System

Abstract

To provide a system for properly managing a user's emotional state and assisting to improve communication skills.SOLUTION: Provided is a system including: means for acquiring a text-based message created by a user; means for recognizing the user's emotion from the message by using an emotion engine; and means for proposing improvements on the message by using a prompt sentence indicating proposition of improvements on the message and a generative AI model, on the basis of the recognized emotion of the user when the message includes an aggressive nuance.SELECTED DRAWING: Figure 1

Description

[Technical field]

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

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including a description and related instruction sentence regarding 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] JP 2022-180282 A Summary of the Invention [Problem to be solved by the invention]

[0004] In modern society, most communication is done through text. However, text communication can make it difficult to convey nuances of expression, and can give the impression of being aggressive or indifferent. This can cause problems in human relationships. [Means for solving the problem]

[0005] The present invention provides a system that detects aggressive or indifferent nuances in text-based communication and makes suggestions to improve the text to be more "kind". The system uses natural language processing technology to understand the meaning of the text and generate appropriate suggestions to improve it. Furthermore, the system aims to improve users' communication skills and enable them to live vibrantly in a place of mutual consideration and psychological safety. [Brief description of the drawings]

[0006] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Diagram 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. FIG. [Diagram 3] FIG. 11 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Diagram 5] FIG. 13 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. 13 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 13 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 2 is a sequence diagram showing a process flow of the data processing system in the first embodiment of the first form example. [Figure 12] FIG. 11 is a sequence diagram showing a process flow of the data processing system in the application example 1 of the first embodiment. [Figure 13] FIG. 11 is a sequence diagram showing a process flow of the data processing system in the second embodiment of the second embodiment. [Figure 14] FIG. 13 is a sequence diagram showing a process flow of the data processing system in application example 2 of embodiment 2. [Figure 15] FIG. 13 is a sequence diagram showing a process flow of a data processing system in a third embodiment of the present invention. [Figure 16] FIG. 13 is a sequence diagram showing a process flow of the data processing system in the third application example of the third embodiment. [Figure 17] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in the first embodiment of the first form example when an emotion engine is combined. [Figure 18] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Form Example 1 when combined with an emotion engine. [Figure 19] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system according to the embodiment 2 of the second embodiment in the case where an emotion engine is combined. [Figure 20] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Form Example 2 when combined with an emotion engine. [Figure 21] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in the embodiment 3 of the form example 3 in the case where an emotion engine is combined. [Figure 22] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Form Example 3 when combined with an emotion engine. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0007] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0008] First, the terms used in the following description will be explained.

[0009] In the following embodiments, a signed processor (hereinafter simply referred to as a "processor") may be one arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be one 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 (registered trademark)).

[0010] In the following embodiments, a signed RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0011] 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.

[0012] In the following embodiments, a communication I / F (Interface) with a code is an interface including a communication processor and an antenna. The communication I / F controls communication between multiple computers. An example of a communication standard applied to the communication I / F is a wireless communication standard including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0013] 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. In addition, in this specification, the same idea as "A and / or B" is also applied when three or more things are expressed by connecting them with "and / or."

[0014] [First embodiment]

[0015] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0016] 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.

[0017] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a wide area network (WAN) and / or a local area network (LAN).

[0018] 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.

[0019] 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 (e.g., 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 acquires the data indicating the user input.

[0020] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (e.g., voice and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs voice according to instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, an aperture, and a shutter, and an imaging element such as a Complementary Metal-Oxide-Semiconductor (CMOS) image sensor or a Charge Coupled Device (CCD) image sensor.

[0021] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54.

[0022] FIG. 2 shows an example of main functions of the data processing device 12 and the smart device 14.

[0023] As shown in Fig. 2, 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. The specific process program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific process program 56 from the storage 32, and executes the read specific process 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 process program 56 executed on the RAM 30.

[0024] 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.

[0025] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores a reception output program 60. The reception output program 60 is used together with the specific processing program 56 by the data processing system 10. The processor 46 reads out the reception output program 60 from the storage 50, and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0026] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0027] "Example 1"

[0028] In one embodiment of the present invention, a means for text-based communication may be email, a messenger app, etc. When a user communicates using these means, the system receives the text and uses natural language processing technology to understand the meaning of the text.

[0029] "Example 2"

[0030] To detect offensive or insensitive nuances, the system uses sentiment analysis technology. Specifically, it analyzes words, phrases, and contexts in a text to determine whether the text contains offensive or insensitive nuances.

[0031] "Example 3"

[0032] As a means of making suggestions to improve sentences to be more "compassionate," the system generates improvement suggestions based on the detected nuances. Specifically, the system makes suggestions to replace aggressive or indifferent expressions with "compassionate" expressions. For example, the system makes suggestions to improve the expression "You are wrong" to a more "compassionate" expression such as "I understand your point of view, but I think from a slightly different perspective."

[0033] The process flow of each embodiment will be described below.

[0034] "Example 1"

[0035] Step 1: The user engages in text-based communication via email, messenger app, etc.

[0036] Step 2: The system receives the text and uses natural language processing techniques to understand the meaning of the text.

[0037] "Example 2"

[0038] Step 1: The system receives the text.

[0039] Step 2: Using sentiment analysis techniques, analyze the words, phrases, and context in the text. Step 3: Determine whether the text contains offensive or insensitive nuances.

[0040] "Example 3"

[0041] Step 1: If offensive or insensitive nuances are detected, the system generates improvement suggestions.

[0042] Step 2: Make suggestions to replace aggressive or insensitive expressions with more "compassionate" ones. Step 3: Make suggestions to improve the expression, for example, "You are wrong" to "I understand your point of view, but I think from a slightly different perspective."

[0043] Example 1

[0044] Next, a description will be given of Example 1 of the embodiment 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal".

[0045] In conventional text-based communication systems, it was difficult to properly understand the nuances of messages sent by users, detect aggressive or indifferent expressions, and make suggestions for improvement. In addition, there was a lack of technology to generate appropriate responses to messages sent by users. This made it difficult to improve users' communication skills and realize considerate communication.

[0046] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0047] In this invention, the server includes a means for a user to send a message, a means for a terminal to send a message to the server, a means for the server to receive the message, a means for the server to preprocess the message, a means for the server to semantically analyze the message, a means for the server to generate a response, a means for the server to send the response to the terminal, and a means for the terminal to display the response to the user. This makes it possible to properly understand the nuances of messages sent by users, detect aggressive or indifferent expressions, and make suggestions for improvement. In addition, it is possible to generate appropriate responses to user messages, improve the user's communication skills, and realize considerate communication.

[0048] "Text-based communication" refers to the exchange of information using written text, including email and messenger apps.

[0049] "Aggressive or indifferent nuances" refer to expressions or attitudes that show hostility or indifference toward the other person.

[0050] "Thoughtful writing" refers to writing that takes into consideration the other person's feelings and situation and includes expressions that show consideration and kindness.

[0051] "Means for a user to send a message" refers to an interface or function that allows a user to input and send text.

[0052] "Means for a terminal to send a message to a server" refers to a function of a user's terminal to send a message to a server via the Internet.

[0053] "Means for the server to receive a message" refers to the function of the server to receive a message sent from a terminal.

[0054] "Means for the server to preprocess the message" refers to a function for performing processing to convert the received message into a form that is easy to analyze.

[0055] "Means for the server to semantically analyze the message" refers to the functionality for understanding the meaning of the pre-processed message and extracting important information.

[0056] "Means for the server to generate a response" refers to the functionality for generating an appropriate response based on the analysis results.

[0057] The term "means for the server to send a response to the terminal" refers to a function for sending the generated response to the user's terminal.

[0058] "Means for the terminal to display the response to the user" refers to a function for visually displaying the response received from the server to the user.

[0059] The object of this invention is to provide a system for text-based communication that properly understands the nuances of messages sent by users, detects offensive or indifferent expressions, makes suggestions for improvement, and generates appropriate responses.

[0060] System program generation

[0061] First, a user sends a message using a text-based communication method (e.g., email or a messenger app). The device sends this message to the server. The server preprocesses the received message, performs semantic analysis, generates an appropriate response, and sends it to the device. The device displays the received response to the user.

[0062] Hardware and Software Used

[0063] The system uses the following hardware and software:

[0064] Hardware: User's device (smartphone, PC, etc.), server

[0065] Software: Natural language processing libraries (e.g., Python's "NLTK" or "spaCy"), generative AI models (e.g., OpenAI's "GPT-3")

[0066] Data processing and calculation

[0067] The server performs the following data processing and calculations:

[0068] 1. Text preprocessing: The server tokenizes the received message and removes unnecessary symbols and stop words. For this process, we use "NLTK" and "spaCy".

[0069] 2. Semantic Analysis: The server analyzes the semantics of the preprocessed text and extracts keywords and important phrases using natural language processing models such as "spaCy" or "BERT".

[0070] 3. Response Generation: The server generates an appropriate response based on the analysis results. This response generation uses a generative AI model (e.g., GPT-3).

[0071] Examples

[0072] As a concrete example, let's say a user uses a messenger app to send the following message:

[0073] "What time does the meeting start tomorrow?"

[0074] The server receives this message and processes it as follows:

[0075] 1. Text preprocessing: The server tokenizes the text “What time does the meeting start tomorrow?” and breaks it down into the words “tomorrow,” “meeting,” “what time,” “start,” and “is.”

[0076] 2. Semantic analysis: The server extracts the keywords "meeting" and "what time" and understands that the user is asking about the start time of the meeting.

[0077] 3. Response generation: The server uses a generative AI model (e.g., GPT-3) to generate an appropriate response. It inputs the prompt "Message from user: What time does tomorrow's meeting start? Please generate a response." and generates the response "Tomorrow's meeting starts at 10:00 a.m."

[0078] In this manner, a system is implemented in which the server analyzes a user's message and generates and delivers an appropriate response to the user.

[0079] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0080] Step 1:

[0081] A user sends a message.

[0082] Specific actions: A user opens a messenger app, types "What time is the meeting tomorrow?", and hits send.

[0083] Input: The text message entered by the user.

[0084] Output: The command is sent to the terminal.

[0085] Step 2:

[0086] The terminal sends a message to the server.

[0087] Specific operation: The terminal transmits the entered message to a server via the Internet.

[0088] Input: The text message entered by the user.

[0089] Output: A message is sent to the server.

[0090] Step 3:

[0091] The server receives the message.

[0092] Specific operation: The server receives the message sent from the terminal and stores it in a database.

[0093] Input: The text message sent from the device.

[0094] Output: The received message is stored in the database.

[0095] Step 4:

[0096] The server pre-processes the message.

[0097] Specific operation: The server tokenizes the received message and removes unnecessary symbols and stop words. This process is done using "NLTK" and "spaCy".

[0098] Input: The received text message.

[0099] Output: Preprocessed text data.

[0100] Step 5:

[0101] The server semantically parses the message.

[0102] How it works: The server analyzes the preprocessed text and extracts keywords and important phrases using natural language processing models such as spaCy and BERT.

[0103] Input: Preprocessed text data.

[0104] Output: Parsed keywords or phrases.

[0105] Step 6:

[0106] The server generates a response.

[0107] Specific operation: The server generates an appropriate response based on the analysis results. A generative AI model (e.g., GPT-3) is used to generate this response. The server inputs the prompt text "Message from user: What time does the meeting start tomorrow? Please generate a response." and generates a response.

[0108] Input: The parsed keyword or phrase.

[0109] Output: The generated response text.

[0110] Step 7:

[0111] The server sends a response to the terminal.

[0112] Specific operation: The server sends the generated response to the user's terminal via the Internet.

[0113] Input: The generated response text.

[0114] Output: The response is sent to the terminal.

[0115] Step 8:

[0116] The terminal displays the response to the user.

[0117] Specific operation: The terminal visually displays the response received from the server to the user. The user confirms the displayed response.

[0118] Input: The response text sent by the server.

[0119] Output: The response text that is displayed to the user.

[0120] (Application example 1)

[0121] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0122] Conventional text-based communication systems can detect the nuances in the text sent by users and make appropriate suggestions for improvement, but they lack the functionality to recommend related content based on the user's interests and concerns. This limits the information and knowledge that users can obtain through communication, making it difficult to improve the quality of communication.

[0123] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0124] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for suggesting improvements to make the text more "kind" based on the detected nuances, and a means for analyzing the user's text and recommending related content based on the content of the text. This allows the user to not only improve the quality of communication, but also to obtain related content based on their interests and concerns.

[0125] "Text-based communication" is any means by which users exchange information using written text.

[0126] "Aggressive or indifferent undertones" are expressions contained in the text that indicate negative or indifferent feelings or attitudes.

[0127] "Thoughtful writing" is text that includes expressions that show consideration and kindness to the other person.

[0128] "Improvement Suggestions" are suggested modifications to make the content of the text a user sends more appropriate and considerate.

[0129] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language.

[0130] "Related content" is information such as articles, videos, and music that is recommended based on the content of the user's text.

[0131] A "recommendation method" is a method that analyzes a user's text and presents relevant content based on that content.

[0132] A system for implementing the present invention includes means for conducting text-based communication, means for detecting aggressive or insensitive nuances in the text-based communication, means for suggesting improvements to make the text more "kind" based on the detected nuances, and means for analyzing a user's text and recommending related content based on the content of the text.

[0133] The server receives the text sent by the user and uses natural language processing technology to analyze the meaning of the text. Specifically, it uses OpenAI's GPT-3 model to understand the content of the text and detect offensive or insensitive nuances. Based on the detected nuances, the server generates improvement suggestions for "kind" text.

[0134] In addition, the server recommends related content based on the content of the user's text. For example, it uses a news API to retrieve articles related to the user's interests and presents them to the user. This series of processes is implemented using a Python program.

[0135] As a concrete example, if a user sends a message saying, "I would like to know about recent technological advances," the server will input the following prompt sentence into the generative AI model:

[0136] Please recommend related content based on the content of the text below:

[0137] I want to know about the latest advances in technology

[0138] Based on this prompt, the generative AI model analyzes the meaning of the text and recommends relevant content. Article titles and URLs retrieved using a news API are presented to the user.

[0139] The hardware used is a smartphone, and the software used is Python, OpenAI API, and news API. This not only improves the quality of communication, but also allows users to obtain relevant content based on their interests.

[0140] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0141] Step 1:

[0142] A user sends a message using a text-based communication method (e.g. email or a messenger app). The input is the text the user sends, and the output is that text being sent to the server.

[0143] Step 2:

[0144] The server uses natural language processing techniques to analyze the received text. Specifically, it uses OpenAI's GPT-3 model to understand the meaning of the text. The input is the text sent by the user, and the output is the result of the semantic analysis of the text.

[0145] Step 3:

[0146] The server detects offensive or insensitive nuances in the text based on the analysis results. The input is the result of semantic analysis of the text, and the output is the detected nuance information.

[0147] Step 4:

[0148] The server generates improvement suggestions for "kind" sentences based on the detected nuances. The input is the nuance information, and the output is the text of the improvement suggestions.

[0149] Step 5:

[0150] The server recommends related content based on the content of the user's text. Specifically, it uses a news API to get related articles. The input is the user's text and the analysis results, and the output is a list of related articles.

[0151] Step 6:

[0152] The server presents the retrieved related content to the user. The input is a list of related articles, and the output is the titles and URLs of the articles presented to the user.

[0153] Step 7:

[0154] The user reviews the presented related content and browses it as needed. The input is the title and URL of the presented article, and the output is the user's browsing behavior.

[0155] Example 2

[0156] Next, a second embodiment of the second embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0157] When text-based communication contains offensive or insensitive nuances, it can lead to misunderstandings and conflicts. Users may also unintentionally use offensive language, lowering the quality of communication. There is a need to solve these problems and enable users to communicate more considerately.

[0158] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0159] In this invention, the server includes a means for performing text-based communication, a means for preprocessing text data, a means for performing sentiment analysis using a generative AI model, a means for outputting the results of the sentiment analysis, and a means for making suggestions for improving the text to be more "kind" based on the detected nuances. This makes it possible to detect aggressive or indifferent nuances in the text data entered by the user and make appropriate suggestions for improvement.

[0160] "Text-based communication" is a method of exchanging information using written text.

[0161] "Aggressive or indifferent nuances" refer to expressions or attitudes that show hostility or indifference toward others.

[0162] "Suggestions for improving language to be more considerate" are suggestions for changing text that contains aggressive or insensitive nuances into more friendly and considerate language.

[0163] "Text data preprocessing" refers to the process of removing unnecessary spaces and special characters from text data and dividing it into words and phrases.

[0164] A "generative AI model" is a model that uses artificial intelligence technology to analyze the sentiment and meaning of text.

[0165] "Sentiment analysis" is the process of analyzing the emotions and nuances contained in text data to determine whether they are aggressive, indifferent, or some other emotion.

[0166] "Output of sentiment analysis results" refers to providing the results obtained by sentiment analysis to a user.

[0167] The present invention is a system that detects aggressive or indifferent nuances in text-based communication and suggests improvements to make the text more considerate. A specific embodiment of this system will be described below.

[0168] System Configuration

[0169] Hardware and software configuration:

[0170] Server: A computer with high-performance computing power that performs data pre-processing, sentiment analysis, and outputting the results.

[0171] Terminal: A device (e.g., PC, smartphone) through which a user enters text data and receives the results.

[0172] Generative AI model: An artificial intelligence model for performing sentiment analysis (e.g., OpenAI's GPT-3).

[0173] Natural Language Processing library: Software used to preprocess text data (e.g., NLTK in Python).

[0174] Program Processing

[0175] Enter data:

[0176] The user uses the device to input text data to be analyzed, such as comments on social media or the contents of emails.

[0177] Data preprocessing:

[0178] The server preprocesses the received text data by removing unnecessary spaces and special characters and tokenizing the text using the Python NLTK (Natural Language Toolkit) library.

[0179] Sentiment analysis:

[0180] The server performs sentiment analysis on the pre-processed text data, using a generative AI model (e.g., OpenAI's GPT-3) to determine whether the text contains offensive or insensitive nuances.

[0181] The resulting output:

[0182] The server returns the results of the sentiment analysis to the user, indicating whether the text contains aggressive, indifferent, or other sentiments.

[0183] Examples

[0184] Let's say a user wants to analyze the social media comment "I don't care about you."

[0185] 1. Enter your data:

[0186] The user types "I don't care about you" into an input field on the terminal.

[0187] The terminal transmits this text data to the server.

[0188] 2. Data preprocessing:

[0189] The server preprocesses the received text data "I don't care about you."

[0190] Remove unnecessary spaces and special characters.

[0191] Tokenize the text and split it into "you," "something," and "it doesn't matter."

[0192] 3. Sentiment analysis:

[0193] The server creates a prompt sentence to input the preprocessed text data into the generative AI model:

[0194] "Please determine whether the following text contains offensive or insensitive connotations:

[0195] "I don't care about you."

[0196] A generative AI model takes this prompt and analyzes the sentiment of the text.

[0197] The generative AI model determines that the text is "offensive" and returns that result to the server.

[0198] 4. Result output:

[0199] The server returns the result it received from the generative AI model, “aggressive,” to the user.

[0200] The device displays the result to the user, for example, "This comment is offensive."

[0201] The above is a specific embodiment for carrying out the present invention. This system allows users to detect offensive or indifferent nuances contained in text data and receive appropriate suggestions for improvement.

[0202] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0203] Step 1:

[0204] Entering data

[0205] The user inputs the text data to be analyzed using the terminal. For example, the user inputs comments on social media or the contents of an email. The terminal then transmits the input text data to the server.

[0206] Input: Text data entered by the user (e.g. "I don't care about you")

[0207] Output: Text data sent from the terminal to the server

[0208] Step 2:

[0209] Data Preprocessing

[0210] The server preprocesses the received text data by removing unnecessary spaces and special characters and tokenizing the text using the Python NLTK (Natural Language Toolkit) library.

[0211] Input: Text data sent from the terminal

[0212] Data processing: removing unnecessary spaces and special characters, tokenizing text

[0213] Output: Preprocessed text data (e.g. "you", "something", "it doesn't matter")

[0214] Step 3:

[0215] sentiment analysis

[0216] The server performs sentiment analysis on the preprocessed text data, using a generative AI model (e.g., OpenAI's GPT-3) to determine whether the text contains offensive or insensitive nuances. The server inputs the following prompt sentence into the generative AI model:

[0217] "Please determine whether the following text contains offensive or insensitive connotations:

[0218] "I don't care about you."

[0219] A generative AI model takes this prompt and analyzes the sentiment of the text.

[0220] Input: Preprocessed text data, prompt

[0221] Data Computation: Sentiment Analysis with Generative AI Models

[0222] Output: Result of sentiment analysis (e.g. "offensive")

[0223] Step 4:

[0224] Output of results

[0225] The server returns the result of the sentiment analysis to the user, indicating whether the text contains offensive, indifferent, or other sentiments. The terminal displays the result received from the server to the user, for example, "This comment is offensive."

[0226] Input: Sentiment analysis results

[0227] Output: The result that is displayed to the user (e.g. "This comment is offensive")

[0228] The above is the specific processing flow of the program of this system.

[0229] (Application example 2)

[0230] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0231] In modern text-based communication, messages containing aggressive or insensitive nuances are frequently sent and received, which can lead to deterioration of relationships between users and loss of psychological safety. In addition, there is a lack of means to detect messages containing these nuances in real time and warn users, making it difficult to improve the quality of communication.

