System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting voice data; means for converting the voice data into text data; means for analyzing the text data and evaluating an emotion and a psychological state; means for visualizing a mental state based on a result of the evaluation; and means for generating a report indicating the mental state.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] In recent years, customer harassment (kasuhara) has been on the rise in the hospitality and customer support industries, often resulting in significant impacts on employee mental health. While this issue calls for advanced mental health management, current methods make it difficult to efficiently and accurately assess employees' mental health. Furthermore, there is a lack of data that allows managers and supervisors to quickly grasp employees' mental health and provide appropriate care, which means that stress tends to continue to build up among employees. To address these issues, a new and progressive system is needed. [Means for solving the problem]

[0005] The present invention solves the above problem by providing a system including a means for collecting voice data, a means for converting the voice data into text data, a means for analyzing the text data and evaluating emotions and psychological states, a means for visualizing the mental states based on the evaluation results, and a means for generating a report showing the mental states.

[0006] First, voice data from customer service or customer support operations is collected from intercoms or telephone systems. Next, the collected voice data is converted into text data using a voice recognition engine, and the text data is further analyzed using a natural language processing engine to evaluate emotions and psychological states. In addition, tone of voice features are extracted from the voice data, and psychological states are evaluated using these features. Furthermore, mental states are visualized based on the analysis results and evaluation points, and graphs and reports are generated based on data assigned to each employee. In this way, a system is provided that automatically evaluates mental states in real time, allowing managers and supervisors to quickly grasp the mental health of employees.

[0007] "Audio data" is a data format for recording and transmitting audio as a digital signal.

[0008] A "collection means" is a device or process used to obtain voice data from an intercom, telephone system, etc.

[0009] The "means for converting into text data" refers to a process or device that converts voice data into text information using a voice recognition engine.

[0010] The "means for analyzing" is a process or device for analyzing text data through a natural language processing engine to assess emotions and psychological states.

[0011] "Emotions and psychological states" refers to mental health states related to human emotions and mental health.

[0012] The "assessment means" is a process or device for evaluating emotions and psychological states as numerical values ​​or scores based on analytical data.

[0013] A "visualization means" is a process or device for displaying the results of an assessment in a visual format, such as a graph or chart.

[0014] A "report generating means" is a process or device for producing a written or digital report containing visualized mental state information.

[0015] "Vocal features" are data that quantitatively represent vocal attributes such as pitch, tone, and energy.

[0016] A "natural language processing engine" is software or an algorithm that analyzes text data and performs processes such as extracting meaning, emotion, and keywords.

[0017] A "speech recognition engine" is software or algorithms for converting voice data into text data. [Brief explanation of the drawings]

[0018] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0019] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0020] First, the terms used in the following description will be explained.

[0021] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0022] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0023] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0024] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0025] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0026] [First embodiment]

[0027] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0028] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0029] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0030] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0031] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0032] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0033] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0034] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0036] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0037] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0038] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0039] This invention is a system for managing the mental health of employees in customer support and customer service work, and includes a series of processes for collecting, analyzing, and evaluating voice data. Specific embodiments for carrying out this invention are described below.

[0040] 1. Data Collection

[0041] Device:

[0042] It captures voice data from intercoms and telephone systems installed in stores and call centers, which involves recording conversations between operators and customers in real time.

[0043] The captured audio data is sent to the server in real time using a protocol for audio streaming (e.g., WebSocket or RTSP).

[0044] 2. Audio data analysis

[0045] server:

[0046] The server sends the voice data received from the terminal to a voice recognition engine, which converts the data into text data. The voice recognition engine may be, for example, Google (registered trademark) Cloud Speech-to-Text API.

[0047] The converted text data is passed to a natural language processing (NLP) engine for text analysis, which analyzes the sentiment of the conversation, extracts keywords, and understands the context.

[0048] 3. Mental state assessment

[0049] server:

[0050] The analysis results from the NLP engine are used to assess the employee's emotional and psychological state, taking into account positive and negative emotional scoring as well as vocal characteristics (pitch, tone, energy, etc.).

[0051] A machine learning model is used to calculate employees' mental assessment points.

[0052] Examples:

[0053] For example, if negative keywords such as "anger" or "anxiety" frequently appear in the analysis results for a certain operator, the operator's mental evaluation score will be high. This score is also evaluated based on changes in tone of voice during conversations with customers, allowing for an accurate understanding of the operator's mental state.

[0054] 4. Report generation and visualization

[0055] server:

[0056] The mental evaluation points for each employee are aggregated and changes over time are visualized in graphs and charts.

[0057] The report includes mental assessment scores and related conversation highlights, allowing managers to understand specifically what situations increase stress levels.

[0058] Examples:

[0059] For example, if a graph of Operator A's mental health score for the past month shows a spike in stress on a specific date, the system will display details of the conversations that took place on that date, allowing managers to identify the root cause of the problem and take appropriate measures.

[0060] 5. Review and care of reports

[0061] User (Administrator):

[0062] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[0063] Consider consultation and care options and implement specific measures, such as stress management training or temporary reassignment.

[0064] Examples:

[0065] A manager checks the report of Operator B and finds that the mental health score for the past week was high. The manager schedules a meeting with Operator B, identifies the cause of stress, and provides mindfulness training to relieve stress, thereby improving the next mental health score.

[0066] The above is a specific embodiment for carrying out the present invention. By utilizing this system, it is possible to efficiently manage the mental health of employees and improve the quality of work.

[0067] The processing flow will be explained below.

[0068] Step 1:

[0069] Device:

[0070] Voice data is acquired in real time from intercoms and telephone systems installed in stores and call centers.

[0071] The acquired audio data is sent to the server using a protocol for audio streaming (e.g., WebSocket or RTSP).

[0072] Step 2:

[0073] server:

[0074] Stores audio data received from the device in temporary storage, which ensures a consistent data retention for subsequent processing.

[0075] Step 3:

[0076] server:

[0077] Call a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the voice data into text data.

[0078] Step 4:

[0079] server:

[0080] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information from the text.

[0081] Step 5:

[0082] server:

[0083] Extract voice characteristics (pitch, tone, energy, etc.) from the audio data by applying an acoustic feature analysis algorithm.

[0084] Step 6:

[0085] server:

[0086] Based on the analysis results from the NLP engine and the features of the voice tone, the emotional and psychological state is evaluated and a mental evaluation score is calculated. Here, a machine learning model is used to quantify the mental evaluation score.

[0087] Step 7:

[0088] server:

[0089] The mental evaluation points for each employee are aggregated and their changing mental state over time is visualized in graphs and charts.

[0090] Step 8:

[0091] server:

[0092] Automatically generate a mental status report, including mental assessment scores, relevant conversation content, and tone of voice fluctuations.

[0093] Step 9:

[0094] User (Administrator):

[0095] Reports are viewed through the server's dashboard, which visualizes each employee's mental assessment score and conversation highlights.

[0096] Step 10:

[0097] User (Administrator):

[0098] Schedule a meeting with the employee based on the report. Identify the causes of stress and take measures to address them, such as training programs or reassignment.

[0099] This is the specific process flow. This system allows managers to efficiently and accurately evaluate the mental health of their employees and provide appropriate care.

[0100] Example 1

[0101] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0102] In conventional customer support operations, the mental health of employees was not adequately managed, resulting in a decline in the quality of work, and problems such as employee stress and increased turnover. In addition, there was also the issue of the time required to collect, analyze, and evaluate voice data, making it difficult to respond in real time.

[0103] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0104] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for transmitting the collected voice data to the server in real time, means for sending the converted text data to a natural language processing engine for emotion analysis, means for accumulating mental evaluation points over time and displaying them in graphs or charts, and means for providing a dashboard for managers to check mental evaluation reports. This enables real-time collection and immediate analysis of voice data, making it possible to efficiently manage employee mental health and improve the quality of work.

[0105] "Voice data" means digital voice signals that record the conversation between an operator and a customer.

[0106] "Text data" is voice data converted into character information.

[0107] "Means for assessing emotions and psychological states" refers to technological means for analyzing text data and quantifying and assessing employees' emotions (positive and negative) and psychological states.

[0108] "Visualization means" refers to a means for displaying the evaluation results in graphs or charts so that users can understand them visually.

[0109] The "means for generating a report" refers to a means for creating a detailed report on the mental state of an employee based on the evaluation results.

[0110] "Means of transmitting in real time" refers to the technical means of instantly transferring collected voice data to a server.

[0111] A "natural language processing engine" refers to a processing engine that performs sentiment analysis, keyword extraction, context understanding, etc. on text data.

[0112] "Means for accumulating and displaying mental evaluation points over time" refers to a means for aggregating employees' mental evaluation points over a fixed period of time and visually displaying the fluctuations in graphs or charts.

[0113] "Dashboard" refers to a web-based or software-based platform that provides an interface through which managers can view employee mental health assessment reports in real time.

[0114] This invention is a system for managing the mental health of employees in customer support and customer service. The system includes a series of processes for collecting, analyzing, and evaluating voice data. Specific embodiments for implementing this invention are described below.

[0115] Data collection

[0116] Device:

[0117] Intercoms and telephone systems installed in stores and call centers collect voice data, including the ability to record conversations between operators and customers in real time. The collected voice data is then sent to a server in real time. Protocols such as WebSocket and RTSP are used to stream the data.

[0118] Sending audio data

[0119] Device:

[0120] The acquired voice data is sent to the server in real time. This process is carried out using protocols such as WebSocket and RTSP. For example, voice data collected during conversations in a store or during phone calls is immediately transferred to the server.

[0121] Converting audio data to text

[0122] server:

[0123] The server sends the received voice data to a speech recognition engine (e.g., Google Cloud Speech-to-Text API) and converts it into text data. The speech recognition engine converts voice into text with high accuracy, and text data is generated in near real time.

[0124] Text data analysis

[0125] server:

[0126] The converted text data is passed to a natural language processing (NLP) engine (e.g., Google Cloud Natural Language API), which performs sentiment analysis, keyword extraction, and contextual understanding of the text to analyze the content of employee conversations.

[0127] Mental state assessment

[0128] server:

[0129] The server evaluates the employee's emotions and psychological state based on the analysis results obtained from the NLP engine. This evaluation takes into account positive and negative emotion scores and vocal characteristics (pitch, tone, energy, etc.). Furthermore, a machine learning model is used to calculate the employee's mental evaluation score.

[0130] Examples:

[0131] For example, if negative keywords such as "anger" or "anxiety" frequently appear during a conversation with an agent, the agent's mental evaluation score will be high. This score also reflects changes in tone of voice during conversations with customers, allowing for an accurate understanding of the agent's mental state.

[0132] Collection and visualization of mental assessment data

[0133] server:

[0134] The server aggregates the mental evaluation points for each employee and visualizes them in graphs and charts, allowing managers to track fluctuations in employees' mental states over time.

[0135] Examples:

[0136] For example, the mental health score of Operator A over the past month can be graphed, allowing fluctuations in the evaluation to be observed at a glance. If the graph shows that stress spiked on a particular day, the content of the conversations on that day can be checked in detail.

[0137] Administrator report confirmation

[0138] User (Administrator):

[0139] Managers can access a dashboard provided by the server and view mental health assessment reports for each employee, which display graphs, scores, and conversation highlights, allowing managers to identify specific stressors.

[0140] Examples:

[0141] For example, a manager can check Operator B's high mental health score on the dashboard, investigate the cause, and take appropriate measures.

[0142] Stress management

[0143] User (Administrator):

[0144] Managers can use the reports to schedule interviews and stress management training sessions with employees, and, if necessary, take concrete action, such as temporarily reassigning employees.

[0145] Examples:

[0146] For example, a manager might schedule a meeting with Operator B to identify the cause of stress and provide mindfulness training to improve the next mental score.

[0147] Prompt Sentence Examples

[0148] The generative AI model generates a natural language explanation for a mental health analysis by inputting the following prompt:

[0149] Please explain your mental health assessment system for customer support employees. The system involves a series of processes to collect, analyze, and evaluate voice data. Please explain what hardware or software is used, and what data processing or calculation is performed, including specific scenarios and examples of use.

[0150] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0151] Step 1:

[0152] Data collection

[0153] The device collects voice data from intercoms and telephone systems, including the ability to record conversations between operators and customers in real time. The device records and stores data of interactions with clients via intercoms and telephone systems.

[0154] Input: Voice of conversation between operator and customer

[0155] Output: Real-time audio data

[0156] Step 2:

[0157] Sending audio data

[0158] The device transmits the acquired audio data to the server in real time using protocols such as WebSocket and RTSP. For example, when a customer complains that their product has not arrived, the audio is immediately transmitted to the server.

[0159] Input: Real-time audio data

[0160] Output: Audio data sent to the server

[0161] Step 3:

[0162] Converting audio data to text

[0163] The server sends the received voice data to a voice recognition engine and converts it into text data. For example, the voice recognition engine uses the Google Cloud Speech-to-Text API. The server sends the voice data to the API and receives text data as character information.

[0164] Input: Audio data transferred to the server

[0165] Output: Text data generated by the speech recognition engine

[0166] Step 4:

[0167] Text data analysis

[0168] The server then sends the converted text data to a natural language processing engine to perform sentiment analysis, keyword extraction, and context understanding. For example, it detects whether the text contains keywords such as "anger" or "frustration." The server inputs the text data into the analysis engine and obtains an emotion score and a keyword list as output.

[0169] Input: Text data generated by a speech recognition engine

[0170] Output: Sentiment score and keyword list analyzed by a natural language processing engine

[0171] Step 5:

[0172] Mental state assessment

[0173] The server evaluates the employee's emotions and psychological state based on the analysis results obtained from the NLP engine. This evaluation also includes vocal characteristics (pitch, tone, energy, etc.). The server inputs the analysis results and calculates the employee's mental evaluation score. For example, if the emotion "anger" appears frequently, the mental evaluation score will be high.

[0174] Input: Sentiment scores and keyword list analyzed by a natural language processing engine

[0175] Output: Mental evaluation points calculated by the server

[0176] Step 6:

[0177] Collection and visualization of mental assessment data

[0178] The server aggregates the mental evaluation points for each employee and visualizes them in graphs and charts. The evaluation results are displayed in daily, weekly, monthly, etc. For example, Operator A's mental score for the past month is displayed in a graph.

[0179] Input: Mental evaluation points for each employee

[0180] Output: Visualized graphs and charts

[0181] Step 7:

[0182] Administrator report confirmation

[0183] The user (administrator) accesses the dashboard provided by the server and checks the mental assessment report for each employee. The report includes analysis results, graphs, scores, and conversation highlights. The administrator can obtain detailed reports from the dashboard and check stress factors.

[0184] Input: Visualized graphs, charts, detailed reports

[0185] Output: Mental evaluation report reviewed by the administrator

[0186] Step 8:

[0187] Stress management

[0188] Based on the report, the user (manager) schedules interviews with employees and stress management training. If necessary, the user can take specific countermeasures, such as reassignment. For example, if Operator B shows signs of high stress, the user can schedule an interview and take specific countermeasures.

[0189] Input: Mental evaluation report reviewed by administrator

[0190] Output: Plans for scheduled interviews and stress management training

[0191] (Application example 1)

[0192] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0193] Employee mental health has a significant impact on the quality and efficiency of work. However, in physical stores, it is difficult to grasp the mental state of employees in real time and provide appropriate care. There are limited means to accurately assess the level of stress employees are experiencing while interacting with customers and to take immediate countermeasures. This can lead to an accumulation of employee stress, which can lead to reduced work efficiency and mental health problems.

[0194] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0195] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for visualizing the collected mental evaluation points over time, and means for collecting voice data from a microphone built into the smart glasses and transmitting it to the server in real time, thereby enabling employees to evaluate the mental states of customers in real time while interacting with them, and allowing managers to provide immediate care.

[0196] "Voice data" refers to audio information of conversations between employees and customers.

[0197] "Text data" is voice data converted into character information that is easy to analyze.

[0198] The "means for assessing emotions and psychological states" refers to a means for analyzing the emotions and psychological states of employees based on text data and obtaining the assessment results.

[0199] A "means for visualizing mental state" is a means for understanding mental state by visually displaying the evaluation results.

[0200] The "means for generating a report" refers to a means for creating a report summarizing the mental state of employees based on the evaluation results.

[0201] A "means for extracting vocal timbre features" is a means for extracting vocal features such as pitch, tone, and energy from audio data.

[0202] "Real-time means" means means for collecting, converting, and analyzing audio data in real time.

[0203] "Smart glasses" are glasses-type devices for collecting and transmitting audio data in real time.

[0204] A "server" is a computer system for analyzing collected voice data and processing evaluation results.

[0205] "Mental evaluation points" are indicators that quantitatively represent an employee's mental state and are calculated based on the analysis results.

[0206] The present invention relates to a system that utilizes smart glasses to efficiently monitor the mental health of employees in brick-and-mortar stores while they are serving customers. Specific embodiments are described below.

[0207] System configuration

[0208] The system includes a set of means for collecting, analyzing, and evaluating audio data.

[0209] 1. Audio data collection method:

[0210] The system uses microphones built into smart glasses worn by store employees to collect conversations between employees and customers in real time.

[0211] The collected audio data is transmitted to a server using WebSocket or a similar real-time transmission protocol.

[0212] 2. Means of converting audio data to text data:

[0213] The server converts the received voice data into text data using a voice recognition engine (e.g., Google Cloud Speech-to-Text API).

[0214] This conversion process converts the audio data into a format that is easier to analyze.

[0215] 3. Instruments for assessing emotional and psychological states:

[0216] The server passes the converted text data to a natural language processing (NLP) engine to perform sentiment analysis (e.g., Google Cloud Natural Language API).

[0217] The NLP engine extracts emotional scores such as positive or negative and specific keywords from text data to evaluate the mental state of employees.

[0218] 4. Ways to visualize your mental state:

[0219] The server calculates and visualizes mental evaluation points based on the evaluation results obtained from the NLP engine.

[0220] This visualization includes a means for displaying the time-varying change in the evaluation points as a graph or chart.

[0221] 5. To generate a report:

[0222] The server generates reports showing the mental state of each employee and displays them on a dashboard accessible to managers.

[0223] The report includes detailed rating points, graphs of changes over time, and highlights of key conversations.

[0224] Specific examples of implementation

[0225] For example, while Employee A is wearing smart glasses and interacting with customers, their conversation is collected as voice data. The collected voice data is sent to a server in real time and converted into text data by a voice recognition engine. The text data is analyzed by an NLP engine to evaluate Employee A's emotion score and psychological state. The evaluation results are visualized on a dashboard for managers, allowing them to monitor Employee A's mental health and provide care as needed.

[0226] Prompt Sentence Examples

[0227] I would like to build a system that collects customer voice data, converts it to text in real time, and performs sentiment analysis. Specifically, please generate sample code for a system that uses a speech recognition engine to convert speech to text, and a natural language processing engine to perform sentiment analysis.

[0228] The above is a specific embodiment for carrying out the present invention. By utilizing this system, it is possible to monitor the mental health of employees in real time and provide appropriate care.

[0229] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0230] Step 1:

[0231] A microphone built into the smart glasses collects conversations between employees and customers in real time.

[0232] Input: Employee-customer conversation.

[0233] Output: Audio data.

[0234] To send the audio data to the server, the audio data is transferred in real time using the WebSocket protocol, etc. Specifically, the audio data obtained from the smart glasses is received via a WebSocket connection.

[0235] Step 2:

[0236] The server passes the received voice data to a voice recognition engine, which converts the data into text data.

[0237] Input: Audio data.

[0238] Output: Text data.

[0239] Specifically, the voice data is converted into a string of characters (text data) using the Google Cloud Speech-to-Text API, etc. The converted text data is saved in a format suitable for subsequent analysis.

[0240] Step 3:

[0241] The server passes the text data to a natural language processing (NLP) engine for sentiment analysis.

[0242] Input: Text data.

