system

Abstract

We provide the system. [Solution] Means of collecting employee information, A means of identifying data related to operational efficiency from collected employee information, A method for analyzing employee emotions using natural language processing models, A method for detecting anomalies in business efficiency data using machine learning models, A method for integrating and analyzing sentiment analysis results and anomaly detection results using a deep learning model, A means of recommending countermeasures for high-risk employees, A system that includes this.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In companies where the employee turnover rate is increasing and there is a continuous shortage of manpower, it is difficult to easily analyze a large amount of data and identify high-risk employees in order to improve employee engagement. In addition, on-site human resources staff are overwhelmed with normal work and do not have time to analyze individual engagement surveys, productivity data, and attendance data in detail, so there is also a problem that they cannot take prompt and effective countermeasures. This invention is provided to solve the above problems.

Means for Solving the Problems

[0005] The present invention solves the problem by providing a system that includes means for collecting employee information, means for identifying data related to work efficiency from the collected information, means for analyzing employee sentiment using a natural language processing model, means for detecting anomalies in work efficiency data using a machine learning model, means for integrating and analyzing the sentiment analysis results and anomaly detection results using a deep learning model, and means for recommending countermeasures for high-risk employees. This makes it possible to effectively and efficiently improve employee engagement and reduce employee turnover.

[0006] "Means for collecting employee information" refers to functions for acquiring and accumulating relevant information such as employee engagement survey results, productivity data, and attendance data.

[0007] "Means for identifying data related to operational efficiency" refers to a function for extracting and identifying data related to operational productivity and work efficiency from collected employee information.

[0008] A "natural language processing model" refers to an algorithm that analyzes text data written by employees and uses that data to understand their emotions and intentions.

[0009] A "machine learning model" refers to an algorithm designed to detect anomalies in new data based on past data patterns.

[0010] A "deep learning model" refers to an algorithm that uses multi-layered artificial neural networks to analyze the correlations between complex data and perform advanced inference.

[0011] "Sentiment analysis results" refer to the results of evaluating employees' emotional states and satisfaction levels based on data extracted using a natural language processing model.

[0012] "Anomaly detection results" refer to records of abnormal patterns or events in business efficiency data detected using machine learning models.

[0013] A "high-risk employee" refers to an employee who, based on a combination of sentiment analysis results and anomaly detection results, is assessed as having a high probability of negatively impacting the company in the near future.

[0014] "Means of recommending countermeasures" refers to a function for proposing appropriate countermeasures to high-risk employees, and involves presenting specific action plans. [Brief explanation of the drawing]

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

Mode for Carrying Out the Invention

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

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

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

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

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

[0021] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0022] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0023] [First Embodiment]

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

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

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

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

[0028] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0029] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0030] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

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

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

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

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

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

[0036] This invention provides a system for improving employee engagement, and uses multiple modules to collect and analyze employee data.

[0037] First, the server automatically collects employee information from multiple systems. This information includes engagement survey results, productivity data, attendance data, and more. The collected data is stored in a central database and saved in a consistent format.

[0038] Next, the server uses a natural language processing model to analyze the text data obtained from the engagement survey. In this process, for example, it identifies negative emotions from an employee's comment such as, "The atmosphere at work hasn't been good lately." This emotion analysis allows for the evaluation of each employee's satisfaction level and stress level.

[0039] Furthermore, the server uses machine learning models to analyze productivity and attendance data, comparing them to past patterns to detect unusual activity. This anomaly detection allows for quick identification of situations such as an employee suddenly becoming more late or a decline in work efficiency.

[0040] These analysis results are integrated by a deep learning model, and the server identifies high-risk employees. High-risk employees are those who show negative results in both sentiment analysis and anomaly detection, and require special attention.

[0041] Finally, the server uses a recommendation engine to generate specific countermeasures for this high-risk employee. For example, recommendations such as "consult with the mental health support team" or "consider reassigning duties" can be provided to HR personnel via their terminals, enabling a quick response.

[0042] This system allows users to effectively and efficiently reduce employee turnover and improve employee engagement.

[0043] The following describes the processing flow.

[0044] Step 1:

[0045] The server automatically collects employee information from various systems. This information includes engagement survey results, productivity data, attendance data, etc., and the server stores it in a central database. During data collection, the server verifies the integrity of the data format and performs format conversion as needed.

[0046] Step 2:

[0047] The server inputs the text data from the engagement survey into a natural language processing model. This model analyzes the emotions contained in the survey comments and classifies each comment as "positive," "negative," or "neutral." The server then aggregates the analysis results and calculates an individual employee's emotion score.

[0048] Step 3:

[0049] The server inputs productivity and attendance data into a machine learning model. This model detects anomalous patterns by comparing them with historical data, for example, identifying sudden drops in productivity or anomalies in attendance. The server saves these results as an anomaly score for later analysis.

[0050] Step 4:

[0051] The server inputs the sentiment analysis results and anomaly detection results into a deep learning model, integrating them to identify high-risk employees. This analysis selects employees with high sentiment scores and anomaly scores, and the server generates this information as a report.

[0052] Step 5:

[0053] The server uses a recommendation engine to suggest countermeasures for high-risk employees. In this process, the server selects specific actions, such as "reducing workload" or "providing counseling," and displays the recommendations on the employee's device. HR personnel can receive this information via the device and use it to implement appropriate responses.

[0054] (Example 1)

[0055] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0056] In today's work environment, declining employee engagement and increasing stress are significant problems. These factors lead to higher turnover rates and impact corporate productivity, making early detection and countermeasures crucial. A system is needed that efficiently analyzes changes in employees' emotional states and work efficiency, and proposes appropriate responses to high-risk employees.

[0057] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0058] In this invention, the server includes means for collecting employee attributes, means for identifying information related to work efficiency from the collected employee attributes, means for analyzing employee sentiment using natural language processing technology, means for detecting anomalies in work efficiency information using machine learning methods, means for comprehensively analyzing the sentiment analysis results and anomaly detection results using deep learning methods, and means for recommending countermeasures for high-risk employees. This makes it possible to quickly grasp the status of employees and provide appropriate countermeasures, thereby reducing employee turnover and improving overall engagement within the company.

[0059] "Employee attributes" refer to information about individual employees, including data such as job duties, work efficiency, productivity, and attendance.

[0060] "Business efficiency" is an indicator that shows the results and efficiency of employees when performing their tasks.

[0061] "Natural language processing technology" is a technique that enables computers to understand and analyze human language, with the aim of analyzing the emotions and intentions behind text.

[0062] "Machine learning techniques" are technical means that allow computers to automatically learn patterns using large amounts of data and make predictions and judgments.

[0063] "Deep learning techniques" are technologies that use multi-layered neural networks to analyze complex data and provide more accurate analysis results.

[0064] A "high-risk employee" refers to an employee who shows negative results in sentiment analysis or anomaly detection and requires special attention.

[0065] A "means for recommending countermeasures" is a mechanism for proposing the optimal solution to an identified problem.

[0066] A "recommendation system" is an algorithm that suggests the optimal option based on a user's past behavior and tendencies.

[0067] This invention provides a system where a server effectively collects, analyzes, and recommends employee data. The server collects employee attributes from multiple data sources via APIs and database connections. This includes employee productivity information and attendance information. Each piece of information is acquired via general-purpose software such as Microsoft® Excel® or Google® Sheets and stored in a central database in a consistent format.

[0068] Regarding the collected attributes, the server uses natural language processing techniques to analyze the sentiment from employee comments and feedback. In this process, open-source language models such as BERT and GPT are utilized to extract sentiments such as positive, negative, and neutral from the data. For example, a negative sentiment is identified from a comment such as "I'm feeling stressed out by the recent project."

[0069] Next, the server uses machine learning techniques to analyze operational efficiency information. Specifically, it detects abnormal patterns such as increased tardiness or absenteeism, and decreased productivity. At this time, algorithms such as random forests and LSTMs are implemented to compare past data with current data and identify anomalies.

[0070] Furthermore, the sentiment analysis results and anomaly detection results are integrated and analyzed using deep learning techniques to identify high-risk employees. This method allows the server to assess the severity of each employee's problems and, if necessary, notify the HR department.

[0071] Ultimately, the server recommends appropriate countermeasures for high-risk employees. This process utilizes a recommendation system that presents optimal solutions based on past success stories. These include specific suggestions such as "guiding employees to mental health support" or "revising their work content."

[0072] As a concrete example, the following prompt statement can be used.

[0073] "Please perform a sentiment analysis based on the following employee comment and identify any negative emotions. Comment: 'I'm feeling stressed out by the recent project.'"

[0074] "Detect anomalies from employee data and compare them to past patterns to identify unusual behavior. Data: Late arrivals: 5 times / month, Productivity score: 70%"

[0075] Through this system, users can improve employee engagement and reduce employee turnover.

[0076] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0077] Step 1:

[0078] The server uses APIs and database connections to collect employee productivity and attendance information from multiple systems. This includes data acquisition from employee management systems and productivity management tools. Since input data may be provided in multiple formats, the server performs data conversion processing and stores these in a consistent format in a central database. The output is a format-converted dataset, which is then used for further analysis.

[0079] Step 2:

[0080] The server retrieves text data from a central database and applies natural language processing techniques to perform sentiment analysis. Specifically, it inputs the text data into a generative AI model such as BERT, which outputs a sentiment score that includes sentiment labels (positive, negative, neutral). This process identifies keywords within each comment and evaluates the direction of the sentiment to analyze each employee's satisfaction and stress level.

[0081] Step 3:

[0082] The server utilizes machine learning techniques to analyze operational efficiency information. It uses historical attendance data and productivity scores as input to perform anomaly detection. Random forest and LSTM models are used to identify unusual patterns. The output consists of data points flagged as anomalies, which the server uses to determine what is normal and what is abnormal.

[0083] Step 4:

[0084] The server applies deep learning techniques to integrate sentiment analysis results and anomaly detection results. This calculates an overall risk score for each employee. In this process, individual scores are used as input for a weighted, integrated analysis, and an integrated risk assessment is generated as output. This is used to identify high-risk employees.

[0085] Step 5:

[0086] The server uses a recommendation engine to generate countermeasures for employees identified as high-risk. It utilizes an integrated risk assessment as input and outputs optimal suggestions using a model trained on past successes. Specifically, it might present concrete countermeasures to the HR department, such as "introduction to mental health support" or "consideration of changes to job duties."

[0087] (Application Example 1)

[0088] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0089] Traditional employee engagement systems struggled to provide real-time insights into employee and machine performance, making it difficult to quickly implement appropriate countermeasures. Furthermore, they lacked systems that integrated sentiment analysis and anomaly detection, instead relying solely on individual analyses. As a result, overall organizational productivity improvements and optimal collaboration between machines and humans were not achieved.

[0090] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0091] In this invention, the server includes means for collecting employee information and machine information, means for identifying data related to work efficiency from the collected employee information and machine information, and means for analyzing employee sentiment and feedback from workers using a natural language processing model. This enables integrated management of humans and machines within the organization, leading to improved productivity and proper engagement management.

[0092] "Employee information" refers to data about employees' attributes, behavior, and work, including productivity data and attendance data.

[0093] "Machine information" refers to data related to the operating status and performance of equipment, including operational data collected through sensors.

[0094] "Work efficiency" is an indicator that shows the performance level of tasks and operations performed by employees and machines.

[0095] A "natural language processing model" is a collection of algorithms and methods for processing and analyzing human language using computers.

[0096] "Emotional and feedback analysis" is a process of analyzing employees' emotions and opinions from text data and using that information to improve the organization.

[0097] A "machine learning model" is a mathematical model that learns patterns from data and uses them for prediction and classification.

[0098] Anomaly detection is the process of finding patterns or behaviors that are different from the norm within data.

[0099] A "deep learning model" is a technique that uses a neural network, which has a structure similar to the human brain, to learn complex patterns.

[0100] "Measures for high-risk employees and equipment" refer to proposed improvements and countermeasures for employees or equipment that pose risks beyond normal operations.

[0101] A "recommendation engine" is a system that provides individually optimized suggestions and actions based on user data and context.

[0102] This invention provides a system that improves overall organizational work efficiency and engagement by efficiently collecting and analyzing employee and machine operation data. The server first automatically collects employee and machine information from various sensors and IoT devices installed in the factory and workplace environment. This information, including productivity data, attendance data, and equipment operation data, is stored in a central database in a consistent format.

[0103] The server utilizes natural language processing technology to analyze text-based feedback provided by employees and operators. This allows for the extraction and evaluation of employees' emotional states and opinions on the work environment. OpenAI® language models are used for natural language processing.

[0104] Furthermore, the server uses machine learning models to detect anomalies in the collected work efficiency data. This process identifies patterns that deviate from the normal operation of employees and equipment, aiming for early detection of risks. Suitable libraries for this purpose include scikit-learn.

[0105] The results of both anomaly detection and sentiment analysis are integrated and analyzed by a deep learning model. At this stage, TENSORFLOW® is used to learn the complex relationships between data and derive effective improvement strategies.

[0106] For employees and equipment identified as high-risk, a recommendation engine will provide specific countermeasures. Based on past analysis results, this engine proposes the optimal action to help improve organizational efficiency.

[0107] For example, if an operator inputs feedback such as "Recently, work has been falling behind schedule," the system will respond to that feedback and efficiently make suggestions. An example of a prompt using a generative AI model is: "Perform sentiment analysis on the operator's feedback and generate recommendations for improving the environment. Input feedback: 'Recently, work has been falling behind schedule.'"

[0108] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0109] Step 1:

[0110] The server collects employee and machine information from IoT devices and sensors installed in factories and workplaces. This information includes productivity data, attendance data, and equipment operation data. This data is aggregated in real time and stored in a central database in a consistent format.