[0232] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0233] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for suggesting improvements to make the text more "considerate" based on the detected nuances, and a means for analyzing messages in real time, detecting messages containing aggressive or indifferent nuances, and issuing a warning to the user. This allows the user to receive a warning before sending a message containing aggressive or indifferent nuances, thereby improving the quality of communication.

[0234] "Text-based communication" is a method of exchanging information using written text.

[0235] An "aggressive nuance" is a nuance that includes words or expressions that show hostility or aggression toward the other person.

[0236] An "indifferent nuance" is a nuance that includes words and expressions that show no interest or concern for the other person.

[0237] "Thoughtful writing" is writing that takes into consideration the other person's feelings and situation and includes words and expressions that show consideration and kindness.

[0238] "Analyzing messages in real time" means analyzing the content of a message immediately at the moment it is sent.

[0239] "Warning the user" means notifying the user about a detected message that contains offensive or insensitive nuances.

[0240] "Natural language processing technology" is a technology that enables computers to process and understand human language.

[0241] "Sentiment analysis" is a technology that analyzes emotions and nuances in text and classifies and evaluates its content.

[0242] In order to implement the present invention, the following system configuration and processing procedures are required.

[0243] System Configuration

[0244] 1. Hardware:

[0245] Smartphone: A device that allows users to engage in text-based communication.

[0246] Server: Provides the computational resources for sentiment analysis and nuance detection.

[0247] 2. Software:

[0248] Python: A programming language for running programs for sentiment analysis.

[0249] NLTK (Natural Language Toolkit): A library for natural language processing.

[0250] VADER (Valence Aware Dictionary and sEntiment Reasoner): A model for performing sentiment analysis.

[0251] Processing Steps

[0252] 1. Data Entry:

[0253] A user enters a text-based message using a smartphone.

[0254] 2. Data preprocessing:

[0255] The smartphone sends the entered message in text format to the server.

[0256] 3. Sentiment analysis:

[0257] The server uses NLTK's VADER sentiment analysis model to calculate a sentiment score for each received message.

[0258] 4. Nuance detection:

[0259] The server uses the sentiment score to determine whether the message contains offensive or insensitive overtones.

[0260] 5. Warning Issued:

[0261] Based on the detection results, the server will display a warning to the user about messages that contain offensive or insensitive nuances.

[0262] Examples

[0263] For example, if a user tries to send a message saying "You're really incompetent," the smartphone will send this message to the server. The server will use NLTK's VADER sentiment analysis model to calculate the sentiment score of this message and determine that it contains offensive nuances. As a result, the server will display a warning to the user saying "Offensive message detected."

[0264] Examples of prompt statements

[0265] Here is an example of a prompt to be input to the generative AI model:

[0266] Please judge whether the following messages contain offensive or insensitive connotations:

[0267] Message: "You're really incompetent."

[0268] In this way, users can be warned before sending messages that may be offensive or insensitive, improving the quality of communication.

[0269] The flow of the specific process in the application example 2 will be described with reference to FIG.

[0270] Step 1:

[0271] A user enters a text-based message using a smartphone.

[0272] Input: The text message entered by the user.

[0273] Output: The typed text message displayed on the smartphone.

[0274] Specific behavior: A user opens a chat application and enters text into the message input field.

[0275] Step 2:

[0276] The terminal transmits the input message in text format to the server.

[0277] Input: The text message entered by the user.

[0278] Output: The text message sent to the server.

[0279] Specific operation: The smartphone application sends the entered text message to the server as an HTTP request.

[0280] Step 3:

[0281] The server uses NLTK's VADER sentiment analysis model to calculate a sentiment score for each received message.

[0282] Input: The text message sent to the server.

[0283] Output: Sentiment scores (positive, negative and neutral scores).

[0284] Specific operation: The server inputs the received text message into the VADER sentiment analysis model and calculates positive, negative, and neutral scores.

[0285] Step 4:

[0286] The server uses the sentiment score to determine whether the message contains offensive or insensitive overtones.

[0287] Input: Sentiment score.

[0288] Output: Nuance judgment result (aggressive, indifferent, normal).

[0289] Specific operation: The server analyzes the emotion score, and if the negative score is high, it is determined to be "aggressive," if the neutral score is high, it is determined to be "indifferent," and otherwise it is determined to be "normal."

[0290] Step 5:

[0291] Based on the detection results, the server will display a warning to the user about messages that contain offensive or insensitive nuances.

[0292] Input: Nuance judgment result.

[0293] Output: The warning message that is displayed to the user.

[0294] Specific operation: The server sends the judgment result to the smartphone, and the smartphone application displays a warning to the user such as "Offensive message has been detected."

[0295] Example 3

[0296] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0297] In text-based communication, aggressive or indifferent nuances are often included, which can lead to misunderstandings and conflicts. In addition, there is a lack of concrete ways for users to improve their expressions to be more "considerate." This makes it difficult for users to communicate smoothly and considerately.

[0298] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0299] In this invention, the server includes a means for a user to input text, a means for a terminal to transmit the input text to the server, a means for the server to analyze the nuances of the text, a means for the server to generate improvement suggestions using a generative AI model, a means for the server to transmit the improvement suggestions to the terminal, and a means for the terminal to display the improvement suggestions to the user. This enables the user to receive specific suggestions for improving aggressive or insensitive expressions to "considerate" expressions.

[0300] "Text-based communication" is a method of exchanging information using written text.

[0301] "Nuance" refers to the subtle meanings and emotions contained in text.

[0302] "Offensive language" is language that has the potential to offend others or provoke negative feelings.

[0303] "Indifferent expressions" are language that makes the other person feel that you lack interest or consideration.

[0304] "Considerate expression" is language that shows consideration for the other person's feelings and position and shows kindness and understanding.

[0305] "User" means any individual or entity that uses the System.

[0306] A "Terminal" is a device through which a User accesses the System.

[0307] A "server" is a computer system that receives and processes requests from users.

[0308] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[0309] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate new text or suggestions.

[0310] "Improvement suggestions" are specific suggestions for replacing aggressive or insensitive language with more "compassionate" language.

[0311] The present invention is a system that detects aggressive or indifferent nuances in text-based communication and suggests ways to improve the text to be more “considerate.” A specific embodiment of this system will be described below.

[0312] First, the user inputs text using the terminal. For example, the user inputs the text "You are wrong." The terminal transmits the input text to the server. For transmission, a communication protocol such as an HTTP request is used.

[0313] The server analyzes the received text using natural language processing (NLP) techniques, specifically using NLP tools such as Python's NLTK library and spaCy to identify offensive or insensitive content in the text.

[0314] Next, the server generates improvement suggestions using a generative AI model based on the analysis results. For example, OpenAI's GPT-4 (registered trademark) is used as this generative AI model. The server inputs the following prompt sentence to the generative AI model:

[0315] Improve your statement from "You're wrong" to something more "compassionate."

[0316] The generative AI model generates an improvement suggestion such as, "I understand your point of view, but I'm thinking about it from a slightly different perspective." The server sends this improvement suggestion to the terminal. For sending, a communication protocol such as HTTP response is used.

[0317] The terminal displays the received improvement suggestions to the user. The user can check the displayed improvement suggestions and modify the text as necessary. For example, the user can use the displayed improvement suggestions as they are or can make further modifications.

[0318] This system allows users to receive specific suggestions for improving aggressive or indifferent expressions into "kind" expressions. This allows smooth and kind communication to be realized. The flow of the identification process in the third embodiment will be described with reference to FIG. 15.

[0319] Step 1:

[0320] The user enters text.

[0321] The user inputs the text to be improved in the input field of the terminal, for example, "You are wrong." The input text is temporarily stored in the terminal's memory.

[0322] Step 2:

[0323] The terminal sends the entered text to the server.

[0324] The terminal sends the entered text as an HTTP request to the server. The input data is packaged in JSON format and sent over the network to the server. For example, the following JSON data is sent:

[0325] json

[0326] {

[0327] "text": "You are wrong"

[0328] }

[0329] Step 3:

[0330] The server analyzes the text for nuances.

[0331] The server analyzes the received text using natural language processing (NLP) techniques. Specifically, it uses NLP tools such as Python's NLTK library and spaCy to identify offensive or insensitive expressions in the text. The input data is the text, and the output data is nuanced information as a result of the analysis. For example, the part that says "wrong" is determined to be offensive.

[0332] Step 4:

[0333] The server generates improvement suggestions using the generative AI model.

[0334] The server uses a generative AI model (e.g., OpenAI's GPT-4) based on the analysis results to generate improvement suggestions for more "compassionate" expressions. The server inputs the following prompt sentence into the generative AI model:

[0335] Improve your statement from "You're wrong" to something more "compassionate."

[0336] The generative AI model generates improvement suggestions such as, "I understand your point of view, but I think about it from a slightly different perspective." The input data is the prompt sentence, and the output data is the improvement suggestions.

[0337] Step 5:

[0338] The server sends improvement suggestions to the device.

[0339] The server sends the generated improvement proposal to the terminal as an HTTP response. The output data is packaged in JSON format and sent to the terminal via the network. For example, the following JSON data is sent:

[0340] json

[0341] {

[0342] "suggestion": "I understand your point of view, but I think about it from a slightly different perspective."

[0343] }

[0344] Step 6:

[0345] The terminal displays improvement suggestions to the user.

[0346] The terminal displays the improvement suggestions received from the server to the user. The user can check the displayed improvement suggestions and modify the text as necessary. For example, the user can use the displayed improvement suggestions as they are or modify them further. The input data is the improvement suggestions, and the output data is the text displayed to the user.

[0347] (Application example 3)

[0348] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0349] Text-based communication often contains aggressive or insensitive expressions, which can lead to misunderstandings and conflicts. Users may also unintentionally use expressions that hurt others. This leads to problems in that the quality of communication declines and psychological safety is lost. In addition, there is a lack of ways for users to receive concrete feedback to improve their expressions, which makes it difficult to improve communication skills.

[0350] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0351] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making improvement suggestions to make the text more "kind" based on the detected nuances, a means for analyzing messages sent by users in real time and displaying improvement suggestions, and a means for generating improvement suggestions using a generative AI model. This allows users to receive specific feedback for improving their own expressions in real time, improving the quality of communication and ensuring psychological safety.

[0352] "Text-based communication" is a method of exchanging information using written text.

[0353] "Nuance" refers to the subtle meanings and emotions contained in words and expressions.

[0354] "Offensive language" is any word or phrase intended to injure or provoke another person.

[0355] "Indifferent expressions" are cold words or phrases that show no interest in the other person's feelings or opinions.

[0356] "Thoughtful writing" is expression that shows respect and consideration for the other person's feelings and position.

[0357] "Improvement suggestions" are specific advice or proposed revisions to improve the current expression.

[0358] "Real-time analysis" means that the content of a message is analyzed immediately as the user types it.

[0359] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate new data or information.

[0360] "Feedback" means evaluation or advice provided to a user in response to their actions or expressions.

[0361] "Psychological safety" refers to an environment in which individuals feel safe to express their opinions and take action.

[0362] A system for implementing this invention includes means for conducting text-based communication, means for detecting aggressive or insensitive nuances, means for making suggestions for improvements to make the text more "kind" based on the detected nuances, means for analyzing messages sent by users in real time and displaying suggestions for improvement, and means for generating suggestions for improvement using a generative AI model.

[0363] Hardware and Software Configuration

[0364] Hardware:

[0365] Smartphone (iOS or ANDROID (registered trademark))

[0366] software:

[0367] Python

[0368] OpenAI API

[0369] Data processing and calculation

[0370] server:

[0371] 1. A user types a text message using a smartphone.

[0372] 2. The entered message is sent to the server.

[0373] 3. The server analyzes the received message and detects offensive or insensitive nuances.

[0374] 4. Based on the detected nuances, a generative AI model (e.g., OpenAI’s GPT-3) is used to generate improvement suggestions for making the text more “kind.”

[0375] 5. Improvement suggestions are displayed in real time on the user's smartphone.

[0376] Examples

[0377] When a user types "You're wrong," the server analyzes the message to detect offensive nuances, then uses a generative AI model to generate a suggestion for improvement, such as "I understand your point of view, but I think about it from a slightly different perspective," and displays it to the user.

[0378] Example prompt for a generative AI model:

[0379] Please improve the following message to make it more considerate: You are wrong

[0380] In this way, users can get specific feedback to improve their expression in real time, which improves the quality of communication and ensures psychological safety.

[0381] The flow of the specific process in the application example 3 will be described with reference to FIG.

[0382] Step 1:

[0383] A user enters a text message using a smartphone.

[0384] Input: A text message entered by the user (e.g. "You are wrong")

[0385] Output: The entered text message is sent to the server by pressing the send button on the smartphone.

[0386] Step 2:

[0387] The server analyzes the messages it receives and detects offensive or indifferent nuances.

[0388] Input: Text message sent from smartphone

[0389] Data processing: Using natural language processing techniques, we analyze the content of messages to detect offensive or insensitive nuances.

[0390] Output: Detected nuance information (e.g. "contains offensive language")

[0391] Step 3:

[0392] Based on the nuances detected, the server uses a generative AI model to generate improvement suggestions for making the text more "considerate."

[0393] Input: Detected nuance information and original text message

[0394] Data computation: Send prompts to a generative AI model (e.g., OpenAI’s GPT-3) to generate improvement suggestions.

[0395] Output: Text of the suggestion for improvement (e.g. "I understand your point of view, but I have a slightly different perspective")

[0396] Step 4:

[0397] The server displays the generated improvement suggestions on the user's smartphone in real time.

[0398] Input: Text of improvement suggestion

[0399] Data processing: The improvement suggestions are converted into a data format for sending to the user's smartphone.

[0400] Output: Suggested improvements displayed on the user's smartphone (e.g., "I understand your point of view, but I have a slightly different perspective")

[0401] Step 5:

[0402] The user reviews the improvement suggestions and, if necessary, submits a revised message.

[0403] Input: User's correction message after reviewing the improvement suggestions

[0404] Output: The corrected text message is finally sent.

[0405] In this way, users can get specific feedback to improve their expression in real time, which improves the quality of communication and ensures psychological safety.

[0406] Furthermore, an emotion engine that estimates the emotion of the user may be combined. That is, the identification processing unit 290 may estimate the emotion of the user using the emotion identification model 59, and perform identification processing using the emotion of the user.

[0407] "Example 1"

[0408] In one embodiment of the present invention, the system includes an emotion engine that recognizes the emotion of a user. The emotion engine infers the user's emotional state from the text when the user engages in text-based communication. For example, if the user enters the text "I am angry," the emotion engine infers that the user is in an angry emotional state.

[0409] "Example 2"

[0410] In one embodiment of the present invention, the emotion engine adjusts the remediation suggestions based on the estimated emotional state. For example, if the user is estimated to be in an angry emotional state, the emotion engine adjusts the remediation suggestions for "kind" text to appropriately express the anger, thereby respecting the user's emotional state and reducing aggressive or indifferent overtones.

[0411] "Example 3"

[0412] In one embodiment of the present invention, the emotion engine tracks the user's emotional state over time and

[0413] For example, if a user is estimated to be in an angry emotional state for a certain period of time, the emotion engine can use that information to understand why the user is feeling angry and suggest ways to deal with it, thereby helping the user improve their communication skills.

[0414] The process flow of each embodiment will be described below.

[0415] "Example 1"

[0416] Step 1: Users engage in text-based communication.

[0417] Step 2: The emotion engine infers the emotional state of the user from the text. For example, from the text "I am angry", it infers that the user is in an angry emotional state.

[0418] "Example 2"

[0419] Step 1: The emotion engine estimates the user's emotional state.

[0420] Step 2: The system adjusts the remediation suggestions based on the estimated emotional state. For example, if the user is estimated to be in an angry emotional state, the system will adjust the remediation suggestions for "kind" sentences to better express the anger.

[0421] "Example 3"

[0422] Step 1: The emotion engine estimates the user's emotional state and records that information.

[0423] Step 2: The emotion engine tracks the user's emotional state over time.

[0424] Step 3: Based on the long-term emotional state tracking results, the system helps users improve their communication skills. For example, if the user is estimated to be in an angry emotional state for a certain period of time, the system uses the information to understand the reasons why the user feels angry and suggests ways to deal with it.

[0425] Example 1

[0426] Next, a description will be given of Example 1 of the embodiment 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal".

[0427] In modern text-based communication, it is difficult to accurately understand the user's emotions and nuances and provide appropriate feedback and suggestions for improvement. In particular, messages that contain aggressive or indifferent nuances can reduce the quality of communication and undermine psychological safety. It is also difficult for users to accurately recognize and appropriately express their own emotional state, so there is a demand for smoother communication.

[0428] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0429] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making suggestions for improving the text to be more "kind" based on the detected nuances, a means for receiving a text message input by a user, a means for analyzing the text message using natural language processing technology, a means for estimating the user's emotional state based on the analysis results, and a means for notifying the user of the estimation results. This makes it possible to accurately understand the user's emotions and nuances and provide appropriate feedback and suggestions for improvement.

[0430] "Text-based communication" refers to communication methods that use written information, including email and messenger apps.

[0431] "Aggressive or indifferent nuances" refers to expressions contained in text messages that convey negative or indifferent feelings or attitudes.

[0432] "Suggestions for improving considerate language" are suggestions for changing text messages that contain aggressive or insensitive nuances into more positive and considerate language.

[0433] The "means for receiving a text message input by a user" is a function that enables the server to receive a text message sent by a user through a terminal.

[0434] "Means for analyzing text messages using natural language processing techniques" refers to techniques for understanding the content of text messages and analyzing their meaning and context, and includes tokenization, part-of-speech tagging, context analysis, etc.

[0435] The "means for inferring the emotional state of the user based on the analysis results" is a function for inferring the emotional state of the user by using the analysis results obtained by natural language processing technology.

[0436] The "means for notifying the user of the estimation result" is a function for notifying the user of the result of the estimated emotional state.

[0437] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0438] The present invention is a system for text-based communication that analyzes text messages entered by a user, estimates the user's emotional state, and provides appropriate feedback and suggestions for improvement. A specific embodiment of the system will be described below.

[0439] Hardware and Software Used

[0440] server

[0441] The server plays a central role in receiving and analyzing text messages sent by users. The following software is installed on the server:

[0442] Natural Language Processing (NLP) technology: Google® Natural Language API, Microsoft® Text Analytics API

[0443] Emotion engine: IBM's Watson(R) Tone Analyzer, Amazon's Comprehend

[0444] Terminal

[0445] A terminal is a device that allows a user to input text messages and communicate with a server, such as a smartphone or PC.

[0446] Program Processing

[0447] User Actions

[0448] A user inputs a text message using a text-based communication method such as email or a messenger app. For example, the user inputs the message "I am angry."

[0449] Device operation

[0450] The device transmits the text message entered by the user to the server, which requires an Internet connection.

[0451] Server Operations

[0452] The server receives text messages sent from the terminals and temporarily stores the received messages in a database.

[0453] The server then analyzes the text messages using natural language processing (NLP) techniques, such as Google's Natural Language API and Microsoft's Text Analytics API, to tokenize the text, tag parts of speech, and perform contextual analysis.

[0454] Based on the analysis results, the server uses an emotion engine to estimate the user's emotional state. This emotion engine uses IBM's Watson Tone Analyzer and Amazon's Comprehend. For example, it estimates that the user is in an angry emotional state from the keyword "angry."

[0455] The estimation results are sent from the server to the terminal using a secure communication protocol (e.g., HTTPS).

[0456] Device operation

[0457] The terminal displays the estimated emotional state received from the server to the user. For example, if the user inputs "I am angry," the terminal displays "You are in an angry emotional state."

[0458] Examples

[0459] As a concrete example, consider a case where a user uses a messenger app to send a message saying "I'm angry."

[0460] 1. A user types "I'm angry" into a messenger app.

[0461] 2. The device sends this message to the server.

[0462] 3. The server receives the message and stores it in a database.

[0463] 4. The server uses Google's natural language API to extract the keyword "angry."

[0464] 5. The server uses IBM's Watson Tone Analyzer to estimate that the user is in an angry emotional state.

[0465] 6. The server sends the estimation results to the device.

[0466] 7. The device displays to the user, "You are in an angry emotional state."

[0467] Examples of prompts to be input to a generative AI model include:

[0468] A user sends a message saying "I am angry". Analyze this message and infer the user's emotional state.

[0469] By feeding this prompt into a generative AI model, the model can infer the user's emotional state and generate an appropriate response.

[0470] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0471] Step 1:

[0472] The user enters a text message.

[0473] A user enters a text message using a text-based means of communication such as email or a messenger app. For example, the user enters the message "I am angry." The entered text message is the output of this step.

[0474] Step 2:

[0475] The device sends a text message to the server.

[0476] The terminal transmits the text message entered by the user to the server. This transmission requires an Internet connection. The input is the text message entered by the user and the output is the text message transmitted to the server.

[0477] Step 3:

[0478] A server receives the text message.

[0479] The server receives text messages sent from the terminal. The received messages are temporarily stored in a database. The input is the text message sent from the terminal, and the output is the text message stored in the database.

[0480] Step 4:

[0481] The server analyzes the text message using natural language processing (NLP) techniques.

[0482] The server analyzes the received text messages using natural language processing (NLP) technology. Google's natural language API and Microsoft's Text Analytics API are used for this analysis. Specifically, text tokenization, part-of-speech tagging, context analysis, etc. are performed. The input is the text message stored in the database, and the output is the analysis result.

[0483] Step 5:

[0484] The server uses an emotion engine to estimate the user's emotional state.

