System

Abstract

A system is provided.SOLUTION: A system comprising: means for acquiring information from an internal bulletin board; means for analyzing the acquired information and classifying the information into categories; means for collecting a browsing history and a click history of an employee; means for analyzing the collected history information and learning a field of interest of the employee; means for picking up and notifying important information to the employee based on an analysis result; and means for displaying information considered to be of particular interest on a dashboard of the employee as recommended information.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Describe the "problem that the invention aims to solve" and the "means for solving the problem."

[0005] Corporate internal bulletin boards contain a wide range of information, from business information to social information. However, due to the large amount of information and its diverse content, employees may miss information they need or are interested in. This increases the time and effort required to search for information, which can lead to reduced work efficiency and employee satisfaction. The present invention aims to improve work efficiency and employee satisfaction by enabling employees to quickly and efficiently obtain the information they need. [Means for solving the problem]

[0006] The present invention is a system that includes a means for acquiring information from an internal bulletin board, a means for analyzing the acquired information and classifying it by category, a means for collecting employees' browsing history and click history, a means for analyzing the collected history information and learning employees' areas of interest, a means for selecting and notifying employees of important information based on the analysis results, and a means for displaying information that is likely to be of particular interest to employees as recommended information on the employee's dashboard. This system prevents employees from missing necessary information and reduces the time and effort required to search for information. In this way, the present invention improves work efficiency and employee satisfaction.

[0007] An "internal bulletin board" is an electronic bulletin board system used to share information within a company.

[0008] "Generative AI" is a type of artificial intelligence that refers to models and algorithms that have the ability to generate new information based on data.

[0009] "Scraping" is the technique of automatically extracting information from a website.

[0010] "Natural language processing" is a technical field that enables computers to understand, analyze, and generate human language.

[0011] "Employee browsing history" is a record of the pages and links that employees have accessed on the Internet or internal company systems.

[0012] "Click history" is a record of the links and buttons that employees click on the Internet or on internal company systems.

[0013] The "means of learning" is the process by which an AI model extracts patterns and features from data and accumulates analytical results.

[0014] A "means of notification" is the mechanism the system uses to notify employees of specific information.

[0015] "Featured Information" is information selected and presented with special emphasis based on employee interests and needs.

[0016] A "dashboard" is a user interface that allows employees to view work-related information and notifications at a glance. [Brief explanation of the drawings]

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

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

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

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

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

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

[0023] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0024] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0025] [First embodiment]

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

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

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

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

[0030] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0031] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0032] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

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

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

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

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

[0038] The present invention provides a system for acquiring information from an internal bulletin board, extracting information related to employees' areas of interest or necessary for their work, and notifying them of the information. This system functions as follows.

[0039] System configuration and operation

[0040] Data collection and analysis

[0041] 1. (Server) The server periodically scrapes information from the internal bulletin board using a web scraping tool such as Beautiful Soup or Scrapy.

[0042] 2. (Server) The collected text data is analyzed using natural language processing (NLP) techniques. Specifically, NLP libraries such as spaCy and BERT are used to classify the text into categories (e.g., business information, social information).

[0043] Collecting employee behavior data

[0044] 3. (Device) When an employee logs in to an internal system and views information, their access history and click history are recorded. This is done using JavaScript and web tracking technology.

[0045] 4. (Terminal) The collected historical data is sent to the server in real time and stored in a database.

[0046] Learn about areas of interest and important information

[0047] 5. (Server) The server analyzes the collected employee history data and uses a generative AI model to learn each employee's areas of interest and work-related information. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used.

[0048] 6. (Server) Based on the analysis results, the newly collected information from the internal bulletin board is scrutinized and information that is important to employees is selected.

[0049] Displaying notifications and recommendations

[0050] 7. (Server) When important information is picked up, employees are notified in real time using a notification system (e.g., Firebase Cloud Messaging, WebSocket).

[0051] 8. (Device) A notification will pop up on the employee's PC or smartphone, providing a link to view more information.

[0052] 9. (Server) Furthermore, information that is deemed to be of particular interest is displayed as recommended information on the dashboard that is displayed when employees log in.

[0053] Specific examples

[0054] Example 1: Information collection and analysis

[0055] (Server) Every morning at 9:00, the server accesses the company bulletin board and scrapes the latest posts, which include information such as "New project recruitment," "Weekend club activities," and "Drinking parties next week."

[0056] In the (server) analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "project recruitment" is classified as business information, while "club activities" and "drinking parties" are classified as social information.

[0057] Example 2: Learning and notifying employees about information of interest

[0058] (Terminal) Employee A logs in to the system and the history of clicking on the page recruiting for new projects is recorded.

[0059] (Server) Based on this historical data, the AI ​​model learns that Employee A has a high interest in "Project Recruitment."

[0060] (Server) The next day, when a new project recruitment post is posted on the company bulletin board, the server picks up the information and notifies employee A's device via Firebase Cloud Messaging.

[0061] (Device) Employee A's PC displays a notification saying "A new project recruitment post has been made," and he can click on the link to view more information.

[0062] Example 3: Displaying recommendations

[0063] (Server) By analyzing employee B's history data, it is discovered that employee B frequently views pages related to club activities.

[0064] (Server) Based on the results, the AI ​​model picks out information about club activities that it determines Employee B will be particularly interested in, and displays this information on the dashboard every Monday as "Recommended club activities for this weekend."

[0065] (Device) When employee B logs in on Monday, the dashboard will display "Recommended club activities for this weekend."

[0066] In this way, the system of the present invention is designed to enable employees to quickly and effectively obtain information necessary for their work and information of interest, thereby improving work efficiency and employee satisfaction.

[0067] The processing flow will be explained below.

[0068] Step 1:

[0069] (Server) Set up regular access to an internal bulletin board. For example, access the bulletin board every morning at 9:00 using Scrapy or Beautiful Soup to collect the latest posts.

[0070] Step 2:

[0071] (Server) Analyze the collected text data using natural language processing (NLP) techniques. Specifically, use spaCy or BERT to tokenize the text and identify entities and classify them by category.

[0072] Step 3:

[0073] (Server) Analyzed text data is stored in a database by category, e.g., business information, social information, etc.

[0074] Step 4:

[0075] (Device) When an employee logs into the company's internal system, their access history and click history are recorded in real time using JavaScript and web tracking technology.

[0076] Step 5:

[0077] (Terminal) The recorded history data is sent to the server in real time.

[0078] Step 6:

[0079] (Server) Stores the submitted historical data in a database and prepares it for analysis.

[0080] Step 7:

[0081] (Server) Use a generative AI model to analyze collected historical data and learn employee interests and work-related information. Example: Use Python machine learning libraries (TensorFlow, PyTorch).

[0082] Step 8:

[0083] (Server) Based on the analysis results, the newly collected information from the internal bulletin board is scrutinized and important information for employees is selected, with priority given to information that is of high interest to employees.

[0084] Step 9:

[0085] (Server) The collected information is notified to employees in real time using Firebase Cloud Messaging or WebSocket.

[0086] Step 10:

[0087] (Device) A notification pops up on the employee's PC or smartphone. For example, a message saying "A new project recruitment post has been made" appears.

[0088] Step 11:

[0089] (Device) When an employee clicks on the pop-up notification, they are taken to a screen that displays more information.

[0090] Step 12:

[0091] (Server) In addition, prepare to display information that is thought to be of particular interest as recommended information on employees' dashboards.

[0092] Step 13:

[0093] (Device) When an employee logs in, recommended information is displayed on the dashboard. For example, information is presented as "recommended club activities for this weekend."

[0094] In this way, the system quickly and efficiently provides employees with the information they need and are most interested in, thereby improving work efficiency and employee satisfaction.

[0095] Example 1

[0096] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0097] Conventional internal information sharing systems have made it difficult to efficiently collect and provide information that employees need or are interested in. As a result, employees may miss important information or waste time on information that is not relevant to their work. This can lead to problems such as reduced work efficiency and lower employee satisfaction.

[0098] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0099] In this invention, the server includes means for acquiring information from an internal bulletin board, means for analyzing the acquired information and classifying it by category, means for collecting employees' browsing history and click history, means for analyzing the collected history information and learning employees' areas of interest, means for selecting and notifying employees of important information based on the analysis results, means for displaying information that is likely to be of particular interest to employees as recommended information on their dashboards, means for periodically collecting information using a scraping tool, means for analyzing text data using natural language processing technology and classifying it after tokenization and part-of-speech tagging, and means for machine learning employee areas of interest using a generative AI model. This makes it possible to quickly and effectively collect and provide information that employees need or are interested in.

[0100] An "internal bulletin board" is a digital bulletin board set up within a company or organization for the purpose of sharing information among employees.

[0101] "Means of obtaining information" refers to the mechanism by which the server automatically collects the content posted on the internal bulletin board.

[0102] "Means of analysis" refers to techniques that use natural language processing technology to analyze collected text data and understand its content.

[0103] "Means for categorizing by category" refers to a method for sorting analyzed text data into specific categories (e.g., business information, social information).

[0104] "Means for collecting employees' browsing and click histories" refers to technology for recording the operations performed by employees within the company's internal systems.

[0105] "Means for learning areas of interest" refers to a machine learning model that analyzes and learns from employee interests and concerns from collected historical data.

[0106] "Means of notification" refers to a system for informing employees of important information in real time based on the analysis results.

[0107] "Means for displaying recommended information on the dashboard" refers to an interface that displays information related to employees' areas of interest when they log in.

[0108] A "scraping tool" is software or a library used to automatically extract data from websites.

[0109] "Natural language processing technology" is a technology that enables computers to understand and process human language.

[0110] "Tokenization" refers to the process of dividing text data into smaller units such as words or sentences.

[0111] "Part-of-speech tagging" refers to labeling tokenized words with their parts of speech (e.g., nouns, verbs, adjectives).

[0112] A "generative AI model" is an artificial intelligence model that uses machine learning to learn patterns and rules from data and make predictions and classifications.

[0113] The present invention is a system that acquires information from an in-house bulletin board, extracts information that is relevant to employees' fields of interest or their work, and notifies them of the information. This system is configured as follows.

[0114] System configuration and operation procedures

[0115] This system is mainly composed of servers, terminals, and users, each of which has a specific role and functions in cooperation with each other.

[0116] Data collection and analysis methods

[0117] The server periodically scrapes information from the company's internal bulletin board. The scraping tools used are Beautiful Soup and Scrapy. The collected data is then analyzed using natural language processing (NLP) techniques. Specifically, the NLP libraries spaCy and BERT are used to tokenize the text, tag parts of speech, and analyze dependencies. Finally, the text data is classified by category (e.g., business information, social information).

[0118] How to collect employee behavioral data

[0119] (Device) records access history and click history every time an employee logs in to the company's internal system or browses information. This is done using JavaScript and web tracking technology. The collected history data is sent to (server) in real time and stored in a database.

[0120] Areas of interest and how to learn important information

[0121] The server analyzes the collected employee behavioral data and uses a generative AI model to learn each employee's areas of interest and work-related information. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used. The AI ​​model learns patterns and areas of interest based on this data.

[0122] How notifications and recommendations are displayed

[0123] (Server) scrutinizes newly collected information from internal bulletin boards based on the analysis results and selects information that is important to employees. (Server) notifies employees in real time using a notification system (e.g., Firebase Cloud Messaging, WebSocket). (Terminal) displays a notification as a pop-up on employees' PCs or smartphones and provides a link to view more information. (Server) also displays information of particular interest as recommended information on the dashboard that is displayed when employees log in.

[0124] Specific examples

[0125] Example 1: Information collection and analysis

[0126] (Server) accesses the company's internal bulletin board at 9:00 every morning and scrapes the latest posts. For example, this includes information such as "new project recruitment," "weekend club activities," and "drinking party next week." In the analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "project recruitment" is classified as business information, while "club activities" and "drinking party" are classified as social information.

[0127] Example 2: Learning and notifying employees about information of interest

[0128] On (terminal), employee A logs in to the system and records the history of clicking on the new project recruitment page. Based on this history data, (server) uses its AI model to learn that employee A has a high interest in "project recruitment." The next day, when a new project recruitment post is posted on the company bulletin board, (server) picks up the information and notifies employee A's device via Firebase Cloud Messaging. (terminal) displays a notification on employee A's PC stating "A new project recruitment post has been made," and employee A can click on the link to view more information.

[0129] Example 3: Displaying recommendations

[0130] When (Server) analyzes employee B's historical data, it finds that employee B frequently views pages related to club activities. Based on the results, (Server) selects club activity information that the AI ​​model determines to be of particular interest to employee B, and displays this information on the dashboard every Monday as "Recommended club activities for this weekend." When employee B logs in on Monday, (Device) displays "Recommended club activities for this weekend" on the dashboard.

[0131] This system will enable employees to obtain the information they need quickly and efficiently, which is expected to improve work efficiency and employee satisfaction.

[0132] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0133] Step 1:

[0134] (Server) accesses the company bulletin board every morning at 9:00 and scrapes new posts. The URL of the bulletin board is given as input, and the HTML structure is analyzed using a scraping tool (Beautiful Soup or Scrapy). The output is the extracted text data. Specifically, it performs the following operations: , , ) to extract the text.

[0135] Step 2:

[0136] The text data acquired by the server is preprocessed using natural language processing (NLP) techniques. The input is the raw data acquired in step 1, and the output is the preprocessed text data. Specifically, NLP libraries (spaCy and BERT) are used to tokenize the data and remove unnecessary tags and spaces.

[0137] Step 3:

[0138] The server tokenizes the preprocessed text data and performs part-of-speech tagging and dependency analysis. The input is the data preprocessed in step 2, and the output is tokenized and tagged text. Specifically, it uses spaCy to perform structural analysis of the sentence and tag each word with a part-of-speech tag.

[0139] Step 4:

[0140] The server classifies the analyzed text data into categories (business information, social information). The input is the text tagged with part-of-speech tags in step 3, and the output is the classified data. Specifically, it uses a machine learning algorithm to classify each text into a predefined category.

[0141] Step 5:

[0142] (Terminal) records logins to the company's internal system, access history, and click history. The input is the user's operation log, and the output is the recorded access data. Specifically, JavaScript code captures the user's operations and sends them to the server.

[0143] Step 6:

[0144] The historical data collected by (the terminal) is sent to the server in real time and stored in the database. The input is the access data recorded in step 5, and the output is the historical data stored on the server. Specifically, the data is sent to the server via an HTTP request and stored in the database.

[0145] Step 7:

[0146] The server analyzes the collected employee behavioral data and uses a generative AI model to learn each employee's areas of interest and work-related information. The input is the historical data saved in step 6, and the output is the learned model and interest information. Specifically, the AI ​​model is trained using TensorFlow or PyTorch to learn patterns from the data.

[0147] Step 8:

[0148] (Server) scrutinizes the newly collected information from the internal bulletin board and picks out information that is important to employees. The input is the data from Step 1 and Step 7, and the output is the picked-up important information. Specifically, it filters the information based on the prediction results of the AI ​​model.

[0149] Step 9:

[0150] The server notifies employees in real time using a notification system (Firebase Cloud Messaging or WebSocket). The input is the information picked up in step 8, and the output is a notification message. Specifically, the server sends a message in real time through the notification system.

[0151] Step 10:

[0152] The device displays a pop-up notification and provides a link to view more information. The input is the notification message sent in step 9, and the output is the user's browsing behavior. The specific behavior is to display a pop-up window in a browser or application.

[0153] Step 11:

[0154] Information that the server determines to be of particular interest is displayed as recommended information on the dashboard when employees log in. The input is the data from steps 7 and 8, and the output is the recommended information displayed on the dashboard. Specifically, the settings are configured to list and display specific information on the dashboard.

[0155] Through the above processing steps, a system is provided that allows employees to quickly and efficiently obtain the information they need or are interested in.

[0156] (Application example 1)

[0157] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0158] Conventional systems that collect information from internal bulletin boards and notify employees have difficulty in properly identifying employees' areas of interest and necessary information. Furthermore, at production sites, proper information collection and notification is not performed, making it difficult to provide efficient work support or timely maintenance. For this reason, there is a need for a system that efficiently collects production schedules and maintenance information at factories and notifies workers appropriately.

[0159] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0160] In this invention, the server includes means for acquiring information from an in-house bulletin board, means for analyzing the acquired information and classifying it by category, means for collecting employees' browsing history and click history, means for analyzing the collected history information and learning employees' areas of interest, means for selecting and notifying employees of important information based on the analysis results, means for displaying information that is likely to be of particular interest as recommended information on the employee's dashboard, means for scraping production plans and function check information from an in-house management system, and means for analyzing the collected information and notifying it in conjunction with sensor data and work logs acquired during work. This allows employees and workers to quickly and efficiently receive the information they need, improving work efficiency and enabling appropriate maintenance.

[0161] An "internal bulletin board" is an electronic bulletin board system set up within a company for the purpose of sharing information and communicating.

[0162] "Means of obtaining information" refers to the methods and technologies used to collect the necessary information from various data sources.

[0163] "Means for analyzing information and classifying it into categories" refers to technology that analyzes acquired data and classifies it into specific categories based on its content.

[0164] "Means of collecting employees' browsing and click histories" refers to technology that collects the behavioral history of employees when they use web systems.

[0165] "Means of analyzing collected historical information to learn employees' areas of interest" refers to technology that analyzes collected behavioral data to learn what information a particular employee is interested in.

[0166] "Means of selecting and notifying employees of important information based on the analysis results" refers to methods and technologies for selecting important information from the analysis results and notifying employees of it.

[0167] "Means for displaying recommended information on employee dashboards" refers to a method for displaying information that is determined to be of particular interest to employees on their interface.

[0168] "Factory management system" refers to a comprehensive system for managing factory production and equipment.

[0169] "Means for scraping production plan and function check information" refers to technology that automatically extracts necessary information from factory management systems.

[0170] "Sensor data and work logs obtained during work" refers to the various data generated by robots and workers while they are working, as well as the recorded work history.

[0171] "Generative AI models" refer to algorithms and technologies that use machine learning techniques to generate knowledge and predictive models from data.

[0172] A "prompt" is an instruction given to a generative AI model to make it perform a specific task.

[0173] A "notification system" refers to technology that notifies individual devices of important information in real time.

[0174] The following describes an embodiment of the present invention. The present invention is a system that acquires information from a factory management system, notifies robots and workers of appropriate information, and improves work efficiency. This system functions as follows.

[0175] System configuration and operation

[0176] Data collection and analysis

[0177] 1. Server: The server periodically scrapes production plans and functional check information from the factory management system using web scraping tools such as Beautiful Soup and Scrapy.

[0178] 2. Server: The collected text data is analyzed using natural language processing (NLP) techniques. Specifically, NLP libraries such as spaCy and BERT are used to classify the text into categories (e.g., production information, maintenance information).

[0179] User behavior data collection

[0180] 3. Robot: The sensor data and work logs acquired by the robot while working are recorded and sent to the server in real time. The data is collected using JavaScript and web tracking technology.

[0181] 4. Server: The collected historical data is stored on the server.

[0182] Learn about areas of interest and important information

[0183] 5. Server: The server analyzes the collected robot history data and uses generative AI models to learn important information for each robot and worker. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used.

[0184] 6. Server: Based on the analysis results, the newly collected information from the factory management system is scrutinized and important information is picked up.

[0185] Displaying notifications and recommendations

[0186] 7. Server: When important information is picked up, it is notified to the robot or worker in real time using a notification system (e.g. MQTT, WebSocket).

[0187] 8. Robot: Notifications are sent via the robot's display and voice output, which can be checked by the worker. Work instructions and changes are also displayed here.

[0188] Specific examples

[0189] Example 1: Information collection and analysis

[0190] Server: Every morning at 9:00, the server accesses the factory management system and scrapes the latest production schedule and maintenance schedule, including information such as "new production line setup," "machine maintenance next month," and "equipment inspection this weekend."

[0191] Server: In the analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "production line settings" is classified as production information, while "machine maintenance" and "equipment inspection" are classified as maintenance information.

[0192] Example 2: Learning and notifying robot behavior information

[0193] Robot: Sensor data and work logs acquired by specific robots while performing machine maintenance are recorded.

[0194] Server: Based on this historical data, the AI ​​model learns the information necessary for the robot.

[0195] Server: When new machine maintenance information is posted on the factory management system, the server picks up the information and notifies the robot via MQTT.

[0196] Robot: A notification saying "New machine maintenance information has been added" appears on the robot's display and can be confirmed by the worker.

[0197] Example 3: Displaying recommendations

[0198] Server: By analyzing historical data for a particular robot, you may discover that the robot frequently performs tasks on a particular production line setup.

[0199] Server: Based on the results, the AI ​​model picks out information about the production line settings that it deems particularly important for that robot.

[0200] Server: Every morning, the server notifies the robot of this information and displays it on the robot's display as "Today's recommended production line settings."

[0201] In this way, the system of the present invention is designed to enable robots and workers to quickly and effectively obtain the information they need for their work, thereby improving work efficiency and ensuring appropriate maintenance.

[0202] Prompt Sentence Examples

[0203] "Next week's machine maintenance schedule"

[0204] "New Production Plan for Project X"

[0205] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0206] Step 1:

[0207] The server acquires production plan information and maintenance schedule information from the factory management system.

[0208] Input: Production plan information and maintenance schedule information from the factory management system.

[0209] Data processing: Use web scraping tools such as Beautiful Soup or Scrapy to extract the necessary information and save it as text data.

[0210] Output: Text data of extracted production plan information and maintenance schedule information.

[0211] Step 2:

[0212] The server analyzes the acquired text data and classifies it into categories.

[0213] Input: The text data extracted in step 1.

[0214] Data processing: Using natural language processing (NLP) techniques, we tokenize the text and classify it into categories (e.g., production information, maintenance information). Specifically, we use NLP libraries such as spaCy and BERT.

[0215] Output: Text data classified by category.

[0216] Step 3:

[0217] The robot records sensor data and work logs acquired during work and sends them to a server.

[0218] Input: Sensor data and work logs acquired while the robot is working.

[0219] Data processing: Collect data in real time and send it to the server. Record data using JavaScript and web tracking technology.

[0220] Output: Sensor data and work logs sent to the server.