[0111] Step 2:

[0112] The server applies natural language processing to the employee feedback it collects. It analyzes the text-based feedback entered by users and extracts opinions on emotional states and work performance. This analysis uses a generative AI model to evaluate the positive and negative aspects of emotions and opinions in particular. The input is text data, and the output is an emotional score and keywords.

[0113] Step 3:

[0114] The server applies machine learning models to collected work efficiency data and equipment operation data to perform anomaly detection. The input is raw numerical data, and statistical methods and machine learning algorithms are used to detect deviations from normal patterns. The output provides the anomaly detection results and their associated parameters.

[0115] Step 4:

[0116] The server integrates the results of sentiment analysis and anomaly detection using a deep learning model. This model learns the complex relationships between data and identifies risk patterns. The input is the sentiment score and anomaly detection results obtained in the previous step, and the output is data for risk assessment and improvement suggestions.

[0117] Step 5:

[0118] The device generates specific countermeasures for employees and equipment identified as high-risk. Using a recommendation engine, optimal actions are suggested and presented to the user based on past analysis results. The input is the result of integrated analysis, and the output is specific action suggestions and action plans.

[0119] Step 6:

[0120] Users evaluate the proposed solutions and select actionable steps. This improves on-site work efficiency and employee engagement, and enables immediate corrective measures.

[0121] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0122] This invention is a system for improving employee engagement, combining an emotion engine with various analytical models to achieve more advanced data analysis.

[0123] First, the server automatically collects employee information. This information includes productivity data, attendance data, and feedback from engagement surveys. This data is stored consistently in a central database.

[0124] Next, the server uses a natural language processing model to analyze the emotions expressed in the survey comments. For example, it detects negative emotions from a comment such as, "My motivation at work has been low lately." This result is then aggregated as an emotion score for each individual employee.

[0125] Furthermore, the server utilizes machine learning models to analyze productivity and attendance data to detect anomalies. For example, it might detect an anomaly if an employee unexpectedly starts being absent frequently.

[0126] In addition to these analysis processes, an emotion engine is introduced to analyze the user's emotions in real time from their facial expressions and voice. The data provided by the emotion engine is integrated with other emotion analysis results, enabling more comprehensive analysis based on richer information.

[0127] A deep learning model integrates sentiment analysis results and anomaly detection results to accurately identify high-risk employees. Based on this, the server uses a recommendation engine to suggest appropriate countermeasures for high-risk employees.

[0128] For example, if an employee is identified whose emotional score has decreased and whose work efficiency has also declined, specific actions such as "providing mental support" or "proposing flexible working hours" can be taken. These recommendations are provided to HR personnel via their devices, enabling quick decision-making and response.

[0129] In this way, the system of the present invention achieves more accurate employee support through a comprehensive analysis that takes user emotions into account.

[0130] The following describes the processing flow.

[0131] Step 1:

[0132] The server automatically collects employee information from each system. This includes a wide range of data, such as productivity data, attendance data, and engagement survey results. The server stores the collected information in a central database and standardizes the format to maintain data consistency.

[0133] Step 2:

[0134] The server uses a natural language processing model to analyze the text data from the engagement survey. This analysis extracts emotions from each comment and assigns them emotion labels such as "positive," "negative," or "neutral." The server then aggregates these results to calculate an emotion score for each employee.

[0135] Step 3:

[0136] The server feeds the collected productivity and attendance data into a machine learning model. This model detects anomalies by comparing them to past data patterns, for example, identifying decreased productivity or increased tardiness for specific employees. The server records the results of anomaly detection as an anomaly score.

[0137] Step 4:

[0138] The emotion engine analyzes the user's facial expressions and voice data in real time. While the user is in a video conference or call, the emotion engine processes this data, evaluates the user's emotional state, and generates a real-time emotion score.

[0139] Step 5:

[0140] The server inputs sentiment scores, anomaly scores, and real-time user sentiment scores into a deep learning model for integrated analysis. This analysis identifies high-risk employees and assesses their impact.

[0141] Step 6:

[0142] The server uses a recommendation engine to generate countermeasures for identified high-risk employees. These recommendations may include, for example, "referral to a professional counselor" or "suggestions for improving the work environment," and the server displays this information on the terminal and notifies the HR personnel.

[0143] Step 7:

[0144] HR personnel review the recommendations received on their devices, select appropriate countermeasures, and implement them. This speeds up responses within the organization and leads to improved employee engagement.

[0145] (Example 2)

[0146] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0147] For many companies, effectively monitoring employee work efficiency and emotional states, and identifying and addressing high-risk employees early on, is a challenge. While changes in employee emotions impact productivity, understanding emotions is subjective and difficult. Furthermore, early detection of abnormal behavior and the provision of effective countermeasures are crucial, but traditional methods lack real-time and multifaceted analysis. Therefore, there is a need for the development of systems that efficiently and comprehensively manage employee conditions and enable early intervention.

[0148] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0149] In this invention, the server includes means for collecting employee information, means for identifying data related to the efficiency of work activities, means for analyzing employee emotions using a language analysis model, means for detecting anomalies in work activity efficiency data using a learning algorithm, means for comprehensively analyzing emotion analysis results and anomaly detection results using deep learning technology, a device for analyzing user emotions in real time, and a device for recommending countermeasures for high-risk employees. This makes it possible to quickly and accurately grasp changes in employee emotions and anomalies in work activities and to suggest appropriate countermeasures.

[0150] "Means for collecting employee information" refers to a device or program that has the function of automatically collecting and storing data related to individual employees within an organization.

[0151] "Means for identifying data related to the efficiency of job activities" refers to methods or devices for extracting and analyzing work efficiency and performance indicators from collected employee information.

[0152] A "language analysis model" is an algorithm or program that uses natural language processing technology to analyze text data and interpret emotions and intentions.

[0153] A "learning algorithm" is a computational method or program that uses machine learning to recognize patterns in data and detect anomalies and trends.

[0154] "Deep learning technology" is an advanced data analysis technique that uses neural networks, aiming to automatically analyze complex data structures and make integrated decisions.

[0155] A "device that analyzes user emotions in real time" refers to hardware and software that instantly collects user facial expressions and voice data and uses this data to score their emotional state.

[0156] A "countermeasure recommendation system" is an algorithm or program that proposes the most suitable action plan or support measures for employees identified as high-risk.

[0157] To implement this invention, a server-centered data processing system is first required. The server first centrally collects employee information, including productivity indicators and attendance data. The server can use general database management software to store this data in a central database.

[0158] Next, the server uses the collected data to analyze the efficiency of work activities and employee sentiment. For language analysis models, general models applying natural language processing techniques (e.g., BERT or GPT) can be used. This quantifies sentiment from survey comments and forms a sentiment score. For the learning algorithm, machine learning techniques are utilized, and models for detecting anomalies (e.g., Random Forest or SVM) are used. For deep learning techniques, neural networks using TensorFlow or PyTorch are effective. This allows for the integration of sentiment analysis results and anomaly detection results, enabling risk identification.

[0159] To understand the user's emotions in real time, the device collects data via its camera and microphone. An emotion engine analyzes this data and instantly updates the emotion score. This data is then sent to a server where it is integrated with other results.

[0160] For example, by inputting a prompt such as, "Please suggest specific actions to address declining employee motivation," into the AI ​​model, appropriate countermeasures can be obtained. These suggestions, as support measures for high-risk employees, are output using the suggestion engine and provided to HR personnel via their terminals.

[0161] In this way, the collaboration between servers, terminals, and users enables comprehensive management of employee status and facilitates quick and appropriate responses.

[0162] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0163] Step 1:

[0164] The server collects employee information. Inputs include productivity metrics, attendance data, and engagement survey feedback. This data is automatically collected and stored in a central database in a structured format. The output is a unified dataset. Standardizing and storing inputs from each data source ensures consistency for subsequent data analysis.

[0165] Step 2:

[0166] The server uses natural language processing techniques to analyze employee sentiment. The input is comments from an engagement survey. The analysis uses language processing models (e.g., BERT or GPT) to extract sentiment from each comment. The output is a sentiment score, such as negative, positive, or neutral. This allows for the quantification of each employee's emotional state.

[0167] Step 3:

[0168] The server uses machine learning techniques to detect anomalies. Inputs include productivity metrics and attendance data. This data is analyzed using learning algorithms (e.g., Random Forest or SVM) to detect anomalous patterns. The output is the data points where anomalies were detected. This allows for the early detection of abnormal behavior and sudden changes.

[0169] Step 4:

[0170] The device analyzes the user's emotions in real time. As input, the user's facial expressions and voice data are collected via the camera and microphone. An emotion engine analyzes this data and instantly scores the emotional state. This data is sent to a server and integrated with other results. The output is the user's current emotion score. Tracking changes in real time allows for more accurate emotion assessment.

[0171] Step 5:

[0172] The server integrates the results using deep learning technology. Inputs include sentiment scores and anomaly detection results. Deep learning techniques (e.g., TensorFlow or PyTorch) are used to integrate and analyze this data, identifying high-risk employees. The output is a list of employees identified as high-risk. This allows for a clear prioritization of those requiring action.

[0173] Step 6:

[0174] The server generates countermeasures and notifies the user via the terminal. The input is information on high-risk employees. The suggestion engine uses a generation AI model to create specific countermeasures. The output may include countermeasures such as "providing mental support" or "proposing flexible working hours." This notification, received on the terminal, allows HR personnel to respond quickly.

[0175] (Application Example 2)

[0176] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0177] A decline in employee work efficiency and motivation at a logistics center is a serious problem that has a significant impact on overall productivity. In such an environment, it is necessary to understand employees' emotional states and work efficiency in real time and to quickly provide appropriate improvement measures. However, conventional methods have made it difficult to effectively analyze such complex human emotional and behavioral data and propose countermeasures.

[0178] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0179] In this invention, the server includes means for collecting employee attributes, means for identifying information related to work efficiency from the collected employee attributes, means for analyzing employee emotions using natural language processing technology, means for detecting anomalies in work efficiency using machine learning technology, means for comprehensively analyzing the emotion analysis results and anomaly detection results using deep learning technology, means for recommending countermeasures for high-risk employees, and means for providing a user interface for visualizing real-time emotion and productivity data. This enables real-time monitoring of employee emotions and work efficiency within a logistics center, and allows for the provision of rapid and accurate support measures.

[0180] "Employee attributes" is a general term for personal information and work performance information related to employees.

[0181] "Work efficiency" refers to the proportion and quality of work results achieved by employees within a certain time frame.

[0182] "Natural language processing technology" is a general term for technologies used by computers to understand and analyze human language.

[0183] "Machine learning technology" is a technique that enables computers to learn from data and develop the ability to identify patterns and anomalies.

[0184] "Deep learning technology" is a technique that uses a multi-layered neural network to automatically extract sophisticated features from large-scale data.

[0185] A "high-risk employee" refers to an employee whose work performance or emotional state falls outside the standard, potentially requiring special measures or support.

[0186] "Means of recommending countermeasures" refers to methods for presenting appropriate support measures and improvement measures for high-risk employees.

[0187] A "user interface" refers to an interface that provides a visual or manipulative means for a system to interact with a user and exchange information.

[0188] The system implementing this invention aims to monitor employee emotions and work efficiency in real time at a logistics center and to suggest appropriate countermeasures to high-risk employees. To achieve this objective, the system includes the following configuration.

[0189] The server automatically collects employee attributes, such as the labor resources to which each employee belongs and attendance records. This involves using sensors and RFID readers as hardware, and Python as the software. The collected employee data is consistently stored in a central database.

[0190] Next, the server uses a generative AI model to perform natural language processing and analyze emotions from text data such as employee feedback and comments. During this process, the emotion engine calculates an emotion score and reveals the emotional state.

[0191] Furthermore, the server utilizes machine learning techniques to detect anomalies in the collected work efficiency data. Deep learning methods such as TensorFlow integrate sentiment analysis results and anomaly detection results to identify employees deemed high-risk.

[0192] The data visualized via the terminal is provided to the user in real time, and the recommendation engine generates concrete measures to support countermeasures for high-risk employees. This process supports rapid and efficient business collaboration.

[0193] For example, if a particular employee leaves feedback stating that their work motivation has recently decreased, the system will analyze that comment and recommend essential support measures such as "conducting individual interviews" or "flexibly adjusting working hours."

[0194] An example of a prompt to input into the generating AI model is: "Analyze the employee's sentiment score from the following data and suggest necessary support measures: {Employee comments, attendance records, sentiment data}."

[0195] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0196] Step 1:

[0197] The server collects employee attributes such as labor resources and attendance records from sensors and attendance management systems placed within the logistics center. Inputs include employee fingerprint data and RFID tag information, and output is that this data is stored in a central database. Data collection enables the visualization of basic employee information.

[0198] Step 2:

[0199] The server analyzes text data using a generative AI model based on collected employee data. Input includes employee daily report comments and feedback. Natural language processing techniques are used to analyze emotions from this text data. The output provides an analyzed emotion score and its trends. This analysis makes it possible to predict employee emotional fluctuations.

[0200] Step 3:

[0201] The server uses machine learning techniques to validate collected work efficiency data and detect anomalies. Inputs include employee working hours, work progress, and the number of work errors. Through data analysis, patterned anomalies are identified, and the presence and frequency of these anomalies are clearly indicated as output. This step allows for the proactive detection of potential employee problems.

[0202] Step 4:

[0203] The server uses deep learning technology to comprehensively analyze sentiment analysis results and anomaly detection results to identify high-risk employees. The inputs are sentiment scores and anomaly detection results. The output of the integrated analysis is a risk assessment, indicating the likelihood of high risk. This enables the early detection of employees who require attention.