[0485] The server uses an emotion engine based on the analysis results to estimate the user's emotional state. IBM's Watson Tone Analyzer and Amazon's Comprehend are used for this emotion engine. For example, it estimates that the user is in an angry emotional state based on the keyword "angry." The input is the analysis results, and the output is the estimated emotional state.

[0486] Step 6:

[0487] The server transmits the estimation result to the terminal.

[0488] The server transmits the result of the emotional state estimated by the emotion engine to the terminal. A secure communication protocol (e.g., HTTPS) is used for this transmission. The input is the estimated result of the emotional state, and the output is the estimation result sent to the terminal.

[0489] Step 7:

[0490] The terminal displays the estimation results to the user.

[0491] The terminal displays the estimation result of the emotional state received from the server to the user. For example, if the user inputs "I am angry," the terminal displays "You are in an angry emotional state." The input is the estimation result sent from the server, and the output is the estimation result of the emotional state displayed to the user.

[0492] (Application example 1)

[0493] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0494] In modern text-based communication, it is necessary to properly recognize the user's emotional state and respond quickly when an abnormal emotional state is detected. However, conventional systems often have difficulty accurately grasping the user's emotional state, and appropriate responses are often delayed. They also lack the ability to detect aggressive or indifferent nuances and suggest improvements to make the text more considerate. This may undermine the user's psychological safety.

[0495] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0496] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for suggesting improvements to make the text more "kind" based on the detected nuances, a means for estimating the emotional state of the user, and a means for issuing an alert if the emotional state is abnormal. This makes it possible to monitor the emotional state of the user in real time and quickly respond when an abnormality is detected.

[0497] "Text-based communication" refers to the exchange of information using written text, including email and messenger apps.

[0498] "Aggressive or indifferent nuances" are words or expressions that show hostility or indifference toward the other person, and can have a negative impact on communication.

[0499] "Suggestions for improving language to be more considerate" refers to making suggestions to change text that contains aggressive or insensitive nuances into more friendly and considerate language.

[0500] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language, and is used to understand the meaning of text.

[0501] "Means for inferring a user's emotional state" refers to technologies or algorithms for analyzing a user's emotions from text-based communications and inferring that emotional state.

[0502] "Means of sending an alert" refers to a function for sending a notification to an administrator or security officer if a user's emotional state is determined to be abnormal.

[0503] As an embodiment of the present invention, the following system can be constructed.

[0504] First, the server provides a means for text-based communication. This means can be in the form of email, messenger apps, etc. When a user communicates using these means, the server receives the text and uses natural language processing technology to understand the meaning of the text.

[0505] Next, the server has a means for detecting offensive or insensitive nuances in the text-based communication, which is realized using a natural language processing library (e.g., the nltk library), and specifically, a SentimentIntensityAnalyzer is used to analyze the sentiment score of the text.

[0506] In addition, the server provides a means to suggest improvements to make sentences more "kind" based on the detected nuances. This means uses a generative AI model to generate appropriate suggestions. For example, it generates prompt sentences to convert aggressive expressions into friendly expressions.

[0507] The server also includes a means for estimating the user's emotional state, which includes techniques and algorithms for analyzing the user's emotions from text-based communications and estimating the emotional state. Specifically, the server uses a SentimentIntensityAnalyzer to analyze the user's emotion score.

[0508] Finally, the server provides a means for sending an alert if the emotional state is abnormal, including functionality for sending a notification to an administrator or security officer if the emotional score falls below a certain threshold, by sending an email using the smtplib library.

[0509] For example, if a user types "I'm so angry," the emotion engine will recognize this text as a negative emotion and send an alert to the administrator, with the following example prompt:

[0510] If a user types "I am so angry", the sentiment engine should recognize this text as a negative sentiment and send an alert to an admin.

[0511] In this way, emotion-aware security apps can monitor a user's emotional state in real time and respond quickly if anomalies are detected.

[0512] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0513] Step 1:

[0514] The users engage in text-based communication.

[0515] Input: Text typed by the user (e.g. "I'm so angry")

[0516] Output: The text data entered

[0517] Specific action: A user types and sends text via email or a messenger app.

[0518] Step 2:

[0519] The server receives the entered text.

[0520] Input: Text data submitted by the user

[0521] Output: Received text data

[0522] Specific operation: The server receives text data through a mail server or messenger server.

[0523] Step 3:

[0524] The server uses natural language processing techniques to analyze the meaning of the text.

[0525] Input: Received text data

[0526] Output: Sentiment score of the text (e.g. {'neg': 0.8, 'neu': 0.1, 'pos': 0.1, 'compound': -0.7})

[0527] Specific behavior: The server uses the SentimentIntensityAnalyzer from the nltk library to calculate the sentiment score of the text.

[0528] Step 4:

[0529] The server detects nuances in the text and determines whether it contains offensive or insensitive content.

[0530] Input: Sentiment score of the text

[0531] Output: Nuance detection result (e.g. aggressive)

[0532] Specific operation: The server analyzes the emotion scores and detects aggressive nuances when the negative scores are high.

[0533] Step 5:

[0534] Based on the nuances detected by the server, suggestions for improving the text to be more "considerate" are generated.

[0535] Input: Nuance detection results, original text data

[0536] Output: Text of the improvement suggestion (e.g. "I am a bit dissatisfied")

[0537] How it works: The server uses a generative AI model to generate prompts that convert aggressive expressions into friendly ones.

[0538] Step 6:

[0539] The server estimates the user's emotional state.

[0540] Input: Sentiment score of the text

[0541] Output: Estimated emotional state (e.g. anger)

[0542] Specific operation: The server estimates the user's emotional state based on the emotion score.

[0543] Step 7:

[0544] The server issues an alert if the emotional state is abnormal.

[0545] Input: Emotion state estimation result

[0546] Output: Alert notification (e.g. email notification to administrator)

[0547] Specific behavior: If the server determines that the emotional state is abnormal (e.g., very angry), it uses the smtplib library to send an alert via email to the administrator.

[0548] In this way, emotion-aware security apps can monitor a user's emotional state in real time and respond quickly if anomalies are detected.

[0549] Example 2

[0550] Next, a second embodiment of the second embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0551] In text-based communication, aggressive or indifferent nuances may be included, which may reduce the quality of communication. Also, if users cannot express their emotional state appropriately, misunderstandings and conflicts may occur. This may worsen relationships between users and damage psychological safety.

[0552] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for acquiring a text-based message created by a user, a means for recognizing the user's emotion from the message using an emotion engine, and a means for making an improvement proposal for the message using a prompt sentence for instructing to suggest an improvement of the message based on the recognized emotion of the user when the message contains an aggressive nuance, and a generative AI model. This makes it possible to improve the quality of communication by mitigating aggressive or indifferent nuances while respecting the emotional state of the user. In addition, the means for making an improvement proposal for the message in the server determines whether the message created by the user contains an aggressive nuance based on an exchange between the message created by the user and a message created by the other party, and makes an improvement proposal for the message using a prompt sentence for instructing to suggest an improvement of the message based on the recognized emotion of the user when the aggressive nuance is included, and a generative AI model.

[0553] "Text-based communication" refers to the exchange of information using written information.

[0554] "Aggressive or indifferent nuances" refer to words and expressions that show hostility or indifference toward the other person.

[0555] "Thoughtful writing" refers to words and expressions that take into consideration the other person's feelings and situation and show consideration and kindness.

[0556] "Sentiment analysis technology" refers to technology that analyzes words, phrases, and context in text and infers emotional states.

[0557] "Words, phrases, and context" refer to the basic linguistic elements that make up a text and their relationships.

[0558] "Emotional state" refers to the type and intensity of emotion expressed through text.

[0559] "Improvement suggestions" refer to suggesting more appropriate expressions or words based on detected nuances.

[0560] "Natural language processing technology" refers to technology that enables computers to understand and generate human language.

[0561] A "generative AI model" is a model for generating text using artificial intelligence.

[0562] "Inference" refers to predicting a state or outcome based on given data.

[0563] This invention is a system that detects aggressive or indifferent nuances in text-based communication and suggests ways to improve the text to be more "kind" based on the emotional state. This system works in cooperation with a server, a terminal, and a user.

[0564] The server uses sentiment analysis techniques to analyze words, phrases, and context in the text to infer the emotional state. Specifically, it uses a sentiment analysis engine built using Python and TENSORFLOW. The sentiment analysis engine uses pre-trained generative AI models (e.g., BERT and GPT-3) to analyze the text.

[0565] The device is responsible for sending the text entered by the user to the server. When the user enters text into the input field of the device, the device sends the text in JSON format to the server. The server analyzes the received text and estimates the emotional state.

[0566] For example, if a user enters "I don't understand your opinion at all," the server passes this text to a sentiment analysis engine. The sentiment analysis engine analyzes the text and infers that the user is in an angry emotional state. Next, the server uses a prompt sentence that instructs the server to suggest improvements to the message based on the recognized user's emotions if the message contains aggressive nuances, and a generative AI model to make suggestions for improving the message, generating suggestions for improving the message to be more "kind." For example, the server generates a suggestion such as, "Could you please explain your opinion in more detail?"

[0567] Here is an example of a prompt to be input to the generative AI model:

[0568] "Input text: "I don't understand your opinion at all" Emotional state: "Anger" Determine whether the input text contains offensive nuances, and if so, suggest improvements to the message based on the perceived emotional state of the user."

[0569] Example of a suggestion for improvement: "Can you explain your opinion in more detail?"

[0570] The server sends the generated improvement proposal in JSON format to the terminal. The terminal displays the received improvement proposal to the user. The user can check the displayed improvement proposal and modify the text if necessary.

[0571] In this way, it is possible to improve the quality of communication by respecting the user's emotional state while mitigating aggressive or indifferent nuances. When making an improvement suggestion, it is preferable to further consider the interaction between the user's message and the other party's message. In this case, the server judges whether the message created by the user contains an aggressive nuance based on the interaction between the message created by the user and the message created by the other party, and if an aggressive nuance is included, makes an improvement suggestion for the message using a prompt sentence instructing to suggest an improvement of the message based on the recognized emotion of the user and a generative AI model. Here is an example of a prompt to be input to the generative AI model:

[0572] After entering the exchange between your message and the other person's message, "Input text: "I don't understand your opinion at all" Emotional state: "Anger" Given the messaging conversation, determine whether the input text contains offensive nuances and, if so, suggest improvements to the message based on the perceived emotional state of the user.”

[0573] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0574] Step 1:

[0575] A user inputs text into an input field on a terminal. For example, the user inputs "I don't understand your opinion at all." This input text becomes the input data for the next processing step.

[0576] Step 2:

[0577] The terminal converts the text entered by the user into JSON format and sends it to the server as an HTTP request. At this time, the input text is sent to the server as JSON format data.

[0578] Step 3:

[0579] The server parses the JSON data received from the terminal and extracts the text part. Specifically, it extracts the "text" field from the received JSON data and passes its contents to the next processing step. This extracted text becomes the input data for the next processing step.

[0580] Step 4:

[0581] The server passes the extracted text to a sentiment analysis engine, which is built using Python and TensorFlow and analyzes the text using pre-trained generative AI models (e.g., BERT and GPT-3). The sentiment analysis engine considers words, phrases, and context in the text to infer an emotional state. This inferred emotional state becomes the input data for the next processing step.

[0582] Step 5:

[0583] The server generates an appropriate improvement suggestion based on the emotional state obtained from the sentiment analysis engine. For example, if the user is estimated to be in an angry emotional state, the server generates an improvement suggestion in a considerate text such as "Can you please explain your opinion in more detail?". This generated improvement suggestion becomes the input data for the next processing step.

[0584] Step 6:

[0585] The server converts the generated improvement proposal into JSON format and sends it to the terminal as an HTTP response. At this time, the improvement proposal is sent to the terminal as JSON format data.

[0586] Step 7:

[0587] The terminal parses the JSON data received from the server and extracts the improvement suggestion part. Specifically, it extracts the "improvement suggestion" field from the received JSON data and displays its contents to the user. The user can check the displayed improvement suggestion and modify the text if necessary.

[0588] In this way, it is possible to improve the quality of communication by respecting the user's emotional state while mitigating aggressive or indifferent nuances. The server may also issue an alert to the user or a third party if the recognized emotion is abnormal.

[0589] (Application example 2)

[0590] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0591] In modern corporate communications, messages containing aggressive or indifferent nuances are frequently generated, which causes a bad atmosphere in the workplace. Such communications can damage employees' psychological safety and lead to reduced work efficiency and increased turnover. Therefore, there is a need for a system that can monitor chat and email exchanges within a company in real time, detect aggressive or indifferent nuances, and make appropriate improvement suggestions.

[0592] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0593] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making improvement suggestions to make the text more "kind" based on the detected nuances, a means for monitoring chat and email exchanges within the company in real time, a means for analyzing words, phrases, and contexts in the text using sentiment analysis technology, and a means for generating improvement suggestions using a generative AI model. This improves the quality of communication within the company, ensures psychological safety for employees, and enables improved business efficiency and reduced turnover.

[0594] "Text-based communication" is a method of exchanging information using written text.

[0595] "Aggressive or indifferent nuances" refer to words and expressions that show hostility or indifference toward the other person.

[0596] "Thoughtful writing" refers to expressions that take into consideration the other person's feelings and situation and show kindness and consideration.

[0597] "Intra-company chat and email communications" refers to electronic message exchanges that take place within a company.

[0598] "Real-time monitoring" means checking the contents of a message the moment it is sent or received.

[0599] "Sentiment analysis technology" is a technology that analyzes words, phrases, and context in text and estimates the emotions contained in the text.

[0600] A "generative AI model" is a model that uses artificial intelligence to generate new text and suggestions.

[0601] "Generating improvement suggestions" means creating suggestions for transforming the original text into more appropriate expressions.

[0602] A system for carrying out the present invention is configured as follows: A server includes a means for conducting text-based communication, a means for detecting offensive or insensitive nuances, a means for making suggestions for improving text to be more "kind" based on the detected nuances, a means for monitoring chat and email exchanges within a company in real time, a means for analyzing words, phrases, and contexts in the text using sentiment analysis technology, and a means for generating suggestions for improvement using a generative AI model.

[0603] Program processing explanation

[0604] The server first provides an interface for text-based communication, where users can send and receive chats and emails. The server then monitors these messages in real time to detect offensive or insensitive comments, a process that involves analyzing words, phrases and context within the text using sentiment analysis techniques.

[0605] For sentiment analysis, we use natural language processing libraries such as TextBlob. This allows us to obtain the sentiment polarity of the text (positive, negative, neutral) as a numerical value. For example, when we use TextBlob to analyze the text "Your job is no good. Do it better," we determine that the sentiment polarity is negative.

[0606] The server then uses a generative AI model (e.g., OpenAI's GPT-3) to generate improvement suggestions based on the detected nuances. In this process, the prompt sentence is input to the generative AI model to get appropriate improvement suggestions. Examples of prompt sentences include:

[0607] Please improve the following text: The emotional state is anger.

[0608] Your job is no good. Do a better job.

[0609] Improved text:

[0610] The generated improvement suggestions are presented to the user to help them refine the original text into a more thoughtful one.

[0611] Examples

[0612] For example, in a company's chat system, if an employee sends a message saying, "Your work is terrible. Do it better," the server monitors this message in real time and performs sentiment analysis using TextBlob. As a result, it is determined that this message contains negative sentiment. Next, the server uses a generative AI model to input the following prompt sentence and generate improvement suggestions:

[0613] Please improve the following text: The emotional state is anger.

[0614] Your job is no good. Do a better job.

[0615] Improved text:

[0616] The generated improvement suggestions are presented to the user in the form of text such as "There is room for improvement in your work. Please pay more attention next time." This improves the quality of communication within the company and ensures psychological safety for employees.

[0617] The flow of the identification process in the application example 2 will be described with reference to FIG.

[0618] Step 1:

[0619] The server provides an interface for text-based communication, including input fields and a send button that allow users to chat and send / receive email. Through this interface, users enter and send messages. The input is the text entered by the user, and the output is the text data sent to the server.

[0620] Step 2:

[0621] The server monitors the messages sent in real time. It receives the sent text data and starts processing to detect offensive or insensitive nuances. The input is the sent text data and the output is the text data for sentiment analysis.

[0622] Step 3:

[0623] The server uses sentiment analysis technology to analyze words, phrases, and context in the text. Specifically, it uses a natural language processing library such as TextBlob to obtain the sentiment polarity of the text (positive, negative, neutral) as a numerical value. The input is text data, and the output is numerical data indicating the sentiment polarity.

[0624] Step 4:

[0625] If the sentiment polarity is too negative or too positive, the server uses a generative AI model to generate improvement suggestions. Specifically, it uses OpenAI's GPT-3 and inputs the prompt sentence into the generative AI model. The input is the sentiment polarity and the original text data, and the output is the text data containing the improvement suggestions.

[0626] Step 5:

[0627] The server presents the generated improvement suggestions to the user. The user refers to the proposed improvement suggestions and revises the original text to make it more considerate. The input is text data including improvement suggestions, and the output is the improvement suggestions presented to the user.

[0628] Step 6:

[0629] The user corrects the original text based on the proposed improvement suggestions and resends it. This improves the quality of communication within the company and ensures psychological safety for employees. The input is the text corrected by the user, and the output is the corrected text data that is resent to the server.

[0630] Example 3

[0631] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0632] In modern text-based communication, the frequent use of aggressive or indifferent expressions is a problem. This leads to poor communication quality and makes misunderstandings and conflicts more likely to occur. In addition, there is a lack of means to properly manage users' emotional states and improve their communication skills. To solve these issues, a system is needed that can suggest more "kind" expressions and track and manage users' emotional states.

[0633] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0634] In this invention, the server includes a means for a user to input text, a means for transmitting the input text to the server, a means for the server to analyze the nuance of the text, a means for detecting aggressive or indifferent expressions based on the analysis result, a means for generating improvement suggestions for replacing the detected expressions with "kind" expressions, a means for presenting the generated improvement suggestions to the user, and a means for tracking the user's emotional state over the long term and proposing ways to deal with the situation. This allows the user to use more "kind" expressions, improves the quality of communication, and reduces misunderstandings and conflicts. It is also possible to appropriately manage the user's emotional state and support the improvement of communication skills.

[0635] A "user" is an individual or entity that utilizes the system to input text and receive improvement suggestions.

[0636] "Text" refers to a character string or sentence entered by a user and used as a means of communication.

[0637] A "server" is a computer system for receiving and analyzing user-submitted text and generating improvement suggestions.

[0638] "Nuance" refers to subtle meanings that reveal the emotion or intent contained in text.

[0639] "Analysis" refers to the process the server performs to understand the nuances of the text, often using natural language processing techniques.

[0640] "Offensive language" is any word or phrase that may offend others or provoke negative feelings.

[0641] "Insensitive speech" refers to words or phrases that show a lack of consideration or concern for others.

[0642] "Thoughtful expressions" are words and phrases that show consideration and empathy for others.

[0643] "Improvement suggestions" are suggestions for replacing aggressive or insensitive expressions with more "compassionate" expressions.

[0644] An "emotional state" is the emotion a user is feeling at a particular point in time, and may include anger, joy, sadness, etc.

[0645] "Tracking" refers to the long-term observation and recording of a user's emotional state.

[0646] A "cope" is a method or advice for a user to deal with a particular emotional state.

[0647] "Natural language processing technology" refers to technology that enables computers to understand and generate human language.

[0648] A "generative AI model" is a model for generating text using artificial intelligence, including, for example, a specific algorithm or a neural network.

[0649] This invention is a system in which a user inputs text, analyzes the nuances of the text, and improves it to a more "considerate" expression. The system is composed of three main elements: a user, a terminal, and a server.

[0650] First, a user inputs text using a terminal (e.g., a PC or smartphone). The input text is sent to the server via the terminal. At this time, the HTTPS protocol is used to ensure the security of the data.

[0651] The server passes the received text to an emotion engine, which uses natural language processing technologies such as IBM Watson or Google Cloud Natural Language API to analyze the nuances of the text and determine whether the text contains aggressive, indifferent, or other emotions.

[0652] Next, the server generates improvement suggestions based on the analysis results to replace offensive or insensitive expressions with "kind" expressions. This process uses a generative AI model (for example, OpenAI's GPT-4). The generative AI model understands the context of the input text and generates appropriate "kind" expressions.

[0653] Specifically, the server converts the text "You're always late" into an improvement suggestion: "I understand that you try hard to be on time, but it would be helpful if you could come a little earlier."

[0654] The generated improvement proposals are sent from the server to the terminal and presented to the user. The user can check the proposed improvement proposals and make corrections as necessary.

[0655] In addition, the server tracks the user's emotional state over time. The emotion engine uses this information to suggest ways to deal with situations when the user is estimated to be in an angry emotional state over a period of time. For example, if the user frequently expresses anger, the server will suggest ways to deal with the situation, such as "take a deep breath and take some time to calm down."

[0656] Examples of concrete examples and prompts

[0657] Examples

[0658] If a user types "You're always late," the device sends this text to the server. The server uses an emotion engine to analyze the nuances of the text and determines that it is an offensive expression. Next, the server uses a generative AI model to generate an improvement suggestion such as "I understand that you try hard to be on time, but it would be helpful if you could come a little earlier." Finally, the server presents this improvement suggestion to the user, who confirms it.

[0659] Examples of prompt statements

[0660] "Please improve the following text to be more considerate: 'You are always late.'"

[0661] In this way, the system analyzes the text entered by the user, generates and presents appropriate improvement suggestions, and thereby improves the quality of communication. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0662] Step 1:

[0663] The user enters text.