[0221] Step 4:

[0222] The server analyzes the collected robot history data and uses a generative AI model to learn important information.

[0223] Input: Sensor data and work logs sent in step 3.

[0224] Data processing: Using Python machine learning libraries (e.g., TensorFlow, PyTorch), the data is analyzed to learn information that is important to each robot.

[0225] Output: Key information identified by the generative AI model.

[0226] Step 5:

[0227] The server picks out important information based on the analysis results and notifies you.

[0228] Input: Key information learned in step 4.

[0229] Data processing: Newly collected production plans and maintenance information are compared with the analysis results to select important information.

[0230] Output: Important information picked up.

[0231] Step 6:

[0232] The server uses a notification system to notify robots and workers of important information in real time.

[0233] Input: Important information picked up in step 5.

[0234] Data processing: Use MQTT or WebSocket to send notifications in real time.

[0235] Output: Notification message that will be displayed to the robot and worker.

[0236] Step 7:

[0237] The robot will display the notified information on a screen or output it as audio so that the worker can check it.

[0238] Input: Information provided in step 6.

[0239] Data processing: Display and audio output are performed so that workers can check the information.

[0240] Output: Notifications are displayed on the robot's display and audio output.

[0241] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0242] The following describes an embodiment of the present invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the areas of interest and emotional state of employees. The system is operated by combining a generative AI and an emotion engine.

[0243] System configuration and operation

[0244] Data collection and analysis

[0245] 1. (Server) Regularly scrape information from internal bulletin boards. For example, use tools like Scrapy or Beautiful Soup to collect the latest posts every morning at 9am.

[0246] 2. (Server) Analyze the collected text data using natural language processing (NLP) techniques, such as spaCy or BERT, to tokenize the text and classify it by category (e.g., business information, social information).

[0247] Collecting employee behavioral and emotional data

[0248] 3. (Device) Every time an employee logs in to an internal system and views information, their access history and click history are recorded. This is done using JavaScript and web tracking technology.

[0249] 4. (Device) Obtain recorded history data, as well as user-entered text and interaction data.

[0250] 5. (Terminal) Send the collected data to the server in real time.

[0251] Learning areas of interest and emotional states

[0252] 6. (Server) Stores the submitted historical and interaction data in a database and prepares it for analysis.

[0253] 7. (Server) Use a generative AI model to analyze historical data to learn employee interests and job-related information. Use Python machine learning libraries (e.g., TensorFlow, PyTorch).

[0254] 8. (Server) An emotion engine is used to analyze the user's text input and click behavior to identify their emotional state. Emotion recognition technology combines NLP technology with machine learning models (e.g., BERT, Keras).

[0255] Providing notifications and recommendations

[0256] 9. (Server) Based on the learning results, the server scrutinizes the newly collected information from the internal bulletin board and selects information that is important to employees. The server also adjusts the timing and content of notifications based on the results of the emotion engine.

[0257] 10. (Server) Use Firebase Cloud Messaging or WebSocket to send notifications to employees in real time.

[0258] 11. (Device) Notifications are displayed as pop-ups on employees' PCs or smartphones. The pop-ups display specific messages such as "A new project recruitment post has been made," and employees can click on a link to access more information.

[0259] 12. (Server) Information that is thought to be of particular interest is displayed as recommended information on the dashboard that employees log in to.

[0260] Specific examples

[0261] Example 1: Information collection and analysis

[0262] (Server) Every morning at 9:00, information is scraped from the company bulletin board to obtain information such as "New project recruitment," "Weekend club activities," and "Drinking parties next week."

[0263] (Server) Analyze the acquired text using spaCy or BERT and classify it into categories. For example, "Project recruitment" is classified as business information, while "Club activities" and "Drinking parties" are classified as social information.

[0264] Example 2: Learning and notifying employee interests and emotional states

[0265] (Device) Employee A logs in and clicks on the new project recruitment page. The text input and click history are recorded.

[0266] (Server) The generative AI model analyzes the historical data and learns that Employee A has a high interest in the "Project Recruitment" campaign. The emotion engine also detects "excitement" and "interest" from Employee A's input text.

[0267] (Server) The next day, if a new project recruitment post is made, a notification is sent to Employee A's device using Firebase Cloud Messaging. Based on the results of the emotion engine, the notification message is adjusted to read, "This project seems interesting to you."

[0268] (Device) A pop-up message appears on Employee A's PC saying "A new project recruitment post has been made," and clicking the link displays more information.

[0269] Example 3: Displaying recommendations

[0270] (Server) Analyzes employee B's historical data and emotional state and finds that he frequently views pages related to club activities and feels "excited" and "expected."

[0271] (Server) Based on this, information on club activities that are of particular interest to employee B is selected and displayed on the dashboard every Monday as "Recommended club activities for this weekend."

[0272] (Device) When employee B logs in on Monday, the dashboard will display "Recommended club activities for this weekend."

[0273] In this way, the system of the present invention provides employees with information necessary for their work and information of interest quickly and efficiently, and also provides appropriate information according to the user's emotional state, thereby improving work efficiency and employee satisfaction.

[0274] The processing flow will be explained below.

[0275] Step 1:

[0276] (Server) Set up periodic access to the internal bulletin board. For example, use Scrapy or Beautiful Soup to collect the latest postings every morning at 9am.

[0277] Step 2:

[0278] (Server) Analyze the collected text data using natural language processing (NLP) technology. Specifically, use spaCy or BERT to tokenize the text and classify it by category (e.g., business information, social information).

[0279] Step 3:

[0280] (Server) The analyzed text data is stored in a database by category. The stored data is used for later information review.

[0281] Step 4:

[0282] (Device) When an employee logs in to the company's internal system, JavaScript and web tracking technology are used to record access history and click history.

[0283] Step 5:

[0284] (Device) Recorded historical data is sent to the server in real time, making it immediately available for analysis.

[0285] Step 6:

[0286] (Server) Stores the submitted historical data in a database so that it can be used by the generative AI model and emotion engine for analysis.

[0287] Step 7:

[0288] (Server) Analyze historical data using a generative AI model to learn employee interests and work-related information. Build the learning model using Python machine learning libraries (e.g., TensorFlow, PyTorch).

[0289] Step 8:

[0290] (Server) An emotion engine is used to analyze user input data and click history to identify the user's emotional state. For example, NLP techniques and machine learning models are combined to identify emotions such as "excitement," "interest," and "boredom" from user text input.

[0291] Step 9:

[0292] (Server) Based on the analysis results, the server scrutinizes newly collected information from internal bulletin boards and selects information that is important to employees. At this time, the server adjusts the timing and content of notifications according to the user's emotional state as determined by the emotion engine.

[0293] Step 10:

[0294] (Server) Notifications are sent to employees in real time using Firebase Cloud Messaging and WebSockets. The content and format of the notifications are customized taking into account the results of the emotion engine.

[0295] Step 11:

[0296] (Device) A notification pops up on the employee's PC or smartphone. For example, a message saying "A new project recruitment post has been made" appears. When the user clicks on the link, more information is displayed.

[0297] Step 12:

[0298] (Server) In addition, prepare to display information that is thought to be of particular interest as "recommended information" on employees' dashboards.

[0299] Step 13:

[0300] (Device) When an employee logs in, "recommended information" is displayed on the dashboard. For example, information such as "recommended club activities for this weekend" is displayed.

[0301] This system allows employees to quickly and efficiently obtain information necessary for their work and information of interest. In addition, because it provides appropriate information based on the user's emotional state, it is expected to improve work efficiency and employee satisfaction.

[0302] Example 2

[0303] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0304] Conventional internal information sharing systems make it difficult for employees to efficiently obtain the information they need, resulting in problems such as missing information or failing to grasp important information. Furthermore, to improve the quality of communication within a company and increase the work efficiency of each employee, it is necessary to provide information that takes into account the interests and emotional state of employees, but there were few systems that could achieve this. Furthermore, the lack of information provided that was tailored to the interests of each employee sometimes resulted in information overload, which in turn reduced productivity.

[0305] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0306] In this invention, the server includes a means for acquiring information from the in-house bulletin board, a means for analyzing the acquired information using natural language processing technology and classifying it into categories, and a means for collecting collected history information and user input text in real time. This makes it possible to efficiently collect information needed by employees and provide important information at an appropriate time, taking into account the employees' areas of interest and emotional state.

[0307] 1. An "internal bulletin board" is an online platform for sharing information within a company or organization.

[0308] 2. "Means of obtaining information" refers to tools and programs for automatically collecting necessary information from internal bulletin boards.

[0309] 3. "Natural language processing technology" is an artificial intelligence technology for analyzing and understanding human language.

[0310] 4. "Categorization methods" refers to algorithms or programs that group collected information based on specific criteria.

[0311] 5. "Browsing history" refers to a record of web pages and documents that a user has previously viewed.

[0312] 6. "Click History" refers to a record of links and buttons that a user has previously clicked.

[0313] 7. "History information" refers to all data relating to a user's behavioral history.

[0314] 8. "User Input Text" means any characters or sentences entered by a User into the System.

[0315] 9. "Real-time collection means" refers to technologies or methods that capture data as it is transmitted.

[0316] 10. "Database" refers to a system for efficiently storing, managing, and retrieving large amounts of data.

[0317] 11. "Generative AI model" refers to an artificial intelligence model that learns from generated data and performs tasks such as prediction and classification.

[0318] 12. "Emotion Engine" refers to technology or programs that identify a user's emotional state from their text input or behavior.

[0319] 13. "Interest Areas" refers to the themes or topics that interest a particular user.

[0320] 14. "Business-Related Information" means information related to a User's job function.

[0321] 15. "Means of notification" refers to tools and technologies used to communicate information to users.

[0322] 16. "Recommended Information" refers to the provision of specific information selected based on a user's interests and behavior.

[0323] The following describes an embodiment of the present invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the areas of interest and emotional state of employees. The system is operated by combining a generative artificial intelligence model and an emotion analysis engine.

[0324] System configuration and operation

[0325] Data collection and analysis

[0326] 1. The server periodically retrieves information from the company bulletin board. To collect the information, use a web scraping tool such as Scrapy or Beautiful Soup. For example, set it to collect the latest posts every morning at 9:00.

[0327] 2. The server analyzes the collected text data using natural language processing (NLP) techniques, such as spaCy and BERT, to tokenize the text and classify it into categories (e.g., business information, social information).

[0328] Collecting employee behavioral and emotional data

[0329] 3. The device records access and click histories every time an employee logs into the company's internal systems and views information. This recording is done using JavaScript and web tracking technology.

[0330] 4. The device acquires the recorded history data, text entered by the user, and interaction data, and transmits them to the server in real time.

[0331] Saving data and preparing for analysis

[0332] 5. The server stores the transmitted history and interaction data in a database and prepares it for analysis, using a database system such as PostgreSQL or MongoDB.

[0333] Learning areas of interest and emotional states

[0334] 6. The server uses generative artificial intelligence models (e.g., TensorFlow, PyTorch) to analyze historical data and learn employee interests and work-related information.

[0335] 7. The server uses an emotion engine (e.g., BERT, Keras) to identify the user's emotional state from their text input and click behavior.

[0336] Providing notifications and recommendations

[0337] 8. Based on the learning results and sentiment analysis results, the server examines the newly collected information from the internal bulletin board and notifies employees of important information. For example, it prepares a tailored message such as, "This project seems interesting to you."

[0338] 9. The server uses Firebase Cloud Messaging or WebSocket to send notifications in real time to employee devices, which are displayed as pop-ups on PCs and smartphones.

[0339] 10. The server displays information of particular interest as recommended information on the dashboard screen when employees log in. For example, information on club activities of high interest is displayed on the dashboard every Monday as "Recommended activities for this weekend."

[0340] Specific examples

[0341] Example 1: Information collection and analysis

[0342] The server scrapes information from the company's internal bulletin boards every morning at 9 a.m., obtaining information such as "new project recruitment," "weekend social events," and "next week's meetings."

[0343] The server analyzes the acquired text using natural language processing technology (spaCy or BERT) and classifies it into categories. For example, "project recruitment" is classified as business information, while "social events" and "meetings" are classified as social information.

[0344] Example 2: Learning and notifying employee interests and emotional states

[0345] The device records access history and click history when employee A logs in and clicks on the page recruiting for new projects.

[0346] The server analyzes the historical data based on a generative artificial intelligence model and learns that Employee A has a high interest in the "Project Recruitment." The emotion engine also detects "excitement" and "interest" from Employee A's input text.

[0347] The next day, if there is a new project opening, the server will use Firebase Cloud Messaging to notify employee A's device, saying, "We think this project might be interesting to you."

[0348] The device displays a pop-up on Employee A's PC saying, "A new project recruitment post has been made," and when Employee A clicks on the link, more information will be displayed.

[0349] Example 3: Displaying recommendations

[0350] The server analyzes employee B's historical data and emotional state, and determines that he frequently views pages related to club activities and feels emotions of "excitement" and "expectation."

[0351] Based on this, the server selects information about club activities that employee B is particularly interested in and displays it on the dashboard every Monday as "Recommended club activities for this weekend."

[0352] When employee B logs in to the device on Monday, the dashboard will display "Recommended club activities for this weekend."

[0353] Example prompt sentence:

[0354] "Collect information on new project openings and notify employees."

[0355] "Identify areas of interest from employees' click history"

[0356] "Analyze the user's emotional state using an emotion engine and provide appropriate notifications."

[0357] This system allows employees to quickly and efficiently receive not only the information they need for their work, but also appropriate information tailored to their individual interests and emotional state, thereby improving work efficiency and employee satisfaction.

[0358] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0359] Step 1:

[0360] The server scrapes information from the company bulletin board every morning at 9:00. It uses the bulletin board URL as input and generates data including the title, body, and posting date and time of each post as output. Specifically, it uses Scrapy and Beautiful Soup to parse the HTML and extract the required text information.

[0361] Step 2:

[0362] The server analyzes the collected text data using natural language processing technology and classifies it into categories. It uses the acquired text data as input and obtains data classified into categories (e.g., business information, social information) as output. Specifically, it uses spaCy and BERT to tokenize the text and perform contextual analysis.

[0363] Step 3:

[0364] The device records access and click histories every time an employee logs into the company's internal systems and browses information. It uses the employee's browsing and clicking behavior as input and generates the recorded history data as output. Specifically, it captures user actions using JavaScript and web tracking technology.

[0365] Step 4:

[0366] The device sends the recorded history data and the user's input text and interaction data to the server in real time. The device uses the recorded history data and the user's text data as input and generates the data sent to the server as output. Specifically, the device converts the data format to JSON or similar and sends it to the server using the HTTPS protocol.

[0367] Step 5:

[0368] The server stores the submitted historical and interaction data in a database and prepares it for analysis. It uses the received historical and interaction data as input and generates the data stored in the database as output, specifically by inserting data using database queries.

[0369] Step 6:

[0370] The server uses a generative AI model to analyze historical data and learn employee interests and work-related information. It uses the stored historical data as input and obtains data identifying employee interests as output. Specifically, it trains the model using TensorFlow or PyTorch and makes predictions based on the historical data.

[0371] Step 7:

[0372] The server uses an emotion engine to identify the user's emotional state from their text input and click behavior. It uses the user's text data and click data as input and obtains the identified emotional state as output. Specifically, it uses BERT and Keras to perform emotion recognition.

[0373] Step 8:

[0374] Based on the learning results and sentiment analysis results, the server scrutinizes the newly collected information from the internal bulletin board, picks out information that is important to employees, and prepares the notification content. It uses the data obtained from the internal bulletin board and the analysis results as input, and generates a notification message as output. Specifically, it selects an appropriate template and creates the message.

[0375] Step 9:

[0376] The server sends notifications to employee devices in real time using Firebase Cloud Messaging or WebSocket. It uses the prepared notification message as input and gets the sent notification as output. Specifically, it sends messages using the Firebase API or WebSocket protocol.

[0377] Step 10:

[0378] The device displays the received notification as a pop-up on the employee's PC or smartphone. It uses the received notification message as input and obtains the displayed pop-up notification as output. Specifically, it uses the browser's Notification API or the OS's notification function to display the message.

[0379] Step 11:

[0380] The server then displays the most interesting information as recommendations on the employee's dashboard. It uses information based on the areas of interest as input and generates recommendations that are displayed on the dashboard as output. Specifically, it dynamically updates the content of the web page and applies templates to display the information of interest.

[0381] (Application example 2)

[0382] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0383] Conventional internal bulletin board systems and notification systems do not provide information that takes into account the interests and emotional state of individual employees, limiting their ability to improve work efficiency and employee satisfaction. Furthermore, it is difficult to provide employees with the important information they need at the right time, leading to problems with missed information and excessive notifications. There is a need for a system that can solve these issues and provide optimal information to employees while improving work efficiency.

[0384] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring information from the in-house bulletin board, means for analyzing the acquired information and classifying it by category, and means for analyzing the collected history data and interaction data to identify the user's emotional state. This makes it possible to pick out important information based on the employee's areas of interest and emotional state, appropriately adjust the content of notification messages, and provide them to employees in real time.

[0385] An "internal bulletin board" is an online bulletin board for the purpose of sharing information and communicating among employees.

[0386] "Means of obtaining information" refers to the techniques and methods used to collect the necessary information from the internal bulletin board.

[0387] "Means of analyzing information" refers to the techniques and methods used to structure collected information and convert it into an easily understandable format.

[0388] "Categorization means" refers to techniques or methods for classifying analyzed information into different categories.

[0389] "Means of collecting employee browsing and click history" refers to technologies and methods that record what information employees view and what links they click.

[0390] "Means for analyzing historical information and learning about employees' areas of interest" refers to techniques and methods for identifying employees' areas of interest and concern from collected historical information.

[0391] "Means for analyzing interaction data and identifying the user's emotional state" refers to technologies and methods for analyzing the emotions of employees at that time based on their interaction data.

[0392] "Means of notification" refers to the technology and methods used to inform employees of necessary information based on the analysis results.

[0393] "Means for adjusting the content of notification messages" refers to techniques and methods for changing notification messages to optimal content depending on the emotional state of employees.

[0394] "Means for displaying recommended information on the dashboard" refers to the technology and methods for displaying information of particular interest on employees' operation screens.

[0395] "Means of real-time notification" refers to technologies and methods that instantly inform employees of the information they need at that moment.

[0396] The following describes an embodiment of this invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the employee's areas of interest and emotional state. The system is operated by combining a generative AI and an emotion engine.

[0397] System configuration and operation

[0398] Data collection and analysis

[0399] The server periodically collects information from internal bulletin boards using scraping technologies such as Scrapy and Beautiful Soup. The collected information is analyzed and categorized using natural language processing (NLP) techniques. Specifically, text is tokenized using spaCy and BERT, and then classified into categories such as business information and social information.

[0400] Collecting employee behavioral and emotional data

[0401] The device uses JavaScript and web tracking technology to record historical data and click history when employees access internal systems. It also collects text entered by users and interaction data, which it then transmits in real time to a server where it is prepared for analysis.

[0402] Learning areas of interest and emotional states

[0403] The server uses a generative AI model built with TensorFlow and PyTorch to learn employee interests and work-related information from historical data. It also uses an emotion engine using BERT and Keras to analyze users' text input and click behavior to identify their emotional state. For example, if an employee expresses emotions such as "excitement" or "interest," that information is identified by the emotion recognition model.

[0404] Providing notifications and recommendations

[0405] Based on the analysis results, the server scrutinizes newly collected information from internal bulletin boards and selects information that is important to employees. The timing and content of notifications are adjusted taking into account the results of the emotion engine. Notifications are sent to employees in real time using Firebase Cloud Messaging and WebSocket. These notifications are displayed as pop-ups on employees' PCs or smartphones. Additionally, information that is deemed to be of particular interest is displayed as recommended information on the dashboard that employees log in to.

[0406] Specific examples

[0407] Information collection and analysis

[0408] The server scrapes information from the company's internal bulletin board at 9 a.m. every morning, obtaining information such as new project recruitment, weekend club activities, and next week's drinking party. The obtained text is analyzed using spaCy and BERT and classified by category. For example, "project recruitment" is classified as business information, and "club activities" is classified as social information.

[0409] Learn and notify employee interests and emotional states

[0410] The device records the text input and click history when Employee A logs in and clicks on the new project recruitment page. The server analyzes the historical data using a generative AI model and learns that Employee A has a high interest in the "project recruitment." The emotion engine also detects "excitement" and "interest" from Employee A's input text. The next day, when a new project recruitment post is made, a notification is sent to Employee A's device using Firebase Cloud Messaging. Based on the results of the emotion engine, the notification message is adjusted to read, "We think this project might be interesting to you." A pop-up appears on Employee A's PC saying, "A new project recruitment post has been made," and the user can click the link to access more information.

[0411] Displaying recommendations

[0412] The server analyzes employee B's historical data and emotional state, and determines that he frequently views pages related to club activities and that he feels "excitement" and "expectations." Based on this, information on club activities that are of particular interest to employee B is selected and displayed on the dashboard every Monday as "Recommended club activities for this weekend." When employee B logs in on Monday, "Recommended club activities for this weekend" is displayed on the dashboard.

[0413] Prompt Sentence Examples

[0414] "Have a generative AI model perform sentiment analysis on food delivery requests and send relevant notifications for appropriate delivery jobs if the sentiment score is high."

[0415] In this way, the system of the present invention can quickly and efficiently provide employees with information necessary for their work and information of interest, and can also provide appropriate information according to the user's emotional state, thereby improving work efficiency and employee satisfaction.

[0416] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0417] Step 1:

[0418] Collect information from an internal bulletin board. The server uses a scraping tool such as Scrapy or Beautiful Soup to retrieve the contents of the internal bulletin board at a set time (for example, 9:00 every morning). The input is a URL and necessary initial information, and the output is the latest posted information as text data.

[0419] Step 2:

[0420] The collected information is analyzed and classified by category. The server uses natural language processing (NLP) technology to analyze the acquired text data and perform tokenization and parsing. Specifically, it uses spaCy and BERT to classify it into categories such as business information and social information. The input is the collected text data, and the output is information classified by category.

[0421] Step 3:

[0422] This system collects employee behavioral and emotional data. The device uses JavaScript and web tracking technology to record the browsing and click history of employees when they access internal systems in real time. The input is the employee's interactions (clicks, page transitions, etc.), and the output is the recorded behavioral data.