[0204] Step 5:

[0205] Users view real-time data using the user interface on their device and receive suggestions from the recommendation engine. Inputs include analysis results and recommendation results. Outputs display the suggested content, serving as a source of information for developing concrete action plans. In this step, more specific support measures are formulated for high-risk employees.

[0206] Step 6:

[0207] For example, the terminal receives feedback from a specific employee, such as "Recently, their work motivation has decreased," and passes this feedback to a recommendation engine, which then presents the analysis results to the user. In this process, feedback and historical data are submitted as input, and the output is an optimal solution based on machine learning. This allows for the rapid provision of appropriate support measures to employees.

[0208] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0209] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0210] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0211] [Second Embodiment]

[0212] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0213] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0214] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0215] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0216] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0217] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0218] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0219] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0220] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0221] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0222] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0223] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0224] This invention provides a system for improving employee engagement, and uses multiple modules to collect and analyze employee data.

[0225] First, the server automatically collects employee information from multiple systems. This information includes engagement survey results, productivity data, attendance data, and more. The collected data is stored in a central database and saved in a consistent format.

[0226] Next, the server uses a natural language processing model to analyze the text data obtained from the engagement survey. In this process, for example, it identifies negative emotions from an employee's comment such as, "The atmosphere at work hasn't been good lately." This emotion analysis allows for the evaluation of each employee's satisfaction level and stress level.

[0227] Furthermore, the server uses machine learning models to analyze productivity and attendance data, comparing them to past patterns to detect unusual activity. This anomaly detection allows for quick identification of situations such as an employee suddenly becoming more late or a decline in work efficiency.

[0228] These analysis results are integrated by a deep learning model, and the server identifies high-risk employees. High-risk employees are those who show negative results in both sentiment analysis and anomaly detection, and require special attention.

[0229] Finally, the server uses a recommendation engine to generate specific countermeasures for this high-risk employee. For example, recommendations such as "consult with the mental health support team" or "consider reassigning duties" can be provided to HR personnel via their terminals, enabling a quick response.

[0230] This system allows users to effectively and efficiently reduce employee turnover and improve employee engagement.

[0231] The following describes the processing flow.

[0232] Step 1:

[0233] The server automatically collects employee information from various systems. This information includes engagement survey results, productivity data, attendance data, etc., and the server stores it in a central database. During data collection, the server verifies the integrity of the data format and performs format conversion as needed.

[0234] Step 2:

[0235] The server inputs the text data from the engagement survey into a natural language processing model. This model analyzes the emotions contained in the survey comments and classifies each comment as "positive," "negative," or "neutral." The server then aggregates the analysis results and calculates an individual employee's emotion score.

[0236] Step 3:

[0237] The server inputs productivity and attendance data into a machine learning model. This model detects anomalous patterns by comparing them with historical data, for example, identifying sudden drops in productivity or anomalies in attendance. The server saves these results as an anomaly score for later analysis.

[0238] Step 4:

[0239] The server inputs the sentiment analysis results and anomaly detection results into a deep learning model, integrating them to identify high-risk employees. This analysis selects employees with high sentiment scores and anomaly scores, and the server generates this information as a report.

[0240] Step 5:

[0241] The server uses a recommendation engine to suggest countermeasures for high-risk employees. In this process, the server selects specific actions, such as "reducing workload" or "providing counseling," and displays the recommendations on the employee's device. HR personnel can receive this information via the device and use it to implement appropriate responses.

[0242] (Example 1)

[0243] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0244] In today's work environment, declining employee engagement and increasing stress are significant problems. These factors lead to higher turnover rates and impact corporate productivity, making early detection and countermeasures crucial. A system is needed that efficiently analyzes changes in employees' emotional states and work efficiency, and proposes appropriate responses to high-risk employees.

[0245] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0246] In this invention, the server includes means for collecting employee attributes, means for identifying information related to work efficiency from the collected employee attributes, means for analyzing employee sentiment using natural language processing technology, means for detecting anomalies in work efficiency information using machine learning methods, means for comprehensively analyzing the sentiment analysis results and anomaly detection results using deep learning methods, and means for recommending countermeasures for high-risk employees. This makes it possible to quickly grasp the status of employees and provide appropriate countermeasures, thereby reducing employee turnover and improving overall engagement within the company.

[0247] "Employee attributes" refer to information about individual employees, including data such as job duties, work efficiency, productivity, and attendance.

[0248] "Business efficiency" is an indicator that shows the results and efficiency of employees when performing their tasks.

[0249] "Natural language processing technology" is a technique that enables computers to understand and analyze human language, with the aim of analyzing the emotions and intentions behind text.

[0250] "Machine learning techniques" are technical means that allow computers to automatically learn patterns using large amounts of data and make predictions and judgments.

[0251] "Deep learning techniques" are technologies that use multi-layered neural networks to analyze complex data and provide more accurate analysis results.

[0252] A "high-risk employee" refers to an employee who shows negative results in sentiment analysis or anomaly detection and requires special attention.

[0253] A "means for recommending countermeasures" is a mechanism for proposing the optimal solution to an identified problem.

[0254] A "recommendation system" is an algorithm that suggests the optimal option based on a user's past behavior and tendencies.

[0255] This invention provides a system where a server effectively collects, analyzes, and recommends employee data. The server collects employee attributes from multiple data sources via APIs and database connections. This includes employee productivity information and attendance information. Each piece of information is retrieved via general-purpose software such as Microsoft Excel or Google Sheets and stored in a central database in a consistent format.

[0256] Regarding the collected attributes, the server uses natural language processing techniques to analyze the sentiment from employee comments and feedback. In this process, open-source language models such as BERT and GPT are utilized to extract sentiments such as positive, negative, and neutral from the data. For example, a negative sentiment is identified from a comment such as "I'm feeling stressed out by the recent project."

[0257] Next, the server uses machine learning techniques to analyze operational efficiency information. Specifically, it detects abnormal patterns such as increased tardiness or absenteeism, and decreased productivity. At this time, algorithms such as random forests and LSTMs are implemented to compare past data with current data and identify anomalies.

[0258] Furthermore, the sentiment analysis results and anomaly detection results are integrated and analyzed using deep learning techniques to identify high-risk employees. This method allows the server to assess the severity of each employee's problems and, if necessary, notify the HR department.

[0259] Ultimately, the server recommends appropriate countermeasures for high-risk employees. This process utilizes a recommendation system that presents optimal solutions based on past success stories. These include specific suggestions such as "guiding employees to mental health support" or "revising their work content."

[0260] As a concrete example, the following prompt statement can be used.

[0261] "Please perform a sentiment analysis based on the following employee comment and identify any negative emotions. Comment: 'I'm feeling stressed out by the recent project.'"

[0262] "Detect anomalies from employee data and compare them to past patterns to identify unusual behavior. Data: Late arrivals: 5 times / month, Productivity score: 70%"

[0263] Through this system, users can improve employee engagement and reduce employee turnover.

[0264] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0265] Step 1:

[0266] The server uses APIs and database connections to collect employee productivity and attendance information from multiple systems. This includes data acquisition from employee management systems and productivity management tools. Since input data may be provided in multiple formats, the server performs data conversion processing and stores these in a consistent format in a central database. The output is a format-converted dataset, which is then used for further analysis.

[0267] Step 2:

[0268] The server retrieves text data from a central database and applies natural language processing techniques to perform sentiment analysis. Specifically, it inputs the text data into a generative AI model such as BERT, which outputs a sentiment score that includes sentiment labels (positive, negative, neutral). This process identifies keywords within each comment and evaluates the direction of the sentiment to analyze each employee's satisfaction and stress level.

[0269] Step 3:

[0270] The server utilizes machine learning techniques to analyze operational efficiency information. It uses historical attendance data and productivity scores as input to perform anomaly detection. Random forest and LSTM models are used to identify unusual patterns. The output consists of data points flagged as anomalies, which the server uses to determine what is normal and what is abnormal.

[0271] Step 4:

[0272] The server applies deep learning techniques to integrate sentiment analysis results and anomaly detection results. This calculates an overall risk score for each employee. In this process, individual scores are used as input for a weighted, integrated analysis, and an integrated risk assessment is generated as output. This is used to identify high-risk employees.

[0273] Step 5:

[0274] The server uses a recommendation engine to generate countermeasures for employees identified as high-risk. It utilizes an integrated risk assessment as input and outputs optimal suggestions using a model trained on past successes. Specifically, it might present concrete countermeasures to the HR department, such as "introduction to mental health support" or "consideration of changes to job duties."

[0275] (Application Example 1)

[0276] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0277] Traditional employee engagement systems struggled to provide real-time insights into employee and machine performance, making it difficult to quickly implement appropriate countermeasures. Furthermore, they lacked systems that integrated sentiment analysis and anomaly detection, instead relying solely on individual analyses. As a result, overall organizational productivity improvements and optimal collaboration between machines and humans were not achieved.

[0278] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0279] In this invention, the server includes means for collecting employee information and machine information, means for identifying data related to work efficiency from the collected employee information and machine information, and means for analyzing employee sentiment and feedback from workers using a natural language processing model. This enables integrated management of humans and machines within the organization, leading to improved productivity and proper engagement management.

[0280] "Employee information" refers to data about employees' attributes, behavior, and work, including productivity data and attendance data.

[0281] "Machine information" refers to data related to the operating status and performance of equipment, including operational data collected through sensors.

[0282] "Work efficiency" is an indicator that shows the performance level of tasks and operations performed by employees and machines.

[0283] A "natural language processing model" is a collection of algorithms and methods for processing and analyzing human language using computers.

[0284] "Analysis of Emotions and Feedback" is a process of analyzing employees' emotions and opinions from text data and utilizing them for organizational improvement.

[0285] "Machine learning model" is a mathematical model that learns patterns from data for prediction and classification.

[0286] "Anomaly detection" is a process of finding patterns and behaviors different from normal in data.

[0287] "Deep learning model" is a method of learning complex patterns using neural networks with a structure similar to the human brain.

[0288] "Countermeasures for high-risk employees and equipment" are improvement measures and coping methods proposed for employees and equipment with risks deviating from normal operations.

[0289] "Recommendation engine" is a system that provides individually optimized suggestions and actions based on user data and context.

[0290] This invention provides a system that improves the work efficiency and engagement of the entire organization by efficiently collecting and analyzing the operation data of employees and machines. The server first automatically collects employee information and machine information from various sensors and IoT devices installed in the factory or workplace environment. This information includes productivity data, attendance data, machine operation data, etc., and is stored in a central database in a consistent format.

[0291] The server utilizes natural language processing technology to analyze text-based feedback provided by employees and operators. As a result, it is possible to extract and evaluate employees' emotional states and opinions on the working environment. For natural language processing, language models such as OpenAI's are used.

[0292] Furthermore, the server uses machine learning models to detect anomalies in the collected work efficiency data. This process identifies patterns that deviate from the normal operation of employees and equipment, aiming for early detection of risks. Suitable libraries for this purpose include scikit-learn.

[0293] The results of both anomaly detection and sentiment analysis are integrated and analyzed by a deep learning model. At this stage, TensorFlow is used to learn the complex relationships between the data and derive effective improvement strategies.

[0294] For employees and equipment identified as high-risk, a recommendation engine will provide specific countermeasures. Based on past analysis results, this engine proposes the optimal action to help improve organizational efficiency.

[0295] For example, if an operator inputs feedback such as "Recently, work has been falling behind schedule," the system will respond to that feedback and efficiently make suggestions. An example of a prompt using a generative AI model is: "Perform sentiment analysis on the operator's feedback and generate recommendations for improving the environment. Input feedback: 'Recently, work has been falling behind schedule.'"

[0296] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0297] Step 1:

[0298] The server collects employee and machine information from IoT devices and sensors installed in factories and workplaces. This information includes productivity data, attendance data, and equipment operation data. This data is aggregated in real time and stored in a central database in a consistent format.

[0299] Step 2:

[0300] The server applies natural language processing to the collected feedback from employees. It analyzes the text-based feedback entered by the user and extracts the emotional state and opinions regarding work. For this analysis, a generative AI model is used to evaluate particularly the positivity and negativity of emotions and opinions. The input is text data, and the output is an emotion score and keywords.

[0301] Step 3:

[0302] The server applies a machine learning model to the collected work efficiency data and equipment operation data for anomaly detection. The input is numerical information of raw data, and statistical methods and machine learning algorithms are used to detect deviations from normal patterns. As output, the results of anomaly detection and their related parameters are obtained.

[0303] Step 4:

[0304] The server integratively analyzes the results of sentiment analysis and anomaly detection using a deep learning model. This model learns the complex relationships between data and identifies risk patterns. The input is the emotion score and anomaly detection results obtained in the previous steps, and the output is data for risk assessment and improvement suggestions.

[0305] [[ID=2,0]]Step 5:

[0306] The terminal generates specific countermeasures for employees and equipment identified as high risk. Using a recommendation engine, the optimal actions are proposed from the previous analysis results and presented to the user. The input is the result of the integrated analysis, and the output is specific action proposals and action plans.

[0307] Step 6:

[0308] The user evaluates the proposed countermeasures and selects executable actions. Thereby, the work efficiency at the site and the engagement of employees are improved, and immediate improvement measures are taken.

[0309] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0310] This invention is a system for improving employee engagement, combining an emotion engine with various analytical models to achieve more advanced data analysis.

[0311] First, the server automatically collects employee information. This information includes productivity data, attendance data, and feedback from engagement surveys. This data is stored consistently in a central database.

[0312] Next, the server uses a natural language processing model to analyze the emotions expressed in the survey comments. For example, it detects negative emotions from a comment such as, "My motivation at work has been low lately." This result is then aggregated as an emotion score for each individual employee.

