerns were discussed and what was decided as the next steps. Sentimental data is an important indicator of which topics generated more emotional discussion.

[0375] The following describes the processing flow.

[0376] Step 1:

[0377] The terminal uses a microphone to capture audio in real time at the start of the meeting. The captured audio data is compressed and then sent to the server. Noise cancellation technology is applied to ensure audio quality.

[0378] Step 2:

[0379] The server converts the received audio data into text data using a speech recognition engine. The speech recognition uses a highly accurate deep learning model to handle a wide variety of speech patterns.

[0380] Step 3:

[0381] The server passes the audio data to the emotion engine, which analyzes the user's emotions from the conversation. It identifies the emotional tone in each utterance, such as joy, anger, or sadness, and stores this data along with other analysis data.

[0382] Step 4:

[0383] The server analyzes text data using natural language processing algorithms to identify the context of the discussion and the intentions of the participants. Sentiment data plays a crucial role in this analysis, deepening the understanding of how the discussion is progressing.

[0384] Step 5:

[0385] The server generates feedback based on the analysis results to improve the members' progress. This includes summaries of key points from the discussion and instructions on the direction of the meeting based on sentiment analysis.

[0386] Step 6:

[0387] The device displays generated feedback and real-time analyzed sentiment data to the user. This allows the user to understand the emotional state of the participants while facilitating the discussion.

[0388] Step 7:

[0389] After the meeting concludes, the server integrates all data to create detailed minutes and action plans. Sentimental data is used to assess the impact of the dialogue and show how the discussion evolved.

[0390] Step 8:

[0391] The completed meeting minutes and action plans are saved to cloud storage and made accessible to users as needed. This allows meeting participants to clearly define their next steps and follow up efficiently.

[0392] (Example 2)

[0393] 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".

[0394] Traditional meeting systems made it difficult to provide feedback that took participants' emotions into account, potentially influencing the progress of discussions and decision-making processes. Furthermore, it was challenging to quickly and accurately analyze individual statements and their intentions, and to effectively utilize emotional data. This resulted in decreased meeting efficiency and productivity.

[0395] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0396] In this invention, the server includes means for acquiring audio information during a meeting as data, means for converting the acquired audio information into text information in real time, means for analyzing the converted text information to identify the context of the discussion and the intentions of the participants, means for recognizing the emotions contained in the analyzed text information and storing them as data, means for generating feedback to optimize the progress of the meeting based on the analysis results, and means for compiling the emotion data and analysis results after the meeting to create a meeting record and an activity plan. This enables the provision of feedback that takes into account the emotions of the participants in real time, and makes it possible to grasp intentions and optimize the progress of the meeting using emotion data.

[0397] "Audio information" refers to a series of sound data collected during a meeting, used to record the progress of the meeting and the statements made.

[0398] "Textual information" refers to text data converted in real time based on audio information, and is used to visually recognize and analyze the content of meetings.

[0399] "Context" refers to the linguistic background information of statements made during a meeting, as well as the circumstances and flow in which those statements were made, and is of great importance to the progress of the discussion.

[0400] "Intention" refers to the purpose or aim that a speaker is trying to convey through their words, and it is a factor that influences the direction of the discussion.

[0401] "Emotions" are psychological states underpinned by words and actions, such as expressing feelings like joy, anger, and sadness.

[0402] "Feedback" refers to opinions and suggestions provided to participants based on analyzed data to optimize the flow of a meeting, and is important information for indicating areas for improvement.

[0403] "Meeting minutes" are documents that summarize the discussions, statements, and decisions made during a meeting, and are used as reference material for future reference.

[0404] An "activity plan" is a document that outlines specific future action plans and schedules based on the results of a meeting, and it supports the organization's strategic execution.

[0405] "Data" refers to facts and information that have been measured, collected, and analyzed, and includes numerical values, text, and other information that are processed and stored within a system.

[0406] This system optimizes meeting progress by analyzing audio during meetings in real time, interpreting participants' emotions and intentions from their statements, and providing feedback.

[0407] The terminal uses a built-in high-performance microphone to capture meeting audio information in real time. A dedicated application is installed that has the ability to continuously monitor audio input and acquire data. Audio data is transmitted to the server immediately using a low-latency protocol (e.g., WebSocket).

[0408] The server converts the received audio information into text using a speech recognition engine (e.g., a speech recognition API from a major cloud service provider). Furthermore, the server uses an emotion analysis engine and natural language processing technology (e.g., a natural language model API from a commercial AI provider) to identify emotions from the spoken content and to deeply analyze the context of the discussion and the intentions of the participants.

[0409] The analyzed information is used to generate feedback that takes into account the emotional data contained in each statement and the flow of the discussion. This feedback is provided to participants in real time and serves as an important indicator for the progress of the meeting. After the meeting, meeting records and activity plans are automatically generated based on the accumulated data and saved to shared network storage.

[0410] As a concrete example, in a meeting about the market strategy for a new product, if a participant expresses concerns about the product's design, the system analyzes the emotion of "anxiety" contained in their statement and provides feedback such as "further design evaluation is needed." This feedback is instantly displayed to the participants, allowing everyone to share their concerns.

[0411] An example of a prompt could be, "In a meeting about a new product launch, please tell me how to identify participants' anxieties and concerns and generate appropriate feedback." This system makes it easier for users to understand the flow of emotions and obtain information to make better decisions.

[0412] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0413] Step 1:

[0414] The terminal captures audio in real time using its built-in microphone as soon as the meeting starts. The input is ambient sound, and this audio is sent to the server as data. Low-latency protocols are used to efficiently capture audio.

[0415] Step 2:

[0416] The server receives audio data sent from the terminal. The input is audio data. This data is input to the speech recognition engine and processed to convert it into text information. The output is text information in which the spoken content has been converted into text format.

[0417] Step 3:

[0418] The server inputs the obtained text information into the sentiment analysis engine. The input is text information. Here, the engine identifies emotions and generates sentiment data such as "joy," "anger," and "sadness." The output is the sentiment data corresponding to each statement.

[0419] Step 4:

[0420] The server analyzes the context of a discussion and the intentions of the participants using textual and sentiment data. The input consists of textual and sentiment data. It utilizes a generative AI model to clarify participants' intentions and identify which statements are important. The output is the analysis results, including the discussion context and participants' intentions.

[0421] Step 5:

[0422] The server generates real-time feedback based on the analysis results. The input is the context and sentiment data of the discussion. Based on this, it extracts suggestions and points to note for improving the meeting's progress and creates a feedback message as output.

[0423] Step 6:

[0424] The terminal displays the generated feedback to participants in real time. The input is feedback messages from the server. In operation, it visually displays feedback and changes in emotion on the display to help participants respond immediately.

[0425] Step 7:

[0426] The server integrates all data after the meeting and creates meeting minutes and an action plan. Inputs include all spoken statements and associated sentiment data. It integrates and summarizes the data, generating meeting minutes and an action plan outlining the next steps as output.

[0427] (Application Example 2)

[0428] 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."

[0429] Conventional conversation analysis systems only identify context and understand intent based on the content of the conversation, and do not adequately provide real-time analysis and feedback on the emotional state of participants. Therefore, there is a need for a mechanism that facilitates appropriate responses in situations where emotions such as tension and stress are abnormally heightened. In particular, in workplace environments, there is a need to detect sudden changes in emotions early and respond appropriately.

[0430] 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.

[0431] In this invention, the server includes means for acquiring audio data of a conversation, means for converting the acquired audio data into text data in real time, means for analyzing the converted text data to identify the context of the discussion and the intentions of the participants, means for generating feedback based on the analysis results and participant sentiment analysis to improve the progress of the conversation, means for providing the generated feedback to the participants in real time, means for creating a record and plan after the conversation has ended based on the analysis results and sentiment data, and means for detecting abnormal sentiment levels and issuing an alarm based on sentiment data. This makes it possible to quickly detect changes in sentiment and take immediate action as needed.

[0432] "Conversation audio data" refers to data that electronically records participants' speech in its original form.

[0433] "Real-time conversion to text data" refers to the process of instantly converting acquired audio data into text information.

[0434] "The context of the discussion" refers to information that indicates the topic and progress of a conversation.

[0435] "Participant's intent" refers to the speaker's purpose or information that expresses their objective in a conversation.

[0436] "Generating feedback" means creating information that suggests areas for improvement and action guidelines based on the analysis results.

[0437] "Providing feedback to participants in real time" means notifying participants of the generated feedback immediately.

[0438] "Creating records and plans" means organizing the data obtained after the conversation and developing documents and guidelines for future actions that can be referenced later.

[0439] "Emotional data" refers to information that indicates the emotional state of a person, extracted from the content of their speech.

[0440] "Detecting abnormal emotional levels" means identifying a state in which a participant's emotions deviate from the normal range.

[0441] "Issuing an alarm" means notifying relevant parties to take precautions when an anomaly is detected.

[0442] The system that implements this application includes a program that acquires and analyzes conversational audio data in real time. This system mainly consists of three elements: a server, a terminal, and a user.

