system

Abstract

The system according to the embodiment aims to automatically reduce the risk of information leaks and flame wars caused by employee posts. [Solution] A system according to an embodiment includes a learning unit, a checking unit, a correcting unit, a warning unit, and a generating unit. The learning unit learns a security policy. The checking unit checks posted text based on the security policy learned by the learning unit. The correcting unit automatically corrects posted text checked by the checking unit. The warning unit displays a warning based on confidential information or a risk of flaming detected by the checking unit. The generating unit generates a report based on the results of the checking unit and the correcting unit.

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] With conventional technology, checks of employee posts to reduce the risk of information leaks and online outrage were carried out manually, making it difficult to respond efficiently and accurately.

[0005] The system according to the embodiment aims to automatically reduce the risk of information leaks and flame wars caused by employee posts. [Means for solving the problem]

[0006] The system according to the embodiment includes a learning unit, a checking unit, a correcting unit, a warning unit, and a generating unit. The learning unit learns a security policy. The checking unit checks posted text based on the security policy learned by the learning unit. The correcting unit automatically corrects posted text checked by the checking unit. The warning unit displays a warning based on confidential information or a risk of flaming detected by the checking unit. The generating unit generates a report based on the results of the checking unit and the correcting unit. [Effects of the Invention]

[0007] The system according to the embodiment can automatically reduce the risk of information leaks and online outrage caused by employee posts. [Brief explanation of the drawings]

[0008] [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. DETAILED DESCRIPTION OF THE INVENTION

[0009] 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.

[0010] First, the terms used in the following description will be explained.

[0011] 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, the 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), an APU (Accelerated Processing Unit), or a TPU (Tensor Processing Unit).

[0012] 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.

[0013] 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.

[0014] 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), and Bluetooth (registered trademark).

[0015] 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."

[0016] [First embodiment] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0017] 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.

[0018] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, RAM 30, and 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).

[0019] 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.

[0020] The reception device 38 includes a touch panel 38A and a microphone 38B, and receives user input. The touch panel 38A detects contact with a pointer (for example, a pen or a finger) to receive user input by the touch of the pointer. 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 (see FIG. 2) acquires the data indicating the user input.

[0021] Output device 40 includes a display 40A and a speaker 40B, and presents data to a user by outputting the data in a form of expression that the user can perceive (e.g., audio and / or text). Display 40A displays visible information such as text and images in accordance with instructions from processor 46. Speaker 40B outputs audio in accordance with instructions from processor 46. 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.

[0022] 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.

[0023] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0024] 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.

[0025] 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. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0026] In the smart device 14, the specific processing is performed by the processor 46. The storage 50 stores a specific processing program 60. The specific processing program 60 is used together with the specific processing program 56 by the data processing system 10. The processor 46 reads the specific processing program 60 from the storage 50 and executes the read specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as the control unit 46A in accordance with the specific processing program 60 executed on the RAM 48. Note that the smart device 14 has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can also perform processing similar to that of the specific processing unit 290 using these models.

[0027] Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device (e.g., a generation server) may have the data generation model 58. In this case, the data processing device 12 obtains a processing result (prediction result, etc.) using the data generation model 58 by communicating with the server device having the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device owned by a user (e.g., a mobile phone, a robot, a home appliance, etc.). Next, an example of processing by the data processing system 10 according to the first embodiment will be described.

[0028] (Example 1) An information leak and flame war risk reduction system according to an embodiment of the present invention uses a generation AI to automatically check employee posts and reduce the risk of information leaks and flame wars. In this system, the generation AI automatically learns the company's security policy, pre-checks the posts when employees create them, and automatically corrects them as necessary. If the posts contain confidential information or contain expressions that pose a high risk of flame wars, a warning is displayed, and a report is generated based on the results of the post check and provided to the administrator. This enables accurate and appropriate decisions to be made at all times, reducing the risk of information leaks and flame wars. Furthermore, when combined with employee training, more reliable security measures are realized. For example, when the generation AI automatically learns the company's security policy, it analyzes the contents of the security policy in detail and understands its contents. For example, it learns how to handle confidential information and lists information that must not be made public. Next, when an employee creates a post, the generation AI pre-checks the post. The generation AI analyzes the content of the post to ensure it does not violate the security policy. For example, if the post contains confidential information or contains expressions that pose a high risk of flame wars, a warning is displayed. Furthermore, the generation AI automatically corrects posts as needed. For example, if confidential information is included, it deletes that portion or replaces it with appropriate wording. This reduces the risk of employees unintentionally disclosing confidential information. The generation AI also generates a report based on the results of the post check. This report includes the content of the post, the check results, and any corrections. The report is provided to administrators, allowing them to understand the status of employees' posts. This enables them to make accurate and appropriate decisions at all times, reducing the risk of information leaks and online outrages. Furthermore, when combined with employee training, more reliable security measures can be achieved. For example, educating employees about the importance of security policies and providing feedback on the generation AI's check results can raise employee awareness. This allows the information leak and online outrage risk reduction system to automatically check employees' posts and reduce the risk of information leaks and online outrages.

[0029] An information leakage and flame war risk reduction system according to an embodiment includes a learning unit, a checking unit, a correcting unit, a warning unit, and a generating unit. The learning unit learns a security policy. Examples of security policies include, but are not limited to, information leakage prevention, access control, and data protection. The learning unit learns, for example, a list of confidential information, a list of prohibited information, and a list of expressions that pose a high risk of flame wars. The checking unit checks posted messages based on the security policy learned by the learning unit. The checking unit checks posted messages using, for example, a natural language processing algorithm. Examples of natural language processing algorithms include, but are not limited to, morphological analysis, grammatical analysis, and semantic analysis. The correcting unit automatically corrects posted messages checked by the checking unit. For example, the correcting unit deletes messages that contain confidential information and replaces them with appropriate expressions that pose a high risk of flame wars. Examples of appropriate expressions include, but are not limited to, positive expressions and neutral expressions. The warning unit displays a warning based on the confidential information and flame war risks detected by the checking unit. The warning unit displays a warning, for example, in the form of a pop-up. Examples of pop-ups include, but are not limited to, warning messages and alerts. The generation unit generates a report based on the results of the check unit and the correction unit. The generation unit generates a report including, for example, the content of the posted message, the check results, the corrections, and the warning, and provides the report to the administrator by email. In this way, the information leak and flame war risk reduction system according to the embodiment can automatically check the messages posted by employees and reduce the risk of information leaks and flame wars.

[0030] The learning unit can learn a list of confidential information, a list of information that should not be disclosed, or a list of expressions that have a high risk of causing an uproar. The learning unit, for example, learns a list of confidential information. Examples of confidential information include, but are not limited to, personal information, trade secrets, and financial information. The learning unit can also learn a list of information that should not be disclosed. Examples of information that should not be disclosed include, but are not limited to, legal regulations and corporate policies. The learning unit can also learn a list of expressions that have a high risk of causing an uproar. Examples of expressions that have a high risk of causing an uproar include, but are not limited to, past cases of uproars and sentiment analysis results. By learning a specific list, the accuracy of the security policy is improved. Some or all of the above-described processing in the learning unit may be performed using, for example, AI, or may be performed without using AI. For example, the learning unit can input a list of confidential information into a generation AI and have it learn the contents of the list.

[0031] The check unit can check the posted text using a natural language processing algorithm. The check unit checks the posted text using, for example, morphological analysis. Morphological analysis is a technique that divides a sentence into words and analyzes the part of speech of each word. The check unit can also check the posted text using grammatical analysis. Grammatical analysis is a technique that analyzes the grammatical structure of a sentence and detects grammatical errors. The check unit can also check the posted text using semantic analysis. Semantic analysis is a technique that analyzes the meaning of a sentence and determines the appropriate meaning based on the context. Thus, using a natural language processing algorithm improves the accuracy of checking the posted text. Some or all of the above-mentioned processing in the check unit may be performed using, for example, AI, or may be performed without using AI. For example, the check unit can input the posted text to a generation AI and have the generation AI perform natural language processing.

[0032] The correction unit can delete confidential information if it contains confidential information, and replace expressions that pose a risk of causing an uproar with appropriate expressions if it contains confidential information. For example, the correction unit can delete confidential information if it contains confidential information. Confidential information includes, but is not limited to, personal information, trade secrets, and financial information. The correction unit can also replace expressions that pose a high risk of causing an uproar with appropriate expressions if it contains expressions that pose a high risk of causing an uproar. Appropriate expressions include, but are not limited to, positive expressions and neutral expressions. This reduces the risk by automatically correcting confidential information and expressions that pose a high risk of causing an uproar. Some or all of the above-described processing by the correction unit can be performed using, for example, AI, or without AI. For example, the correction unit can input the posted text into a generation AI and have the generation AI perform automatic corrections.

[0033] The warning unit can display a warning in a pop-up. For example, the warning unit displays a warning in a pop-up if the posted text contains confidential information. The pop-up displays a warning message such as, "This posted text contains confidential information. Do not make it public." The warning unit can also display a warning in a pop-up if the posted text contains a high risk of causing an uproar. The pop-up displays a warning message such as, "This posted text contains a high risk of causing an uproar. Please correct it." By displaying a warning in a pop-up, the user can be immediately notified of the risk. Some or all of the above-described processing in the warning unit may be performed using, or without, AI, for example. For example, the warning unit can input the posted text into a generation AI and cause the generation AI to display a warning.