[0664] Input: A user uses a terminal to enter text, for example, "You are always late."

[0665] Output: The entered text is saved to the terminal.

[0666] Specific operation: A user inputs text using a keyboard on a PC or smartphone.

[0667] Step 2:

[0668] The device sends the text to the server.

[0669] Input: The text entered by the user.

[0670] Output: The text is sent to the server.

[0671] Specific operation: The terminal sends the entered text to the server using the HTTPS protocol.

[0672] Step 3:

[0673] The server analyzes the text for nuances.

[0674] Input: The text sent from the terminal.

[0675] Output: Nuance analysis of the text.

[0676] Specific behavior: The server uses an emotion engine (e.g., an API that uses natural language processing technology) to analyze the nuances of the text. As a result of the analysis, it determines whether the text contains aggressive, indifferent, or other emotions.

[0677] Step 4:

[0678] The server generates improvement suggestions based on the analysis results.

[0679] Input: Nuance analysis results for text.

[0680] Output: Improvement suggestions.

[0681] Specific behavior: The server uses a generative AI model (e.g., a generative AI model) to generate improvement suggestions to replace aggressive or insensitive expressions with "kind" expressions. For example, converting the text "You are always late" into the improvement suggestion "I understand that you try hard to be on time, but it would be helpful if you could come a little earlier."

[0682] Step 5:

[0683] The server presents the generated improvement suggestions to the user.

[0684] Input: The generated improvement suggestions.

[0685] Output: The improvement suggestions are sent to the terminal and presented to the user.

[0686] Specific operation: The server uses the HTTPS protocol to send the generated improvement suggestions to the terminal, and the terminal displays the received improvement suggestions to the user.

[0687] Step 6:

[0688] The user reviews the improvement suggestions.

[0689] Input: Improvement suggestions sent by the server.

[0690] Output: The user confirmation result.

[0691] Specific actions: The user checks the improvement suggestions displayed on the device screen and makes corrections if necessary.

[0692] Step 7:

[0693] The server tracks the user's emotional state over time.

[0694] Input: User's text input history and sentiment analysis results.

[0695] Output: Tracking data of the user's emotional state and suggestions on how to deal with it.

[0696] Specific behavior: The server uses the emotion engine to track the user's emotional state over the long term. For example, if the user is estimated to be in an angry emotional state for a certain period of time, the server will suggest countermeasures such as "take a deep breath and take some time to calm down."

[0697] (Application example 3)

[0698] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0699] In text-based communication, aggressive or indifferent nuances can make the other person feel uncomfortable. In electronic payment services, if communication with the other party is not smooth, problems and misunderstandings may occur. This poses the challenge of making it difficult for users to improve their communication skills and ensure psychological safety.

[0700] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0701] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making suggestions for improving the text to be more "considerate" based on the detected nuances, and a means for making suggestions for improving communication with a transaction partner in an electronic payment service to be more "considerate." This enables a user to have more friendly and constructive communication when using an electronic payment service.

[0702] "Text-based communication" is a method of exchanging information using written text.

[0703] "Aggressive or indifferent nuances" refer to expressions or attitudes that show hostility or indifference toward the other person.

[0704] "Thoughtful writing" is writing that takes into consideration the other person's feelings and position and uses friendly and considerate language.

[0705] An "electronic payment service" is a service that enables monetary transactions via the Internet or mobile devices.

[0706] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[0707] "Communication skills" are the ability to effectively convey information and understand others.

[0708] "Psychological safety" refers to an environment in which individuals feel safe to express their opinions and ask questions.

[0709] "Transaction partner" refers to a party with whom money, goods, or services are exchanged in an electronic payment service.

[0710] "Friendly and constructive communication" is communication that shows a friendly attitude toward the other person and promotes problem solving and cooperation.

[0711] A system for implementing this invention includes means for conducting text-based communication, means for detecting aggressive or indifferent nuances, means for making suggestions for improving communication to make it more "considerate" based on the detected nuances, and means for making suggestions for improving communication with a transaction partner in an electronic payment service to make it more "considerate."

[0712] The server receives text-based communications from users and analyzes their content. The analysis uses natural language processing techniques. Specifically, it uses OpenAI's generative AI models (e.g., GPT-3) to understand the meaning of the text and detect offensive or insensitive nuances.

[0713] Based on the detected nuances, the server generates suggestions for improving "kind" text, which will help users to communicate more friendly and constructively with their counterparts when using electronic payment services.

[0714] If the message entered by the user is "You're wrong. Please refund me.", the server will analyze it and detect the offensive nuance. Then, using a generative AI model, it will generate suggestions for improving the message to be more "kind" like the following:

[0715] "Thank you for your help. I've had a slight problem with the transaction and would like to discuss a refund."

[0716] In this way, users can have a more friendly and constructive communication.

[0717] The hardware used is a server and the user's smartphone. The software uses Python and the OpenAI API. Data processing and data calculation involves using generative AI models to analyze the meaning of the text and generate appropriate improvement suggestions.

[0718] An example of a prompt is as follows:

[0719] Please improve the following sentences to make them more considerate:

[0720] You are wrong. Please refund my money.

[0721] Improved sentence:

[0722] The system will enable users to have more friendly and constructive communication when using electronic payment services.

[0723] The flow of the specific process in the application example 3 will be described with reference to FIG.

[0724] Step 1:

[0725] A user inputs a text-based message using a smartphone, and the input message is sent to a server.

[0726] Input: A text message entered by the user.

[0727] Output: The text message sent to the server.

[0728] Step 2:

[0729] The server analyzes the received text message. It uses natural language processing technology to understand the meaning of the text. Specifically, it uses a generative AI model (e.g. GPT-3) to analyze the content of the text.

[0730] Input: The text message sent to the server.

[0731] Output: Semantic information of the parsed text

[0732] Step 3:

[0733] The server detects offensive or insensitive nuances based on the semantic information of the analyzed text. It uses a generative AI model to evaluate the nuances of the text and determine whether it contains offensive or insensitive expressions.

[0734] Input: Semantic information of the parsed text

[0735] Output: Nuance detection result (aggressive, indifferent, etc.)

[0736] Step 4:

[0737] The server generates improvement suggestions for "kind" sentences based on the detected nuances. Using a generative AI model, it generates appropriate improvement suggestions and presents them to the user.

[0738] Input: Nuance detection result

[0739] Output: Suggestions for improving the text to be more "kind"

[0740] Step 5:

[0741] The server sends the generated improvement proposals to the user's smartphone, where the user can check the proposals and make corrections if necessary.

[0742] Input: Suggestions for improving the wording to be more "kind"

[0743] Output: Improvement suggestions sent to the user's smartphone

[0744] Step 6:

[0745] The user reviews the suggested improvements, makes any necessary corrections, and then sends the final message to the trading partner.

[0746] Input: The final message that the user has confirmed and modified.

[0747] Output: The final message sent to the trading partner

[0748] In this way, users can have more friendly and constructive communication when using electronic payment services.

[0749] 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 a voice indicating a user input for the result of the specific processing. The control unit 46A transmits the voice 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 voice data.

[0750] 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<URL: https: / / openai.com / blog / chatgpt> ) and other generative AIs. Data generation model 58 is

[0751] , obtained by making a neural network perform deep learning. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image is input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0752] Another example of generative AI is Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) are mentioned.

[0753] In the above embodiment, an example was given in which the specific process was performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0754] [Second embodiment]

[0755] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0756] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0757] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a wide area network (WAN) and / or a local area network (LAN).

[0758] 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.

[0759] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs the voice according to instructions from the processor 46.

[0760] Camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (e.g., an imaging range defined by an angle of view equivalent to the width of the field of vision of an average healthy person).

[0761] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54. The transmission and reception of various types of information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is performed in a secure state.

[0762] Fig. 4 shows an example of 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.

[0763] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32, and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0764] 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.

[0765] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50, and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0766] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0767] "Example 1"

[0768] In one embodiment of the present invention, a means for text-based communication may be email, a messenger app, etc. When a user communicates using these means, the system receives the text and uses natural language processing technology to understand the meaning of the text.

[0769] "Example 2"

[0770] To detect offensive or insensitive nuances, the system uses sentiment analysis technology. Specifically, it analyzes words, phrases, and contexts in a text to determine whether the text contains offensive or insensitive nuances.

[0771] "Example 3"

[0772] As a means of making suggestions to improve sentences to be more "compassionate," the system generates improvement suggestions based on the detected nuances. Specifically, the system makes suggestions to replace aggressive or indifferent expressions with "compassionate" expressions. For example, the system makes suggestions to improve the expression "You are wrong" to a more "compassionate" expression such as "I understand your point of view, but I think from a slightly different perspective."

[0773] The process flow of each embodiment will be described below.

[0774] "Example 1"

[0775] Step 1: The user engages in text-based communication via email, messenger app, etc.

[0776] Step 2: The system receives the text and uses natural language processing techniques to understand the meaning of the text.

[0777] "Example 2"

[0778] Step 1: The system receives the text.

[0779] Step 2: Using sentiment analysis techniques, analyze the words, phrases, and context in the text. Step 3: Determine whether the text contains offensive or insensitive nuances.

[0780] "Example 3"

[0781] Step 1: If offensive or insensitive nuances are detected, the system generates improvement suggestions.

[0782] Step 2: Make suggestions to replace aggressive or insensitive expressions with more "compassionate" ones. Step 3: Make suggestions to improve the expression, for example, "You are wrong" to "I understand your point of view, but I think from a slightly different perspective."

[0783] Example 1

[0784] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[0785] In conventional text-based communication systems, it was difficult to properly understand the nuances of messages sent by users, detect aggressive or indifferent expressions, and make suggestions for improvement. In addition, there was a lack of technology to generate appropriate responses to messages sent by users. This made it difficult to improve users' communication skills and realize considerate communication.

[0786] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0787] In this invention, the server includes a means for a user to send a message, a means for a terminal to send a message to the server, a means for the server to receive the message, a means for the server to preprocess the message, a means for the server to semantically analyze the message, a means for the server to generate a response, a means for the server to send the response to the terminal, and a means for the terminal to display the response to the user. This makes it possible to properly understand the nuances of messages sent by users, detect aggressive or indifferent expressions, and make suggestions for improvement. In addition, it is possible to generate appropriate responses to user messages, improve the user's communication skills, and realize considerate communication.

[0788] "Text-based communication" refers to the exchange of information using written text, including email and messenger apps.

[0789] "Aggressive or indifferent nuances" refer to expressions or attitudes that show hostility or indifference toward the other person.

[0790] "Thoughtful writing" refers to writing that takes into consideration the other person's feelings and situation and includes expressions that show consideration and kindness.

[0791] "Means for a user to send a message" refers to an interface or function that allows a user to input and send text.

[0792] "Means for a terminal to send a message to a server" refers to a function of a user's terminal to send a message to a server via the Internet.

[0793] "Means for the server to receive a message" refers to the function of the server to receive a message sent from a terminal.

[0794] "Means for the server to preprocess the message" refers to a function for performing processing to convert the received message into a form that is easy to analyze.

[0795] "Means for the server to semantically analyze the message" refers to the functionality for understanding the meaning of the pre-processed message and extracting important information.

[0796] "Means for the server to generate a response" refers to the functionality for generating an appropriate response based on the analysis results.

[0797] The term "means for the server to send a response to the terminal" refers to a function for sending the generated response to the user's terminal.

[0798] "Means for the terminal to display the response to the user" refers to a function for visually displaying the response received from the server to the user.

[0799] The object of this invention is to provide a system for text-based communication that properly understands the nuances of messages sent by users, detects offensive or indifferent expressions, makes suggestions for improvement, and generates appropriate responses.

[0800] System program generation

[0801] First, a user sends a message using a text-based communication method (e.g., email or a messenger app). The device sends this message to the server. The server preprocesses the received message, performs semantic analysis, generates an appropriate response, and sends it to the device. The device displays the received response to the user.

[0802] Hardware and Software Used

[0803] The system uses the following hardware and software:

[0804] Hardware: User's device (smartphone, PC, etc.), server

[0805] Software: Natural language processing libraries (e.g., Python's "NLTK" or "spaCy"), generative AI models (e.g., OpenAI's "GPT-3")

[0806] Data processing and calculation

[0807] The server performs the following data processing and calculations:

[0808] 1. Text preprocessing: The server tokenizes the received message and removes unnecessary symbols and stop words. For this process, we use "NLTK" and "spaCy".

[0809] 2. Semantic Analysis: The server analyzes the semantics of the preprocessed text and extracts keywords and important phrases using natural language processing models such as "spaCy" or "BERT".

[0810] 3. Response Generation: The server generates an appropriate response based on the analysis results. This response generation uses a generative AI model (e.g., GPT-3).

[0811] Examples

[0812] As a concrete example, let's say a user uses a messenger app to send the following message:

[0813] "What time does the meeting start tomorrow?"

[0814] The server receives this message and processes it as follows:

[0815] 1. Text preprocessing: The server tokenizes the text “What time does the meeting start tomorrow?” and breaks it down into the words “tomorrow,” “meeting,” “what time,” “start,” and “is.”

[0816] 2. Semantic analysis: The server extracts the keywords "meeting" and "what time" and understands that the user is asking about the start time of the meeting.

[0817] 3. Response generation: The server uses a generative AI model (e.g., GPT-3) to generate an appropriate response. It inputs the prompt "Message from user: What time does tomorrow's meeting start? Please generate a response." and generates the response "Tomorrow's meeting starts at 10:00 a.m."

[0818] In this manner, a system is implemented in which the server analyzes a user's message and generates and delivers an appropriate response to the user.

[0819] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0820] Step 1:

[0821] A user sends a message.

[0822] Specific actions: A user opens a messenger app, types "What time is the meeting tomorrow?", and hits send.

[0823] Input: The text message entered by the user.

[0824] Output: The instruction to send is sent to the terminal.

[0825] Step 2:

[0826] The terminal sends a message to the server.

[0827] Specific operation: The terminal transmits the entered message to a server via the Internet.

[0828] Input: The text message entered by the user.

[0829] Output: A message is sent to the server.

[0830] Step 3:

[0831] The server receives the message.

[0832] Specific operation: The server receives the message sent from the terminal and stores it in a database.

[0833] Input: The text message sent from the device.

[0834] Output: The received message is stored in the database.

[0835] Step 4:

[0836] The server pre-processes the message.

[0837] Specific operation: The server tokenizes the received message and removes unnecessary symbols and stop words. This process is done using "NLTK" and "spaCy".

[0838] Input: The received text message.

[0839] Output: Preprocessed text data.

[0840] Step 5:

[0841] The server semantically parses the message.

[0842] How it works: The server analyzes the preprocessed text and extracts keywords and important phrases using natural language processing models such as spaCy and BERT.

[0843] Input: Preprocessed text data.

[0844] Output: Parsed keywords or phrases.

[0845] Step 6:

[0846] The server generates a response.

[0847] Specific operation: The server generates an appropriate response based on the analysis results. A generative AI model (e.g., GPT-3) is used to generate this response. The server inputs the prompt text "Message from user: What time does the meeting start tomorrow? Please generate a response." and generates a response.

[0848] Input: The parsed keyword or phrase.

[0849] Output: The generated response text.

[0850] Step 7:

[0851] The server sends a response to the terminal.

[0852] Specific operation: The server sends the generated response to the user's terminal via the Internet.

[0853] Input: The generated response text.

[0854] Output: The response is sent to the terminal.

[0855] Step 8:

[0856] The terminal displays the response to the user.

[0857] Specific operation: The terminal visually displays the response received from the server to the user. The user confirms the displayed response.

[0858] Input: The response text sent by the server.

[0859] Output: The response text that is displayed to the user.

[0860] (Application example 1)

[0861] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[0862] Conventional text-based communication systems can detect the nuances in the text sent by users and make appropriate suggestions for improvement, but they lack the functionality to recommend related content based on the user's interests and concerns. This limits the information and knowledge that users can obtain through communication, making it difficult to improve the quality of communication.

[0863] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0864] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for suggesting improvements to make the text more "kind" based on the detected nuances, and a means for analyzing the user's text and recommending related content based on the content of the text. This allows the user to not only improve the quality of communication, but also to obtain related content based on their interests and concerns.

[0865] "Text-based communication" is any means by which users exchange information using written text.

[0866] "Aggressive or indifferent undertones" are expressions contained in the text that indicate negative or indifferent feelings or attitudes.

[0867] "Thoughtful writing" is text that includes expressions that show consideration and kindness to the other person.

[0868] "Improvement Suggestions" are suggested modifications to make the content of the text a user sends more appropriate and considerate.

[0869] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language.

[0870] "Related content" is information such as articles, videos, and music that is recommended based on the content of the user's text.

[0871] A "recommendation method" is a method that analyzes a user's text and presents relevant content based on that content.

[0872] A system for implementing the present invention includes means for conducting text-based communication, means for detecting aggressive or insensitive nuances in the text-based communication, means for suggesting improvements to make the text more "kind" based on the detected nuances, and means for analyzing a user's text and recommending related content based on the content of the text.

[0873] The server receives the text sent by the user and uses natural language processing technology to analyze the meaning of the text. Specifically, it uses OpenAI's GPT-3 model to understand the content of the text and detect offensive or insensitive nuances. Based on the detected nuances, the server generates improvement suggestions for "kind" text.

[0874] In addition, the server recommends related content based on the content of the user's text. For example, it uses a news API to retrieve articles related to the user's interests and presents them to the user. This series of processes is implemented using a Python program.

[0875] As a concrete example, if a user sends a message saying, "I would like to know about recent technological advances," the server will input the following prompt sentence into the generative AI model:

[0876] Please recommend related content based on the content of the text below:

[0877] I want to know about the latest advances in technology

[0878] Based on this prompt, the generative AI model analyzes the meaning of the text and recommends relevant content. Article titles and URLs retrieved using a news API are presented to the user.

[0879] The hardware used is a smartphone, and the software used is Python, OpenAI API, and news API. This not only improves the quality of communication, but also allows users to obtain relevant content based on their interests.

[0880] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0881] Step 1:

[0882] A user sends a message using a text-based communication method (e.g. email or a messenger app). The input is the text the user sends, and the output is that text being sent to the server.

[0883] Step 2:

[0884] The server uses natural language processing techniques to analyze the received text. Specifically, it uses OpenAI's GPT-3 model to understand the meaning of the text. The input is the text submitted by the user, and the output is the result of the semantic analysis of the text.

[0885] Step 3:

[0886] The server detects offensive or insensitive nuances in the text based on the analysis results. The input is the result of semantic analysis of the text, and the output is the detected nuance information.

[0887] Step 4:

[0888] The server generates improvement suggestions for "kind" sentences based on the detected nuances. The input is the nuance information, and the output is the text of the improvement suggestions.

[0889] Step 5:

[0890] The server recommends related content based on the content of the user's text. Specifically, it uses a news API to get related articles. The input is the user's text and the analysis results, and the output is a list of related articles.

[0891] Step 6:

[0892] The server presents the retrieved related content to the user. The input is a list of related articles, and the output is the titles and URLs of the articles presented to the user.

[0893] Step 7:

[0894] The user reviews the presented related content and browses it as needed. The input is the title and URL of the presented article, and the output is the user's browsing behavior.

[0895] Example 2

[0896] Next, a description will be given of Example 2 of the embodiment 2. In the following description, the data processing device 12 is referred to as a "server" and the smart glasses 214 are referred to as a "terminal".

[0897] When text-based communication contains offensive or insensitive nuances, it can lead to misunderstandings and conflicts. Users may also unintentionally use offensive language, lowering the quality of communication. There is a need to solve these problems and enable users to communicate more considerately.

[0898] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0899] In this invention, the server includes a means for performing text-based communication, a means for preprocessing text data, a means for performing sentiment analysis using a generative AI model, a means for outputting the results of the sentiment analysis, and a means for making suggestions for improving the text to be more "kind" based on the detected nuances. This makes it possible to detect aggressive or indifferent nuances in the text data entered by the user and make appropriate suggestions for improvement.

[0900] "Text-based communication" is a method of exchanging information using written text.

[0901] "Aggressive or indifferent nuances" refer to expressions or attitudes that show hostility or indifference toward others.

[0902] "Suggestions for improving language to be more considerate" are suggestions for changing text that contains aggressive or insensitive nuances into more friendly and considerate language.

[0903] "Text data preprocessing" refers to the process of removing unnecessary spaces and special characters from text data and dividing it into words and phrases.

[0904] A "generative AI model" is a model that uses artificial intelligence technology to analyze the sentiment and meaning of text.

[0905] "Sentiment analysis" is the process of analyzing the emotions and nuances contained in text data to determine whether they are aggressive, indifferent, or some other emotion.

[0906] "Output of sentiment analysis results" refers to providing the results obtained by sentiment analysis to a user.

[0907] The present invention is a system that detects aggressive or indifferent nuances in text-based communication and suggests improvements to make the text more considerate. A specific embodiment of this system will be described below.

[0908] System Configuration

[0909] Hardware and software configuration:

[0910] Server: A computer with high-performance computing power that performs data pre-processing, sentiment analysis, and outputting the results.

[0911] Terminal: A device (e.g., PC, smartphone) through which a user enters text data and receives the results.

[0912] Generative AI model: An artificial intelligence model for performing sentiment analysis (e.g., OpenAI's GPT-3).

[0913] Natural Language Processing library: Software used to preprocess text data (e.g., NLTK in Python).

[0914] Program Processing

[0915] Enter data:

[0916] The user uses the device to input text data to be analyzed, such as comments on social media or the contents of emails.