[0243] Output: Sentiment analysis results (sentiment score, keywords, etc.).

[0244] Specifically, it uses an NLP engine such as Google Cloud Natural Language API to extract sentiment scores and important keywords from text data to assess employee emotions and psychological states.

[0245] Step 4:

[0246] The server calculates the employee's mental evaluation points based on the emotion analysis results.

[0247] Input: Sentiment analysis results.

[0248] Output: Mental evaluation points.

[0249] Specifically, the mental state of employees is quantified based on the emotion score and the frequency of keywords, and a mental evaluation point is calculated, allowing for a quantitative evaluation of the mental state of employees.

[0250] Step 5:

[0251] The server generates graphs and charts to visualize the mental assessment points and displays them on an administrator dashboard.

[0252] Input: Mental Assessment Points.

[0253] Output: Visualized data (graphs, charts).

[0254] Specifically, a web-based dashboard is used to display the collected mental health assessment scores over time in graphs and charts, allowing managers to provide appropriate care based on this information.

[0255] Step 6:

[0256] The server generates reports showing each employee's mental state, which managers can check regularly.

[0257] Input: Mental evaluation points, conversation highlights.

[0258] Output: Report.

[0259] Specifically, the system generates regular reports containing detailed mental health assessment scores, graphs of changes over time, and highlights of important conversations, and makes them accessible to managers, allowing them to understand trends in employee mental health and take necessary measures.

[0260] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0261] This invention is a system for managing the mental health of employees in customer support and customer service, and includes a series of processes for collecting voice data in real time and evaluating their emotions and psychological state. In particular, by combining it with an emotion engine, the system can more precisely recognize the user's emotions and reflect them in the mental health evaluation.

[0262] Specific Embodiments

[0263] Device:

[0264] It captures voice data from intercoms and telephone systems installed in stores and call centers, which involves recording conversations between operators and customers in real time.

[0265] The captured audio data is sent to the server in real time using a protocol for audio streaming (e.g., WebSocket or RTSP).

[0266] server:

[0267] 1. Audio data conversion:

[0268] The server stores the voice data received from the device in temporary storage and converts it into text data by calling a voice recognition engine, such as the Google Cloud Speech-to-Text API.

[0269] 2. Text data and NLP analysis:

[0270] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information from the text.

[0271] 3. Voice timbre feature extraction:

[0272] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[0273] 4. Applying the Emotion Engine:

[0274] The extracted vocal features are input into an emotion engine to recognize the corresponding emotion, such as anger, joy, or sadness.

[0275] 5. Mental status assessment:

[0276] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state. The evaluation includes scoring positive and negative emotions and calculating mental evaluation points.

[0277] Examples:

[0278] For example, if an agent frequently uses keywords that express negative emotions, such as "anxiety" or "anger," while talking to customers, the emotion engine will recognize this and, if the characteristics of the agent's voice (e.g., a sudden rise in pitch or an emphasized tone) indicate the corresponding emotion, the agent's mental evaluation score will be high. This evaluation is displayed as a quantitative score and can be checked by managers.

[0279] 6. Report generation and visualization:

[0280] The system aggregates each employee's mental assessment score and visualizes their mental state over time in graphs and charts. The report includes the mental assessment score, related conversation content, and changes in tone of voice.

[0281] Examples:

[0282] For example, if a graph of Operator A's mental health score for the past month shows a spike in stress on a specific date, the system will display details of the conversations that took place on that date, allowing managers to identify the root cause of the problem and take appropriate measures.

[0283] User (Administrator):

[0284] 1. Report Review and Care:

[0285] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[0286] Consider consultation and care options and implement specific measures, such as stress management training or temporary reassignment.

[0287] Examples:

[0288] A manager reviews Operator B's report and finds that his mental health score for the past week was high. The manager can schedule a meeting with Operator B to identify the cause of his stress and provide mindfulness training to relieve stress, thereby improving his mental health score next time.

[0289] The above is a specific embodiment of the invention that combines an emotion engine. This system allows managers to efficiently and accurately evaluate the mental health of employees and provide appropriate care.

[0290] The processing flow will be explained below.

[0291] Step 1:

[0292] Device:

[0293] Voice data is acquired in real time from intercoms and telephone systems installed in stores and call centers.

[0294] The acquired audio data is sent to the server using an audio streaming protocol (e.g., WebSocket or RTSP).

[0295] Step 2:

[0296] server:

[0297] Stores audio data received from the device in temporary storage, which ensures a consistent data retention for subsequent processing.

[0298] Step 3:

[0299] server:

[0300] Call a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the voice data into text data.

[0301] Step 4:

[0302] server:

[0303] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information.

[0304] Step 5:

[0305] server:

[0306] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[0307] Step 6:

[0308] server:

[0309] The extracted vocal features are input into an emotion engine to recognize the corresponding emotion, such as anger, joy, or sadness.

[0310] Step 7:

[0311] server:

[0312] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state, including scoring positive and negative emotions and calculating mental evaluation points.

[0313] Examples:

[0314] For example, if an agent frequently uses words like "anxiety" or "anger" while talking to customers, the emotion engine will recognize this as a negative emotion. Furthermore, if the tone of voice features (higher pitch or emphasized tone) indicate the corresponding emotion, the agent's mental evaluation score will be higher. This evaluation is displayed as a quantitative score and can be checked by managers.

[0315] Step 8:

[0316] server:

[0317] The mental evaluation points for each employee are aggregated and their changing mental state over time is visualized in graphs and charts.

[0318] Step 9:

[0319] server:

[0320] Automatically generate a mental status report, including mental assessment scores, relevant conversation content, and tone of voice fluctuations.

[0321] Examples:

[0322] For example, if the mental health score of Operator A for the past month is graphed and stress spikes on a particular day, the details of the conversations that took place on that day will be displayed in detail. This display allows managers to identify the root cause of the problem and take appropriate measures.

[0323] Step 10:

[0324] User (Administrator):

[0325] Reports are viewed through a server dashboard, which visualizes each employee's mental assessment score and conversation highlights.

[0326] Step 11:

[0327] User (Administrator):

[0328] Based on the report, meetings with employees can be scheduled to identify the causes of stress, and specific countermeasures such as stress management training or temporary reassignment can be considered.

[0329] Examples:

[0330] The manager checks Operator B's report and finds that his mental health score for the past week is high. The manager schedules a meeting with Operator B to identify the cause of his stress. Based on this, the manager provides mindfulness training to relieve stress and improve his mental health score for the next time.

[0331] The above is the specific processing flow of the invention that combines the emotion engine and the detailed operation of each step. This system allows managers to efficiently and accurately evaluate the mental health of employees and provide appropriate care.

[0332] Example 2

[0333] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0334] To manage employee mental health, there is a need for a method to accurately and in real time assess emotions and psychological states from on-site voice data. However, conventional methods have the problem that they cannot analyze voice data in real time and are not precise enough in recognizing emotions and assessing mental states, making it difficult for managers to take appropriate action quickly.

[0335] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for temporarily saving voice data, means for converting voice data into text data, means for analyzing the text data and extracting emotion and context information, means for extracting tone of voice features from the voice data, means for recognizing emotion based on the tone of voice features, means for evaluating mental states by integrating emotion recognition results and text analysis results, means for visualizing mental states based on evaluation results, and means for generating reports showing mental states. This allows employees' emotions and psychological states to be accurately evaluated in real time, enabling managers to take prompt and appropriate action.

[0336] "Audio data" refers to information that records audio in digital format.

[0337] "Temporarily storing" means holding the received data in a storage area for a short period of time to prevent it from being lost.

[0338] "Converting to text data" refers to the process of converting voice data into character information.

[0339] "Analyzing text data and extracting emotional and contextual information" means using natural language processing technology to extract emotional, keyword, and contextual information from text.

[0340] "Extracting vocal features" means extracting acoustic features such as pitch, tone, and energy from speech data.

[0341] "Emotion recognition" means identifying the speaker's emotion based on the extracted vocal timbre features.

[0342] "Evaluating the mental state" means quantitatively evaluating the speaker's psychological state by combining the emotion recognition results and text analysis results.

[0343] "Visualizing" means converting data into a format that is easy to understand visually, such as a graph or chart.

[0344] "Generating a report" means creating a report based on the evaluation results so that the administrator can check it.

[0345] "Real-time" means that data is processed immediately after it is generated, with little delay.

[0346] This system uses hardware installed in stores or call centers, a server, and specialized software to process a series of data in order to manage the mental health of employees. A specific embodiment of this system is described below.

[0347] Terminal

[0348] The device acquires voice data from intercoms and telephone systems installed in stores and call centers, and transmits the acquired voice data to a server in real time using an audio streaming protocol such as WebSocket or RTSP.

[0349] server

[0350] The server processes the audio data in the following steps:

[0351] 1. Temporary storage of audio data:

[0352] The server temporarily stores the voice data received from the terminal in storage to ensure data security and facilitate subsequent processing.

[0353] 2. Audio to text conversion:

[0354] The server uses the Google Cloud Speech-to-Text API to convert the temporarily stored voice data into text data. For example, a conversation between an operator and a customer, such as "Hello, how can I help you?", is converted directly into text.

[0355] 3. NLP analysis of text data:

[0356] The server passes the converted text data to an NLP engine (e.g., spaCy or BERT) to analyze the sentiment, keywords, and context of the text. For example, from the phrase "I'm feeling very stressed today," "stress" is extracted as an emotional keyword.

[0357] 4. Extracting voice timbre features from speech data:

[0358] The server applies an acoustic feature analysis algorithm to extract vocal timbre features such as pitch, tone, and energy from the audio data. For example, a rise in vocal pitch indicates a change in emotion.

[0359] 5. Emotion recognition based on voice features:

[0360] The server inputs the extracted vocal features into an emotion engine to recognize emotions, such as "tension," "joy," and "anger."

[0361] 6. Mental Status Assessment:

[0362] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to calculate positive and negative emotion scores and mental evaluation points. For example, the result may be "positive score 70%" or "negative score 30%."

[0363] 7. Visualization of mental state and report generation:

[0364] The server aggregates the mental assessment scores for each employee and visualizes fluctuations in their mental state over time in graphs and charts. For example, it can display the weekly fluctuations in mental scores in a line graph, and show the reason for a sudden rise in stress on a particular date along with detailed conversation content.

[0365] User (Administrator)

[0366] Managers access a dashboard provided by the server and check the mental health assessment report for each employee. The report displays graphs, scores, conversation highlights, and more. Based on this, managers can consider interviews and care measures and take specific countermeasures. For example, for employees with high levels of stress, they can suggest "mindfulness training" or "temporary reassignment."

[0367] Examples:

[0368] If an agent frequently uses keywords indicating negative emotions, such as "anxiety" or "anger," while talking to a customer, the emotion engine will identify this. Furthermore, if characteristics indicating anxiety or anger are detected from an analysis of the agent's tone of voice, the agent's mental state score will be higher. This score will be reported to the manager. The report will display a graph showing the "progression of mental state over the past month," visualizing, for example, spikes in stress on specific days.

[0369] Example prompt sentence:

[0370] "To assess the mental health of employees over the past month, please generate a report based on the following information: speech-to-text transcription results, NLP analysis of emotions and keywords, and voice timbre features. Please also include the mental health assessment score for each date and the relevant conversation content."

[0371] This system clarifies the roles of servers, terminals, and users, making it possible to efficiently manage employee mental health in real time.

[0372] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0373] Step 1:

[0374] The terminal acquires voice data from intercoms or telephone systems installed in stores or call centers. At this time, conversations between operators and customers are recorded in real time. The input is voice data from the intercom or telephone system, and the output is an audio file acquired in real time.

[0375] Step 2:

[0376] The audio data captured by the device is sent to the server in real time. This process uses audio streaming protocols such as WebSocket and RTSP. The input is the audio data captured by the device, and the output is the audio streaming data sent to the server.

[0377] Step 3:

[0378] The server temporarily stores the audio data received from the device in storage. This storage process prevents loss of audio data and makes subsequent processing easier. The input is audio data acquired as streaming data, and the output is an audio file stored in temporary storage.

[0379] Step 4:

[0380] The server uses the Google Cloud Speech-to-Text API to convert the temporarily stored voice data into text data. The input is the audio file stored in temporary storage, and the output is the converted text data. This conversion process allows the voice information to be treated as text information.

[0381] Step 5:

[0382] The server passes the text data to a natural language processing (NLP) engine (e.g., spaCy or BERT) to extract sentiment, keywords, and contextual information from the text. The input is the text data, and the output is the sentiment, keywords, and contextual information extracted through analysis. For example, the keyword "stress" is extracted from the text "I'm feeling very stressed today."

[0383] Step 6:

[0384] The server applies an acoustic feature analysis algorithm to extract vocal timbre features such as pitch, tone, and energy from the audio data. The input is a temporarily saved audio file, and the output is the extracted vocal timbre features. For example, a rise in voice pitch indicates a change in emotion.

[0385] Step 7:

[0386] The server inputs the extracted voice timbre features into an emotion engine to recognize the corresponding emotion. The input is the voice timbre features obtained from acoustic feature analysis, and the output is the recognized emotion (e.g., tension, joy, anger). The emotion engine then quantifies the extracted features as an emotion.

[0387] Step 8:

[0388] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to calculate positive and negative emotion scores and mental evaluation points for employees. The input is the text analysis results and emotion recognition results, and the output is the emotion score and mental evaluation points for each employee. For example, evaluations such as "positive score 70%" and "negative score 30%" are possible.

[0389] Step 9:

[0390] The server aggregates the mental assessment points for each employee and visualizes their mental state over time in graphs and charts. The input is each employee's mental assessment point, and the output is their mental state visualized as a graph or chart. For example, a line graph can be used to display the progress of mental state by week, showing the reason for a sudden rise in stress on a particular day.

[0391] Step 10:

[0392] The user (manager) accesses the dashboard provided by the server and checks the mental health assessment report for each employee. The input is the mental health assessment report provided by the server, and the output is report information that the manager can check and use. Based on this, the manager considers interviews and care measures and takes specific countermeasures (e.g., mindfulness training or temporary reassignment).

[0393] The above are the processing steps of the program for this system, and the specific operations performed at each step, as well as the inputs and outputs, have been explained.

[0394] (Application example 2)

[0395] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0396] Efficiently managing employee mental health is important in customer support and customer service. However, conventional systems have difficulty evaluating mental health in real time, making it difficult to immediately grasp changes in employees' stress levels and emotions. In addition, there are limited ways for employees to check their own mental state in real time, making it difficult to implement appropriate stress management.

[0397] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0398] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for visualizing the mental states based on the evaluation results, means for generating a report showing the mental states, and means for displaying the results on a smart device. This allows emotions and psychological states to be evaluated in real time from the voice data, and allows employees to check their mental states themselves via their smart devices.

[0399] "Voice data" means the acoustic signals of a conversation between an operator and a customer.

[0400] "Means for collecting" refers to a device or method for capturing and inputting audio data into the system.

[0401] "Text data" is voice data converted into character information.

[0402] "Means for converting into text data" refers to a device or method that converts voice data into text information using voice recognition technology.

[0403] "Means for analyzing and assessing emotions and psychological states" means a device or method for analyzing and assessing an employee's emotions and psychological states using text data and tone of voice features.

[0404] "Visualization means" refers to a device or method for displaying the evaluation results in a visual form such as a graph or chart.

[0405] A "means for generating a report" is a device or method for creating a detailed written report based on the results of the evaluation.

[0406] "Means for displaying results on a smart device" refers to a device or method for displaying the results of a mental state assessment in real time on a device such as smart glasses or a wearable device.

[0407] The system for implementing this invention mainly performs a series of processes, including collecting voice data, analyzing emotions, and evaluating mental states. The system is realized by linking terminals, a server, and users.

[0408] Terminal

[0409] The terminals are mainly smart devices (e.g., smart glasses). These terminals have built-in microphones that collect voice data from workers. The collected voice data is transmitted to the server in real time. WebSocket or RTSP is used as the protocol for audio streaming.

[0410] server

[0411] The server performs the following steps:

[0412] 1. Audio data conversion:

[0413] The server temporarily stores the voice data acquired from the device and converts it into text data using the Google Cloud Speech-to-Text API.

[0414] 2. Text data analysis:

[0415] The converted text data is analyzed using a natural language processing (NLP) engine, using spaCy and BERT to extract sentiment, keywords, and contextual information from the text.

[0416] 3. Voice timbre feature extraction:

[0417] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[0418] 4. Applying the Emotion Engine:

[0419] The extracted voice features are input into an emotion engine to recognize the corresponding emotion. BERT or other generative AI models may be used as emotion models.

[0420] 5. Mental status assessment:

[0421] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state, and the evaluation results are displayed as a positive or negative emotion score.

[0422] 6. Real-time display:

[0423] The evaluation results are sent in real time to smart devices so that employees can view them.

[0424] User (Administrator)

[0425] Users are responsible for managing the mental health of their employees and providing appropriate care when necessary. Specifically, this includes:

[0426] 1. Report Review and Care:

[0427] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[0428] We will consider options for interviews and care, and provide stress management training and temporary reassignments.

[0429] Specific examples

[0430] For example, if a store employee wearing smart glasses speaks in a "high-pitched, emphasized tone" while interacting with a customer, the voice data is sent to a server, where real-time emotion analysis is performed. As a result, the employee's stress level is assessed as rising, and the smart glasses display shows "high stress." Based on this information, managers can instruct employees to take a break. Managers can also take countermeasures based on this data later.

[0431] Prompt Sentence Examples

[0432] "The employee's voice is high-pitched and stressed. We suggest they take a 10-minute break."

[0433] "Your mental health score for the past week has been high. I suggest you take stress management training."

[0434] Thus, in order to implement this invention, it is necessary to build a system in which the components of the server, terminal, and user work together to evaluate and manage the mental state of employees in real time.

[0435] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0436] Step 1:

[0437] Audio data collection

[0438] The device collects employee voice data.

[0439] Input: Audio data captured from the built-in microphone of the smart glasses.

[0440] How it works: The built-in microphone records the conversation between the operator and the customer and sends the audio data to the server in real time.

[0441] Step 2:

[0442] Converting audio data to text data

[0443] The server converts the received voice data into text data.

[0444] Input: Audio data sent from the device.

[0445] Output: Text data.

[0446] What it does: It uses the Google Cloud Speech-to-Text API to convert audio data into text. A speech recognition model analyzes the acoustic signal and generates a corresponding string of characters.

[0447] Step 3:

[0448] Text data analysis

[0449] The server analyzes the converted text data and evaluates the emotions and psychological state.

[0450] Input: Text data.

[0451] Output: Emotion score and mental state assessment results.

[0452] Specific operation: Text data is input into a natural language processing (NLP) engine (e.g., spaCy, BERT) to analyze and extract sentiment, keywords, and contextual information.

[0453] Step 4:

[0454] Voice timbre feature extraction

[0455] The server extracts voice characteristics from the voice data.

[0456] Input: Audio data.

[0457] Output: Voice characteristics (pitch, tone, energy, etc.).

[0458] Specific operation: Apply an acoustic feature analysis algorithm (e.g., librosa) to extract features such as pitch, tone, and energy from the audio signal.

[0459] Step 5:

[0460] Applying the Emotion Engine

[0461] The server inputs the voice features into the emotion engine and recognizes the corresponding emotion.

[0462] Input: Voice timbre features.

[0463] Output: Emotion recognition results.

[0464] Specific operation: Using a generative AI model (e.g., BERT), emotions are classified from tone of voice features, and emotions such as anger, joy, and sadness are recognized.

[0465] Step 6:

[0466] Mental state assessment

[0467] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to evaluate the mental state of employees.

[0468] Input: Emotion score, voice features, and text analysis results.

[0469] Output: Mental state assessment score.

[0470] Specific operations: Evaluate overall mental state by assigning positive and negative emotional scores and calculating mental evaluation points.