[0034] The generation unit can generate a report including the content of the posted message, the check results, the corrections, and the warnings, and provide it to the administrator by email. The generation unit generates a report including, for example, the content of the posted message, the check results, the corrections, and the warnings. The report includes, for example, the full text of the posted message, details of the check results, a list of corrections, and details of the warnings. The generation unit can also provide the generated report to the administrator by email. The email includes, for example, an overview of the report and important points. By generating a report and providing it to the administrator, the status of the posted message can be understood. Some or all of the above-described processing by the generation unit may be performed using, for example, AI, or may be performed without using AI. For example, the generation unit can input the content of the posted message into a generation AI and cause the generation AI to generate a report.

[0035] During learning, the learning unit can optimize the learning algorithm by referring to past security incidents. For example, the learning unit refers to past information leak cases and reflects similar patterns in the learning algorithm. Examples of information leak cases include, but are not limited to, unauthorized access to databases and erroneous email transmissions. The learning unit can also refer to past flaming cases and add high-risk expressions to the learning algorithm. Examples of flaming cases include, but are not limited to, inappropriate comments on social media and misleading expressions in blog posts. Furthermore, the learning unit can analyze the causes of past security incidents and improve the learning algorithm. Examples of the causes of security incidents include, but are not limited to, system vulnerabilities and user errors. By referring to past security incidents, the accuracy of the learning algorithm can be improved. Some or all of the above-described processing in the learning unit may be performed using, or without, AI. For example, the learning unit can input past security incident data into the generation AI and cause the generation AI to optimize the algorithm.

[0036] The learning unit can take into account differences in security policies between industries during learning. For example, the learning unit can learn security policies for the financial industry to consider risks specific to the financial industry. Security policies for the financial industry include, but are not limited to, protection of customer information and encryption of transaction data. The learning unit can also learn security policies for the medical industry to consider risks related to the handling of medical data. Security policies for the medical industry include, but are not limited to, protection of patient information and security of electronic medical records. The learning unit can also learn security policies for the IT industry to consider technical risks. Security policies for the IT industry include, but are not limited to, system vulnerability management and data backup. This improves the accuracy of learning by taking into account security policies for each industry. Some or all of the above-described processing in the learning unit can be performed using, for example, AI, or without AI. For example, the learning unit can input industry-specific security policy data into a generation AI and cause the generation AI to perform learning.

[0037] During learning, the learning unit can customize the learning content according to the company's growth stage. For example, the learning unit provides learning content on basic security policies for a start-up company. Basic security policies include, but are not limited to, password management and basic data protection methods. The learning unit can also provide learning content on detailed and advanced security policies for a mature company. Detailed and advanced security policies include, but are not limited to, the application of encryption technology and security incident response methods. Furthermore, the learning unit can provide learning content focused on risk management for a company in the expansion phase. Learning content focused on risk management includes, but is not limited to, risk assessment methods and risk mitigation measures. This enhances the effectiveness of learning by providing learning content tailored to the company's growth stage. Some or all of the above-described processing in the learning unit may be performed using, or without, AI. For example, the learning unit can input company growth stage data into the generation AI and cause the generation AI to customize the learning content.

[0038] The learning unit can take into account regional laws and regulations during learning. For example, the learning unit can learn security policies based on the European General Data Protection Regulation (GDPR). GDPR includes, but is not limited to, data protection principles and data subject rights. The learning unit can also learn medical data handling based on the American HIPAA. HIPAA includes, but is not limited to, protection of medical information and patient privacy rights. The learning unit can also learn data management based on the Japanese Act on the Protection of Personal Information. The Act includes, but is not limited to, proper handling of personal information and safety management measures for personal information. This improves the accuracy of learning by taking regional laws and regulations into consideration. Some or all of the above-described processing in the learning unit can be performed using AI, for example, or without AI. For example, the learning unit can input regional legal and regulatory data into the generation AI and cause the generation AI to perform learning.

[0039] The checking unit can improve the accuracy of the check by taking into account the context of the posted text. For example, the checking unit analyzes the context before and after the posted text and performs an appropriate check. The context before and after the posted text includes, but is not limited to, the preceding and following sentences, related topics, etc. The checking unit can also understand the subject of the posted text and perform a check based on the context. The subject includes, but is not limited to, the theme of the posted text, the main points of discussion, etc. Furthermore, the checking unit can also check the posted text to avoid misunderstandings by taking into account the context of the posted text. Checks to avoid misunderstandings include, but are not limited to, correcting ambiguous expressions and expressions that do not fit the context. This improves the accuracy of the check by taking into account the context of the posted text. Some or all of the above-described processing by the checking unit may be performed using, for example, AI, or without AI. For example, the checking unit can input context data of the posted text into a generation AI and have the generation AI perform context analysis and check.

[0040] The checking unit can apply different checking algorithms depending on the category of the posted text during the check. For example, the checking unit can apply a checking algorithm that takes into account technical risks to technical documents. Examples of technical documents include, but are not limited to, technical details and terminology. The checking unit can also apply a checking algorithm that takes into account brand image to marketing documents. Examples of marketing documents include, but are not limited to, brand messages and advertising expressions. The checking unit can also apply a checking algorithm that takes into account the risk of confidential information leaks to internal documents. Examples of internal documents include, but are not limited to, internal reports and meeting records. This improves the accuracy of the check by applying a checking algorithm according to the category of the posted text. Some or all of the above-described processing by the checking unit can be performed using, for example, AI, or without AI. For example, the checking unit can input category data of the posted text into a generation AI and cause the generation AI to apply a checking algorithm.

[0041] The checking unit may determine the priority of checks based on the submission date of each post. For example, the checking unit may prioritize urgent posts. Urgent posts include, but are not limited to, important announcements and emergency notifications. The checking unit may also check regularly posted posts with normal priority. Regular posts include, but are not limited to, regular reports and periodic updates. The checking unit may also postpone checking past posts. Past posts include, but are not limited to, archived documents and past reports. This improves the effectiveness of checks by determining the priority of checks based on the submission date of each post. Some or all of the above-described processing by the checking unit may be performed using, or without, AI. For example, the checking unit may input data on the submission date of each post into the generation AI, causing the generation AI to determine the priority.

[0042] The checking unit can improve the accuracy of the check by referring to literature related to the posted text during the check. For example, the checking unit can improve the accuracy of the check by referring to past literature related to the posted text. Past literature includes, but is not limited to, related academic papers, technical reports, industry news, etc. The checking unit can also refer to literature related to the subject of the posted text to perform an appropriate check. Literature related to the subject includes, but is not limited to, research papers on the same topic and related technical documents. The checking unit can also refer to literature related to the context of the posted text to check to avoid misunderstandings. Literature related to the context includes, but is not limited to, related background information, past cases, etc. In this way, referring to literature related to the posted text improves the accuracy of the check. Some or all of the above-described processing by the checking unit can be performed using, for example, AI, or without AI. For example, the checking unit can input related literature data into the generation AI and have the generation AI perform literature reference and check.

[0043] When making a correction, the correction unit can select the optimal correction method by referring to the past correction history of the posted text. For example, the correction unit selects the optimal correction method by referring to the history of similar corrections made in the past. Past correction history includes, but is not limited to, a correction log, a correction history database, etc. The correction unit can also learn and apply effective correction methods from the past correction history. Effective correction methods include, but are not limited to, specific correction patterns, successful correction cases, etc. Furthermore, the correction unit can analyze the past correction history and select the most appropriate correction method. Appropriate correction methods include, but are not limited to, accuracy and speed of correction. By referring to the past correction history, the accuracy of corrections can be improved. Some or all of the above-described processing by the correction unit can be performed using, for example, AI, or without AI. For example, the correction unit can input past correction history data into a generation AI and cause the generation AI to select a correction method.

[0044] The correction unit can apply different correction algorithms depending on the category of the posted text during correction. For example, the correction unit applies a correction algorithm suited to technical content to technical documents. Technical documents include, but are not limited to, technical details and terminology. The correction unit can also apply a correction algorithm that takes brand image into account to marketing documents. Marketing documents include, but are not limited to, brand messages and advertising expressions. The correction unit can also apply a correction algorithm that takes into account the risk of confidential information leaks to internal documents. Internal documents include, but are not limited to, internal reports and meeting records. This improves the accuracy of corrections by applying a correction algorithm suited to the category of the posted text. Some or all of the above-described processing by the correction unit can be performed using, or without, AI. For example, the correction unit can input category data of the posted text into a generation AI and cause the generation AI to apply a correction algorithm.

[0045] The correction unit may take into account the geographical distribution of the posted text when making corrections. The correction unit may, for example, take into account regional cultures and customs when making corrections. Regional cultures and customs include, but are not limited to, differences in language and expressions. The correction unit may also take into account regional laws and regulations when making corrections. Regional laws and regulations include, but are not limited to, data protection laws and privacy laws. The correction unit may also take into account regional differences in language and expressions when making corrections. Differences in language and expressions include, but are not limited to, dialects and regional expressions. This takes geographical distribution into account, thereby improving the accuracy of corrections. Some or all of the above-described processing by the correction unit may be performed using, or without, AI. For example, the correction unit may input geographical distribution data into a generation AI and cause the generation AI to perform corrections.

[0046] The correction unit can improve the accuracy of the correction by referring to literature related to the posted text during correction. For example, the correction unit can improve the accuracy of the correction by referring to past literature related to the posted text. Past literature includes, but is not limited to, related academic papers, technical reports, industry news, etc. The correction unit can also refer to literature related to the subject of the posted text to make appropriate corrections. Literature related to the subject includes, but is not limited to, research papers on the same topic and related technical documents. The correction unit can also refer to literature related to the context of the posted text to make corrections to avoid misunderstandings. Literature related to the context includes, but is not limited to, related background information, past cases, etc. In this way, referring to literature related to the posted text improves the accuracy of the correction. Some or all of the above-described processing by the correction unit can be performed using, for example, AI, or without AI. For example, the correction unit can input related literature data into the generation AI and cause the generation AI to perform literature reference and correction.