[0313] Furthermore, the server utilizes machine learning models to analyze productivity and attendance data to detect anomalies. For example, it might detect an anomaly if an employee unexpectedly starts being absent frequently.

[0314] In addition to these analysis processes, an emotion engine is introduced to analyze the user's emotions in real time from their facial expressions and voice. The data provided by the emotion engine is integrated with other emotion analysis results, enabling more comprehensive analysis based on richer information.

[0315] A deep learning model integrates sentiment analysis results and anomaly detection results to accurately identify high-risk employees. Based on this, the server uses a recommendation engine to suggest appropriate countermeasures for high-risk employees.

[0316] For example, if an employee is identified whose emotional score has decreased and whose work efficiency has also declined, specific actions such as "providing mental support" or "proposing flexible working hours" can be taken. These recommendations are provided to HR personnel via their devices, enabling quick decision-making and response.

[0317] In this way, the system of the present invention achieves more accurate employee support through a comprehensive analysis that takes user emotions into account.

[0318] The following describes the processing flow.

[0319] Step 1:

[0320] The server automatically collects employee information from each system. This includes a wide range of data, such as productivity data, attendance data, and engagement survey results. The server stores the collected information in a central database and standardizes the format to maintain data consistency.

[0321] Step 2:

[0322] The server uses a natural language processing model to analyze the text data from the engagement survey. This analysis extracts emotions from each comment and assigns them emotion labels such as "positive," "negative," or "neutral." The server then aggregates these results to calculate an emotion score for each employee.

[0323] Step 3:

[0324] The server feeds the collected productivity and attendance data into a machine learning model. This model detects anomalies by comparing them to past data patterns, for example, identifying decreased productivity or increased tardiness for specific employees. The server records the results of anomaly detection as an anomaly score.

[0325] Step 4:

[0326] The emotion engine analyzes the user's facial expressions and voice data in real time. While the user is in a video conference or call, the emotion engine processes this data, evaluates the user's emotional state, and generates a real-time emotion score.

[0327] Step 5:

[0328] The server inputs sentiment scores, anomaly scores, and real-time user sentiment scores into a deep learning model for integrated analysis. This analysis identifies high-risk employees and assesses their impact.

[0329] Step 6:

[0330] The server uses a recommendation engine to generate countermeasures for identified high-risk employees. These recommendations may include, for example, "referral to a professional counselor" or "suggestions for improving the work environment," and the server displays this information on the terminal and notifies the HR personnel.

[0331] Step 7:

[0332] HR personnel review the recommendations received on their devices, select appropriate countermeasures, and implement them. This speeds up responses within the organization and leads to improved employee engagement.

[0333] (Example 2)

[0334] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0335] For many companies, effectively monitoring employee work efficiency and emotional states, and identifying and addressing high-risk employees early on, is a challenge. While changes in employee emotions impact productivity, understanding emotions is subjective and difficult. Furthermore, early detection of abnormal behavior and the provision of effective countermeasures are crucial, but traditional methods lack real-time and multifaceted analysis. Therefore, there is a need for the development of systems that efficiently and comprehensively manage employee conditions and enable early intervention.

[0336] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0337] In this invention, the server includes means for collecting employee information, means for identifying data related to the efficiency of work activities, means for analyzing employee emotions using a language analysis model, means for detecting anomalies in work activity efficiency data using a learning algorithm, means for comprehensively analyzing emotion analysis results and anomaly detection results using deep learning technology, a device for analyzing user emotions in real time, and a device for recommending countermeasures for high-risk employees. This makes it possible to quickly and accurately grasp changes in employee emotions and anomalies in work activities and to suggest appropriate countermeasures.

[0338] "Means for collecting employee information" refers to a device or program that has the function of automatically collecting and storing data related to individual employees within an organization.

[0339] "Means for identifying data related to the efficiency of job activities" refers to methods or devices for extracting and analyzing work efficiency and performance indicators from collected employee information.

[0340] A "language analysis model" is an algorithm or program that uses natural language processing technology to analyze text data and interpret emotions and intentions.

[0341] A "learning algorithm" is a computational method or program that uses machine learning to recognize patterns in data and detect anomalies and trends.

[0342] "Deep learning technology" is an advanced data analysis technique that uses neural networks, aiming to automatically analyze complex data structures and make integrated decisions.

[0343] A "device that analyzes user emotions in real time" refers to hardware and software that instantly collects user facial expressions and voice data and uses this data to score their emotional state.

[0344] A "countermeasure recommendation system" is an algorithm or program that proposes the most suitable action plan or support measures for employees identified as high-risk.

[0345] To implement this invention, a server-centered data processing system is first required. The server first centrally collects employee information, including productivity indicators and attendance data. The server can use general database management software to store this data in a central database.

[0346] Next, the server uses the collected data to analyze the efficiency of work activities and employee sentiment. For language analysis models, general models applying natural language processing techniques (e.g., BERT or GPT) can be used. This quantifies sentiment from survey comments and forms a sentiment score. For the learning algorithm, machine learning techniques are utilized, and models for detecting anomalies (e.g., Random Forest or SVM) are used. For deep learning techniques, neural networks using TensorFlow or PyTorch are effective. This allows for the integration of sentiment analysis results and anomaly detection results, enabling risk identification.

[0347] To understand the user's emotions in real time, the device collects data via its camera and microphone. An emotion engine analyzes this data and instantly updates the emotion score. This data is then sent to a server where it is integrated with other results.

[0348] For example, by inputting a prompt such as, "Please suggest specific actions to address declining employee motivation," into the AI ​​model, appropriate countermeasures can be obtained. These suggestions, as support measures for high-risk employees, are output using the suggestion engine and provided to HR personnel via their terminals.

[0349] In this way, the collaboration between servers, terminals, and users enables comprehensive management of employee status and facilitates quick and appropriate responses.

[0350] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0351] Step 1:

[0352] The server collects employee information. Inputs include productivity metrics, attendance data, and engagement survey feedback. This data is automatically collected and stored in a central database in a structured format. The output is a unified dataset. Standardizing and storing inputs from each data source ensures consistency for subsequent data analysis.

[0353] Step 2:

[0354] The server uses natural language processing techniques to analyze employee sentiment. The input is comments from an engagement survey. The analysis uses language processing models (e.g., BERT or GPT) to extract sentiment from each comment. The output is a sentiment score, such as negative, positive, or neutral. This allows for the quantification of each employee's emotional state.

[0355] Step 3:

[0356] The server uses machine learning techniques to detect anomalies. Inputs include productivity metrics and attendance data. This data is analyzed using learning algorithms (e.g., Random Forest or SVM) to detect anomalous patterns. The output is the data points where anomalies were detected. This allows for the early detection of abnormal behavior and sudden changes.

[0357] Step 4:

[0358] The device analyzes the user's emotions in real time. As input, the user's facial expressions and voice data are collected via the camera and microphone. An emotion engine analyzes this data and instantly scores the emotional state. This data is sent to a server and integrated with other results. The output is the user's current emotion score. Tracking changes in real time allows for more accurate emotion assessment.

[0359] Step 5:

[0360] The server integrates the results using deep learning technology. Inputs include sentiment scores and anomaly detection results. Deep learning techniques (e.g., TensorFlow or PyTorch) are used to integrate and analyze this data, identifying high-risk employees. The output is a list of employees identified as high-risk. This allows for a clear prioritization of those requiring action.

[0361] Step 6:

[0362] The server generates countermeasures and notifies the user via the terminal. The input is information on high-risk employees. The suggestion engine uses a generation AI model to create specific countermeasures. The output may include countermeasures such as "providing mental support" or "proposing flexible working hours." This notification, received on the terminal, allows HR personnel to respond quickly.

[0363] (Application Example 2)

[0364] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0365] A decline in employee work efficiency and motivation at a logistics center is a serious problem that has a significant impact on overall productivity. In such an environment, it is necessary to understand employees' emotional states and work efficiency in real time and to quickly provide appropriate improvement measures. However, conventional methods have made it difficult to effectively analyze such complex human emotional and behavioral data and propose countermeasures.

[0366] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0367] In this invention, the server includes means for collecting employee attributes, means for identifying information related to work efficiency from the collected employee attributes, means for analyzing employee emotions using natural language processing technology, means for detecting anomalies in work efficiency using machine learning technology, means for comprehensively analyzing the emotion analysis results and anomaly detection results using deep learning technology, means for recommending countermeasures for high-risk employees, and means for providing a user interface for visualizing real-time emotion and productivity data. This enables real-time monitoring of employee emotions and work efficiency within a logistics center, and allows for the provision of rapid and accurate support measures.

[0368] "Employee attributes" is a general term for personal information and work performance information related to employees.

[0369] "Work efficiency" refers to the proportion and quality of work results achieved by employees within a certain time frame.

[0370] "Natural language processing technology" is a general term for technologies used by computers to understand and analyze human language.

[0371] "Machine learning technology" is a technique that enables computers to learn from data and develop the ability to identify patterns and anomalies.

[0372] "Deep learning technology" is a technique that uses a multi-layered neural network to automatically extract sophisticated features from large-scale data.

[0373] A "high-risk employee" refers to an employee whose work performance or emotional state falls outside the standard, potentially requiring special measures or support.

[0374] "Means of recommending countermeasures" refers to methods for presenting appropriate support measures and improvement measures for high-risk employees.

[0375] A "user interface" refers to an interface that provides a visual or manipulative means for a system to interact with a user and exchange information.

[0376] The system implementing this invention aims to monitor employee emotions and work efficiency in real time at a logistics center and to suggest appropriate countermeasures to high-risk employees. To achieve this objective, the system includes the following configuration.

[0377] The server automatically collects employee attributes, such as the labor resources to which each employee belongs and attendance records. This involves using sensors and RFID readers as hardware, and Python as the software. The collected employee data is consistently stored in a central database.

[0378] Next, the server uses a generative AI model to perform natural language processing and analyze emotions from text data such as employee feedback and comments. During this process, the emotion engine calculates an emotion score and reveals the emotional state.

[0379] Furthermore, the server utilizes machine learning techniques to detect anomalies in the collected work efficiency data. Deep learning methods such as TensorFlow integrate sentiment analysis results and anomaly detection results to identify employees deemed high-risk.

[0380] The data visualized via the terminal is provided to the user in real time, and the recommendation engine generates concrete measures to support countermeasures for high-risk employees. This process supports rapid and efficient business collaboration.

[0381] For example, if a particular employee leaves feedback stating that their work motivation has recently decreased, the system will analyze that comment and recommend essential support measures such as "conducting individual interviews" or "flexibly adjusting working hours."

[0382] An example of a prompt to input into the generating AI model is: "Analyze the employee's sentiment score from the following data and suggest necessary support measures: {Employee comments, attendance records, sentiment data}."

[0383] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0384] Step 1:

[0385] The server collects employee attributes such as labor resources and attendance records from sensors and attendance management systems placed within the logistics center. Inputs include employee fingerprint data and RFID tag information, and output is that this data is stored in a central database. Data collection enables the visualization of basic employee information.

[0386] Step 2:

[0387] The server analyzes text data using a generative AI model based on collected employee data. Input includes employee daily report comments and feedback. Natural language processing techniques are used to analyze emotions from this text data. The output provides an analyzed emotion score and its trends. This analysis makes it possible to predict employee emotional fluctuations.

[0388] Step 3:

[0389] The server uses machine learning techniques to validate collected work efficiency data and detect anomalies. Inputs include employee working hours, work progress, and the number of work errors. Through data analysis, patterned anomalies are identified, and the presence and frequency of these anomalies are clearly indicated as output. This step allows for the proactive detection of potential employee problems.

[0390] Step 4:

[0391] The server uses deep learning technology to comprehensively analyze sentiment analysis results and anomaly detection results to identify high-risk employees. The inputs are sentiment scores and anomaly detection results. The output of the integrated analysis is a risk assessment, indicating the likelihood of high risk. This enables the early detection of employees who require attention.

[0392] Step 5:

[0393] Users view real-time data using the user interface on their device and receive suggestions from the recommendation engine. Inputs include analysis results and recommendation results. Outputs display the suggested content, serving as a source of information for developing concrete action plans. In this step, more specific support measures are formulated for high-risk employees.

[0394] Step 6:

[0395] For example, the terminal receives feedback from a specific employee, such as "Recently, their work motivation has decreased," and passes this feedback to a recommendation engine, which then presents the analysis results to the user. In this process, feedback and historical data are submitted as input, and the output is an optimal solution based on machine learning. This allows for the rapid provision of appropriate support measures to employees.

[0396] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0397] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0398] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0399] [Third Embodiment]

[0400] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0401] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0402] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0403] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0404] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0405] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0406] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0407] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0408] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0409] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0410] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0411] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0412] This invention provides a system for improving employee engagement, and uses multiple modules to collect and analyze employee data.

[0413] First, the server automatically collects employee information from multiple systems. This information includes engagement survey results, productivity data, attendance data, and more. The collected data is stored in a central database and saved in a consistent format.

[0414] Next, the server uses a natural language processing model to analyze the text data obtained from the engagement survey. In this process, for example, it identifies negative emotions from an employee's comment such as, "The atmosphere at work hasn't been good lately." This emotion analysis allows for the evaluation of each employee's satisfaction level and stress level.

[0415] Furthermore, the server uses machine learning models to analyze productivity and attendance data, comparing them to past patterns to detect unusual activity. This anomaly detection allows for quick identification of situations such as an employee suddenly becoming more late or a decline in work efficiency.

[0416] These analysis results are integrated by a deep learning model, and the server identifies high-risk employees. High-risk employees are those who show negative results in both sentiment analysis and anomaly detection, and require special attention.