[0471] Step 7:

[0472] Real-time display

[0473] The server transmits the evaluation results to the smart device in real time and displays them.

[0474] Input: Mental State Assessment score.

[0475] Output: Evaluation results displayed on smart devices.

[0476] Specific operation: The evaluation results are displayed on the smart glasses display, allowing employees to check their own mental state in real time.

[0477] Step 8:

[0478] Report Generation

[0479] The server generates a report based on the mental state evaluation results and provides it to the user.

[0480] Input: Accumulated data of evaluation results.

[0481] Output: Detailed mental state report.

[0482] Specific actions: Generate graphs and charts to create reports that visualize fluctuations in mental state for each employee, and display them on a dashboard for managers to review.

[0483] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0484] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0485] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0486] [Second embodiment]

[0487] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0488] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0489] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0490] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0491] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0492] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0493] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0494] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0495] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0496] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0497] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0498] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0499] This invention is a system for managing the mental health of employees in customer support and customer service work, and includes a series of processes for collecting, analyzing, and evaluating voice data. Specific embodiments for carrying out this invention are described below.

[0500] 1. Data Collection

[0501] Device:

[0502] It captures voice data from intercoms and telephone systems installed in stores and call centers, which involves recording conversations between operators and customers in real time.

[0503] The captured audio data is sent to the server in real time using a protocol for audio streaming (e.g., WebSocket or RTSP).

[0504] 2. Audio data analysis

[0505] server:

[0506] The server sends the voice data received from the device to a voice recognition engine, such as the Google Cloud Speech-to-Text API, and converts it into text data.

[0507] The converted text data is passed to a natural language processing (NLP) engine for text analysis, which analyzes the sentiment of the conversation, extracts keywords, and understands the context.

[0508] 3. Mental state assessment

[0509] server:

[0510] The analysis results from the NLP engine are used to assess the employee's emotional and psychological state, taking into account positive and negative emotional scoring as well as vocal characteristics (pitch, tone, energy, etc.).

[0511] A machine learning model is used to calculate employees' mental assessment points.

[0512] Examples:

[0513] For example, if negative keywords such as "anger" or "anxiety" frequently appear in the analysis results for a certain operator, the operator's mental evaluation score will be high. This score is also evaluated based on changes in tone of voice during conversations with customers, allowing for an accurate understanding of the operator's mental state.

[0514] 4. Report generation and visualization

[0515] server:

[0516] The mental evaluation points for each employee are aggregated and changes over time are visualized in graphs and charts.

[0517] The report includes mental assessment scores and related conversation highlights, allowing managers to understand specifically what situations increase stress levels.

[0518] Examples:

[0519] For example, if a graph of Operator A's mental health score for the past month shows a spike in stress on a specific date, the system will display details of the conversations that took place on that date, allowing managers to identify the root cause of the problem and take appropriate measures.

[0520] 5. Review and care of reports

[0521] User (Administrator):

[0522] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[0523] Consider consultation and care options and implement specific measures, such as stress management training or temporary reassignment.

[0524] Examples:

[0525] A manager checks the report of Operator B and finds that the mental health score for the past week was high. The manager schedules a meeting with Operator B, identifies the cause of stress, and provides mindfulness training to relieve stress, thereby improving the next mental health score.

[0526] The above is a specific embodiment for carrying out the present invention. By utilizing this system, it is possible to efficiently manage the mental health of employees and improve the quality of work.

[0527] The processing flow will be explained below.

[0528] Step 1:

[0529] Device:

[0530] Voice data is acquired in real time from intercoms and telephone systems installed in stores and call centers.

[0531] The acquired audio data is sent to the server using a protocol for audio streaming (e.g., WebSocket or RTSP).

[0532] Step 2:

[0533] server:

[0534] Stores audio data received from the device in temporary storage, which ensures a consistent data retention for subsequent processing.

[0535] Step 3:

[0536] server:

[0537] Call a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the voice data into text data.

[0538] Step 4:

[0539] server:

[0540] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information from the text.

[0541] Step 5:

[0542] server:

[0543] Extract voice characteristics (pitch, tone, energy, etc.) from the audio data by applying an acoustic feature analysis algorithm.

[0544] Step 6:

[0545] server:

[0546] Based on the analysis results from the NLP engine and the features of the voice tone, the emotional and psychological state is evaluated and a mental evaluation score is calculated. Here, a machine learning model is used to quantify the mental evaluation score.

[0547] Step 7:

[0548] server:

[0549] The mental evaluation points for each employee are aggregated and their changing mental state over time is visualized in graphs and charts.

[0550] Step 8:

[0551] server:

[0552] Automatically generate a mental status report, including mental assessment scores, relevant conversation content, and tone of voice fluctuations.

[0553] Step 9:

[0554] User (Administrator):

[0555] Reports are viewed through the server's dashboard, which visualizes each employee's mental assessment score and conversation highlights.

[0556] Step 10:

[0557] User (Administrator):

[0558] Schedule a meeting with the employee based on the report. Identify the causes of stress and take measures to address them, such as training programs or reassignment.

[0559] This is the specific process flow. This system allows managers to efficiently and accurately evaluate the mental health of their employees and provide appropriate care.

[0560] Example 1

[0561] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0562] In conventional customer support operations, the mental health of employees was not adequately managed, resulting in a decline in the quality of work, and problems such as employee stress and increased turnover. In addition, there was also the issue of the time required to collect, analyze, and evaluate voice data, making it difficult to respond in real time.

[0563] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0564] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for transmitting the collected voice data to the server in real time, means for sending the converted text data to a natural language processing engine for emotion analysis, means for accumulating mental evaluation points over time and displaying them in graphs or charts, and means for providing a dashboard for managers to check mental evaluation reports. This enables real-time collection and immediate analysis of voice data, making it possible to efficiently manage employee mental health and improve the quality of work.

[0565] "Voice data" means digital voice signals that record the conversation between an operator and a customer.

[0566] "Text data" is voice data converted into character information.

[0567] "Means for assessing emotions and psychological states" refers to technological means for analyzing text data and quantifying and assessing employees' emotions (positive and negative) and psychological states.

[0568] "Visualization means" refers to a means for displaying the evaluation results in graphs or charts so that users can understand them visually.

[0569] The "means for generating a report" refers to a means for creating a detailed report on the mental state of an employee based on the evaluation results.

[0570] "Means of transmitting in real time" refers to the technical means of instantly transferring collected voice data to a server.

[0571] A "natural language processing engine" refers to a processing engine that performs sentiment analysis, keyword extraction, context understanding, etc. on text data.

[0572] "Means for accumulating and displaying mental evaluation points over time" refers to a means for aggregating employees' mental evaluation points over a fixed period of time and visually displaying the fluctuations in graphs or charts.

[0573] "Dashboard" refers to a web-based or software-based platform that provides an interface through which managers can view employee mental health assessment reports in real time.

[0574] This invention is a system for managing the mental health of employees in customer support and customer service. The system includes a series of processes for collecting, analyzing, and evaluating voice data. Specific embodiments for implementing this invention are described below.

[0575] Data collection

[0576] Device:

[0577] Intercoms and telephone systems installed in stores and call centers collect voice data, including the ability to record conversations between operators and customers in real time. The collected voice data is then sent to a server in real time. Protocols such as WebSocket and RTSP are used to stream the data.

[0578] Sending audio data

[0579] Device:

[0580] The acquired voice data is sent to the server in real time. This process is carried out using protocols such as WebSocket and RTSP. For example, voice data collected during conversations in a store or during phone calls is immediately transferred to the server.

[0581] Converting audio data to text

[0582] server:

[0583] The server sends the received voice data to a speech recognition engine (e.g., Google Cloud Speech-to-Text API) and converts it into text data. The speech recognition engine converts voice into text with high accuracy, and text data is generated in near real time.

[0584] Text data analysis

[0585] server:

[0586] The converted text data is passed to a natural language processing (NLP) engine (e.g., Google Cloud Natural Language API), which performs sentiment analysis, keyword extraction, and contextual understanding of the text to analyze the content of employee conversations.

[0587] Mental state assessment

[0588] server:

[0589] The server evaluates the employee's emotions and psychological state based on the analysis results obtained from the NLP engine. This evaluation takes into account positive and negative emotion scores and vocal characteristics (pitch, tone, energy, etc.). Furthermore, a machine learning model is used to calculate the employee's mental evaluation score.

[0590] Examples:

[0591] For example, if negative keywords such as "anger" or "anxiety" frequently appear during a conversation with an agent, the agent's mental evaluation score will be high. This score also reflects changes in tone of voice during conversations with customers, allowing for an accurate understanding of the agent's mental state.

[0592] Collection and visualization of mental assessment data

[0593] server:

[0594] The server aggregates the mental evaluation points for each employee and visualizes them in graphs and charts, allowing managers to track fluctuations in employees' mental states over time.

[0595] Examples:

[0596] For example, the mental health score of Operator A over the past month can be graphed, allowing fluctuations in the evaluation to be observed at a glance. If the graph shows that stress spiked on a particular day, the content of the conversations on that day can be checked in detail.

[0597] Administrator report confirmation

[0598] User (Administrator):

[0599] Managers can access a dashboard provided by the server and view mental health assessment reports for each employee, which display graphs, scores, and conversation highlights, allowing managers to identify specific stressors.

[0600] Examples:

[0601] For example, a manager can check Operator B's high mental health score on the dashboard, investigate the cause, and take appropriate measures.

[0602] Stress management

[0603] User (Administrator):

[0604] Managers can use the reports to schedule interviews and stress management training sessions with employees, and, if necessary, take concrete action, such as temporarily reassigning employees.

[0605] Examples:

[0606] For example, a manager might schedule a meeting with Operator B to identify the cause of stress and provide mindfulness training to improve the next mental score.

[0607] Prompt Sentence Examples

[0608] The generative AI model generates a natural language explanation for a mental health analysis by inputting the following prompt:

[0609] Please explain your mental health assessment system for customer support employees. The system involves a series of processes to collect, analyze, and evaluate voice data. Please explain what hardware or software is used, and what data processing or calculation is performed, including specific scenarios and examples of use.

[0610] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0611] Step 1:

[0612] Data collection

[0613] The device collects voice data from intercoms and telephone systems, including the ability to record conversations between operators and customers in real time. The device records and stores data of interactions with clients via intercoms and telephone systems.

[0614] Input: Voice of conversation between operator and customer

[0615] Output: Real-time audio data

[0616] Step 2:

[0617] Sending audio data

[0618] The device transmits the acquired audio data to the server in real time using protocols such as WebSocket and RTSP. For example, when a customer complains that their product has not arrived, the audio is immediately transmitted to the server.

[0619] Input: Real-time audio data

[0620] Output: Audio data sent to the server

[0621] Step 3:

[0622] Converting audio data to text

[0623] The server sends the received voice data to a voice recognition engine and converts it into text data. For example, the voice recognition engine uses the Google Cloud Speech-to-Text API. The server sends the voice data to the API and receives text data as character information.

[0624] Input: Audio data transferred to the server

[0625] Output: Text data generated by the speech recognition engine

[0626] Step 4:

[0627] Text data analysis

[0628] The server then sends the converted text data to a natural language processing engine to perform sentiment analysis, keyword extraction, and context understanding. For example, it detects whether the text contains keywords such as "anger" or "frustration." The server inputs the text data into the analysis engine and obtains an emotion score and a keyword list as output.

[0629] Input: Text data generated by a speech recognition engine

[0630] Output: Sentiment score and keyword list analyzed by a natural language processing engine

[0631] Step 5:

[0632] Mental state assessment

[0633] The server evaluates the employee's emotions and psychological state based on the analysis results obtained from the NLP engine. This evaluation also includes vocal characteristics (pitch, tone, energy, etc.). The server inputs the analysis results and calculates the employee's mental evaluation score. For example, if the emotion "anger" appears frequently, the mental evaluation score will be high.

[0634] Input: Sentiment scores and keyword list analyzed by a natural language processing engine

[0635] Output: Mental evaluation points calculated by the server

[0636] Step 6:

[0637] Collection and visualization of mental assessment data

[0638] The server aggregates the mental evaluation points for each employee and visualizes them in graphs and charts. The evaluation results are displayed in daily, weekly, monthly, etc. For example, Operator A's mental score for the past month is displayed in a graph.

[0639] Input: Mental evaluation points for each employee

[0640] Output: Visualized graphs and charts

[0641] Step 7:

[0642] Administrator report confirmation

[0643] The user (administrator) accesses the dashboard provided by the server and checks the mental assessment report for each employee. The report includes analysis results, graphs, scores, and conversation highlights. The administrator can obtain detailed reports from the dashboard and check stress factors.

[0644] Input: Visualized graphs, charts, detailed reports

[0645] Output: Mental evaluation report reviewed by the administrator

[0646] Step 8:

[0647] Stress management

[0648] Based on the report, the user (manager) schedules interviews with employees and stress management training. If necessary, the user can take specific countermeasures, such as reassignment. For example, if Operator B shows signs of high stress, the user can schedule an interview and take specific countermeasures.

[0649] Input: Mental evaluation report reviewed by administrator

[0650] Output: Plans for scheduled interviews and stress management training

[0651] (Application example 1)

[0652] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0653] Employee mental health has a significant impact on the quality and efficiency of work. However, in physical stores, it is difficult to grasp the mental state of employees in real time and provide appropriate care. There are limited means to accurately assess the level of stress employees are experiencing while interacting with customers and to take immediate countermeasures. This can lead to an accumulation of employee stress, which can lead to reduced work efficiency and mental health problems.

[0654] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0655] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for visualizing the collected mental evaluation points over time, and means for collecting voice data from a microphone built into the smart glasses and transmitting it to the server in real time, thereby enabling employees to evaluate the mental states of customers in real time while interacting with them, and allowing managers to provide immediate care.

[0656] "Voice data" refers to audio information of conversations between employees and customers.

[0657] "Text data" is voice data converted into character information that is easy to analyze.

[0658] The "means for assessing emotions and psychological states" refers to a means for analyzing the emotions and psychological states of employees based on text data and obtaining the assessment results.

[0659] A "means for visualizing mental state" is a means for understanding mental state by visually displaying the evaluation results.

[0660] The "means for generating a report" refers to a means for creating a report summarizing the mental state of employees based on the evaluation results.

[0661] A "means for extracting vocal timbre features" is a means for extracting vocal features such as pitch, tone, and energy from audio data.

[0662] "Real-time means" means means for collecting, converting, and analyzing audio data in real time.

[0663] "Smart glasses" are glasses-type devices for collecting and transmitting audio data in real time.

[0664] A "server" is a computer system for analyzing collected voice data and processing evaluation results.

[0665] "Mental evaluation points" are indicators that quantitatively represent an employee's mental state and are calculated based on the analysis results.

[0666] The present invention relates to a system that utilizes smart glasses to efficiently monitor the mental health of employees in brick-and-mortar stores while they are serving customers. Specific embodiments are described below.

[0667] System configuration

[0668] The system includes a set of means for collecting, analyzing, and evaluating audio data.

[0669] 1. Audio data collection method:

[0670] The system uses microphones built into smart glasses worn by store employees to collect conversations between employees and customers in real time.

[0671] The collected audio data is transmitted to a server using WebSocket or a similar real-time transmission protocol.

[0672] 2. Means of converting audio data to text data:

[0673] The server converts the received voice data into text data using a voice recognition engine (e.g., Google Cloud Speech-to-Text API).

[0674] This conversion process converts the audio data into a format that is easier to analyze.

[0675] 3. Instruments for assessing emotional and psychological states:

[0676] The server passes the converted text data to a natural language processing (NLP) engine to perform sentiment analysis (e.g., Google Cloud Natural Language API).

[0677] The NLP engine extracts emotional scores such as positive or negative and specific keywords from text data to evaluate the mental state of employees.

[0678] 4. Ways to visualize your mental state:

[0679] The server calculates and visualizes mental evaluation points based on the evaluation results obtained from the NLP engine.

[0680] This visualization includes a means for displaying the time-varying change in the evaluation points as a graph or chart.

[0681] 5. To generate a report:

[0682] The server generates reports showing the mental state of each employee and displays them on a dashboard accessible to managers.

[0683] The report includes detailed rating points, graphs of changes over time, and highlights of key conversations.

[0684] Specific examples of implementation

[0685] For example, while Employee A is wearing smart glasses and interacting with customers, their conversation is collected as voice data. The collected voice data is sent to a server in real time and converted into text data by a voice recognition engine. The text data is analyzed by an NLP engine to evaluate Employee A's emotion score and psychological state. The evaluation results are visualized on a dashboard for managers, allowing them to monitor Employee A's mental health and provide care as needed.

[0686] Prompt Sentence Examples

[0687] I would like to build a system that collects customer voice data, converts it to text in real time, and performs sentiment analysis. Specifically, please generate sample code for a system that uses a speech recognition engine to convert speech to text, and a natural language processing engine to perform sentiment analysis.

[0688] The above is a specific embodiment for carrying out the present invention. By utilizing this system, it is possible to monitor the mental health of employees in real time and provide appropriate care.

[0689] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0690] Step 1:

[0691] A microphone built into the smart glasses collects conversations between employees and customers in real time.

[0692] Input: Employee-customer conversation.

[0693] Output: Audio data.

[0694] To send the audio data to the server, the audio data is transferred in real time using the WebSocket protocol, etc. Specifically, the audio data obtained from the smart glasses is received via a WebSocket connection.

[0695] Step 2:

[0696] The server passes the received voice data to a voice recognition engine, which converts the data into text data.

[0697] Input: Audio data.

[0698] Output: Text data.

[0699] Specifically, the voice data is converted into a string of characters (text data) using the Google Cloud Speech-to-Text API, etc. The converted text data is saved in a format suitable for subsequent analysis.

[0700] Step 3:

[0701] The server passes the text data to a natural language processing (NLP) engine for sentiment analysis.

[0702] Input: Text data.

[0703] Output: Sentiment analysis results (sentiment score, keywords, etc.).

[0704] Specifically, it uses an NLP engine such as Google Cloud Natural Language API to extract sentiment scores and important keywords from text data to assess employee emotions and psychological states.

[0705] Step 4:

[0706] The server calculates the employee's mental evaluation points based on the emotion analysis results.

[0707] Input: Sentiment analysis results.

[0708] Output: Mental evaluation points.

[0709] Specifically, the mental state of employees is quantified based on the emotion score and the frequency of keywords, and a mental evaluation point is calculated, allowing for a quantitative evaluation of the mental state of employees.

[0710] Step 5:

[0711] The server generates graphs and charts to visualize the mental assessment points and displays them on an administrator dashboard.

[0712] Input: Mental Assessment Points.

[0713] Output: Visualized data (graphs, charts).

[0714] Specifically, a web-based dashboard is used to display the collected mental health assessment scores over time in graphs and charts, allowing managers to provide appropriate care based on this information.

[0715] Step 6:

[0716] The server generates reports showing each employee's mental state, which managers can check regularly.

[0717] Input: Mental evaluation points, conversation highlights.

[0718] Output: Report.

[0719] Specifically, the system generates regular reports containing detailed mental health assessment scores, graphs of changes over time, and highlights of important conversations, and makes them accessible to managers, allowing them to understand trends in employee mental health and take necessary measures.

[0720] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0721] This invention is a system for managing the mental health of employees in customer support and customer service, and includes a series of processes for collecting voice data in real time and evaluating their emotions and psychological state. In particular, by combining it with an emotion engine, the system can more precisely recognize the user's emotions and reflect them in the mental health evaluation.

[0722] Specific Embodiments

[0723] Device:

[0724] It captures voice data from intercoms and telephone systems installed in stores and call centers, which involves recording conversations between operators and customers in real time.

[0725] The captured audio data is sent to the server in real time using a protocol for audio streaming (e.g., WebSocket or RTSP).