[0047] The warning unit can improve the accuracy of the warning by referring to past warning history when issuing a warning. The warning unit, for example, refers to a history of similar warnings issued in the past to improve the accuracy of the warning. Past warning history includes, for example, a warning log, a warning history database, etc., but is not limited to these examples. The warning unit can also learn and apply effective warning methods from past warning history. Effective warning methods include, for example, specific warning patterns, successful warning cases, etc., but are not limited to these examples. Furthermore, the warning unit can analyze past warning history and select the most appropriate warning method. Appropriate warning methods include, for example, warning accuracy and promptness, etc., but are not limited to these examples. By referring to past warning history, the accuracy of the warning can be improved. Some or all of the above-described processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input past warning history data to a generation AI and cause the generation AI to select a warning method.

[0048] When issuing a warning, the warning unit can apply different warning algorithms depending on the category of the posted message. For example, the warning unit can apply a warning algorithm that takes into account technical risks to technical documents. Examples of technical documents include, but are not limited to, technical details and terminology. The warning unit can also apply a warning algorithm that takes into account brand image to marketing documents. Examples of marketing documents include, but are not limited to, brand messages and advertising expressions. The warning unit can also apply a warning algorithm that takes into account the risk of confidential information leaks to internal documents. Examples of internal documents include, but are not limited to, internal reports and meeting records. This improves the accuracy of warnings by applying a warning algorithm according to the category of the posted message. Some or all of the above-described processing by the warning unit can be performed using, or without, AI. For example, the warning unit can input category data of the posted message into a generation AI and cause the generation AI to apply a warning algorithm.

[0049] When issuing a warning, the warning unit can adjust the importance of the warning based on the time the post was submitted. For example, the warning unit displays a high-importance warning for urgent posts. Urgent posts include, but are not limited to, important announcements and emergency notifications. The warning unit can also display a normal-importance warning for regularly posted posts. Periodic posts include, but are not limited to, regular reports and periodic updates. The warning unit can also display a low-importance warning for past posts. Past posts include, but are not limited to, archived documents and past reports. This adjusts the importance of the warning based on the time the post was submitted, thereby enhancing the effectiveness of the warning. Some or all of the above-described processing by the warning unit may be performed using, or without, AI. For example, the warning unit can input data on the time the post was submitted to the generation AI and cause the generation AI to adjust the importance.

[0050] When issuing a warning, the warning unit can improve the accuracy of the warning by referring to literature related to the posted text. For example, the warning unit can refer to past literature related to the posted text to improve the accuracy of the warning. Past literature includes, but is not limited to, related academic papers, technical reports, industry news, etc. The warning unit can also refer to literature related to the subject of the posted text to issue an appropriate warning. Literature related to the subject includes, but is not limited to, research papers on the same topic and related technical documents. The warning unit can also refer to literature related to the context of the posted text to issue a warning to avoid misunderstandings. Literature related to the context includes, but is not limited to, related background information, past cases, etc. In this way, referring to literature related to the posted text improves the accuracy of the warning. Some or all of the above-described processing by the warning unit can be performed using, for example, AI, or without AI. For example, the warning unit can input related literature data into the generation AI and cause the generation AI to perform literature reference and issue a warning.

[0051] When generating a report, the generation unit can select an optimal report generation method by referring to past report history. The generation unit, for example, refers to a history of similar reports generated in the past and selects the optimal report generation method. Past report history includes, but is not limited to, a report log, a report history database, etc. The generation unit can also learn and apply effective report generation methods from past report history. Effective report generation methods include, but are not limited to, specific report patterns, successful report cases, etc. The generation unit can also analyze past report history and select the most appropriate report generation method. Appropriate report generation methods include, but are not limited to, report accuracy, report speed, etc. By referring to past report history, the accuracy of the report is improved. Some or all of the above-described processing in the generation unit may be performed using, for example, AI, or without AI. For example, the generation unit can input past report history data into a generation AI and cause the generation AI to select a report generation method.

[0052] When generating a report, the generation unit can apply different report generation algorithms depending on the category of the posted message. For example, the generation unit applies a report generation algorithm suitable for technical content to technical documents. Technical documents include, but are not limited to, technical details and technical terminology. The generation unit can also apply a report generation algorithm that takes brand image into account to marketing documents. Marketing documents include, but are not limited to, brand messages and advertising expressions. The generation unit can also apply a report generation algorithm that takes into account the risk of confidential information leakage to internal documents. Internal documents include, but are not limited to, internal reports and meeting records. This improves the accuracy of the report by applying a report generation algorithm according to the category of the posted message. Some or all of the above-described processing by the generation unit can be performed using, for example, AI, or without AI. For example, the generation unit can input category data of the posted message into the generation AI and cause the generation AI to apply the report generation algorithm.

[0053] When generating a report, the generation unit can adjust the importance of the report based on the time of submission of the posted message. For example, the generation unit generates a report with a high importance level for urgent posts. Urgent posts include, but are not limited to, important announcements and emergency notifications. The generation unit can also generate a report with a normal importance level for periodic posts. Periodic posts include, but are not limited to, regular reports and periodic update information. The generation unit can also generate a report with a low importance level for past posts. Past posts include, but are not limited to, archived documents and past reports. This improves the effectiveness of the report by adjusting the importance level of the report based on the time of submission of the post. Some or all of the above-described processing by the generation unit may be performed using, for example, AI, or without AI. For example, the generation unit can input data on the time of submission of the post into the generation AI and cause the generation AI to adjust the importance level.

[0054] When generating a report, the generation unit can improve the accuracy of the report by referring to literature related to the posted text. For example, the generation unit can improve the accuracy of the report by referring to past literature related to the posted text. Past literature includes, but is not limited to, related academic papers, technical reports, industry news, etc. The generation unit can also generate an appropriate report by referring to literature related to the subject of the posted text. Literature related to the subject includes, but is not limited to, research papers on the same topic and related technical documents. Furthermore, the generation unit can generate a report to avoid misunderstandings by referring to literature related to the context of the posted text. Literature related to the context includes, but is not limited to, related background information, past cases, etc. In this way, referring to literature related to the posted text improves the accuracy of the report. Some or all of the above-described processing by the generation unit can be performed using, for example, AI, or without AI. For example, the generation unit can input related literature data into a generation AI and cause the generation AI to perform literature reference and report generation.

[0055] The system according to the embodiment is not limited to the above-described example, and various modifications are possible, for example, as follows.

[0056] The information leak and flame war risk reduction system can further include a behavior analysis unit that analyzes a user's behavioral history. The behavior analysis unit analyzes, for example, what types of posts a user has made in the past and what time of day they tend to post. This makes it possible to display a warning in advance if there is a high risk during a specific time period or if a specific user tends to make high-risk posts. The behavior analysis unit can also monitor changes in the frequency and content of users' posts and notify an administrator if abnormal behavior is detected. This makes it possible to predict risks based on users' behavior patterns and take appropriate measures.

[0057] The learning unit can also learn best practices for each industry. For example, protecting customer information is important in the financial industry, and protecting patient information is important in the medical industry. The learning unit can learn these industry-specific best practices and reflect them in security policies. This makes it possible to take effective measures against industry-specific risks. The learning unit can also learn industry-specific laws, regulations, and guidelines and optimize security policies based on these. This makes it possible to implement security measures that comply with laws and regulations.

[0058] The correction unit can further learn the user's posting style and provide the optimal correction method for each individual user. For example, if a user prefers a formal writing style, the correction unit can make corrections that suit that writing style. Also, for a user who prefers a casual writing style, the correction unit can suggest appropriate casual expressions. This makes it possible to make corrections that suit the user's posting style, thereby increasing user satisfaction. Furthermore, the correction unit can refer to the user's past correction history and automatically apply similar corrections. This makes it possible to achieve consistent corrections.

[0059] The generation unit can further include a report visualization function. For example, by visually displaying the contents of a report using graphs and charts, it can be made easier for the administrator to intuitively understand. It can also highlight important points in the report, allowing the administrator to quickly grasp important information. Furthermore, the generation unit can include a function for customizing the contents of the report. For example, the administrator can generate a report by focusing on specific items. This allows for flexible report generation that meets the needs of the administrator.

[0060] The checking unit can further improve the accuracy of the check by taking into account the context of the posted message. For example, it can analyze the context before and after the posted message and perform an appropriate check. It can also understand the subject of the posted message and perform a check that is appropriate to the context. This improves the accuracy of the check by taking into account the context of the posted message. Furthermore, the checking unit can also take into account the context of the posted message and perform a check to avoid misunderstandings. This can ensure that the content of the posted message is accurately conveyed.

[0061] The processing flow of the first embodiment will be briefly explained below.

[0062] Step 1: The learning module learns security policies, including information leakage prevention, access control, and data protection. The learning module learns lists of confidential information, lists of prohibited information, and lists of expressions that pose a high risk of causing an uproar. Step 2: The checker checks the posted text based on the security policy learned by the trainer. The checker checks the posted text using a natural language processing algorithm, which includes morphological analysis, grammatical analysis, and semantic analysis. Step 3: The Editing Department automatically edits the posts checked by the Checking Department. If confidential information is included, the Editing Department deletes it, and if it contains language that poses a high risk of causing an uproar, it replaces it with appropriate language. Appropriate language includes positive and neutral language. Step 4: The warning unit displays a warning based on the confidential information or risk of a controversy detected by the checking unit. The warning unit displays a warning in a pop-up. The pop-up contains a warning message, a warning notice, etc. Step 5: The generation unit generates a report based on the results of the checking unit and the correction unit. The generation unit generates a report including the content of the post, the check results, the corrections, and the warnings, and sends it to the administrator by email.