[0423] Step 4:

[0424] The collected history data and interaction data are sent to the server. The terminal transfers and stores the behavioral data in real time to the server. The input is the behavioral data recorded in real time, and the output is the behavioral data stored on the server side.

[0425] Step 5:

[0426] Analyze areas of interest and emotional state. The server uses a generative AI model (using TensorFlow or PyTorch) to analyze historical data and identify employees' areas of interest. It then uses an emotion engine (using BERT or Keras) to analyze emotional states from interaction data. The input is historical data and interaction data, and the output is the identified areas of interest and emotional state.

[0427] Step 6:

[0428] Important information is extracted and notification messages are generated. Based on the analysis results, the server scrutinizes the newly collected information from the internal bulletin board and selects information that is important to employees. It also adjusts the content of the notification message taking into account the results of the emotion engine. The input is the analyzed data and new internal bulletin board information, and the output is an optimized notification message.

[0429] Step 7:

[0430] Send notifications in real time. The server uses Firebase Cloud Messaging or WebSocket to send optimized notification messages to employee devices in real time. The input is the generated notification message, and the output is the notification displayed on the employee's device.

[0431] Step 8:

[0432] Display as recommended information. The server configures the information that is deemed to be of particular interest to be displayed on the employee's dashboard. The input is the identified areas of interest and selected information, and the output is the recommended information displayed on the employee's dashboard.

[0433] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0434] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0435] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0436] [Second embodiment]

[0437] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0438] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0439] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0440] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0441] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0442] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0443] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0444] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0445] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0446] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0447] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0448] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0449] The present invention provides a system for acquiring information from an internal bulletin board, extracting information related to employees' areas of interest or necessary for their work, and notifying them of the information. This system functions as follows.

[0450] System configuration and operation

[0451] Data collection and analysis

[0452] 1. (Server) The server periodically scrapes information from the internal bulletin board using a web scraping tool such as Beautiful Soup or Scrapy.

[0453] 2. (Server) The collected text data is analyzed using natural language processing (NLP) techniques. Specifically, NLP libraries such as spaCy and BERT are used to classify the text into categories (e.g., business information, social information).

[0454] Collecting employee behavior data

[0455] 3. (Device) When an employee logs in to an internal system and views information, their access history and click history are recorded. This is done using JavaScript and web tracking technology.

[0456] 4. (Terminal) The collected historical data is sent to the server in real time and stored in a database.

[0457] Learn about areas of interest and important information

[0458] 5. (Server) The server analyzes the collected employee history data and uses a generative AI model to learn each employee's areas of interest and work-related information. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used.

[0459] 6. (Server) Based on the analysis results, the newly collected information from the internal bulletin board is scrutinized and information that is important to employees is selected.

[0460] Displaying notifications and recommendations

[0461] 7. (Server) When important information is picked up, employees are notified in real time using a notification system (e.g., Firebase Cloud Messaging, WebSocket).

[0462] 8. (Device) A notification will pop up on the employee's PC or smartphone, providing a link to view more information.

[0463] 9. (Server) Furthermore, information that is deemed to be of particular interest is displayed as recommended information on the dashboard that is displayed when employees log in.

[0464] Specific examples

[0465] Example 1: Information collection and analysis

[0466] (Server) Every morning at 9:00, the server accesses the company bulletin board and scrapes the latest posts, which include information such as "New project recruitment," "Weekend club activities," and "Drinking parties next week."

[0467] In the (server) analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "project recruitment" is classified as business information, while "club activities" and "drinking parties" are classified as social information.

[0468] Example 2: Learning and notifying employees about information of interest

[0469] (Terminal) Employee A logs in to the system and the history of clicking on the page recruiting for new projects is recorded.

[0470] (Server) Based on this historical data, the AI ​​model learns that Employee A has a high interest in "Project Recruitment."

[0471] (Server) The next day, when a new project recruitment post is posted on the company bulletin board, the server picks up the information and notifies employee A's device via Firebase Cloud Messaging.

[0472] (Device) Employee A's PC displays a notification saying "A new project recruitment post has been made," and he can click on the link to view more information.

[0473] Example 3: Displaying recommendations

[0474] (Server) By analyzing employee B's history data, it is discovered that employee B frequently views pages related to club activities.

[0475] (Server) Based on the results, the AI ​​model picks out information about club activities that it determines Employee B will be particularly interested in, and displays this information on the dashboard every Monday as "Recommended club activities for this weekend."

[0476] (Device) When employee B logs in on Monday, the dashboard will display "Recommended club activities for this weekend."

[0477] In this way, the system of the present invention is designed to enable employees to quickly and effectively obtain information necessary for their work and information of interest, thereby improving work efficiency and employee satisfaction.

[0478] The processing flow will be explained below.

[0479] Step 1:

[0480] (Server) Set up regular access to an internal bulletin board. For example, access the bulletin board every morning at 9:00 using Scrapy or Beautiful Soup to collect the latest posts.

[0481] Step 2:

[0482] (Server) Analyze the collected text data using natural language processing (NLP) techniques. Specifically, use spaCy or BERT to tokenize the text and identify entities and classify them by category.

[0483] Step 3:

[0484] (Server) Analyzed text data is stored in a database by category, e.g., business information, social information, etc.

[0485] Step 4:

[0486] (Device) When an employee logs into the company's internal system, their access history and click history are recorded in real time using JavaScript and web tracking technology.

[0487] Step 5:

[0488] (Terminal) The recorded history data is sent to the server in real time.

[0489] Step 6:

[0490] (Server) Stores the submitted historical data in a database and prepares it for analysis.

[0491] Step 7:

[0492] (Server) Use a generative AI model to analyze collected historical data and learn employee interests and work-related information. Example: Use Python machine learning libraries (TensorFlow, PyTorch).

[0493] Step 8:

[0494] (Server) Based on the analysis results, the newly collected information from the internal bulletin board is scrutinized and important information for employees is selected, with priority given to information that is of high interest to employees.

[0495] Step 9:

[0496] (Server) The collected information is notified to employees in real time using Firebase Cloud Messaging or WebSocket.

[0497] Step 10:

[0498] (Device) A notification pops up on the employee's PC or smartphone. For example, a message saying "A new project recruitment post has been made" appears.

[0499] Step 11:

[0500] (Device) When an employee clicks on the pop-up notification, they are taken to a screen that displays more information.

[0501] Step 12:

[0502] (Server) In addition, prepare to display information that is thought to be of particular interest as recommended information on employees' dashboards.

[0503] Step 13:

[0504] (Device) When an employee logs in, recommended information is displayed on the dashboard. For example, information is presented as "recommended club activities for this weekend."

[0505] In this way, the system quickly and efficiently provides employees with the information they need and are most interested in, thereby improving work efficiency and employee satisfaction.

[0506] Example 1

[0507] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0508] Conventional internal information sharing systems have made it difficult to efficiently collect and provide information that employees need or are interested in. As a result, employees may miss important information or waste time on information that is not relevant to their work. This can lead to problems such as reduced work efficiency and lower employee satisfaction.

[0509] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0510] In this invention, the server includes means for acquiring information from an internal bulletin board, means for analyzing the acquired information and classifying it by category, means for collecting employees' browsing history and click history, means for analyzing the collected history information and learning employees' areas of interest, means for selecting and notifying employees of important information based on the analysis results, means for displaying information that is likely to be of particular interest to employees as recommended information on their dashboards, means for periodically collecting information using a scraping tool, means for analyzing text data using natural language processing technology and classifying it after tokenization and part-of-speech tagging, and means for machine learning employee areas of interest using a generative AI model. This makes it possible to quickly and effectively collect and provide information that employees need or are interested in.

[0511] An "internal bulletin board" is a digital bulletin board set up within a company or organization for the purpose of sharing information among employees.

[0512] "Means of obtaining information" refers to the mechanism by which the server automatically collects the content posted on the internal bulletin board.

[0513] "Means of analysis" refers to techniques that use natural language processing technology to analyze collected text data and understand its content.

[0514] "Means for categorizing by category" refers to a method for sorting analyzed text data into specific categories (e.g., business information, social information).

[0515] "Means for collecting employees' browsing and click histories" refers to technology for recording the operations performed by employees within the company's internal systems.

[0516] "Means for learning areas of interest" refers to a machine learning model that analyzes and learns from employee interests and concerns from collected historical data.

[0517] "Means of notification" refers to a system for informing employees of important information in real time based on the analysis results.

[0518] "Means for displaying recommended information on the dashboard" refers to an interface that displays information related to employees' areas of interest when they log in.

[0519] A "scraping tool" is software or a library used to automatically extract data from websites.

[0520] "Natural language processing technology" is a technology that enables computers to understand and process human language.

[0521] "Tokenization" refers to the process of dividing text data into smaller units such as words or sentences.

[0522] "Part-of-speech tagging" refers to labeling tokenized words with their parts of speech (e.g., nouns, verbs, adjectives).

[0523] A "generative AI model" is an artificial intelligence model that uses machine learning to learn patterns and rules from data and make predictions and classifications.

[0524] The present invention is a system that acquires information from an in-house bulletin board, extracts information that is relevant to employees' fields of interest or their work, and notifies them of the information. This system is configured as follows.

[0525] System configuration and operation procedures

[0526] This system is mainly composed of servers, terminals, and users, each of which has a specific role and functions in cooperation with each other.

[0527] Data collection and analysis methods

[0528] The server periodically scrapes information from the company's internal bulletin board. The scraping tools used are Beautiful Soup and Scrapy. The collected data is then analyzed using natural language processing (NLP) techniques. Specifically, the NLP libraries spaCy and BERT are used to tokenize the text, tag parts of speech, and analyze dependencies. Finally, the text data is classified by category (e.g., business information, social information).

[0529] How to collect employee behavioral data

[0530] (Device) records access history and click history every time an employee logs in to the company's internal system or browses information. This is done using JavaScript and web tracking technology. The collected history data is sent to (server) in real time and stored in a database.

[0531] Areas of interest and how to learn important information

[0532] The server analyzes the collected employee behavioral data and uses a generative AI model to learn each employee's areas of interest and work-related information. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used. The AI ​​model learns patterns and areas of interest based on this data.

[0533] How notifications and recommendations are displayed

[0534] (Server) scrutinizes newly collected information from internal bulletin boards based on the analysis results and selects information that is important to employees. (Server) notifies employees in real time using a notification system (e.g., Firebase Cloud Messaging, WebSocket). (Terminal) displays a notification as a pop-up on employees' PCs or smartphones and provides a link to view more information. (Server) also displays information of particular interest as recommended information on the dashboard that is displayed when employees log in.

[0535] Specific examples

[0536] Example 1: Information collection and analysis

[0537] (Server) accesses the company's internal bulletin board at 9:00 every morning and scrapes the latest posts. For example, this includes information such as "new project recruitment," "weekend club activities," and "drinking party next week." In the analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "project recruitment" is classified as business information, while "club activities" and "drinking party" are classified as social information.

[0538] Example 2: Learning and notifying employees about information of interest

[0539] On (terminal), employee A logs in to the system and records the history of clicking on the new project recruitment page. Based on this history data, (server) uses its AI model to learn that employee A has a high interest in "project recruitment." The next day, when a new project recruitment post is posted on the company bulletin board, (server) picks up the information and notifies employee A's device via Firebase Cloud Messaging. (terminal) displays a notification on employee A's PC stating "A new project recruitment post has been made," and employee A can click on the link to view more information.

[0540] Example 3: Displaying recommendations

[0541] When (Server) analyzes employee B's historical data, it finds that employee B frequently views pages related to club activities. Based on the results, (Server) selects club activity information that the AI ​​model determines to be of particular interest to employee B, and displays this information on the dashboard every Monday as "Recommended club activities for this weekend." When employee B logs in on Monday, (Device) displays "Recommended club activities for this weekend" on the dashboard.

[0542] This system will enable employees to obtain the information they need quickly and efficiently, which is expected to improve work efficiency and employee satisfaction.

[0543] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0544] Step 1:

[0545] (Server) accesses the company bulletin board every morning at 9:00 and scrapes new posts. The URL of the bulletin board is given as input, and the HTML structure is analyzed using a scraping tool (Beautiful Soup or Scrapy). The output is the extracted text data. Specifically, it performs the following operations:< / url:> , , ) to extract the text.

[0546] Step 2:

[0547] The text data acquired by the server is preprocessed using natural language processing (NLP) techniques. The input is the raw data acquired in step 1, and the output is the preprocessed text data. Specifically, NLP libraries (spaCy and BERT) are used to tokenize the data and remove unnecessary tags and spaces.

[0548] Step 3:

[0549] The server tokenizes the preprocessed text data and performs part-of-speech tagging and dependency analysis. The input is the data preprocessed in step 2, and the output is tokenized and tagged text. Specifically, it uses spaCy to perform structural analysis of the sentence and tag each word with a part-of-speech tag.

[0550] Step 4:

[0551] The server classifies the analyzed text data into categories (business information, social information). The input is the text tagged with part-of-speech tags in step 3, and the output is the classified data. Specifically, it uses a machine learning algorithm to classify each text into a predefined category.

[0552] Step 5:

[0553] (Terminal) records logins to the company's internal system, access history, and click history. The input is the user's operation log, and the output is the recorded access data. Specifically, JavaScript code captures the user's operations and sends them to the server.

[0554] Step 6:

[0555] The historical data collected by (the terminal) is sent to the server in real time and stored in the database. The input is the access data recorded in step 5, and the output is the historical data stored on the server. Specifically, the data is sent to the server via an HTTP request and stored in the database.

[0556] Step 7:

[0557] The server analyzes the collected employee behavioral data and uses a generative AI model to learn each employee's areas of interest and work-related information. The input is the historical data saved in step 6, and the output is the learned model and interest information. Specifically, the AI ​​model is trained using TensorFlow or PyTorch to learn patterns from the data.

[0558] Step 8:

[0559] (Server) scrutinizes the newly collected information from the internal bulletin board and picks out information that is important to employees. The input is the data from Step 1 and Step 7, and the output is the picked-up important information. Specifically, it filters the information based on the prediction results of the AI ​​model.

[0560] Step 9:

[0561] The server notifies employees in real time using a notification system (Firebase Cloud Messaging or WebSocket). The input is the information picked up in step 8, and the output is a notification message. Specifically, the server sends a message in real time through the notification system.

[0562] Step 10:

[0563] The device displays a pop-up notification and provides a link to view more information. The input is the notification message sent in step 9, and the output is the user's browsing behavior. The specific behavior is to display a pop-up window in a browser or application.

[0564] Step 11:

[0565] Information that the server determines to be of particular interest is displayed as recommended information on the dashboard when employees log in. The input is the data from steps 7 and 8, and the output is the recommended information displayed on the dashboard. Specifically, the settings are configured to list and display specific information on the dashboard.

[0566] Through the above processing steps, a system is provided that allows employees to quickly and efficiently obtain the information they need or are interested in.

[0567] (Application example 1)

[0568] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0569] Conventional systems that collect information from internal bulletin boards and notify employees have difficulty in properly identifying employees' areas of interest and necessary information. Furthermore, at production sites, proper information collection and notification is not performed, making it difficult to provide efficient work support or timely maintenance. For this reason, there is a need for a system that efficiently collects production schedules and maintenance information at factories and notifies workers appropriately.

[0570] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0571] In this invention, the server includes means for acquiring information from an in-house bulletin board, means for analyzing the acquired information and classifying it by category, means for collecting employees' browsing history and click history, means for analyzing the collected history information and learning employees' areas of interest, means for selecting and notifying employees of important information based on the analysis results, means for displaying information that is likely to be of particular interest as recommended information on the employee's dashboard, means for scraping production plans and function check information from an in-house management system, and means for analyzing the collected information and notifying it in conjunction with sensor data and work logs acquired during work. This allows employees and workers to quickly and efficiently receive the information they need, improving work efficiency and enabling appropriate maintenance.

[0572] An "internal bulletin board" is an electronic bulletin board system set up within a company for the purpose of sharing information and communicating.

[0573] "Means of obtaining information" refers to the methods and technologies used to collect the necessary information from various data sources.

[0574] "Means for analyzing information and classifying it into categories" refers to technology that analyzes acquired data and classifies it into specific categories based on its content.

[0575] "Means of collecting employees' browsing and click histories" refers to technology that collects the behavioral history of employees when they use web systems.

[0576] "Means of analyzing collected historical information to learn employees' areas of interest" refers to technology that analyzes collected behavioral data to learn what information a particular employee is interested in.

[0577] "Means of selecting and notifying employees of important information based on the analysis results" refers to methods and technologies for selecting important information from the analysis results and notifying employees of it.

[0578] "Means for displaying recommended information on employee dashboards" refers to a method for displaying information that is determined to be of particular interest to employees on their interface.

[0579] "Factory management system" refers to a comprehensive system for managing factory production and equipment.

[0580] "Means for scraping production plan and function check information" refers to technology that automatically extracts necessary information from factory management systems.

[0581] "Sensor data and work logs obtained during work" refers to the various data generated by robots and workers while they are working, as well as the recorded work history.

[0582] "Generative AI models" refer to algorithms and technologies that use machine learning techniques to generate knowledge and predictive models from data.

[0583] A "prompt" is an instruction given to a generative AI model to make it perform a specific task.

[0584] A "notification system" refers to technology that notifies individual devices of important information in real time.

[0585] The following describes an embodiment of the present invention. The present invention is a system that acquires information from a factory management system, notifies robots and workers of appropriate information, and improves work efficiency. This system functions as follows.

[0586] System configuration and operation

[0587] Data collection and analysis

[0588] 1. Server: The server periodically scrapes production plans and functional check information from the factory management system using web scraping tools such as Beautiful Soup and Scrapy.

[0589] 2. Server: The collected text data is analyzed using natural language processing (NLP) techniques. Specifically, NLP libraries such as spaCy and BERT are used to classify the text into categories (e.g., production information, maintenance information).

[0590] User behavior data collection

[0591] 3. Robot: The sensor data and work logs acquired by the robot while working are recorded and sent to the server in real time. The data is collected using JavaScript and web tracking technology.

[0592] 4. Server: The collected historical data is stored on the server.

[0593] Learn about areas of interest and important information

[0594] 5. Server: The server analyzes the collected robot history data and uses generative AI models to learn important information for each robot and worker. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used.

[0595] 6. Server: Based on the analysis results, the newly collected information from the factory management system is scrutinized and important information is picked up.

[0596] Displaying notifications and recommendations

[0597] 7. Server: When important information is picked up, it is notified to the robot or worker in real time using a notification system (e.g. MQTT, WebSocket).

[0598] 8. Robot: Notifications are sent via the robot's display and voice output, which can be checked by the worker. Work instructions and changes are also displayed here.

[0599] Specific examples

[0600] Example 1: Information collection and analysis

[0601] Server: Every morning at 9:00, the server accesses the factory management system and scrapes the latest production schedule and maintenance schedule, including information such as "new production line setup," "machine maintenance next month," and "equipment inspection this weekend."

[0602] Server: In the analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "production line settings" is classified as production information, while "machine maintenance" and "equipment inspection" are classified as maintenance information.

[0603] Example 2: Learning and notifying robot behavior information

[0604] Robot: Sensor data and work logs acquired by specific robots while performing machine maintenance are recorded.

[0605] Server: Based on this historical data, the AI ​​model learns the information necessary for the robot.

[0606] Server: When new machine maintenance information is posted on the factory management system, the server picks up the information and notifies the robot via MQTT.

[0607] Robot: A notification saying "New machine maintenance information has been added" appears on the robot's display and can be confirmed by the worker.

[0608] Example 3: Displaying recommendations

[0609] Server: By analyzing historical data for a particular robot, you may discover that the robot frequently performs tasks on a particular production line setup.

[0610] Server: Based on the results, the AI ​​model picks out information about the production line settings that it deems particularly important for that robot.

[0611] Server: Every morning, the server notifies the robot of this information and displays it on the robot's display as "Today's recommended production line settings."

[0612] In this way, the system of the present invention is designed to enable robots and workers to quickly and effectively obtain the information they need for their work, thereby improving work efficiency and ensuring appropriate maintenance.

[0613] Prompt Sentence Examples

[0614] "Next week's machine maintenance schedule"

[0615] "New Production Plan for Project X"

[0616] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0617] Step 1:

[0618] The server acquires production plan information and maintenance schedule information from the factory management system.

[0619] Input: Production plan information and maintenance schedule information from the factory management system.

[0620] Data processing: Use web scraping tools such as Beautiful Soup or Scrapy to extract the necessary information and save it as text data.

[0621] Output: Text data of extracted production plan information and maintenance schedule information.

[0622] Step 2:

[0623] The server analyzes the acquired text data and classifies it into categories.

[0624] Input: The text data extracted in step 1.

[0625] Data processing: Using natural language processing (NLP) techniques, we tokenize the text and classify it into categories (e.g., production information, maintenance information). Specifically, we use NLP libraries such as spaCy and BERT.

[0626] Output: Text data classified by category.

[0627] Step 3:

[0628] The robot records sensor data and work logs acquired during work and sends them to a server.

[0629] Input: Sensor data and work logs acquired while the robot is working.

[0630] Data processing: Collect data in real time and send it to the server. Record data using JavaScript and web tracking technology.

[0631] Output: Sensor data and work logs sent to the server.

[0632] Step 4:

[0633] The server analyzes the collected robot history data and uses a generative AI model to learn important information.

[0634] Input: Sensor data and work logs sent in step 3.

[0635] Data processing: Using Python machine learning libraries (e.g., TensorFlow, PyTorch), the data is analyzed to learn information that is important to each robot.

[0636] Output: Key information identified by the generative AI model.

[0637] Step 5:

[0638] The server picks out important information based on the analysis results and notifies you.

[0639] Input: Key information learned in step 4.

[0640] Data processing: Newly collected production plans and maintenance information are compared with the analysis results to select important information.

[0641] Output: Important information picked up.

[0642] Step 6:

[0643] The server uses a notification system to notify robots and workers of important information in real time.

[0644] Input: Important information picked up in step 5.

[0645] Data processing: Use MQTT or WebSocket to send notifications in real time.

[0646] Output: Notification message that will be displayed to the robot and worker.