[0917] Data preprocessing:

[0918] The server preprocesses the received text data by removing unnecessary spaces and special characters and tokenizing the text using the Python NLTK (Natural Language Toolkit) library.

[0919] Sentiment analysis:

[0920] The server performs sentiment analysis on the pre-processed text data, using a generative AI model (e.g., OpenAI's GPT-3) to determine whether the text contains offensive or insensitive nuances.

[0921] The resulting output:

[0922] The server returns the results of the sentiment analysis to the user, indicating whether the text contains aggressive, indifferent, or other sentiments.

[0923] Examples

[0924] Let's say a user wants to analyze the social media comment "I don't care about you."

[0925] 1. Enter your data:

[0926] The user types "I don't care about you" into an input field on the terminal.

[0927] The terminal transmits this text data to the server.

[0928] 2. Data preprocessing:

[0929] The server preprocesses the received text data "I don't care about you."

[0930] Remove unnecessary spaces and special characters.

[0931] Tokenize the text and split it into "you," "something," and "it doesn't matter."

[0932] 3. Sentiment analysis:

[0933] The server creates a prompt sentence to input the preprocessed text data into the generative AI model:

[0934] "Please determine whether the following text contains offensive or insensitive connotations:

[0935] "I don't care about you."

[0936] A generative AI model takes this prompt and analyzes the sentiment of the text.

[0937] The generative AI model determines that the text is "offensive" and returns that result to the server.

[0938] 4. Result output:

[0939] The server returns the result it received from the generative AI model, “aggressive,” to the user.

[0940] The device displays the result to the user, for example, "This comment is offensive."

[0941] The above is a specific embodiment for carrying out the present invention. This system allows users to detect offensive or indifferent nuances contained in text data and receive appropriate suggestions for improvement.

[0942] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0943] Step 1:

[0944] Entering data

[0945] The user inputs the text data to be analyzed using the device. For example, the user inputs comments on social media or the contents of an email. The device then transmits the input text data to the server.

[0946] Input: Text data entered by the user (e.g. "I don't care about you")

[0947] Output: Text data sent from the terminal to the server

[0948] Step 2:

[0949] Data Preprocessing

[0950] The server preprocesses the received text data by removing unnecessary spaces and special characters and tokenizing the text using the Python NLTK (Natural Language Toolkit) library.

[0951] Input: Text data sent from the terminal

[0952] Data processing: removing unnecessary spaces and special characters, tokenizing text

[0953] Output: Preprocessed text data (e.g. "you", "something", "it doesn't matter")

[0954] Step 3:

[0955] sentiment analysis

[0956] The server performs sentiment analysis on the preprocessed text data, using a generative AI model (e.g., OpenAI's GPT-3) to determine whether the text contains offensive or insensitive nuances. The server inputs the following prompt sentence into the generative AI model:

[0957] "Please determine whether the following text contains offensive or insensitive connotations:

[0958] "I don't care about you."

[0959] A generative AI model takes this prompt and analyzes the sentiment of the text.

[0960] Input: Preprocessed text data, prompt

[0961] Data Computation: Sentiment Analysis with Generative AI Models

[0962] Output: Result of sentiment analysis (e.g. "offensive")

[0963] Step 4:

[0964] Output of results

[0965] The server returns the result of the sentiment analysis to the user, indicating whether the text contains offensive, indifferent, or other sentiments. The terminal displays the result received from the server to the user, for example, "This comment is offensive."

[0966] Input: Sentiment analysis results

[0967] Output: The result that is displayed to the user (e.g. "This comment is offensive")

[0968] The above is the specific processing flow of the program of this system.

[0969] (Application example 2)

[0970] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server", and the smart glasses 214 will be referred to as a "terminal".

[0971] In modern text-based communication, messages containing aggressive or insensitive nuances are frequently sent and received, which can lead to deterioration of relationships between users and loss of psychological safety. In addition, there is a lack of means to detect messages containing these nuances in real time and warn users, making it difficult to improve the quality of communication.

[0972] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0973] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for suggesting improvements to make the text more "considerate" based on the detected nuances, and a means for analyzing messages in real time, detecting messages containing aggressive or indifferent nuances, and issuing a warning to the user. This allows the user to receive a warning before sending a message containing aggressive or indifferent nuances, thereby improving the quality of communication.

[0974] "Text-based communication" is a method of exchanging information using written text.

[0975] An "aggressive nuance" is a nuance that includes words or expressions that show hostility or aggression toward the other person.

[0976] An "indifferent nuance" is a nuance that includes words and expressions that show no interest or concern for the other person.

[0977] "Thoughtful writing" is writing that takes into consideration the other person's feelings and situation and includes words and expressions that show consideration and kindness.

[0978] "Analyzing messages in real time" means analyzing the content of a message immediately at the moment it is sent.

[0979] "Warning the user" means notifying the user about a detected message that contains offensive or insensitive nuances.

[0980] "Natural language processing technology" is a technology that enables computers to process and understand human language.

[0981] "Sentiment analysis" is a technology that analyzes emotions and nuances in text and classifies and evaluates its content.

[0982] In order to implement the present invention, the following system configuration and processing procedures are required.

[0983] System Configuration

[0984] 1. Hardware:

[0985] Smartphone: A device that allows users to engage in text-based communication.

[0986] Server: Provides the computational resources for sentiment analysis and nuance detection.

[0987] 2. Software:

[0988] Python: A programming language for running programs for sentiment analysis.

[0989] NLTK (Natural Language Toolkit): A library for natural language processing.

[0990] VADER (Valence Aware Dictionary and sEntiment Reasoner): A model for performing sentiment analysis.

[0991] Processing Steps

[0992] 1. Data Entry:

[0993] A user enters a text-based message using a smartphone.

[0994] 2. Data preprocessing:

[0995] The smartphone sends the entered message in text format to the server.

[0996] 3. Sentiment analysis:

[0997] The server uses NLTK's VADER sentiment analysis model to calculate a sentiment score for each received message.

[0998] 4. Nuance detection:

[0999] The server uses the sentiment score to determine whether the message contains offensive or insensitive overtones.

[1000] 5. Warning Issued:

[1001] Based on the detection results, the server will display a warning to the user about messages that contain offensive or insensitive nuances.

[1002] Examples

[1003] For example, if a user tries to send a message saying "You're really incompetent," the smartphone will send this message to the server. The server will use NLTK's VADER sentiment analysis model to calculate the sentiment score of this message and determine that it contains offensive nuances. As a result, the server will display a warning to the user saying "Offensive message detected."

[1004] Examples of prompt statements

[1005] Here is an example of a prompt to be input to the generative AI model:

[1006] Please judge whether the following messages contain offensive or insensitive connotations:

[1007] Message: "You're really incompetent."

[1008] In this way, users can be warned before sending messages that may be offensive or insensitive, improving the quality of communication.

[1009] The flow of the specific process in the application example 2 will be described with reference to FIG.

[1010] Step 1:

[1011] A user enters a text-based message using a smartphone.

[1012] Input: The text message entered by the user.

[1013] Output: The typed text message displayed on the smartphone.

[1014] Specific behavior: A user opens a chat application and enters text into the message input field.

[1015] Step 2:

[1016] The terminal transmits the input message in text format to the server.

[1017] Input: The text message entered by the user.

[1018] Output: The text message sent to the server.

[1019] Specific operation: The smartphone application sends the entered text message to the server as an HTTP request.

[1020] Step 3:

[1021] The server uses NLTK's VADER sentiment analysis model to calculate a sentiment score for each received message.

[1022] Input: The text message sent to the server.

[1023] Output: Sentiment scores (positive, negative and neutral scores).

[1024] Specific operation: The server inputs the received text message into the VADER sentiment analysis model and calculates positive, negative, and neutral scores.

[1025] Step 4:

[1026] The server uses the sentiment score to determine whether the message contains offensive or insensitive overtones.

[1027] Input: Sentiment score.

[1028] Output: Nuance judgment result (aggressive, indifferent, normal).

[1029] Specific operation: The server analyzes the emotion score, and if the negative score is high, it is determined to be "aggressive," if the neutral score is high, it is determined to be "indifferent," and otherwise it is determined to be "normal."

[1030] Step 5:

[1031] Based on the detection results, the server will display a warning to the user about messages that contain offensive or insensitive nuances.

[1032] Input: Nuance judgment result.

[1033] Output: The warning message that is displayed to the user.

[1034] Specific operation: The server sends the judgment result to the smartphone, and the smartphone application displays a warning to the user such as "Offensive message has been detected."

[1035] Example 3

[1036] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[1037] In text-based communication, aggressive or indifferent nuances are often included, which can lead to misunderstandings and conflicts. In addition, there is a lack of concrete ways for users to improve their expressions to be more "considerate." This makes it difficult for users to communicate smoothly and considerately.

[1038] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1039] In this invention, the server includes a means for a user to input text, a means for a terminal to transmit the input text to the server, a means for the server to analyze the nuances of the text, a means for the server to generate improvement suggestions using a generative AI model, a means for the server to transmit the improvement suggestions to the terminal, and a means for the terminal to display the improvement suggestions to the user. This enables the user to receive specific suggestions for improving aggressive or insensitive expressions to "considerate" expressions.

[1040] "Text-based communication" is a method of exchanging information using written text.

[1041] "Nuance" refers to the subtle meanings and emotions contained in text.

[1042] "Offensive language" is language that has the potential to offend others or provoke negative feelings.

[1043] "Indifferent expressions" are language that makes the other person feel that you lack interest or consideration.

[1044] "Considerate expression" is language that shows consideration for the other person's feelings and position and shows kindness and understanding.

[1045] "User" means any individual or entity that uses the System.

[1046] A "Terminal" is a device through which a User accesses the System.

[1047] A "server" is a computer system that receives and processes requests from users.

[1048] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[1049] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate new text or suggestions.

[1050] "Improvement suggestions" are specific suggestions for replacing aggressive or insensitive language with more "compassionate" language.

[1051] The present invention is a system that detects aggressive or indifferent nuances in text-based communication and suggests ways to improve the text to be more “considerate.” A specific embodiment of this system will be described below.

[1052] First, the user inputs text using the terminal. For example, the user inputs the text "You are wrong." The terminal transmits the input text to the server. For transmission, a communication protocol such as an HTTP request is used.

[1053] The server analyzes the received text using natural language processing (NLP) techniques, specifically using NLP tools such as Python's NLTK library and spaCy to identify offensive or insensitive content within the text.

[1054] Next, the server uses a generative AI model to generate improvement suggestions based on the analysis results. For example, OpenAI's GPT-4 is used as the generative AI model. The server inputs the following prompt sentence to the generative AI model:

[1055] Improve your statement from "You're wrong" to something more "compassionate."

[1056] The generative AI model generates an improvement suggestion such as, "I understand your point of view, but I'm thinking about it from a slightly different perspective." The server sends this improvement suggestion to the terminal. For sending, a communication protocol such as HTTP response is used.

[1057] The terminal displays the received improvement suggestions to the user. The user can check the displayed improvement suggestions and modify the text as necessary. For example, the user can use the displayed improvement suggestions as they are or can make further modifications.

[1058] This system allows users to receive specific suggestions for improving aggressive or indifferent expressions into "kind" expressions. This allows smooth and kind communication to be realized. The flow of the identification process in the third embodiment will be described with reference to FIG. 15.

[1059] Step 1:

[1060] The user enters text.

[1061] The user inputs the text to be improved in the input field of the terminal, for example, "You are wrong." The input text is temporarily stored in the terminal's memory.

[1062] Step 2:

[1063] The terminal sends the entered text to the server.

[1064] The terminal sends the entered text as an HTTP request to the server. The input data is packaged in JSON format and sent over the network to the server. For example, the following JSON data is sent:

[1065] json

[1066] {

[1067] "text": "You are wrong"

[1068] }

[1069] Step 3:

[1070] The server analyzes the text for nuances.

[1071] The server analyzes the received text using natural language processing (NLP) techniques. Specifically, it uses NLP tools such as Python's NLTK library and spaCy to identify offensive or insensitive expressions in the text. The input data is the text, and the output data is nuanced information as a result of the analysis. For example, the part that says "wrong" is determined to be offensive.

[1072] Step 4:

[1073] The server generates improvement suggestions using the generative AI model.

[1074] The server uses a generative AI model (e.g., OpenAI's GPT-4) based on the analysis results to generate improvement suggestions for more "compassionate" expressions. The server inputs the following prompt sentence into the generative AI model:

[1075] Improve your statement from "You're wrong" to something more "compassionate."

[1076] The generative AI model generates improvement suggestions such as, "I understand your point of view, but I think about it from a slightly different perspective." The input data is the prompt sentence, and the output data is the improvement suggestions.

[1077] Step 5:

[1078] The server sends improvement suggestions to the device.

[1079] The server sends the generated improvement proposal to the terminal as an HTTP response. The output data is packaged in JSON format and sent to the terminal via the network. For example, the following JSON data is sent:

[1080] json

[1081] {

[1082] "suggestion": "I understand your point of view, but I think about it from a slightly different perspective."

[1083] }

[1084] Step 6:

[1085] The terminal displays improvement suggestions to the user.

[1086] The terminal displays the improvement suggestions received from the server to the user. The user can check the displayed improvement suggestions and modify the text as necessary. For example, the user can use the displayed improvement suggestions as they are or modify them further. The input data is the improvement suggestions, and the output data is the text displayed to the user.

[1087] (Application example 3)

[1088] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server", and the smart glasses 214 will be referred to as a "terminal".

[1089] Text-based communication often contains aggressive or insensitive expressions, which can lead to misunderstandings and conflicts. Users may also unintentionally use expressions that hurt others. This leads to problems in that the quality of communication declines and psychological safety is lost. In addition, there is a lack of ways for users to receive concrete feedback to improve their expressions, which makes it difficult to improve communication skills.

[1090] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1091] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making improvement suggestions to make the text more "kind" based on the detected nuances, a means for analyzing messages sent by users in real time and displaying improvement suggestions, and a means for generating improvement suggestions using a generative AI model. This allows users to receive specific feedback for improving their own expressions in real time, improving the quality of communication and ensuring psychological safety.

[1092] "Text-based communication" is a method of exchanging information using written text.

[1093] "Nuance" refers to the subtle meanings and emotions contained in words and expressions.

[1094] "Offensive language" is any word or phrase intended to injure or provoke another person.

[1095] "Indifferent expressions" are cold words or phrases that show no interest in the other person's feelings or opinions.

[1096] "Thoughtful writing" is expression that shows respect and consideration for the other person's feelings and position.

[1097] "Improvement suggestions" are specific advice or proposed revisions to improve the current expression.

[1098] "Real-time analysis" means that the content of a message is analyzed immediately as the user types it.

[1099] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate new data or information.

[1100] "Feedback" means evaluation or advice provided to a user in response to their actions or expressions.

[1101] "Psychological safety" refers to an environment in which individuals feel safe to express their opinions and take action.

[1102] A system for implementing this invention includes means for conducting text-based communication, means for detecting aggressive or insensitive nuances, means for making suggestions for improvements to make the text more "kind" based on the detected nuances, means for analyzing messages sent by users in real time and displaying suggestions for improvement, and means for generating suggestions for improvement using a generative AI model.

[1103] Hardware and Software Configuration

[1104] Hardware:

[1105] Smartphone (iOS or Android)

[1106] software:

[1107] Python

[1108] OpenAI API

[1109] Data processing and calculation

[1110] server:

[1111] 1. A user types a text message using a smartphone.

[1112] 2. The entered message is sent to the server.

[1113] 3. The server analyzes the received message and detects offensive or insensitive nuances.

[1114] 4. Based on the detected nuances, a generative AI model (e.g., OpenAI’s GPT-3) is used to generate improvement suggestions for making the text more “kind.”

[1115] 5. Improvement suggestions are displayed in real time on the user's smartphone.

[1116] Examples

[1117] When a user types "You're wrong," the server analyzes the message to detect offensive nuances, then uses a generative AI model to generate a suggestion for improvement, such as "I understand your point of view, but I think about it from a slightly different perspective," and displays it to the user.

[1118] Example prompt for a generative AI model:

[1119] Please improve the following message to make it more considerate: You are wrong

[1120] In this way, users can get specific feedback to improve their expression in real time, which improves the quality of communication and ensures psychological safety.

[1121] The flow of the specific process in the application example 3 will be described with reference to FIG.

[1122] Step 1:

[1123] A user enters a text message using a smartphone.

[1124] Input: A text message entered by the user (e.g. "You are wrong")

[1125] Output: The entered text message is sent to the server by pressing the send button on the smartphone.

[1126] Step 2:

[1127] The server analyzes the messages it receives and detects offensive or indifferent nuances.

[1128] Input: Text message sent from smartphone

[1129] Data processing: Using natural language processing techniques, we analyze the content of messages to detect offensive or insensitive nuances.

[1130] Output: Detected nuance information (e.g. "contains offensive language")

[1131] Step 3:

[1132] Based on the nuances detected, the server uses a generative AI model to generate improvement suggestions for making the text more "considerate."

[1133] Input: Detected nuance information and original text message

[1134] Data computation: Send prompts to a generative AI model (e.g., OpenAI’s GPT-3) to generate improvement suggestions.

[1135] Output: Text of the suggestion for improvement (e.g. "I understand your point of view, but I have a slightly different perspective")

[1136] Step 4:

[1137] The server displays the generated improvement suggestions on the user's smartphone in real time.

[1138] Input: Text of improvement suggestion

[1139] Data processing: The improvement suggestions are converted into a data format for sending to the user's smartphone.

[1140] Output: Suggested improvements displayed on the user's smartphone (e.g., "I understand your point of view, but I have a slightly different perspective")

[1141] Step 5:

[1142] The user reviews the improvement suggestions and, if necessary, submits a revised message.

[1143] Input: User's correction message after reviewing the improvement suggestions

[1144] Output: The corrected text message is finally sent.

[1145] In this way, users can get specific feedback to improve their expression in real time, which improves the quality of communication and ensures psychological safety.

[1146] In addition, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1147] "Example 1"

[1148] In one embodiment of the present invention, the system includes an emotion engine that recognizes the emotion of a user. The emotion engine infers the user's emotional state from the text when the user engages in text-based communication. For example, if the user enters the text "I am angry," the emotion engine infers that the user is in an angry emotional state.

[1149] "Example 2"

[1150] In one embodiment of the present invention, the emotion engine adjusts the remediation suggestions based on the estimated emotional state. For example, if the user is estimated to be in an angry emotional state, the emotion engine adjusts the remediation suggestions for "kind" text to appropriately express the anger, thereby respecting the user's emotional state and reducing aggressive or indifferent overtones.

[1151] "Example 3"

[1152] In one embodiment of the present invention, the emotion engine tracks a user's emotional state over time to help improve the user's communication skills. For example, if the user is estimated to be in an angry emotional state over a period of time, the emotion engine uses that information to understand what makes the user angry and suggests ways to address it, thereby helping the user improve their communication skills.

[1153] The process flow of each embodiment will be described below.

[1154] "Example 1"

[1155] Step 1: Users engage in text-based communication.

[1156] Step 2: The emotion engine infers the emotional state of the user from the text. For example, from the text "I am angry", it infers that the user is in an angry emotional state.

[1157] "Example 2"

[1158] Step 1: The emotion engine estimates the user's emotional state.

[1159] Step 2: The system adjusts the remediation suggestions based on the estimated emotional state. For example, if the user is estimated to be in an angry emotional state, the system will adjust the remediation suggestions for "kind" sentences to better express the anger.

[1160] "Example 3"

[1161] Step 1: The emotion engine estimates the user's emotional state and records that information.

[1162] Step 2: The emotion engine tracks the user's emotional state over time.

[1163] Step 3: Based on the long-term emotional state tracking results, the system helps users improve their communication skills. For example, if the user is estimated to be in an angry emotional state for a certain period of time, the system uses the information to understand the reasons why the user feels angry and suggests ways to deal with it.

[1164] Example 1

[1165] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[1166] In modern text-based communication, it is difficult to accurately understand the user's emotions and nuances and provide appropriate feedback and suggestions for improvement. In particular, messages that contain aggressive or indifferent nuances can reduce the quality of communication and undermine psychological safety. It is also difficult for users to accurately recognize and appropriately express their own emotional state, so there is a demand for smoother communication.

[1167] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1168] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making suggestions for improving the text to be more "kind" based on the detected nuances, a means for receiving a text message input by a user, a means for analyzing the text message using natural language processing technology, a means for estimating the user's emotional state based on the analysis results, and a means for notifying the user of the estimation results. This makes it possible to accurately understand the user's emotions and nuances and provide appropriate feedback and suggestions for improvement.

[1169] "Text-based communication" refers to communication methods that use written information, including email and messenger apps.

[1170] "Aggressive or indifferent nuances" refers to expressions contained in text messages that convey negative or indifferent feelings or attitudes.

[1171] "Suggestions for improving considerate language" are suggestions for changing text messages that contain aggressive or insensitive nuances into more positive and considerate language.

[1172] The "means for receiving a text message entered by a user" is a function that enables the server to receive a text message sent by a user through a terminal.

[1173] "Means for analyzing text messages using natural language processing techniques" refers to techniques for understanding the content of text messages and analyzing their meaning and context, and includes tokenization, part-of-speech tagging, context analysis, etc.

[1174] The "means for inferring the emotional state of the user based on the analysis results" is a function for inferring the emotional state of the user by using the analysis results obtained by natural language processing technology.

[1175] The "means for notifying the user of the estimation result" is a function for notifying the user of the result of the estimated emotional state.