[0443] The server's role is to receive audio data transmitted from terminals and convert it into text data using speech recognition software. Specifically, the audio data is transcribed using Microsoft Azure's speech recognition API, and then analyzed as emotion data using Amazon AWS's emotion analysis API. This data analysis clarifies the emotional state and intentions of the conversation participants and generates real-time feedback based on this. Furthermore, if an abnormally high emotional level is detected, a mechanism is in place to issue an alert to designated parties.

[0444] The terminal is a smart device worn by the user (e.g., smart glasses), which is used to capture audio data. This audio data is immediately transferred to a server for analysis. The user receives feedback that is displayed in real time based on the analysis results, providing information that is useful for facilitating meetings and discussions.

[0445] For example, if an employee's stress level becomes abnormally high during a workplace meeting, the system can detect this and notify the mental health officer, enabling early follow-up. This system is expected to improve the workplace environment and maintain employees' mental health.

[0446] Examples of prompt messages include the following:

[0447] "I want to develop a system that analyzes emotions from the following audio data and immediately notifies users if their stress levels are abnormal. Please provide specific program code and examples of API usage."

[0448] This makes it possible to appropriately manage the emotional state of participants in a conversational environment and promote good communication.

[0449] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0450] Step 1:

[0451] The device captures audio data of the conversation. During this process, the smart device records the surrounding conversation, and the acquired audio is prepared as digital data in its original format. The input is ambient sound, and the output is digital audio data.

[0452] Step 2:

[0453] The terminal transfers the captured audio data to the server. The audio data is sent to and received by the server in real time using a secure protocol. The input is digital audio data, and the output is the audio data sent to the server.

[0454] Step 3:

[0455] The server converts the received audio data into text data using speech recognition software. Here, Microsoft Azure's speech recognition API is used to process the audio into text information. The input is audio data, and the output is text data.

[0456] Step 4:

[0457] The server analyzes the received text data using Amazon AWS sentiment analysis API and generates sentiment data for each statement. The input is text data, and the output after data analysis is sentiment data. In this process, the user's emotional state is classified into multiple categories such as "surprise," "joy," and "fear."

[0458] Step 5:

[0459] The server uses the analysis results to construct feedback generated from the conversation context and sentiment data, and sends it to the terminal. It compiles information useful for facilitating the process for each participant. The input is the analyzed text and sentiment data, and the output is feedback data.

[0460] Step 6:

[0461] The device displays feedback received by the user in real time. Based on this information, the user can understand the progress of the conversation and their own or others' emotional states. The input is feedback data, and the output is the information displayed on the user's screen.

[0462] Step 7:

[0463] At the end of each session, the server automatically creates a record and plan based on the retained data. This record integrates sentiment data and feedback, and includes an action plan for the next conversation. Inputs are the text and sentiment data recorded during the session, and outputs are the recorded document and the plan for the next steps.

[0464] 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.

[0465] 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.

[0466] 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.

[0467] [Third Embodiment]

[0468] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0469] 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.

[0470] 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).

[0471] 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.

[0472] 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.

[0473] 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).

[0474] 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.

[0475] 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.

[0476] 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.

[0477] 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.

[0478] 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.

[0479] 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".

[0480] This invention provides technology for conducting meetings and discussions more efficiently, based on an AI system for analyzing discussion intent. Specifically, it accurately grasps the content of meetings and generates feedback by converting audio during meetings into text in real time and analyzing the context of the discussion.

[0481] Acquisition and conversion of audio data

[0482] When a meeting begins, the terminal uses its built-in microphone to acquire audio. The acquired audio data is compressed using compression technology and sent to the server via the internet.

[0483] The server uses a speech recognition engine to convert speech data into text data. It also implements processes that enhance pronunciation clarity by utilizing noise reduction and speech enhancement technologies.

[0484] Text data analysis

[0485] The server analyzes text data using natural language processing techniques. A contextual analysis engine identifies the speaker's intent and the topic of discussion, and extracts important keywords.

[0486] To classify each utterance by speaker, speaker identification technology is applied. This is achieved using a model that learns unique linguistic patterns contained in the text.

[0487] Generating feedback and instructions

[0488] Based on the analysis results, the server generates feedback and specific instructions to improve the meeting's progress. This may include changing topics that have been discussed too much or identifying unresolved issues.

[0489] The generated feedback is immediately sent to the device and presented to the user visually or audibly.

[0490] Creating meeting minutes and action plans

[0491] After the meeting concludes, the server integrates the accumulated data to create reviewable meeting minutes and a clear action plan. This includes summaries of each topic discussed and conclusions, as well as specific actions recommended as next steps.

[0492] The completed meeting minutes and action plan are saved to cloud storage, making them accessible to users and shareable with other participants as needed.

[0493] Specific example

[0494] For example, in a product development meeting at a certain company, this system transcribes participants' comments in real time, identifying topics such as changes in product specifications, development schedules, and marketing strategies. If a comment such as "The release date of this product may be delayed" is made during the meeting, the system picks up on this comment and presents it as feedback suggesting a review of the milestones. After the meeting, meeting minutes accessible to all participants are provided on the cloud. This allows participants to review the tasks and actions they need to prepare for the next meeting.

[0495] The following describes the processing flow.

[0496] Step 1:

[0497] The device captures meeting audio in real time using its built-in microphone. The audio data is compressed and sent to the server over the network. A noise reduction algorithm is applied to optimize audio quality.

[0498] Step 2:

[0499] The server inputs the received audio data into a speech recognition engine and converts it into text data. Deep learning technology is used for speech recognition, and the speaker's accent and vocabulary are analyzed to improve accuracy.

[0500] Step 3:

[0501] The server uses a natural language processing engine to analyze the generated text data. This involves contextual analysis, keyword extraction, and sentiment analysis to identify the intent behind participants' statements and the topics of discussion.

[0502] Step 4:

[0503] Based on the results of the text analysis, the server generates feedback and instructions for the next steps to improve the progress of the meeting. The generated information is immediately sent to the terminal and presented to meeting participants visually or audibly.

[0504] Step 5:

[0505] The terminal displays feedback received from the server in the user interface. Users use this information to make decisions that will advance the discussion.

[0506] Step 6:

[0507] After the meeting concludes, the server integrates all analytical data to create detailed meeting minutes and an action plan for the next meeting. This includes a summary of the discussion, key conclusions, and recommended actions.

[0508] Step 7:

[0509] The final meeting minutes and action plan will be saved to cloud storage and made accessible to users. Users will use this information to prepare for and follow up on future meetings.

[0510] (Example 1)

[0511] 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."

[0512] Meetings and discussions often face challenges such as misunderstandings of what is said and the intentions behind it, leading to stalled progress. Furthermore, it is difficult to quickly and accurately create meeting minutes and action plans afterward. This can result in insufficient information sharing among participants and delays in decision-making.

[0513] 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.

[0514] In this invention, the server includes means for acquiring speech during a meeting as audio information, means for converting the acquired audio information into text information in real time, means for analyzing the converted text information to identify the context of the discussion and the intentions of the participants, means for applying a noise reduction filter and voice enhancement to the audio information, means for generating instructions to improve the progress of the meeting based on the analysis results, means for immediately providing the generated instructions to the participants, and means for compiling the analysis results after the meeting to create a report and an action plan. This improves the efficiency of meetings and enables rapid and accurate information sharing and decision-making.

[0515] "Audio information" refers to data recorded in digital format from speeches such as those heard in meetings or conversations.

[0516] "Acquiring" refers to the act of collecting audio information from a device.

[0517] "Textual information" refers to string data that has been converted from audio information.

[0518] "Real-time" means that processing and conversion occur simultaneously with the utterance.

[0519] "Analysis" refers to the process of processing information based on converted textual data to understand its context and intent.

[0520] "Context" refers to the framework that gives meaning and relevance within a statement or dialogue.

[0521] A "noise reduction filter" refers to a technology that reduces unwanted background noise from audio information.

[0522] "Voice enhancement" refers to the process of adjusting sound to make necessary sounds in audio information easier to hear.

[0523] "Instructions" refers to specific suggestions or commands generated based on the analysis results to support the progress of the meeting.

[0524] A "report" refers to a document that summarizes the content of discussions, conclusions, and action plans during a meeting.

[0525] An "action plan" refers to specific guidelines or procedures developed based on the matters discussed after the meeting.

[0526] "Participants" refers to those who speak or participate in a meeting or discussion.

[0527] A specific description will be given of embodiments for carrying out this invention.

[0528] This discussion intent analysis AI system aims to facilitate efficient meeting management and record-keeping. First, the terminal uses its built-in microphone to capture speech during the meeting as audio information. The audio information is converted into a digital format and processed efficiently through data compression technology. This data is then transmitted to a server via the internet.

[0529] The server uses an advanced speech recognition engine to convert speech information into text in real time. This conversion process is enhanced with noise reduction filters and voice enhancement techniques to improve accuracy. The converted text is time-stamped to maintain the order of speech.