[0647] Step 7:

[0648] The robot will display the notified information on a screen or output it as audio so that the worker can check it.

[0649] Input: Information provided in step 6.

[0650] Data processing: Display and audio output are performed so that workers can check the information.

[0651] Output: Notifications are displayed on the robot's display and audio output.

[0652] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0653] The following describes an embodiment of the present invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the areas of interest and emotional state of employees. The system is operated by combining a generative AI and an emotion engine.

[0654] System configuration and operation

[0655] Data collection and analysis

[0656] 1. (Server) Regularly scrape information from internal bulletin boards. For example, use tools like Scrapy or Beautiful Soup to collect the latest posts every morning at 9am.

[0657] 2. (Server) Analyze the collected text data using natural language processing (NLP) techniques, such as spaCy or BERT, to tokenize the text and classify it by category (e.g., business information, social information).

[0658] Collecting employee behavioral and emotional data

[0659] 3. (Device) Every time an employee logs in to an internal system and views information, their access history and click history are recorded. This is done using JavaScript and web tracking technology.

[0660] 4. (Device) Obtain recorded history data, as well as user-entered text and interaction data.

[0661] 5. (Terminal) Send the collected data to the server in real time.

[0662] Learning areas of interest and emotional states

[0663] 6. (Server) Stores the submitted historical and interaction data in a database and prepares it for analysis.

[0664] 7. (Server) Use a generative AI model to analyze historical data to learn employee interests and job-related information. Use Python machine learning libraries (e.g., TensorFlow, PyTorch).

[0665] 8. (Server) An emotion engine is used to analyze the user's text input and click behavior to identify their emotional state. Emotion recognition technology combines NLP technology with machine learning models (e.g., BERT, Keras).

[0666] Providing notifications and recommendations

[0667] 9. (Server) Based on the learning results, the server scrutinizes the newly collected information from the internal bulletin board and selects information that is important to employees. The server also adjusts the timing and content of notifications based on the results of the emotion engine.

[0668] 10. (Server) Use Firebase Cloud Messaging or WebSocket to send notifications to employees in real time.

[0669] 11. (Device) Notifications are displayed as pop-ups on employees' PCs or smartphones. The pop-ups display specific messages such as "A new project recruitment post has been made," and employees can click on a link to access more information.

[0670] 12. (Server) Information that is thought to be of particular interest is displayed as recommended information on the dashboard that employees log in to.

[0671] Specific examples

[0672] Example 1: Information collection and analysis

[0673] (Server) Every morning at 9:00, information is scraped from the company bulletin board to obtain information such as "New project recruitment," "Weekend club activities," and "Drinking parties next week."

[0674] (Server) Analyze the acquired text using spaCy or BERT and classify it into categories. For example, "Project recruitment" is classified as business information, while "Club activities" and "Drinking parties" are classified as social information.

[0675] Example 2: Learning and notifying employee interests and emotional states

[0676] (Device) Employee A logs in and clicks on the new project recruitment page. The text input and click history are recorded.

[0677] (Server) The generative AI model analyzes the historical data and learns that Employee A has a high interest in the "Project Recruitment" campaign. The emotion engine also detects "excitement" and "interest" from Employee A's input text.

[0678] (Server) The next day, if a new project recruitment post is made, a notification is sent to Employee A's device using Firebase Cloud Messaging. Based on the results of the emotion engine, the notification message is adjusted to read, "This project seems interesting to you."

[0679] (Device) A pop-up message appears on Employee A's PC saying "A new project recruitment post has been made," and clicking the link displays more information.

[0680] Example 3: Displaying recommendations

[0681] (Server) Analyzes employee B's historical data and emotional state and finds that he frequently views pages related to club activities and feels "excited" and "expected."

[0682] (Server) Based on this, information on club activities that are of particular interest to employee B is selected and displayed on the dashboard every Monday as "Recommended club activities for this weekend."

[0683] (Device) When employee B logs in on Monday, the dashboard will display "Recommended club activities for this weekend."

[0684] In this way, the system of the present invention provides employees with information necessary for their work and information of interest quickly and efficiently, and also provides appropriate information according to the user's emotional state, thereby improving work efficiency and employee satisfaction.

[0685] The processing flow will be explained below.

[0686] Step 1:

[0687] (Server) Set up periodic access to the internal bulletin board. For example, use Scrapy or Beautiful Soup to collect the latest postings every morning at 9am.

[0688] Step 2:

[0689] (Server) Analyze the collected text data using natural language processing (NLP) technology. Specifically, use spaCy or BERT to tokenize the text and classify it by category (e.g., business information, social information).

[0690] Step 3:

[0691] (Server) The analyzed text data is stored in a database by category. The stored data is used for later information review.

[0692] Step 4:

[0693] (Device) When an employee logs in to the company's internal system, JavaScript and web tracking technology are used to record access history and click history.

[0694] Step 5:

[0695] (Device) Recorded historical data is sent to the server in real time, making it immediately available for analysis.

[0696] Step 6:

[0697] (Server) Stores the submitted historical data in a database so that it can be used by the generative AI model and emotion engine for analysis.

[0698] Step 7:

[0699] (Server) Analyze historical data using a generative AI model to learn employee interests and work-related information. Build the learning model using Python machine learning libraries (e.g., TensorFlow, PyTorch).

[0700] Step 8:

[0701] (Server) An emotion engine is used to analyze user input data and click history to identify the user's emotional state. For example, NLP techniques and machine learning models are combined to identify emotions such as "excitement," "interest," and "boredom" from user text input.

[0702] Step 9:

[0703] (Server) Based on the analysis results, the server scrutinizes newly collected information from internal bulletin boards and selects information that is important to employees. At this time, the server adjusts the timing and content of notifications according to the user's emotional state as determined by the emotion engine.

[0704] Step 10:

[0705] (Server) Notifications are sent to employees in real time using Firebase Cloud Messaging and WebSockets. The content and format of the notifications are customized taking into account the results of the emotion engine.

[0706] Step 11:

[0707] (Device) A notification pops up on the employee's PC or smartphone. For example, a message saying "A new project recruitment post has been made" appears. When the user clicks on the link, more information is displayed.

[0708] Step 12:

[0709] (Server) In addition, prepare to display information that is thought to be of particular interest as "recommended information" on employees' dashboards.

[0710] Step 13:

[0711] (Device) When an employee logs in, "recommended information" is displayed on the dashboard. For example, information such as "recommended club activities for this weekend" is displayed.

[0712] This system allows employees to quickly and efficiently obtain information necessary for their work and information of interest. In addition, because it provides appropriate information based on the user's emotional state, it is expected to improve work efficiency and employee satisfaction.

[0713] Example 2

[0714] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0715] Conventional internal information sharing systems make it difficult for employees to efficiently obtain the information they need, resulting in problems such as missing information or failing to grasp important information. Furthermore, to improve the quality of communication within a company and increase the work efficiency of each employee, it is necessary to provide information that takes into account the interests and emotional state of employees, but there were few systems that could achieve this. Furthermore, the lack of information provided that was tailored to the interests of each employee sometimes resulted in information overload, which in turn reduced productivity.

[0716] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0717] In this invention, the server includes a means for acquiring information from the in-house bulletin board, a means for analyzing the acquired information using natural language processing technology and classifying it into categories, and a means for collecting collected history information and user input text in real time. This makes it possible to efficiently collect information needed by employees and provide important information at an appropriate time, taking into account the employees' areas of interest and emotional state.

[0718] 1. An "internal bulletin board" is an online platform for sharing information within a company or organization.

[0719] 2. "Means of obtaining information" refers to tools and programs for automatically collecting necessary information from internal bulletin boards.

[0720] 3. "Natural language processing technology" is an artificial intelligence technology for analyzing and understanding human language.

[0721] 4. "Categorization methods" refers to algorithms or programs that group collected information based on specific criteria.

[0722] 5. "Browsing history" refers to a record of web pages and documents that a user has previously viewed.

[0723] 6. "Click History" refers to a record of links and buttons that a user has previously clicked.

[0724] 7. "History information" refers to all data relating to a user's behavioral history.

[0725] 8. "User Input Text" means any characters or sentences entered by a User into the System.

[0726] 9. "Real-time collection means" refers to technologies or methods that capture data as it is transmitted.

[0727] 10. "Database" refers to a system for efficiently storing, managing, and retrieving large amounts of data.

[0728] 11. "Generative AI model" refers to an artificial intelligence model that learns from generated data and performs tasks such as prediction and classification.

[0729] 12. "Emotion Engine" refers to technology or programs that identify a user's emotional state from their text input or behavior.

[0730] 13. "Interest Areas" refers to the themes or topics that interest a particular user.

[0731] 14. "Business-Related Information" means information related to a User's job function.

[0732] 15. "Means of notification" refers to tools and technologies used to communicate information to users.

[0733] 16. "Recommended Information" refers to the provision of specific information selected based on a user's interests and behavior.

[0734] The following describes an embodiment of the present invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the areas of interest and emotional state of employees. The system is operated by combining a generative artificial intelligence model and an emotion analysis engine.

[0735] System configuration and operation

[0736] Data collection and analysis

[0737] 1. The server periodically retrieves information from the company bulletin board. To collect the information, use a web scraping tool such as Scrapy or Beautiful Soup. For example, set it to collect the latest posts every morning at 9:00.

[0738] 2. The server analyzes the collected text data using natural language processing (NLP) techniques, such as spaCy and BERT, to tokenize the text and classify it into categories (e.g., business information, social information).

[0739] Collecting employee behavioral and emotional data

[0740] 3. The device records access and click histories every time an employee logs into the company's internal systems and views information. This recording is done using JavaScript and web tracking technology.

[0741] 4. The device acquires the recorded history data, text entered by the user, and interaction data, and transmits them to the server in real time.

[0742] Saving data and preparing for analysis

[0743] 5. The server stores the transmitted history and interaction data in a database and prepares it for analysis, using a database system such as PostgreSQL or MongoDB.

[0744] Learning areas of interest and emotional states

[0745] 6. The server uses generative artificial intelligence models (e.g., TensorFlow, PyTorch) to analyze historical data and learn employee interests and work-related information.

[0746] 7. The server uses an emotion engine (e.g., BERT, Keras) to identify the user's emotional state from their text input and click behavior.

[0747] Providing notifications and recommendations

[0748] 8. Based on the learning results and sentiment analysis results, the server examines the newly collected information from the internal bulletin board and notifies employees of important information. For example, it prepares a tailored message such as, "This project seems interesting to you."

[0749] 9. The server uses Firebase Cloud Messaging or WebSocket to send notifications in real time to employee devices, which are displayed as pop-ups on PCs and smartphones.

[0750] 10. The server displays information of particular interest as recommended information on the dashboard screen when employees log in. For example, information on club activities of high interest is displayed on the dashboard every Monday as "Recommended activities for this weekend."

[0751] Specific examples

[0752] Example 1: Information collection and analysis

[0753] The server scrapes information from the company's internal bulletin boards every morning at 9 a.m., obtaining information such as "new project recruitment," "weekend social events," and "next week's meetings."

[0754] The server analyzes the acquired text using natural language processing technology (spaCy or BERT) and classifies it into categories. For example, "project recruitment" is classified as business information, while "social events" and "meetings" are classified as social information.

[0755] Example 2: Learning and notifying employee interests and emotional states

[0756] The device records access history and click history when employee A logs in and clicks on the page recruiting for new projects.

[0757] The server analyzes the historical data based on a generative artificial intelligence model and learns that Employee A has a high interest in the "Project Recruitment." The emotion engine also detects "excitement" and "interest" from Employee A's input text.

[0758] The next day, if there is a new project opening, the server will use Firebase Cloud Messaging to notify employee A's device, saying, "We think this project might be interesting to you."

[0759] The device displays a pop-up on Employee A's PC saying, "A new project recruitment post has been made," and when Employee A clicks on the link, more information will be displayed.

[0760] Example 3: Displaying recommendations

[0761] The server analyzes employee B's historical data and emotional state, and determines that he frequently views pages related to club activities and feels emotions of "excitement" and "expectation."

[0762] Based on this, the server selects information about club activities that employee B is particularly interested in and displays it on the dashboard every Monday as "Recommended club activities for this weekend."

[0763] When employee B logs in to the device on Monday, the dashboard will display "Recommended club activities for this weekend."

[0764] Example prompt sentence:

[0765] "Collect information on new project openings and notify employees."

[0766] "Identify areas of interest from employees' click history"

[0767] "Analyze the user's emotional state using an emotion engine and provide appropriate notifications."

[0768] This system allows employees to quickly and efficiently receive not only the information they need for their work, but also appropriate information tailored to their individual interests and emotional state, thereby improving work efficiency and employee satisfaction.

[0769] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0770] Step 1:

[0771] The server scrapes information from the company bulletin board every morning at 9:00. It uses the bulletin board URL as input and generates data including the title, body, and posting date and time of each post as output. Specifically, it uses Scrapy and Beautiful Soup to parse the HTML and extract the required text information.

[0772] Step 2:

[0773] The server analyzes the collected text data using natural language processing technology and classifies it into categories. It uses the acquired text data as input and obtains data classified into categories (e.g., business information, social information) as output. Specifically, it uses spaCy and BERT to tokenize the text and perform contextual analysis.

[0774] Step 3:

[0775] The device records access and click histories every time an employee logs into the company's internal systems and browses information. It uses the employee's browsing and clicking behavior as input and generates the recorded history data as output. Specifically, it captures user actions using JavaScript and web tracking technology.

[0776] Step 4:

[0777] The device sends the recorded history data and the user's input text and interaction data to the server in real time. The device uses the recorded history data and the user's text data as input and generates the data sent to the server as output. Specifically, the device converts the data format to JSON or similar and sends it to the server using the HTTPS protocol.

[0778] Step 5:

[0779] The server stores the submitted historical and interaction data in a database and prepares it for analysis. It uses the received historical and interaction data as input and generates the data stored in the database as output, specifically by inserting data using database queries.

[0780] Step 6:

[0781] The server uses a generative AI model to analyze historical data and learn employee interests and work-related information. It uses the stored historical data as input and obtains data identifying employee interests as output. Specifically, it trains the model using TensorFlow or PyTorch and makes predictions based on the historical data.

[0782] Step 7:

[0783] The server uses an emotion engine to identify the user's emotional state from their text input and click behavior. It uses the user's text data and click data as input and obtains the identified emotional state as output. Specifically, it uses BERT and Keras to perform emotion recognition.

[0784] Step 8:

[0785] Based on the learning results and sentiment analysis results, the server scrutinizes the newly collected information from the internal bulletin board, picks out information that is important to employees, and prepares the notification content. It uses the data obtained from the internal bulletin board and the analysis results as input, and generates a notification message as output. Specifically, it selects an appropriate template and creates the message.

[0786] Step 9:

[0787] The server sends notifications to employee devices in real time using Firebase Cloud Messaging or WebSocket. It uses the prepared notification message as input and gets the sent notification as output. Specifically, it sends messages using the Firebase API or WebSocket protocol.

[0788] Step 10:

[0789] The device displays the received notification as a pop-up on the employee's PC or smartphone. It uses the received notification message as input and obtains the displayed pop-up notification as output. Specifically, it uses the browser's Notification API or the OS's notification function to display the message.

[0790] Step 11:

[0791] The server then displays the most interesting information as recommendations on the employee's dashboard. It uses information based on the areas of interest as input and generates recommendations that are displayed on the dashboard as output. Specifically, it dynamically updates the content of the web page and applies templates to display the information of interest.

[0792] (Application example 2)

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

[0794] Conventional internal bulletin board systems and notification systems do not provide information that takes into account the interests and emotional state of individual employees, limiting their ability to improve work efficiency and employee satisfaction. Furthermore, it is difficult to provide employees with the important information they need at the right time, leading to problems with missed information and excessive notifications. There is a need for a system that can solve these issues and provide optimal information to employees while improving work efficiency.

[0795] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring information from the in-house bulletin board, means for analyzing the acquired information and classifying it by category, and means for analyzing the collected history data and interaction data to identify the user's emotional state. This makes it possible to pick out important information based on the employee's areas of interest and emotional state, appropriately adjust the content of notification messages, and provide them to employees in real time.

[0796] An "internal bulletin board" is an online bulletin board for the purpose of sharing information and communicating among employees.

[0797] "Means of obtaining information" refers to the techniques and methods used to collect the necessary information from the internal bulletin board.

[0798] "Means of analyzing information" refers to the techniques and methods used to structure collected information and convert it into an easily understandable format.

[0799] "Categorization means" refers to techniques or methods for classifying analyzed information into different categories.

[0800] "Means of collecting employee browsing and click history" refers to technologies and methods that record what information employees view and what links they click.

[0801] "Means for analyzing historical information and learning about employees' areas of interest" refers to techniques and methods for identifying employees' areas of interest and concern from collected historical information.

[0802] "Means for analyzing interaction data and identifying the user's emotional state" refers to technologies and methods for analyzing the emotions of employees at that time based on their interaction data.

[0803] "Means of notification" refers to the technology and methods used to inform employees of necessary information based on the analysis results.

[0804] "Means for adjusting the content of notification messages" refers to techniques and methods for changing notification messages to optimal content depending on the emotional state of employees.

[0805] "Means for displaying recommended information on the dashboard" refers to the technology and methods for displaying information of particular interest on employees' operation screens.

[0806] "Means of real-time notification" refers to technologies and methods that instantly inform employees of the information they need at that moment.

[0807] The following describes an embodiment of this invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the employee's areas of interest and emotional state. The system is operated by combining a generative AI and an emotion engine.

[0808] System configuration and operation

[0809] Data collection and analysis

[0810] The server periodically collects information from internal bulletin boards using scraping technologies such as Scrapy and Beautiful Soup. The collected information is analyzed and categorized using natural language processing (NLP) techniques. Specifically, text is tokenized using spaCy and BERT, and then classified into categories such as business information and social information.

[0811] Collecting employee behavioral and emotional data

[0812] The device uses JavaScript and web tracking technology to record historical data and click history when employees access internal systems. It also collects text entered by users and interaction data, which it then transmits in real time to a server where it is prepared for analysis.

[0813] Learning areas of interest and emotional states

[0814] The server uses a generative AI model built with TensorFlow and PyTorch to learn employee interests and work-related information from historical data. It also uses an emotion engine using BERT and Keras to analyze users' text input and click behavior to identify their emotional state. For example, if an employee expresses emotions such as "excitement" or "interest," that information is identified by the emotion recognition model.

[0815] Providing notifications and recommendations

[0816] Based on the analysis results, the server scrutinizes newly collected information from internal bulletin boards and selects information that is important to employees. The timing and content of notifications are adjusted taking into account the results of the emotion engine. Notifications are sent to employees in real time using Firebase Cloud Messaging and WebSocket. These notifications are displayed as pop-ups on employees' PCs or smartphones. Additionally, information that is deemed to be of particular interest is displayed as recommended information on the dashboard that employees log in to.

[0817] Specific examples

[0818] Information collection and analysis

[0819] The server scrapes information from the company's internal bulletin board at 9 a.m. every morning, obtaining information such as new project recruitment, weekend club activities, and next week's drinking party. The obtained text is analyzed using spaCy and BERT and classified by category. For example, "project recruitment" is classified as business information, and "club activities" is classified as social information.

[0820] Learn and notify employee interests and emotional states

[0821] The device records the text input and click history when Employee A logs in and clicks on the new project recruitment page. The server analyzes the historical data using a generative AI model and learns that Employee A has a high interest in the "project recruitment." The emotion engine also detects "excitement" and "interest" from Employee A's input text. The next day, when a new project recruitment post is made, a notification is sent to Employee A's device using Firebase Cloud Messaging. Based on the results of the emotion engine, the notification message is adjusted to read, "We think this project might be interesting to you." A pop-up appears on Employee A's PC saying, "A new project recruitment post has been made," and the user can click the link to access more information.

[0822] Displaying recommendations

[0823] The server analyzes employee B's historical data and emotional state, and determines that he frequently views pages related to club activities and that he feels "excitement" and "expectations." Based on this, information on club activities that are of particular interest to employee B is selected and displayed on the dashboard every Monday as "Recommended club activities for this weekend." When employee B logs in on Monday, "Recommended club activities for this weekend" is displayed on the dashboard.

[0824] Prompt Sentence Examples

[0825] "Have a generative AI model perform sentiment analysis on food delivery requests and send relevant notifications for appropriate delivery jobs if the sentiment score is high."

[0826] In this way, the system of the present invention can quickly and efficiently provide employees with information necessary for their work and information of interest, and can also provide appropriate information according to the user's emotional state, thereby improving work efficiency and employee satisfaction.

[0827] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0828] Step 1:

[0829] Collect information from an internal bulletin board. The server uses a scraping tool such as Scrapy or Beautiful Soup to retrieve the contents of the internal bulletin board at a set time (for example, 9:00 every morning). The input is a URL and necessary initial information, and the output is the latest posted information as text data.

[0830] Step 2:

[0831] The collected information is analyzed and classified by category. The server uses natural language processing (NLP) technology to analyze the acquired text data and perform tokenization and parsing. Specifically, it uses spaCy and BERT to classify it into categories such as business information and social information. The input is the collected text data, and the output is information classified by category.

[0832] Step 3:

[0833] This system collects employee behavioral and emotional data. The device uses JavaScript and web tracking technology to record the browsing and click history of employees when they access internal systems in real time. The input is the employee's interactions (clicks, page transitions, etc.), and the output is the recorded behavioral data.

[0834] Step 4:

[0835] The collected history data and interaction data are sent to the server. The terminal transfers and stores the behavioral data in real time to the server. The input is the behavioral data recorded in real time, and the output is the behavioral data stored on the server side.

[0836] Step 5:

[0837] Analyze areas of interest and emotional state. The server uses a generative AI model (using TensorFlow or PyTorch) to analyze historical data and identify employees' areas of interest. It then uses an emotion engine (using BERT or Keras) to analyze emotional states from interaction data. The input is historical data and interaction data, and the output is the identified areas of interest and emotional state.

[0838] Step 6:

[0839] Important information is extracted and notification messages are generated. Based on the analysis results, the server scrutinizes the newly collected information from the internal bulletin board and selects information that is important to employees. It also adjusts the content of the notification message taking into account the results of the emotion engine. The input is the analyzed data and new internal bulletin board information, and the output is an optimized notification message.