[1176] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[1177] The present invention is a system for text-based communication that analyzes text messages entered by a user, estimates the user's emotional state, and provides appropriate feedback and suggestions for improvement. A specific embodiment of the system will be described below.

[1178] Hardware and Software Used

[1179] server

[1180] The server plays a central role in receiving and analyzing text messages sent by users. The following software is installed on the server:

[1181] Natural Language Processing (NLP) technology: Google's Natural Language API, Microsoft's Text Analytics API

[1182] Emotion engines: IBM's Watson Tone Analyzer, Amazon's Comprehend

[1183] Terminal

[1184] A terminal is a device that allows a user to input text messages and communicate with a server, such as a smartphone or PC.

[1185] Program Processing

[1186] User Actions

[1187] A user inputs a text message using a text-based communication method such as email or a messenger app. For example, the user inputs the message "I am angry."

[1188] Device operation

[1189] The device transmits the text message entered by the user to the server, which requires an Internet connection.

[1190] Server Operations

[1191] The server receives text messages sent from the terminals and temporarily stores the received messages in a database.

[1192] The server then analyzes the text messages using natural language processing (NLP) techniques, such as Google's Natural Language API and Microsoft's Text Analytics API, to tokenize the text, tag parts of speech, and perform contextual analysis.

[1193] Based on the analysis results, the server uses an emotion engine to estimate the user's emotional state. This emotion engine uses IBM's Watson Tone Analyzer and Amazon's Comprehend. For example, it estimates that the user is in an angry emotional state from the keyword "angry."

[1194] The estimation results are sent from the server to the terminal using a secure communication protocol (e.g., HTTPS).

[1195] Device operation

[1196] The terminal displays the estimated emotional state received from the server to the user. For example, if the user inputs "I am angry," the terminal displays "You are in an angry emotional state."

[1197] Examples

[1198] As a concrete example, consider a case where a user uses a messenger app to send a message saying "I'm angry."

[1199] 1. A user types "I'm angry" into a messenger app.

[1200] 2. The device sends this message to the server.

[1201] 3. The server receives the message and stores it in a database.

[1202] 4. The server uses Google's natural language API to extract the keyword "angry."

[1203] 5. The server uses IBM's Watson Tone Analyzer to estimate that the user is in an angry emotional state.

[1204] 6. The server sends the estimation results to the device.

[1205] 7. The device displays to the user, "You are in an angry emotional state."

[1206] Examples of prompts to be input to a generative AI model include:

[1207] A user sends a message saying "I am angry". Analyze this message and infer the user's emotional state.

[1208] By feeding this prompt into a generative AI model, the model can infer the user's emotional state and generate an appropriate response.

[1209] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1210] Step 1:

[1211] The user enters a text message.

[1212] A user enters a text message using a text-based means of communication such as email or a messenger app. For example, the user enters the message "I am angry." The entered text message is the output of this step.

[1213] Step 2:

[1214] The device sends a text message to the server.

[1215] The terminal transmits the text message entered by the user to the server. This transmission requires an Internet connection. The input is the text message entered by the user and the output is the text message transmitted to the server.

[1216] Step 3:

[1217] A server receives the text message.

[1218] The server receives text messages sent from the terminal. The received messages are temporarily stored in a database. The input is the text message sent from the terminal, and the output is the text message stored in the database.

[1219] Step 4:

[1220] The server analyzes the text message using natural language processing (NLP) techniques.

[1221] The server analyzes the received text messages using natural language processing (NLP) technology. Google's natural language API and Microsoft's Text Analytics API are used for this analysis. Specifically, text tokenization, part-of-speech tagging, context analysis, etc. are performed. The input is the text message stored in the database, and the output is the analysis result.

[1222] Step 5:

[1223] The server uses an emotion engine to estimate the user's emotional state.

[1224] The server uses an emotion engine based on the analysis results to estimate the user's emotional state. IBM's Watson Tone Analyzer and Amazon's Comprehend are used for this emotion engine. For example, it estimates that the user is in an angry emotional state based on the keyword "angry." The input is the analysis results, and the output is the estimated emotional state.

[1225] Step 6:

[1226] The server transmits the estimation result to the terminal.

[1227] The server transmits the result of the emotional state estimated by the emotion engine to the terminal. A secure communication protocol (e.g., HTTPS) is used for this transmission. The input is the estimated result of the emotional state, and the output is the estimation result sent to the terminal.

[1228] Step 7:

[1229] The terminal displays the estimation results to the user.

[1230] The terminal displays the estimation result of the emotional state received from the server to the user. For example, if the user inputs "I am angry," the terminal displays "You are in an angry emotional state." The input is the estimation result sent from the server, and the output is the estimation result of the emotional state displayed to the user.

[1231] (Application example 1)

[1232] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[1233] In modern text-based communication, it is necessary to properly recognize the user's emotional state and respond quickly when an abnormal emotional state is detected. However, conventional systems often have difficulty accurately grasping the user's emotional state, and appropriate responses are often delayed. They also lack the ability to detect aggressive or indifferent nuances and suggest improvements to make the text more considerate. This may undermine the user's psychological safety.

[1234] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1235] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for suggesting improvements to make the text more "kind" based on the detected nuances, a means for estimating the emotional state of the user, and a means for issuing an alert if the emotional state is abnormal. This makes it possible to monitor the emotional state of the user in real time and quickly respond when an abnormality is detected.

[1236] "Text-based communication" refers to the exchange of information using written text, including email and messenger apps.

[1237] "Aggressive or indifferent nuances" are words or expressions that show hostility or indifference toward the other person, and can have a negative impact on communication.

[1238] "Suggestions for improving language to be more considerate" refers to making suggestions to change text that contains aggressive or insensitive nuances into more friendly and considerate language.

[1239] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language, and is used to understand the meaning of text.

[1240] "Means for inferring a user's emotional state" refers to technologies or algorithms for analyzing a user's emotions from text-based communications and inferring that emotional state.

[1241] "Means of sending an alert" refers to a function for sending a notification to an administrator or security officer if a user's emotional state is determined to be abnormal.

[1242] As an embodiment of the present invention, the following system can be constructed.

[1243] First, the server provides a means for text-based communication. This means can be in the form of email, messenger apps, etc. When a user communicates using these means, the server receives the text and uses natural language processing technology to understand the meaning of the text.

[1244] Next, the server has a means for detecting offensive or insensitive nuances in the text-based communication, which is realized using a natural language processing library (e.g., the nltk library), and specifically, a SentimentIntensityAnalyzer is used to analyze the sentiment score of the text.

[1245] In addition, the server provides a means to suggest improvements to make sentences more "kind" based on the detected nuances. This means uses a generative AI model to generate appropriate suggestions. For example, it generates prompt sentences to convert aggressive expressions into friendly expressions.

[1246] The server also includes a means for estimating the user's emotional state, which includes techniques and algorithms for analyzing the user's emotions from text-based communications and estimating the emotional state. Specifically, the server uses a SentimentIntensityAnalyzer to analyze the user's emotion score.

[1247] Finally, the server provides a means for sending an alert if the emotional state is abnormal, including functionality for sending a notification to an administrator or security officer if the emotional score falls below a certain threshold, by sending an email using the smtplib library.

[1248] For example, if a user types "I'm so angry," the emotion engine will recognize this text as a negative emotion and send an alert to the administrator, with the following example prompt:

[1249] If a user types "I am so angry", the sentiment engine should recognize this text as a negative sentiment and send an alert to an admin.

[1250] In this way, emotion-aware security apps can monitor a user's emotional state in real time and respond quickly if anomalies are detected.

[1251] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1252] Step 1:

[1253] The users engage in text-based communication.

[1254] Input: Text typed by the user (e.g. "I'm so angry")

[1255] Output: The text data entered

[1256] Specific action: A user types and sends text via email or a messenger app.

[1257] Step 2:

[1258] The server receives the entered text.

[1259] Input: Text data submitted by the user

[1260] Output: Received text data

[1261] Specific operation: The server receives text data through a mail server or messenger server.

[1262] Step 3:

[1263] The server uses natural language processing techniques to analyze the meaning of the text.

[1264] Input: Received text data

[1265] Output: Sentiment score of the text (e.g. {'neg': 0.8, 'neu': 0.1, 'pos': 0.1, 'compound': -0.7})

[1266] Specific behavior: The server uses the SentimentIntensityAnalyzer from the nltk library to calculate the sentiment score of the text.

[1267] Step 4:

[1268] The server detects nuances in the text and determines whether it contains offensive or insensitive content.

[1269] Input: Sentiment score of the text

[1270] Output: Nuance detection result (e.g. aggressive)

[1271] Specific operation: The server analyzes the emotion scores and detects aggressive nuances when the negative scores are high.

[1272] Step 5:

[1273] Based on the nuances detected by the server, suggestions for improving the text to be more "considerate" are generated.

[1274] Input: Nuance detection results, original text data

[1275] Output: Text of the improvement suggestion (e.g. "I am a bit dissatisfied")

[1276] How it works: The server uses a generative AI model to generate prompts that convert aggressive expressions into friendly ones.

[1277] Step 6:

[1278] The server estimates the user's emotional state.

[1279] Input: Sentiment score of the text

[1280] Output: Estimated emotional state (e.g. anger)

[1281] Specific operation: The server estimates the user's emotional state based on the emotion score.

[1282] Step 7:

[1283] The server issues an alert if the emotional state is abnormal.

[1284] Input: Emotion state estimation result

[1285] Output: Alert notification (e.g. email notification to administrator)

[1286] Specific behavior: If the server determines that the emotional state is abnormal (e.g., very angry), it uses the smtplib library to send an alert via email to the administrator.

[1287] In this way, an emotion-aware security app can monitor a user's emotional state in real time and respond quickly if anomalies are detected.

[1288] Example 2

[1289] Next, a description will be given of Example 2 of the embodiment 2. In the following description, the data processing device 12 is referred to as a "server" and the smart glasses 214 are referred to as a "terminal".

[1290] In text-based communication, aggressive or indifferent nuances may be included, which may reduce the quality of communication. Also, if users cannot express their emotional state appropriately, misunderstandings and conflicts may occur. This may worsen relationships between users and damage psychological safety.

[1291] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for acquiring a text-based message created by a user, a means for recognizing the user's emotion from the message using an emotion engine, and a means for making an improvement proposal for the message using a prompt sentence for instructing to suggest an improvement of the message based on the recognized emotion of the user when the message contains an aggressive nuance, and a generative AI model. This makes it possible to improve the quality of communication by mitigating aggressive or indifferent nuances while respecting the emotional state of the user. In addition, the means for making an improvement proposal for the message in the server determines whether the message created by the user contains an aggressive nuance based on an exchange between the message created by the user and a message created by the other party, and makes an improvement proposal for the message using a prompt sentence for instructing to suggest an improvement of the message based on the recognized emotion of the user when the aggressive nuance is included, and a generative AI model.

[1292] "Text-based communication" refers to the exchange of information using written information.

[1293] "Aggressive or indifferent nuances" refer to words and expressions that show hostility or indifference toward the other person.

[1294] "Thoughtful writing" refers to words and expressions that take into consideration the other person's feelings and situation and show consideration and kindness.

[1295] "Sentiment analysis technology" refers to technology that analyzes words, phrases, and context in text and infers emotional states.

[1296] "Words, phrases, and context" refer to the basic linguistic elements that make up a text and their relationships.

[1297] "Emotional state" refers to the type and intensity of emotion expressed through text.

[1298] "Improvement suggestions" refer to suggesting more appropriate expressions or words based on detected nuances.

[1299] "Natural language processing technology" refers to technology that enables computers to understand and generate human language.

[1300] A "generative AI model" is a model for generating text using artificial intelligence.

[1301] "Inference" refers to predicting a state or outcome based on given data.

[1302] This invention is a system that detects aggressive or indifferent nuances in text-based communication and suggests ways to improve the text to be more "kind" based on the emotional state. This system works in cooperation with a server, a terminal, and a user.

[1303] The server uses sentiment analysis techniques to analyze words, phrases, and context in the text to infer the emotional state. Specifically, it uses a sentiment analysis engine built using Python and TENSORFLOW. The sentiment analysis engine uses pre-trained generative AI models (e.g., BERT and GPT-3) to analyze the text.

[1304] The device is responsible for sending the text entered by the user to the server. When the user enters text into the input field of the device, the device sends the text in JSON format to the server. The server analyzes the received text and estimates the emotional state.

[1305] For example, if a user enters "I don't understand your opinion at all," the server passes this text to a sentiment analysis engine. The sentiment analysis engine analyzes the text and infers that the user is in an angry emotional state. Next, the server uses a prompt sentence that instructs the server to suggest improvements to the message based on the recognized user's emotions if the message contains aggressive nuances, and a generative AI model to make suggestions for improving the message, generating suggestions for improving the message to be more "kind." For example, the server generates a suggestion such as, "Could you please explain your opinion in more detail?"

[1306] Here is an example of a prompt to be input to the generative AI model:

[1307] "Input text: "I don't understand your opinion at all" Emotional state: "Anger" Determine whether the input text contains offensive nuances, and if so, suggest improvements to the message based on the perceived emotional state of the user."

[1308] Example of a suggestion for improvement: "Can you explain your opinion in more detail?"

[1309] The server sends the generated improvement proposal in JSON format to the terminal. The terminal displays the received improvement proposal to the user. The user can check the displayed improvement proposal and modify the text if necessary.

[1310] In this way, it is possible to improve the quality of communication by respecting the user's emotional state while mitigating aggressive or indifferent nuances. When making an improvement suggestion, it is preferable to further consider the interaction between the user's message and the other party's message. In this case, the server judges whether the message created by the user contains an aggressive nuance based on the interaction between the message created by the user and the message created by the other party, and if an aggressive nuance is included, makes an improvement suggestion for the message using a prompt sentence instructing to suggest an improvement of the message based on the recognized emotion of the user and a generative AI model. Here is an example of a prompt to be input to the generative AI model:

[1311] After entering the exchange between your message and the other person's message, "Input text: "I don't understand your opinion at all" Emotional state: "Anger" Given the messaging conversation, determine whether the input text contains offensive nuances and, if so, suggest improvements to the message based on the perceived emotional state of the user.”

[1312] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1313] Step 1:

[1314] A user inputs text into an input field on a terminal. For example, the user inputs "I don't understand your opinion at all." This input text becomes the input data for the next processing step.

[1315] Step 2:

[1316] The terminal converts the text entered by the user into JSON format and sends it to the server as an HTTP request. At this time, the input text is sent to the server as JSON format data.

[1317] Step 3:

[1318] The server parses the JSON data received from the terminal and extracts the text part. Specifically, it extracts the "text" field from the received JSON data and passes its contents to the next processing step. This extracted text becomes the input data for the next processing step.

[1319] Step 4:

[1320] The server passes the extracted text to a sentiment analysis engine, which is built using Python and TensorFlow and analyzes the text using pre-trained generative AI models (e.g., BERT and GPT-3). The sentiment analysis engine considers words, phrases, and context in the text to infer an emotional state. This inferred emotional state becomes the input data for the next processing step.

[1321] Step 5:

[1322] The server generates an appropriate improvement suggestion based on the emotional state obtained from the sentiment analysis engine. For example, if the user is estimated to be in an angry emotional state, the server generates an improvement suggestion in a considerate text such as "Can you please explain your opinion in more detail?". This generated improvement suggestion becomes the input data for the next processing step.

[1323] Step 6:

[1324] The server converts the generated improvement proposal into JSON format and sends it to the terminal as an HTTP response. At this time, the improvement proposal is sent to the terminal as JSON format data.

[1325] Step 7:

[1326] The terminal parses the JSON data received from the server and extracts the improvement suggestion part. Specifically, it extracts the "improvement suggestion" field from the received JSON data and displays its contents to the user. The user can check the displayed improvement suggestion and modify the text if necessary.

[1327] In this way, it is possible to improve the quality of communication by respecting the user's emotional state while mitigating aggressive or indifferent nuances. The server may also issue an alert to the user or a third party if the recognized emotion is abnormal.

[1328] (Application example 2)

[1329] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server", and the smart glasses 214 will be referred to as a "terminal".

[1330] In modern corporate communications, messages containing aggressive or indifferent nuances are frequently generated, which causes a bad atmosphere in the workplace. Such communications can damage employees' psychological safety and lead to reduced work efficiency and increased turnover. Therefore, there is a need for a system that can monitor chat and email exchanges within a company in real time, detect aggressive or indifferent nuances, and make appropriate improvement suggestions.

[1331] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1332] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making improvement suggestions to make the text more "kind" based on the detected nuances, a means for monitoring chat and email exchanges within the company in real time, a means for analyzing words, phrases, and contexts in the text using sentiment analysis technology, and a means for generating improvement suggestions using a generative AI model. This improves the quality of communication within the company, ensures psychological safety for employees, and enables improved business efficiency and reduced turnover.

[1333] "Text-based communication" is a method of exchanging information using written text.

[1334] "Aggressive or indifferent nuances" refer to words and expressions that show hostility or indifference toward the other person.

[1335] "Thoughtful writing" refers to expressions that take into consideration the other person's feelings and situation and show kindness and consideration.

[1336] "Intra-company chat and email communications" refers to electronic message exchanges that take place within a company.

[1337] "Real-time monitoring" means checking the contents of a message the moment it is sent or received.

[1338] "Sentiment analysis technology" is a technology that analyzes words, phrases, and context in text and estimates the emotions contained in the text.

[1339] A "generative AI model" is a model that uses artificial intelligence to generate new text and suggestions.

[1340] "Generating improvement suggestions" means creating suggestions for transforming the original text into more appropriate expressions.

[1341] A system for carrying out the present invention is configured as follows: A server includes a means for conducting text-based communication, a means for detecting offensive or insensitive nuances, a means for making suggestions for improving text to be more "kind" based on the detected nuances, a means for monitoring chat and email exchanges within a company in real time, a means for analyzing words, phrases, and contexts in the text using sentiment analysis technology, and a means for generating suggestions for improvement using a generative AI model.

[1342] Program processing explanation

[1343] The server first provides an interface for text-based communication, where users can send and receive chats and emails. The server then monitors these messages in real time to detect offensive or insensitive comments, a process that involves analyzing words, phrases and context within the text using sentiment analysis techniques.

[1344] For sentiment analysis, we use natural language processing libraries such as TextBlob. This allows us to obtain the sentiment polarity of the text (positive, negative, neutral) as a numerical value. For example, when we use TextBlob to analyze the text "Your job is no good. Do it better," we determine that the sentiment polarity is negative.

[1345] The server then uses a generative AI model (e.g., OpenAI's GPT-3) to generate improvement suggestions based on the detected nuances. In this process, the prompt sentence is input to the generative AI model to get appropriate improvement suggestions. Examples of prompt sentences include:

[1346] Please improve the following text: The emotional state is anger.

[1347] Your job is no good. Do a better job.

[1348] Improved text:

[1349] The generated improvement suggestions are presented to the user to help them refine the original text into a more thoughtful one.

[1350] Examples

[1351] For example, in a company's chat system, if an employee sends a message saying, "Your job is terrible. Do it better," the server monitors this message in real time and performs sentiment analysis using TextBlob. As a result, it is determined that this message contains negative sentiment. Next, the server uses a generative AI model to input the following prompt sentence and generate improvement suggestions:

[1352] Please improve the following text: The emotional state is anger.

[1353] Your job is no good. Do a better job.

[1354] Improved text:

[1355] The generated improvement suggestions are presented to the user in the form of text such as "There is room for improvement in your work. Please pay more attention next time." This improves the quality of communication within the company and ensures psychological safety for employees.

[1356] The flow of the identification process in the application example 2 will be described with reference to FIG.

[1357] Step 1:

[1358] The server provides an interface for text-based communication, including input fields and a send button that allow users to chat and send / receive email. Through this interface, users enter and send messages. The input is the text entered by the user, and the output is the text data sent to the server.

[1359] Step 2:

[1360] The server monitors the messages sent in real time. It receives the sent text data and starts processing to detect offensive or insensitive nuances. The input is the sent text data and the output is the text data for sentiment analysis.

[1361] Step 3:

[1362] The server uses sentiment analysis technology to analyze words, phrases, and context in the text. Specifically, it uses a natural language processing library such as TextBlob to obtain the sentiment polarity of the text (positive, negative, neutral) as a numerical value. The input is text data, and the output is numerical data indicating the sentiment polarity.

[1363] Step 4:

[1364] If the sentiment polarity is too negative or too positive, the server uses a generative AI model to generate improvement suggestions. Specifically, it uses OpenAI's GPT-3 and inputs the prompt sentence into the generative AI model. The input is the sentiment polarity and the original text data, and the output is the text data containing the improvement suggestions.

[1365] Step 5:

[1366] The server presents the generated improvement suggestions to the user. The user refers to the proposed improvement suggestions and modifies the original text to make it more considerate. The input is text data including improvement suggestions, and the output is the improvement suggestions presented to the user.

[1367] Step 6:

[1368] The user corrects the original text based on the proposed improvement suggestions and resends it. This improves the quality of communication within the company and ensures psychological safety for employees. The input is the text corrected by the user, and the output is the corrected text data that is resent to the server.

[1369] Example 3

[1370] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal".

[1371] In modern text-based communication, the frequent use of aggressive or indifferent expressions is a problem. This leads to poor communication quality and makes misunderstandings and conflicts more likely to occur. In addition, there is a lack of means to properly manage users' emotional states and improve their communication skills. To solve these issues, a system is needed that can suggest more "kind" expressions and track and manage users' emotional states.