[0417] Finally, the server uses a recommendation engine to generate specific countermeasures for this high-risk employee. For example, recommendations such as "consult with the mental health support team" or "consider reassigning duties" can be provided to HR personnel via their terminals, enabling a quick response.

[0418] This system allows users to effectively and efficiently reduce employee turnover and improve employee engagement.

[0419] The following describes the processing flow.

[0420] Step 1:

[0421] The server automatically collects employee information from various systems. This information includes engagement survey results, productivity data, attendance data, etc., and the server stores it in a central database. During data collection, the server verifies the integrity of the data format and performs format conversion as needed.

[0422] Step 2:

[0423] The server inputs the text data from the engagement survey into a natural language processing model. This model analyzes the emotions contained in the survey comments and classifies each comment as "positive," "negative," or "neutral." The server then aggregates the analysis results and calculates an individual employee's emotion score.

[0424] Step 3:

[0425] The server inputs productivity and attendance data into a machine learning model. This model detects anomalous patterns by comparing them with historical data, for example, identifying sudden drops in productivity or anomalies in attendance. The server saves these results as an anomaly score for later analysis.

[0426] Step 4:

[0427] The server inputs the sentiment analysis results and anomaly detection results into a deep learning model, integrating them to identify high-risk employees. This analysis selects employees with high sentiment scores and anomaly scores, and the server generates this information as a report.

[0428] Step 5:

[0429] The server uses a recommendation engine to suggest countermeasures for high-risk employees. In this process, the server selects specific actions, such as "reducing workload" or "providing counseling," and displays the recommendations on the employee's device. HR personnel can receive this information via the device and use it to implement appropriate responses.

[0430] (Example 1)

[0431] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0432] In today's work environment, declining employee engagement and increasing stress are significant problems. These factors lead to higher turnover rates and impact corporate productivity, making early detection and countermeasures crucial. A system is needed that efficiently analyzes changes in employees' emotional states and work efficiency, and proposes appropriate responses to high-risk employees.

[0433] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0434] In this invention, the server includes means for collecting employee attributes, means for identifying information related to work efficiency from the collected employee attributes, means for analyzing employee sentiment using natural language processing technology, means for detecting anomalies in work efficiency information using machine learning methods, means for comprehensively analyzing the sentiment analysis results and anomaly detection results using deep learning methods, and means for recommending countermeasures for high-risk employees. This makes it possible to quickly grasp the status of employees and provide appropriate countermeasures, thereby reducing employee turnover and improving overall engagement within the company.

[0435] "Employee attributes" refer to information about individual employees, including data such as job duties, work efficiency, productivity, and attendance.

[0436] "Business efficiency" is an indicator that shows the results and efficiency of employees when performing their tasks.

[0437] "Natural language processing technology" is a technique that enables computers to understand and analyze human language, with the aim of analyzing the emotions and intentions behind text.

[0438] "Machine learning techniques" are technical means that allow computers to automatically learn patterns using large amounts of data and make predictions and judgments.

[0439] "Deep learning techniques" are technologies that use multi-layered neural networks to analyze complex data and provide more accurate analysis results.

[0440] A "high-risk employee" refers to an employee who shows negative results in sentiment analysis or anomaly detection and requires special attention.

[0441] A "means for recommending countermeasures" is a mechanism for proposing the optimal solution to an identified problem.

[0442] A "recommendation system" is an algorithm that suggests the optimal option based on a user's past behavior and tendencies.

[0443] This invention provides a system where a server effectively collects, analyzes, and recommends employee data. The server collects employee attributes from multiple data sources via APIs and database connections. This includes employee productivity information and attendance information. Each piece of information is retrieved via general-purpose software such as Microsoft Excel or Google Sheets and stored in a central database in a consistent format.

[0444] Regarding the collected attributes, the server uses natural language processing techniques to analyze the sentiment from employee comments and feedback. In this process, open-source language models such as BERT and GPT are utilized to extract sentiments such as positive, negative, and neutral from the data. For example, a negative sentiment is identified from a comment such as "I'm feeling stressed out by the recent project."

[0445] Next, the server uses machine learning techniques to analyze operational efficiency information. Specifically, it detects abnormal patterns such as increased tardiness or absenteeism, and decreased productivity. At this time, algorithms such as random forests and LSTMs are implemented to compare past data with current data and identify anomalies.

[0446] Furthermore, the sentiment analysis results and anomaly detection results are integrated and analyzed using deep learning techniques to identify high-risk employees. This method allows the server to assess the severity of each employee's problems and, if necessary, notify the HR department.

[0447] Ultimately, the server recommends appropriate countermeasures for high-risk employees. This process utilizes a recommendation system that presents optimal solutions based on past success stories. These include specific suggestions such as "guiding employees to mental health support" or "revising their work content."

[0448] As a concrete example, the following prompt statement can be used.

[0449] "Please perform a sentiment analysis based on the following employee comment and identify any negative emotions. Comment: 'I'm feeling stressed out by the recent project.'"

[0450] "Detect anomalies from employee data and compare them to past patterns to identify unusual behavior. Data: Late arrivals: 5 times / month, Productivity score: 70%"

[0451] Through this system, users can improve employee engagement and reduce employee turnover.

[0452] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0453] Step 1:

[0454] The server uses APIs and database connections to collect employee productivity and attendance information from multiple systems. This includes data acquisition from employee management systems and productivity management tools. Since input data may be provided in multiple formats, the server performs data conversion processing and stores these in a consistent format in a central database. The output is a format-converted dataset, which is then used for further analysis.

[0455] Step 2:

[0456] The server retrieves text data from a central database and applies natural language processing techniques to perform sentiment analysis. Specifically, it inputs the text data into a generative AI model such as BERT, which outputs a sentiment score that includes sentiment labels (positive, negative, neutral). This process identifies keywords within each comment and evaluates the direction of the sentiment to analyze each employee's satisfaction and stress level.

[0457] Step 3:

[0458] The server utilizes machine learning techniques to analyze operational efficiency information. It uses historical attendance data and productivity scores as input to perform anomaly detection. Random forest and LSTM models are used to identify unusual patterns. The output consists of data points flagged as anomalies, which the server uses to determine what is normal and what is abnormal.

[0459] Step 4:

[0460] The server applies deep learning techniques to integrate sentiment analysis results and anomaly detection results. This calculates an overall risk score for each employee. In this process, individual scores are used as input for a weighted, integrated analysis, and an integrated risk assessment is generated as output. This is used to identify high-risk employees.

[0461] Step 5:

[0462] The server uses a recommendation engine to generate countermeasures for employees identified as high-risk. It utilizes an integrated risk assessment as input and outputs optimal suggestions using a model trained on past successes. Specifically, it might present concrete countermeasures to the HR department, such as "introduction to mental health support" or "consideration of changes to job duties."

[0463] (Application Example 1)

[0464] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0465] Traditional employee engagement systems struggled to provide real-time insights into employee and machine performance, making it difficult to quickly implement appropriate countermeasures. Furthermore, they lacked systems that integrated sentiment analysis and anomaly detection, instead relying solely on individual analyses. As a result, overall organizational productivity improvements and optimal collaboration between machines and humans were not achieved.

[0466] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0467] In this invention, the server includes means for collecting employee information and machine information, means for identifying data related to work efficiency from the collected employee information and machine information, and means for analyzing employee sentiment and feedback from workers using a natural language processing model. This enables integrated management of humans and machines within the organization, leading to improved productivity and proper engagement management.

[0468] "Employee information" refers to data about employees' attributes, behavior, and work, including productivity data and attendance data.

[0469] "Machine information" refers to data related to the operating status and performance of equipment, including operational data collected through sensors.

[0470] "Work efficiency" is an indicator that shows the performance level of tasks and operations performed by employees and machines.

[0471] A "natural language processing model" is a collection of algorithms and methods for processing and analyzing human language using computers.

[0472] "Emotional and feedback analysis" is a process of analyzing employees' emotions and opinions from text data and using that information to improve the organization.

[0473] A "machine learning model" is a mathematical model that learns patterns from data and uses them for prediction and classification.

[0474] Anomaly detection is the process of finding patterns or behaviors that are different from the norm within data.

[0475] A "deep learning model" is a technique that uses a neural network, which has a structure similar to the human brain, to learn complex patterns.

[0476] "Measures for high-risk employees and equipment" refer to proposed improvements and countermeasures for employees or equipment that pose risks beyond normal operations.

[0477] A "recommendation engine" is a system that provides individually optimized suggestions and actions based on user data and context.

[0478] This invention provides a system that improves overall organizational work efficiency and engagement by efficiently collecting and analyzing employee and machine operation data. The server first automatically collects employee and machine information from various sensors and IoT devices installed in the factory and workplace environment. This information, including productivity data, attendance data, and equipment operation data, is stored in a central database in a consistent format.

[0479] The server utilizes natural language processing technology to analyze text-based feedback provided by employees and operators. This allows for the extraction and evaluation of employees' emotional states and opinions on the work environment. OpenAI's language models are used for natural language processing.

[0480] Furthermore, the server uses machine learning models to detect anomalies in the collected work efficiency data. This process identifies patterns that deviate from the normal operation of employees and equipment, aiming for early detection of risks. Suitable libraries for this purpose include scikit-learn.

[0481] The results of both anomaly detection and sentiment analysis are integrated and analyzed by a deep learning model. At this stage, TensorFlow is used to learn the complex relationships between the data and derive effective improvement strategies.

[0482] For employees and equipment identified as high-risk, a recommendation engine will provide specific countermeasures. Based on past analysis results, this engine proposes the optimal action to help improve organizational efficiency.

[0483] For example, if an operator inputs feedback such as "Recently, work has been falling behind schedule," the system will respond to that feedback and efficiently make suggestions. An example of a prompt using a generative AI model is: "Perform sentiment analysis on the operator's feedback and generate recommendations for improving the environment. Input feedback: 'Recently, work has been falling behind schedule.'"

[0484] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0485] Step 1:

[0486] The server collects employee and machine information from IoT devices and sensors installed in factories and workplaces. This information includes productivity data, attendance data, and equipment operation data. This data is aggregated in real time and stored in a central database in a consistent format.

[0487] Step 2:

[0488] The server applies natural language processing to the employee feedback it collects. It analyzes the text-based feedback entered by users and extracts opinions on emotional states and work performance. This analysis uses a generative AI model to evaluate the positive and negative aspects of emotions and opinions in particular. The input is text data, and the output is an emotional score and keywords.

[0489] Step 3:

[0490] The server applies machine learning models to collected work efficiency data and equipment operation data to perform anomaly detection. The input is raw numerical data, and statistical methods and machine learning algorithms are used to detect deviations from normal patterns. The output provides the anomaly detection results and their associated parameters.

[0491] Step 4:

[0492] The server integrates the results of sentiment analysis and anomaly detection using a deep learning model. This model learns the complex relationships between data and identifies risk patterns. The input is the sentiment score and anomaly detection results obtained in the previous step, and the output is data for risk assessment and improvement suggestions.

[0493] Step 5:

[0494] The device generates specific countermeasures for employees and equipment identified as high-risk. Using a recommendation engine, optimal actions are suggested and presented to the user based on past analysis results. The input is the result of integrated analysis, and the output is specific action suggestions and action plans.

[0495] Step 6:

[0496] Users evaluate the proposed solutions and select actionable steps. This improves on-site work efficiency and employee engagement, and enables immediate corrective measures.

[0497] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0498] This invention is a system for improving employee engagement, combining an emotion engine with various analytical models to achieve more advanced data analysis.

[0499] First, the server automatically collects employee information. This information includes productivity data, attendance data, and feedback from engagement surveys. This data is stored consistently in a central database.

[0500] Next, the server uses a natural language processing model to analyze the emotions expressed in the survey comments. For example, it detects negative emotions from a comment such as, "My motivation at work has been low lately." This result is then aggregated as an emotion score for each individual employee.

[0501] Furthermore, the server utilizes machine learning models to analyze productivity and attendance data to detect anomalies. For example, it might detect an anomaly if an employee unexpectedly starts being absent frequently.

[0502] In addition to these analysis processes, an emotion engine is introduced to analyze the user's emotions in real time from their facial expressions and voice. The data provided by the emotion engine is integrated with other emotion analysis results, enabling more comprehensive analysis based on richer information.

[0503] A deep learning model integrates sentiment analysis results and anomaly detection results to accurately identify high-risk employees. Based on this, the server uses a recommendation engine to suggest appropriate countermeasures for high-risk employees.

[0504] For example, if an employee is identified whose emotional score has decreased and whose work efficiency has also declined, specific actions such as "providing mental support" or "proposing flexible working hours" can be taken. These recommendations are provided to HR personnel via their devices, enabling quick decision-making and response.

[0505] In this way, the system of the present invention achieves more accurate employee support through a comprehensive analysis that takes user emotions into account.

[0506] The following describes the processing flow.

[0507] Step 1:

[0508] The server automatically collects employee information from each system. This includes a wide range of data, such as productivity data, attendance data, and engagement survey results. The server stores the collected information in a central database and standardizes the format to maintain data consistency.

[0509] Step 2:

[0510] The server uses a natural language processing model to analyze the text data from the engagement survey. This analysis extracts emotions from each comment and assigns them emotion labels such as "positive," "negative," or "neutral." The server then aggregates these results to calculate an emotion score for each employee.

[0511] Step 3:

[0512] The server feeds the collected productivity and attendance data into a machine learning model. This model detects anomalies by comparing them to past data patterns, for example, identifying decreased productivity or increased tardiness for specific employees. The server records the results of anomaly detection as an anomaly score.