[0063] (Example 2) An information leak and flame war risk reduction system according to an embodiment of the present invention uses a generation AI to automatically check employee posts and reduce the risk of information leaks and flame wars. In this system, the generation AI automatically learns the company's security policy, pre-checks the posts when employees create them, and automatically corrects them as necessary. If the posts contain confidential information or contain expressions that pose a high risk of flame wars, a warning is displayed, and a report is generated based on the results of the post check and provided to the administrator. This enables accurate and appropriate decisions to be made at all times, reducing the risk of information leaks and flame wars. Furthermore, when combined with employee training, more reliable security measures are realized. For example, when the generation AI automatically learns the company's security policy, it analyzes the contents of the security policy in detail and understands its contents. For example, it learns how to handle confidential information and lists information that must not be made public. Next, when an employee creates a post, the generation AI pre-checks the post. The generation AI analyzes the content of the post to ensure it does not violate the security policy. For example, if the post contains confidential information or contains expressions that pose a high risk of flame wars, a warning is displayed. Furthermore, the generation AI automatically corrects posts as needed. For example, if confidential information is included, it deletes that portion or replaces it with appropriate wording. This reduces the risk of employees unintentionally disclosing confidential information. The generation AI also generates a report based on the results of the post check. This report includes the content of the post, the check results, and any corrections. The report is provided to administrators, allowing them to understand the status of employees' posts. This enables them to make accurate and appropriate decisions at all times, reducing the risk of information leaks and online outrages. Furthermore, when combined with employee training, more reliable security measures can be achieved. For example, educating employees about the importance of security policies and providing feedback on the generation AI's check results can raise employee awareness. This allows the information leak and online outrage risk reduction system to automatically check employees' posts and reduce the risk of information leaks and online outrages.

[0064] An information leakage and flame war risk reduction system according to an embodiment includes a learning unit, a checking unit, a correcting unit, a warning unit, and a generating unit. The learning unit learns a security policy. Examples of security policies include, but are not limited to, information leakage prevention, access control, and data protection. The learning unit learns, for example, a list of confidential information, a list of prohibited information, and a list of expressions that pose a high risk of flame wars. The checking unit checks posted messages based on the security policy learned by the learning unit. The checking unit checks posted messages using, for example, a natural language processing algorithm. Examples of natural language processing algorithms include, but are not limited to, morphological analysis, grammatical analysis, and semantic analysis. The correcting unit automatically corrects posted messages checked by the checking unit. For example, the correcting unit deletes messages that contain confidential information and replaces them with appropriate expressions that pose a high risk of flame wars. Examples of appropriate expressions include, but are not limited to, positive expressions and neutral expressions. The warning unit displays a warning based on the confidential information and flame war risks detected by the checking unit. The warning unit displays a warning, for example, in the form of a pop-up. Examples of pop-ups include, but are not limited to, warning messages and alerts. The generation unit generates a report based on the results of the check unit and the correction unit. The generation unit generates a report including, for example, the content of the posted message, the check results, the corrections, and the warning, and provides the report to the administrator by email. In this way, the information leak and flame war risk reduction system according to the embodiment can automatically check the messages posted by employees and reduce the risk of information leaks and flame wars.

[0065] The learning unit can learn a list of confidential information, a list of information that should not be disclosed, or a list of expressions that have a high risk of causing an uproar. The learning unit, for example, learns a list of confidential information. Examples of confidential information include, but are not limited to, personal information, trade secrets, and financial information. The learning unit can also learn a list of information that should not be disclosed. Examples of information that should not be disclosed include, but are not limited to, legal regulations and corporate policies. The learning unit can also learn a list of expressions that have a high risk of causing an uproar. Examples of expressions that have a high risk of causing an uproar include, but are not limited to, past cases of uproars and sentiment analysis results. By learning a specific list, the accuracy of the security policy is improved. Some or all of the above-described processing in the learning unit may be performed using, for example, AI, or may be performed without using AI. For example, the learning unit can input a list of confidential information into a generation AI and have it learn the contents of the list.

[0066] The check unit can check the posted text using a natural language processing algorithm. The check unit checks the posted text using, for example, morphological analysis. Morphological analysis is a technique that divides a sentence into words and analyzes the part of speech of each word. The check unit can also check the posted text using grammatical analysis. Grammatical analysis is a technique that analyzes the grammatical structure of a sentence and detects grammatical errors. The check unit can also check the posted text using semantic analysis. Semantic analysis is a technique that analyzes the meaning of a sentence and determines the appropriate meaning based on the context. Thus, using a natural language processing algorithm improves the accuracy of checking the posted text. Some or all of the above-mentioned processing in the check unit may be performed using, for example, AI, or may be performed without using AI. For example, the check unit can input the posted text to a generation AI and have the generation AI perform natural language processing.

[0067] The correction unit can delete confidential information if it contains confidential information, and replace expressions that pose a risk of causing an uproar with appropriate expressions if it contains confidential information. For example, the correction unit can delete confidential information if it contains confidential information. Confidential information includes, but is not limited to, personal information, trade secrets, and financial information. The correction unit can also replace expressions that pose a high risk of causing an uproar with appropriate expressions if it contains expressions that pose a high risk of causing an uproar. Appropriate expressions include, but are not limited to, positive expressions and neutral expressions. This reduces the risk by automatically correcting confidential information and expressions that pose a high risk of causing an uproar. Some or all of the above-described processing by the correction unit can be performed using, for example, AI, or without AI. For example, the correction unit can input the posted text into a generation AI and have the generation AI perform automatic corrections.

[0068] The warning unit can display a warning in a pop-up. For example, the warning unit displays a warning in a pop-up if the posted text contains confidential information. The pop-up displays a warning message such as, "This posted text contains confidential information. Do not make it public." The warning unit can also display a warning in a pop-up if the posted text contains a high risk of causing an uproar. The pop-up displays a warning message such as, "This posted text contains a high risk of causing an uproar. Please correct it." By displaying a warning in a pop-up, the user can be immediately notified of the risk. Some or all of the above-described processing in the warning unit may be performed using, or without, AI, for example. For example, the warning unit can input the posted text into a generation AI and cause the generation AI to display a warning.

[0069] The generation unit can generate a report including the content of the posted message, the check results, the corrections, and the warnings, and provide it to the administrator by email. The generation unit generates a report including, for example, the content of the posted message, the check results, the corrections, and the warnings. The report includes, for example, the full text of the posted message, details of the check results, a list of corrections, and details of the warnings. The generation unit can also provide the generated report to the administrator by email. The email includes, for example, an overview of the report and important points. By generating a report and providing it to the administrator, the status of the posted message can be understood. Some or all of the above-described processing by the generation unit may be performed using, for example, AI, or may be performed without using AI. For example, the generation unit can input the content of the posted message into a generation AI and cause the generation AI to generate a report.

[0070] The learning unit can estimate the user's emotions and select learning data based on the estimated user emotions. For example, when the user is feeling stressed, the learning unit prioritizes the selection of learning data with content that promotes relaxation. Examples of learning data with content that promotes relaxation include, but are not limited to, relaxation techniques and stress management methods. Furthermore, when the user is relaxed, the learning unit can also select learning data with detailed and complex content. Examples of learning data with detailed and complex content include, but are not limited to, advanced technical explanations and specialized knowledge. Furthermore, when the user is concentrating, the learning unit can prioritize the selection of learning data with high importance. Examples of high-importance learning data include, but are not limited to, skills directly related to work and information about important projects. This enhances the effectiveness of learning by selecting learning data based on the user's emotions. Some or all of the above-described processing in the learning unit may be performed using, or without, AI. For example, the learning unit can input the user's emotion data into a generation AI and cause the generation AI to estimate emotions and select learning data.

[0071] During learning, the learning unit can optimize the learning algorithm by referring to past security incidents. For example, the learning unit refers to past information leak cases and reflects similar patterns in the learning algorithm. Examples of information leak cases include, but are not limited to, unauthorized access to databases and erroneous email transmissions. The learning unit can also refer to past flaming cases and add high-risk expressions to the learning algorithm. Examples of flaming cases include, but are not limited to, inappropriate comments on social media and misleading expressions in blog posts. Furthermore, the learning unit can analyze the causes of past security incidents and improve the learning algorithm. Examples of the causes of security incidents include, but are not limited to, system vulnerabilities and user errors. By referring to past security incidents, the accuracy of the learning algorithm can be improved. Some or all of the above-described processing in the learning unit may be performed using, or without, AI. For example, the learning unit can input past security incident data into the generation AI and cause the generation AI to optimize the algorithm.

[0072] The learning unit can take into account differences in security policies between industries during learning. For example, the learning unit can learn security policies for the financial industry to consider risks specific to the financial industry. Security policies for the financial industry include, but are not limited to, protection of customer information and encryption of transaction data. The learning unit can also learn security policies for the medical industry to consider risks related to the handling of medical data. Security policies for the medical industry include, but are not limited to, protection of patient information and security of electronic medical records. The learning unit can also learn security policies for the IT industry to consider technical risks. Security policies for the IT industry include, but are not limited to, system vulnerability management and data backup. This improves the accuracy of learning by taking into account security policies for each industry. Some or all of the above-described processing in the learning unit can be performed using, for example, AI, or without AI. For example, the learning unit can input industry-specific security policy data into a generation AI and cause the generation AI to perform learning.