[0726] server:

[0727] 1. Audio data conversion:

[0728] The server stores the voice data received from the device in temporary storage and converts it into text data by calling a voice recognition engine, such as the Google Cloud Speech-to-Text API.

[0729] 2. Text data and NLP analysis:

[0730] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information from the text.

[0731] 3. Voice timbre feature extraction:

[0732] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[0733] 4. Applying the Emotion Engine:

[0734] The extracted vocal features are input into an emotion engine to recognize the corresponding emotion, such as anger, joy, or sadness.

[0735] 5. Mental status assessment:

[0736] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state. The evaluation includes scoring positive and negative emotions and calculating mental evaluation points.

[0737] Examples:

[0738] For example, if an agent frequently uses keywords that express negative emotions, such as "anxiety" or "anger," while talking to customers, the emotion engine will recognize this and, if the characteristics of the agent's voice (e.g., a sudden rise in pitch or an emphasized tone) indicate the corresponding emotion, the agent's mental evaluation score will be high. This evaluation is displayed as a quantitative score and can be checked by managers.

[0739] 6. Report generation and visualization:

[0740] The system aggregates each employee's mental assessment score and visualizes their mental state over time in graphs and charts. The report includes the mental assessment score, related conversation content, and changes in tone of voice.

[0741] Examples:

[0742] For example, if a graph of Operator A's mental health score for the past month shows a spike in stress on a specific date, the system will display details of the conversations that took place on that date, allowing managers to identify the root cause of the problem and take appropriate measures.

[0743] User (Administrator):

[0744] 1. Report Review and Care:

[0745] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[0746] Consider consultation and care options and implement specific measures, such as stress management training or temporary reassignment.

[0747] Examples:

[0748] A manager reviews Operator B's report and finds that his mental health score for the past week was high. The manager can schedule a meeting with Operator B to identify the cause of his stress and provide mindfulness training to relieve stress, thereby improving his mental health score next time.

[0749] The above is a specific embodiment of the invention that combines an emotion engine. This system allows managers to efficiently and accurately evaluate the mental health of employees and provide appropriate care.

[0750] The processing flow will be explained below.

[0751] Step 1:

[0752] Device:

[0753] Voice data is acquired in real time from intercoms and telephone systems installed in stores and call centers.

[0754] The acquired audio data is sent to the server using an audio streaming protocol (e.g., WebSocket or RTSP).

[0755] Step 2:

[0756] server:

[0757] Stores audio data received from the device in temporary storage, which ensures a consistent data retention for subsequent processing.

[0758] Step 3:

[0759] server:

[0760] Call a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the voice data into text data.

[0761] Step 4:

[0762] server:

[0763] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information.

[0764] Step 5:

[0765] server:

[0766] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[0767] Step 6:

[0768] server:

[0769] The extracted vocal features are input into an emotion engine to recognize the corresponding emotion, such as anger, joy, or sadness.

[0770] Step 7:

[0771] server:

[0772] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state, including scoring positive and negative emotions and calculating mental evaluation points.

[0773] Examples:

[0774] For example, if an agent frequently uses words like "anxiety" or "anger" while talking to customers, the emotion engine will recognize this as a negative emotion. Furthermore, if the tone of voice features (higher pitch or emphasized tone) indicate the corresponding emotion, the agent's mental evaluation score will be higher. This evaluation is displayed as a quantitative score and can be checked by managers.

[0775] Step 8:

[0776] server:

[0777] The mental evaluation points for each employee are aggregated and their changing mental state over time is visualized in graphs and charts.

[0778] Step 9:

[0779] server:

[0780] Automatically generate a mental status report, including mental assessment scores, relevant conversation content, and tone of voice fluctuations.

[0781] Examples:

[0782] For example, if the mental health score of Operator A for the past month is graphed and stress spikes on a particular day, the details of the conversations that took place on that day will be displayed in detail. This display allows managers to identify the root cause of the problem and take appropriate measures.

[0783] Step 10:

[0784] User (Administrator):

[0785] Reports are viewed through a server dashboard, which visualizes each employee's mental assessment score and conversation highlights.

[0786] Step 11:

[0787] User (Administrator):

[0788] Based on the report, meetings with employees can be scheduled to identify the causes of stress, and specific countermeasures such as stress management training or temporary reassignment can be considered.

[0789] Examples:

[0790] The manager checks Operator B's report and finds that his mental health score for the past week is high. The manager schedules a meeting with Operator B to identify the cause of his stress. Based on this, the manager provides mindfulness training to relieve stress and improve his mental health score for the next time.

[0791] The above is the specific processing flow of the invention that combines the emotion engine and the detailed operation of each step. This system allows managers to efficiently and accurately evaluate the mental health of employees and provide appropriate care.

[0792] Example 2

[0793] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0794] To manage employee mental health, there is a need for a method to accurately and in real time assess emotions and psychological states from on-site voice data. However, conventional methods have the problem that they cannot analyze voice data in real time and are not precise enough in recognizing emotions and assessing mental states, making it difficult for managers to take appropriate action quickly.

[0795] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for temporarily saving voice data, means for converting voice data into text data, means for analyzing the text data and extracting emotion and context information, means for extracting tone of voice features from the voice data, means for recognizing emotion based on the tone of voice features, means for evaluating mental states by integrating emotion recognition results and text analysis results, means for visualizing mental states based on evaluation results, and means for generating reports showing mental states. This allows employees' emotions and psychological states to be accurately evaluated in real time, enabling managers to take prompt and appropriate action.

[0796] "Audio data" refers to information that records audio in digital format.

[0797] "Temporarily storing" means holding the received data in a storage area for a short period of time to prevent it from being lost.

[0798] "Converting to text data" refers to the process of converting voice data into character information.

[0799] "Analyzing text data and extracting emotional and contextual information" means using natural language processing technology to extract emotional, keyword, and contextual information from text.

[0800] "Extracting vocal features" means extracting acoustic features such as pitch, tone, and energy from speech data.

[0801] "Emotion recognition" means identifying the speaker's emotion based on the extracted vocal timbre features.

[0802] "Evaluating the mental state" means quantitatively evaluating the speaker's psychological state by combining the emotion recognition results and text analysis results.

[0803] "Visualizing" means converting data into a format that is easy to understand visually, such as a graph or chart.

[0804] "Generating a report" means creating a report based on the evaluation results so that the administrator can check it.

[0805] "Real-time" means that data is processed immediately after it is generated, with little delay.

[0806] This system uses hardware installed in stores or call centers, a server, and specialized software to process a series of data in order to manage the mental health of employees. A specific embodiment of this system is described below.

[0807] Terminal

[0808] The device acquires voice data from intercoms and telephone systems installed in stores and call centers, and transmits the acquired voice data to a server in real time using an audio streaming protocol such as WebSocket or RTSP.

[0809] server

[0810] The server processes the audio data in the following steps:

[0811] 1. Temporary storage of audio data:

[0812] The server temporarily stores the voice data received from the terminal in storage to ensure data security and facilitate subsequent processing.

[0813] 2. Audio to text conversion:

[0814] The server uses the Google Cloud Speech-to-Text API to convert the temporarily stored voice data into text data. For example, a conversation between an operator and a customer, such as "Hello, how can I help you?", is converted directly into text.

[0815] 3. NLP analysis of text data:

[0816] The server passes the converted text data to an NLP engine (e.g., spaCy or BERT) to analyze the sentiment, keywords, and context of the text. For example, from the phrase "I'm feeling very stressed today," "stress" is extracted as an emotional keyword.

[0817] 4. Extracting voice timbre features from speech data:

[0818] The server applies an acoustic feature analysis algorithm to extract vocal timbre features such as pitch, tone, and energy from the audio data. For example, a rise in vocal pitch indicates a change in emotion.

[0819] 5. Emotion recognition based on voice features:

[0820] The server inputs the extracted vocal features into an emotion engine to recognize emotions, such as "tension," "joy," and "anger."

[0821] 6. Mental Status Assessment:

[0822] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to calculate positive and negative emotion scores and mental evaluation points. For example, the result may be "positive score 70%" or "negative score 30%."

[0823] 7. Visualization of mental state and report generation:

[0824] The server aggregates the mental assessment scores for each employee and visualizes fluctuations in their mental state over time in graphs and charts. For example, it can display the weekly fluctuations in mental scores in a line graph, and show the reason for a sudden rise in stress on a particular date along with detailed conversation content.

[0825] User (Administrator)

[0826] Managers access a dashboard provided by the server and check the mental health assessment report for each employee. The report displays graphs, scores, conversation highlights, and more. Based on this, managers can consider interviews and care measures and take specific countermeasures. For example, for employees with high levels of stress, they can suggest "mindfulness training" or "temporary reassignment."

[0827] Examples:

[0828] If an agent frequently uses keywords indicating negative emotions, such as "anxiety" or "anger," while talking to a customer, the emotion engine will identify this. Furthermore, if characteristics indicating anxiety or anger are detected from an analysis of the agent's tone of voice, the agent's mental state score will be higher. This score will be reported to the manager. The report will display a graph showing the "progression of mental state over the past month," visualizing, for example, spikes in stress on specific days.

[0829] Example prompt sentence:

[0830] "To assess the mental health of employees over the past month, please generate a report based on the following information: speech-to-text transcription results, NLP analysis of emotions and keywords, and voice timbre features. Please also include the mental health assessment score for each date and the relevant conversation content."

[0831] This system clarifies the roles of servers, terminals, and users, making it possible to efficiently manage employee mental health in real time.

[0832] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0833] Step 1:

[0834] The terminal acquires voice data from intercoms or telephone systems installed in stores or call centers. At this time, conversations between operators and customers are recorded in real time. The input is voice data from the intercom or telephone system, and the output is an audio file acquired in real time.

[0835] Step 2:

[0836] The audio data captured by the device is sent to the server in real time. This process uses audio streaming protocols such as WebSocket and RTSP. The input is the audio data captured by the device, and the output is the audio streaming data sent to the server.

[0837] Step 3:

[0838] The server temporarily stores the audio data received from the device in storage. This storage process prevents loss of audio data and makes subsequent processing easier. The input is audio data acquired as streaming data, and the output is an audio file stored in temporary storage.

[0839] Step 4:

[0840] The server uses the Google Cloud Speech-to-Text API to convert the temporarily stored voice data into text data. The input is the audio file stored in temporary storage, and the output is the converted text data. This conversion process allows the voice information to be treated as text information.

[0841] Step 5:

[0842] The server passes the text data to a natural language processing (NLP) engine (e.g., spaCy or BERT) to extract sentiment, keywords, and contextual information from the text. The input is the text data, and the output is the sentiment, keywords, and contextual information extracted through analysis. For example, the keyword "stress" is extracted from the text "I'm feeling very stressed today."

[0843] Step 6:

[0844] The server applies an acoustic feature analysis algorithm to extract vocal timbre features such as pitch, tone, and energy from the audio data. The input is a temporarily saved audio file, and the output is the extracted vocal timbre features. For example, a rise in voice pitch indicates a change in emotion.

[0845] Step 7:

[0846] The server inputs the extracted voice timbre features into an emotion engine to recognize the corresponding emotion. The input is the voice timbre features obtained from acoustic feature analysis, and the output is the recognized emotion (e.g., tension, joy, anger). The emotion engine then quantifies the extracted features as an emotion.

[0847] Step 8:

[0848] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to calculate positive and negative emotion scores and mental evaluation points for employees. The input is the text analysis results and emotion recognition results, and the output is the emotion score and mental evaluation points for each employee. For example, evaluations such as "positive score 70%" and "negative score 30%" are possible.

[0849] Step 9:

[0850] The server aggregates the mental assessment points for each employee and visualizes their mental state over time in graphs and charts. The input is each employee's mental assessment point, and the output is their mental state visualized as a graph or chart. For example, a line graph can be used to display the progress of mental state by week, showing the reason for a sudden rise in stress on a particular day.

[0851] Step 10:

[0852] The user (manager) accesses the dashboard provided by the server and checks the mental health assessment report for each employee. The input is the mental health assessment report provided by the server, and the output is report information that the manager can check and use. Based on this, the manager considers interviews and care measures and takes specific countermeasures (e.g., mindfulness training or temporary reassignment).

[0853] The above are the processing steps of the program for this system, and the specific operations performed at each step, as well as the inputs and outputs, have been explained.

[0854] (Application example 2)

[0855] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0856] Efficiently managing employee mental health is important in customer support and customer service. However, conventional systems have difficulty evaluating mental health in real time, making it difficult to immediately grasp changes in employees' stress levels and emotions. In addition, there are limited ways for employees to check their own mental state in real time, making it difficult to implement appropriate stress management.

[0857] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0858] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for visualizing the mental states based on the evaluation results, means for generating a report showing the mental states, and means for displaying the results on a smart device. This allows emotions and psychological states to be evaluated in real time from the voice data, and allows employees to check their mental states themselves via their smart devices.

[0859] "Voice data" means the acoustic signals of a conversation between an operator and a customer.

[0860] "Means for collecting" refers to a device or method for capturing and inputting audio data into the system.

[0861] "Text data" is voice data converted into character information.

[0862] "Means for converting into text data" refers to a device or method that converts voice data into text information using voice recognition technology.

[0863] "Means for analyzing and assessing emotions and psychological states" means a device or method for analyzing and assessing an employee's emotions and psychological states using text data and tone of voice features.

[0864] "Visualization means" refers to a device or method for displaying the evaluation results in a visual form such as a graph or chart.

[0865] A "means for generating a report" is a device or method for creating a detailed written report based on the results of the evaluation.

[0866] "Means for displaying results on a smart device" refers to a device or method for displaying the results of a mental state assessment in real time on a device such as smart glasses or a wearable device.

[0867] The system for implementing this invention mainly performs a series of processes, including collecting voice data, analyzing emotions, and evaluating mental states. The system is realized by linking terminals, a server, and users.

[0868] Terminal

[0869] The terminals are mainly smart devices (e.g., smart glasses). These terminals have built-in microphones that collect voice data from workers. The collected voice data is transmitted to the server in real time. WebSocket or RTSP is used as the protocol for audio streaming.

[0870] server

[0871] The server performs the following steps:

[0872] 1. Audio data conversion:

[0873] The server temporarily stores the voice data acquired from the device and converts it into text data using the Google Cloud Speech-to-Text API.

[0874] 2. Text data analysis:

[0875] The converted text data is analyzed using a natural language processing (NLP) engine, using spaCy and BERT to extract sentiment, keywords, and contextual information from the text.

[0876] 3. Voice timbre feature extraction:

[0877] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[0878] 4. Applying the Emotion Engine:

[0879] The extracted voice features are input into an emotion engine to recognize the corresponding emotion. BERT or other generative AI models may be used as emotion models.

[0880] 5. Mental status assessment:

[0881] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state, and the evaluation results are displayed as a positive or negative emotion score.

[0882] 6. Real-time display:

[0883] The evaluation results are sent in real time to smart devices so that employees can view them.

[0884] User (Administrator)

[0885] Users are responsible for managing the mental health of their employees and providing appropriate care when necessary. Specifically, this includes:

[0886] 1. Report Review and Care:

[0887] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[0888] We will consider options for interviews and care, and provide stress management training and temporary reassignments.

[0889] Specific examples

[0890] For example, if a store employee wearing smart glasses speaks in a "high-pitched, emphasized tone" while interacting with a customer, the voice data is sent to a server, where real-time emotion analysis is performed. As a result, the employee's stress level is assessed as rising, and the smart glasses display shows "high stress." Based on this information, managers can instruct employees to take a break. Managers can also take countermeasures based on this data later.

[0891] Prompt Sentence Examples

[0892] "The employee's voice is high-pitched and stressed. We suggest they take a 10-minute break."

[0893] "Your mental health score for the past week has been high. I suggest you take stress management training."

[0894] Thus, in order to implement this invention, it is necessary to build a system in which the components of the server, terminal, and user work together to evaluate and manage the mental state of employees in real time.

[0895] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0896] Step 1:

[0897] Audio data collection

[0898] The device collects employee voice data.

[0899] Input: Audio data captured from the built-in microphone of the smart glasses.

[0900] How it works: The built-in microphone records the conversation between the operator and the customer and sends the audio data to the server in real time.

[0901] Step 2:

[0902] Converting audio data to text data

[0903] The server converts the received voice data into text data.

[0904] Input: Audio data sent from the device.

[0905] Output: Text data.

[0906] What it does: It uses the Google Cloud Speech-to-Text API to convert audio data into text. A speech recognition model analyzes the acoustic signal and generates a corresponding string of characters.

[0907] Step 3:

[0908] Text data analysis

[0909] The server analyzes the converted text data and evaluates the emotions and psychological state.

[0910] Input: Text data.

[0911] Output: Emotion score and mental state assessment results.

[0912] Specific operation: Text data is input into a natural language processing (NLP) engine (e.g., spaCy, BERT) to analyze and extract sentiment, keywords, and contextual information.

[0913] Step 4:

[0914] Voice timbre feature extraction

[0915] The server extracts voice characteristics from the voice data.

[0916] Input: Audio data.

[0917] Output: Voice characteristics (pitch, tone, energy, etc.).

[0918] Specific operation: Apply an acoustic feature analysis algorithm (e.g., librosa) to extract features such as pitch, tone, and energy from the audio signal.

[0919] Step 5:

[0920] Applying the Emotion Engine

[0921] The server inputs the voice features into the emotion engine and recognizes the corresponding emotion.

[0922] Input: Voice timbre features.

[0923] Output: Emotion recognition results.

[0924] Specific operation: Using a generative AI model (e.g., BERT), emotions are classified from tone of voice features, and emotions such as anger, joy, and sadness are recognized.

[0925] Step 6:

[0926] Mental state assessment

[0927] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to evaluate the mental state of employees.

[0928] Input: Emotion score, voice features, and text analysis results.

[0929] Output: Mental state assessment score.

[0930] Specific operations: Evaluate overall mental state by assigning positive and negative emotional scores and calculating mental evaluation points.

[0931] Step 7:

[0932] Real-time display

[0933] The server transmits the evaluation results to the smart device in real time and displays them.

[0934] Input: Mental State Assessment score.

[0935] Output: Evaluation results displayed on smart devices.

[0936] Specific operation: The evaluation results are displayed on the smart glasses display, allowing employees to check their own mental state in real time.

[0937] Step 8:

[0938] Report Generation

[0939] The server generates a report based on the mental state evaluation results and provides it to the user.

[0940] Input: Accumulated data of evaluation results.

[0941] Output: Detailed mental state report.

[0942] Specific actions: Generate graphs and charts to create reports that visualize fluctuations in mental state for each employee, and display them on a dashboard for managers to review.

[0943] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0944] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0945] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0946] [Third embodiment]

[0947] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0948] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0949] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0950] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0951] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0952] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0953] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0954] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0955] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0956] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0957] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0958] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0959] This invention is a system for managing the mental health of employees in customer support and customer service work, and includes a series of processes for collecting, analyzing, and evaluating voice data. Specific embodiments for carrying out this invention are described below.

[0960] 1. Data Collection

[0961] Device:

[0962] It captures voice data from intercoms and telephone systems installed in stores and call centers, which involves recording conversations between operators and customers in real time.

[0963] The captured audio data is sent to the server in real time using a protocol for audio streaming (e.g., WebSocket or RTSP).

[0964] 2. Audio data analysis

[0965] server:

[0966] The server sends the voice data received from the device to a voice recognition engine, such as the Google Cloud Speech-to-Text API, and converts it into text data.

[0967] The converted text data is passed to a natural language processing (NLP) engine for text analysis, which analyzes the sentiment of the conversation, extracts keywords, and understands the context.

[0968] 3. Mental state assessment

[0969] server:

[0970] The analysis results from the NLP engine are used to assess the employee's emotional and psychological state, taking into account positive and negative emotional scoring as well as vocal characteristics (pitch, tone, energy, etc.).

[0971] A machine learning model is used to calculate employees' mental assessment points.