[0513] Step 4:

[0514] The emotion engine analyzes the user's facial expressions and voice data in real time. While the user is in a video conference or call, the emotion engine processes this data, evaluates the user's emotional state, and generates a real-time emotion score.

[0515] Step 5:

[0516] The server inputs sentiment scores, anomaly scores, and real-time user sentiment scores into a deep learning model for integrated analysis. This analysis identifies high-risk employees and assesses their impact.

[0517] Step 6:

[0518] The server uses a recommendation engine to generate countermeasures for identified high-risk employees. These recommendations may include, for example, "referral to a professional counselor" or "suggestions for improving the work environment," and the server displays this information on the terminal and notifies the HR personnel.

[0519] Step 7:

[0520] HR personnel review the recommendations received on their devices, select appropriate countermeasures, and implement them. This speeds up responses within the organization and leads to improved employee engagement.

[0521] (Example 2)

[0522] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0523] For many companies, effectively monitoring employee work efficiency and emotional states, and identifying and addressing high-risk employees early on, is a challenge. While changes in employee emotions impact productivity, understanding emotions is subjective and difficult. Furthermore, early detection of abnormal behavior and the provision of effective countermeasures are crucial, but traditional methods lack real-time and multifaceted analysis. Therefore, there is a need for the development of systems that efficiently and comprehensively manage employee conditions and enable early intervention.

[0524] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0525] In this invention, the server includes means for collecting employee information, means for identifying data related to the efficiency of work activities, means for analyzing employee emotions using a language analysis model, means for detecting anomalies in work activity efficiency data using a learning algorithm, means for comprehensively analyzing emotion analysis results and anomaly detection results using deep learning technology, a device for analyzing user emotions in real time, and a device for recommending countermeasures for high-risk employees. This makes it possible to quickly and accurately grasp changes in employee emotions and anomalies in work activities and to suggest appropriate countermeasures.

[0526] "Means for collecting employee information" refers to a device or program that has the function of automatically collecting and storing data related to individual employees within an organization.

[0527] "Means for identifying data related to the efficiency of job activities" refers to methods or devices for extracting and analyzing work efficiency and performance indicators from collected employee information.

[0528] A "language analysis model" is an algorithm or program that uses natural language processing technology to analyze text data and interpret emotions and intentions.

[0529] A "learning algorithm" is a computational method or program that uses machine learning to recognize patterns in data and detect anomalies and trends.

[0530] "Deep learning technology" is an advanced data analysis technique that uses neural networks, aiming to automatically analyze complex data structures and make integrated decisions.

[0531] A "device that analyzes user emotions in real time" refers to hardware and software that instantly collects user facial expressions and voice data and uses this data to score their emotional state.

[0532] A "countermeasure recommendation system" is an algorithm or program that proposes the most suitable action plan or support measures for employees identified as high-risk.

[0533] To implement this invention, a server-centered data processing system is first required. The server first centrally collects employee information, including productivity indicators and attendance data. The server can use general database management software to store this data in a central database.

[0534] Next, the server uses the collected data to analyze the efficiency of work activities and employee sentiment. For language analysis models, general models applying natural language processing techniques (e.g., BERT or GPT) can be used. This quantifies sentiment from survey comments and forms a sentiment score. For the learning algorithm, machine learning techniques are utilized, and models for detecting anomalies (e.g., Random Forest or SVM) are used. For deep learning techniques, neural networks using TensorFlow or PyTorch are effective. This allows for the integration of sentiment analysis results and anomaly detection results, enabling risk identification.

[0535] To understand the user's emotions in real time, the device collects data via its camera and microphone. An emotion engine analyzes this data and instantly updates the emotion score. This data is then sent to a server where it is integrated with other results.

[0536] For example, by inputting a prompt such as, "Please suggest specific actions to address declining employee motivation," into the AI ​​model, appropriate countermeasures can be obtained. These suggestions, as support measures for high-risk employees, are output using the suggestion engine and provided to HR personnel via their terminals.

[0537] In this way, the collaboration between servers, terminals, and users enables comprehensive management of employee status and facilitates quick and appropriate responses.

[0538] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0539] Step 1:

[0540] The server collects employee information. Inputs include productivity metrics, attendance data, and engagement survey feedback. This data is automatically collected and stored in a central database in a structured format. The output is a unified dataset. Standardizing and storing inputs from each data source ensures consistency for subsequent data analysis.

[0541] Step 2:

[0542] The server uses natural language processing techniques to analyze employee sentiment. The input is comments from an engagement survey. The analysis uses language processing models (e.g., BERT or GPT) to extract sentiment from each comment. The output is a sentiment score, such as negative, positive, or neutral. This allows for the quantification of each employee's emotional state.

[0543] Step 3:

[0544] The server uses machine learning techniques to detect anomalies. Inputs include productivity metrics and attendance data. This data is analyzed using learning algorithms (e.g., Random Forest or SVM) to detect anomalous patterns. The output is the data points where anomalies were detected. This allows for the early detection of abnormal behavior and sudden changes.

[0545] Step 4:

[0546] The device analyzes the user's emotions in real time. As input, the user's facial expressions and voice data are collected via the camera and microphone. An emotion engine analyzes this data and instantly scores the emotional state. This data is sent to a server and integrated with other results. The output is the user's current emotion score. Tracking changes in real time allows for more accurate emotion assessment.

[0547] Step 5:

[0548] The server integrates the results using deep learning technology. Inputs include sentiment scores and anomaly detection results. Deep learning techniques (e.g., TensorFlow or PyTorch) are used to integrate and analyze this data, identifying high-risk employees. The output is a list of employees identified as high-risk. This allows for a clear prioritization of those requiring action.

[0549] Step 6:

[0550] The server generates countermeasures and notifies the user via the terminal. The input is information on high-risk employees. The suggestion engine uses a generation AI model to create specific countermeasures. The output may include countermeasures such as "providing mental support" or "proposing flexible working hours." This notification, received on the terminal, allows HR personnel to respond quickly.

[0551] (Application Example 2)

[0552] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0553] A decline in employee work efficiency and motivation at a logistics center is a serious problem that has a significant impact on overall productivity. In such an environment, it is necessary to understand employees' emotional states and work efficiency in real time and to quickly provide appropriate improvement measures. However, conventional methods have made it difficult to effectively analyze such complex human emotional and behavioral data and propose countermeasures.

[0554] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0555] In this invention, the server includes means for collecting employee attributes, means for identifying information related to work efficiency from the collected employee attributes, means for analyzing employee emotions using natural language processing technology, means for detecting anomalies in work efficiency using machine learning technology, means for comprehensively analyzing the emotion analysis results and anomaly detection results using deep learning technology, means for recommending countermeasures for high-risk employees, and means for providing a user interface for visualizing real-time emotion and productivity data. This enables real-time monitoring of employee emotions and work efficiency within a logistics center, and allows for the provision of rapid and accurate support measures.

[0556] "Employee attributes" is a general term for personal information and work performance information related to employees.

[0557] "Work efficiency" refers to the proportion and quality of work results achieved by employees within a certain time frame.

[0558] "Natural language processing technology" is a general term for technologies used by computers to understand and analyze human language.

[0559] "Machine learning technology" is a technique that enables computers to learn from data and develop the ability to identify patterns and anomalies.

[0560] "Deep learning technology" is a technique that uses a multi-layered neural network to automatically extract sophisticated features from large-scale data.

[0561] A "high-risk employee" refers to an employee whose work performance or emotional state falls outside the standard, potentially requiring special measures or support.

[0562] "Means of recommending countermeasures" refers to methods for presenting appropriate support measures and improvement measures for high-risk employees.

[0563] A "user interface" refers to an interface that provides a visual or manipulative means for a system to interact with a user and exchange information.

[0564] The system implementing this invention aims to monitor employee emotions and work efficiency in real time at a logistics center and to suggest appropriate countermeasures to high-risk employees. To achieve this objective, the system includes the following configuration.

[0565] The server automatically collects employee attributes, such as the labor resources to which each employee belongs and attendance records. This involves using sensors and RFID readers as hardware, and Python as the software. The collected employee data is consistently stored in a central database.

[0566] Next, the server uses a generative AI model to perform natural language processing and analyze emotions from text data such as employee feedback and comments. During this process, the emotion engine calculates an emotion score and reveals the emotional state.

[0567] Furthermore, the server utilizes machine learning techniques to detect anomalies in the collected work efficiency data. Deep learning methods such as TensorFlow integrate sentiment analysis results and anomaly detection results to identify employees deemed high-risk.

[0568] The data visualized via the terminal is provided to the user in real time, and the recommendation engine generates concrete measures to support countermeasures for high-risk employees. This process supports rapid and efficient business collaboration.

[0569] For example, if a particular employee leaves feedback stating that their work motivation has recently decreased, the system will analyze that comment and recommend essential support measures such as "conducting individual interviews" or "flexibly adjusting working hours."

[0570] An example of a prompt to input into the generating AI model is: "Analyze the employee's sentiment score from the following data and suggest necessary support measures: {Employee comments, attendance records, sentiment data}."

[0571] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0572] Step 1:

[0573] The server collects employee attributes such as labor resources and attendance records from sensors and attendance management systems placed within the logistics center. Inputs include employee fingerprint data and RFID tag information, and output is that this data is stored in a central database. Data collection enables the visualization of basic employee information.

[0574] Step 2:

[0575] The server analyzes text data using a generative AI model based on collected employee data. Input includes employee daily report comments and feedback. Natural language processing techniques are used to analyze emotions from this text data. The output provides an analyzed emotion score and its trends. This analysis makes it possible to predict employee emotional fluctuations.

[0576] Step 3:

[0577] The server uses machine learning techniques to validate collected work efficiency data and detect anomalies. Inputs include employee working hours, work progress, and the number of work errors. Through data analysis, patterned anomalies are identified, and the presence and frequency of these anomalies are clearly indicated as output. This step allows for the proactive detection of potential employee problems.

[0578] Step 4:

[0579] The server uses deep learning technology to comprehensively analyze sentiment analysis results and anomaly detection results to identify high-risk employees. The inputs are sentiment scores and anomaly detection results. The output of the integrated analysis is a risk assessment, indicating the likelihood of high risk. This enables the early detection of employees who require attention.

[0580] Step 5:

[0581] Users view real-time data using the user interface on their device and receive suggestions from the recommendation engine. Inputs include analysis results and recommendation results. Outputs display the suggested content, serving as a source of information for developing concrete action plans. In this step, more specific support measures are formulated for high-risk employees.

[0582] Step 6:

[0583] For example, the terminal receives feedback from a specific employee, such as "Recently, their work motivation has decreased," and passes this feedback to a recommendation engine, which then presents the analysis results to the user. In this process, feedback and historical data are submitted as input, and the output is an optimal solution based on machine learning. This allows for the rapid provision of appropriate support measures to employees.

[0584] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0585] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0586] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0587] [Fourth Embodiment]

[0588] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0589] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0590] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0591] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0592] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0593] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0594] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0595] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0596] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0597] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0598] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0599] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0600] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0601] This invention provides a system for improving employee engagement, and uses multiple modules to collect and analyze employee data.

[0602] First, the server automatically collects employee information from multiple systems. This information includes engagement survey results, productivity data, attendance data, and more. The collected data is stored in a central database and saved in a consistent format.

[0603] Next, the server uses a natural language processing model to analyze the text data obtained from the engagement survey. In this process, for example, it identifies negative emotions from an employee's comment such as, "The atmosphere at work hasn't been good lately." This emotion analysis allows for the evaluation of each employee's satisfaction level and stress level.

[0604] Furthermore, the server uses machine learning models to analyze productivity and attendance data, comparing them to past patterns to detect unusual activity. This anomaly detection allows for quick identification of situations such as an employee suddenly becoming more late or a decline in work efficiency.

[0605] These analysis results are integrated by a deep learning model, and the server identifies high-risk employees. High-risk employees are those who show negative results in both sentiment analysis and anomaly detection, and require special attention.

[0606] Finally, the server uses a recommendation engine to generate specific countermeasures for this high-risk employee. For example, recommendations such as "consult with the mental health support team" or "consider reassigning duties" can be provided to HR personnel via their terminals, enabling a quick response.

[0607] This system allows users to effectively and efficiently reduce employee turnover and improve employee engagement.

[0608] The following describes the processing flow.

[0609] Step 1:

[0610] The server automatically collects employee information from various systems. This information includes engagement survey results, productivity data, attendance data, etc., and the server stores it in a central database. During data collection, the server verifies the integrity of the data format and performs format conversion as needed.

[0611] Step 2:

[0612] The server inputs the text data from the engagement survey into a natural language processing model. This model analyzes the emotions contained in the survey comments and classifies each comment as "positive," "negative," or "neutral." The server then aggregates the analysis results and calculates an individual employee's emotion score.

[0613] Step 3:

[0614] The server inputs productivity and attendance data into a machine learning model. This model detects anomalous patterns by comparing them with historical data, for example, identifying sudden drops in productivity or anomalies in attendance. The server saves these results as an anomaly score for later analysis.

[0615] Step 4:

[0616] The server inputs the sentiment analysis results and anomaly detection results into a deep learning model, integrating them to identify high-risk employees. This analysis selects employees with high sentiment scores and anomaly scores, and the server generates this information as a report.

[0617] Step 5:

[0618] The server uses a recommendation engine to suggest countermeasures for high-risk employees. In this process, the server selects specific actions, such as "reducing workload" or "providing counseling," and displays the recommendations on the employee's device. HR personnel can receive this information via the device and use it to implement appropriate responses.