[0073] The learning unit can estimate the user's emotions and adjust the frequency of learning based on the estimated user emotions. For example, if the user is feeling stressed, the learning unit can reduce the frequency of learning to alleviate the burden. Examples of reducing the frequency of learning include, but are not limited to, increasing the interval between learning sessions or simplifying the learning content. Furthermore, if the user is relaxed, the learning unit can increase the frequency of learning to efficiently progress with learning. Examples of increasing the frequency of learning include, but are not limited to, shortening the interval between learning sessions or providing more detailed learning content. Furthermore, if the user is concentrating, the learning unit can optimize the frequency of learning to effectively study. Examples of optimizing the frequency of learning include, but are not limited to, adjusting the timing of learning sessions according to the user's level of concentration or appropriately allocating the learning content. This adjusts the frequency of learning based on the user's emotions, thereby improving the effectiveness of learning. Some or all of the above-described processing in the learning unit may be performed using, or without, AI. For example, the learning unit can input the user's emotion data into the generation AI and cause the generation AI to adjust the learning frequency.

[0074] During learning, the learning unit can customize the learning content according to the company's growth stage. For example, the learning unit provides learning content on basic security policies for a start-up company. Basic security policies include, but are not limited to, password management and basic data protection methods. The learning unit can also provide learning content on detailed and advanced security policies for a mature company. Detailed and advanced security policies include, but are not limited to, the application of encryption technology and security incident response methods. Furthermore, the learning unit can provide learning content focused on risk management for a company in the expansion phase. Learning content focused on risk management includes, but is not limited to, risk assessment methods and risk mitigation measures. This enhances the effectiveness of learning by providing learning content tailored to the company's growth stage. Some or all of the above-described processing in the learning unit may be performed using, or without, AI. For example, the learning unit can input company growth stage data into the generation AI and cause the generation AI to customize the learning content.

[0075] The learning unit can take into account regional laws and regulations during learning. For example, the learning unit can learn security policies based on the European General Data Protection Regulation (GDPR). GDPR includes, but is not limited to, data protection principles and data subject rights. The learning unit can also learn medical data handling based on the American HIPAA. HIPAA includes, but is not limited to, protection of medical information and patient privacy rights. The learning unit can also learn data management based on the Japanese Act on the Protection of Personal Information. The Act includes, but is not limited to, proper handling of personal information and safety management measures for personal information. This improves the accuracy of learning by taking regional laws and regulations into consideration. Some or all of the above-described processing in the learning unit can be performed using AI, for example, or without AI. For example, the learning unit can input regional legal and regulatory data into the generation AI and cause the generation AI to perform learning.

[0076] The check unit can estimate the user's emotions and adjust the strictness of the check based on the estimated user emotions. For example, the check unit can ease the strictness of the check when the user is stressed. Examples of easing the strictness of the check include, but are not limited to, reducing the number of check items or loosening the check criteria. The check unit can also tighten the strictness of the check when the user is relaxed. Examples of tightening the strictness of the check include, but are not limited to, increasing the number of check items or tightening the check criteria. Furthermore, the check unit can optimize the strictness of the check when the user is concentrating. Examples of optimizing the strictness of the check include, but are not limited to, adjusting the check items according to the user's level of concentration or appropriately setting the check criteria. This improves the effectiveness of the check by adjusting the strictness of the check based on the user's emotions. Some or all of the above-described processing by the check unit can be performed using, for example, AI, or without AI. For example, the check unit can input the user's emotion data into the generation AI and have the generation AI adjust the strictness of the check.

[0077] The checking unit can improve the accuracy of the check by taking into account the context of the posted text. For example, the checking unit analyzes the context before and after the posted text and performs an appropriate check. The context before and after the posted text includes, but is not limited to, the preceding and following sentences, related topics, etc. The checking unit can also understand the subject of the posted text and perform a check based on the context. The subject includes, but is not limited to, the theme of the posted text, the main points of discussion, etc. Furthermore, the checking unit can also check the posted text to avoid misunderstandings by taking into account the context of the posted text. Checks to avoid misunderstandings include, but are not limited to, correcting ambiguous expressions and expressions that do not fit the context. This improves the accuracy of the check by taking into account the context of the posted text. Some or all of the above-described processing by the checking unit may be performed using, for example, AI, or without AI. For example, the checking unit can input context data of the posted text into a generation AI and have the generation AI perform context analysis and check.

[0078] The checking unit can apply different checking algorithms depending on the category of the posted text during the check. For example, the checking unit can apply a checking algorithm that takes into account technical risks to technical documents. Examples of technical documents include, but are not limited to, technical details and terminology. The checking unit can also apply a checking algorithm that takes into account brand image to marketing documents. Examples of marketing documents include, but are not limited to, brand messages and advertising expressions. The checking unit can also apply a checking algorithm that takes into account the risk of confidential information leaks to internal documents. Examples of internal documents include, but are not limited to, internal reports and meeting records. This improves the accuracy of the check by applying a checking algorithm according to the category of the posted text. Some or all of the above-described processing by the checking unit can be performed using, for example, AI, or without AI. For example, the checking unit can input category data of the posted text into a generation AI and cause the generation AI to apply a checking algorithm.

[0079] The check unit can estimate the user's emotions and adjust the display method of the check results based on the estimated user emotions. For example, if the user is feeling stressed, the check unit can provide a concise and easy-to-understand display method. Examples of concise and easy-to-understand display methods include, but are not limited to, displaying only the main points or using visually easy-to-understand graphics. Furthermore, if the user is relaxed, the check unit can display detailed check results. Examples of detailed check results include, but are not limited to, detailed explanations of each check item and suggestions for correction. Furthermore, if the user is concentrating, the check unit can provide a display method that emphasizes important points. Examples of display methods that emphasize important points include, but are not limited to, highlighting important check items and highlighting important corrections. This adjusts the display method of the check results based on the user's emotions, thereby enhancing the effectiveness of the check. Some or all of the above-described processing in the check unit may be performed using AI, for example, or without AI. For example, the check unit can input the user's emotion data into a generation AI and cause the generation AI to adjust the display method.

[0080] The checking unit may determine the priority of checks based on the submission date of each post. For example, the checking unit may prioritize urgent posts. Urgent posts include, but are not limited to, important announcements and emergency notifications. The checking unit may also check regularly posted posts with normal priority. Regular posts include, but are not limited to, regular reports and periodic updates. The checking unit may also postpone checking past posts. Past posts include, but are not limited to, archived documents and past reports. This improves the effectiveness of checks by determining the priority of checks based on the submission date of each post. Some or all of the above-described processing by the checking unit may be performed using, or without, AI. For example, the checking unit may input data on the submission date of each post into the generation AI, causing the generation AI to determine the priority.

[0081] The checking unit can improve the accuracy of the check by referring to literature related to the posted text during the check. For example, the checking unit can improve the accuracy of the check by referring to past literature related to the posted text. Past literature includes, but is not limited to, related academic papers, technical reports, industry news, etc. The checking unit can also refer to literature related to the subject of the posted text to perform an appropriate check. Literature related to the subject includes, but is not limited to, research papers on the same topic and related technical documents. The checking unit can also refer to literature related to the context of the posted text to check to avoid misunderstandings. Literature related to the context includes, but is not limited to, related background information, past cases, etc. In this way, referring to literature related to the posted text improves the accuracy of the check. Some or all of the above-described processing by the checking unit can be performed using, for example, AI, or without AI. For example, the checking unit can input related literature data into the generation AI and have the generation AI perform literature reference and check.

[0082] The correction unit can estimate the user's emotions and adjust the correction method based on the estimated user emotions. For example, if the user is stressed, the correction unit can provide a concise and easy-to-understand correction method. Examples of concise and easy-to-understand correction methods include, but are not limited to, correcting only the main points and using visually easy-to-understand graphics. Furthermore, if the user is relaxed, the correction unit can provide a detailed correction method. Examples of detailed correction methods include, but are not limited to, detailed explanations of each correction item and correction suggestions. Furthermore, if the user is concentrating, the correction unit can provide a correction method that emphasizes important points. Examples of correction methods that emphasize important points include, but are not limited to, highlighting important items to be corrected and emphasizing important correction points. This adjusts the correction method based on the user's emotions, thereby enhancing the effectiveness of the correction. Some or all of the above-described processing in the correction unit can be performed using AI, for example, or without AI. For example, the correction unit can input the user's emotion data into the generation AI and cause the generation AI to adjust the correction method.

[0083] When making a correction, the correction unit can select the optimal correction method by referring to the past correction history of the posted text. For example, the correction unit selects the optimal correction method by referring to the history of similar corrections made in the past. Past correction history includes, but is not limited to, a correction log, a correction history database, etc. The correction unit can also learn and apply effective correction methods from the past correction history. Effective correction methods include, but are not limited to, specific correction patterns, successful correction cases, etc. Furthermore, the correction unit can analyze the past correction history and select the most appropriate correction method. Appropriate correction methods include, but are not limited to, accuracy and speed of correction. By referring to the past correction history, the accuracy of corrections can be improved. Some or all of the above-described processing by the correction unit can be performed using, for example, AI, or without AI. For example, the correction unit can input past correction history data into a generation AI and cause the generation AI to select a correction method.

[0084] The correction unit can apply different correction algorithms depending on the category of the posted text during correction. For example, the correction unit applies a correction algorithm suited to technical content to technical documents. Technical documents include, but are not limited to, technical details and terminology. The correction unit can also apply a correction algorithm that takes brand image into account to marketing documents. Marketing documents include, but are not limited to, brand messages and advertising expressions. The correction unit can also apply a correction algorithm that takes into account the risk of confidential information leaks to internal documents. Internal documents include, but are not limited to, internal reports and meeting records. This improves the accuracy of corrections by applying a correction algorithm suited to the category of the posted text. Some or all of the above-described processing by the correction unit can be performed using, or without, AI. For example, the correction unit can input category data of the posted text into a generation AI and cause the generation AI to apply a correction algorithm.