[1372] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1373] In this invention, the server includes a means for a user to input text, a means for transmitting the input text to the server, a means for the server to analyze the nuance of the text, a means for detecting aggressive or indifferent expressions based on the analysis result, a means for generating improvement suggestions for replacing the detected expressions with "kind" expressions, a means for presenting the generated improvement suggestions to the user, and a means for tracking the user's emotional state over the long term and proposing ways to deal with the situation. This allows the user to use more "kind" expressions, improves the quality of communication, and reduces misunderstandings and conflicts. It is also possible to appropriately manage the user's emotional state and support the improvement of communication skills.

[1374] A "user" is an individual or entity that utilizes the system to input text and receive improvement suggestions.

[1375] "Text" refers to a character string or sentence entered by a user and used as a means of communication.

[1376] A "server" is a computer system for receiving and analyzing user-submitted text and generating improvement suggestions.

[1377] "Nuance" refers to subtle meanings that reveal the emotion or intent contained in text.

[1378] "Analysis" refers to the process the server performs to understand the nuances of the text, often using natural language processing techniques.

[1379] "Offensive language" is any word or phrase that may offend others or provoke negative feelings.

[1380] "Insensitive speech" refers to words or phrases that show a lack of consideration or concern for others.

[1381] "Thoughtful expressions" are words and phrases that show consideration and empathy for others.

[1382] "Improvement suggestions" are suggestions for replacing aggressive or insensitive expressions with more "compassionate" expressions.

[1383] An "emotional state" is the emotion a user is feeling at a particular point in time, and may include anger, joy, sadness, etc.

[1384] "Tracking" refers to the long-term observation and recording of a user's emotional state.

[1385] A "cope" is a method or advice for a user to deal with a particular emotional state.

[1386] "Natural language processing technology" refers to technology that enables computers to understand and generate human language.

[1387] A "generative AI model" is a model for generating text using artificial intelligence, including, for example, a specific algorithm or a neural network.

[1388] This invention is a system in which a user inputs text, analyzes the nuances of the text, and improves it to a more "considerate" expression. The system is composed of three main elements: a user, a terminal, and a server.

[1389] First, a user inputs text using a terminal (e.g., a PC or smartphone). The input text is sent to the server via the terminal. At this time, the HTTPS protocol is used to ensure the security of the data.

[1390] The server passes the received text to an emotion engine, which uses natural language processing technologies such as IBM Watson or Google Cloud Natural Language API to analyze the nuances of the text and determine whether the text contains aggressive, indifferent, or other emotions.

[1391] Next, the server generates improvement suggestions based on the analysis results to replace offensive or insensitive expressions with "kind" expressions. This process uses a generative AI model (for example, OpenAI's GPT-4). The generative AI model understands the context of the input text and generates appropriate "kind" expressions.

[1392] Specifically, the server converts the text "You're always late" into an improvement suggestion: "I understand that you try hard to be on time, but it would be helpful if you could come a little earlier."

[1393] The generated improvement proposals are sent from the server to the terminal and presented to the user. The user can check the proposed improvement proposals and make corrections as necessary.

[1394] In addition, the server tracks the user's emotional state over time. The emotion engine uses this information to suggest ways to deal with situations when the user is estimated to be in an angry emotional state over a period of time. For example, if the user frequently expresses anger, the server will suggest ways to deal with the situation, such as "take a deep breath and take some time to calm down."

[1395] Examples of concrete examples and prompts

[1396] Examples

[1397] If a user types "You're always late," the device sends this text to the server. The server uses an emotion engine to analyze the nuances of the text and determines that it is an offensive expression. Next, the server uses a generative AI model to generate an improvement suggestion such as "I understand that you try hard to be on time, but it would be helpful if you could come a little earlier." Finally, the server presents this improvement suggestion to the user, who confirms it.

[1398] Examples of prompt statements

[1399] "Please improve the following text to be more considerate: 'You are always late.'"

[1400] In this way, the system analyzes the text entered by the user, generates and presents appropriate improvement suggestions, and thereby improves the quality of communication. The flow of the identification process in the third embodiment will be described with reference to FIG.

[1401] Step 1:

[1402] The user enters text.

[1403] Input: A user uses a terminal to enter text, for example, "You are always late."

[1404] Output: The entered text is saved to the terminal.

[1405] Specific operation: A user inputs text using a keyboard on a PC or smartphone.

[1406] Step 2:

[1407] The device sends the text to the server.

[1408] Input: The text entered by the user.

[1409] Output: The text is sent to the server.

[1410] Specific operation: The terminal sends the entered text to the server using the HTTPS protocol.

[1411] Step 3:

[1412] The server analyzes the text for nuances.

[1413] Input: The text sent from the terminal.

[1414] Output: Nuance analysis of the text.

[1415] Specific behavior: The server uses an emotion engine (e.g., an API that uses natural language processing technology) to analyze the nuances of the text. As a result of the analysis, it determines whether the text contains aggressive, indifferent, or other emotions.

[1416] Step 4:

[1417] The server generates improvement suggestions based on the analysis results.

[1418] Input: Nuance analysis results for text.

[1419] Output: Improvement suggestions.

[1420] Specific behavior: The server uses a generative AI model (e.g., a generative AI model) to generate improvement suggestions to replace aggressive or insensitive expressions with "kind" expressions. For example, converting the text "You are always late" into the improvement suggestion "I understand that you try hard to be on time, but it would be helpful if you could come a little earlier."

[1421] Step 5:

[1422] The server presents the generated improvement suggestions to the user.

[1423] Input: The generated improvement suggestions.

[1424] Output: The improvement suggestions are sent to the terminal and presented to the user.

[1425] Specific operation: The server uses the HTTPS protocol to send the generated improvement suggestions to the terminal, and the terminal displays the received improvement suggestions to the user.

[1426] Step 6:

[1427] The user reviews the improvement suggestions.

[1428] Input: Improvement suggestions sent by the server.

[1429] Output: The user confirmation result.

[1430] Specific actions: The user checks the improvement suggestions displayed on the device screen and makes corrections if necessary.

[1431] Step 7:

[1432] The server tracks the user's emotional state over time.

[1433] Input: User's text input history and sentiment analysis results.

[1434] Output: Tracking data of the user's emotional state and suggestions on how to deal with it.

[1435] Specific behavior: The server uses the emotion engine to track the user's emotional state over the long term. For example, if the user is estimated to be in an angry emotional state for a certain period of time, the server will suggest countermeasures such as "take a deep breath and take some time to calm down."

[1436] (Application example 3)

[1437] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server", and the smart glasses 214 will be referred to as a "terminal".

[1438] In text-based communication, aggressive or indifferent nuances can make the other person feel uncomfortable. In electronic payment services, if communication with the other party is not smooth, problems and misunderstandings may occur. This poses the challenge of making it difficult for users to improve their communication skills and ensure psychological safety.

[1439] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1440] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making suggestions for improving the text to be more "considerate" based on the detected nuances, and a means for making suggestions for improving communication with a transaction partner in an electronic payment service to be more "considerate." This enables a user to have more friendly and constructive communication when using an electronic payment service.

[1441] "Text-based communication" is a method of exchanging information using written text.

[1442] "Aggressive or indifferent nuances" refer to expressions or attitudes that show hostility or indifference toward the other person.

[1443] "Thoughtful writing" is writing that takes into consideration the other person's feelings and position and uses friendly and considerate language.

[1444] An "electronic payment service" is a service that enables monetary transactions via the Internet or mobile devices.

[1445] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[1446] "Communication skills" are the ability to effectively convey information and understand others.

[1447] "Psychological safety" refers to an environment in which individuals feel safe to express their opinions and ask questions.

[1448] "Transaction partner" refers to a party with whom money, goods, or services are exchanged in an electronic payment service.

[1449] "Friendly and constructive communication" is communication that shows a friendly attitude toward the other person and promotes problem solving and cooperation.

[1450] A system for implementing this invention includes means for conducting text-based communication, means for detecting aggressive or indifferent nuances, means for making suggestions for improving communication to make it more "considerate" based on the detected nuances, and means for making suggestions for improving communication with a transaction partner in an electronic payment service to make it more "considerate."

[1451] The server receives text-based communications from users and analyzes their content. The analysis uses natural language processing techniques. Specifically, it uses OpenAI's generative AI models (e.g., GPT-3) to understand the meaning of the text and detect offensive or insensitive nuances.

[1452] Based on the detected nuances, the server generates suggestions for improving "kind" text, which will help users to communicate more friendly and constructively with their counterparts when using electronic payment services.

[1453] If the message entered by the user is "You're wrong. Please refund me.", the server will analyze it and detect the offensive nuance. Then, using a generative AI model, it will generate suggestions for improving the message to be more "kind" like the following:

[1454] "Thank you for your help. I've had a slight problem with the transaction and would like to discuss a refund."

[1455] In this way, users can have a more friendly and constructive communication.

[1456] The hardware used is a server and the user's smartphone. The software uses Python and the OpenAI API. Data processing and data calculation involves using generative AI models to analyze the meaning of the text and generate appropriate improvement suggestions.

[1457] An example of a prompt is as follows:

[1458] Please improve the following sentences to make them more considerate:

[1459] You are wrong. Please refund my money.

[1460] Improved sentence:

[1461] The system will enable users to have more friendly and constructive communication when using electronic payment services.

[1462] The flow of the specific process in the application example 3 will be described with reference to FIG.

[1463] Step 1:

[1464] A user inputs a text-based message using a smartphone, and the input message is sent to a server.

[1465] Input: A text message entered by the user.

[1466] Output: The text message sent to the server.

[1467] Step 2:

[1468] The server analyzes the received text message. It uses natural language processing technology to understand the meaning of the text. Specifically, it uses a generative AI model (e.g. GPT-3) to analyze the content of the text.

[1469] Input: The text message sent to the server.

[1470] Output: Semantic information of the parsed text

[1471] Step 3:

[1472] The server detects offensive or insensitive nuances based on the semantic information of the analyzed text. It uses a generative AI model to evaluate the nuances of the text and determine whether it contains offensive or insensitive expressions.

[1473] Input: Semantic information of the parsed text

[1474] Output: Nuance detection result (aggressive, indifferent, etc.)

[1475] Step 4:

[1476] The server generates improvement suggestions for "kind" sentences based on the detected nuances. Using a generative AI model, it generates appropriate improvement suggestions and presents them to the user.

[1477] Input: Nuance detection result

[1478] Output: Suggestions for improving the text to be more "kind"

[1479] Step 5:

[1480] The server sends the generated improvement proposals to the user's smartphone, where the user can check the proposals and make corrections if necessary.

[1481] Input: Suggestions for improving the wording to be more "kind"

[1482] Output: Improvement suggestions sent to the user's smartphone

[1483] Step 6:

[1484] The user reviews the suggested improvements, makes any necessary corrections, and then sends the final message to the trading partner.

[1485] Input: The final message that the user has confirmed and modified.

[1486] Output: The final message sent to the trading partner

[1487] In this way, users can have more friendly and constructive communication when using electronic payment services.

[1488] 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 a voice indicating a user input for the result of the specific processing. The control unit 46A transmits the 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.

[1489] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by making a neural network perform deep learning. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image is input. The data generation model 58 performs inference on the input inference data according to the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1490] Another example of generative AI is Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) are mentioned.

[1491] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1492] [Third embodiment]

[1493] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1494] 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.

[1495] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a wide area network (WAN) and / or a local area network (LAN).

[1496] 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.

[1497] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs the voice according to instructions from the processor 46.

[1498] Camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (e.g., an imaging range defined by an angle of view equivalent to the width of the field of vision of an average healthy person).

[1499] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54. The transmission and reception of various types of information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is performed in a secure state.

[1500] Fig. 6 shows an example of 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.

[1501] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32, and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1502] 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.

[1503] In the headset type terminal 314, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50, and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1504] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1505] "Example 1"

[1506] In one embodiment of the present invention, a means for text-based communication may be email, a messenger app, etc. When a user communicates using these means, the system receives the text and uses natural language processing technology to understand the meaning of the text.

[1507] "Example 2"

[1508] To detect offensive or insensitive nuances, the system uses sentiment analysis technology. Specifically, it analyzes words, phrases, and contexts in a text to determine whether the text contains offensive or insensitive nuances.

[1509] "Example 3"

[1510] As a means of making suggestions to improve sentences to be more "compassionate," the system generates improvement suggestions based on the detected nuances. Specifically, the system makes suggestions to replace aggressive or indifferent expressions with "compassionate" expressions. For example, the system makes suggestions to improve the expression "You are wrong" to a more "compassionate" expression such as "I understand your point of view, but I think from a slightly different perspective."

[1511] The process flow of each embodiment will be described below.

[1512] "Example 1"

[1513] Step 1: The user engages in text-based communication via email, messenger app, etc.

[1514] Step 2: The system receives the text and uses natural language processing techniques to understand the meaning of the text.

[1515] "Example 2"

[1516] Step 1: The system receives the text.

[1517] Step 2: Using sentiment analysis techniques, analyze the words, phrases, and context in the text. Step 3: Determine whether the text contains offensive or insensitive nuances.

[1518] "Example 3"

[1519] Step 1: If offensive or insensitive nuances are detected, the system generates improvement suggestions.

[1520] Step 2: Make suggestions to replace aggressive or insensitive expressions with more "compassionate" ones. Step 3: Make suggestions to improve the expression, for example, "You are wrong" to "I understand your point of view, but I think from a slightly different perspective."

[1521] Example 1

[1522] Next, a description will be given of Example 1 of the embodiment 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal".

[1523] In conventional text-based communication systems, it was difficult to properly understand the nuances of messages sent by users, detect aggressive or indifferent expressions, and make suggestions for improvement. In addition, there was a lack of technology to generate appropriate responses to messages sent by users. This made it difficult to improve users' communication skills and realize considerate communication.

[1524] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1525] In this invention, the server includes a means for a user to send a message, a means for a terminal to send a message to the server, a means for the server to receive the message, a means for the server to preprocess the message, a means for the server to semantically analyze the message, a means for the server to generate a response, a means for the server to send the response to the terminal, and a means for the terminal to display the response to the user. This makes it possible to properly understand the nuances of messages sent by users, detect aggressive or indifferent expressions, and make suggestions for improvement. In addition, it is possible to generate appropriate responses to user messages, improve the user's communication skills, and realize considerate communication.

[1526] "Text-based communication" refers to the exchange of information using written text, including email and messenger apps.

[1527] "Aggressive or indifferent nuances" refer to expressions or attitudes that show hostility or indifference toward the other person.

[1528] "Thoughtful writing" refers to writing that takes into consideration the other person's feelings and situation and includes expressions that show consideration and kindness.

[1529] "Means for a user to send a message" refers to an interface or function that allows a user to input and send text.

[1530] "Means for a terminal to send a message to a server" refers to a function of a user's terminal to send a message to a server via the Internet.

[1531] "Means for the server to receive a message" refers to the function of the server to receive a message sent from a terminal.

[1532] "Means for the server to preprocess the message" refers to a function for performing processing to convert the received message into a form that is easy to analyze.

[1533] "Means for the server to semantically analyze the message" refers to the functionality for understanding the meaning of the pre-processed message and extracting important information.

[1534] "Means for the server to generate a response" refers to the functionality for generating an appropriate response based on the analysis results.

[1535] The term "means for the server to send a response to the terminal" refers to a function for sending the generated response to the user's terminal.

[1536] "Means for the terminal to display the response to the user" refers to a function for visually displaying the response received from the server to the user.

[1537] The object of this invention is to provide a system for text-based communication that properly understands the nuances of messages sent by users, detects offensive or indifferent expressions, makes suggestions for improvement, and generates appropriate responses.

[1538] System program generation

[1539] First, a user sends a message using a text-based communication method (e.g., email or a messenger app). The device sends this message to the server. The server preprocesses the received message, performs semantic analysis, generates an appropriate response, and sends it to the device. The device displays the received response to the user.

[1540] Hardware and Software Used

[1541] The system uses the following hardware and software:

[1542] Hardware: User's device (smartphone, PC, etc.), server

[1543] Software: Natural language processing libraries (e.g., Python's "NLTK" or "spaCy"), generative AI models (e.g., OpenAI's "GPT-3")

[1544] Data processing and calculation

[1545] The server performs the following data processing and calculations:

[1546] 1. Text preprocessing: The server tokenizes the received message and removes unnecessary symbols and stop words. For this process, we use "NLTK" and "spaCy".

[1547] 2. Semantic Analysis: The server analyzes the semantics of the preprocessed text and extracts keywords and important phrases using natural language processing models such as "spaCy" or "BERT".

[1548] 3. Response Generation: The server generates an appropriate response based on the analysis results. This response generation uses a generative AI model (e.g., GPT-3).

[1549] Examples

[1550] As a concrete example, let's say a user uses a messenger app to send the following message:

[1551] "What time does the meeting start tomorrow?"

[1552] The server receives this message and processes it as follows:

[1553] 1. Text preprocessing: The server tokenizes the text “What time does the meeting start tomorrow?” and breaks it down into the words “tomorrow,” “meeting,” “what time,” “start,” and “is.”

[1554] 2. Semantic analysis: The server extracts the keywords "meeting" and "what time" and understands that the user is asking about the start time of the meeting.

[1555] 3. Response generation: The server uses a generative AI model (e.g., GPT-3) to generate an appropriate response. It inputs the prompt "Message from user: What time does tomorrow's meeting start? Please generate a response." and generates the response "Tomorrow's meeting starts at 10:00 a.m."

[1556] In this manner, a system is implemented in which the server analyzes a user's message and generates and delivers an appropriate response to the user.

[1557] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1558] Step 1:

[1559] A user sends a message.

[1560] Specific actions: A user opens a messenger app, types "What time is the meeting tomorrow?", and hits send.

[1561] Input: The text message entered by the user.

[1562] Output: The instruction to send is sent to the terminal.

[1563] Step 2:

[1564] The terminal sends a message to the server.

[1565] Specific operation: The terminal transmits the entered message to a server via the Internet.

[1566] Input: The text message entered by the user.

[1567] Output: A message is sent to the server.

[1568] Step 3:

[1569] The server receives the message.

[1570] Specific operation: The server receives the message sent from the terminal and stores it in a database.

[1571] Input: The text message sent from the device.

[1572] Output: The received message is stored in the database.

[1573] Step 4:

[1574] The server pre-processes the message.

[1575] Specific operation: The server tokenizes the received message and removes unnecessary symbols and stop words. This process is done using "NLTK" and "spaCy".

[1576] Input: The received text message.

[1577] Output: Preprocessed text data.

[1578] Step 5:

[1579] The server semantically parses the message.

[1580] How it works: The server analyzes the preprocessed text and extracts keywords and important phrases using natural language processing models such as spaCy and BERT.

[1581] Input: Preprocessed text data.

[1582] Output: Parsed keywords or phrases.

[1583] Step 6:

[1584] The server generates a response.

[1585] Specific operation: The server generates an appropriate response based on the analysis results. A generative AI model (e.g., GPT-3) is used to generate this response. The server inputs the prompt text "Message from user: What time does the meeting start tomorrow? Please generate a response." and generates a response.

[1586] Input: The parsed keyword or phrase.

[1587] Output: The generated response text.

[1588] Step 7:

[1589] The server sends a response to the terminal.

[1590] Specific operation: The server sends the generated response to the user's terminal via the Internet.

[1591] Input: The generated response text.

[1592] Output: The response is sent to the terminal.

[1593] Step 8:

[1594] The terminal displays the response to the user.

[1595] Specific operation: The terminal visually displays the response received from the server to the user. The user confirms the displayed response.

[1596] Input: The response text sent by the server.

[1597] Output: The response text that is displayed to the user.

[1598] (Application example 1)

[1599] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server", and the headset type terminal 314 will be referred to as a "terminal".

[1600] Conventional text-based communication systems can detect the nuances in the text sent by users and make appropriate suggestions for improvement, but they lack the functionality to recommend related content based on the user's interests and concerns. This limits the information and knowledge that users can obtain through communication, making it difficult to improve the quality of communication.

[1601] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1602] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for suggesting improvements to make the text more "kind" based on the detected nuances, and a means for analyzing the user's text and recommending related content based on the content of the text. This allows the user to not only improve the quality of communication, but also to obtain related content based on their interests and concerns.

[1603] "Text-based communication" is any means by which users exchange information using written text.

[1604] "Aggressive or indifferent undertones" are expressions contained in the text that indicate negative or indifferent feelings or attitudes.

[1605] "Thoughtful writing" is text that includes expressions that show consideration and kindness to the other person.

[1606] "Improvement Suggestions" are suggested modifications to make the content of the text a user sends more appropriate and considerate.

[1607] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language.

[1608] "Related content" is information such as articles, videos, and music that is recommended based on the content of the user's text.

[1609] A "recommendation method" is a method that analyzes a user's text and presents relevant content based on that content.

[1610] A system for implementing the present invention includes means for conducting text-based communication, means for detecting aggressive or insensitive nuances in the text-based communication, means for suggesting improvements to make the text more "kind" based on the detected nuances, and means for analyzing a user's text and recommending related content based on the content of the text.

[1611] The server receives the text sent by the user and uses natural language processing technology to analyze the meaning of the text. Specifically, it uses OpenAI's GPT-3 model to understand the content of the text and detect offensive or insensitive nuances. Based on the detected nuances, the server generates improvement suggestions for "kind" text.

[1612] In addition, the server recommends related content based on the content of the user's text. For example, it uses a news API to retrieve articles related to the user's interests and presents them to the user. This series of processes is implemented using a Python program.