[0530] Next, the server uses natural language processing (NLP) techniques to analyze the textual information. This analysis identifies the context of the discussion and the intentions of the speakers. Machine learning models are also used to identify speakers and classify each statement by speaker. This enables detailed tracking and analysis of the discussion.

[0531] Based on the analysis results, the server generates instructions to support the progress of the meeting. These instructions are presented to the user from the terminal as digital or audio output, providing real-time feedback.

[0532] After the meeting ends, the server automatically generates a report and action plan based on the data accumulated so far. The generated report is saved to cloud storage, making it easily accessible to the user and shareable with other participants.

[0533] As a concrete example, in a product development meeting, this system immediately picks up on a statement such as "The product launch date may be delayed" and instructs the system to re-evaluate milestones as appropriate feedback. An example of a prompt to be input into the generating AI model is, "Transcribe the statements made in this meeting in real time, analyze the speaker's intent, and generate feedback."

[0534] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0535] Step 1:

[0536] The device acquires audio during a meeting via its built-in microphone. This audio information is converted into a digital signal, and the data size is reduced using compression technology (e.g., MP3). The input is raw audio, and the output is compressed digital audio data.

[0537] Step 2:

[0538] Compressed audio data is transmitted to a server via the internet. The server utilizes a speech recognition engine to convert this data into text. The input is compressed digital audio data, and the output is text information with a timestamp. Noise reduction and speech enhancement algorithms are applied to optimize sound quality.

[0539] Step 3:

[0540] The server analyzes the converted text information using natural language processing (NLP) techniques. It performs contextual analysis and keyword extraction to identify participants' intentions and discussion topics. The input is text information, and the output is contextual data as a result of the analysis, extracted keywords, and speaker identification information.

[0541] Step 4:

[0542] Based on the analysis results, the server generates instructions to improve the meeting progress. This involves intent analysis using prompt sentences and utilizing a generative AI model to create specific feedback. The input is the analysis results from step 3, and the output is instructions and feedback presented in real time.

[0543] Step 5:

[0544] The generated instructions and feedback are provided to the user through the device. These are presented either on a screen or verbally using speech synthesis technology. The input is the instructions from step 4, and the output is the visual or auditory information received by the user.

[0545] Step 6:

[0546] After the meeting concludes, the server integrates all data to create a report and action plan. This process includes summarizing each generated instruction and statement. Input is data from all sessions, and output is the report and action plan as documents.

[0547] Step 7:

[0548] Finally, the generated report and execution plan are saved to cloud storage and configured to be accessible to the user. This process is for saving the output of step 6 to the cloud.

[0549] (Application Example 1)

[0550] 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."

[0551] A challenge arises when information sharing among staff at a store is not conducted quickly and accurately, as this can hinder customer service and operational efficiency. This problem stems particularly from the inability to respond quickly to real-time changes in customer requests and operational tasks, potentially resulting in a decline in service quality.

[0552] 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.

[0553] In this invention, the server includes a device for acquiring speech as audio information, a device for immediately converting the acquired audio information into text information, and a device for analyzing the converted text information to identify the background of the discussion and the intentions of the participants. This enables immediate information sharing among staff, real-time acquisition of information necessary for customer service and business operations, and efficient business operations.

[0554] "Speech" refers to linguistic expression used by people to convey intentions and information through sound.

[0555] "Audio information" refers to data obtained by acquiring and processing sound as a digital signal, making it analyzable.

[0556] "Textual information" refers to text data converted from audio information, and is used for analysis.

[0557] A "device" is a system of machinery or software designed to perform a specific function.

[0558] "The background of the discussion" refers to the context and situation in which participants speak, and it serves as the foundation for understanding their intentions.

[0559] "Instantly" means that processing or results can be obtained in near real-time with no delay.

[0560] "Analysis" is the process of organizing and breaking down information to understand its structure and meaning.

[0561] An "information terminal" is an electronic device used to display and manipulate data.

[0562] A "record document" is a formal written document in which information is organized and preserved for later reference.

[0563] An "action plan" is a detailed document outlining specific steps and methods for achieving a particular objective.

[0564] "Memory space" refers to digital space used to store data and information.

[0565] The system that implements this application consists of multiple components. The server uses a combination of appropriate hardware and software to instantly acquire and process audio information. Specifically, it uses a mobile device such as a smartphone or tablet to acquire audio, utilizing the microphone built into it. The device compresses the acquired audio information and transmits it to the server via the network.

[0566] The server uses an advanced speech recognition engine to convert audio information into text. This process leverages the Google Cloud Speech-to-Text API, improving recognition accuracy through noise reduction and speech enhancement techniques. The server then analyzes the converted text, using natural language processing techniques to identify the context of the discussion and the participants' intentions. This utilizes natural language processing libraries such as spaCy.

[0567] The analysis results are provided to the user as real-time feedback via an information terminal. Users can use this interface to immediately check the information and take necessary actions. After completion, the analysis data is compiled into a record document and action plan, and saved to cloud storage. Users can access and share this information at any time via their digital devices. Google Cloud Platform is used for the cloud service.

[0568] As a concrete example, consider information sharing regarding the arrival date of new products at a store. Using this system, a voice command such as "Check when the new products will arrive and inform everyone" would be instantly analyzed, and the results would be shared with the staff. This would allow each staff member to quickly prepare countermeasures and improve the quality of service provided to customers.

[0569] An example of a prompt for a generative AI model is: "Explain how to build a system that extracts important business tasks from store staff conversations and provides real-time feedback."

[0570] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0571] Step 1:

[0572] The device acquires conversational audio information using a microphone. The audio data, as input, is captured directly from the microphone as raw sound. The device then uses a compression algorithm to reduce the size of the audio data and convert it into a format suitable for transfer to the server. This process reduces the data size while maintaining the quality of the audio signal.

[0573] Step 2:

[0574] The server receives audio data sent from the terminal. The input is compressed audio data. The server uses the Google Cloud Speech-to-Text API to convert the audio data into text. During the conversion process, it uses techniques to suppress noise and emphasize important audio components. The output is text data that reflects the content of the discussion.

[0575] Step 3:

[0576] The server analyzes the text data. The input is the text data generated in step 2. It utilizes natural language processing techniques, specifically the spaCy library, to analyze the background of the discussion and the intentions of the participants. It extracts keywords and important contexts and structures the information based on them. The output is metadata of the analysis results, i.e., a set of identified intentions and keywords.

[0577] Step 4:

[0578] The server generates feedback based on the analysis results. The input is the metadata obtained in step 3. This generates effective feedback to improve the meeting's progress. Using a generation AI model, specific improvement suggestions and points are created, and the results are immediately transmitted to the information terminal. The output is the proposed feedback message and points.

[0579] Step 5:

[0580] Users receive this feedback through their information terminals and use it in real time to aid in decision-making. Users can modify their actions based on the feedback as needed, streamlining their work processes. Input is feedback messages, and output is improvement measures and corresponding actions.

[0581] Step 6:

[0582] The server generates a record document and an action plan based on the data analyzed after completion. The input is the data processed in all previous steps. The record document includes an overview of the meeting, a summary of the discussions, and decisions made, while the action plan outlines specific next steps and responsibilities. The output is the saved record document and action plan.

[0583] Step 7:

[0584] Users access record documents and action plans stored in cloud storage, and individually review and share them. Google Cloud Platform is used as the cloud service. Inputs are record documents and action plans, and outputs are the standardization of work through checking and sharing of verified information.

[0585] 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.

[0586] This invention provides a system characterized by its ability to analyze intentions and provide feedback during discussions, as well as to recognize participants' emotions. By incorporating an emotion engine, it enables real-time analysis of user emotions from speech during meetings and allows for the use of the results to improve the progress of the meeting.

[0587] Acquisition of voice data and sentiment analysis

[0588] The terminal begins capturing audio in real time as soon as the meeting starts and transmits it to the server. The audio data is then passed to the emotion engine simultaneously with speech recognition.

[0589] The server converts the received audio data into text using a speech recognition engine and analyzes the emotions contained in the utterances using an emotion engine. It identifies emotions such as joy, anger, and sadness from individual utterances and stores them as data.

[0590] Text data analysis and feedback generation

[0591] The server analyzes text data using natural language processing techniques to identify the context of the discussion and the intentions of the participants. This analysis is then combined with sentiment data to generate feedback that optimizes the meeting's progress.

[0592] The analyzed emotional data is also included as an important factor in creating meeting minutes and developing action plans. In particular, in situations where discussions are heated or there is no agreement, emotional data plays a role in regulating the progress of the meeting.

[0593] Real-time emotion change notifications and meeting minute creation

[0594] The terminal displays feedback and sentiment analysis results sent from the server to the user in real time. This allows the user to make decisions based not only on the context of the discussion but also on the flow of emotions.

[0595] After the meeting concludes, the server integrates all data to create meeting minutes and an action plan. Sentimental data is used as an indicator to assess the impact of each statement, contributing to an objective meeting record.