[0085] The correction unit can estimate the user's emotions and determine the priority of corrections based on the estimated user emotions. For example, when the user is stressed, the correction unit can prioritize important corrections. Important corrections include, but are not limited to, deleting confidential information and correcting expressions that pose a high risk of causing controversy. Furthermore, when the user is relaxed, the correction unit can prioritize detailed corrections. Detailed corrections include, but are not limited to, improving sentence structure and optimizing expressions. Furthermore, when the user is focused, the correction unit can prioritize overall corrections. Overall corrections include, but are not limited to, reviewing the entire sentence and adjusting the overall tone. This improves the effectiveness of corrections by prioritizing corrections based on the user's emotions. Some or all of the above-described processing by the correction unit can be performed using, or without, AI. For example, the correction unit can input the user's emotion data into a generation AI and have the generation AI determine the priority of corrections.

[0086] The correction unit may take into account the geographical distribution of the posted text when making corrections. The correction unit may, for example, take into account regional cultures and customs when making corrections. Regional cultures and customs include, but are not limited to, differences in language and expressions. The correction unit may also take into account regional laws and regulations when making corrections. Regional laws and regulations include, but are not limited to, data protection laws and privacy laws. The correction unit may also take into account regional differences in language and expressions when making corrections. Differences in language and expressions include, but are not limited to, dialects and regional expressions. This takes geographical distribution into account, thereby improving the accuracy of corrections. Some or all of the above-described processing by the correction unit may be performed using, or without, AI. For example, the correction unit may input geographical distribution data into a generation AI and cause the generation AI to perform corrections.

[0087] The correction unit can improve the accuracy of the correction by referring to literature related to the posted text during correction. For example, the correction unit can improve the accuracy of the correction by referring to past literature related to the posted text. Past literature includes, but is not limited to, related academic papers, technical reports, industry news, etc. The correction unit can also refer to literature related to the subject of the posted text to make appropriate corrections. Literature related to the subject includes, but is not limited to, research papers on the same topic and related technical documents. The correction unit can also refer to literature related to the context of the posted text to make corrections to avoid misunderstandings. Literature related to the context includes, but is not limited to, related background information, past cases, etc. In this way, referring to literature related to the posted text improves the accuracy of the correction. Some or all of the above-described processing by the correction unit can be performed using, for example, AI, or without AI. For example, the correction unit can input related literature data into the generation AI and cause the generation AI to perform literature reference and correction.

[0088] The warning unit can estimate the user's emotions and adjust the warning display method based on the estimated user emotions. For example, if the user is feeling stressed, the warning unit can provide a concise and easy-to-understand warning display. Examples of concise and easy-to-understand warning displays include, but are not limited to, displaying only the main points or using visually easy-to-understand graphics. Furthermore, if the user is relaxed, the warning unit can provide a detailed warning display. Examples of detailed warning displays include, but are not limited to, detailed explanations of each warning item and warning suggestions. Furthermore, if the user is concentrating, the warning unit can provide a warning display that emphasizes important points. Examples of warning displays that emphasize important points include, but are not limited to, highlighting important warning items or important warning points. This adjusts the warning display method based on the user's emotions, thereby enhancing the effectiveness of the warning. Some or all of the above-described processing in the warning unit can be performed using, or without, AI. For example, the warning unit can input the user's emotion data into a generation AI and cause the generation AI to adjust the display method.

[0089] The warning unit can improve the accuracy of the warning by referring to past warning history when issuing a warning. The warning unit, for example, refers to a history of similar warnings issued in the past to improve the accuracy of the warning. Past warning history includes, for example, a warning log, a warning history database, etc., but is not limited to these examples. The warning unit can also learn and apply effective warning methods from past warning history. Effective warning methods include, for example, specific warning patterns, successful warning cases, etc., but are not limited to these examples. Furthermore, the warning unit can analyze past warning history and select the most appropriate warning method. Appropriate warning methods include, for example, warning accuracy and promptness, etc., but are not limited to these examples. By referring to past warning history, the accuracy of the warning can be improved. Some or all of the above-described processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input past warning history data to a generation AI and cause the generation AI to select a warning method.

[0090] When issuing a warning, the warning unit can apply different warning algorithms depending on the category of the posted message. For example, the warning unit can apply a warning algorithm that takes into account technical risks to technical documents. Examples of technical documents include, but are not limited to, technical details and terminology. The warning unit can also apply a warning algorithm that takes into account brand image to marketing documents. Examples of marketing documents include, but are not limited to, brand messages and advertising expressions. The warning unit can also apply a warning algorithm that takes into account the risk of confidential information leaks to internal documents. Examples of internal documents include, but are not limited to, internal reports and meeting records. This improves the accuracy of warnings by applying a warning algorithm according to the category of the posted message. Some or all of the above-described processing by the warning unit can be performed using, or without, AI. For example, the warning unit can input category data of the posted message into a generation AI and cause the generation AI to apply a warning algorithm.

[0091] The warning unit can estimate the user's emotions and prioritize warnings based on the estimated user emotions. For example, when the user is stressed, the warning unit can prioritize displaying important warnings. Important warnings include, but are not limited to, expressions that pose a high risk of leaking confidential information or causing a controversy. Furthermore, when the user is relaxed, the warning unit can prioritize displaying detailed warnings. Detailed warnings include, but are not limited to, detailed explanations of each warning item and warning suggestions. Furthermore, when the user is concentrating, the warning unit can prioritize displaying overall warnings. Overall warnings include, but are not limited to, reviewing the entire post and adjusting the overall tone. This improves the effectiveness of warnings by prioritizing warnings based on the user's emotions. Some or all of the above-described processing in the warning unit can be performed using, or without, AI. For example, the warning unit can input user emotion data into a generation AI and have the generation AI determine the priority of warnings.

[0092] When issuing a warning, the warning unit can adjust the importance of the warning based on the time the post was submitted. For example, the warning unit displays a high-importance warning for urgent posts. Urgent posts include, but are not limited to, important announcements and emergency notifications. The warning unit can also display a normal-importance warning for regularly posted posts. Periodic posts include, but are not limited to, regular reports and periodic updates. The warning unit can also display a low-importance warning for past posts. Past posts include, but are not limited to, archived documents and past reports. This adjusts the importance of the warning based on the time the post was submitted, thereby enhancing the effectiveness of the warning. Some or all of the above-described processing by the warning unit may be performed using, or without, AI. For example, the warning unit can input data on the time the post was submitted to the generation AI and cause the generation AI to adjust the importance.

[0093] When issuing a warning, the warning unit can improve the accuracy of the warning by referring to literature related to the posted text. For example, the warning unit can refer to past literature related to the posted text to improve the accuracy of the warning. Past literature includes, but is not limited to, related academic papers, technical reports, industry news, etc. The warning unit can also refer to literature related to the subject of the posted text to issue an appropriate warning. Literature related to the subject includes, but is not limited to, research papers on the same topic and related technical documents. The warning unit can also refer to literature related to the context of the posted text to issue a warning to avoid misunderstandings. Literature related to the context includes, but is not limited to, related background information, past cases, etc. In this way, referring to literature related to the posted text improves the accuracy of the warning. Some or all of the above-described processing by the warning unit can be performed using, for example, AI, or without AI. For example, the warning unit can input related literature data into the generation AI and cause the generation AI to perform literature reference and issue a warning.

[0094] The generation unit can estimate the user's emotions and adjust the presentation style of the report based on the estimated user's emotions. For example, when the user is stressed, the generation unit can provide a concise and easy-to-understand report. Examples of concise and easy-to-understand reports include, but are not limited to, displaying only the main points or using visually easy-to-understand graphics. The generation unit can also provide a detailed report when the user is relaxed. Examples of detailed reports include, but are not limited to, detailed explanations of each report item and report suggestions. The generation unit can also provide a report that emphasizes important points when the user is focused. Examples of reports that emphasize important points include, but are not limited to, highlighting important report items and highlighting important report points. This adjusts the presentation style of the report based on the user's emotions, thereby enhancing the effectiveness of the report. Some or all of the above-described processing by the generation unit can be performed using, for example, AI, or without AI. For example, the generation unit can input the user's emotion data into the generation AI and cause the generation AI to adjust the presentation style.

[0095] When generating a report, the generation unit can select an optimal report generation method by referring to past report history. The generation unit, for example, refers to a history of similar reports generated in the past and selects the optimal report generation method. Past report history includes, but is not limited to, a report log, a report history database, etc. The generation unit can also learn and apply effective report generation methods from past report history. Effective report generation methods include, but are not limited to, specific report patterns, successful report cases, etc. The generation unit can also analyze past report history and select the most appropriate report generation method. Appropriate report generation methods include, but are not limited to, report accuracy, report speed, etc. By referring to past report history, the accuracy of the report is improved. Some or all of the above-described processing in the generation unit may be performed using, for example, AI, or without AI. For example, the generation unit can input past report history data into a generation AI and cause the generation AI to select a report generation method.

[0096] When generating a report, the generation unit can apply different report generation algorithms depending on the category of the posted message. For example, the generation unit applies a report generation algorithm suitable for technical content to technical documents. Technical documents include, but are not limited to, technical details and technical terminology. The generation unit can also apply a report generation algorithm that takes brand image into account to marketing documents. Marketing documents include, but are not limited to, brand messages and advertising expressions. The generation unit can also apply a report generation algorithm that takes into account the risk of confidential information leakage to internal documents. Internal documents include, but are not limited to, internal reports and meeting records. This improves the accuracy of the report by applying a report generation algorithm according to the category of the posted message. Some or all of the above-described processing by the generation unit can be performed using, for example, AI, or without AI. For example, the generation unit can input category data of the posted message into the generation AI and cause the generation AI to apply the report generation algorithm.