[0972] Examples:

[0973] For example, if negative keywords such as "anger" or "anxiety" frequently appear in the analysis results for a certain operator, the operator's mental evaluation score will be high. This score is also evaluated based on changes in tone of voice during conversations with customers, allowing for an accurate understanding of the operator's mental state.

[0974] 4. Report generation and visualization

[0975] server:

[0976] The mental evaluation points for each employee are aggregated and changes over time are visualized in graphs and charts.

[0977] The report includes mental assessment scores and related conversation highlights, allowing managers to understand specifically what situations increase stress levels.

[0978] Examples:

[0979] For example, if a graph of Operator A's mental health score for the past month shows a spike in stress on a specific date, the system will display details of the conversations that took place on that date, allowing managers to identify the root cause of the problem and take appropriate measures.

[0980] 5. Review and care of reports

[0981] User (Administrator):

[0982] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[0983] Consider consultation and care options and implement specific measures, such as stress management training or temporary reassignment.

[0984] Examples:

[0985] A manager checks the report of Operator B and finds that the mental health score for the past week was high. The manager schedules a meeting with Operator B, identifies the cause of stress, and provides mindfulness training to relieve stress, thereby improving the next mental health score.

[0986] The above is a specific embodiment for carrying out the present invention. By utilizing this system, it is possible to efficiently manage the mental health of employees and improve the quality of work.

[0987] The processing flow will be explained below.

[0988] Step 1:

[0989] Device:

[0990] Voice data is acquired in real time from intercoms and telephone systems installed in stores and call centers.

[0991] The acquired audio data is sent to the server using a protocol for audio streaming (e.g., WebSocket or RTSP).

[0992] Step 2:

[0993] server:

[0994] Stores audio data received from the device in temporary storage, which ensures a consistent data retention for subsequent processing.

[0995] Step 3:

[0996] server:

[0997] Call a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the voice data into text data.

[0998] Step 4:

[0999] server:

[1000] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information from the text.

[1001] Step 5:

[1002] server:

[1003] Extract voice characteristics (pitch, tone, energy, etc.) from the audio data by applying an acoustic feature analysis algorithm.

[1004] Step 6:

[1005] server:

[1006] Based on the analysis results from the NLP engine and the features of the voice tone, the emotional and psychological state is evaluated and a mental evaluation score is calculated. Here, a machine learning model is used to quantify the mental evaluation score.

[1007] Step 7:

[1008] server:

[1009] The mental evaluation points for each employee are aggregated and their changing mental state over time is visualized in graphs and charts.

[1010] Step 8:

[1011] server:

[1012] Automatically generate a mental status report, including mental assessment scores, relevant conversation content, and tone of voice fluctuations.

[1013] Step 9:

[1014] User (Administrator):

[1015] Reports are viewed through the server's dashboard, which visualizes each employee's mental assessment score and conversation highlights.

[1016] Step 10:

[1017] User (Administrator):

[1018] Schedule a meeting with the employee based on the report. Identify the causes of stress and take measures to address them, such as training programs or reassignment.

[1019] This is the specific process flow. This system allows managers to efficiently and accurately evaluate the mental health of their employees and provide appropriate care.

[1020] Example 1

[1021] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1022] In conventional customer support operations, the mental health of employees was not adequately managed, resulting in a decline in the quality of work, and problems such as employee stress and increased turnover. In addition, there was also the issue of the time required to collect, analyze, and evaluate voice data, making it difficult to respond in real time.

[1023] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1024] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for transmitting the collected voice data to the server in real time, means for sending the converted text data to a natural language processing engine for emotion analysis, means for accumulating mental evaluation points over time and displaying them in graphs or charts, and means for providing a dashboard for managers to check mental evaluation reports. This enables real-time collection and immediate analysis of voice data, making it possible to efficiently manage employee mental health and improve the quality of work.

[1025] "Voice data" means digital voice signals that record the conversation between an operator and a customer.

[1026] "Text data" is voice data converted into character information.

[1027] "Means for assessing emotions and psychological states" refers to technological means for analyzing text data and quantifying and assessing employees' emotions (positive and negative) and psychological states.

[1028] "Visualization means" refers to a means for displaying the evaluation results in graphs or charts so that users can understand them visually.

[1029] The "means for generating a report" refers to a means for creating a detailed report on the mental state of an employee based on the evaluation results.

[1030] "Means of transmitting in real time" refers to the technical means of instantly transferring collected voice data to a server.

[1031] A "natural language processing engine" refers to a processing engine that performs sentiment analysis, keyword extraction, context understanding, etc. on text data.

[1032] "Means for accumulating and displaying mental evaluation points over time" refers to a means for aggregating employees' mental evaluation points over a fixed period of time and visually displaying the fluctuations in graphs or charts.

[1033] "Dashboard" refers to a web-based or software-based platform that provides an interface through which managers can view employee mental health assessment reports in real time.

[1034] This invention is a system for managing the mental health of employees in customer support and customer service. The system includes a series of processes for collecting, analyzing, and evaluating voice data. Specific embodiments for implementing this invention are described below.

[1035] Data collection

[1036] Device:

[1037] Intercoms and telephone systems installed in stores and call centers collect voice data, including the ability to record conversations between operators and customers in real time. The collected voice data is then sent to a server in real time. Protocols such as WebSocket and RTSP are used to stream the data.

[1038] Sending audio data

[1039] Device:

[1040] The acquired voice data is sent to the server in real time. This process is carried out using protocols such as WebSocket and RTSP. For example, voice data collected during conversations in a store or during phone calls is immediately transferred to the server.

[1041] Converting audio data to text

[1042] server:

[1043] The server sends the received voice data to a speech recognition engine (e.g., Google Cloud Speech-to-Text API) and converts it into text data. The speech recognition engine converts voice into text with high accuracy, and text data is generated in near real time.

[1044] Text data analysis

[1045] server:

[1046] The converted text data is passed to a natural language processing (NLP) engine (e.g., Google Cloud Natural Language API), which performs sentiment analysis, keyword extraction, and contextual understanding of the text to analyze the content of employee conversations.

[1047] Mental state assessment

[1048] server:

[1049] The server evaluates the employee's emotions and psychological state based on the analysis results obtained from the NLP engine. This evaluation takes into account positive and negative emotion scores and vocal characteristics (pitch, tone, energy, etc.). Furthermore, a machine learning model is used to calculate the employee's mental evaluation score.

[1050] Examples:

[1051] For example, if negative keywords such as "anger" or "anxiety" frequently appear during a conversation with an agent, the agent's mental evaluation score will be high. This score also reflects changes in tone of voice during conversations with customers, allowing for an accurate understanding of the agent's mental state.

[1052] Collection and visualization of mental assessment data

[1053] server:

[1054] The server aggregates the mental evaluation points for each employee and visualizes them in graphs and charts, allowing managers to track fluctuations in employees' mental states over time.

[1055] Examples:

[1056] For example, the mental health score of Operator A over the past month can be graphed, allowing fluctuations in the evaluation to be observed at a glance. If the graph shows that stress spiked on a particular day, the content of the conversations on that day can be checked in detail.

[1057] Administrator report confirmation

[1058] User (Administrator):

[1059] Managers can access a dashboard provided by the server and view mental health assessment reports for each employee, which display graphs, scores, and conversation highlights, allowing managers to identify specific stressors.

[1060] Examples:

[1061] For example, a manager can check Operator B's high mental health score on the dashboard, investigate the cause, and take appropriate measures.

[1062] Stress management

[1063] User (Administrator):

[1064] Managers can use the reports to schedule interviews and stress management training sessions with employees, and, if necessary, take concrete action, such as temporarily reassigning employees.

[1065] Examples:

[1066] For example, a manager might schedule a meeting with Operator B to identify the cause of stress and provide mindfulness training to improve the next mental score.

[1067] Prompt Sentence Examples

[1068] The generative AI model generates a natural language explanation for a mental health analysis by inputting the following prompt:

[1069] Please explain your mental health assessment system for customer support employees. The system involves a series of processes to collect, analyze, and evaluate voice data. Please explain what hardware or software is used, and what data processing or calculation is performed, including specific scenarios and examples of use.

[1070] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1071] Step 1:

[1072] Data collection

[1073] The device collects voice data from intercoms and telephone systems, including the ability to record conversations between operators and customers in real time. The device records and stores data of interactions with clients via intercoms and telephone systems.

[1074] Input: Voice of conversation between operator and customer

[1075] Output: Real-time audio data

[1076] Step 2:

[1077] Sending audio data

[1078] The device transmits the acquired audio data to the server in real time using protocols such as WebSocket and RTSP. For example, when a customer complains that their product has not arrived, the audio is immediately transmitted to the server.

[1079] Input: Real-time audio data

[1080] Output: Audio data sent to the server

[1081] Step 3:

[1082] Converting audio data to text

[1083] The server sends the received voice data to a voice recognition engine and converts it into text data. For example, the voice recognition engine uses the Google Cloud Speech-to-Text API. The server sends the voice data to the API and receives text data as character information.

[1084] Input: Audio data transferred to the server

[1085] Output: Text data generated by the speech recognition engine

[1086] Step 4:

[1087] Text data analysis

[1088] The server then sends the converted text data to a natural language processing engine to perform sentiment analysis, keyword extraction, and context understanding. For example, it detects whether the text contains keywords such as "anger" or "frustration." The server inputs the text data into the analysis engine and obtains an emotion score and a keyword list as output.

[1089] Input: Text data generated by a speech recognition engine

[1090] Output: Sentiment score and keyword list analyzed by a natural language processing engine

[1091] Step 5:

[1092] Mental state assessment

[1093] The server evaluates the employee's emotions and psychological state based on the analysis results obtained from the NLP engine. This evaluation also includes vocal characteristics (pitch, tone, energy, etc.). The server inputs the analysis results and calculates the employee's mental evaluation score. For example, if the emotion "anger" appears frequently, the mental evaluation score will be high.

[1094] Input: Sentiment scores and keyword list analyzed by a natural language processing engine

[1095] Output: Mental evaluation points calculated by the server

[1096] Step 6:

[1097] Collection and visualization of mental assessment data

[1098] The server aggregates the mental evaluation points for each employee and visualizes them in graphs and charts. The evaluation results are displayed in daily, weekly, monthly, etc. For example, Operator A's mental score for the past month is displayed in a graph.

[1099] Input: Mental evaluation points for each employee

[1100] Output: Visualized graphs and charts

[1101] Step 7:

[1102] Administrator report confirmation

[1103] The user (administrator) accesses the dashboard provided by the server and checks the mental assessment report for each employee. The report includes analysis results, graphs, scores, and conversation highlights. The administrator can obtain detailed reports from the dashboard and check stress factors.

[1104] Input: Visualized graphs, charts, detailed reports

[1105] Output: Mental evaluation report reviewed by the administrator

[1106] Step 8:

[1107] Stress management

[1108] Based on the report, the user (manager) schedules interviews with employees and stress management training. If necessary, the user can take specific countermeasures, such as reassignment. For example, if Operator B shows signs of high stress, the user can schedule an interview and take specific countermeasures.

[1109] Input: Mental evaluation report reviewed by administrator

[1110] Output: Plans for scheduled interviews and stress management training

[1111] (Application example 1)

[1112] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1113] Employee mental health has a significant impact on the quality and efficiency of work. However, in physical stores, it is difficult to grasp the mental state of employees in real time and provide appropriate care. There are limited means to accurately assess the level of stress employees are experiencing while interacting with customers and to take immediate countermeasures. This can lead to an accumulation of employee stress, which can lead to reduced work efficiency and mental health problems.

[1114] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1115] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for visualizing the collected mental evaluation points over time, and means for collecting voice data from a microphone built into the smart glasses and transmitting it to the server in real time, thereby enabling employees to evaluate the mental states of customers in real time while interacting with them, and allowing managers to provide immediate care.

[1116] "Voice data" refers to audio information of conversations between employees and customers.

[1117] "Text data" is voice data converted into character information that is easy to analyze.

[1118] The "means for assessing emotions and psychological states" refers to a means for analyzing the emotions and psychological states of employees based on text data and obtaining the assessment results.

[1119] A "means for visualizing mental state" is a means for understanding mental state by visually displaying the evaluation results.

[1120] The "means for generating a report" refers to a means for creating a report summarizing the mental state of employees based on the evaluation results.

[1121] A "means for extracting vocal timbre features" is a means for extracting vocal features such as pitch, tone, and energy from audio data.

[1122] "Real-time means" means means for collecting, converting, and analyzing audio data in real time.

[1123] "Smart glasses" are glasses-type devices for collecting and transmitting audio data in real time.

[1124] A "server" is a computer system for analyzing collected voice data and processing evaluation results.

[1125] "Mental evaluation points" are indicators that quantitatively represent an employee's mental state and are calculated based on the analysis results.

[1126] The present invention relates to a system that utilizes smart glasses to efficiently monitor the mental health of employees in brick-and-mortar stores while they are serving customers. Specific embodiments are described below.

[1127] System configuration

[1128] The system includes a set of means for collecting, analyzing, and evaluating audio data.

[1129] 1. Audio data collection method:

[1130] The system uses microphones built into smart glasses worn by store employees to collect conversations between employees and customers in real time.

[1131] The collected audio data is transmitted to a server using WebSocket or a similar real-time transmission protocol.

[1132] 2. Means of converting audio data to text data:

[1133] The server converts the received voice data into text data using a voice recognition engine (e.g., Google Cloud Speech-to-Text API).

[1134] This conversion process converts the audio data into a format that is easier to analyze.

[1135] 3. Instruments for assessing emotional and psychological states:

[1136] The server passes the converted text data to a natural language processing (NLP) engine to perform sentiment analysis (e.g., Google Cloud Natural Language API).

[1137] The NLP engine extracts emotional scores such as positive or negative and specific keywords from text data to evaluate the mental state of employees.

[1138] 4. Ways to visualize your mental state:

[1139] The server calculates and visualizes mental evaluation points based on the evaluation results obtained from the NLP engine.

[1140] This visualization includes a means for displaying the time-varying change in the evaluation points as a graph or chart.

[1141] 5. To generate a report:

[1142] The server generates reports showing the mental state of each employee and displays them on a dashboard accessible to managers.

[1143] The report includes detailed rating points, graphs of changes over time, and highlights of key conversations.

[1144] Specific examples of implementation

[1145] For example, while Employee A is wearing smart glasses and interacting with customers, their conversation is collected as voice data. The collected voice data is sent to a server in real time and converted into text data by a voice recognition engine. The text data is analyzed by an NLP engine to evaluate Employee A's emotion score and psychological state. The evaluation results are visualized on a dashboard for managers, allowing them to monitor Employee A's mental health and provide care as needed.

[1146] Prompt Sentence Examples

[1147] I would like to build a system that collects customer voice data, converts it to text in real time, and performs sentiment analysis. Specifically, please generate sample code for a system that uses a speech recognition engine to convert speech to text, and a natural language processing engine to perform sentiment analysis.

[1148] The above is a specific embodiment for carrying out the present invention. By utilizing this system, it is possible to monitor the mental health of employees in real time and provide appropriate care.

[1149] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1150] Step 1:

[1151] A microphone built into the smart glasses collects conversations between employees and customers in real time.

[1152] Input: Employee-customer conversation.

[1153] Output: Audio data.

[1154] To send the audio data to the server, the audio data is transferred in real time using the WebSocket protocol, etc. Specifically, the audio data obtained from the smart glasses is received via a WebSocket connection.

[1155] Step 2:

[1156] The server passes the received voice data to a voice recognition engine, which converts the data into text data.

[1157] Input: Audio data.

[1158] Output: Text data.

[1159] Specifically, the voice data is converted into a string of characters (text data) using the Google Cloud Speech-to-Text API, etc. The converted text data is saved in a format suitable for subsequent analysis.

[1160] Step 3:

[1161] The server passes the text data to a natural language processing (NLP) engine for sentiment analysis.

[1162] Input: Text data.

[1163] Output: Sentiment analysis results (sentiment score, keywords, etc.).

[1164] Specifically, it uses an NLP engine such as Google Cloud Natural Language API to extract sentiment scores and important keywords from text data to assess employee emotions and psychological states.

[1165] Step 4:

[1166] The server calculates the employee's mental evaluation points based on the emotion analysis results.

[1167] Input: Sentiment analysis results.

[1168] Output: Mental evaluation points.

[1169] Specifically, the mental state of employees is quantified based on the emotion score and the frequency of keywords, and a mental evaluation point is calculated, allowing for a quantitative evaluation of the mental state of employees.

[1170] Step 5:

[1171] The server generates graphs and charts to visualize the mental assessment points and displays them on an administrator dashboard.

[1172] Input: Mental Assessment Points.

[1173] Output: Visualized data (graphs, charts).

[1174] Specifically, a web-based dashboard is used to display the collected mental health assessment scores over time in graphs and charts, allowing managers to provide appropriate care based on this information.

[1175] Step 6:

[1176] The server generates reports showing each employee's mental state, which managers can check regularly.

[1177] Input: Mental evaluation points, conversation highlights.

[1178] Output: Report.

[1179] Specifically, the system generates regular reports containing detailed mental health assessment scores, graphs of changes over time, and highlights of important conversations, and makes them accessible to managers, allowing them to understand trends in employee mental health and take necessary measures.

[1180] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1181] This invention is a system for managing the mental health of employees in customer support and customer service, and includes a series of processes for collecting voice data in real time and evaluating their emotions and psychological state. In particular, by combining it with an emotion engine, the system can more precisely recognize the user's emotions and reflect them in the mental health evaluation.

[1182] Specific Embodiments

[1183] Device:

[1184] It captures voice data from intercoms and telephone systems installed in stores and call centers, which involves recording conversations between operators and customers in real time.

[1185] The captured audio data is sent to the server in real time using a protocol for audio streaming (e.g., WebSocket or RTSP).

[1186] server:

[1187] 1. Audio data conversion:

[1188] The server stores the voice data received from the device in temporary storage and converts it into text data by calling a voice recognition engine, such as the Google Cloud Speech-to-Text API.

[1189] 2. Text data and NLP analysis:

[1190] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information from the text.

[1191] 3. Voice timbre feature extraction:

[1192] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[1193] 4. Applying the Emotion Engine:

[1194] The extracted vocal features are input into an emotion engine to recognize the corresponding emotion, such as anger, joy, or sadness.

[1195] 5. Mental status assessment:

[1196] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state. The evaluation includes scoring positive and negative emotions and calculating mental evaluation points.

[1197] Examples:

[1198] For example, if an agent frequently uses keywords that express negative emotions, such as "anxiety" or "anger," while talking to customers, the emotion engine will recognize this and, if the characteristics of the agent's voice (e.g., a sudden rise in pitch or an emphasized tone) indicate the corresponding emotion, the agent's mental evaluation score will be high. This evaluation is displayed as a quantitative score and can be checked by managers.

[1199] 6. Report generation and visualization:

[1200] The system aggregates each employee's mental assessment score and visualizes their mental state over time in graphs and charts. The report includes the mental assessment score, related conversation content, and changes in tone of voice.

[1201] Examples:

[1202] For example, if a graph of Operator A's mental health score for the past month shows a spike in stress on a specific date, the system will display details of the conversations that took place on that date, allowing managers to identify the root cause of the problem and take appropriate measures.

[1203] User (Administrator):

[1204] 1. Report Review and Care:

[1205] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[1206] Consider consultation and care options and implement specific measures, such as stress management training or temporary reassignment.

[1207] Examples:

[1208] A manager reviews Operator B's report and finds that his mental health score for the past week was high. The manager can schedule a meeting with Operator B to identify the cause of his stress and provide mindfulness training to relieve stress, thereby improving his mental health score next time.

[1209] The above is a specific embodiment of the invention that combines an emotion engine. This system allows managers to efficiently and accurately evaluate the mental health of employees and provide appropriate care.

[1210] The processing flow will be explained below.