[0596] Specific example

[0597] For example, in a meeting about the launch of a new product, if a participant expresses strong concerns about market reaction, this system identifies the emotions such as "anxiety" and "concern" contained in their statement. The analysis results in feedback such as "The team needs to discuss this concern further," which is immediately made public to other participants. Subsequently, at the end of the meeting, detailed minutes are created, including how the concerns were discussed and what was decided as the next steps. Sentimental data is an important indicator of which topics generated more emotional discussion.

[0598] The following describes the processing flow.

[0599] Step 1:

[0600] The terminal uses a microphone to capture audio in real time at the start of the meeting. The captured audio data is compressed and then sent to the server. Noise cancellation technology is applied to ensure audio quality.

[0601] Step 2:

[0602] The server converts the received audio data into text data using a speech recognition engine. The speech recognition uses a highly accurate deep learning model to handle a wide variety of speech patterns.

[0603] Step 3:

[0604] The server passes the audio data to the emotion engine, which analyzes the user's emotions from the conversation. It identifies the emotional tone in each utterance, such as joy, anger, or sadness, and stores this data along with other analysis data.

[0605] Step 4:

[0606] The server analyzes text data using natural language processing algorithms to identify the context of the discussion and the intentions of the participants. Sentiment data plays a crucial role in this analysis, deepening the understanding of how the discussion is progressing.

[0607] Step 5:

[0608] The server generates feedback based on the analysis results to improve the members' progress. This includes summaries of key points from the discussion and instructions on the direction of the meeting based on sentiment analysis.

[0609] Step 6:

[0610] The device displays generated feedback and real-time analyzed sentiment data to the user. This allows the user to understand the emotional state of the participants while facilitating the discussion.

[0611] Step 7:

[0612] After the meeting concludes, the server integrates all data to create detailed minutes and action plans. Sentimental data is used to assess the impact of the dialogue and show how the discussion evolved.

[0613] Step 8:

[0614] The completed meeting minutes and action plans are saved to cloud storage and made accessible to users as needed. This allows meeting participants to clearly define their next steps and follow up efficiently.

[0615] (Example 2)

[0616] 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."

[0617] Traditional meeting systems made it difficult to provide feedback that took participants' emotions into account, potentially influencing the progress of discussions and decision-making processes. Furthermore, it was challenging to quickly and accurately analyze individual statements and their intentions, and to effectively utilize emotional data. This resulted in decreased meeting efficiency and productivity.

[0618] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0619] In this invention, the server includes means for acquiring audio information during a meeting as data, means for converting the acquired audio information into text information in real time, means for analyzing the converted text information to identify the context of the discussion and the intentions of the participants, means for recognizing the emotions contained in the analyzed text information and storing them as data, means for generating feedback to optimize the progress of the meeting based on the analysis results, and means for compiling the emotion data and analysis results after the meeting to create a meeting record and an activity plan. This enables the provision of feedback that takes into account the emotions of the participants in real time, and makes it possible to grasp intentions and optimize the progress of the meeting using emotion data.

[0620] "Audio information" refers to a series of sound data collected during a meeting, used to record the progress of the meeting and the statements made.

[0621] "Textual information" refers to text data converted in real time based on audio information, and is used to visually recognize and analyze the content of meetings.

[0622] "Context" refers to the linguistic background information of statements made during a meeting, as well as the circumstances and flow in which those statements were made, and is of great importance to the progress of the discussion.

[0623] "Intention" refers to the purpose or aim that a speaker is trying to convey through their words, and it is a factor that influences the direction of the discussion.

[0624] "Emotions" are psychological states underpinned by words and actions, such as expressing feelings like joy, anger, and sadness.

[0625] "Feedback" refers to opinions and suggestions provided to participants based on analyzed data to optimize the flow of a meeting, and is important information for indicating areas for improvement.

[0626] "Meeting minutes" are documents that summarize the discussions, statements, and decisions made during a meeting, and are used as reference material for future reference.

[0627] An "activity plan" is a document that outlines specific future action plans and schedules based on the results of a meeting, and it supports the organization's strategic execution.

[0628] "Data" refers to facts and information that have been measured, collected, and analyzed, and includes numerical values, text, and other information that are processed and stored within a system.

[0629] This system optimizes meeting progress by analyzing audio during meetings in real time, interpreting participants' emotions and intentions from their statements, and providing feedback.

[0630] The terminal uses a built-in high-performance microphone to capture meeting audio information in real time. A dedicated application is installed that has the ability to continuously monitor audio input and acquire data. Audio data is transmitted to the server immediately using a low-latency protocol (e.g., WebSocket).

[0631] The server converts the received audio information into text using a speech recognition engine (e.g., a speech recognition API from a major cloud service provider). Furthermore, the server uses an emotion analysis engine and natural language processing technology (e.g., a natural language model API from a commercial AI provider) to identify emotions from the spoken content and to deeply analyze the context of the discussion and the intentions of the participants.

[0632] The analyzed information is used to generate feedback that takes into account the emotional data contained in each statement and the flow of the discussion. This feedback is provided to participants in real time and serves as an important indicator for the progress of the meeting. After the meeting, meeting records and activity plans are automatically generated based on the accumulated data and saved to shared network storage.

[0633] As a concrete example, in a meeting about the market strategy for a new product, if a participant expresses concerns about the product's design, the system analyzes the emotion of "anxiety" contained in their statement and provides feedback such as "further design evaluation is needed." This feedback is instantly displayed to the participants, allowing everyone to share their concerns.

[0634] An example of a prompt could be, "In a meeting about a new product launch, please tell me how to identify participants' anxieties and concerns and generate appropriate feedback." This system makes it easier for users to understand the flow of emotions and obtain information to make better decisions.

[0635] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0636] Step 1:

[0637] The terminal captures audio in real time using its built-in microphone as soon as the meeting starts. The input is ambient sound, and this audio is sent to the server as data. Low-latency protocols are used to efficiently capture audio.

[0638] Step 2:

[0639] The server receives audio data sent from the terminal. The input is audio data. This data is input to the speech recognition engine and processed to convert it into text information. The output is text information in which the spoken content has been converted into text format.

[0640] Step 3:

[0641] The server inputs the obtained text information into the sentiment analysis engine. The input is text information. Here, the engine identifies emotions and generates sentiment data such as "joy," "anger," and "sadness." The output is the sentiment data corresponding to each statement.

[0642] Step 4:

[0643] The server analyzes the context of a discussion and the intentions of the participants using textual and sentiment data. The input consists of textual and sentiment data. It utilizes a generative AI model to clarify participants' intentions and identify which statements are important. The output is the analysis results, including the discussion context and participants' intentions.

[0644] Step 5:

[0645] The server generates real-time feedback based on the analysis results. The input is the context and sentiment data of the discussion. Based on this, it extracts suggestions and points to note for improving the meeting's progress and creates a feedback message as output.

[0646] Step 6:

[0647] The terminal displays the generated feedback to participants in real time. The input is feedback messages from the server. In operation, it visually displays feedback and changes in emotion on the display to help participants respond immediately.

[0648] Step 7:

[0649] The server integrates all data after the meeting and creates meeting minutes and an action plan. Inputs include all spoken statements and associated sentiment data. It integrates and summarizes the data, generating meeting minutes and an action plan outlining the next steps as output.

[0650] (Application Example 2)

[0651] 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."

[0652] Conventional conversation analysis systems only identify context and understand intent based on the content of the conversation, and do not adequately provide real-time analysis and feedback on the emotional state of participants. Therefore, there is a need for a mechanism that facilitates appropriate responses in situations where emotions such as tension and stress are abnormally heightened. In particular, in workplace environments, there is a need to detect sudden changes in emotions early and respond appropriately.

[0653] 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.

[0654] In this invention, the server includes means for acquiring audio data of a conversation, means for converting the acquired audio data into text data in real time, means for analyzing the converted text data to identify the context of the discussion and the intentions of the participants, means for generating feedback based on the analysis results and participant sentiment analysis to improve the progress of the conversation, means for providing the generated feedback to the participants in real time, means for creating a record and plan after the conversation has ended based on the analysis results and sentiment data, and means for detecting abnormal sentiment levels and issuing an alarm based on sentiment data. This makes it possible to quickly detect changes in sentiment and take immediate action as needed.

[0655] "Conversation audio data" refers to data that electronically records participants' speech in its original form.

[0656] "Real-time conversion to text data" refers to the process of instantly converting acquired audio data into text information.

[0657] "The context of the discussion" refers to information that indicates the topic and progress of a conversation.

[0658] "Participant's intent" refers to the speaker's purpose or information that expresses their objective in a conversation.

[0659] "Generating feedback" means creating information that suggests areas for improvement and action guidelines based on the analysis results.

[0660] "Providing feedback to participants in real time" means notifying participants of the generated feedback immediately.

[0661] "Creating records and plans" means organizing the data obtained after the conversation and developing documents and guidelines for future actions that can be referenced later.