[0619] (Example 1)

[0620] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0621] In today's work environment, declining employee engagement and increasing stress are significant problems. These factors lead to higher turnover rates and impact corporate productivity, making early detection and countermeasures crucial. A system is needed that efficiently analyzes changes in employees' emotional states and work efficiency, and proposes appropriate responses to high-risk employees.

[0622] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0623] In this invention, the server includes means for collecting employee attributes, means for identifying information related to work efficiency from the collected employee attributes, means for analyzing employee sentiment using natural language processing technology, means for detecting anomalies in work efficiency information using machine learning methods, means for comprehensively analyzing the sentiment analysis results and anomaly detection results using deep learning methods, and means for recommending countermeasures for high-risk employees. This makes it possible to quickly grasp the status of employees and provide appropriate countermeasures, thereby reducing employee turnover and improving overall engagement within the company.

[0624] "Employee attributes" refer to information about individual employees, including data such as job duties, work efficiency, productivity, and attendance.

[0625] "Business efficiency" is an indicator that shows the results and efficiency of employees when performing their tasks.

[0626] "Natural language processing technology" is a technique that enables computers to understand and analyze human language, with the aim of analyzing the emotions and intentions behind text.

[0627] "Machine learning techniques" are technical means that allow computers to automatically learn patterns using large amounts of data and make predictions and judgments.

[0628] "Deep learning techniques" are technologies that use multi-layered neural networks to analyze complex data and provide more accurate analysis results.

[0629] A "high-risk employee" refers to an employee who shows negative results in sentiment analysis or anomaly detection and requires special attention.

[0630] A "means for recommending countermeasures" is a mechanism for proposing the optimal solution to an identified problem.

[0631] A "recommendation system" is an algorithm that suggests the optimal option based on a user's past behavior and tendencies.

[0632] This invention provides a system where a server effectively collects, analyzes, and recommends employee data. The server collects employee attributes from multiple data sources via APIs and database connections. This includes employee productivity information and attendance information. Each piece of information is retrieved via general-purpose software such as Microsoft Excel or Google Sheets and stored in a central database in a consistent format.

[0633] Regarding the collected attributes, the server uses natural language processing techniques to analyze the sentiment from employee comments and feedback. In this process, open-source language models such as BERT and GPT are utilized to extract sentiments such as positive, negative, and neutral from the data. For example, a negative sentiment is identified from a comment such as "I'm feeling stressed out by the recent project."

[0634] Next, the server uses machine learning techniques to analyze operational efficiency information. Specifically, it detects abnormal patterns such as increased tardiness or absenteeism, and decreased productivity. At this time, algorithms such as random forests and LSTMs are implemented to compare past data with current data and identify anomalies.

[0635] Furthermore, the sentiment analysis results and anomaly detection results are integrated and analyzed using deep learning techniques to identify high-risk employees. This method allows the server to assess the severity of each employee's problems and, if necessary, notify the HR department.

[0636] Ultimately, the server recommends appropriate countermeasures for high-risk employees. This process utilizes a recommendation system that presents optimal solutions based on past success stories. These include specific suggestions such as "guiding employees to mental health support" or "revising their work content."

[0637] As a concrete example, the following prompt statement can be used.

[0638] "Please perform a sentiment analysis based on the following employee comment and identify any negative emotions. Comment: 'I'm feeling stressed out by the recent project.'"

[0639] "Detect anomalies from employee data and compare them to past patterns to identify unusual behavior. Data: Late arrivals: 5 times / month, Productivity score: 70%"

[0640] Through this system, users can improve employee engagement and reduce employee turnover.

[0641] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0642] Step 1:

[0643] The server uses APIs and database connections to collect employee productivity and attendance information from multiple systems. This includes data acquisition from employee management systems and productivity management tools. Since input data may be provided in multiple formats, the server performs data conversion processing and stores these in a consistent format in a central database. The output is a format-converted dataset, which is then used for further analysis.

[0644] Step 2:

[0645] The server retrieves text data from a central database and applies natural language processing techniques to perform sentiment analysis. Specifically, it inputs the text data into a generative AI model such as BERT, which outputs a sentiment score that includes sentiment labels (positive, negative, neutral). This process identifies keywords within each comment and evaluates the direction of the sentiment to analyze each employee's satisfaction and stress level.

[0646] Step 3:

[0647] The server utilizes machine learning techniques to analyze operational efficiency information. It uses historical attendance data and productivity scores as input to perform anomaly detection. Random forest and LSTM models are used to identify unusual patterns. The output consists of data points flagged as anomalies, which the server uses to determine what is normal and what is abnormal.

[0648] Step 4:

[0649] The server applies deep learning techniques to integrate sentiment analysis results and anomaly detection results. This calculates an overall risk score for each employee. In this process, individual scores are used as input for a weighted, integrated analysis, and an integrated risk assessment is generated as output. This is used to identify high-risk employees.

[0650] Step 5:

[0651] The server uses a recommendation engine to generate countermeasures for employees identified as high-risk. It utilizes an integrated risk assessment as input and outputs optimal suggestions using a model trained on past successes. Specifically, it might present concrete countermeasures to the HR department, such as "introduction to mental health support" or "consideration of changes to job duties."

[0652] (Application Example 1)

[0653] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0654] Traditional employee engagement systems struggled to provide real-time insights into employee and machine performance, making it difficult to quickly implement appropriate countermeasures. Furthermore, they lacked systems that integrated sentiment analysis and anomaly detection, instead relying solely on individual analyses. As a result, overall organizational productivity improvements and optimal collaboration between machines and humans were not achieved.

[0655] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0656] In this invention, the server includes means for collecting employee information and machine information, means for identifying data related to work efficiency from the collected employee information and machine information, and means for analyzing employee sentiment and feedback from workers using a natural language processing model. This enables integrated management of humans and machines within the organization, leading to improved productivity and proper engagement management.

[0657] "Employee information" refers to data about employees' attributes, behavior, and work, including productivity data and attendance data.

[0658] "Machine information" refers to data related to the operating status and performance of equipment, including operational data collected through sensors.

[0659] "Work efficiency" is an indicator that shows the performance level of tasks and operations performed by employees and machines.

[0660] A "natural language processing model" is a collection of algorithms and methods for processing and analyzing human language using computers.

[0661] "Emotional and feedback analysis" is a process of analyzing employees' emotions and opinions from text data and using that information to improve the organization.

[0662] A "machine learning model" is a mathematical model that learns patterns from data and uses them for prediction and classification.

[0663] Anomaly detection is the process of finding patterns or behaviors that are different from the norm within data.

[0664] A "deep learning model" is a technique that uses a neural network, which has a structure similar to the human brain, to learn complex patterns.

[0665] "Measures for high-risk employees and equipment" refer to proposed improvements and countermeasures for employees or equipment that pose risks beyond normal operations.

[0666] A "recommendation engine" is a system that provides individually optimized suggestions and actions based on user data and context.

[0667] This invention provides a system that improves overall organizational work efficiency and engagement by efficiently collecting and analyzing employee and machine operation data. The server first automatically collects employee and machine information from various sensors and IoT devices installed in the factory and workplace environment. This information, including productivity data, attendance data, and equipment operation data, is stored in a central database in a consistent format.

[0668] The server utilizes natural language processing technology to analyze text-based feedback provided by employees and operators. This allows for the extraction and evaluation of employees' emotional states and opinions on the work environment. OpenAI's language models are used for natural language processing.

[0669] Furthermore, the server uses machine learning models to detect anomalies in the collected work efficiency data. This process identifies patterns that deviate from the normal operation of employees and equipment, aiming for early detection of risks. Suitable libraries for this purpose include scikit-learn.

[0670] The results of both anomaly detection and sentiment analysis are integrated and analyzed by a deep learning model. At this stage, TensorFlow is used to learn the complex relationships between the data and derive effective improvement strategies.

[0671] For employees and equipment identified as high-risk, a recommendation engine will provide specific countermeasures. Based on past analysis results, this engine proposes the optimal action to help improve organizational efficiency.

[0672] For example, if an operator inputs feedback such as "Recently, work has been falling behind schedule," the system will respond to that feedback and efficiently make suggestions. An example of a prompt using a generative AI model is: "Perform sentiment analysis on the operator's feedback and generate recommendations for improving the environment. Input feedback: 'Recently, work has been falling behind schedule.'"

[0673] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0674] Step 1:

[0675] The server collects employee and machine information from IoT devices and sensors installed in factories and workplaces. This information includes productivity data, attendance data, and equipment operation data. This data is aggregated in real time and stored in a central database in a consistent format.

[0676] Step 2:

[0677] The server applies natural language processing to the employee feedback it collects. It analyzes the text-based feedback entered by users and extracts opinions on emotional states and work performance. This analysis uses a generative AI model to evaluate the positive and negative aspects of emotions and opinions in particular. The input is text data, and the output is an emotional score and keywords.

[0678] Step 3:

[0679] The server applies machine learning models to collected work efficiency data and equipment operation data to perform anomaly detection. The input is raw numerical data, and statistical methods and machine learning algorithms are used to detect deviations from normal patterns. The output provides the anomaly detection results and their associated parameters.

[0680] Step 4:

[0681] The server integrates the results of sentiment analysis and anomaly detection using a deep learning model. This model learns the complex relationships between data and identifies risk patterns. The input is the sentiment score and anomaly detection results obtained in the previous step, and the output is data for risk assessment and improvement suggestions.

[0682] Step 5:

[0683] The device generates specific countermeasures for employees and equipment identified as high-risk. Using a recommendation engine, optimal actions are suggested and presented to the user based on past analysis results. The input is the result of integrated analysis, and the output is specific action suggestions and action plans.

[0684] Step 6:

[0685] Users evaluate the proposed solutions and select actionable steps. This improves on-site work efficiency and employee engagement, and enables immediate corrective measures.

[0686] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0687] This invention is a system for improving employee engagement, combining an emotion engine with various analytical models to achieve more advanced data analysis.

[0688] First, the server automatically collects employee information. This information includes productivity data, attendance data, and feedback from engagement surveys. This data is stored consistently in a central database.

[0689] Next, the server uses a natural language processing model to analyze the emotions expressed in the survey comments. For example, it detects negative emotions from a comment such as, "My motivation at work has been low lately." This result is then aggregated as an emotion score for each individual employee.

[0690] Furthermore, the server utilizes machine learning models to analyze productivity and attendance data to detect anomalies. For example, it might detect an anomaly if an employee unexpectedly starts being absent frequently.

[0691] In addition to these analysis processes, an emotion engine is introduced to analyze the user's emotions in real time from their facial expressions and voice. The data provided by the emotion engine is integrated with other emotion analysis results, enabling more comprehensive analysis based on richer information.

[0692] A deep learning model integrates sentiment analysis results and anomaly detection results to accurately identify high-risk employees. Based on this, the server uses a recommendation engine to suggest appropriate countermeasures for high-risk employees.

[0693] For example, if an employee is identified whose emotional score has decreased and whose work efficiency has also declined, specific actions such as "providing mental support" or "proposing flexible working hours" can be taken. These recommendations are provided to HR personnel via their devices, enabling quick decision-making and response.

[0694] In this way, the system of the present invention achieves more accurate employee support through a comprehensive analysis that takes user emotions into account.

[0695] The following describes the processing flow.

[0696] Step 1:

[0697] The server automatically collects employee information from each system. This includes a wide range of data, such as productivity data, attendance data, and engagement survey results. The server stores the collected information in a central database and standardizes the format to maintain data consistency.

[0698] Step 2:

[0699] The server uses a natural language processing model to analyze the text data from the engagement survey. This analysis extracts emotions from each comment and assigns them emotion labels such as "positive," "negative," or "neutral." The server then aggregates these results to calculate an emotion score for each employee.

[0700] Step 3:

[0701] The server feeds the collected productivity and attendance data into a machine learning model. This model detects anomalies by comparing them to past data patterns, for example, identifying decreased productivity or increased tardiness for specific employees. The server records the results of anomaly detection as an anomaly score.

[0702] Step 4:

[0703] The emotion engine analyzes the user's facial expressions and voice data in real time. While the user is in a video conference or call, the emotion engine processes this data, evaluates the user's emotional state, and generates a real-time emotion score.

[0704] Step 5:

[0705] The server inputs sentiment scores, anomaly scores, and real-time user sentiment scores into a deep learning model for integrated analysis. This analysis identifies high-risk employees and assesses their impact.

[0706] Step 6:

[0707] The server uses a recommendation engine to generate countermeasures for identified high-risk employees. These recommendations may include, for example, "referral to a professional counselor" or "suggestions for improving the work environment," and the server displays this information on the terminal and notifies the HR personnel.

[0708] Step 7:

[0709] HR personnel review the recommendations received on their devices, select appropriate countermeasures, and implement them. This speeds up responses within the organization and leads to improved employee engagement.

[0710] (Example 2)

[0711] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0712] For many companies, effectively monitoring employee work efficiency and emotional states, and identifying and addressing high-risk employees early on, is a challenge. While changes in employee emotions impact productivity, understanding emotions is subjective and difficult. Furthermore, early detection of abnormal behavior and the provision of effective countermeasures are crucial, but traditional methods lack real-time and multifaceted analysis. Therefore, there is a need for the development of systems that efficiently and comprehensively manage employee conditions and enable early intervention.

[0713] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0714] In this invention, the server includes means for collecting employee information, means for identifying data related to the efficiency of work activities, means for analyzing employee emotions using a language analysis model, means for detecting anomalies in work activity efficiency data using a learning algorithm, means for comprehensively analyzing emotion analysis results and anomaly detection results using deep learning technology, a device for analyzing user emotions in real time, and a device for recommending countermeasures for high-risk employees. This makes it possible to quickly and accurately grasp changes in employee emotions and anomalies in work activities and to suggest appropriate countermeasures.