[1211] Step 1:

[1212] Device:

[1213] Voice data is acquired in real time from intercoms and telephone systems installed in stores and call centers.

[1214] The acquired audio data is sent to the server using an audio streaming protocol (e.g., WebSocket or RTSP).

[1215] Step 2:

[1216] server:

[1217] Stores audio data received from the device in temporary storage, which ensures a consistent data retention for subsequent processing.

[1218] Step 3:

[1219] server:

[1220] Call a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the voice data into text data.

[1221] Step 4:

[1222] server:

[1223] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information.

[1224] Step 5:

[1225] server:

[1226] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[1227] Step 6:

[1228] server:

[1229] The extracted vocal features are input into an emotion engine to recognize the corresponding emotion, such as anger, joy, or sadness.

[1230] Step 7:

[1231] server:

[1232] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state, including scoring positive and negative emotions and calculating mental evaluation points.

[1233] Examples:

[1234] For example, if an agent frequently uses words like "anxiety" or "anger" while talking to customers, the emotion engine will recognize this as a negative emotion. Furthermore, if the tone of voice features (higher pitch or emphasized tone) indicate the corresponding emotion, the agent's mental evaluation score will be higher. This evaluation is displayed as a quantitative score and can be checked by managers.

[1235] Step 8:

[1236] server:

[1237] The mental evaluation points for each employee are aggregated and their changing mental state over time is visualized in graphs and charts.

[1238] Step 9:

[1239] server:

[1240] Automatically generate a mental status report, including mental assessment scores, relevant conversation content, and tone of voice fluctuations.

[1241] Examples:

[1242] For example, if the mental health score of Operator A for the past month is graphed and stress spikes on a particular day, the details of the conversations that took place on that day will be displayed in detail. This display allows managers to identify the root cause of the problem and take appropriate measures.

[1243] Step 10:

[1244] User (Administrator):

[1245] Reports are viewed through a server dashboard, which visualizes each employee's mental assessment score and conversation highlights.

[1246] Step 11:

[1247] User (Administrator):

[1248] Based on the report, meetings with employees can be scheduled to identify the causes of stress, and specific countermeasures such as stress management training or temporary reassignment can be considered.

[1249] Examples:

[1250] The manager checks Operator B's report and finds that his mental health score for the past week is high. The manager schedules a meeting with Operator B to identify the cause of his stress. Based on this, the manager provides mindfulness training to relieve stress and improve his mental health score for the next time.

[1251] The above is the specific processing flow of the invention that combines the emotion engine and the detailed operation of each step. This system allows managers to efficiently and accurately evaluate the mental health of employees and provide appropriate care.

[1252] Example 2

[1253] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1254] To manage employee mental health, there is a need for a method to accurately and in real time assess emotions and psychological states from on-site voice data. However, conventional methods have the problem that they cannot analyze voice data in real time and are not precise enough in recognizing emotions and assessing mental states, making it difficult for managers to take appropriate action quickly.

[1255] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for temporarily saving voice data, means for converting voice data into text data, means for analyzing the text data and extracting emotion and context information, means for extracting tone of voice features from the voice data, means for recognizing emotion based on the tone of voice features, means for evaluating mental states by integrating emotion recognition results and text analysis results, means for visualizing mental states based on evaluation results, and means for generating reports showing mental states. This allows employees' emotions and psychological states to be accurately evaluated in real time, enabling managers to take prompt and appropriate action.

[1256] "Audio data" refers to information that records audio in digital format.

[1257] "Temporarily storing" means holding the received data in a storage area for a short period of time to prevent it from being lost.

[1258] "Converting to text data" refers to the process of converting voice data into character information.

[1259] "Analyzing text data and extracting emotional and contextual information" means using natural language processing technology to extract emotional, keyword, and contextual information from text.

[1260] "Extracting vocal features" means extracting acoustic features such as pitch, tone, and energy from speech data.

[1261] "Emotion recognition" means identifying the speaker's emotion based on the extracted vocal timbre features.

[1262] "Evaluating the mental state" means quantitatively evaluating the speaker's psychological state by combining the emotion recognition results and text analysis results.

[1263] "Visualizing" means converting data into a format that is easy to understand visually, such as a graph or chart.

[1264] "Generating a report" means creating a report based on the evaluation results so that the administrator can check it.

[1265] "Real-time" means that data is processed immediately after it is generated, with little delay.

[1266] This system uses hardware installed in stores or call centers, a server, and specialized software to process a series of data in order to manage the mental health of employees. A specific embodiment of this system is described below.

[1267] Terminal

[1268] The device acquires voice data from intercoms and telephone systems installed in stores and call centers, and transmits the acquired voice data to a server in real time using an audio streaming protocol such as WebSocket or RTSP.

[1269] server

[1270] The server processes the audio data in the following steps:

[1271] 1. Temporary storage of audio data:

[1272] The server temporarily stores the voice data received from the terminal in storage to ensure data security and facilitate subsequent processing.

[1273] 2. Audio to text conversion:

[1274] The server uses the Google Cloud Speech-to-Text API to convert the temporarily stored voice data into text data. For example, a conversation between an operator and a customer, such as "Hello, how can I help you?", is converted directly into text.

[1275] 3. NLP analysis of text data:

[1276] The server passes the converted text data to an NLP engine (e.g., spaCy or BERT) to analyze the sentiment, keywords, and context of the text. For example, from the phrase "I'm feeling very stressed today," "stress" is extracted as an emotional keyword.

[1277] 4. Extracting voice timbre features from speech data:

[1278] The server applies an acoustic feature analysis algorithm to extract vocal timbre features such as pitch, tone, and energy from the audio data. For example, a rise in vocal pitch indicates a change in emotion.

[1279] 5. Emotion recognition based on voice features:

[1280] The server inputs the extracted vocal features into an emotion engine to recognize emotions, such as "tension," "joy," and "anger."

[1281] 6. Mental Status Assessment:

[1282] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to calculate positive and negative emotion scores and mental evaluation points. For example, the result may be "positive score 70%" or "negative score 30%."

[1283] 7. Visualization of mental state and report generation:

[1284] The server aggregates the mental assessment scores for each employee and visualizes fluctuations in their mental state over time in graphs and charts. For example, it can display the weekly fluctuations in mental scores in a line graph, and show the reason for a sudden rise in stress on a particular date along with detailed conversation content.

[1285] User (Administrator)

[1286] Managers access a dashboard provided by the server and check the mental health assessment report for each employee. The report displays graphs, scores, conversation highlights, and more. Based on this, managers can consider interviews and care measures and take specific countermeasures. For example, for employees with high levels of stress, they can suggest "mindfulness training" or "temporary reassignment."

[1287] Examples:

[1288] If an agent frequently uses keywords indicating negative emotions, such as "anxiety" or "anger," while talking to a customer, the emotion engine will identify this. Furthermore, if characteristics indicating anxiety or anger are detected from an analysis of the agent's tone of voice, the agent's mental state score will be higher. This score will be reported to the manager. The report will display a graph showing the "progression of mental state over the past month," visualizing, for example, spikes in stress on specific days.

[1289] Example prompt sentence:

[1290] "To assess the mental health of employees over the past month, please generate a report based on the following information: speech-to-text transcription results, NLP analysis of emotions and keywords, and voice timbre features. Please also include the mental health assessment score for each date and the relevant conversation content."

[1291] This system clarifies the roles of servers, terminals, and users, making it possible to efficiently manage employee mental health in real time.

[1292] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1293] Step 1:

[1294] The terminal acquires voice data from intercoms or telephone systems installed in stores or call centers. At this time, conversations between operators and customers are recorded in real time. The input is voice data from the intercom or telephone system, and the output is an audio file acquired in real time.

[1295] Step 2:

[1296] The audio data captured by the device is sent to the server in real time. This process uses audio streaming protocols such as WebSocket and RTSP. The input is the audio data captured by the device, and the output is the audio streaming data sent to the server.

[1297] Step 3:

[1298] The server temporarily stores the audio data received from the device in storage. This storage process prevents loss of audio data and makes subsequent processing easier. The input is audio data acquired as streaming data, and the output is an audio file stored in temporary storage.

[1299] Step 4:

[1300] The server uses the Google Cloud Speech-to-Text API to convert the temporarily stored voice data into text data. The input is the audio file stored in temporary storage, and the output is the converted text data. This conversion process allows the voice information to be treated as text information.

[1301] Step 5:

[1302] The server passes the text data to a natural language processing (NLP) engine (e.g., spaCy or BERT) to extract sentiment, keywords, and contextual information from the text. The input is the text data, and the output is the sentiment, keywords, and contextual information extracted through analysis. For example, the keyword "stress" is extracted from the text "I'm feeling very stressed today."

[1303] Step 6:

[1304] The server applies an acoustic feature analysis algorithm to extract vocal timbre features such as pitch, tone, and energy from the audio data. The input is a temporarily saved audio file, and the output is the extracted vocal timbre features. For example, a rise in voice pitch indicates a change in emotion.

[1305] Step 7:

[1306] The server inputs the extracted voice timbre features into an emotion engine to recognize the corresponding emotion. The input is the voice timbre features obtained from acoustic feature analysis, and the output is the recognized emotion (e.g., tension, joy, anger). The emotion engine then quantifies the extracted features as an emotion.

[1307] Step 8:

[1308] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to calculate positive and negative emotion scores and mental evaluation points for employees. The input is the text analysis results and emotion recognition results, and the output is the emotion score and mental evaluation points for each employee. For example, evaluations such as "positive score 70%" and "negative score 30%" are possible.

[1309] Step 9:

[1310] The server aggregates the mental assessment points for each employee and visualizes their mental state over time in graphs and charts. The input is each employee's mental assessment point, and the output is their mental state visualized as a graph or chart. For example, a line graph can be used to display the progress of mental state by week, showing the reason for a sudden rise in stress on a particular day.

[1311] Step 10:

[1312] The user (manager) accesses the dashboard provided by the server and checks the mental health assessment report for each employee. The input is the mental health assessment report provided by the server, and the output is report information that the manager can check and use. Based on this, the manager considers interviews and care measures and takes specific countermeasures (e.g., mindfulness training or temporary reassignment).

[1313] The above are the processing steps of the program for this system, and the specific operations performed at each step, as well as the inputs and outputs, have been explained.

[1314] (Application example 2)

[1315] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1316] Efficiently managing employee mental health is important in customer support and customer service. However, conventional systems have difficulty evaluating mental health in real time, making it difficult to immediately grasp changes in employees' stress levels and emotions. In addition, there are limited ways for employees to check their own mental state in real time, making it difficult to implement appropriate stress management.

[1317] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1318] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for visualizing the mental states based on the evaluation results, means for generating a report showing the mental states, and means for displaying the results on a smart device. This allows emotions and psychological states to be evaluated in real time from the voice data, and allows employees to check their mental states themselves via their smart devices.

[1319] "Voice data" means the acoustic signals of a conversation between an operator and a customer.

[1320] "Means for collecting" refers to a device or method for capturing and inputting audio data into the system.

[1321] "Text data" is voice data converted into character information.

[1322] "Means for converting into text data" refers to a device or method that converts voice data into text information using voice recognition technology.

[1323] "Means for analyzing and assessing emotions and psychological states" means a device or method for analyzing and assessing an employee's emotions and psychological states using text data and tone of voice features.

[1324] "Visualization means" refers to a device or method for displaying the evaluation results in a visual form such as a graph or chart.

[1325] A "means for generating a report" is a device or method for creating a detailed written report based on the results of the evaluation.

[1326] "Means for displaying results on a smart device" refers to a device or method for displaying the results of a mental state assessment in real time on a device such as smart glasses or a wearable device.

[1327] The system for implementing this invention mainly performs a series of processes, including collecting voice data, analyzing emotions, and evaluating mental states. The system is realized by linking terminals, a server, and users.

[1328] Terminal

[1329] The terminals are mainly smart devices (e.g., smart glasses). These terminals have built-in microphones that collect voice data from workers. The collected voice data is transmitted to the server in real time. WebSocket or RTSP is used as the protocol for audio streaming.

[1330] server

[1331] The server performs the following steps:

[1332] 1. Audio data conversion:

[1333] The server temporarily stores the voice data acquired from the device and converts it into text data using the Google Cloud Speech-to-Text API.

[1334] 2. Text data analysis:

[1335] The converted text data is analyzed using a natural language processing (NLP) engine, using spaCy and BERT to extract sentiment, keywords, and contextual information from the text.

[1336] 3. Voice timbre feature extraction:

[1337] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[1338] 4. Applying the Emotion Engine:

[1339] The extracted voice features are input into an emotion engine to recognize the corresponding emotion. BERT or other generative AI models may be used as emotion models.

[1340] 5. Mental status assessment:

[1341] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state, and the evaluation results are displayed as a positive or negative emotion score.

[1342] 6. Real-time display:

[1343] The evaluation results are sent in real time to smart devices so that employees can view them.

[1344] User (Administrator)

[1345] Users are responsible for managing the mental health of their employees and providing appropriate care when necessary. Specifically, this includes:

[1346] 1. Report Review and Care:

[1347] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[1348] We will consider options for interviews and care, and provide stress management training and temporary reassignments.

[1349] Specific examples

[1350] For example, if a store employee wearing smart glasses speaks in a "high-pitched, emphasized tone" while interacting with a customer, the voice data is sent to a server, where real-time emotion analysis is performed. As a result, the employee's stress level is assessed as rising, and the smart glasses display shows "high stress." Based on this information, managers can instruct employees to take a break. Managers can also take countermeasures based on this data later.

[1351] Prompt Sentence Examples

[1352] "The employee's voice is high-pitched and stressed. We suggest they take a 10-minute break."

[1353] "Your mental health score for the past week has been high. I suggest you take stress management training."

[1354] Thus, in order to implement this invention, it is necessary to build a system in which the components of the server, terminal, and user work together to evaluate and manage the mental state of employees in real time.

[1355] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1356] Step 1:

[1357] Audio data collection

[1358] The device collects employee voice data.

[1359] Input: Audio data captured from the built-in microphone of the smart glasses.

[1360] How it works: The built-in microphone records the conversation between the operator and the customer and sends the audio data to the server in real time.

[1361] Step 2:

[1362] Converting audio data to text data

[1363] The server converts the received voice data into text data.

[1364] Input: Audio data sent from the device.

[1365] Output: Text data.

[1366] What it does: It uses the Google Cloud Speech-to-Text API to convert audio data into text. A speech recognition model analyzes the acoustic signal and generates a corresponding string of characters.

[1367] Step 3:

[1368] Text data analysis

[1369] The server analyzes the converted text data and evaluates the emotions and psychological state.

[1370] Input: Text data.

[1371] Output: Emotion score and mental state assessment results.

[1372] Specific operation: Text data is input into a natural language processing (NLP) engine (e.g., spaCy, BERT) to analyze and extract sentiment, keywords, and contextual information.

[1373] Step 4:

[1374] Voice timbre feature extraction

[1375] The server extracts voice characteristics from the voice data.

[1376] Input: Audio data.

[1377] Output: Voice characteristics (pitch, tone, energy, etc.).

[1378] Specific operation: Apply an acoustic feature analysis algorithm (e.g., librosa) to extract features such as pitch, tone, and energy from the audio signal.

[1379] Step 5:

[1380] Applying the Emotion Engine

[1381] The server inputs the voice features into the emotion engine and recognizes the corresponding emotion.

[1382] Input: Voice timbre features.

[1383] Output: Emotion recognition results.

[1384] Specific operation: Using a generative AI model (e.g., BERT), emotions are classified from tone of voice features, and emotions such as anger, joy, and sadness are recognized.

[1385] Step 6:

[1386] Mental state assessment

[1387] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to evaluate the mental state of employees.

[1388] Input: Emotion score, voice features, and text analysis results.

[1389] Output: Mental state assessment score.

[1390] Specific operations: Evaluate overall mental state by assigning positive and negative emotional scores and calculating mental evaluation points.

[1391] Step 7:

[1392] Real-time display

[1393] The server transmits the evaluation results to the smart device in real time and displays them.

[1394] Input: Mental State Assessment score.

[1395] Output: Evaluation results displayed on smart devices.

[1396] Specific operation: The evaluation results are displayed on the smart glasses display, allowing employees to check their own mental state in real time.

[1397] Step 8:

[1398] Report Generation

[1399] The server generates a report based on the mental state evaluation results and provides it to the user.

[1400] Input: Accumulated data of evaluation results.

[1401] Output: Detailed mental state report.

[1402] Specific actions: Generate graphs and charts to create reports that visualize fluctuations in mental state for each employee, and display them on a dashboard for managers to review.

[1403] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1404] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1405] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1406] [Fourth embodiment]

[1407] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1408] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1409] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1410] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1411] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1412] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1413] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1414] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1415] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1416] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1417] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1418] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1419] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1420] This invention is a system for managing the mental health of employees in customer support and customer service work, and includes a series of processes for collecting, analyzing, and evaluating voice data. Specific embodiments for carrying out this invention are described below.

[1421] 1. Data Collection

[1422] Device:

[1423] It captures voice data from intercoms and telephone systems installed in stores and call centers, which involves recording conversations between operators and customers in real time.

[1424] The captured audio data is sent to the server in real time using a protocol for audio streaming (e.g., WebSocket or RTSP).

[1425] 2. Audio data analysis

[1426] server:

[1427] The server sends the voice data received from the device to a voice recognition engine, such as the Google Cloud Speech-to-Text API, and converts it into text data.

[1428] The converted text data is passed to a natural language processing (NLP) engine for text analysis, which analyzes the sentiment of the conversation, extracts keywords, and understands the context.

[1429] 3. Mental state assessment

[1430] server:

[1431] The analysis results from the NLP engine are used to assess the employee's emotional and psychological state, taking into account positive and negative emotional scoring as well as vocal characteristics (pitch, tone, energy, etc.).

[1432] A machine learning model is used to calculate employees' mental assessment points.

[1433] Examples:

[1434] For example, if negative keywords such as "anger" or "anxiety" frequently appear in the analysis results for a certain operator, the operator's mental evaluation score will be high. This score is also evaluated based on changes in tone of voice during conversations with customers, allowing for an accurate understanding of the operator's mental state.

[1435] 4. Report generation and visualization

[1436] server:

[1437] The mental evaluation points for each employee are aggregated and changes over time are visualized in graphs and charts.

[1438] The report includes mental assessment scores and related conversation highlights, allowing managers to understand specifically what situations increase stress levels.

[1439] Examples:

[1440] For example, if a graph of Operator A's mental health score for the past month shows a spike in stress on a specific date, the system will display details of the conversations that took place on that date, allowing managers to identify the root cause of the problem and take appropriate measures.

[1441] 5. Review and care of reports

[1442] User (Administrator):

[1443] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[1444] Consider consultation and care options and implement specific measures, such as stress management training or temporary reassignment.

[1445] Examples:

[1446] A manager checks the report of Operator B and finds that the mental health score for the past week was high. The manager schedules a meeting with Operator B, identifies the cause of stress, and provides mindfulness training to relieve stress, thereby improving the next mental health score.

[1447] The above is a specific embodiment for carrying out the present invention. By utilizing this system, it is possible to efficiently manage the mental health of employees and improve the quality of work.

[1448] The processing flow will be explained below.

[1449] Step 1:

[1450] Device:

[1451] Voice data is acquired in real time from intercoms and telephone systems installed in stores and call centers.

[1452] The acquired audio data is sent to the server using a protocol for audio streaming (e.g., WebSocket or RTSP).

[1453] Step 2:

[1454] server:

[1455] Stores audio data received from the device in temporary storage, which ensures a consistent data retention for subsequent processing.

[1456] Step 3:

[1457] server:

[1458] Call a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the voice data into text data.

[1459] Step 4:

[1460] server:

[1461] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information from the text.

[1462] Step 5:

[1463] server:

[1464] Extract voice characteristics (pitch, tone, energy, etc.) from the audio data by applying an acoustic feature analysis algorithm.