[0840] Step 7:

[0841] Send notifications in real time. The server uses Firebase Cloud Messaging or WebSocket to send optimized notification messages to employee devices in real time. The input is the generated notification message, and the output is the notification displayed on the employee's device.

[0842] Step 8:

[0843] Display as recommended information. The server configures the information that is deemed to be of particular interest to be displayed on the employee's dashboard. The input is the identified areas of interest and selected information, and the output is the recommended information displayed on the employee's dashboard.

[0844] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0845] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0846] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0847] [Third embodiment]

[0848] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0849] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0850] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0851] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0852] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0853] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0854] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0855] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0856] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0857] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0858] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0859] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0860] The present invention provides a system for acquiring information from an internal bulletin board, extracting information related to employees' areas of interest or necessary for their work, and notifying them of the information. This system functions as follows.

[0861] System configuration and operation

[0862] Data collection and analysis

[0863] 1. (Server) The server periodically scrapes information from the internal bulletin board using a web scraping tool such as Beautiful Soup or Scrapy.

[0864] 2. (Server) The collected text data is analyzed using natural language processing (NLP) techniques. Specifically, NLP libraries such as spaCy and BERT are used to classify the text into categories (e.g., business information, social information).

[0865] Collecting employee behavior data

[0866] 3. (Device) When an employee logs in to an internal system and views information, their access history and click history are recorded. This is done using JavaScript and web tracking technology.

[0867] 4. (Terminal) The collected historical data is sent to the server in real time and stored in a database.

[0868] Learn about areas of interest and important information

[0869] 5. (Server) The server analyzes the collected employee history data and uses a generative AI model to learn each employee's areas of interest and work-related information. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used.

[0870] 6. (Server) Based on the analysis results, the newly collected information from the internal bulletin board is scrutinized and information that is important to employees is selected.

[0871] Displaying notifications and recommendations

[0872] 7. (Server) When important information is picked up, employees are notified in real time using a notification system (e.g., Firebase Cloud Messaging, WebSocket).

[0873] 8. (Device) A notification will pop up on the employee's PC or smartphone, providing a link to view more information.

[0874] 9. (Server) Furthermore, information that is deemed to be of particular interest is displayed as recommended information on the dashboard that is displayed when employees log in.

[0875] Specific examples

[0876] Example 1: Information collection and analysis

[0877] (Server) Every morning at 9:00, the server accesses the company bulletin board and scrapes the latest posts, which include information such as "New project recruitment," "Weekend club activities," and "Drinking parties next week."

[0878] In the (server) analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "project recruitment" is classified as business information, while "club activities" and "drinking parties" are classified as social information.

[0879] Example 2: Learning and notifying employees about information of interest

[0880] (Terminal) Employee A logs in to the system and the history of clicking on the page recruiting for new projects is recorded.

[0881] (Server) Based on this historical data, the AI ​​model learns that Employee A has a high interest in "Project Recruitment."

[0882] (Server) The next day, when a new project recruitment post is posted on the company bulletin board, the server picks up the information and notifies employee A's device via Firebase Cloud Messaging.

[0883] (Device) Employee A's PC displays a notification saying "A new project recruitment post has been made," and he can click on the link to view more information.

[0884] Example 3: Displaying recommendations

[0885] (Server) By analyzing employee B's history data, it is discovered that employee B frequently views pages related to club activities.

[0886] (Server) Based on the results, the AI ​​model picks out information about club activities that it determines Employee B will be particularly interested in, and displays this information on the dashboard every Monday as "Recommended club activities for this weekend."

[0887] (Device) When employee B logs in on Monday, the dashboard will display "Recommended club activities for this weekend."

[0888] In this way, the system of the present invention is designed to enable employees to quickly and effectively obtain information necessary for their work and information of interest, thereby improving work efficiency and employee satisfaction.

[0889] The processing flow will be explained below.

[0890] Step 1:

[0891] (Server) Set up regular access to an internal bulletin board. For example, access the bulletin board every morning at 9:00 using Scrapy or Beautiful Soup to collect the latest posts.

[0892] Step 2:

[0893] (Server) Analyze the collected text data using natural language processing (NLP) techniques. Specifically, use spaCy or BERT to tokenize the text and identify entities and classify them by category.

[0894] Step 3:

[0895] (Server) Analyzed text data is stored in a database by category, e.g., business information, social information, etc.

[0896] Step 4:

[0897] (Device) When an employee logs into the company's internal system, their access history and click history are recorded in real time using JavaScript and web tracking technology.

[0898] Step 5:

[0899] (Terminal) The recorded history data is sent to the server in real time.

[0900] Step 6:

[0901] (Server) Stores the submitted historical data in a database and prepares it for analysis.

[0902] Step 7:

[0903] (Server) Use a generative AI model to analyze collected historical data and learn employee interests and work-related information. Example: Use Python machine learning libraries (TensorFlow, PyTorch).

[0904] Step 8:

[0905] (Server) Based on the analysis results, the newly collected information from the internal bulletin board is scrutinized and important information for employees is selected, with priority given to information that is of high interest to employees.

[0906] Step 9:

[0907] (Server) The collected information is notified to employees in real time using Firebase Cloud Messaging or WebSocket.

[0908] Step 10:

[0909] (Device) A notification pops up on the employee's PC or smartphone. For example, a message saying "A new project recruitment post has been made" appears.

[0910] Step 11:

[0911] (Device) When an employee clicks on the pop-up notification, they are taken to a screen that displays more information.

[0912] Step 12:

[0913] (Server) In addition, prepare to display information that is thought to be of particular interest as recommended information on employees' dashboards.

[0914] Step 13:

[0915] (Device) When an employee logs in, recommended information is displayed on the dashboard. For example, information is presented as "recommended club activities for this weekend."

[0916] In this way, the system quickly and efficiently provides employees with the information they need and are most interested in, thereby improving work efficiency and employee satisfaction.

[0917] Example 1

[0918] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0919] Conventional internal information sharing systems have made it difficult to efficiently collect and provide information that employees need or are interested in. As a result, employees may miss important information or waste time on information that is not relevant to their work. This can lead to problems such as reduced work efficiency and lower employee satisfaction.

[0920] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0921] In this invention, the server includes means for acquiring information from an internal bulletin board, means for analyzing the acquired information and classifying it by category, means for collecting employees' browsing history and click history, means for analyzing the collected history information and learning employees' areas of interest, means for selecting and notifying employees of important information based on the analysis results, means for displaying information that is likely to be of particular interest to employees as recommended information on their dashboards, means for periodically collecting information using a scraping tool, means for analyzing text data using natural language processing technology and classifying it after tokenization and part-of-speech tagging, and means for machine learning employee areas of interest using a generative AI model. This makes it possible to quickly and effectively collect and provide information that employees need or are interested in.

[0922] An "internal bulletin board" is a digital bulletin board set up within a company or organization for the purpose of sharing information among employees.

[0923] "Means of obtaining information" refers to the mechanism by which the server automatically collects the content posted on the internal bulletin board.

[0924] "Means of analysis" refers to techniques that use natural language processing technology to analyze collected text data and understand its content.

[0925] "Means for categorizing by category" refers to a method for sorting analyzed text data into specific categories (e.g., business information, social information).

[0926] "Means for collecting employees' browsing and click histories" refers to technology for recording the operations performed by employees within the company's internal systems.

[0927] "Means for learning areas of interest" refers to a machine learning model that analyzes and learns from employee interests and concerns from collected historical data.

[0928] "Means of notification" refers to a system for informing employees of important information in real time based on the analysis results.

[0929] "Means for displaying recommended information on the dashboard" refers to an interface that displays information related to employees' areas of interest when they log in.

[0930] A "scraping tool" is software or a library used to automatically extract data from websites.

[0931] "Natural language processing technology" is a technology that enables computers to understand and process human language.

[0932] "Tokenization" refers to the process of dividing text data into smaller units such as words or sentences.

[0933] "Part-of-speech tagging" refers to labeling tokenized words with their parts of speech (e.g., nouns, verbs, adjectives).

[0934] A "generative AI model" is an artificial intelligence model that uses machine learning to learn patterns and rules from data and make predictions and classifications.

[0935] The present invention is a system that acquires information from an in-house bulletin board, extracts information that is relevant to employees' fields of interest or their work, and notifies them of the information. This system is configured as follows.

[0936] System configuration and operation procedures

[0937] This system is mainly composed of servers, terminals, and users, each of which has a specific role and functions in cooperation with each other.

[0938] Data collection and analysis methods

[0939] The server periodically scrapes information from the company's internal bulletin board. The scraping tools used are Beautiful Soup and Scrapy. The collected data is then analyzed using natural language processing (NLP) techniques. Specifically, the NLP libraries spaCy and BERT are used to tokenize the text, tag parts of speech, and analyze dependencies. Finally, the text data is classified by category (e.g., business information, social information).

[0940] How to collect employee behavioral data

[0941] (Device) records access history and click history every time an employee logs in to the company's internal system or browses information. This is done using JavaScript and web tracking technology. The collected history data is sent to (server) in real time and stored in a database.

[0942] Areas of interest and how to learn important information

[0943] The server analyzes the collected employee behavioral data and uses a generative AI model to learn each employee's areas of interest and work-related information. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used. The AI ​​model learns patterns and areas of interest based on this data.

[0944] How notifications and recommendations are displayed

[0945] (Server) scrutinizes newly collected information from internal bulletin boards based on the analysis results and selects information that is important to employees. (Server) notifies employees in real time using a notification system (e.g., Firebase Cloud Messaging, WebSocket). (Terminal) displays a notification as a pop-up on employees' PCs or smartphones and provides a link to view more information. (Server) also displays information of particular interest as recommended information on the dashboard that is displayed when employees log in.

[0946] Specific examples

[0947] Example 1: Information collection and analysis

[0948] (Server) accesses the company's internal bulletin board at 9:00 every morning and scrapes the latest posts. For example, this includes information such as "new project recruitment," "weekend club activities," and "drinking party next week." In the analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "project recruitment" is classified as business information, while "club activities" and "drinking party" are classified as social information.

[0949] Example 2: Learning and notifying employees about information of interest

[0950] On (terminal), employee A logs in to the system and records the history of clicking on the new project recruitment page. Based on this history data, (server) uses its AI model to learn that employee A has a high interest in "project recruitment." The next day, when a new project recruitment post is posted on the company bulletin board, (server) picks up the information and notifies employee A's device via Firebase Cloud Messaging. (terminal) displays a notification on employee A's PC stating "A new project recruitment post has been made," and employee A can click on the link to view more information.

[0951] Example 3: Displaying recommendations

[0952] When (Server) analyzes employee B's historical data, it finds that employee B frequently views pages related to club activities. Based on the results, (Server) selects club activity information that the AI ​​model determines to be of particular interest to employee B, and displays this information on the dashboard every Monday as "Recommended club activities for this weekend." When employee B logs in on Monday, (Device) displays "Recommended club activities for this weekend" on the dashboard.

[0953] This system will enable employees to obtain the information they need quickly and efficiently, which is expected to improve work efficiency and employee satisfaction.

[0954] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0955] Step 1:

[0956] (Server) accesses the company bulletin board every morning at 9:00 and scrapes new posts. The URL of the bulletin board is given as input, and the HTML structure is analyzed using a scraping tool (Beautiful Soup or Scrapy). The output is the extracted text data. Specifically, it performs the following operations:< / url:> , , ) to extract the text.

[0957] Step 2:

[0958] The text data acquired by the server is preprocessed using natural language processing (NLP) techniques. The input is the raw data acquired in step 1, and the output is the preprocessed text data. Specifically, NLP libraries (spaCy and BERT) are used to tokenize the data and remove unnecessary tags and spaces.

[0959] Step 3:

[0960] The server tokenizes the preprocessed text data and performs part-of-speech tagging and dependency analysis. The input is the data preprocessed in step 2, and the output is tokenized and tagged text. Specifically, it uses spaCy to perform structural analysis of the sentence and tag each word with a part-of-speech tag.

[0961] Step 4:

[0962] The server classifies the analyzed text data into categories (business information, social information). The input is the text tagged with part-of-speech tags in step 3, and the output is the classified data. Specifically, it uses a machine learning algorithm to classify each text into a predefined category.

[0963] Step 5:

[0964] (Terminal) records logins to the company's internal system, access history, and click history. The input is the user's operation log, and the output is the recorded access data. Specifically, JavaScript code captures the user's operations and sends them to the server.

[0965] Step 6:

[0966] The historical data collected by (the terminal) is sent to the server in real time and stored in the database. The input is the access data recorded in step 5, and the output is the historical data stored on the server. Specifically, the data is sent to the server via an HTTP request and stored in the database.

[0967] Step 7:

[0968] The server analyzes the collected employee behavioral data and uses a generative AI model to learn each employee's areas of interest and work-related information. The input is the historical data saved in step 6, and the output is the learned model and interest information. Specifically, the AI ​​model is trained using TensorFlow or PyTorch to learn patterns from the data.

[0969] Step 8:

[0970] (Server) scrutinizes the newly collected information from the internal bulletin board and picks out information that is important to employees. The input is the data from Step 1 and Step 7, and the output is the picked-up important information. Specifically, it filters the information based on the prediction results of the AI ​​model.

[0971] Step 9:

[0972] The server notifies employees in real time using a notification system (Firebase Cloud Messaging or WebSocket). The input is the information picked up in step 8, and the output is a notification message. Specifically, the server sends a message in real time through the notification system.

[0973] Step 10:

[0974] The device displays a pop-up notification and provides a link to view more information. The input is the notification message sent in step 9, and the output is the user's browsing behavior. The specific behavior is to display a pop-up window in a browser or application.

[0975] Step 11:

[0976] Information that the server determines to be of particular interest is displayed as recommended information on the dashboard when employees log in. The input is the data from steps 7 and 8, and the output is the recommended information displayed on the dashboard. Specifically, the settings are configured to list and display specific information on the dashboard.

[0977] Through the above processing steps, a system is provided that allows employees to quickly and efficiently obtain the information they need or are interested in.

[0978] (Application example 1)

[0979] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0980] Conventional systems that collect information from internal bulletin boards and notify employees have difficulty in properly identifying employees' areas of interest and necessary information. Furthermore, at production sites, proper information collection and notification is not performed, making it difficult to provide efficient work support or timely maintenance. For this reason, there is a need for a system that efficiently collects production schedules and maintenance information at factories and notifies workers appropriately.

[0981] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0982] In this invention, the server includes means for acquiring information from an in-house bulletin board, means for analyzing the acquired information and classifying it by category, means for collecting employees' browsing history and click history, means for analyzing the collected history information and learning employees' areas of interest, means for selecting and notifying employees of important information based on the analysis results, means for displaying information that is likely to be of particular interest as recommended information on the employee's dashboard, means for scraping production plans and function check information from an in-house management system, and means for analyzing the collected information and notifying it in conjunction with sensor data and work logs acquired during work. This allows employees and workers to quickly and efficiently receive the information they need, improving work efficiency and enabling appropriate maintenance.

[0983] An "internal bulletin board" is an electronic bulletin board system set up within a company for the purpose of sharing information and communicating.

[0984] "Means of obtaining information" refers to the methods and technologies used to collect the necessary information from various data sources.

[0985] "Means for analyzing information and classifying it into categories" refers to technology that analyzes acquired data and classifies it into specific categories based on its content.

[0986] "Means of collecting employees' browsing and click histories" refers to technology that collects the behavioral history of employees when they use web systems.

[0987] "Means of analyzing collected historical information to learn employees' areas of interest" refers to technology that analyzes collected behavioral data to learn what information a particular employee is interested in.

[0988] "Means of selecting and notifying employees of important information based on the analysis results" refers to methods and technologies for selecting important information from the analysis results and notifying employees of it.

[0989] "Means for displaying recommended information on employee dashboards" refers to a method for displaying information that is determined to be of particular interest to employees on their interface.

[0990] "Factory management system" refers to a comprehensive system for managing factory production and equipment.

[0991] "Means for scraping production plan and function check information" refers to technology that automatically extracts necessary information from factory management systems.

[0992] "Sensor data and work logs obtained during work" refers to the various data generated by robots and workers while they are working, as well as the recorded work history.

[0993] "Generative AI models" refer to algorithms and technologies that use machine learning techniques to generate knowledge and predictive models from data.

[0994] A "prompt" is an instruction given to a generative AI model to make it perform a specific task.

[0995] A "notification system" refers to technology that notifies individual devices of important information in real time.

[0996] The following describes an embodiment of the present invention. The present invention is a system that acquires information from a factory management system, notifies robots and workers of appropriate information, and improves work efficiency. This system functions as follows.

[0997] System configuration and operation

[0998] Data collection and analysis

[0999] 1. Server: The server periodically scrapes production plans and functional check information from the factory management system using web scraping tools such as Beautiful Soup and Scrapy.

[1000] 2. Server: The collected text data is analyzed using natural language processing (NLP) techniques. Specifically, NLP libraries such as spaCy and BERT are used to classify the text into categories (e.g., production information, maintenance information).

[1001] User behavior data collection

[1002] 3. Robot: The sensor data and work logs acquired by the robot while working are recorded and sent to the server in real time. The data is collected using JavaScript and web tracking technology.

[1003] 4. Server: The collected historical data is stored on the server.

[1004] Learn about areas of interest and important information

[1005] 5. Server: The server analyzes the collected robot history data and uses generative AI models to learn important information for each robot and worker. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used.

[1006] 6. Server: Based on the analysis results, the newly collected information from the factory management system is scrutinized and important information is picked up.

[1007] Displaying notifications and recommendations

[1008] 7. Server: When important information is picked up, it is notified to the robot or worker in real time using a notification system (e.g. MQTT, WebSocket).

[1009] 8. Robot: Notifications are sent via the robot's display and voice output, which can be checked by the worker. Work instructions and changes are also displayed here.

[1010] Specific examples

[1011] Example 1: Information collection and analysis

[1012] Server: Every morning at 9:00, the server accesses the factory management system and scrapes the latest production schedule and maintenance schedule, including information such as "new production line setup," "machine maintenance next month," and "equipment inspection this weekend."

[1013] Server: In the analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "production line settings" is classified as production information, while "machine maintenance" and "equipment inspection" are classified as maintenance information.

[1014] Example 2: Learning and notifying robot behavior information

[1015] Robot: Sensor data and work logs acquired by specific robots while performing machine maintenance are recorded.

[1016] Server: Based on this historical data, the AI ​​model learns the information necessary for the robot.

[1017] Server: When new machine maintenance information is posted on the factory management system, the server picks up the information and notifies the robot via MQTT.

[1018] Robot: A notification saying "New machine maintenance information has been added" appears on the robot's display and can be confirmed by the worker.

[1019] Example 3: Displaying recommendations

[1020] Server: By analyzing historical data for a particular robot, you may discover that the robot frequently performs tasks on a particular production line setup.

[1021] Server: Based on the results, the AI ​​model picks out information about the production line settings that it deems particularly important for that robot.

[1022] Server: Every morning, the server notifies the robot of this information and displays it on the robot's display as "Today's recommended production line settings."

[1023] In this way, the system of the present invention is designed to enable robots and workers to quickly and effectively obtain the information they need for their work, thereby improving work efficiency and ensuring appropriate maintenance.

[1024] Prompt Sentence Examples

[1025] "Next week's machine maintenance schedule"

[1026] "New Production Plan for Project X"

[1027] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1028] Step 1:

[1029] The server acquires production plan information and maintenance schedule information from the factory management system.

[1030] Input: Production plan information and maintenance schedule information from the factory management system.

[1031] Data processing: Use web scraping tools such as Beautiful Soup or Scrapy to extract the necessary information and save it as text data.

[1032] Output: Text data of extracted production plan information and maintenance schedule information.

[1033] Step 2:

[1034] The server analyzes the acquired text data and classifies it into categories.

[1035] Input: The text data extracted in step 1.

[1036] Data processing: Using natural language processing (NLP) techniques, we tokenize the text and classify it into categories (e.g., production information, maintenance information). Specifically, we use NLP libraries such as spaCy and BERT.

[1037] Output: Text data classified by category.

[1038] Step 3:

[1039] The robot records sensor data and work logs acquired during work and sends them to a server.

[1040] Input: Sensor data and work logs acquired while the robot is working.

[1041] Data processing: Collect data in real time and send it to the server. Record data using JavaScript and web tracking technology.

[1042] Output: Sensor data and work logs sent to the server.

[1043] Step 4:

[1044] The server analyzes the collected robot history data and uses a generative AI model to learn important information.

[1045] Input: Sensor data and work logs sent in step 3.

[1046] Data processing: Using Python machine learning libraries (e.g., TensorFlow, PyTorch), the data is analyzed to learn information that is important to each robot.

[1047] Output: Key information identified by the generative AI model.

[1048] Step 5:

[1049] The server picks out important information based on the analysis results and notifies you.

[1050] Input: Key information learned in step 4.

[1051] Data processing: Newly collected production plans and maintenance information are compared with the analysis results to select important information.

[1052] Output: Important information picked up.

[1053] Step 6:

[1054] The server uses a notification system to notify robots and workers of important information in real time.

[1055] Input: Important information picked up in step 5.

[1056] Data processing: Use MQTT or WebSocket to send notifications in real time.

[1057] Output: Notification message that will be displayed to the robot and worker.

[1058] Step 7:

[1059] The robot will display the notified information on a screen or output it as audio so that the worker can check it.

[1060] Input: Information provided in step 6.

[1061] Data processing: Display and audio output are performed so that workers can check the information.

[1062] Output: Notifications are displayed on the robot's display and audio output.

[1063] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1064] The following describes an embodiment of the present invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the areas of interest and emotional state of employees. The system is operated by combining a generative AI and an emotion engine.

[1065] System configuration and operation

[1066] Data collection and analysis

[1067] 1. (Server) Regularly scrape information from internal bulletin boards. For example, use tools like Scrapy or Beautiful Soup to collect the latest posts every morning at 9am.

[1068] 2. (Server) Analyze the collected text data using natural language processing (NLP) techniques, such as spaCy or BERT, to tokenize the text and classify it by category (e.g., business information, social information).