[0097] The generation unit can estimate the user's emotions and prioritize reports based on the estimated user emotions. For example, when the user is stressed, the generation unit prioritizes generating important reports. Important reports include, but are not limited to, expressions that pose a high risk of leaking confidential information or sparking controversy. Furthermore, when the user is relaxed, the generation unit can prioritize generating detailed reports. Detailed reports include, but are not limited to, detailed explanations of each report item and report suggestions. Furthermore, when the user is focused, the generation unit can prioritize generating overall reports. Overall reports include, but are not limited to, reviewing the entire post and adjusting the overall tone. This improves the effectiveness of reports by prioritizing reports based on the user's emotions. Some or all of the above-described processing by the generation unit may be performed using AI, for example, or without AI. For example, the generation unit may input the user's emotion data into the generation AI and cause the generation AI to determine the priority of reports.

[0098] When generating a report, the generation unit can adjust the importance of the report based on the time of submission of the posted message. For example, the generation unit generates a report with a high importance level for urgent posts. Urgent posts include, but are not limited to, important announcements and emergency notifications. The generation unit can also generate a report with a normal importance level for periodic posts. Periodic posts include, but are not limited to, regular reports and periodic update information. The generation unit can also generate a report with a low importance level for past posts. Past posts include, but are not limited to, archived documents and past reports. This improves the effectiveness of the report by adjusting the importance level of the report based on the time of submission of the post. Some or all of the above-described processing by the generation unit may be performed using, for example, AI, or without AI. For example, the generation unit can input data on the time of submission of the post into the generation AI and cause the generation AI to adjust the importance level.

[0099] When generating a report, the generation unit can improve the accuracy of the report by referring to literature related to the posted text. For example, the generation unit can improve the accuracy of the report by referring to past literature related to the posted text. Past literature includes, but is not limited to, related academic papers, technical reports, industry news, etc. The generation unit can also generate an appropriate report by referring to literature related to the subject of the posted text. Literature related to the subject includes, but is not limited to, research papers on the same topic and related technical documents. Furthermore, the generation unit can generate a report to avoid misunderstandings by referring to literature related to the context of the posted text. Literature related to the context includes, but is not limited to, related background information, past cases, etc. In this way, referring to literature related to the posted text improves the accuracy of the report. Some or all of the above-described processing by the generation unit can be performed using, for example, AI, or without AI. For example, the generation unit can input related literature data into a generation AI and cause the generation AI to perform literature reference and report generation. === Hard Collateral 1-1 === Each of the multiple elements including the learning unit, checking unit, correction unit, warning unit, and generating unit described above is realized, for example, by at least one of the smart device 14 and the data processing device 12. For example, the learning unit is realized by the specific processing unit 290 of the data processing device 12 and learns a security policy. The checking unit is realized, for example, by the control unit 46A of the smart device 14 and checks posted text. The correction unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and automatically corrects posted text. The warning unit is realized, for example, by the control unit 46A of the smart device 14 and displays a warning. The generating unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and generates a report. === Hard Collateral 1-2 === Each of the multiple elements including the learning unit, checking unit, correction unit, warning unit, and generating unit described above is realized, for example, by at least one of the smart glasses 214 and the data processing device 12. For example, the learning unit is realized by the specific processing unit 290 of the data processing device 12 and learns a security policy. The checking unit is realized, for example, by the control unit 46A of the smart glasses 214 and checks posted text. The correction unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and automatically corrects posted text. The warning unit is realized, for example, by the control unit 46A of the smart glasses 214 and displays a warning. The generating unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and generates a report. === Hard Collateral 1-3 === Each of the multiple elements including the learning unit, checking unit, correction unit, warning unit, and generating unit described above is realized, for example, by at least one of the headset type terminal 314 and the data processing device 12. For example, the learning unit is realized by the specific processing unit 290 of the data processing device 12 and learns a security policy. The checking unit is realized, for example, by the control unit 46A of the headset type terminal 314 and checks posted text. The correction unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and automatically corrects posted text. The warning unit is realized, for example, by the control unit 46A of the headset type terminal 314 and displays a warning. The generating unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and generates a report. === Hard Collateral 1-4 === Each of the multiple elements including the learning unit, checking unit, correction unit, warning unit, and generating unit described above is realized, for example, by at least one of the robot 414 and the data processing device 12. For example, the learning unit is realized by the specific processing unit 290 of the data processing device 12 and learns a security policy. The checking unit is realized, for example, by the control unit 46A of the robot 414 and checks posted text. The correction unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and automatically corrects posted text. The warning unit is realized, for example, by the control unit 46A of the robot 414 and displays a warning. The generating unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and generates a report.

[0100] The system according to the embodiment is not limited to the above-described example, and various modifications are possible, for example, as follows.

[0101] The information leak and flame war risk reduction system can further include a behavior analysis unit that analyzes a user's behavioral history. The behavior analysis unit analyzes, for example, what types of posts a user has made in the past and what time of day they tend to post. This makes it possible to display a warning in advance if there is a high risk during a specific time period or if a specific user tends to make high-risk posts. The behavior analysis unit can also monitor changes in the frequency and content of users' posts and notify an administrator if abnormal behavior is detected. This makes it possible to predict risks based on users' behavior patterns and take appropriate measures.

[0102] The learning unit can also learn best practices for each industry. For example, protecting customer information is important in the financial industry, and protecting patient information is important in the medical industry. The learning unit can learn these industry-specific best practices and reflect them in security policies. This makes it possible to take effective measures against industry-specific risks. The learning unit can also learn industry-specific laws, regulations, and guidelines and optimize security policies based on these. This makes it possible to implement security measures that comply with laws and regulations.

[0103] The checking unit can also perform sentiment analysis of posted text. For example, if a posted text contains negative sentiment, a warning can be displayed because there is a high risk of a controversy. Furthermore, for posted text containing positive sentiment, the checking can be relaxed if there is no particular problem. This allows the strictness of the checking to be adjusted based on the sentiment of the posted text, enabling efficient risk management. Furthermore, appropriate feedback can be provided to the user based on the results of the sentiment analysis. For example, if there are a lot of negative sentiments, advice on stress management can be provided.

[0104] The correction unit can further learn the user's posting style and provide the optimal correction method for each individual user. For example, if a user prefers a formal writing style, the correction unit can make corrections that suit that writing style. Also, for a user who prefers a casual writing style, the correction unit can suggest appropriate casual expressions. This makes it possible to make corrections that suit the user's posting style, thereby increasing user satisfaction. Furthermore, the correction unit can refer to the user's past correction history and automatically apply similar corrections. This makes it possible to achieve consistent corrections.

[0105] The warning unit can further estimate the user's emotions and adjust the warning display method based on the estimated user's emotions. For example, if the user is feeling stressed, a concise and easy-to-understand warning message can be displayed. On the other hand, if the user is relaxed, a detailed warning message can be displayed. This makes it possible to provide an appropriate warning according to the user's emotions. Furthermore, if the user is concentrating, a warning message that emphasizes important points can be displayed. This maximizes the effectiveness of the warning.

[0106] The generation unit can further include a report visualization function. For example, by visually displaying the contents of a report using graphs and charts, it can be made easier for the administrator to intuitively understand. It can also highlight important points in the report, allowing the administrator to quickly grasp important information. Furthermore, the generation unit can include a function for customizing the contents of the report. For example, the administrator can generate a report by focusing on specific items. This allows for flexible report generation that meets the needs of the administrator.

[0107] The learning unit can further estimate the user's emotions and select learning data based on the estimated user's emotions. For example, if the user is feeling stressed, learning data with relaxing content can be preferentially selected. Also, if the user is relaxed, learning data with detailed and complex content can be selected. In this way, by selecting learning data based on the user's emotions, the effectiveness of learning can be improved. Furthermore, if the user is concentrating, learning data with high importance can be preferentially selected. This enables efficient learning.

[0108] The checking unit can further improve the accuracy of the check by taking into account the context of the posted message. For example, it can analyze the context before and after the posted message and perform an appropriate check. It can also understand the subject of the posted message and perform a check that is appropriate to the context. This improves the accuracy of the check by taking into account the context of the posted message. Furthermore, the checking unit can also take into account the context of the posted message and perform a check to avoid misunderstandings. This can ensure that the content of the posted message is accurately conveyed.

[0109] The correction unit can further estimate the user's emotions and adjust the correction method based on the estimated user's emotions. For example, if the user is feeling stressed, a simple and easy-to-understand correction method can be provided. On the other hand, if the user is relaxed, a detailed correction method can be provided. This makes it possible to improve the effectiveness of correction by adjusting the correction method based on the user's emotions. Furthermore, if the user is concentrating, a correction method that emphasizes important points can be provided. This improves the accuracy of correction.

[0110] The generation unit can further estimate the user's emotions and adjust the presentation method of the report based on the estimated user's emotions. For example, if the user is feeling stressed, a concise and easy-to-understand report can be provided. On the other hand, if the user is relaxed, a detailed report can be provided. In this way, the effectiveness of the report can be increased by adjusting the presentation method of the report based on the user's emotions. Furthermore, if the user is concentrating, a report that emphasizes important points can be provided. This makes it possible to accurately convey the content of the report.

[0111] The processing flow of the second embodiment will be briefly explained below.