[0662] "Emotional data" refers to information that indicates the emotional state of a person, extracted from the content of their speech.

[0663] "Detecting abnormal emotional levels" means identifying a state in which a participant's emotions deviate from the normal range.

[0664] "Issuing an alarm" means notifying relevant parties to take precautions when an anomaly is detected.

[0665] The system that implements this application includes a program that acquires and analyzes conversational audio data in real time. This system mainly consists of three elements: a server, a terminal, and a user.

[0666] The server's role is to receive audio data transmitted from terminals and convert it into text data using speech recognition software. Specifically, the audio data is transcribed using Microsoft Azure's speech recognition API, and then analyzed as emotion data using Amazon AWS's emotion analysis API. This data analysis clarifies the emotional state and intentions of the conversation participants and generates real-time feedback based on this. Furthermore, if an abnormally high emotional level is detected, a mechanism is in place to issue an alert to designated parties.

[0667] The terminal is a smart device worn by the user (e.g., smart glasses), which is used to capture audio data. This audio data is immediately transferred to a server for analysis. The user receives feedback that is displayed in real time based on the analysis results, providing information that is useful for facilitating meetings and discussions.

[0668] For example, if an employee's stress level becomes abnormally high during a workplace meeting, the system can detect this and notify the mental health officer, enabling early follow-up. This system is expected to improve the workplace environment and maintain employees' mental health.

[0669] Examples of prompt messages include the following:

[0670] "I want to develop a system that analyzes emotions from the following audio data and immediately notifies users if their stress levels are abnormal. Please provide specific program code and examples of API usage."

[0671] This makes it possible to appropriately manage the emotional state of participants in a conversational environment and promote good communication.

[0672] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0673] Step 1:

[0674] The device captures audio data of the conversation. During this process, the smart device records the surrounding conversation, and the acquired audio is prepared as digital data in its original format. The input is ambient sound, and the output is digital audio data.

[0675] Step 2:

[0676] The terminal transfers the captured audio data to the server. The audio data is sent to and received by the server in real time using a secure protocol. The input is digital audio data, and the output is the audio data sent to the server.

[0677] Step 3:

[0678] The server converts the received audio data into text data using speech recognition software. Here, Microsoft Azure's speech recognition API is used to process the audio into text information. The input is audio data, and the output is text data.

[0679] Step 4:

[0680] The server analyzes the received text data using Amazon AWS sentiment analysis API and generates sentiment data for each statement. The input is text data, and the output after data analysis is sentiment data. In this process, the user's emotional state is classified into multiple categories such as "surprise," "joy," and "fear."

[0681] Step 5:

[0682] The server uses the analysis results to construct feedback generated from the conversation context and sentiment data, and sends it to the terminal. It compiles information useful for facilitating the process for each participant. The input is the analyzed text and sentiment data, and the output is feedback data.

[0683] Step 6:

[0684] The device displays feedback received by the user in real time. Based on this information, the user can understand the progress of the conversation and their own or others' emotional states. The input is feedback data, and the output is the information displayed on the user's screen.

[0685] Step 7:

[0686] At the end of each session, the server automatically creates a record and plan based on the retained data. This record integrates sentiment data and feedback, and includes an action plan for the next conversation. Inputs are the text and sentiment data recorded during the session, and outputs are the recorded document and the plan for the next steps.

[0687] 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.

[0688] 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.

[0689] 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.

[0690] [Fourth Embodiment]

[0691] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0692] 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.

[0693] 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).

[0694] 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.

[0695] 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.

[0696] 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).

[0697] 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.

[0698] 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.

[0699] 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.

[0700] 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.

[0701] 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.

[0702] 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.

[0703] 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".

[0704] This invention provides technology for conducting meetings and discussions more efficiently, based on an AI system for analyzing discussion intent. Specifically, it accurately grasps the content of meetings and generates feedback by converting audio during meetings into text in real time and analyzing the context of the discussion.

[0705] Acquisition and conversion of audio data

[0706] When a meeting begins, the terminal uses its built-in microphone to acquire audio. The acquired audio data is compressed using compression technology and sent to the server via the internet.

[0707] The server uses a speech recognition engine to convert speech data into text data. It also implements processes that enhance pronunciation clarity by utilizing noise reduction and speech enhancement technologies.

[0708] Text data analysis

[0709] The server analyzes text data using natural language processing techniques. A contextual analysis engine identifies the speaker's intent and the topic of discussion, and extracts important keywords.

[0710] To classify each utterance by speaker, speaker identification technology is applied. This is achieved using a model that learns unique linguistic patterns contained in the text.

[0711] Generating feedback and instructions

[0712] Based on the analysis results, the server generates feedback and specific instructions to improve the meeting's progress. This may include changing topics that have been discussed too much or identifying unresolved issues.

[0713] The generated feedback is immediately sent to the device and presented to the user visually or audibly.

[0714] Creating meeting minutes and action plans

[0715] After the meeting concludes, the server integrates the accumulated data to create reviewable meeting minutes and a clear action plan. This includes summaries of each topic discussed and conclusions, as well as specific actions recommended as next steps.

[0716] The completed meeting minutes and action plan are saved to cloud storage, making them accessible to users and shareable with other participants as needed.

[0717] Specific example

[0718] For example, in a product development meeting at a certain company, this system transcribes participants' comments in real time, identifying topics such as changes in product specifications, development schedules, and marketing strategies. If a comment such as "The release date of this product may be delayed" is made during the meeting, the system picks up on this comment and presents it as feedback suggesting a review of the milestones. After the meeting, meeting minutes accessible to all participants are provided on the cloud. This allows participants to review the tasks and actions they need to prepare for the next meeting.

[0719] The following describes the processing flow.

[0720] Step 1:

[0721] The device captures meeting audio in real time using its built-in microphone. The audio data is compressed and sent to the server over the network. A noise reduction algorithm is applied to optimize audio quality.

[0722] Step 2:

[0723] The server inputs the received audio data into a speech recognition engine and converts it into text data. Deep learning technology is used for speech recognition, and the speaker's accent and vocabulary are analyzed to improve accuracy.

[0724] Step 3:

[0725] The server uses a natural language processing engine to analyze the generated text data. This involves contextual analysis, keyword extraction, and sentiment analysis to identify the intent behind participants' statements and the topics of discussion.

[0726] Step 4:

[0727] Based on the results of the text analysis, the server generates feedback and instructions for the next steps to improve the progress of the meeting. The generated information is immediately sent to the terminal and presented to meeting participants visually or audibly.

[0728] Step 5:

[0729] The terminal displays feedback received from the server in the user interface. Users use this information to make decisions that will advance the discussion.

[0730] Step 6:

[0731] After the meeting concludes, the server integrates all analytical data to create detailed meeting minutes and an action plan for the next meeting. This includes a summary of the discussion, key conclusions, and recommended actions.

[0732] Step 7:

[0733] The final meeting minutes and action plan will be saved to cloud storage and made accessible to users. Users will use this information to prepare for and follow up on future meetings.

[0734] (Example 1)

[0735] 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".

[0736] Meetings and discussions often face challenges such as misunderstandings of what is said and the intentions behind it, leading to stalled progress. Furthermore, it is difficult to quickly and accurately create meeting minutes and action plans afterward. This can result in insufficient information sharing among participants and delays in decision-making.

[0737] 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.

[0738] In this invention, the server includes means for acquiring speech during a meeting as audio information, means for converting the acquired audio information into text information in real time, means for analyzing the converted text information to identify the context of the discussion and the intentions of the participants, means for applying a noise reduction filter and voice enhancement to the audio information, means for generating instructions to improve the progress of the meeting based on the analysis results, means for immediately providing the generated instructions to the participants, and means for compiling the analysis results after the meeting to create a report and an action plan. This improves the efficiency of meetings and enables rapid and accurate information sharing and decision-making.

[0739] "Audio information" refers to data recorded in digital format from speeches such as those heard in meetings or conversations.

[0740] "Acquiring" refers to the act of collecting audio information from a device.

[0741] "Textual information" refers to string data that has been converted from audio information.

[0742] "Real-time" means that processing and conversion occur simultaneously with the utterance.

[0743] "Analysis" refers to the process of processing information based on converted textual data to understand its context and intent.

[0744] "Context" refers to the framework that gives meaning and relevance within a statement or dialogue.

[0745] A "noise reduction filter" refers to a technology that reduces unwanted background noise from audio information.

[0746] "Voice enhancement" refers to the process of adjusting sound to make necessary sounds in audio information easier to hear.

[0747] "Instructions" refers to specific suggestions or commands generated based on the analysis results to support the progress of the meeting.

[0748] A "report" refers to a document that summarizes the content of discussions, conclusions, and action plans during a meeting.

[0749] An "action plan" refers to specific guidelines or procedures developed based on the matters discussed after the meeting.

[0750] "Participants" refers to those who speak or participate in a meeting or discussion.

[0751] A specific description will be given of embodiments for carrying out this invention.