[1069] Collecting employee behavioral and emotional data

[1070] 3. (Device) Every time an employee logs in to an internal system and views information, their access history and click history are recorded. This is done using JavaScript and web tracking technology.

[1071] 4. (Device) Obtain recorded history data, as well as user-entered text and interaction data.

[1072] 5. (Terminal) Send the collected data to the server in real time.

[1073] Learning areas of interest and emotional states

[1074] 6. (Server) Stores the submitted historical and interaction data in a database and prepares it for analysis.

[1075] 7. (Server) Use a generative AI model to analyze historical data to learn employee interests and job-related information. Use Python machine learning libraries (e.g., TensorFlow, PyTorch).

[1076] 8. (Server) An emotion engine is used to analyze the user's text input and click behavior to identify their emotional state. Emotion recognition technology combines NLP technology with machine learning models (e.g., BERT, Keras).

[1077] Providing notifications and recommendations

[1078] 9. (Server) Based on the learning results, the server scrutinizes the newly collected information from the internal bulletin board and selects information that is important to employees. The server also adjusts the timing and content of notifications based on the results of the emotion engine.

[1079] 10. (Server) Use Firebase Cloud Messaging or WebSocket to send notifications to employees in real time.

[1080] 11. (Device) Notifications are displayed as pop-ups on employees' PCs or smartphones. The pop-ups display specific messages such as "A new project recruitment post has been made," and employees can click on a link to access more information.

[1081] 12. (Server) Information that is thought to be of particular interest is displayed as recommended information on the dashboard that employees log in to.

[1082] Specific examples

[1083] Example 1: Information collection and analysis

[1084] (Server) Every morning at 9:00, information is scraped from the company bulletin board to obtain information such as "New project recruitment," "Weekend club activities," and "Drinking parties next week."

[1085] (Server) Analyze the acquired text using spaCy or BERT and classify it into categories. For example, "Project recruitment" is classified as business information, while "Club activities" and "Drinking parties" are classified as social information.

[1086] Example 2: Learning and notifying employee interests and emotional states

[1087] (Device) Employee A logs in and clicks on the new project recruitment page. The text input and click history are recorded.

[1088] (Server) The generative AI model analyzes the historical data and learns that Employee A has a high interest in the "Project Recruitment" campaign. The emotion engine also detects "excitement" and "interest" from Employee A's input text.

[1089] (Server) The next day, if a new project recruitment post is made, a notification is sent to Employee A's device using Firebase Cloud Messaging. Based on the results of the emotion engine, the notification message is adjusted to read, "This project seems interesting to you."

[1090] (Device) A pop-up message appears on Employee A's PC saying "A new project recruitment post has been made," and clicking the link displays more information.

[1091] Example 3: Displaying recommendations

[1092] (Server) Analyzes employee B's historical data and emotional state and finds that he frequently views pages related to club activities and feels "excited" and "expected."

[1093] (Server) Based on this, information on club activities that are of particular interest to employee B is selected and displayed on the dashboard every Monday as "Recommended club activities for this weekend."

[1094] (Device) When employee B logs in on Monday, the dashboard will display "Recommended club activities for this weekend."

[1095] In this way, the system of the present invention provides employees with information necessary for their work and information of interest quickly and efficiently, and also provides appropriate information according to the user's emotional state, thereby improving work efficiency and employee satisfaction.

[1096] The processing flow will be explained below.

[1097] Step 1:

[1098] (Server) Set up periodic access to the internal bulletin board. For example, use Scrapy or Beautiful Soup to collect the latest postings every morning at 9am.

[1099] Step 2:

[1100] (Server) Analyze the collected text data using natural language processing (NLP) technology. Specifically, use spaCy or BERT to tokenize the text and classify it by category (e.g., business information, social information).

[1101] Step 3:

[1102] (Server) The analyzed text data is stored in a database by category. The stored data is used for later information review.

[1103] Step 4:

[1104] (Device) When an employee logs in to the company's internal system, JavaScript and web tracking technology are used to record access history and click history.

[1105] Step 5:

[1106] (Device) Recorded historical data is sent to the server in real time, making it immediately available for analysis.

[1107] Step 6:

[1108] (Server) Stores the submitted historical data in a database so that it can be used by the generative AI model and emotion engine for analysis.

[1109] Step 7:

[1110] (Server) Analyze historical data using a generative AI model to learn employee interests and work-related information. Build the learning model using Python machine learning libraries (e.g., TensorFlow, PyTorch).

[1111] Step 8:

[1112] (Server) An emotion engine is used to analyze user input data and click history to identify the user's emotional state. For example, NLP techniques and machine learning models are combined to identify emotions such as "excitement," "interest," and "boredom" from user text input.

[1113] Step 9:

[1114] (Server) Based on the analysis results, the server scrutinizes newly collected information from internal bulletin boards and selects information that is important to employees. At this time, the server adjusts the timing and content of notifications according to the user's emotional state as determined by the emotion engine.

[1115] Step 10:

[1116] (Server) Notifications are sent to employees in real time using Firebase Cloud Messaging and WebSockets. The content and format of the notifications are customized taking into account the results of the emotion engine.

[1117] Step 11:

[1118] (Device) A notification pops up on the employee's PC or smartphone. For example, a message saying "A new project recruitment post has been made" appears. When the user clicks on the link, more information is displayed.

[1119] Step 12:

[1120] (Server) In addition, prepare to display information that is thought to be of particular interest as "recommended information" on employees' dashboards.

[1121] Step 13:

[1122] (Device) When an employee logs in, "recommended information" is displayed on the dashboard. For example, information such as "recommended club activities for this weekend" is displayed.

[1123] This system allows employees to quickly and efficiently obtain information necessary for their work and information of interest. In addition, because it provides appropriate information based on the user's emotional state, it is expected to improve work efficiency and employee satisfaction.

[1124] Example 2

[1125] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1126] Conventional internal information sharing systems make it difficult for employees to efficiently obtain the information they need, resulting in problems such as missing information or failing to grasp important information. Furthermore, to improve the quality of communication within a company and increase the work efficiency of each employee, it is necessary to provide information that takes into account the interests and emotional state of employees, but there were few systems that could achieve this. Furthermore, the lack of information provided that was tailored to the interests of each employee sometimes resulted in information overload, which in turn reduced productivity.

[1127] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1128] In this invention, the server includes a means for acquiring information from the in-house bulletin board, a means for analyzing the acquired information using natural language processing technology and classifying it into categories, and a means for collecting collected history information and user input text in real time. This makes it possible to efficiently collect information needed by employees and provide important information at an appropriate time, taking into account the employees' areas of interest and emotional state.

[1129] 1. An "internal bulletin board" is an online platform for sharing information within a company or organization.

[1130] 2. "Means of obtaining information" refers to tools and programs for automatically collecting necessary information from internal bulletin boards.

[1131] 3. "Natural language processing technology" is an artificial intelligence technology for analyzing and understanding human language.

[1132] 4. "Categorization methods" refers to algorithms or programs that group collected information based on specific criteria.

[1133] 5. "Browsing history" refers to a record of web pages and documents that a user has previously viewed.

[1134] 6. "Click History" refers to a record of links and buttons that a user has previously clicked.

[1135] 7. "History information" refers to all data relating to a user's behavioral history.

[1136] 8. "User Input Text" means any characters or sentences entered by a User into the System.

[1137] 9. "Real-time collection means" refers to technologies or methods that capture data as it is transmitted.

[1138] 10. "Database" refers to a system for efficiently storing, managing, and retrieving large amounts of data.

[1139] 11. "Generative AI model" refers to an artificial intelligence model that learns from generated data and performs tasks such as prediction and classification.

[1140] 12. "Emotion Engine" refers to technology or programs that identify a user's emotional state from their text input or behavior.

[1141] 13. "Interest Areas" refers to the themes or topics that interest a particular user.

[1142] 14. "Business-Related Information" means information related to a User's job function.

[1143] 15. "Means of notification" refers to tools and technologies used to communicate information to users.

[1144] 16. "Recommended Information" refers to the provision of specific information selected based on a user's interests and behavior.

[1145] The following describes an embodiment of the present invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the areas of interest and emotional state of employees. The system is operated by combining a generative artificial intelligence model and an emotion analysis engine.

[1146] System configuration and operation

[1147] Data collection and analysis

[1148] 1. The server periodically retrieves information from the company bulletin board. To collect the information, use a web scraping tool such as Scrapy or Beautiful Soup. For example, set it to collect the latest posts every morning at 9:00.

[1149] 2. The server analyzes the collected text data using natural language processing (NLP) techniques, such as spaCy and BERT, to tokenize the text and classify it into categories (e.g., business information, social information).

[1150] Collecting employee behavioral and emotional data

[1151] 3. The device records access and click histories every time an employee logs into the company's internal systems and views information. This recording is done using JavaScript and web tracking technology.

[1152] 4. The device acquires the recorded history data, text entered by the user, and interaction data, and transmits them to the server in real time.

[1153] Saving data and preparing for analysis

[1154] 5. The server stores the transmitted history and interaction data in a database and prepares it for analysis, using a database system such as PostgreSQL or MongoDB.

[1155] Learning areas of interest and emotional states

[1156] 6. The server uses generative artificial intelligence models (e.g., TensorFlow, PyTorch) to analyze historical data and learn employee interests and work-related information.

[1157] 7. The server uses an emotion engine (e.g., BERT, Keras) to identify the user's emotional state from their text input and click behavior.

[1158] Providing notifications and recommendations

[1159] 8. Based on the learning results and sentiment analysis results, the server examines the newly collected information from the internal bulletin board and notifies employees of important information. For example, it prepares a tailored message such as, "This project seems interesting to you."

[1160] 9. The server uses Firebase Cloud Messaging or WebSocket to send notifications in real time to employee devices, which are displayed as pop-ups on PCs and smartphones.

[1161] 10. The server displays information of particular interest as recommended information on the dashboard screen when employees log in. For example, information on club activities of high interest is displayed on the dashboard every Monday as "Recommended activities for this weekend."

[1162] Specific examples

[1163] Example 1: Information collection and analysis

[1164] The server scrapes information from the company's internal bulletin boards every morning at 9 a.m., obtaining information such as "new project recruitment," "weekend social events," and "next week's meetings."

[1165] The server analyzes the acquired text using natural language processing technology (spaCy or BERT) and classifies it into categories. For example, "project recruitment" is classified as business information, while "social events" and "meetings" are classified as social information.

[1166] Example 2: Learning and notifying employee interests and emotional states

[1167] The device records access history and click history when employee A logs in and clicks on the page recruiting for new projects.

[1168] The server analyzes the historical data based on a generative artificial intelligence model and learns that Employee A has a high interest in the "Project Recruitment." The emotion engine also detects "excitement" and "interest" from Employee A's input text.

[1169] The next day, if there is a new project opening, the server will use Firebase Cloud Messaging to notify employee A's device, saying, "We think this project might be interesting to you."

[1170] The device displays a pop-up on Employee A's PC saying, "A new project recruitment post has been made," and when Employee A clicks on the link, more information will be displayed.

[1171] Example 3: Displaying recommendations

[1172] The server analyzes employee B's historical data and emotional state, and determines that he frequently views pages related to club activities and feels emotions of "excitement" and "expectation."

[1173] Based on this, the server selects information about club activities that employee B is particularly interested in and displays it on the dashboard every Monday as "Recommended club activities for this weekend."

[1174] When employee B logs in to the device on Monday, the dashboard will display "Recommended club activities for this weekend."

[1175] Example prompt sentence:

[1176] "Collect information on new project openings and notify employees."

[1177] "Identify areas of interest from employees' click history"

[1178] "Analyze the user's emotional state using an emotion engine and provide appropriate notifications."

[1179] This system allows employees to quickly and efficiently receive not only the information they need for their work, but also appropriate information tailored to their individual interests and emotional state, thereby improving work efficiency and employee satisfaction.

[1180] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1181] Step 1:

[1182] The server scrapes information from the company bulletin board every morning at 9:00. It uses the bulletin board URL as input and generates data including the title, body, and posting date and time of each post as output. Specifically, it uses Scrapy and Beautiful Soup to parse the HTML and extract the required text information.

[1183] Step 2:

[1184] The server analyzes the collected text data using natural language processing technology and classifies it into categories. It uses the acquired text data as input and obtains data classified into categories (e.g., business information, social information) as output. Specifically, it uses spaCy and BERT to tokenize the text and perform contextual analysis.

[1185] Step 3:

[1186] The device records access and click histories every time an employee logs into the company's internal systems and browses information. It uses the employee's browsing and clicking behavior as input and generates the recorded history data as output. Specifically, it captures user actions using JavaScript and web tracking technology.

[1187] Step 4:

[1188] The device sends the recorded history data and the user's input text and interaction data to the server in real time. The device uses the recorded history data and the user's text data as input and generates the data sent to the server as output. Specifically, the device converts the data format to JSON or similar and sends it to the server using the HTTPS protocol.

[1189] Step 5:

[1190] The server stores the submitted historical and interaction data in a database and prepares it for analysis. It uses the received historical and interaction data as input and generates the data stored in the database as output, specifically by inserting data using database queries.

[1191] Step 6:

[1192] The server uses a generative AI model to analyze historical data and learn employee interests and work-related information. It uses the stored historical data as input and obtains data identifying employee interests as output. Specifically, it trains the model using TensorFlow or PyTorch and makes predictions based on the historical data.

[1193] Step 7:

[1194] The server uses an emotion engine to identify the user's emotional state from their text input and click behavior. It uses the user's text data and click data as input and obtains the identified emotional state as output. Specifically, it uses BERT and Keras to perform emotion recognition.

[1195] Step 8:

[1196] Based on the learning results and sentiment analysis results, the server scrutinizes the newly collected information from the internal bulletin board, picks out information that is important to employees, and prepares the notification content. It uses the data obtained from the internal bulletin board and the analysis results as input, and generates a notification message as output. Specifically, it selects an appropriate template and creates the message.

[1197] Step 9:

[1198] The server sends notifications to employee devices in real time using Firebase Cloud Messaging or WebSocket. It uses the prepared notification message as input and gets the sent notification as output. Specifically, it sends messages using the Firebase API or WebSocket protocol.

[1199] Step 10:

[1200] The device displays the received notification as a pop-up on the employee's PC or smartphone. It uses the received notification message as input and obtains the displayed pop-up notification as output. Specifically, it uses the browser's Notification API or the OS's notification function to display the message.

[1201] Step 11:

[1202] The server then displays the most interesting information as recommendations on the employee's dashboard. It uses information based on the areas of interest as input and generates recommendations that are displayed on the dashboard as output. Specifically, it dynamically updates the content of the web page and applies templates to display the information of interest.

[1203] (Application example 2)

[1204] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1205] Conventional internal bulletin board systems and notification systems do not provide information that takes into account the interests and emotional state of individual employees, limiting their ability to improve work efficiency and employee satisfaction. Furthermore, it is difficult to provide employees with the important information they need at the right time, leading to problems with missed information and excessive notifications. There is a need for a system that can solve these issues and provide optimal information to employees while improving work efficiency.

[1206] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring information from the in-house bulletin board, means for analyzing the acquired information and classifying it by category, and means for analyzing the collected history data and interaction data to identify the user's emotional state. This makes it possible to pick out important information based on the employee's areas of interest and emotional state, appropriately adjust the content of notification messages, and provide them to employees in real time.

[1207] An "internal bulletin board" is an online bulletin board for the purpose of sharing information and communicating among employees.

[1208] "Means of obtaining information" refers to the techniques and methods used to collect the necessary information from the internal bulletin board.

[1209] "Means of analyzing information" refers to the techniques and methods used to structure collected information and convert it into an easily understandable format.

[1210] "Categorization means" refers to techniques or methods for classifying analyzed information into different categories.

[1211] "Means of collecting employee browsing and click history" refers to technologies and methods that record what information employees view and what links they click.

[1212] "Means for analyzing historical information and learning about employees' areas of interest" refers to techniques and methods for identifying employees' areas of interest and concern from collected historical information.

[1213] "Means for analyzing interaction data and identifying the user's emotional state" refers to technologies and methods for analyzing the emotions of employees at that time based on their interaction data.

[1214] "Means of notification" refers to the technology and methods used to inform employees of necessary information based on the analysis results.

[1215] "Means for adjusting the content of notification messages" refers to techniques and methods for changing notification messages to optimal content depending on the emotional state of employees.

[1216] "Means for displaying recommended information on the dashboard" refers to the technology and methods for displaying information of particular interest on employees' operation screens.

[1217] "Means of real-time notification" refers to technologies and methods that instantly inform employees of the information they need at that moment.

[1218] The following describes an embodiment of this invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the employee's areas of interest and emotional state. The system is operated by combining a generative AI and an emotion engine.

[1219] System configuration and operation

[1220] Data collection and analysis

[1221] The server periodically collects information from internal bulletin boards using scraping technologies such as Scrapy and Beautiful Soup. The collected information is analyzed and categorized using natural language processing (NLP) techniques. Specifically, text is tokenized using spaCy and BERT, and then classified into categories such as business information and social information.

[1222] Collecting employee behavioral and emotional data

[1223] The device uses JavaScript and web tracking technology to record historical data and click history when employees access internal systems. It also collects text entered by users and interaction data, which it then transmits in real time to a server where it is prepared for analysis.

[1224] Learning areas of interest and emotional states

[1225] The server uses a generative AI model built with TensorFlow and PyTorch to learn employee interests and work-related information from historical data. It also uses an emotion engine using BERT and Keras to analyze users' text input and click behavior to identify their emotional state. For example, if an employee expresses emotions such as "excitement" or "interest," that information is identified by the emotion recognition model.

[1226] Providing notifications and recommendations

[1227] Based on the analysis results, the server scrutinizes newly collected information from internal bulletin boards and selects information that is important to employees. The timing and content of notifications are adjusted taking into account the results of the emotion engine. Notifications are sent to employees in real time using Firebase Cloud Messaging and WebSocket. These notifications are displayed as pop-ups on employees' PCs or smartphones. Additionally, information that is deemed to be of particular interest is displayed as recommended information on the dashboard that employees log in to.

[1228] Specific examples

[1229] Information collection and analysis

[1230] The server scrapes information from the company's internal bulletin board at 9 a.m. every morning, obtaining information such as new project recruitment, weekend club activities, and next week's drinking party. The obtained text is analyzed using spaCy and BERT and classified by category. For example, "project recruitment" is classified as business information, and "club activities" is classified as social information.

[1231] Learn and notify employee interests and emotional states

[1232] The device records the text input and click history when Employee A logs in and clicks on the new project recruitment page. The server analyzes the historical data using a generative AI model and learns that Employee A has a high interest in the "project recruitment." The emotion engine also detects "excitement" and "interest" from Employee A's input text. The next day, when a new project recruitment post is made, a notification is sent to Employee A's device using Firebase Cloud Messaging. Based on the results of the emotion engine, the notification message is adjusted to read, "We think this project might be interesting to you." A pop-up appears on Employee A's PC saying, "A new project recruitment post has been made," and the user can click the link to access more information.

[1233] Displaying recommendations

[1234] The server analyzes employee B's historical data and emotional state, and determines that he frequently views pages related to club activities and that he feels "excitement" and "expectations." Based on this, information on club activities that are of particular interest to employee B is selected and displayed on the dashboard every Monday as "Recommended club activities for this weekend." When employee B logs in on Monday, "Recommended club activities for this weekend" is displayed on the dashboard.

[1235] Prompt Sentence Examples

[1236] "Have a generative AI model perform sentiment analysis on food delivery requests and send relevant notifications for appropriate delivery jobs if the sentiment score is high."

[1237] In this way, the system of the present invention can quickly and efficiently provide employees with information necessary for their work and information of interest, and can also provide appropriate information according to the user's emotional state, thereby improving work efficiency and employee satisfaction.

[1238] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1239] Step 1:

[1240] Collect information from an internal bulletin board. The server uses a scraping tool such as Scrapy or Beautiful Soup to retrieve the contents of the internal bulletin board at a set time (for example, 9:00 every morning). The input is a URL and necessary initial information, and the output is the latest posted information as text data.

[1241] Step 2:

[1242] The collected information is analyzed and classified by category. The server uses natural language processing (NLP) technology to analyze the acquired text data and perform tokenization and parsing. Specifically, it uses spaCy and BERT to classify it into categories such as business information and social information. The input is the collected text data, and the output is information classified by category.

[1243] Step 3:

[1244] This system collects employee behavioral and emotional data. The device uses JavaScript and web tracking technology to record the browsing and click history of employees when they access internal systems in real time. The input is the employee's interactions (clicks, page transitions, etc.), and the output is the recorded behavioral data.

[1245] Step 4:

[1246] The collected history data and interaction data are sent to the server. The terminal transfers and stores the behavioral data in real time to the server. The input is the behavioral data recorded in real time, and the output is the behavioral data stored on the server side.

[1247] Step 5:

[1248] Analyze areas of interest and emotional state. The server uses a generative AI model (using TensorFlow or PyTorch) to analyze historical data and identify employees' areas of interest. It then uses an emotion engine (using BERT or Keras) to analyze emotional states from interaction data. The input is historical data and interaction data, and the output is the identified areas of interest and emotional state.

[1249] Step 6:

[1250] Important information is extracted and notification messages are generated. Based on the analysis results, the server scrutinizes the newly collected information from the internal bulletin board and selects information that is important to employees. It also adjusts the content of the notification message taking into account the results of the emotion engine. The input is the analyzed data and new internal bulletin board information, and the output is an optimized notification message.

[1251] Step 7:

[1252] Send notifications in real time. The server uses Firebase Cloud Messaging or WebSocket to send optimized notification messages to employee devices in real time. The input is the generated notification message, and the output is the notification displayed on the employee's device.

[1253] Step 8:

[1254] Display as recommended information. The server configures the information that is deemed to be of particular interest to be displayed on the employee's dashboard. The input is the identified areas of interest and selected information, and the output is the recommended information displayed on the employee's dashboard.

[1255] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1256] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1257] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1258] [Fourth embodiment]

[1259] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1260] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1261] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1262] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1263] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1264] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1265] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1266] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1267] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1268] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1269] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1270] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1271] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1272] The present invention provides a system for acquiring information from an internal bulletin board, extracting information related to employees' areas of interest or necessary for their work, and notifying them of the information. This system functions as follows.