[1465] Step 6:

[1466] server:

[1467] Based on the analysis results from the NLP engine and the features of the voice tone, the emotional and psychological state is evaluated and a mental evaluation score is calculated. Here, a machine learning model is used to quantify the mental evaluation score.

[1468] Step 7:

[1469] server:

[1470] The mental evaluation points for each employee are aggregated and their changing mental state over time is visualized in graphs and charts.

[1471] Step 8:

[1472] server:

[1473] Automatically generate a mental status report, including mental assessment scores, relevant conversation content, and tone of voice fluctuations.

[1474] Step 9:

[1475] User (Administrator):

[1476] Reports are viewed through the server's dashboard, which visualizes each employee's mental assessment score and conversation highlights.

[1477] Step 10:

[1478] User (Administrator):

[1479] Schedule a meeting with the employee based on the report. Identify the causes of stress and take measures to address them, such as training programs or reassignment.

[1480] This is the specific process flow. This system allows managers to efficiently and accurately evaluate the mental health of their employees and provide appropriate care.

[1481] Example 1

[1482] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1483] In conventional customer support operations, the mental health of employees was not adequately managed, resulting in a decline in the quality of work, and problems such as employee stress and increased turnover. In addition, there was also the issue of the time required to collect, analyze, and evaluate voice data, making it difficult to respond in real time.

[1484] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1485] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for transmitting the collected voice data to the server in real time, means for sending the converted text data to a natural language processing engine for emotion analysis, means for accumulating mental evaluation points over time and displaying them in graphs or charts, and means for providing a dashboard for managers to check mental evaluation reports. This enables real-time collection and immediate analysis of voice data, making it possible to efficiently manage employee mental health and improve the quality of work.

[1486] "Voice data" means digital voice signals that record the conversation between an operator and a customer.

[1487] "Text data" is voice data converted into character information.

[1488] "Means for assessing emotions and psychological states" refers to technological means for analyzing text data and quantifying and assessing employees' emotions (positive and negative) and psychological states.

[1489] "Visualization means" refers to a means for displaying the evaluation results in graphs or charts so that users can understand them visually.

[1490] The "means for generating a report" refers to a means for creating a detailed report on the mental state of an employee based on the evaluation results.

[1491] "Means of transmitting in real time" refers to the technical means of instantly transferring collected voice data to a server.

[1492] A "natural language processing engine" refers to a processing engine that performs sentiment analysis, keyword extraction, context understanding, etc. on text data.

[1493] "Means for accumulating and displaying mental evaluation points over time" refers to a means for aggregating employees' mental evaluation points over a fixed period of time and visually displaying the fluctuations in graphs or charts.

[1494] "Dashboard" refers to a web-based or software-based platform that provides an interface through which managers can view employee mental health assessment reports in real time.

[1495] This invention is a system for managing the mental health of employees in customer support and customer service. The system includes a series of processes for collecting, analyzing, and evaluating voice data. Specific embodiments for implementing this invention are described below.

[1496] Data collection

[1497] Device:

[1498] Intercoms and telephone systems installed in stores and call centers collect voice data, including the ability to record conversations between operators and customers in real time. The collected voice data is then sent to a server in real time. Protocols such as WebSocket and RTSP are used to stream the data.

[1499] Sending audio data

[1500] Device:

[1501] The acquired voice data is sent to the server in real time. This process is carried out using protocols such as WebSocket and RTSP. For example, voice data collected during conversations in a store or during phone calls is immediately transferred to the server.

[1502] Converting audio data to text

[1503] server:

[1504] The server sends the received voice data to a speech recognition engine (e.g., Google Cloud Speech-to-Text API) and converts it into text data. The speech recognition engine converts voice into text with high accuracy, and text data is generated in near real time.

[1505] Text data analysis

[1506] server:

[1507] The converted text data is passed to a natural language processing (NLP) engine (e.g., Google Cloud Natural Language API), which performs sentiment analysis, keyword extraction, and contextual understanding of the text to analyze the content of employee conversations.

[1508] Mental state assessment

[1509] server:

[1510] The server evaluates the employee's emotions and psychological state based on the analysis results obtained from the NLP engine. This evaluation takes into account positive and negative emotion scores and vocal characteristics (pitch, tone, energy, etc.). Furthermore, a machine learning model is used to calculate the employee's mental evaluation score.

[1511] Examples:

[1512] For example, if negative keywords such as "anger" or "anxiety" frequently appear during a conversation with an agent, the agent's mental evaluation score will be high. This score also reflects changes in tone of voice during conversations with customers, allowing for an accurate understanding of the agent's mental state.

[1513] Collection and visualization of mental assessment data

[1514] server:

[1515] The server aggregates the mental evaluation points for each employee and visualizes them in graphs and charts, allowing managers to track fluctuations in employees' mental states over time.

[1516] Examples:

[1517] For example, the mental health score of Operator A over the past month can be graphed, allowing fluctuations in the evaluation to be observed at a glance. If the graph shows that stress spiked on a particular day, the content of the conversations on that day can be checked in detail.

[1518] Administrator report confirmation

[1519] User (Administrator):

[1520] Managers can access a dashboard provided by the server and view mental health assessment reports for each employee, which display graphs, scores, and conversation highlights, allowing managers to identify specific stressors.

[1521] Examples:

[1522] For example, a manager can check Operator B's high mental health score on the dashboard, investigate the cause, and take appropriate measures.

[1523] Stress management

[1524] User (Administrator):

[1525] Managers can use the reports to schedule interviews and stress management training sessions with employees, and, if necessary, take concrete action, such as temporarily reassigning employees.

[1526] Examples:

[1527] For example, a manager might schedule a meeting with Operator B to identify the cause of stress and provide mindfulness training to improve the next mental score.

[1528] Prompt Sentence Examples

[1529] The generative AI model generates a natural language explanation for a mental health analysis by inputting the following prompt:

[1530] Please explain your mental health assessment system for customer support employees. The system involves a series of processes to collect, analyze, and evaluate voice data. Please explain what hardware or software is used, and what data processing or calculation is performed, including specific scenarios and examples of use.

[1531] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1532] Step 1:

[1533] Data collection

[1534] The device collects voice data from intercoms and telephone systems, including the ability to record conversations between operators and customers in real time. The device records and stores data of interactions with clients via intercoms and telephone systems.

[1535] Input: Voice of conversation between operator and customer

[1536] Output: Real-time audio data

[1537] Step 2:

[1538] Sending audio data

[1539] The device transmits the acquired audio data to the server in real time using protocols such as WebSocket and RTSP. For example, when a customer complains that their product has not arrived, the audio is immediately transmitted to the server.

[1540] Input: Real-time audio data

[1541] Output: Audio data sent to the server

[1542] Step 3:

[1543] Converting audio data to text

[1544] The server sends the received voice data to a voice recognition engine and converts it into text data. For example, the voice recognition engine uses the Google Cloud Speech-to-Text API. The server sends the voice data to the API and receives text data as character information.

[1545] Input: Audio data transferred to the server

[1546] Output: Text data generated by the speech recognition engine

[1547] Step 4:

[1548] Text data analysis

[1549] The server then sends the converted text data to a natural language processing engine to perform sentiment analysis, keyword extraction, and context understanding. For example, it detects whether the text contains keywords such as "anger" or "frustration." The server inputs the text data into the analysis engine and obtains an emotion score and a keyword list as output.

[1550] Input: Text data generated by a speech recognition engine

[1551] Output: Sentiment score and keyword list analyzed by a natural language processing engine

[1552] Step 5:

[1553] Mental state assessment

[1554] The server evaluates the employee's emotions and psychological state based on the analysis results obtained from the NLP engine. This evaluation also includes vocal characteristics (pitch, tone, energy, etc.). The server inputs the analysis results and calculates the employee's mental evaluation score. For example, if the emotion "anger" appears frequently, the mental evaluation score will be high.

[1555] Input: Sentiment scores and keyword list analyzed by a natural language processing engine

[1556] Output: Mental evaluation points calculated by the server

[1557] Step 6:

[1558] Collection and visualization of mental assessment data

[1559] The server aggregates the mental evaluation points for each employee and visualizes them in graphs and charts. The evaluation results are displayed in daily, weekly, monthly, etc. For example, Operator A's mental score for the past month is displayed in a graph.

[1560] Input: Mental evaluation points for each employee

[1561] Output: Visualized graphs and charts

[1562] Step 7:

[1563] Administrator report confirmation

[1564] The user (administrator) accesses the dashboard provided by the server and checks the mental assessment report for each employee. The report includes analysis results, graphs, scores, and conversation highlights. The administrator can obtain detailed reports from the dashboard and check stress factors.

[1565] Input: Visualized graphs, charts, detailed reports

[1566] Output: Mental evaluation report reviewed by the administrator

[1567] Step 8:

[1568] Stress management

[1569] Based on the report, the user (manager) schedules interviews with employees and stress management training. If necessary, the user can take specific countermeasures, such as reassignment. For example, if Operator B shows signs of high stress, the user can schedule an interview and take specific countermeasures.

[1570] Input: Mental evaluation report reviewed by administrator

[1571] Output: Plans for scheduled interviews and stress management training

[1572] (Application example 1)

[1573] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1574] Employee mental health has a significant impact on the quality and efficiency of work. However, in physical stores, it is difficult to grasp the mental state of employees in real time and provide appropriate care. There are limited means to accurately assess the level of stress employees are experiencing while interacting with customers and to take immediate countermeasures. This can lead to an accumulation of employee stress, which can lead to reduced work efficiency and mental health problems.

[1575] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1576] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for visualizing the collected mental evaluation points over time, and means for collecting voice data from a microphone built into the smart glasses and transmitting it to the server in real time, thereby enabling employees to evaluate the mental states of customers in real time while interacting with them, and allowing managers to provide immediate care.

[1577] "Voice data" refers to audio information of conversations between employees and customers.

[1578] "Text data" is voice data converted into character information that is easy to analyze.

[1579] The "means for assessing emotions and psychological states" refers to a means for analyzing the emotions and psychological states of employees based on text data and obtaining the assessment results.

[1580] A "means for visualizing mental state" is a means for understanding mental state by visually displaying the evaluation results.

[1581] The "means for generating a report" refers to a means for creating a report summarizing the mental state of employees based on the evaluation results.

[1582] A "means for extracting vocal timbre features" is a means for extracting vocal features such as pitch, tone, and energy from audio data.

[1583] "Real-time means" means means for collecting, converting, and analyzing audio data in real time.

[1584] "Smart glasses" are glasses-type devices for collecting and transmitting audio data in real time.

[1585] A "server" is a computer system for analyzing collected voice data and processing evaluation results.

[1586] "Mental evaluation points" are indicators that quantitatively represent an employee's mental state and are calculated based on the analysis results.

[1587] The present invention relates to a system that utilizes smart glasses to efficiently monitor the mental health of employees in brick-and-mortar stores while they are serving customers. Specific embodiments are described below.

[1588] System configuration

[1589] The system includes a set of means for collecting, analyzing, and evaluating audio data.

[1590] 1. Audio data collection method:

[1591] The system uses microphones built into smart glasses worn by store employees to collect conversations between employees and customers in real time.

[1592] The collected audio data is transmitted to a server using WebSocket or a similar real-time transmission protocol.

[1593] 2. Means of converting audio data to text data:

[1594] The server converts the received voice data into text data using a voice recognition engine (e.g., Google Cloud Speech-to-Text API).

[1595] This conversion process converts the audio data into a format that is easier to analyze.

[1596] 3. Instruments for assessing emotional and psychological states:

[1597] The server passes the converted text data to a natural language processing (NLP) engine to perform sentiment analysis (e.g., Google Cloud Natural Language API).

[1598] The NLP engine extracts emotional scores such as positive or negative and specific keywords from text data to evaluate the mental state of employees.

[1599] 4. Ways to visualize your mental state:

[1600] The server calculates and visualizes mental evaluation points based on the evaluation results obtained from the NLP engine.

[1601] This visualization includes a means for displaying the time-varying change in the evaluation points as a graph or chart.

[1602] 5. To generate a report:

[1603] The server generates reports showing the mental state of each employee and displays them on a dashboard accessible to managers.

[1604] The report includes detailed rating points, graphs of changes over time, and highlights of key conversations.

[1605] Specific examples of implementation

[1606] For example, while Employee A is wearing smart glasses and interacting with customers, their conversation is collected as voice data. The collected voice data is sent to a server in real time and converted into text data by a voice recognition engine. The text data is analyzed by an NLP engine to evaluate Employee A's emotion score and psychological state. The evaluation results are visualized on a dashboard for managers, allowing them to monitor Employee A's mental health and provide care as needed.

[1607] Prompt Sentence Examples

[1608] I would like to build a system that collects customer voice data, converts it to text in real time, and performs sentiment analysis. Specifically, please generate sample code for a system that uses a speech recognition engine to convert speech to text, and a natural language processing engine to perform sentiment analysis.

[1609] The above is a specific embodiment for carrying out the present invention. By utilizing this system, it is possible to monitor the mental health of employees in real time and provide appropriate care.

[1610] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1611] Step 1:

[1612] A microphone built into the smart glasses collects conversations between employees and customers in real time.

[1613] Input: Employee-customer conversation.

[1614] Output: Audio data.

[1615] To send the audio data to the server, the audio data is transferred in real time using the WebSocket protocol, etc. Specifically, the audio data obtained from the smart glasses is received via a WebSocket connection.

[1616] Step 2:

[1617] The server passes the received voice data to a voice recognition engine, which converts the data into text data.

[1618] Input: Audio data.

[1619] Output: Text data.

[1620] Specifically, the voice data is converted into a string of characters (text data) using the Google Cloud Speech-to-Text API, etc. The converted text data is saved in a format suitable for subsequent analysis.

[1621] Step 3:

[1622] The server passes the text data to a natural language processing (NLP) engine for sentiment analysis.

[1623] Input: Text data.

[1624] Output: Sentiment analysis results (sentiment score, keywords, etc.).

[1625] Specifically, it uses an NLP engine such as Google Cloud Natural Language API to extract sentiment scores and important keywords from text data to assess employee emotions and psychological states.

[1626] Step 4:

[1627] The server calculates the employee's mental evaluation points based on the emotion analysis results.

[1628] Input: Sentiment analysis results.

[1629] Output: Mental evaluation points.

[1630] Specifically, the mental state of employees is quantified based on the emotion score and the frequency of keywords, and a mental evaluation point is calculated, allowing for a quantitative evaluation of the mental state of employees.

[1631] Step 5:

[1632] The server generates graphs and charts to visualize the mental assessment points and displays them on an administrator dashboard.

[1633] Input: Mental Assessment Points.

[1634] Output: Visualized data (graphs, charts).

[1635] Specifically, a web-based dashboard is used to display the collected mental health assessment scores over time in graphs and charts, allowing managers to provide appropriate care based on this information.

[1636] Step 6:

[1637] The server generates reports showing each employee's mental state, which managers can check regularly.

[1638] Input: Mental evaluation points, conversation highlights.

[1639] Output: Report.

[1640] Specifically, the system generates regular reports containing detailed mental health assessment scores, graphs of changes over time, and highlights of important conversations, and makes them accessible to managers, allowing them to understand trends in employee mental health and take necessary measures.

[1641] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1642] This invention is a system for managing the mental health of employees in customer support and customer service, and includes a series of processes for collecting voice data in real time and evaluating their emotions and psychological state. In particular, by combining it with an emotion engine, the system can more precisely recognize the user's emotions and reflect them in the mental health evaluation.

[1643] Specific Embodiments

[1644] Device:

[1645] It captures voice data from intercoms and telephone systems installed in stores and call centers, which involves recording conversations between operators and customers in real time.

[1646] The captured audio data is sent to the server in real time using a protocol for audio streaming (e.g., WebSocket or RTSP).

[1647] server:

[1648] 1. Audio data conversion:

[1649] The server stores the voice data received from the device in temporary storage and converts it into text data by calling a voice recognition engine, such as the Google Cloud Speech-to-Text API.

[1650] 2. Text data and NLP analysis:

[1651] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information from the text.

[1652] 3. Voice timbre feature extraction:

[1653] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[1654] 4. Applying the Emotion Engine:

[1655] The extracted vocal features are input into an emotion engine to recognize the corresponding emotion, such as anger, joy, or sadness.

[1656] 5. Mental status assessment:

[1657] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state. The evaluation includes scoring positive and negative emotions and calculating mental evaluation points.

[1658] Examples:

[1659] For example, if an agent frequently uses keywords that express negative emotions, such as "anxiety" or "anger," while talking to customers, the emotion engine will recognize this and, if the characteristics of the agent's voice (e.g., a sudden rise in pitch or an emphasized tone) indicate the corresponding emotion, the agent's mental evaluation score will be high. This evaluation is displayed as a quantitative score and can be checked by managers.

[1660] 6. Report generation and visualization:

[1661] The system aggregates each employee's mental assessment score and visualizes their mental state over time in graphs and charts. The report includes the mental assessment score, related conversation content, and changes in tone of voice.

[1662] Examples:

[1663] For example, if a graph of Operator A's mental health score for the past month shows a spike in stress on a specific date, the system will display details of the conversations that took place on that date, allowing managers to identify the root cause of the problem and take appropriate measures.

[1664] User (Administrator):

[1665] 1. Report Review and Care:

[1666] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[1667] Consider consultation and care options and implement specific measures, such as stress management training or temporary reassignment.

[1668] Examples:

[1669] A manager reviews Operator B's report and finds that his mental health score for the past week was high. The manager can schedule a meeting with Operator B to identify the cause of his stress and provide mindfulness training to relieve stress, thereby improving his mental health score next time.

[1670] The above is a specific embodiment of the invention that combines an emotion engine. This system allows managers to efficiently and accurately evaluate the mental health of employees and provide appropriate care.

[1671] The processing flow will be explained below.

[1672] Step 1:

[1673] Device:

[1674] Voice data is acquired in real time from intercoms and telephone systems installed in stores and call centers.

[1675] The acquired audio data is sent to the server using an audio streaming protocol (e.g., WebSocket or RTSP).

[1676] Step 2:

[1677] server:

[1678] Stores audio data received from the device in temporary storage, which ensures a consistent data retention for subsequent processing.

[1679] Step 3:

[1680] server:

[1681] Call a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the voice data into text data.

[1682] Step 4:

[1683] server:

[1684] The converted text data is passed to a natural language processing (NLP) engine (e.g., spaCy or BERT) to analyze the text, extracting sentiment, keywords, and contextual information.

[1685] Step 5:

[1686] server:

[1687] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[1688] Step 6:

[1689] server:

[1690] The extracted vocal features are input into an emotion engine to recognize the corresponding emotion, such as anger, joy, or sadness.

[1691] Step 7:

[1692] server:

[1693] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state, including scoring positive and negative emotions and calculating mental evaluation points.

[1694] Examples:

[1695] For example, if an agent frequently uses words like "anxiety" or "anger" while talking to customers, the emotion engine will recognize this as a negative emotion. Furthermore, if the tone of voice features (higher pitch or emphasized tone) indicate the corresponding emotion, the agent's mental evaluation score will be higher. This evaluation is displayed as a quantitative score and can be checked by managers.

[1696] Step 8:

[1697] server:

[1698] The mental evaluation points for each employee are aggregated and their changing mental state over time is visualized in graphs and charts.

[1699] Step 9:

[1700] server:

[1701] Automatically generate a mental status report, including mental assessment scores, relevant conversation content, and tone of voice fluctuations.

[1702] Examples:

[1703] For example, if the mental health score of Operator A for the past month is graphed and stress spikes on a particular day, the details of the conversations that took place on that day will be displayed in detail. This display allows managers to identify the root cause of the problem and take appropriate measures.

[1704] Step 10:

[1705] User (Administrator):

[1706] Reports are viewed through a server dashboard, which visualizes each employee's mental assessment score and conversation highlights.