[0752] This discussion intent analysis AI system aims to facilitate efficient meeting management and record-keeping. First, the terminal uses its built-in microphone to capture speech during the meeting as audio information. The audio information is converted into a digital format and processed efficiently through data compression technology. This data is then transmitted to a server via the internet.

[0753] The server uses an advanced speech recognition engine to convert speech information into text in real time. This conversion process is enhanced with noise reduction filters and voice enhancement techniques to improve accuracy. The converted text is time-stamped to maintain the order of speech.

[0754] Next, the server uses natural language processing (NLP) techniques to analyze the textual information. This analysis identifies the context of the discussion and the intentions of the speakers. Machine learning models are also used to identify speakers and classify each statement by speaker. This enables detailed tracking and analysis of the discussion.

[0755] Based on the analysis results, the server generates instructions to support the progress of the meeting. These instructions are presented to the user from the terminal as digital or audio output, providing real-time feedback.

[0756] After the meeting ends, the server automatically generates a report and action plan based on the data accumulated so far. The generated report is saved to cloud storage, making it easily accessible to the user and shareable with other participants.

[0757] As a concrete example, in a product development meeting, this system immediately picks up on a statement such as "The product launch date may be delayed" and instructs the system to re-evaluate milestones as appropriate feedback. An example of a prompt to be input into the generating AI model is, "Transcribe the statements made in this meeting in real time, analyze the speaker's intent, and generate feedback."

[0758] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0759] Step 1:

[0760] The device acquires audio during a meeting via its built-in microphone. This audio information is converted into a digital signal, and the data size is reduced using compression technology (e.g., MP3). The input is raw audio, and the output is compressed digital audio data.

[0761] Step 2:

[0762] Compressed audio data is transmitted to a server via the internet. The server utilizes a speech recognition engine to convert this data into text. The input is compressed digital audio data, and the output is text information with a timestamp. Noise reduction and speech enhancement algorithms are applied to optimize sound quality.

[0763] Step 3:

[0764] The server analyzes the converted text information using natural language processing (NLP) techniques. It performs contextual analysis and keyword extraction to identify participants' intentions and discussion topics. The input is text information, and the output is contextual data as a result of the analysis, extracted keywords, and speaker identification information.

[0765] Step 4:

[0766] Based on the analysis results, the server generates instructions to improve the meeting progress. This involves intent analysis using prompt sentences and utilizing a generative AI model to create specific feedback. The input is the analysis results from step 3, and the output is instructions and feedback presented in real time.

[0767] Step 5:

[0768] The generated instructions and feedback are provided to the user through the device. These are presented either on a screen or verbally using speech synthesis technology. The input is the instructions from step 4, and the output is the visual or auditory information received by the user.

[0769] Step 6:

[0770] After the meeting concludes, the server integrates all data to create a report and action plan. This process includes summarizing each generated instruction and statement. Input is data from all sessions, and output is the report and action plan as documents.

[0771] Step 7:

[0772] Finally, the generated report and execution plan are saved to cloud storage and configured to be accessible to the user. This process is for saving the output of step 6 to the cloud.

[0773] (Application Example 1)

[0774] 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".

[0775] A challenge arises when information sharing among staff at a store is not conducted quickly and accurately, as this can hinder customer service and operational efficiency. This problem stems particularly from the inability to respond quickly to real-time changes in customer requests and operational tasks, potentially resulting in a decline in service quality.

[0776] 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.

[0777] In this invention, the server includes a device for acquiring speech as audio information, a device for immediately converting the acquired audio information into text information, and a device for analyzing the converted text information to identify the background of the discussion and the intentions of the participants. This enables immediate information sharing among staff, real-time acquisition of information necessary for customer service and business operations, and efficient business operations.

[0778] "Speech" refers to linguistic expression used by people to convey intentions and information through sound.

[0779] "Audio information" refers to data obtained by acquiring and processing sound as a digital signal, making it analyzable.

[0780] "Textual information" refers to text data converted from audio information, and is used for analysis.

[0781] A "device" is a system of machinery or software designed to perform a specific function.

[0782] "The background of the discussion" refers to the context and situation in which participants speak, and it serves as the foundation for understanding their intentions.

[0783] "Instantly" means that processing or results can be obtained in near real-time with no delay.

[0784] "Analysis" is the process of organizing and breaking down information to understand its structure and meaning.

[0785] An "information terminal" is an electronic device used to display and manipulate data.

[0786] A "record document" is a formal written document in which information is organized and preserved for later reference.

[0787] An "action plan" is a detailed document outlining specific steps and methods for achieving a particular objective.

[0788] "Memory space" refers to digital space used to store data and information.

[0789] The system that implements this application consists of multiple components. The server uses a combination of appropriate hardware and software to instantly acquire and process audio information. Specifically, it uses a mobile device such as a smartphone or tablet to acquire audio, utilizing the microphone built into it. The device compresses the acquired audio information and transmits it to the server via the network.

[0790] The server uses an advanced speech recognition engine to convert audio information into text. This process leverages the Google Cloud Speech-to-Text API, improving recognition accuracy through noise reduction and speech enhancement techniques. The server then analyzes the converted text, using natural language processing techniques to identify the context of the discussion and the participants' intentions. This utilizes natural language processing libraries such as spaCy.

[0791] The analysis results are provided to the user as real-time feedback via an information terminal. Users can use this interface to immediately check the information and take necessary actions. After completion, the analysis data is compiled into a record document and action plan, and saved to cloud storage. Users can access and share this information at any time via their digital devices. Google Cloud Platform is used for the cloud service.

[0792] As a concrete example, consider information sharing regarding the arrival date of new products at a store. Using this system, a voice command such as "Check when the new products will arrive and inform everyone" would be instantly analyzed, and the results would be shared with the staff. This would allow each staff member to quickly prepare countermeasures and improve the quality of service provided to customers.

[0793] An example of a prompt for a generative AI model is: "Explain how to build a system that extracts important business tasks from store staff conversations and provides real-time feedback."

[0794] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0795] Step 1:

[0796] The device acquires conversational audio information using a microphone. The audio data, as input, is captured directly from the microphone as raw sound. The device then uses a compression algorithm to reduce the size of the audio data and convert it into a format suitable for transfer to the server. This process reduces the data size while maintaining the quality of the audio signal.

[0797] Step 2:

[0798] The server receives audio data sent from the terminal. The input is compressed audio data. The server uses the Google Cloud Speech-to-Text API to convert the audio data into text. During the conversion process, it uses techniques to suppress noise and emphasize important audio components. The output is text data that reflects the content of the discussion.

[0799] Step 3:

[0800] The server analyzes the text data. The input is the text data generated in step 2. It utilizes natural language processing techniques, specifically the spaCy library, to analyze the background of the discussion and the intentions of the participants. It extracts keywords and important contexts and structures the information based on them. The output is metadata of the analysis results, i.e., a set of identified intentions and keywords.

[0801] Step 4:

[0802] The server generates feedback based on the analysis results. The input is the metadata obtained in step 3. This generates effective feedback to improve the meeting's progress. Using a generation AI model, specific improvement suggestions and points are created, and the results are immediately transmitted to the information terminal. The output is the proposed feedback message and points.

[0803] Step 5:

[0804] Users receive this feedback through their information terminals and use it in real time to aid in decision-making. Users can modify their actions based on the feedback as needed, streamlining their work processes. Input is feedback messages, and output is improvement measures and corresponding actions.

[0805] Step 6:

[0806] The server generates a record document and an action plan based on the data analyzed after completion. The input is the data processed in all previous steps. The record document includes an overview of the meeting, a summary of the discussions, and decisions made, while the action plan outlines specific next steps and responsibilities. The output is the saved record document and action plan.

[0807] Step 7:

[0808] Users access record documents and action plans stored in cloud storage, and individually review and share them. Google Cloud Platform is used as the cloud service. Inputs are record documents and action plans, and outputs are the standardization of work through checking and sharing of verified information.

[0809] 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.

[0810] This invention provides a system characterized by its ability to analyze intentions and provide feedback during discussions, as well as to recognize participants' emotions. By incorporating an emotion engine, it enables real-time analysis of user emotions from speech during meetings and allows for the use of the results to improve the progress of the meeting.

[0811] Acquisition of voice data and sentiment analysis

[0812] The terminal begins capturing audio in real time as soon as the meeting starts and transmits it to the server. The audio data is then passed to the emotion engine simultaneously with speech recognition.

[0813] The server converts the received audio data into text using a speech recognition engine and analyzes the emotions contained in the utterances using an emotion engine. It identifies emotions such as joy, anger, and sadness from individual utterances and stores them as data.

[0814] Text data analysis and feedback generation

[0815] The server analyzes text data using natural language processing techniques to identify the context of the discussion and the intentions of the participants. This analysis is then combined with sentiment data to generate feedback that optimizes the meeting's progress.

[0816] The analyzed emotional data is also included as an important factor in creating meeting minutes and developing action plans. In particular, in situations where discussions are heated or there is no agreement, emotional data plays a role in regulating the progress of the meeting.