[1273] System configuration and operation

[1274] Data collection and analysis

[1275] 1. (Server) The server periodically scrapes information from the internal bulletin board using a web scraping tool such as Beautiful Soup or Scrapy.

[1276] 2. (Server) The collected text data is analyzed using natural language processing (NLP) techniques. Specifically, NLP libraries such as spaCy and BERT are used to classify the text into categories (e.g., business information, social information).

[1277] Collecting employee behavior data

[1278] 3. (Device) When an employee logs in to an internal system and views information, their access history and click history are recorded. This is done using JavaScript and web tracking technology.

[1279] 4. (Terminal) The collected historical data is sent to the server in real time and stored in a database.

[1280] Learn about areas of interest and important information

[1281] 5. (Server) The server analyzes the collected employee history data and uses a generative AI model to learn each employee's areas of interest and work-related information. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used.

[1282] 6. (Server) Based on the analysis results, the newly collected information from the internal bulletin board is scrutinized and information that is important to employees is selected.

[1283] Displaying notifications and recommendations

[1284] 7. (Server) When important information is picked up, employees are notified in real time using a notification system (e.g., Firebase Cloud Messaging, WebSocket).

[1285] 8. (Device) A notification will pop up on the employee's PC or smartphone, providing a link to view more information.

[1286] 9. (Server) Furthermore, information that is deemed to be of particular interest is displayed as recommended information on the dashboard that is displayed when employees log in.

[1287] Specific examples

[1288] Example 1: Information collection and analysis

[1289] (Server) Every morning at 9:00, the server accesses the company bulletin board and scrapes the latest posts, which include information such as "New project recruitment," "Weekend club activities," and "Drinking parties next week."

[1290] In the (server) analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "project recruitment" is classified as business information, while "club activities" and "drinking parties" are classified as social information.

[1291] Example 2: Learning and notifying employees about information of interest

[1292] (Terminal) Employee A logs in to the system and the history of clicking on the page recruiting for new projects is recorded.

[1293] (Server) Based on this historical data, the AI ​​model learns that Employee A has a high interest in "Project Recruitment."

[1294] (Server) The next day, when a new project recruitment post is posted on the company bulletin board, the server picks up the information and notifies employee A's device via Firebase Cloud Messaging.

[1295] (Device) Employee A's PC displays a notification saying "A new project recruitment post has been made," and he can click on the link to view more information.

[1296] Example 3: Displaying recommendations

[1297] (Server) By analyzing employee B's history data, it is discovered that employee B frequently views pages related to club activities.

[1298] (Server) Based on the results, the AI ​​model picks out information about club activities that it determines Employee B will be particularly interested in, and displays this information on the dashboard every Monday as "Recommended club activities for this weekend."

[1299] (Device) When employee B logs in on Monday, the dashboard will display "Recommended club activities for this weekend."

[1300] In this way, the system of the present invention is designed to enable employees to quickly and effectively obtain information necessary for their work and information of interest, thereby improving work efficiency and employee satisfaction.

[1301] The processing flow will be explained below.

[1302] Step 1:

[1303] (Server) Set up regular access to an internal bulletin board. For example, access the bulletin board every morning at 9:00 using Scrapy or Beautiful Soup to collect the latest posts.

[1304] Step 2:

[1305] (Server) Analyze the collected text data using natural language processing (NLP) techniques. Specifically, use spaCy or BERT to tokenize the text and identify entities and classify them by category.

[1306] Step 3:

[1307] (Server) Analyzed text data is stored in a database by category, e.g., business information, social information, etc.

[1308] Step 4:

[1309] (Device) When an employee logs into the company's internal system, their access history and click history are recorded in real time using JavaScript and web tracking technology.

[1310] Step 5:

[1311] (Terminal) The recorded history data is sent to the server in real time.

[1312] Step 6:

[1313] (Server) Stores the submitted historical data in a database and prepares it for analysis.

[1314] Step 7:

[1315] (Server) Use a generative AI model to analyze collected historical data and learn employee interests and work-related information. Example: Use Python machine learning libraries (TensorFlow, PyTorch).

[1316] Step 8:

[1317] (Server) Based on the analysis results, the newly collected information from the internal bulletin board is scrutinized and important information for employees is selected, with priority given to information that is of high interest to employees.

[1318] Step 9:

[1319] (Server) The collected information is notified to employees in real time using Firebase Cloud Messaging or WebSocket.

[1320] Step 10:

[1321] (Device) A notification pops up on the employee's PC or smartphone. For example, a message saying "A new project recruitment post has been made" appears.

[1322] Step 11:

[1323] (Device) When an employee clicks on the pop-up notification, they are taken to a screen that displays more information.

[1324] Step 12:

[1325] (Server) In addition, prepare to display information that is thought to be of particular interest as recommended information on employees' dashboards.

[1326] Step 13:

[1327] (Device) When an employee logs in, recommended information is displayed on the dashboard. For example, information is presented as "recommended club activities for this weekend."

[1328] In this way, the system quickly and efficiently provides employees with the information they need and are most interested in, thereby improving work efficiency and employee satisfaction.

[1329] Example 1

[1330] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1331] Conventional internal information sharing systems have made it difficult to efficiently collect and provide information that employees need or are interested in. As a result, employees may miss important information or waste time on information that is not relevant to their work. This can lead to problems such as reduced work efficiency and lower employee satisfaction.

[1332] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1333] In this invention, the server includes means for acquiring information from an internal bulletin board, means for analyzing the acquired information and classifying it by category, means for collecting employees' browsing history and click history, means for analyzing the collected history information and learning employees' areas of interest, means for selecting and notifying employees of important information based on the analysis results, means for displaying information that is likely to be of particular interest to employees as recommended information on their dashboards, means for periodically collecting information using a scraping tool, means for analyzing text data using natural language processing technology and classifying it after tokenization and part-of-speech tagging, and means for machine learning employee areas of interest using a generative AI model. This makes it possible to quickly and effectively collect and provide information that employees need or are interested in.

[1334] An "internal bulletin board" is a digital bulletin board set up within a company or organization for the purpose of sharing information among employees.

[1335] "Means of obtaining information" refers to the mechanism by which the server automatically collects the content posted on the internal bulletin board.

[1336] "Means of analysis" refers to techniques that use natural language processing technology to analyze collected text data and understand its content.

[1337] "Means for categorizing by category" refers to a method for sorting analyzed text data into specific categories (e.g., business information, social information).

[1338] "Means for collecting employees' browsing and click histories" refers to technology for recording the operations performed by employees within the company's internal systems.

[1339] "Means for learning areas of interest" refers to a machine learning model that analyzes and learns from employee interests and concerns from collected historical data.

[1340] "Means of notification" refers to a system for informing employees of important information in real time based on the analysis results.

[1341] "Means for displaying recommended information on the dashboard" refers to an interface that displays information related to employees' areas of interest when they log in.

[1342] A "scraping tool" is software or a library used to automatically extract data from websites.

[1343] "Natural language processing technology" is a technology that enables computers to understand and process human language.

[1344] "Tokenization" refers to the process of dividing text data into smaller units such as words or sentences.

[1345] "Part-of-speech tagging" refers to labeling tokenized words with their parts of speech (e.g., nouns, verbs, adjectives).

[1346] A "generative AI model" is an artificial intelligence model that uses machine learning to learn patterns and rules from data and make predictions and classifications.

[1347] The present invention is a system that acquires information from an in-house bulletin board, extracts information that is relevant to employees' fields of interest or their work, and notifies them of the information. This system is configured as follows.

[1348] System configuration and operation procedures

[1349] This system is mainly composed of servers, terminals, and users, each of which has a specific role and functions in cooperation with each other.

[1350] Data collection and analysis methods

[1351] The server periodically scrapes information from the company's internal bulletin board. The scraping tools used are Beautiful Soup and Scrapy. The collected data is then analyzed using natural language processing (NLP) techniques. Specifically, the NLP libraries spaCy and BERT are used to tokenize the text, tag parts of speech, and analyze dependencies. Finally, the text data is classified by category (e.g., business information, social information).

[1352] How to collect employee behavioral data

[1353] (Device) records access history and click history every time an employee logs in to the company's internal system or browses information. This is done using JavaScript and web tracking technology. The collected history data is sent to (server) in real time and stored in a database.

[1354] Areas of interest and how to learn important information

[1355] The server analyzes the collected employee behavioral data and uses a generative AI model to learn each employee's areas of interest and work-related information. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used. The AI ​​model learns patterns and areas of interest based on this data.

[1356] How notifications and recommendations are displayed

[1357] (Server) scrutinizes newly collected information from internal bulletin boards based on the analysis results and selects information that is important to employees. (Server) notifies employees in real time using a notification system (e.g., Firebase Cloud Messaging, WebSocket). (Terminal) displays a notification as a pop-up on employees' PCs or smartphones and provides a link to view more information. (Server) also displays information of particular interest as recommended information on the dashboard that is displayed when employees log in.

[1358] Specific examples

[1359] Example 1: Information collection and analysis

[1360] (Server) accesses the company's internal bulletin board at 9:00 every morning and scrapes the latest posts. For example, this includes information such as "new project recruitment," "weekend club activities," and "drinking party next week." In the analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "project recruitment" is classified as business information, while "club activities" and "drinking party" are classified as social information.

[1361] Example 2: Learning and notifying employees about information of interest

[1362] On (terminal), employee A logs in to the system and records the history of clicking on the new project recruitment page. Based on this history data, (server) uses its AI model to learn that employee A has a high interest in "project recruitment." The next day, when a new project recruitment post is posted on the company bulletin board, (server) picks up the information and notifies employee A's device via Firebase Cloud Messaging. (terminal) displays a notification on employee A's PC stating "A new project recruitment post has been made," and employee A can click on the link to view more information.

[1363] Example 3: Displaying recommendations

[1364] When (Server) analyzes employee B's historical data, it finds that employee B frequently views pages related to club activities. Based on the results, (Server) selects club activity information that the AI ​​model determines to be of particular interest to employee B, and displays this information on the dashboard every Monday as "Recommended club activities for this weekend." When employee B logs in on Monday, (Device) displays "Recommended club activities for this weekend" on the dashboard.

[1365] This system will enable employees to obtain the information they need quickly and efficiently, which is expected to improve work efficiency and employee satisfaction.

[1366] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1367] Step 1:

[1368] (Server) accesses the company bulletin board every morning at 9:00 and scrapes new posts. The URL of the bulletin board is given as input, and the HTML structure is analyzed using a scraping tool (Beautiful Soup or Scrapy). The output is the extracted text data. Specifically, it performs the following operations:< / url:> , , ) to extract the text.

[1369] Step 2:

[1370] The text data acquired by the server is preprocessed using natural language processing (NLP) techniques. The input is the raw data acquired in step 1, and the output is the preprocessed text data. Specifically, NLP libraries (spaCy and BERT) are used to tokenize the data and remove unnecessary tags and spaces.

[1371] Step 3:

[1372] The server tokenizes the preprocessed text data and performs part-of-speech tagging and dependency analysis. The input is the data preprocessed in step 2, and the output is tokenized and tagged text. Specifically, it uses spaCy to perform structural analysis of the sentence and tag each word with a part-of-speech tag.

[1373] Step 4:

[1374] The server classifies the analyzed text data into categories (business information, social information). The input is the text tagged with part-of-speech tags in step 3, and the output is the classified data. Specifically, it uses a machine learning algorithm to classify each text into a predefined category.

[1375] Step 5:

[1376] (Terminal) records logins to the company's internal system, access history, and click history. The input is the user's operation log, and the output is the recorded access data. Specifically, JavaScript code captures the user's operations and sends them to the server.

[1377] Step 6:

[1378] The historical data collected by (the terminal) is sent to the server in real time and stored in the database. The input is the access data recorded in step 5, and the output is the historical data stored on the server. Specifically, the data is sent to the server via an HTTP request and stored in the database.

[1379] Step 7:

[1380] The server analyzes the collected employee behavioral data and uses a generative AI model to learn each employee's areas of interest and work-related information. The input is the historical data saved in step 6, and the output is the learned model and interest information. Specifically, the AI ​​model is trained using TensorFlow or PyTorch to learn patterns from the data.

[1381] Step 8:

[1382] (Server) scrutinizes the newly collected information from the internal bulletin board and picks out information that is important to employees. The input is the data from Step 1 and Step 7, and the output is the picked-up important information. Specifically, it filters the information based on the prediction results of the AI ​​model.

[1383] Step 9:

[1384] The server notifies employees in real time using a notification system (Firebase Cloud Messaging or WebSocket). The input is the information picked up in step 8, and the output is a notification message. Specifically, the server sends a message in real time through the notification system.

[1385] Step 10:

[1386] The device displays a pop-up notification and provides a link to view more information. The input is the notification message sent in step 9, and the output is the user's browsing behavior. The specific behavior is to display a pop-up window in a browser or application.

[1387] Step 11:

[1388] Information that the server determines to be of particular interest is displayed as recommended information on the dashboard when employees log in. The input is the data from steps 7 and 8, and the output is the recommended information displayed on the dashboard. Specifically, the settings are configured to list and display specific information on the dashboard.

[1389] Through the above processing steps, a system is provided that allows employees to quickly and efficiently obtain the information they need or are interested in.

[1390] (Application example 1)

[1391] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1392] Conventional systems that collect information from internal bulletin boards and notify employees have difficulty in properly identifying employees' areas of interest and necessary information. Furthermore, at production sites, proper information collection and notification is not performed, making it difficult to provide efficient work support or timely maintenance. For this reason, there is a need for a system that efficiently collects production schedules and maintenance information at factories and notifies workers appropriately.

[1393] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1394] In this invention, the server includes means for acquiring information from an in-house bulletin board, means for analyzing the acquired information and classifying it by category, means for collecting employees' browsing history and click history, means for analyzing the collected history information and learning employees' areas of interest, means for selecting and notifying employees of important information based on the analysis results, means for displaying information that is likely to be of particular interest as recommended information on the employee's dashboard, means for scraping production plans and function check information from an in-house management system, and means for analyzing the collected information and notifying it in conjunction with sensor data and work logs acquired during work. This allows employees and workers to quickly and efficiently receive the information they need, improving work efficiency and enabling appropriate maintenance.

[1395] An "internal bulletin board" is an electronic bulletin board system set up within a company for the purpose of sharing information and communicating.

[1396] "Means of obtaining information" refers to the methods and technologies used to collect the necessary information from various data sources.

[1397] "Means for analyzing information and classifying it into categories" refers to technology that analyzes acquired data and classifies it into specific categories based on its content.

[1398] "Means of collecting employees' browsing and click histories" refers to technology that collects the behavioral history of employees when they use web systems.

[1399] "Means of analyzing collected historical information to learn employees' areas of interest" refers to technology that analyzes collected behavioral data to learn what information a particular employee is interested in.

[1400] "Means of selecting and notifying employees of important information based on the analysis results" refers to methods and technologies for selecting important information from the analysis results and notifying employees of it.

[1401] "Means for displaying recommended information on employee dashboards" refers to a method for displaying information that is determined to be of particular interest to employees on their interface.

[1402] "Factory management system" refers to a comprehensive system for managing factory production and equipment.

[1403] "Means for scraping production plan and function check information" refers to technology that automatically extracts necessary information from factory management systems.

[1404] "Sensor data and work logs obtained during work" refers to the various data generated by robots and workers while they are working, as well as the recorded work history.

[1405] "Generative AI models" refer to algorithms and technologies that use machine learning techniques to generate knowledge and predictive models from data.

[1406] A "prompt" is an instruction given to a generative AI model to make it perform a specific task.

[1407] A "notification system" refers to technology that notifies individual devices of important information in real time.

[1408] The following describes an embodiment of the present invention. The present invention is a system that acquires information from a factory management system, notifies robots and workers of appropriate information, and improves work efficiency. This system functions as follows.

[1409] System configuration and operation

[1410] Data collection and analysis

[1411] 1. Server: The server periodically scrapes production plans and functional check information from the factory management system using web scraping tools such as Beautiful Soup and Scrapy.

[1412] 2. Server: The collected text data is analyzed using natural language processing (NLP) techniques. Specifically, NLP libraries such as spaCy and BERT are used to classify the text into categories (e.g., production information, maintenance information).

[1413] User behavior data collection

[1414] 3. Robot: The sensor data and work logs acquired by the robot while working are recorded and sent to the server in real time. The data is collected using JavaScript and web tracking technology.

[1415] 4. Server: The collected historical data is stored on the server.

[1416] Learn about areas of interest and important information

[1417] 5. Server: The server analyzes the collected robot history data and uses generative AI models to learn important information for each robot and worker. Python machine learning libraries (e.g., TensorFlow, PyTorch) are used.

[1418] 6. Server: Based on the analysis results, the newly collected information from the factory management system is scrutinized and important information is picked up.

[1419] Displaying notifications and recommendations

[1420] 7. Server: When important information is picked up, it is notified to the robot or worker in real time using a notification system (e.g. MQTT, WebSocket).

[1421] 8. Robot: Notifications are sent via the robot's display and voice output, which can be checked by the worker. Work instructions and changes are also displayed here.

[1422] Specific examples

[1423] Example 1: Information collection and analysis

[1424] Server: Every morning at 9:00, the server accesses the factory management system and scrapes the latest production schedule and maintenance schedule, including information such as "new production line setup," "machine maintenance next month," and "equipment inspection this weekend."

[1425] Server: In the analysis process, the collected text is tokenized using spaCy and classified into categories. For example, "production line settings" is classified as production information, while "machine maintenance" and "equipment inspection" are classified as maintenance information.

[1426] Example 2: Learning and notifying robot behavior information

[1427] Robot: Sensor data and work logs acquired by specific robots while performing machine maintenance are recorded.

[1428] Server: Based on this historical data, the AI ​​model learns the information necessary for the robot.

[1429] Server: When new machine maintenance information is posted on the factory management system, the server picks up the information and notifies the robot via MQTT.

[1430] Robot: A notification saying "New machine maintenance information has been added" appears on the robot's display and can be confirmed by the worker.

[1431] Example 3: Displaying recommendations

[1432] Server: By analyzing historical data for a particular robot, you may discover that the robot frequently performs tasks on a particular production line setup.

[1433] Server: Based on the results, the AI ​​model picks out information about the production line settings that it deems particularly important for that robot.

[1434] Server: Every morning, the server notifies the robot of this information and displays it on the robot's display as "Today's recommended production line settings."

[1435] In this way, the system of the present invention is designed to enable robots and workers to quickly and effectively obtain the information they need for their work, thereby improving work efficiency and ensuring appropriate maintenance.

[1436] Prompt Sentence Examples

[1437] "Next week's machine maintenance schedule"

[1438] "New Production Plan for Project X"

[1439] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1440] Step 1:

[1441] The server acquires production plan information and maintenance schedule information from the factory management system.

[1442] Input: Production plan information and maintenance schedule information from the factory management system.

[1443] Data processing: Use web scraping tools such as Beautiful Soup or Scrapy to extract the necessary information and save it as text data.

[1444] Output: Text data of extracted production plan information and maintenance schedule information.

[1445] Step 2:

[1446] The server analyzes the acquired text data and classifies it into categories.

[1447] Input: The text data extracted in step 1.

[1448] Data processing: Using natural language processing (NLP) techniques, we tokenize the text and classify it into categories (e.g., production information, maintenance information). Specifically, we use NLP libraries such as spaCy and BERT.

[1449] Output: Text data classified by category.

[1450] Step 3:

[1451] The robot records sensor data and work logs acquired during work and sends them to a server.

[1452] Input: Sensor data and work logs acquired while the robot is working.

[1453] Data processing: Collect data in real time and send it to the server. Record data using JavaScript and web tracking technology.

[1454] Output: Sensor data and work logs sent to the server.

[1455] Step 4:

[1456] The server analyzes the collected robot history data and uses a generative AI model to learn important information.

[1457] Input: Sensor data and work logs sent in step 3.

[1458] Data processing: Using Python machine learning libraries (e.g., TensorFlow, PyTorch), the data is analyzed to learn information that is important to each robot.

[1459] Output: Key information identified by the generative AI model.

[1460] Step 5:

[1461] The server picks out important information based on the analysis results and notifies you.

[1462] Input: Key information learned in step 4.

[1463] Data processing: Newly collected production plans and maintenance information are compared with the analysis results to select important information.

[1464] Output: Important information picked up.

[1465] Step 6:

[1466] The server uses a notification system to notify robots and workers of important information in real time.

[1467] Input: Important information picked up in step 5.

[1468] Data processing: Use MQTT or WebSocket to send notifications in real time.

[1469] Output: Notification message that will be displayed to the robot and worker.

[1470] Step 7:

[1471] The robot will display the notified information on a screen or output it as audio so that the worker can check it.

[1472] Input: Information provided in step 6.

[1473] Data processing: Display and audio output are performed so that workers can check the information.

[1474] Output: Notifications are displayed on the robot's display and audio output.

[1475] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1476] The following describes an embodiment of the present invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the areas of interest and emotional state of employees. The system is operated by combining a generative AI and an emotion engine.

[1477] System configuration and operation

[1478] Data collection and analysis

[1479] 1. (Server) Regularly scrape information from internal bulletin boards. For example, use tools like Scrapy or Beautiful Soup to collect the latest posts every morning at 9am.

[1480] 2. (Server) Analyze the collected text data using natural language processing (NLP) techniques, such as spaCy or BERT, to tokenize the text and classify it by category (e.g., business information, social information).

[1481] Collecting employee behavioral and emotional data

[1482] 3. (Device) Every time an employee logs in to an internal system and views information, their access history and click history are recorded. This is done using JavaScript and web tracking technology.

[1483] 4. (Device) Obtain recorded history data, as well as user-entered text and interaction data.

[1484] 5. (Terminal) Send the collected data to the server in real time.

[1485] Learning areas of interest and emotional states

[1486] 6. (Server) Stores the submitted historical and interaction data in a database and prepares it for analysis.

[1487] 7. (Server) Use a generative AI model to analyze historical data to learn employee interests and job-related information. Use Python machine learning libraries (e.g., TensorFlow, PyTorch).