[0715] "Means for collecting employee information" refers to a device or program that has the function of automatically collecting and storing data related to individual employees within an organization.

[0716] "Means for identifying data related to the efficiency of job activities" refers to methods or devices for extracting and analyzing work efficiency and performance indicators from collected employee information.

[0717] A "language analysis model" is an algorithm or program that uses natural language processing technology to analyze text data and interpret emotions and intentions.

[0718] A "learning algorithm" is a computational method or program that uses machine learning to recognize patterns in data and detect anomalies and trends.

[0719] "Deep learning technology" is an advanced data analysis technique that uses neural networks, aiming to automatically analyze complex data structures and make integrated decisions.

[0720] A "device that analyzes user emotions in real time" refers to hardware and software that instantly collects user facial expressions and voice data and uses this data to score their emotional state.

[0721] A "countermeasure recommendation system" is an algorithm or program that proposes the most suitable action plan or support measures for employees identified as high-risk.

[0722] To implement this invention, a server-centered data processing system is first required. The server first centrally collects employee information, including productivity indicators and attendance data. The server can use general database management software to store this data in a central database.

[0723] Next, the server uses the collected data to analyze the efficiency of work activities and employee sentiment. For language analysis models, general models applying natural language processing techniques (e.g., BERT or GPT) can be used. This quantifies sentiment from survey comments and forms a sentiment score. For the learning algorithm, machine learning techniques are utilized, and models for detecting anomalies (e.g., Random Forest or SVM) are used. For deep learning techniques, neural networks using TensorFlow or PyTorch are effective. This allows for the integration of sentiment analysis results and anomaly detection results, enabling risk identification.

[0724] To understand the user's emotions in real time, the device collects data via its camera and microphone. An emotion engine analyzes this data and instantly updates the emotion score. This data is then sent to a server where it is integrated with other results.

[0725] For example, by inputting a prompt such as, "Please suggest specific actions to address declining employee motivation," into the AI ​​model, appropriate countermeasures can be obtained. These suggestions, as support measures for high-risk employees, are output using the suggestion engine and provided to HR personnel via their terminals.

[0726] In this way, the collaboration between servers, terminals, and users enables comprehensive management of employee status and facilitates quick and appropriate responses.

[0727] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0728] Step 1:

[0729] The server collects employee information. Inputs include productivity metrics, attendance data, and engagement survey feedback. This data is automatically collected and stored in a central database in a structured format. The output is a unified dataset. Standardizing and storing inputs from each data source ensures consistency for subsequent data analysis.

[0730] Step 2:

[0731] The server uses natural language processing techniques to analyze employee sentiment. The input is comments from an engagement survey. The analysis uses language processing models (e.g., BERT or GPT) to extract sentiment from each comment. The output is a sentiment score, such as negative, positive, or neutral. This allows for the quantification of each employee's emotional state.

[0732] Step 3:

[0733] The server uses machine learning techniques to detect anomalies. Inputs include productivity metrics and attendance data. This data is analyzed using learning algorithms (e.g., Random Forest or SVM) to detect anomalous patterns. The output is the data points where anomalies were detected. This allows for the early detection of abnormal behavior and sudden changes.

[0734] Step 4:

[0735] The device analyzes the user's emotions in real time. As input, the user's facial expressions and voice data are collected via the camera and microphone. An emotion engine analyzes this data and instantly scores the emotional state. This data is sent to a server and integrated with other results. The output is the user's current emotion score. Tracking changes in real time allows for more accurate emotion assessment.

[0736] Step 5:

[0737] The server integrates the results using deep learning technology. Inputs include sentiment scores and anomaly detection results. Deep learning techniques (e.g., TensorFlow or PyTorch) are used to integrate and analyze this data, identifying high-risk employees. The output is a list of employees identified as high-risk. This allows for a clear prioritization of those requiring action.

[0738] Step 6:

[0739] The server generates countermeasures and notifies the user via the terminal. The input is information on high-risk employees. The suggestion engine uses a generation AI model to create specific countermeasures. The output may include countermeasures such as "providing mental support" or "proposing flexible working hours." This notification, received on the terminal, allows HR personnel to respond quickly.

[0740] (Application Example 2)

[0741] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0742] A decline in employee work efficiency and motivation at a logistics center is a serious problem that has a significant impact on overall productivity. In such an environment, it is necessary to understand employees' emotional states and work efficiency in real time and to quickly provide appropriate improvement measures. However, conventional methods have made it difficult to effectively analyze such complex human emotional and behavioral data and propose countermeasures.

[0743] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0744] In this invention, the server includes means for collecting employee attributes, means for identifying information related to work efficiency from the collected employee attributes, means for analyzing employee emotions using natural language processing technology, means for detecting anomalies in work efficiency using machine learning technology, means for comprehensively analyzing the emotion analysis results and anomaly detection results using deep learning technology, means for recommending countermeasures for high-risk employees, and means for providing a user interface for visualizing real-time emotion and productivity data. This enables real-time monitoring of employee emotions and work efficiency within a logistics center, and allows for the provision of rapid and accurate support measures.

[0745] "Employee attributes" is a general term for personal information and work performance information related to employees.

[0746] "Work efficiency" refers to the proportion and quality of work results achieved by employees within a certain time frame.

[0747] "Natural language processing technology" is a general term for technologies used by computers to understand and analyze human language.

[0748] "Machine learning technology" is a technique that enables computers to learn from data and develop the ability to identify patterns and anomalies.

[0749] "Deep learning technology" is a technique that uses a multi-layered neural network to automatically extract sophisticated features from large-scale data.

[0750] A "high-risk employee" refers to an employee whose work performance or emotional state falls outside the standard, potentially requiring special measures or support.

[0751] "Means of recommending countermeasures" refers to methods for presenting appropriate support measures and improvement measures for high-risk employees.

[0752] A "user interface" refers to an interface that provides a visual or manipulative means for a system to interact with a user and exchange information.

[0753] The system implementing this invention aims to monitor employee emotions and work efficiency in real time at a logistics center and to suggest appropriate countermeasures to high-risk employees. To achieve this objective, the system includes the following configuration.

[0754] The server automatically collects employee attributes, such as the labor resources to which each employee belongs and attendance records. This involves using sensors and RFID readers as hardware, and Python as the software. The collected employee data is consistently stored in a central database.

[0755] Next, the server uses a generative AI model to perform natural language processing and analyze emotions from text data such as employee feedback and comments. During this process, the emotion engine calculates an emotion score and reveals the emotional state.

[0756] Furthermore, the server utilizes machine learning techniques to detect anomalies in the collected work efficiency data. Deep learning methods such as TensorFlow integrate sentiment analysis results and anomaly detection results to identify employees deemed high-risk.

[0757] The data visualized via the terminal is provided to the user in real time, and the recommendation engine generates concrete measures to support countermeasures for high-risk employees. This process supports rapid and efficient business collaboration.

[0758] For example, if a particular employee leaves feedback stating that their work motivation has recently decreased, the system will analyze that comment and recommend essential support measures such as "conducting individual interviews" or "flexibly adjusting working hours."

[0759] An example of a prompt to input into the generating AI model is: "Analyze the employee's sentiment score from the following data and suggest necessary support measures: {Employee comments, attendance records, sentiment data}."

[0760] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0761] Step 1:

[0762] The server collects employee attributes such as labor resources and attendance records from sensors and attendance management systems placed within the logistics center. Inputs include employee fingerprint data and RFID tag information, and output is that this data is stored in a central database. Data collection enables the visualization of basic employee information.

[0763] Step 2:

[0764] The server analyzes text data using a generative AI model based on collected employee data. Input includes employee daily report comments and feedback. Natural language processing techniques are used to analyze emotions from this text data. The output provides an analyzed emotion score and its trends. This analysis makes it possible to predict employee emotional fluctuations.

[0765] Step 3:

[0766] The server uses machine learning techniques to validate collected work efficiency data and detect anomalies. Inputs include employee working hours, work progress, and the number of work errors. Through data analysis, patterned anomalies are identified, and the presence and frequency of these anomalies are clearly indicated as output. This step allows for the proactive detection of potential employee problems.

[0767] Step 4:

[0768] The server uses deep learning technology to comprehensively analyze sentiment analysis results and anomaly detection results to identify high-risk employees. The inputs are sentiment scores and anomaly detection results. The output of the integrated analysis is a risk assessment, indicating the likelihood of high risk. This enables the early detection of employees who require attention.

[0769] Step 5:

[0770] Users view real-time data using the user interface on their device and receive suggestions from the recommendation engine. Inputs include analysis results and recommendation results. Outputs display the suggested content, serving as a source of information for developing concrete action plans. In this step, more specific support measures are formulated for high-risk employees.

[0771] Step 6:

[0772] For example, the terminal receives feedback from a specific employee, such as "Recently, their work motivation has decreased," and passes this feedback to a recommendation engine, which then presents the analysis results to the user. In this process, feedback and historical data are submitted as input, and the output is an optimal solution based on machine learning. This allows for the rapid provision of appropriate support measures to employees.

[0773] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0774] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0775] 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 this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0776] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0777] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0778] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0779] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0780] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0781] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0782] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0783] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0784] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0785] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0786] 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.

[0787] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0788] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0789] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0790] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0791] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0792] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0793] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0794] The following is further disclosed regarding the embodiments described above.

[0795] (Claim 1)

[0796] Means of collecting employee information,

[0797] A means of identifying data related to operational efficiency from collected employee information,

[0798] A method for analyzing employee emotions using natural language processing models,

[0799] A method for detecting anomalies in business efficiency data using machine learning models,

[0800] A method for integrating and analyzing sentiment analysis results and anomaly detection results using a deep learning model,

[0801] A means of recommending countermeasures for high-risk employees,

[0802] A system that includes this.

[0803] (Claim 2)

[0804] The system according to claim 1, wherein the employee information includes productivity data and attendance data.

[0805] (Claim 3)

[0806] The system according to claim 1, wherein the means for recommending countermeasures uses a recommendation engine.

[0807] "Example 1"

[0808] (Claim 1)

[0809] Means of collecting employee attributes,

[0810] A means of identifying information regarding operational efficiency from collected employee attributes,

[0811] A method for analyzing employee emotions using natural language processing technology,

[0812] A means of detecting anomalies in business efficiency information using machine learning techniques,

[0813] A method for integrating and analyzing sentiment analysis results and anomaly detection results using deep learning techniques,

[0814] A means of recommending countermeasures for high-risk employees,

[0815] A system that includes this.

[0816] (Claim 2)

[0817] The system according to claim 1, wherein the employee attributes include productivity information and attendance information.

[0818] (Claim 3)

[0819] The system according to claim 1, wherein the countermeasure recommendation means uses a recommendation system.

[0820] "Application Example 1"

[0821] (Claim 1)

[0822] Means for collecting employee information and machine information,

[0823] A means for identifying data related to work efficiency from collected employee information and machine information,

[0824] A method for analyzing employee emotions and worker feedback using natural language processing models,

[0825] A means for detecting anomalies in work efficiency data and machine operation data using machine learning models,

[0826] A means for comprehensively analyzing sentiment analysis results, feedback analysis results, and anomaly detection results using a deep learning model,

[0827] A means of recommending countermeasures for high-risk employees and equipment,

[0828] A system that includes this.

[0829] (Claim 2)

[0830] The system according to claim 1, wherein the employee information and machine information include productivity data, attendance data, and equipment operation data.

[0831] (Claim 3)

[0832] The system according to claim 1, wherein the means for recommending countermeasures uses a recommendation engine.

[0833] "Example 2 of combining an emotion engine"

[0834] (Claim 1)

[0835] Means of collecting employee information,

[0836] A means of identifying data related to the efficiency of job activities from collected employee information,

[0837] A method for analyzing employee emotions using language analysis models,

[0838] A means for detecting anomalies in job activity efficiency data using a learning algorithm,

[0839] A means for integrating and analyzing emotion analysis results and anomaly detection results using deep learning technology,

[0840] A device that analyzes users' emotions in real time,

[0841] A device that recommends countermeasures for high-risk employees,

[0842] A system that includes this.

[0843] (Claim 2)

[0844] The system according to claim 1, wherein the employee information includes productivity indicators and attendance data.

[0845] (Claim 3)

[0846] The system according to claim 1, wherein the countermeasure recommendation device uses a proposed engine.

[0847] "Application example 2 when combining with an emotional engine"

[0848] (Claim 1)

[0849] Means for collecting employee attributes,

[0850] A means of identifying information related to work efficiency from collected employee attributes,

[0851] A method for analyzing employee emotions using natural language processing technology,

[0852] A means of detecting anomalies in work efficiency using machine learning technology,

[0853] A means for integrating and analyzing emotion analysis results and anomaly detection results using deep learning technology,

[0854] A means of recommending countermeasures for high-risk employees,

[0855] A means of providing a user interface for visualizing real-time sentiment and productivity data,

[0856] A system that includes this.

[0857] (Claim 2)

[0858] The system according to claim 1, wherein the employee attributes include work volume data and attendance records.

[0859] (Claim 3)

[0860] The system according to claim 1, wherein the countermeasure recommendation means uses a recommendation system engine. [Explanation of Symbols]

[0861] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. Means of collecting employee information, A means of identifying data related to operational efficiency from collected employee information, A method for analyzing employee emotions using natural language processing models, A method for detecting anomalies in business efficiency data using machine learning models, A method for integrating and analyzing sentiment analysis results and anomaly detection results using a deep learning model, A means of recommending countermeasures for high-risk employees, A system that includes this.

2. The system according to claim 1, wherein the employee information includes productivity data and attendance data.

3. The system according to claim 1, wherein the means for recommending countermeasures uses a recommendation engine.

Images