[0817] Real-time emotion change notifications and meeting minute creation

[0818] The terminal displays feedback and sentiment analysis results sent from the server to the user in real time. This allows the user to make decisions based not only on the context of the discussion but also on the flow of emotions.

[0819] After the meeting concludes, the server integrates all data to create meeting minutes and an action plan. Sentimental data is used as an indicator to assess the impact of each statement, contributing to an objective meeting record.

[0820] Specific example

[0821] For example, in a meeting about the launch of a new product, if a participant expresses strong concerns about market reaction, this system identifies the emotions such as "anxiety" and "concern" contained in their statement. The analysis results in feedback such as "The team needs to discuss this concern further," which is immediately made public to other participants. Subsequently, at the end of the meeting, detailed minutes are created, including how the concerns were discussed and what was decided as the next steps. Sentimental data is an important indicator of which topics generated more emotional discussion.

[0822] The following describes the processing flow.

[0823] Step 1:

[0824] The terminal uses a microphone to capture audio in real time at the start of the meeting. The captured audio data is compressed and then sent to the server. Noise cancellation technology is applied to ensure audio quality.

[0825] Step 2:

[0826] The server converts the received audio data into text data using a speech recognition engine. The speech recognition uses a highly accurate deep learning model to handle a wide variety of speech patterns.

[0827] Step 3:

[0828] The server passes the audio data to the emotion engine, which analyzes the user's emotions from the conversation. It identifies the emotional tone in each utterance, such as joy, anger, or sadness, and stores this data along with other analysis data.

[0829] Step 4:

[0830] The server analyzes text data using natural language processing algorithms to identify the context of the discussion and the intentions of the participants. Sentiment data plays a crucial role in this analysis, deepening the understanding of how the discussion is progressing.

[0831] Step 5:

[0832] The server generates feedback based on the analysis results to improve the members' progress. This includes summaries of key points from the discussion and instructions on the direction of the meeting based on sentiment analysis.

[0833] Step 6:

[0834] The device displays generated feedback and real-time analyzed sentiment data to the user. This allows the user to understand the emotional state of the participants while facilitating the discussion.

[0835] Step 7:

[0836] After the meeting concludes, the server integrates all data to create detailed minutes and action plans. Sentimental data is used to assess the impact of the dialogue and show how the discussion evolved.

[0837] Step 8:

[0838] The completed meeting minutes and action plans are saved to cloud storage and made accessible to users as needed. This allows meeting participants to clearly define their next steps and follow up efficiently.

[0839] (Example 2)

[0840] 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".

[0841] Traditional meeting systems made it difficult to provide feedback that took participants' emotions into account, potentially influencing the progress of discussions and decision-making processes. Furthermore, it was challenging to quickly and accurately analyze individual statements and their intentions, and to effectively utilize emotional data. This resulted in decreased meeting efficiency and productivity.

[0842] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0843] In this invention, the server includes means for acquiring audio information during a meeting as data, means for converting the acquired audio information into text information in real time, means for analyzing the converted text information to identify the context of the discussion and the intentions of the participants, means for recognizing the emotions contained in the analyzed text information and storing them as data, means for generating feedback to optimize the progress of the meeting based on the analysis results, and means for compiling the emotion data and analysis results after the meeting to create a meeting record and an activity plan. This enables the provision of feedback that takes into account the emotions of the participants in real time, and makes it possible to grasp intentions and optimize the progress of the meeting using emotion data.

[0844] "Audio information" refers to a series of sound data collected during a meeting, used to record the progress of the meeting and the statements made.

[0845] "Textual information" refers to text data converted in real time based on audio information, and is used to visually recognize and analyze the content of meetings.

[0846] "Context" refers to the linguistic background information of statements made during a meeting, as well as the circumstances and flow in which those statements were made, and is of great importance to the progress of the discussion.

[0847] "Intention" refers to the purpose or aim that a speaker is trying to convey through their words, and it is a factor that influences the direction of the discussion.

[0848] "Emotions" are psychological states underpinned by words and actions, such as expressing feelings like joy, anger, and sadness.

[0849] "Feedback" refers to opinions and suggestions provided to participants based on analyzed data to optimize the flow of a meeting, and is important information for indicating areas for improvement.

[0850] "Meeting minutes" are documents that summarize the discussions, statements, and decisions made during a meeting, and are used as reference material for future reference.

[0851] An "activity plan" is a document that outlines specific future action plans and schedules based on the results of a meeting, and it supports the organization's strategic execution.

[0852] "Data" refers to facts and information that have been measured, collected, and analyzed, and includes numerical values, text, and other information that are processed and stored within a system.

[0853] This system optimizes meeting progress by analyzing audio during meetings in real time, interpreting participants' emotions and intentions from their statements, and providing feedback.

[0854] The terminal uses a built-in high-performance microphone to capture meeting audio information in real time. A dedicated application is installed that has the ability to continuously monitor audio input and acquire data. Audio data is transmitted to the server immediately using a low-latency protocol (e.g., WebSocket).

[0855] The server converts the received audio information into text using a speech recognition engine (e.g., a speech recognition API from a major cloud service provider). Furthermore, the server uses an emotion analysis engine and natural language processing technology (e.g., a natural language model API from a commercial AI provider) to identify emotions from the spoken content and to deeply analyze the context of the discussion and the intentions of the participants.

[0856] The analyzed information is used to generate feedback that takes into account the emotional data contained in each statement and the flow of the discussion. This feedback is provided to participants in real time and serves as an important indicator for the progress of the meeting. After the meeting, meeting records and activity plans are automatically generated based on the accumulated data and saved to shared network storage.

[0857] As a concrete example, in a meeting about the market strategy for a new product, if a participant expresses concerns about the product's design, the system analyzes the emotion of "anxiety" contained in their statement and provides feedback such as "further design evaluation is needed." This feedback is instantly displayed to the participants, allowing everyone to share their concerns.

[0858] An example of a prompt could be, "In a meeting about a new product launch, please tell me how to identify participants' anxieties and concerns and generate appropriate feedback." This system makes it easier for users to understand the flow of emotions and obtain information to make better decisions.

[0859] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0860] Step 1:

[0861] The terminal captures audio in real time using its built-in microphone as soon as the meeting starts. The input is ambient sound, and this audio is sent to the server as data. Low-latency protocols are used to efficiently capture audio.

[0862] Step 2:

[0863] The server receives audio data sent from the terminal. The input is audio data. This data is input to the speech recognition engine and processed to convert it into text information. The output is text information in which the spoken content has been converted into text format.

[0864] Step 3:

[0865] The server inputs the obtained text information into the sentiment analysis engine. The input is text information. Here, the engine identifies emotions and generates sentiment data such as "joy," "anger," and "sadness." The output is the sentiment data corresponding to each statement.

[0866] Step 4:

[0867] The server analyzes the context of a discussion and the intentions of the participants using textual and sentiment data. The input consists of textual and sentiment data. It utilizes a generative AI model to clarify participants' intentions and identify which statements are important. The output is the analysis results, including the discussion context and participants' intentions.

[0868] Step 5:

[0869] The server generates real-time feedback based on the analysis results. The input is the context and sentiment data of the discussion. Based on this, it extracts suggestions and points to note for improving the meeting's progress and creates a feedback message as output.

[0870] Step 6:

[0871] The terminal displays the generated feedback to participants in real time. The input is feedback messages from the server. In operation, it visually displays feedback and changes in emotion on the display to help participants respond immediately.

[0872] Step 7:

[0873] The server integrates all data after the meeting and creates meeting minutes and an action plan. Inputs include all spoken statements and associated sentiment data. It integrates and summarizes the data, generating meeting minutes and an action plan outlining the next steps as output.

[0874] (Application Example 2)

[0875] 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".

[0876] Conventional conversation analysis systems only identify context and understand intent based on the content of the conversation, and do not adequately provide real-time analysis and feedback on the emotional state of participants. Therefore, there is a need for a mechanism that facilitates appropriate responses in situations where emotions such as tension and stress are abnormally heightened. In particular, in workplace environments, there is a need to detect sudden changes in emotions early and respond appropriately.

[0877] 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.

[0878] In this invention, the server includes means for acquiring audio data of a conversation, means for converting the acquired audio data into text data in real time, means for analyzing the converted text data to identify the context of the discussion and the intentions of the participants, means for generating feedback based on the analysis results and participant sentiment analysis to improve the progress of the conversation, means for providing the generated feedback to the participants in real time, means for creating a record and plan after the conversation has ended based on the analysis results and sentiment data, and means for detecting abnormal sentiment levels and issuing an alarm based on sentiment data. This makes it possible to quickly detect changes in sentiment and take immediate action as needed.

[0879] "Conversation audio data" refers to data that electronically records participants' speech in its original form.

[0880] "Real-time conversion to text data" refers to the process of instantly converting acquired audio data into text information.

[0881] "The context of the discussion" refers to information that indicates the topic and progress of a conversation.

[0882] "Participant's intent" refers to the speaker's purpose or information that expresses their objective in a conversation.