[1488] 8. (Server) An emotion engine is used to analyze the user's text input and click behavior to identify their emotional state. Emotion recognition technology combines NLP technology with machine learning models (e.g., BERT, Keras).

[1489] Providing notifications and recommendations

[1490] 9. (Server) Based on the learning results, the server scrutinizes the newly collected information from the internal bulletin board and selects information that is important to employees. The server also adjusts the timing and content of notifications based on the results of the emotion engine.

[1491] 10. (Server) Use Firebase Cloud Messaging or WebSocket to send notifications to employees in real time.

[1492] 11. (Device) Notifications are displayed as pop-ups on employees' PCs or smartphones. The pop-ups display specific messages such as "A new project recruitment post has been made," and employees can click on a link to access more information.

[1493] 12. (Server) Information that is thought to be of particular interest is displayed as recommended information on the dashboard that employees log in to.

[1494] Specific examples

[1495] Example 1: Information collection and analysis

[1496] (Server) Every morning at 9:00, information is scraped from the company bulletin board to obtain information such as "New project recruitment," "Weekend club activities," and "Drinking parties next week."

[1497] (Server) Analyze the acquired text using spaCy or BERT and classify it into categories. For example, "Project recruitment" is classified as business information, while "Club activities" and "Drinking parties" are classified as social information.

[1498] Example 2: Learning and notifying employee interests and emotional states

[1499] (Device) Employee A logs in and clicks on the new project recruitment page. The text input and click history are recorded.

[1500] (Server) The generative AI model analyzes the historical data and learns that Employee A has a high interest in the "Project Recruitment" campaign. The emotion engine also detects "excitement" and "interest" from Employee A's input text.

[1501] (Server) The next day, if a new project recruitment post is made, a notification is sent to Employee A's device using Firebase Cloud Messaging. Based on the results of the emotion engine, the notification message is adjusted to read, "This project seems interesting to you."

[1502] (Device) A pop-up message appears on Employee A's PC saying "A new project recruitment post has been made," and clicking the link displays more information.

[1503] Example 3: Displaying recommendations

[1504] (Server) Analyzes employee B's historical data and emotional state and finds that he frequently views pages related to club activities and feels "excited" and "expected."

[1505] (Server) Based on this, information on club activities that are of particular interest to employee B is selected and displayed on the dashboard every Monday as "Recommended club activities for this weekend."

[1506] (Device) When employee B logs in on Monday, the dashboard will display "Recommended club activities for this weekend."

[1507] In this way, the system of the present invention provides employees with information necessary for their work and information of interest quickly and efficiently, and also provides appropriate information according to the user's emotional state, thereby improving work efficiency and employee satisfaction.

[1508] The processing flow will be explained below.

[1509] Step 1:

[1510] (Server) Set up periodic access to the internal bulletin board. For example, use Scrapy or Beautiful Soup to collect the latest postings every morning at 9am.

[1511] Step 2:

[1512] (Server) Analyze the collected text data using natural language processing (NLP) technology. Specifically, use spaCy or BERT to tokenize the text and classify it by category (e.g., business information, social information).

[1513] Step 3:

[1514] (Server) The analyzed text data is stored in a database by category. The stored data is used for later information review.

[1515] Step 4:

[1516] (Device) When an employee logs in to the company's internal system, JavaScript and web tracking technology are used to record access history and click history.

[1517] Step 5:

[1518] (Device) Recorded historical data is sent to the server in real time, making it immediately available for analysis.

[1519] Step 6:

[1520] (Server) Stores the submitted historical data in a database so that it can be used by the generative AI model and emotion engine for analysis.

[1521] Step 7:

[1522] (Server) Analyze historical data using a generative AI model to learn employee interests and work-related information. Build the learning model using Python machine learning libraries (e.g., TensorFlow, PyTorch).

[1523] Step 8:

[1524] (Server) An emotion engine is used to analyze user input data and click history to identify the user's emotional state. For example, NLP techniques and machine learning models are combined to identify emotions such as "excitement," "interest," and "boredom" from user text input.

[1525] Step 9:

[1526] (Server) Based on the analysis results, the server scrutinizes newly collected information from internal bulletin boards and selects information that is important to employees. At this time, the server adjusts the timing and content of notifications according to the user's emotional state as determined by the emotion engine.

[1527] Step 10:

[1528] (Server) Notifications are sent to employees in real time using Firebase Cloud Messaging and WebSockets. The content and format of the notifications are customized taking into account the results of the emotion engine.

[1529] Step 11:

[1530] (Device) A notification pops up on the employee's PC or smartphone. For example, a message saying "A new project recruitment post has been made" appears. When the user clicks on the link, more information is displayed.

[1531] Step 12:

[1532] (Server) In addition, prepare to display information that is thought to be of particular interest as "recommended information" on employees' dashboards.

[1533] Step 13:

[1534] (Device) When an employee logs in, "recommended information" is displayed on the dashboard. For example, information such as "recommended club activities for this weekend" is displayed.

[1535] This system allows employees to quickly and efficiently obtain information necessary for their work and information of interest. In addition, because it provides appropriate information based on the user's emotional state, it is expected to improve work efficiency and employee satisfaction.

[1536] Example 2

[1537] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1538] Conventional internal information sharing systems make it difficult for employees to efficiently obtain the information they need, resulting in problems such as missing information or failing to grasp important information. Furthermore, to improve the quality of communication within a company and increase the work efficiency of each employee, it is necessary to provide information that takes into account the interests and emotional state of employees, but there were few systems that could achieve this. Furthermore, the lack of information provided that was tailored to the interests of each employee sometimes resulted in information overload, which in turn reduced productivity.

[1539] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1540] In this invention, the server includes a means for acquiring information from the in-house bulletin board, a means for analyzing the acquired information using natural language processing technology and classifying it into categories, and a means for collecting collected history information and user input text in real time. This makes it possible to efficiently collect information needed by employees and provide important information at an appropriate time, taking into account the employees' areas of interest and emotional state.

[1541] 1. An "internal bulletin board" is an online platform for sharing information within a company or organization.

[1542] 2. "Means of obtaining information" refers to tools and programs for automatically collecting necessary information from internal bulletin boards.

[1543] 3. "Natural language processing technology" is an artificial intelligence technology for analyzing and understanding human language.

[1544] 4. "Categorization methods" refers to algorithms or programs that group collected information based on specific criteria.

[1545] 5. "Browsing history" refers to a record of web pages and documents that a user has previously viewed.

[1546] 6. "Click History" refers to a record of links and buttons that a user has previously clicked.

[1547] 7. "History information" refers to all data relating to a user's behavioral history.

[1548] 8. "User Input Text" means any characters or sentences entered by a User into the System.

[1549] 9. "Real-time collection means" refers to technologies or methods that capture data as it is transmitted.

[1550] 10. "Database" refers to a system for efficiently storing, managing, and retrieving large amounts of data.

[1551] 11. "Generative AI model" refers to an artificial intelligence model that learns from generated data and performs tasks such as prediction and classification.

[1552] 12. "Emotion Engine" refers to technology or programs that identify a user's emotional state from their text input or behavior.

[1553] 13. "Interest Areas" refers to the themes or topics that interest a particular user.

[1554] 14. "Business-Related Information" means information related to a User's job function.

[1555] 15. "Means of notification" refers to tools and technologies used to communicate information to users.

[1556] 16. "Recommended Information" refers to the provision of specific information selected based on a user's interests and behavior.

[1557] The following describes an embodiment of the present invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the areas of interest and emotional state of employees. The system is operated by combining a generative artificial intelligence model and an emotion analysis engine.

[1558] System configuration and operation

[1559] Data collection and analysis

[1560] 1. The server periodically retrieves information from the company bulletin board. To collect the information, use a web scraping tool such as Scrapy or Beautiful Soup. For example, set it to collect the latest posts every morning at 9:00.

[1561] 2. The server analyzes the collected text data using natural language processing (NLP) techniques, such as spaCy and BERT, to tokenize the text and classify it into categories (e.g., business information, social information).

[1562] Collecting employee behavioral and emotional data

[1563] 3. The device records access and click histories every time an employee logs into the company's internal systems and views information. This recording is done using JavaScript and web tracking technology.

[1564] 4. The device acquires the recorded history data, text entered by the user, and interaction data, and transmits them to the server in real time.

[1565] Saving data and preparing for analysis

[1566] 5. The server stores the transmitted history and interaction data in a database and prepares it for analysis, using a database system such as PostgreSQL or MongoDB.

[1567] Learning areas of interest and emotional states

[1568] 6. The server uses generative artificial intelligence models (e.g., TensorFlow, PyTorch) to analyze historical data and learn employee interests and work-related information.

[1569] 7. The server uses an emotion engine (e.g., BERT, Keras) to identify the user's emotional state from their text input and click behavior.

[1570] Providing notifications and recommendations

[1571] 8. Based on the learning results and sentiment analysis results, the server examines the newly collected information from the internal bulletin board and notifies employees of important information. For example, it prepares a tailored message such as, "This project seems interesting to you."

[1572] 9. The server uses Firebase Cloud Messaging or WebSocket to send notifications in real time to employee devices, which are displayed as pop-ups on PCs and smartphones.

[1573] 10. The server displays information of particular interest as recommended information on the dashboard screen when employees log in. For example, information on club activities of high interest is displayed on the dashboard every Monday as "Recommended activities for this weekend."

[1574] Specific examples

[1575] Example 1: Information collection and analysis

[1576] The server scrapes information from the company's internal bulletin boards every morning at 9 a.m., obtaining information such as "new project recruitment," "weekend social events," and "next week's meetings."

[1577] The server analyzes the acquired text using natural language processing technology (spaCy or BERT) and classifies it into categories. For example, "project recruitment" is classified as business information, while "social events" and "meetings" are classified as social information.

[1578] Example 2: Learning and notifying employee interests and emotional states

[1579] The device records access history and click history when employee A logs in and clicks on the page recruiting for new projects.

[1580] The server analyzes the historical data based on a generative artificial intelligence model and learns that Employee A has a high interest in the "Project Recruitment." The emotion engine also detects "excitement" and "interest" from Employee A's input text.

[1581] The next day, if there is a new project opening, the server will use Firebase Cloud Messaging to notify employee A's device, saying, "We think this project might be interesting to you."

[1582] The device displays a pop-up on Employee A's PC saying, "A new project recruitment post has been made," and when Employee A clicks on the link, more information will be displayed.

[1583] Example 3: Displaying recommendations

[1584] The server analyzes employee B's historical data and emotional state, and determines that he frequently views pages related to club activities and feels emotions of "excitement" and "expectation."

[1585] Based on this, the server selects information about club activities that employee B is particularly interested in and displays it on the dashboard every Monday as "Recommended club activities for this weekend."

[1586] When employee B logs in to the device on Monday, the dashboard will display "Recommended club activities for this weekend."

[1587] Example prompt sentence:

[1588] "Collect information on new project openings and notify employees."

[1589] "Identify areas of interest from employees' click history"

[1590] "Analyze the user's emotional state using an emotion engine and provide appropriate notifications."

[1591] This system allows employees to quickly and efficiently receive not only the information they need for their work, but also appropriate information tailored to their individual interests and emotional state, thereby improving work efficiency and employee satisfaction.

[1592] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1593] Step 1:

[1594] The server scrapes information from the company bulletin board every morning at 9:00. It uses the bulletin board URL as input and generates data including the title, body, and posting date and time of each post as output. Specifically, it uses Scrapy and Beautiful Soup to parse the HTML and extract the required text information.

[1595] Step 2:

[1596] The server analyzes the collected text data using natural language processing technology and classifies it into categories. It uses the acquired text data as input and obtains data classified into categories (e.g., business information, social information) as output. Specifically, it uses spaCy and BERT to tokenize the text and perform contextual analysis.

[1597] Step 3:

[1598] The device records access and click histories every time an employee logs into the company's internal systems and browses information. It uses the employee's browsing and clicking behavior as input and generates the recorded history data as output. Specifically, it captures user actions using JavaScript and web tracking technology.

[1599] Step 4:

[1600] The device sends the recorded history data and the user's input text and interaction data to the server in real time. The device uses the recorded history data and the user's text data as input and generates the data sent to the server as output. Specifically, the device converts the data format to JSON or similar and sends it to the server using the HTTPS protocol.

[1601] Step 5:

[1602] The server stores the submitted historical and interaction data in a database and prepares it for analysis. It uses the received historical and interaction data as input and generates the data stored in the database as output, specifically by inserting data using database queries.

[1603] Step 6:

[1604] The server uses a generative AI model to analyze historical data and learn employee interests and work-related information. It uses the stored historical data as input and obtains data identifying employee interests as output. Specifically, it trains the model using TensorFlow or PyTorch and makes predictions based on the historical data.

[1605] Step 7:

[1606] The server uses an emotion engine to identify the user's emotional state from their text input and click behavior. It uses the user's text data and click data as input and obtains the identified emotional state as output. Specifically, it uses BERT and Keras to perform emotion recognition.

[1607] Step 8:

[1608] Based on the learning results and sentiment analysis results, the server scrutinizes the newly collected information from the internal bulletin board, picks out information that is important to employees, and prepares the notification content. It uses the data obtained from the internal bulletin board and the analysis results as input, and generates a notification message as output. Specifically, it selects an appropriate template and creates the message.

[1609] Step 9:

[1610] The server sends notifications to employee devices in real time using Firebase Cloud Messaging or WebSocket. It uses the prepared notification message as input and gets the sent notification as output. Specifically, it sends messages using the Firebase API or WebSocket protocol.

[1611] Step 10:

[1612] The device displays the received notification as a pop-up on the employee's PC or smartphone. It uses the received notification message as input and obtains the displayed pop-up notification as output. Specifically, it uses the browser's Notification API or the OS's notification function to display the message.

[1613] Step 11:

[1614] The server then displays the most interesting information as recommendations on the employee's dashboard. It uses information based on the areas of interest as input and generates recommendations that are displayed on the dashboard as output. Specifically, it dynamically updates the content of the web page and applies templates to display the information of interest.

[1615] (Application example 2)

[1616] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1617] Conventional internal bulletin board systems and notification systems do not provide information that takes into account the interests and emotional state of individual employees, limiting their ability to improve work efficiency and employee satisfaction. Furthermore, it is difficult to provide employees with the important information they need at the right time, leading to problems with missed information and excessive notifications. There is a need for a system that can solve these issues and provide optimal information to employees while improving work efficiency.

[1618] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring information from the in-house bulletin board, means for analyzing the acquired information and classifying it by category, and means for analyzing the collected history data and interaction data to identify the user's emotional state. This makes it possible to pick out important information based on the employee's areas of interest and emotional state, appropriately adjust the content of notification messages, and provide them to employees in real time.

[1619] An "internal bulletin board" is an online bulletin board for the purpose of sharing information and communicating among employees.

[1620] "Means of obtaining information" refers to the techniques and methods used to collect the necessary information from the internal bulletin board.

[1621] "Means of analyzing information" refers to the techniques and methods used to structure collected information and convert it into an easily understandable format.

[1622] "Categorization means" refers to techniques or methods for classifying analyzed information into different categories.

[1623] "Means of collecting employee browsing and click history" refers to technologies and methods that record what information employees view and what links they click.

[1624] "Means for analyzing historical information and learning about employees' areas of interest" refers to techniques and methods for identifying employees' areas of interest and concern from collected historical information.

[1625] "Means for analyzing interaction data and identifying the user's emotional state" refers to technologies and methods for analyzing the emotions of employees at that time based on their interaction data.

[1626] "Means of notification" refers to the technology and methods used to inform employees of necessary information based on the analysis results.

[1627] "Means for adjusting the content of notification messages" refers to techniques and methods for changing notification messages to optimal content depending on the emotional state of employees.

[1628] "Means for displaying recommended information on the dashboard" refers to the technology and methods for displaying information of particular interest on employees' operation screens.

[1629] "Means of real-time notification" refers to technologies and methods that instantly inform employees of the information they need at that moment.

[1630] The following describes an embodiment of this invention. The present invention is a system that acquires information from an internal bulletin board, extracts information necessary for work, and notifies it, taking into consideration the employee's areas of interest and emotional state. The system is operated by combining a generative AI and an emotion engine.

[1631] System configuration and operation

[1632] Data collection and analysis

[1633] The server periodically collects information from internal bulletin boards using scraping technologies such as Scrapy and Beautiful Soup. The collected information is analyzed and categorized using natural language processing (NLP) techniques. Specifically, text is tokenized using spaCy and BERT, and then classified into categories such as business information and social information.

[1634] Collecting employee behavioral and emotional data

[1635] The device uses JavaScript and web tracking technology to record historical data and click history when employees access internal systems. It also collects text entered by users and interaction data, which it then transmits in real time to a server where it is prepared for analysis.

[1636] Learning areas of interest and emotional states

[1637] The server uses a generative AI model built with TensorFlow and PyTorch to learn employee interests and work-related information from historical data. It also uses an emotion engine using BERT and Keras to analyze users' text input and click behavior to identify their emotional state. For example, if an employee expresses emotions such as "excitement" or "interest," that information is identified by the emotion recognition model.

[1638] Providing notifications and recommendations

[1639] Based on the analysis results, the server scrutinizes newly collected information from internal bulletin boards and selects information that is important to employees. The timing and content of notifications are adjusted taking into account the results of the emotion engine. Notifications are sent to employees in real time using Firebase Cloud Messaging and WebSocket. These notifications are displayed as pop-ups on employees' PCs or smartphones. Additionally, information that is deemed to be of particular interest is displayed as recommended information on the dashboard that employees log in to.

[1640] Specific examples

[1641] Information collection and analysis

[1642] The server scrapes information from the company's internal bulletin board at 9 a.m. every morning, obtaining information such as new project recruitment, weekend club activities, and next week's drinking party. The obtained text is analyzed using spaCy and BERT and classified by category. For example, "project recruitment" is classified as business information, and "club activities" is classified as social information.

[1643] Learn and notify employee interests and emotional states

[1644] The device records the text input and click history when Employee A logs in and clicks on the new project recruitment page. The server analyzes the historical data using a generative AI model and learns that Employee A has a high interest in the "project recruitment." The emotion engine also detects "excitement" and "interest" from Employee A's input text. The next day, when a new project recruitment post is made, a notification is sent to Employee A's device using Firebase Cloud Messaging. Based on the results of the emotion engine, the notification message is adjusted to read, "We think this project might be interesting to you." A pop-up appears on Employee A's PC saying, "A new project recruitment post has been made," and the user can click the link to access more information.

[1645] Displaying recommendations

[1646] The server analyzes employee B's historical data and emotional state, and determines that he frequently views pages related to club activities and that he feels "excitement" and "expectations." Based on this, information on club activities that are of particular interest to employee B is selected and displayed on the dashboard every Monday as "Recommended club activities for this weekend." When employee B logs in on Monday, "Recommended club activities for this weekend" is displayed on the dashboard.

[1647] Prompt Sentence Examples

[1648] "Have a generative AI model perform sentiment analysis on food delivery requests and send relevant notifications for appropriate delivery jobs if the sentiment score is high."

[1649] In this way, the system of the present invention can quickly and efficiently provide employees with information necessary for their work and information of interest, and can also provide appropriate information according to the user's emotional state, thereby improving work efficiency and employee satisfaction.

[1650] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1651] Step 1:

[1652] Collect information from an internal bulletin board. The server uses a scraping tool such as Scrapy or Beautiful Soup to retrieve the contents of the internal bulletin board at a set time (for example, 9:00 every morning). The input is a URL and necessary initial information, and the output is the latest posted information as text data.

[1653] Step 2:

[1654] The collected information is analyzed and classified by category. The server uses natural language processing (NLP) technology to analyze the acquired text data and perform tokenization and parsing. Specifically, it uses spaCy and BERT to classify it into categories such as business information and social information. The input is the collected text data, and the output is information classified by category.

[1655] Step 3:

[1656] This system collects employee behavioral and emotional data. The device uses JavaScript and web tracking technology to record the browsing and click history of employees when they access internal systems in real time. The input is the employee's interactions (clicks, page transitions, etc.), and the output is the recorded behavioral data.

[1657] Step 4:

[1658] The collected history data and interaction data are sent to the server. The terminal transfers and stores the behavioral data in real time to the server. The input is the behavioral data recorded in real time, and the output is the behavioral data stored on the server side.

[1659] Step 5:

[1660] Analyze areas of interest and emotional state. The server uses a generative AI model (using TensorFlow or PyTorch) to analyze historical data and identify employees' areas of interest. It then uses an emotion engine (using BERT or Keras) to analyze emotional states from interaction data. The input is historical data and interaction data, and the output is the identified areas of interest and emotional state.

[1661] Step 6:

[1662] Important information is extracted and notification messages are generated. Based on the analysis results, the server scrutinizes the newly collected information from the internal bulletin board and selects information that is important to employees. It also adjusts the content of the notification message taking into account the results of the emotion engine. The input is the analyzed data and new internal bulletin board information, and the output is an optimized notification message.

[1663] Step 7:

[1664] Send notifications in real time. The server uses Firebase Cloud Messaging or WebSocket to send optimized notification messages to employee devices in real time. The input is the generated notification message, and the output is the notification displayed on the employee's device.

[1665] Step 8:

[1666] Display as recommended information. The server configures the information that is deemed to be of particular interest to be displayed on the employee's dashboard. The input is the identified areas of interest and selected information, and the output is the recommended information displayed on the employee's dashboard.

[1667] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1668] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1669] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1670] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1671] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1672] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1673] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1674] Human emotio...

Claims

1. A means of obtaining information from the company bulletin board; A means for analyzing the acquired information and classifying it into categories; A means of collecting employees' browsing and clicking histories, A means of analyzing the collected historical information to learn about employee areas of interest; Based on the analysis results, a means of picking out and notifying employees of important information, A way to display information that is likely to be of particular interest as recommended information on employees' dashboards, and A system including:

2. A method for recording employee access history and click history using JavaScript and web tracking technology and sending them to a server. Includes a means to analyze employee history data recorded on a server and identify areas of interest using a generative AI model. The system of claim 1 .

3. A method for periodically scraping and collecting information from internal bulletin boards, A means of analyzing the collected text data using natural language processing technology and classifying it into categories; Includes a way to notify employees in real time using Firebase Cloud Messaging and WebSockets The system of claim 1 .

Images