[1707] Step 11:

[1708] User (Administrator):

[1709] Based on the report, meetings with employees can be scheduled to identify the causes of stress, and specific countermeasures such as stress management training or temporary reassignment can be considered.

[1710] Examples:

[1711] The manager checks Operator B's report and finds that his mental health score for the past week is high. The manager schedules a meeting with Operator B to identify the cause of his stress. Based on this, the manager provides mindfulness training to relieve stress and improve his mental health score for the next time.

[1712] The above is the specific processing flow of the invention that combines the emotion engine and the detailed operation of each step. This system allows managers to efficiently and accurately evaluate the mental health of employees and provide appropriate care.

[1713] Example 2

[1714] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1715] To manage employee mental health, there is a need for a method to accurately and in real time assess emotions and psychological states from on-site voice data. However, conventional methods have the problem that they cannot analyze voice data in real time and are not precise enough in recognizing emotions and assessing mental states, making it difficult for managers to take appropriate action quickly.

[1716] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for temporarily saving voice data, means for converting voice data into text data, means for analyzing the text data and extracting emotion and context information, means for extracting tone of voice features from the voice data, means for recognizing emotion based on the tone of voice features, means for evaluating mental states by integrating emotion recognition results and text analysis results, means for visualizing mental states based on evaluation results, and means for generating reports showing mental states. This allows employees' emotions and psychological states to be accurately evaluated in real time, enabling managers to take prompt and appropriate action.

[1717] "Audio data" refers to information that records audio in digital format.

[1718] "Temporarily storing" means holding the received data in a storage area for a short period of time to prevent it from being lost.

[1719] "Converting to text data" refers to the process of converting voice data into character information.

[1720] "Analyzing text data and extracting emotional and contextual information" means using natural language processing technology to extract emotional, keyword, and contextual information from text.

[1721] "Extracting vocal features" means extracting acoustic features such as pitch, tone, and energy from speech data.

[1722] "Emotion recognition" means identifying the speaker's emotion based on the extracted vocal timbre features.

[1723] "Evaluating the mental state" means quantitatively evaluating the speaker's psychological state by combining the emotion recognition results and text analysis results.

[1724] "Visualizing" means converting data into a format that is easy to understand visually, such as a graph or chart.

[1725] "Generating a report" means creating a report based on the evaluation results so that the administrator can check it.

[1726] "Real-time" means that data is processed immediately after it is generated, with little delay.

[1727] This system uses hardware installed in stores or call centers, a server, and specialized software to process a series of data in order to manage the mental health of employees. A specific embodiment of this system is described below.

[1728] Terminal

[1729] The device acquires voice data from intercoms and telephone systems installed in stores and call centers, and transmits the acquired voice data to a server in real time using an audio streaming protocol such as WebSocket or RTSP.

[1730] server

[1731] The server processes the audio data in the following steps:

[1732] 1. Temporary storage of audio data:

[1733] The server temporarily stores the voice data received from the terminal in storage to ensure data security and facilitate subsequent processing.

[1734] 2. Audio to text conversion:

[1735] The server uses the Google Cloud Speech-to-Text API to convert the temporarily stored voice data into text data. For example, a conversation between an operator and a customer, such as "Hello, how can I help you?", is converted directly into text.

[1736] 3. NLP analysis of text data:

[1737] The server passes the converted text data to an NLP engine (e.g., spaCy or BERT) to analyze the sentiment, keywords, and context of the text. For example, from the phrase "I'm feeling very stressed today," "stress" is extracted as an emotional keyword.

[1738] 4. Extracting voice timbre features from speech data:

[1739] The server applies an acoustic feature analysis algorithm to extract vocal timbre features such as pitch, tone, and energy from the audio data. For example, a rise in vocal pitch indicates a change in emotion.

[1740] 5. Emotion recognition based on voice features:

[1741] The server inputs the extracted vocal features into an emotion engine to recognize emotions, such as "tension," "joy," and "anger."

[1742] 6. Mental Status Assessment:

[1743] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to calculate positive and negative emotion scores and mental evaluation points. For example, the result may be "positive score 70%" or "negative score 30%."

[1744] 7. Visualization of mental state and report generation:

[1745] The server aggregates the mental assessment scores for each employee and visualizes fluctuations in their mental state over time in graphs and charts. For example, it can display the weekly fluctuations in mental scores in a line graph, and show the reason for a sudden rise in stress on a particular date along with detailed conversation content.

[1746] User (Administrator)

[1747] Managers access a dashboard provided by the server and check the mental health assessment report for each employee. The report displays graphs, scores, conversation highlights, and more. Based on this, managers can consider interviews and care measures and take specific countermeasures. For example, for employees with high levels of stress, they can suggest "mindfulness training" or "temporary reassignment."

[1748] Examples:

[1749] If an agent frequently uses keywords indicating negative emotions, such as "anxiety" or "anger," while talking to a customer, the emotion engine will identify this. Furthermore, if characteristics indicating anxiety or anger are detected from an analysis of the agent's tone of voice, the agent's mental state score will be higher. This score will be reported to the manager. The report will display a graph showing the "progression of mental state over the past month," visualizing, for example, spikes in stress on specific days.

[1750] Example prompt sentence:

[1751] "To assess the mental health of employees over the past month, please generate a report based on the following information: speech-to-text transcription results, NLP analysis of emotions and keywords, and voice timbre features. Please also include the mental health assessment score for each date and the relevant conversation content."

[1752] This system clarifies the roles of servers, terminals, and users, making it possible to efficiently manage employee mental health in real time.

[1753] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1754] Step 1:

[1755] The terminal acquires voice data from intercoms or telephone systems installed in stores or call centers. At this time, conversations between operators and customers are recorded in real time. The input is voice data from the intercom or telephone system, and the output is an audio file acquired in real time.

[1756] Step 2:

[1757] The audio data captured by the device is sent to the server in real time. This process uses audio streaming protocols such as WebSocket and RTSP. The input is the audio data captured by the device, and the output is the audio streaming data sent to the server.

[1758] Step 3:

[1759] The server temporarily stores the audio data received from the device in storage. This storage process prevents loss of audio data and makes subsequent processing easier. The input is audio data acquired as streaming data, and the output is an audio file stored in temporary storage.

[1760] Step 4:

[1761] The server uses the Google Cloud Speech-to-Text API to convert the temporarily stored voice data into text data. The input is the audio file stored in temporary storage, and the output is the converted text data. This conversion process allows the voice information to be treated as text information.

[1762] Step 5:

[1763] The server passes the text data to a natural language processing (NLP) engine (e.g., spaCy or BERT) to extract sentiment, keywords, and contextual information from the text. The input is the text data, and the output is the sentiment, keywords, and contextual information extracted through analysis. For example, the keyword "stress" is extracted from the text "I'm feeling very stressed today."

[1764] Step 6:

[1765] The server applies an acoustic feature analysis algorithm to extract vocal timbre features such as pitch, tone, and energy from the audio data. The input is a temporarily saved audio file, and the output is the extracted vocal timbre features. For example, a rise in voice pitch indicates a change in emotion.

[1766] Step 7:

[1767] The server inputs the extracted voice timbre features into an emotion engine to recognize the corresponding emotion. The input is the voice timbre features obtained from acoustic feature analysis, and the output is the recognized emotion (e.g., tension, joy, anger). The emotion engine then quantifies the extracted features as an emotion.

[1768] Step 8:

[1769] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to calculate positive and negative emotion scores and mental evaluation points for employees. The input is the text analysis results and emotion recognition results, and the output is the emotion score and mental evaluation points for each employee. For example, evaluations such as "positive score 70%" and "negative score 30%" are possible.

[1770] Step 9:

[1771] The server aggregates the mental assessment points for each employee and visualizes their mental state over time in graphs and charts. The input is each employee's mental assessment point, and the output is their mental state visualized as a graph or chart. For example, a line graph can be used to display the progress of mental state by week, showing the reason for a sudden rise in stress on a particular day.

[1772] Step 10:

[1773] The user (manager) accesses the dashboard provided by the server and checks the mental health assessment report for each employee. The input is the mental health assessment report provided by the server, and the output is report information that the manager can check and use. Based on this, the manager considers interviews and care measures and takes specific countermeasures (e.g., mindfulness training or temporary reassignment).

[1774] The above are the processing steps of the program for this system, and the specific operations performed at each step, as well as the inputs and outputs, have been explained.

[1775] (Application example 2)

[1776] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1777] Efficiently managing employee mental health is important in customer support and customer service. However, conventional systems have difficulty evaluating mental health in real time, making it difficult to immediately grasp changes in employees' stress levels and emotions. In addition, there are limited ways for employees to check their own mental state in real time, making it difficult to implement appropriate stress management.

[1778] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1779] In this invention, the server includes means for collecting voice data, means for converting the voice data into text data, means for analyzing the text data and evaluating emotions and psychological states, means for visualizing the mental states based on the evaluation results, means for generating a report showing the mental states, and means for displaying the results on a smart device. This allows emotions and psychological states to be evaluated in real time from the voice data, and allows employees to check their mental states themselves via their smart devices.

[1780] "Voice data" means the acoustic signals of a conversation between an operator and a customer.

[1781] "Means for collecting" refers to a device or method for capturing and inputting audio data into the system.

[1782] "Text data" is voice data converted into character information.

[1783] "Means for converting into text data" refers to a device or method that converts voice data into text information using voice recognition technology.

[1784] "Means for analyzing and assessing emotions and psychological states" means a device or method for analyzing and assessing an employee's emotions and psychological states using text data and tone of voice features.

[1785] "Visualization means" refers to a device or method for displaying the evaluation results in a visual form such as a graph or chart.

[1786] A "means for generating a report" is a device or method for creating a detailed written report based on the results of the evaluation.

[1787] "Means for displaying results on a smart device" refers to a device or method for displaying the results of a mental state assessment in real time on a device such as smart glasses or a wearable device.

[1788] The system for implementing this invention mainly performs a series of processes, including collecting voice data, analyzing emotions, and evaluating mental states. The system is realized by linking terminals, a server, and users.

[1789] Terminal

[1790] The terminals are mainly smart devices (e.g., smart glasses). These terminals have built-in microphones that collect voice data from workers. The collected voice data is transmitted to the server in real time. WebSocket or RTSP is used as the protocol for audio streaming.

[1791] server

[1792] The server performs the following steps:

[1793] 1. Audio data conversion:

[1794] The server temporarily stores the voice data acquired from the device and converts it into text data using the Google Cloud Speech-to-Text API.

[1795] 2. Text data analysis:

[1796] The converted text data is analyzed using a natural language processing (NLP) engine, using spaCy and BERT to extract sentiment, keywords, and contextual information from the text.

[1797] 3. Voice timbre feature extraction:

[1798] An acoustic feature analysis algorithm is applied to extract voice characteristics (pitch, tone, energy, etc.) from the audio data.

[1799] 4. Applying the Emotion Engine:

[1800] The extracted voice features are input into an emotion engine to recognize the corresponding emotion. BERT or other generative AI models may be used as emotion models.

[1801] 5. Mental status assessment:

[1802] The analysis results of the NLP engine and the recognition results of the emotion engine are combined to evaluate the employee's emotional and psychological state, and the evaluation results are displayed as a positive or negative emotion score.

[1803] 6. Real-time display:

[1804] The evaluation results are sent in real time to smart devices so that employees can view them.

[1805] User (Administrator)

[1806] Users are responsible for managing the mental health of their employees and providing appropriate care when necessary. Specifically, this includes:

[1807] 1. Report Review and Care:

[1808] The server provides a dashboard where users can view mental health assessment reports for each employee, including graphs, scores, and conversation highlights.

[1809] We will consider options for interviews and care, and provide stress management training and temporary reassignments.

[1810] Specific examples

[1811] For example, if a store employee wearing smart glasses speaks in a "high-pitched, emphasized tone" while interacting with a customer, the voice data is sent to a server, where real-time emotion analysis is performed. As a result, the employee's stress level is assessed as rising, and the smart glasses display shows "high stress." Based on this information, managers can instruct employees to take a break. Managers can also take countermeasures based on this data later.

[1812] Prompt Sentence Examples

[1813] "The employee's voice is high-pitched and stressed. We suggest they take a 10-minute break."

[1814] "Your mental health score for the past week has been high. I suggest you take stress management training."

[1815] Thus, in order to implement this invention, it is necessary to build a system in which the components of the server, terminal, and user work together to evaluate and manage the mental state of employees in real time.

[1816] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1817] Step 1:

[1818] Audio data collection

[1819] The device collects employee voice data.

[1820] Input: Audio data captured from the built-in microphone of the smart glasses.

[1821] How it works: The built-in microphone records the conversation between the operator and the customer and sends the audio data to the server in real time.

[1822] Step 2:

[1823] Converting audio data to text data

[1824] The server converts the received voice data into text data.

[1825] Input: Audio data sent from the device.

[1826] Output: Text data.

[1827] What it does: It uses the Google Cloud Speech-to-Text API to convert audio data into text. A speech recognition model analyzes the acoustic signal and generates a corresponding string of characters.

[1828] Step 3:

[1829] Text data analysis

[1830] The server analyzes the converted text data and evaluates the emotions and psychological state.

[1831] Input: Text data.

[1832] Output: Emotion score and mental state assessment results.

[1833] Specific operation: Text data is input into a natural language processing (NLP) engine (e.g., spaCy, BERT) to analyze and extract sentiment, keywords, and contextual information.

[1834] Step 4:

[1835] Voice timbre feature extraction

[1836] The server extracts voice characteristics from the voice data.

[1837] Input: Audio data.

[1838] Output: Voice characteristics (pitch, tone, energy, etc.).

[1839] Specific operation: Apply an acoustic feature analysis algorithm (e.g., librosa) to extract features such as pitch, tone, and energy from the audio signal.

[1840] Step 5:

[1841] Applying the Emotion Engine

[1842] The server inputs the voice features into the emotion engine and recognizes the corresponding emotion.

[1843] Input: Voice timbre features.

[1844] Output: Emotion recognition results.

[1845] Specific operation: Using a generative AI model (e.g., BERT), emotions are classified from tone of voice features, and emotions such as anger, joy, and sadness are recognized.

[1846] Step 6:

[1847] Mental state assessment

[1848] The server combines the analysis results of the NLP engine and the recognition results of the emotion engine to evaluate the mental state of employees.

[1849] Input: Emotion score, voice features, and text analysis results.

[1850] Output: Mental state assessment score.

[1851] Specific operations: Evaluate overall mental state by assigning positive and negative emotional scores and calculating mental evaluation points.

[1852] Step 7:

[1853] Real-time display

[1854] The server transmits the evaluation results to the smart device in real time and displays them.

[1855] Input: Mental State Assessment score.

[1856] Output: Evaluation results displayed on smart devices.

[1857] Specific operation: The evaluation results are displayed on the smart glasses display, allowing employees to check their own mental state in real time.

[1858] Step 8:

[1859] Report Generation

[1860] The server generates a report based on the mental state evaluation results and provides it to the user.

[1861] Input: Accumulated data of evaluation results.

[1862] Output: Detailed mental state report.

[1863] Specific actions: Generate graphs and charts to create reports that visualize fluctuations in mental state for each employee, and display them on a dashboard for managers to review.

[1864] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1865] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1866] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1867] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1868] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1869] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1870] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1871] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1872] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1873] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1874] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1875] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1876] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1877] 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.

[1878] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1879] The hardware resource for executing a specific process can be any of the following types of processors: A processor, for example, is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1880] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1881] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1882] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1883] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1884] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1885] The following is further disclosed regarding the above embodiment.

[1886] (Claim 1)

[1887] means for collecting audio data;

[1888] means for converting voice data into text data;

[1889] means for analyzing text data and assessing emotions and psychological states;

[1890] A means of visualizing mental state based on the evaluation results, and

[1891] means for generating a report indicative of a mental state;

[1892] A system including:

[1893] (Claim 2)

[1894] 2. The system according to claim 1, further comprising means for extracting a feature of a tone of voice from the voice data and evaluating a psychological state based on the feature.

[1895] (Claim 3)

[1896] 10. The system of claim 1, further comprising means for collecting, converting and analyzing audio data in real time.

[1897] "Example 1"

[1898] (Claim 1)

[1899] means for collecting audio data;

[1900] means for converting voice data into text data;

[1901] means for analyzing text data and assessing emotions and psychological states;

[1902] A means of visualizing mental state based on the evaluation results, and

[1903] means for generating a report indicative of a mental state;

[1904] A means for transmitting the collected voice data to a server in real time;

[1905] A means for sending the converted text data to a natural language processing engine for sentiment analysis;

[1906] A means to accumulate mental evaluation points over time and display them in graphs and charts,

[1907] A means to provide a dashboard for managers to review mental assessment reports;

[1908] A system including:

[1909] (Claim 2)

[1910] 2. The system according to claim 1, further comprising means for extracting a feature of a tone of voice from the voice data and evaluating a psychological state based on the feature.

[1911] (Claim 3)

[1912] 10. The system of claim 1, further comprising means for collecting, converting and analyzing audio data in real time.

[1913] "Application Example 1"

[1914] (Claim 1)

[1915] means for collecting audio data;

[1916] means for converting voice data into text data;

[1917] means for analyzing text data and assessing emotions and psychological states;

[1918] A means of visualizing mental state based on the evaluation results, and

[1919] means for generating a report indicative of a mental state;

[1920] A means for visualizing the collected mental evaluation points over time;

[1921] A system including:

[1922] (Claim 2)

[1923] 2. The system according to claim 1, further comprising means for extracting a feature of a tone of voice from the voice data and evaluating a psychological state based on the feature.

[1924] (Claim 3)

[1925] 10. The system of claim 1, further comprising: means for collecting audio data from a microphone built into the smart glasses and transmitting the audio data to a server in real time.

[1926] "Example 2: Combining Emotion Engines"

[1927] (Claim 1)

[1928] means for collecting audio data;

[1929] means for temporarily storing the audio data;

[1930] means for converting voice data into text data;

[1931] A means of analyzing text data and extracting sentiment and contextual information;

[1932] A means for extracting voice timbre features from speech data;

[1933] A means for recognizing emotions based on vocal timbre features;

[1934] a means for evaluating mental states by integrating emotion recognition results and text analysis results;

[1935] A means of visualizing mental state based on the evaluation results, and

[1936] means for generating a report indicative of a mental state;

[1937] A system including:

[1938] (Claim 2)

[1939] 10. The system of claim 1, further comprising means for collecting, converting and analyzing audio data in real time.

[1940] (Claim 3)

[1941] 2. The system of claim 1, further comprising means for an administrator to check the mental assessment report through a dashboard and take appropriate countermeasures.

[1942] "Application example 2 when combining emotion engines"

[1943] (Claim 1)

[1944] means for collecting audio data;

[1945] means for converting voice data into text data;

[1946] means for analyzing text data and assessing emotions and psychological states;

[1947] A means of visualizing mental state based on the evaluation results, and

[1948] means for generating a report indicative of a mental state;

[1949] a means for displaying the results on a smart device;

[1950] A system including:

[1951] (Claim 2)

[1952] 2. The system according to claim 1, further comprising means for extracting a feature of a tone of voice from the voice data and evaluating a psychological state based on the feature.

[1953] (Claim 3)

[1954] 10. The system of claim 1, further comprising means for collecting, converting and analyzing audio data in real time. [Explanation of symbols]

[1955] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for collecting audio data; means for converting voice data into text data; means for analyzing text data and assessing emotions and psychological states; A means of visualizing mental state based on the evaluation results, and means for generating a report indicative of a mental state; A system including:

2. 2. The system according to claim 1, further comprising means for extracting a feature of a tone of voice from the voice data and evaluating a psychological state based on the feature.

3. 10. The system of claim 1, further comprising means for collecting, converting and analyzing audio data in real time.

Images