[0883] "Generating feedback" means creating information that suggests areas for improvement and action guidelines based on the analysis results.

[0884] "Providing feedback to participants in real time" means notifying participants of the generated feedback immediately.

[0885] "Creating records and plans" means organizing the data obtained after the conversation and developing documents and guidelines for future actions that can be referenced later.

[0886] "Emotional data" refers to information that indicates the emotional state of a person, extracted from the content of their speech.

[0887] "Detecting abnormal emotional levels" means identifying a state in which a participant's emotions deviate from the normal range.

[0888] "Issuing an alarm" means notifying relevant parties to take precautions when an anomaly is detected.

[0889] The system that implements this application includes a program that acquires and analyzes conversational audio data in real time. This system mainly consists of three elements: a server, a terminal, and a user.

[0890] The server's role is to receive audio data transmitted from terminals and convert it into text data using speech recognition software. Specifically, the audio data is transcribed using Microsoft Azure's speech recognition API, and then analyzed as emotion data using Amazon AWS's emotion analysis API. This data analysis clarifies the emotional state and intentions of the conversation participants and generates real-time feedback based on this. Furthermore, if an abnormally high emotional level is detected, a mechanism is in place to issue an alert to designated parties.

[0891] The terminal is a smart device worn by the user (e.g., smart glasses), which is used to capture audio data. This audio data is immediately transferred to a server for analysis. The user receives feedback that is displayed in real time based on the analysis results, providing information that is useful for facilitating meetings and discussions.

[0892] For example, if an employee's stress level becomes abnormally high during a workplace meeting, the system can detect this and notify the mental health officer, enabling early follow-up. This system is expected to improve the workplace environment and maintain employees' mental health.

[0893] Examples of prompt messages include the following:

[0894] "I want to develop a system that analyzes emotions from the following audio data and immediately notifies users if their stress levels are abnormal. Please provide specific program code and examples of API usage."

[0895] This makes it possible to appropriately manage the emotional state of participants in a conversational environment and promote good communication.

[0896] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0897] Step 1:

[0898] The device captures audio data of the conversation. During this process, the smart device records the surrounding conversation, and the acquired audio is prepared as digital data in its original format. The input is ambient sound, and the output is digital audio data.

[0899] Step 2:

[0900] The terminal transfers the captured audio data to the server. The audio data is sent to and received by the server in real time using a secure protocol. The input is digital audio data, and the output is the audio data sent to the server.

[0901] Step 3:

[0902] The server converts the received audio data into text data using speech recognition software. Here, Microsoft Azure's speech recognition API is used to process the audio into text information. The input is audio data, and the output is text data.

[0903] Step 4:

[0904] The server analyzes the received text data using Amazon AWS sentiment analysis API and generates sentiment data for each statement. The input is text data, and the output after data analysis is sentiment data. In this process, the user's emotional state is classified into multiple categories such as "surprise," "joy," and "fear."

[0905] Step 5:

[0906] The server uses the analysis results to construct feedback generated from the conversation context and sentiment data, and sends it to the terminal. It compiles information useful for facilitating the process for each participant. The input is the analyzed text and sentiment data, and the output is feedback data.

[0907] Step 6:

[0908] The device displays feedback received by the user in real time. Based on this information, the user can understand the progress of the conversation and their own or others' emotional states. The input is feedback data, and the output is the information displayed on the user's screen.

[0909] Step 7:

[0910] At the end of each session, the server automatically creates a record and plan based on the retained data. This record integrates sentiment data and feedback, and includes an action plan for the next conversation. Inputs are the text and sentiment data recorded during the session, and outputs are the recorded document and the plan for the next steps.

[0911] 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.

[0912] 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.

[0913] 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.

[0914] 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.

[0915] 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.

[0916] 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.

[0917] 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.

[0918] 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.

[0919] 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."

[0920] 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.

[0921] 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.

[0922] 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.

[0923] 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.

[0924] 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.

[0925] 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.

[0926] 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.

[0927] 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.

[0928] 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.

[0929] 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.

[0930] 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.

[0931] 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 to be incorporated by reference.

[0932] The following is further disclosed regarding the embodiments described above.

[0933] (Claim 1)

[0934] A means of acquiring speech during a meeting as audio data,

[0935] A means of converting acquired audio data into text data in real time,

[0936] A means of analyzing the converted text data to identify the context of the discussion and the intentions of the participants,

[0937] A means of generating feedback to improve the progress of meetings based on the analysis results,

[0938] A means of providing the generated feedback to participants in real time,

[0939] A system that includes a means of compiling analysis results after a meeting and creating meeting minutes and an action plan.

[0940] (Claim 2)

[0941] The system according to claim 1, comprising means for identifying the speaker from acquired audio data and classifying each statement by speaker.

[0942] (Claim 3)

[0943] The system according to claim 1, comprising means for storing and making accessible the feedback of analysis results and action plans in cloud storage.

[0944] "Example 1"

[0945] (Claim 1)

[0946] A means of acquiring speech during a meeting as audio information,

[0947] A means of converting acquired audio information into text information in real time,

[0948] A means of analyzing the converted textual information to identify the context of the discussion and the intentions of the participants,

[0949] A method for applying a noise reduction filter and voice enhancement to audio information,

[0950] A means for generating instructions to improve the progress of a meeting based on the analysis results,

[0951] A means of immediately providing the generated instructions to participants,

[0952] A system that includes a means of compiling analysis results after a meeting and creating a report and action plan.

[0953] (Claim 2)

[0954] The system according to claim 1, comprising means for identifying speakers from acquired audio information and classifying each utterance by speaker.

[0955] (Claim 3)

[0956] The system according to claim 1, comprising means for storing and making accessible the analysis results, instructions, and execution plan in a cloud storage device.

[0957] "Application Example 1"

[0958] (Claim 1)

[0959] A device that acquires speech as audio information,

[0960] A device that instantly converts acquired audio information into text information,

[0961] A device that analyzes converted text information to identify the background of the discussion and the intentions of the participants,

[0962] A device that generates notifications to improve progress based on analysis results,

[0963] A device that immediately provides the generated notification,

[0964] A device that, after completion, compiles the analysis results and creates a record document and action plan,

[0965] A system that includes a device for sharing instant notifications on information terminals.

[0966] (Claim 2)

[0967] The system according to claim 1, comprising a device for identifying speakers from acquired audio information and classifying each statement by speaker.

[0968] (Claim 3)

[0969] The system according to claim 1, comprising a device for storing and making accessible the notification of analysis results and action plans in a memory area.

[0970] "Example 2 of combining an emotion engine"

[0971] (Claim 1)

[0972] A means of acquiring audio information during a meeting as data,

[0973] A means of converting acquired audio information into text information in real time,

[0974] A means of analyzing the converted text information to identify the context of the discussion and the intentions of the participants,

[0975] A means of recognizing the emotions contained in the analyzed textual information and storing them as data,

[0976] A means for generating feedback to optimize the progress of a meeting based on the analysis results,

[0977] A means of providing the generated feedback to participants in real time,

[0978] A system that includes a means of compiling emotional data and analysis results after a meeting to create meeting records and action plans.

[0979] (Claim 2)

[0980] The system according to claim 1, comprising means for identifying speakers from acquired audio information and classifying each utterance by speaker.

[0981] (Claim 3)

[0982] The system according to claim 1, comprising means for storing and making accessible the generated feedback and activity plan in a shared network storage.

[0983] "Application example 2 when combining with an emotional engine"

[0984] (Claim 1)

[0985] A means of acquiring audio data of a conversation,

[0986] A means of converting acquired audio data into text data in real time,

[0987] A means of analyzing the converted text data to identify the context of the discussion and the intentions of the participants,

[0988] A means of generating feedback based on analysis results and participant sentiment analysis to improve the flow of conversation,

[0989] A means of providing the generated feedback to participants in real time,

[0990] A means of creating a record and plan after the conversation has ended, based on the analysis results and emotional data.

[0991] A system that includes means for detecting abnormal emotional levels and issuing alarms based on emotional data.

[0992] (Claim 2)

[0993] The system according to claim 1, comprising means for identifying the speaker from acquired audio data and classifying each statement by speaker.

[0994] (Claim 3)

[0995] The system according to claim 1, comprising means for storing and making accessible the feedback of analysis results, action plans, and sentiment data in an information storage. [Explanation of Symbols]

[0996] 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. A device that acquires speech as audio information, A device that instantly converts acquired audio information into text information, A device that analyzes converted text information to identify the background of the discussion and the intentions of the participants, A device that generates notifications to improve progress based on analysis results, A device that immediately provides the generated notification, A device that, after completion, compiles the analysis results and creates a record document and action plan, A system that includes a device for sharing instant notifications on information terminals.

2. The system according to claim 1, comprising a device for identifying speakers from acquired audio information and classifying each statement by speaker.

3. The system according to claim 1, comprising a device for storing and making accessible the notification of analysis results and action plans in a memory area.

Images