[0112] Step 1: The learning module learns security policies, including information leakage prevention, access control, and data protection. The learning module learns lists of confidential information, lists of prohibited information, and lists of expressions that pose a high risk of causing an uproar. Step 2: The checker checks the posted text based on the security policy learned by the trainer. The checker checks the posted text using a natural language processing algorithm, which includes morphological analysis, grammatical analysis, and semantic analysis. Step 3: The Editing Department automatically edits the posts checked by the Checking Department. If confidential information is included, the Editing Department deletes it, and if it contains language that poses a high risk of causing an uproar, it replaces it with appropriate language. Appropriate language includes positive and neutral language. Step 4: The warning unit displays a warning based on the confidential information or risk of a controversy detected by the checking unit. The warning unit displays a warning in a pop-up. The pop-up contains a warning message, a warning notice, etc. Step 5: The generation unit generates a report based on the results of the checking unit and the correction unit. The generation unit generates a report including the content of the post, the check results, the corrections, and the warnings, and sends it to the administrator by email.

[0113] 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.

[0114] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> Examples of the generative AI include a neural network (NN) and a neural network (NN). The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 receives a prompt containing an instruction, as well as inference data such as voice data representing speech, text data representing text, and image data representing an image (e.g., still image data or video data). 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 one or more data formats of voice data, text data, image data, etc. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specification processing unit 290 performs the above-mentioned specification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation model 58 includes AIs other than the generative AI. The AI ​​other than the generative AI may be, for example, linear regression, logistic regression, decision tree, random forest, support vector machine (SVM), k-means clustering, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), or naive Bayes, and may perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processing of each of the above-mentioned parts is performed by an AI, the processing may be performed in part or entirely by the AI, but is not limited to these examples. The processing performed by an AI including the generative AI may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by an AI including the generative AI.

[0115] Furthermore, the processing by the data processing system 10 described above is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart device 14, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart device 14. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the smart device 14 or an external device, and the smart device 14 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0116] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0117] [Second embodiment] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0118] 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0119] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, RAM 30, and 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 and / or a LAN.

[0120] 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.

[0121] The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to instructions from the processor 46.

[0122] 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 user's surroundings (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0123] 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.

[0124] 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.

[0125] 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.

[0126] 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. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0127] In the smart glasses 214, the specific processing is performed by the processor 46. A specific processing program 60 is stored in the storage 50. The processor 46 reads the specific processing program 60 from the storage 50 and executes the read specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as the control unit 46A in accordance with the specific processing program 60 executed on the RAM 48. The smart glasses 214 also have a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can perform processing similar to that of the specific processing unit 290 using these models.

[0128] Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 communicates with the server device having the data generation model 58 to obtain a processing result (such as a prediction result) using the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device (for example, a mobile phone, a robot, a home appliance, etc.) owned by a user.

[0129] 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.

[0130] The data generation model 58 is a so-called generative AI. An example of the data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 receives a prompt including an instruction, as well as inference data such as audio data indicating speech, text data indicating text, and image data indicating an image (e.g., still image data or video data). 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 one or more data formats, such as audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The identification processing unit 290 performs the above-mentioned identification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation models 58 include AIs other than the generation AI. Examples of AIs other than the generation AI include, but are not limited to, linear regression, logistic regression, decision trees, random forests, support vector machines (SVMs), k-means clustering, convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial networks (GANs), and naive Bayes. These AIs can perform various types of processing, but are not limited to these examples. The AI ​​may also be an AI agent. When the processing of each of the above-described parts is performed by an AI, the processing may be performed in part or entirely by the AI, but is not limited to these examples. Processing performed by an AI, including the generation AI, may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by an AI, including the generation AI.

[0131] The data processing system 210 according to the second embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 210 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart glasses 214, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart glasses 214. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the smart glasses 214 or an external device, etc., and the smart glasses 214 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

[0132] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0133] [Third embodiment] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0134] 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.

[0135] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, RAM 30, and 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 and / or a LAN.

[0136] 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.

[0137] The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to instructions from the processor 46.

[0138] 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 user's surroundings (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0139] 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.

[0140] 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.

[0141] 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.

[0142] 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. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0143] In the headset type terminal 314, the identification process is performed by the processor 46. A identification program 60 is stored in the storage 50. The processor 46 reads the identification program 60 from the storage 50 and executes the read identification program 60 on the RAM 48. The identification process is realized by the processor 46 operating as a control unit 46A in accordance with the identification program 60 executed on the RAM 48. Note that the headset type terminal 314 has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can also perform processing similar to that of the identification processing unit 290 using these models.

[0144] Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 communicates with the server device having the data generation model 58 to obtain a processing result (such as a prediction result) using the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device (for example, a mobile phone, a robot, a home appliance, etc.) owned by a user.

[0145] 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.

[0146] The data generation model 58 is a so-called generative AI. An example of the data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 receives a prompt including an instruction, as well as inference data such as audio data indicating speech, text data indicating text, and image data indicating an image (e.g., still image data or video data). 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 one or more data formats, such as audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The identification processing unit 290 performs the above-mentioned identification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation models 58 include AIs other than the generation AI. Examples of AIs other than the generation AI include, but are not limited to, linear regression, logistic regression, decision trees, random forests, support vector machines (SVMs), k-means clustering, convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial networks (GANs), and naive Bayes. These AIs can perform various types of processing, but are not limited to these examples. The AI ​​may also be an AI agent. When the processing of each of the above-described parts is performed by an AI, the processing may be performed in part or entirely by the AI, but is not limited to these examples. Processing performed by an AI, including the generation AI, may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by an AI, including the generation AI.

[0147] The data processing system 310 according to the third embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 310 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the headset type terminal 314, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the headset type terminal 314. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the headset type terminal 314 or an external device, etc., and the headset type terminal 314 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

[0148] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0149] [Fourth embodiment] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0150] 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.

[0151] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, RAM 30, and 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 and / or a LAN.

[0152] 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.

[0153] The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to instructions from the processor 46.

[0154] 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 image sensor or a CCD image sensor, and captures images of the user's surroundings (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0155] 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.

[0156] The control object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the 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.

[0157] 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.

[0158] 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.

[0159] 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. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0160] In the robot 414, the processor 46 performs the identification process. The storage 50 stores the identification program 60. The processor 46 reads the identification program 60 from the storage 50 and executes the read identification program 60 on the RAM 48. The identification process is realized by the processor 46 operating as the control unit 46A in accordance with the identification program 60 executed on the RAM 48. The robot 414 also has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can perform the same process as the identification processing unit 290 using these models.

[0161] Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 communicates with the server device having the data generation model 58 to obtain a processing result (such as a prediction result) using the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device (for example, a mobile phone, a robot, a home appliance, etc.) owned by a user.

[0162] 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.

[0163] The data generation model 58 is a so-called generative AI. An example of the data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 receives a prompt including an instruction, as well as inference data such as audio data indicating speech, text data indicating text, and image data indicating an image (e.g., still image data or video data). 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 one or more data formats, such as audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The identification processing unit 290 performs the above-mentioned identification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation models 58 include AIs other than the generation AI. Examples of AIs other than the generation AI include, but are not limited to, linear regression, logistic regression, decision trees, random forests, support vector machines (SVMs), k-means clustering, convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial networks (GANs), and naive Bayes. These AIs can perform various types of processing, but are not limited to these examples. The AI ​​may also be an AI agent. When the processing of each of the above-described parts is performed by an AI, the processing may be performed in part or entirely by the AI, but is not limited to these examples. Processing performed by an AI, including the generation AI, may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by an AI, including the generation AI.

[0164] The data processing system 410 according to the fourth embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 410 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the robot 414, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the robot 414. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the robot 414 or an external device, etc., and the robot 414 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

[0165] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0166] 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.

[0167] FIG. 9 illustrates 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 behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion encompasses both emotions 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.

[0168] 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.

[0169] 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).

[0170] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, cars, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on speech emotion recognition and brain physiological signal analysis systems for emotions, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the area called "reaction," where sensation is dominant. The right half of the emotion map lists emotions belonging to the area called "situation," where situational awareness is dominant.

[0171] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[0172] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[0173] In the above embodiment, an example was given in which a specific process is performed by one computer 22, but the technology disclosed herein is not limited to this, and distributed processing of the specific process may be performed by multiple computers including computer 22.

[0174] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[0175] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[0176] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[0177] The hardware resource for executing a specific process can be any of the following types of processors: A processor, for example, is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. A processor also includes a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[0178] The hardware resource that executes the specific process may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific process may be a single processor.

[0179] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[0180] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[0181] In the above example, the first to fourth embodiments have been described separately, but some or all of these embodiments may be combined. The smart device 14, smart glasses 214, headset terminal 314, and robot 414 are merely examples, and they may be combined, or other devices may be used. In the above example, the first and second embodiments have been described separately, but they may be combined.

[0182] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[0183] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[0184] [Explanation of symbols]

[0185] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot

Claims

1. a learning unit that learns a security policy; a checking unit that checks posted text based on the security policy learned by the learning unit; a correction unit that automatically corrects the posted text checked by the checking unit; a warning unit that displays a warning based on the confidential information or the risk of a controversy detected by the checking unit; a generating unit that generates a report based on the results of the checking unit and the correcting unit; Equipped with A system characterized by:

2. The learning unit Learn lists of confidential or prohibited information, and lists of potentially inflammatory expressions 2. The system of claim 1.

3. The checking unit Use natural language processing algorithms to check posts 2. The system of claim 1.

4. The correction unit If confidential information is included, it will be deleted. If it contains expressions that could cause controversy, they will be replaced with appropriate expressions.

2. The system of claim 1.

5. The warning unit Display a pop-up warning 2. The system of claim 1.

6. The generation unit A report containing the content of the post, check results, corrections, and warnings is generated and sent to the administrator via email.

2. The system of claim 1.

7. The learning unit Estimate the user's emotions and select training data based on the estimated user emotions.

2. The system of claim 1.

8. The learning unit During learning, the learning algorithm is optimized by referencing past security incidents.

2. The system of claim 1.

Images