[1613] As a concrete example, if a user sends a message saying, "I would like to know about recent technological advances," the server will input the following prompt sentence into the generative AI model:

[1614] Please recommend related content based on the content of the text below:

[1615] I want to know about the latest advances in technology

[1616] Based on this prompt, the generative AI model analyzes the meaning of the text and recommends relevant content. Article titles and URLs retrieved using a news API are presented to the user.

[1617] The hardware used is a smartphone, and the software used is Python, OpenAI API, and news API. This not only improves the quality of communication, but also allows users to obtain relevant content based on their interests.

[1618] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1619] Step 1:

[1620] A user sends a message using a text-based communication method (e.g. email or a messenger app). The input is the text the user sends, and the output is that text being sent to the server.

[1621] Step 2:

[1622] The server uses natural language processing techniques to analyze the received text. Specifically, it uses OpenAI's GPT-3 model to understand the meaning of the text. The input is the text sent by the user, and the output is the result of the semantic analysis of the text.

[1623] Step 3:

[1624] The server detects offensive or insensitive nuances in the text based on the analysis results. The input is the result of semantic analysis of the text, and the output is the detected nuance information.

[1625] Step 4:

[1626] The server generates improvement suggestions for "kind" sentences based on the detected nuances. The input is the nuance information, and the output is the text of the improvement suggestions.

[1627] Step 5:

[1628] The server recommends related content based on the content of the user's text. Specifically, it uses a news API to get related articles. The input is the user's text and the analysis results, and the output is a list of related articles.

[1629] Step 6:

[1630] The server presents the retrieved related content to the user. The input is a list of related articles, and the output is the titles and URLs of the articles presented to the user.

[1631] Step 7:

[1632] The user reviews the presented related content and browses it as needed. The input is the title and URL of the presented article, and the output is the user's browsing behavior.

[1633] Example 2

[1634] Next, a second embodiment of the second form will be described. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal".

[1635] When text-based communication contains offensive or insensitive nuances, it can lead to misunderstandings and conflicts. Users may also unintentionally use offensive language, lowering the quality of communication. There is a need to solve these problems and enable users to communicate more considerately.

[1636] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1637] In this invention, the server includes a means for performing text-based communication, a means for preprocessing text data, a means for performing sentiment analysis using a generative AI model, a means for outputting the results of the sentiment analysis, and a means for making suggestions for improving the text to be more "kind" based on the detected nuances. This makes it possible to detect aggressive or indifferent nuances in the text data entered by the user and make appropriate suggestions for improvement.

[1638] "Text-based communication" is a method of exchanging information using written text.

[1639] "Aggressive or indifferent nuances" refer to expressions or attitudes that show hostility or indifference toward others.

[1640] "Suggestions for improving language to be more considerate" are suggestions for changing text that contains aggressive or insensitive nuances into more friendly and considerate language.

[1641] "Text data preprocessing" refers to the process of removing unnecessary spaces and special characters from text data and dividing it into words and phrases.

[1642] A "generative AI model" is a model that uses artificial intelligence technology to analyze the sentiment and meaning of text.

[1643] "Sentiment analysis" is the process of analyzing the emotions and nuances contained in text data to determine whether they are aggressive, indifferent, or some other emotion.

[1644] "Output of sentiment analysis results" refers to providing the results obtained by sentiment analysis to a user.

[1645] The present invention is a system that detects aggressive or indifferent nuances in text-based communication and suggests improvements to make the text more considerate. A specific embodiment of this system will be described below.

[1646] System Configuration

[1647] Hardware and software configuration:

[1648] Server: A computer with high-performance computing power that performs data pre-processing, sentiment analysis, and outputting the results.

[1649] Terminal: A device (e.g., PC, smartphone) through which a user enters text data and receives the results.

[1650] Generative AI model: An artificial intelligence model for performing sentiment analysis (e.g., OpenAI's GPT-3).

[1651] Natural Language Processing library: Software used to preprocess text data (e.g., NLTK in Python).

[1652] Program Processing

[1653] Enter data:

[1654] The user uses the device to input text data to be analyzed, such as comments on social media or the contents of emails.

[1655] Data preprocessing:

[1656] The server preprocesses the received text data by removing unnecessary spaces and special characters and tokenizing the text using the Python NLTK (Natural Language Toolkit) library.

[1657] Sentiment analysis:

[1658] The server performs sentiment analysis on the pre-processed text data, using a generative AI model (e.g., OpenAI's GPT-3) to determine whether the text contains offensive or insensitive nuances.

[1659] The resulting output:

[1660] The server returns the results of the sentiment analysis to the user, indicating whether the text contains aggressive, indifferent, or other sentiments.

[1661] Examples

[1662] Let's say a user wants to analyze the social media comment "I don't care about you."

[1663] 1. Enter your data:

[1664] The user types "I don't care about you" into an input field on the terminal.

[1665] The terminal transmits this text data to the server.

[1666] 2. Data preprocessing:

[1667] The server preprocesses the received text data "I don't care about you."

[1668] Remove unnecessary spaces and special characters.

[1669] Tokenize the text and split it into "you," "something," and "it doesn't matter."

[1670] 3. Sentiment analysis:

[1671] The server creates a prompt sentence to input the preprocessed text data into the generative AI model:

[1672] "Please determine whether the following text contains offensive or insensitive connotations:

[1673] "I don't care about you."

[1674] A generative AI model takes this prompt and analyzes the sentiment of the text.

[1675] The generative AI model determines that the text is "offensive" and returns that result to the server.

[1676] 4. Result output:

[1677] The server returns the result it received from the generative AI model, “aggressive,” to the user.

[1678] The device displays the result to the user, for example, "This comment is offensive."

[1679] The above is a specific embodiment for carrying out the present invention. This system allows users to detect offensive or indifferent nuances contained in text data and receive appropriate suggestions for improvement.

[1680] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1681] Step 1:

[1682] Entering data

[1683] The user inputs the text data to be analyzed using the terminal. For example, the user inputs comments on social media or the contents of an email. The terminal then transmits the input text data to the server.

[1684] Input: Text data entered by the user (e.g. "I don't care about you")

[1685] Output: Text data sent from the terminal to the server

[1686] Step 2:

[1687] Data Preprocessing

[1688] The server preprocesses the received text data by removing unnecessary spaces and special characters and tokenizing the text using the Python NLTK (Natural Language Toolkit) library.

[1689] Input: Text data sent from the terminal

[1690] Data processing: removing unnecessary spaces and special characters, tokenizing text

[1691] Output: Preprocessed text data (e.g. "you", "something", "it doesn't matter")

[1692] Step 3:

[1693] sentiment analysis

[1694] The server performs sentiment analysis on the preprocessed text data, using a generative AI model (e.g., OpenAI's GPT-3) to determine whether the text contains offensive or insensitive nuances. The server inputs the following prompt sentence into the generative AI model:

[1695] "Please determine whether the following text contains offensive or insensitive connotations:

[1696] "I don't care about you."

[1697] A generative AI model takes this prompt and analyzes the sentiment of the text.

[1698] Input: Preprocessed text data, prompt

[1699] Data Computation: Sentiment Analysis with Generative AI Models

[1700] Output: Result of sentiment analysis (e.g. "offensive")

[1701] Step 4:

[1702] Output of results

[1703] The server returns the result of the sentiment analysis to the user, indicating whether the text contains offensive, indifferent, or other sentiments. The terminal displays the result received from the server to the user, for example, "This comment is offensive."

[1704] Input: Sentiment analysis results

[1705] Output: The result that is displayed to the user (e.g. "This comment is offensive")

[1706] The above is the specific processing flow of the program of this system.

[1707] (Application example 2)

[1708] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server", and the headset type terminal 314 will be referred to as a "terminal".

[1709] In modern text-based communication, messages containing aggressive or insensitive nuances are frequently sent and received, which can lead to deterioration of relationships between users and loss of psychological safety. In addition, there is a lack of means to detect messages containing these nuances in real time and warn users, making it difficult to improve the quality of communication.

[1710] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1711] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for suggesting improvements to make the text more "considerate" based on the detected nuances, and a means for analyzing messages in real time, detecting messages containing aggressive or indifferent nuances, and issuing a warning to the user. This allows the user to receive a warning before sending a message containing aggressive or indifferent nuances, thereby improving the quality of communication.

[1712] "Text-based communication" is a method of exchanging information using written text.

[1713] An "aggressive nuance" is a nuance that includes words or expressions that show hostility or aggression toward the other person.

[1714] An "indifferent nuance" is a nuance that includes words and expressions that show no interest or concern for the other person.

[1715] "Thoughtful writing" is writing that takes into consideration the other person's feelings and situation and includes words and expressions that show consideration and kindness.

[1716] "Analyzing messages in real time" means analyzing the content of a message immediately at the moment it is sent.

[1717] "Warning the user" means notifying the user about a detected message that contains offensive or insensitive nuances.

[1718] "Natural language processing technology" is a technology that enables computers to process and understand human language.

[1719] "Sentiment analysis" is a technology that analyzes emotions and nuances in text and classifies and evaluates its content.

[1720] In order to implement the present invention, the following system configuration and processing procedures are required.

[1721] System Configuration

[1722] 1. Hardware:

[1723] Smartphone: A device that allows users to engage in text-based communication.

[1724] Server: Provides the computational resources for sentiment analysis and nuance detection.

[1725] 2. Software:

[1726] Python: A programming language for running programs for sentiment analysis.

[1727] NLTK (Natural Language Toolkit): A library for natural language processing.

[1728] VADER (Valence Aware Dictionary and sEntiment Reasoner): A model for performing sentiment analysis.

[1729] Processing Steps

[1730] 1. Data Entry:

[1731] A user enters a text-based message using a smartphone.

[1732] 2. Data preprocessing:

[1733] The smartphone sends the entered message in text format to the server.

[1734] 3. Sentiment analysis:

[1735] The server uses NLTK's VADER sentiment analysis model to calculate a sentiment score for each received message.

[1736] 4. Nuance detection:

[1737] The server uses the sentiment score to determine whether the message contains offensive or insensitive overtones.

[1738] 5. Warning Issued:

[1739] Based on the detection results, the server will display a warning to the user about messages that contain offensive or insensitive nuances.

[1740] Examples

[1741] For example, if a user tries to send a message saying "You're really incompetent," the smartphone will send this message to the server. The server will use NLTK's VADER sentiment analysis model to calculate the sentiment score of this message and determine that it contains offensive nuances. As a result, the server will display a warning to the user saying "Offensive message detected."

[1742] Examples of prompt statements

[1743] Here is an example of a prompt to be input to the generative AI model:

[1744] Please judge whether the following messages contain offensive or insensitive connotations:

[1745] Message: "You're really incompetent."

[1746] In this way, users can be warned before sending messages that may be offensive or insensitive, improving the quality of communication.

[1747] The flow of the specific process in the application example 2 will be described with reference to FIG.

[1748] Step 1:

[1749] A user enters a text-based message using a smartphone.

[1750] Input: The text message entered by the user.

[1751] Output: The typed text message displayed on the smartphone.

[1752] Specific behavior: A user opens a chat application and enters text into the message input field.

[1753] Step 2:

[1754] The terminal transmits the input message in text format to the server.

[1755] Input: The text message entered by the user.

[1756] Output: The text message sent to the server.

[1757] Specific operation: The smartphone application sends the entered text message to the server as an HTTP request.

[1758] Step 3:

[1759] The server uses NLTK's VADER sentiment analysis model to calculate a sentiment score for each received message.

[1760] Input: The text message sent to the server.

[1761] Output: Sentiment scores (positive, negative and neutral scores).

[1762] Specific operation: The server inputs the received text message into the VADER sentiment analysis model and calculates positive, negative, and neutral scores.

[1763] Step 4:

[1764] The server uses the sentiment score to determine whether the message contains offensive or insensitive overtones.

[1765] Input: Sentiment score.

[1766] Output: Nuance judgment result (aggressive, indifferent, normal).

[1767] Specific operation: The server analyzes the emotion score, and if the negative score is high, it is determined to be "aggressive," if the neutral score is high, it is determined to be "indifferent," and otherwise it is determined to be "normal."

[1768] Step 5:

[1769] Based on the detection results, the server will display a warning to the user about messages that contain offensive or insensitive nuances.

[1770] Input: Nuance judgment result.

[1771] Output: The warning message that is displayed to the user.

[1772] Specific operation: The server sends the judgment result to the smartphone, and the smartphone application displays a warning to the user such as "Offensive message has been detected."

[1773] Example 3

[1774] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal".

[1775] In text-based communication, aggressive or indifferent nuances are often included, which can lead to misunderstandings and conflicts. In addition, there is a lack of concrete ways for users to improve their expressions to be more "considerate." This makes it difficult for users to communicate smoothly and considerately.

[1776] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1777] In this invention, the server includes a means for a user to input text, a means for a terminal to transmit the input text to the server, a means for the server to analyze the nuances of the text, a means for the server to generate improvement suggestions using a generative AI model, a means for the server to transmit the improvement suggestions to the terminal, and a means for the terminal to display the improvement suggestions to the user. This enables the user to receive specific suggestions for improving aggressive or insensitive expressions to "considerate" expressions.

[1778] "Text-based communication" is a method of exchanging information using written text.

[1779] "Nuance" refers to the subtle meanings and emotions contained in text.

[1780] "Offensive language" is language that has the potential to offend others or provoke negative feelings.

[1781] "Indifferent expressions" are language that makes the other person feel that you lack interest or consideration.

[1782] "Considerate expression" is language that shows consideration for the other person's feelings and position and shows kindness and understanding.

[1783] "User" means any individual or entity that uses the System.

[1784] A "Terminal" is a device through which a User accesses the System.

[1785] A "server" is a computer system that receives and processes requests from users.

[1786] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[1787] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate new text or suggestions.

[1788] "Improvement suggestions" are specific suggestions for replacing aggressive or insensitive language with more "compassionate" language.

[1789] The present invention is a system that detects aggressive or indifferent nuances in text-based communication and suggests ways to improve the text to be more “considerate.” A specific embodiment of this system will be described below.

[1790] First, the user inputs text using the terminal. For example, the user inputs the text "You are wrong." The terminal transmits the input text to the server. For transmission, a communication protocol such as an HTTP request is used.

[1791] The server analyzes the received text using natural language processing (NLP) techniques, specifically using NLP tools such as Python's NLTK library and spaCy to identify offensive or insensitive content in the text.

[1792] Next, the server uses a generative AI model to generate improvement suggestions based on the analysis results. For example, OpenAI's GPT-4 is used as the generative AI model. The server inputs the following prompt sentence to the generative AI model:

[1793] Improve your statement from "You're wrong" to something more "compassionate."

[1794] The generative AI model generates an improvement suggestion such as, "I understand your point of view, but I'm thinking about it from a slightly different perspective." The server sends this improvement suggestion to the terminal. For sending, a communication protocol such as HTTP response is used.

[1795] The terminal displays the received improvement suggestions to the user. The user can check the displayed improvement suggestions and modify the text as necessary. For example, the user can use the displayed improvement suggestions as they are or can make further modifications.

[1796] This system allows users to receive specific suggestions for improving aggressive or indifferent expressions into "kind" expressions. This allows smooth and kind communication to be realized. The flow of the identification process in the third embodiment will be described with reference to FIG. 15.

[1797] Step 1:

[1798] The user enters text.

[1799] The user inputs the text to be improved in the input field of the terminal, for example, "You are wrong." The input text is temporarily stored in the terminal's memory.

[1800] Step 2:

[1801] The terminal sends the entered text to the server.

[1802] The terminal sends the entered text as an HTTP request to the server. The input data is packaged in JSON format and sent over the network to the server. For example, the following JSON data is sent:

[1803] json

[1804] {

[1805] "text": "You are wrong"

[1806] }

[1807] Step 3:

[1808] The server analyzes the text for nuances.

[1809] The server analyzes the received text using natural language processing (NLP) techniques. Specifically, it uses NLP tools such as Python's NLTK library and spaCy to identify offensive or insensitive expressions in the text. The input data is the text, and the output data is nuanced information as a result of the analysis. For example, the part that says "wrong" is determined to be offensive.

[1810] Step 4:

[1811] The server generates improvement suggestions using the generative AI model.

[1812] The server uses a generative AI model (e.g., OpenAI's GPT-4) based on the analysis results to generate improvement suggestions for more "compassionate" expressions. The server inputs the following prompt sentence into the generative AI model:

[1813] Improve your statement from "You're wrong" to something more "compassionate."

[1814] The generative AI model generates improvement suggestions such as, "I understand your point of view, but I think about it from a slightly different perspective." The input data is the prompt sentence, and the output data is the improvement suggestions.

[1815] Step 5:

[1816] The server sends improvement suggestions to the device.

[1817] The server sends the generated improvement proposal to the terminal as an HTTP response. The output data is packaged in JSON format and sent to the terminal via the network. For example, the following JSON data is sent:

[1818] json

[1819] {

[1820] "suggestion": "I understand your point of view, but I think about it from a slightly different perspective."

[1821] }

[1822] Step 6:

[1823] The terminal displays improvement suggestions to the user.

[1824] The terminal displays the improvement suggestions received from the server to the user. The user can check the displayed improvement suggestions and modify the text as necessary. For example, the user can use the displayed improvement suggestions as they are or modify them further. The input data is the improvement suggestions, and the output data is the text displayed to the user.

[1825] (Application example 3)

[1826] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server", and the headset type terminal 314 will be referred to as a "terminal".

[1827] Text-based communication often contains aggressive or insensitive expressions, which can lead to misunderstandings and conflicts. Users may also unintentionally use expressions that hurt others. This leads to problems in that the quality of communication declines and psychological safety is lost. In addition, there is a lack of ways for users to receive concrete feedback to improve their expressions, which makes it difficult to improve communication skills.

[1828] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1829] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making improvement suggestions to make the text more "kind" based on the detected nuances, a means for analyzing messages sent by users in real time and displaying improvement suggestions, and a means for generating improvement suggestions using a generative AI model. This allows users to receive specific feedback for improving their own expressions in real time, improving the quality of communication and ensuring psychological safety.

[1830] "Text-based communication" is a method of exchanging information using written text.

[1831] "Nuance" refers to the subtle meanings and emotions contained in words and expressions.

[1832] "Offensive language" is any word or phrase intended to injure or provoke another person.

[1833] "Indifferent expressions" are cold words or phrases that show no interest in the other person's feelings or opinions.

[1834] "Thoughtful writing" is expression that shows respect and consideration for the other person's feelings and position.

[1835] "Improvement suggestions" are specific advice or proposed revisions to improve the current expression.

[1836] "Real-time analysis" means that the content of a message is analyzed immediately as the user types it.

[1837] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate new data or information.

[1838] "Feedback" means evaluation or advice provided to a user in response to their actions or expressions.

[1839] "Psychological safety" refers to an environment in which individuals feel safe to express their opinions and take action.

[1840] A system for implementing this invention includes means for conducting text-based communication, means for detecting aggressive or insensitive nuances, means for making suggestions for improvements to make the text more "kind" based on the detected nuances, means for analyzing messages sent by users in real time and displaying suggestions for improvement, and means for generating suggestions for improvement using a generative AI model.

[1841] Hardware and Software Configuration

[1842] Hardware:

[1843] Smartphone (iOS or Android)

[1844] software:

[1845] Python

[1846] OpenAI API

[1847] Data processing and calculation

[1848] server:

[1849] 1. A user types a text message using a smartphone.

[1850] 2. The entered message is sent to the server.

[1851] 3. The server analyzes the received message and detects offensive or insensitive nuances.

[1852] 4. Based on the detected nuances, a generative AI model (e.g., OpenAI’s GPT-3) is used to generate improvement suggestions for making the text more “kind.”

[1853] 5. Improvement suggestions are displayed in real time on the user's smartphone.

[1854] Examples

[1855] When a user types "You're wrong," the server analyzes the message to detect offensive nuances, then uses a generative AI model to generate a suggestion for improvement, such as "I understand your point of view, but I think about it from a slightly different perspective," and displays it to the user.

[1856] Example prompt for a generative AI model:

[1857] Please improve the following message to make it more considerate: You are wrong

[1858] In this way, users can get specific feedback to improve their expression in real time, which improves the quality of communication and ensures psychological safety.

[1859] The flow of the specific process in the application example 3 will be described with reference to FIG.

[1860] Step 1:

[1861] A user enters a text message using a smartphone.

[1862] Input: A text message entered by the user (e.g. "You are wrong")

[1863] Output: The entered text message is sent to the server by pressing the send button on the smartphone.

[1864] Step 2:

[1865] The server analyzes the messages it receives and detects offensive or indifferent nuances.

[1866] Input: Text message sent from smartphone

[1867] Data processing: Using natural language processing techniques, we analyze the content of messages to detect offensive or insensitive nuances.

[1868] Output: Detected nuance information (e.g. "contains offensive language")

[1869] Step 3:

[1870] Based on the nuances detected, the server uses a generative AI model to generate improvement suggestions for making the text more "considerate."

[1871] Input: Detected nuance information and original text message

[1872] Data computation: Send prompts to a generative AI model (e.g., OpenAI’s GPT-3) to generate improvement suggestions.

[1873] Output: Text of the suggestion for improvement (e.g. "I understand your point of view, but I have a slightly different perspective")

[1874] Step 4:

[1875] The server displays the generated improvement suggestions on the user's smartphone in real time.

[1876] Input: Text of improvement suggestion

[1877] Data processing: The improvement suggestions are converted into a data format for sending to the user's smartphone.

[1878] Output: Suggested improvements displayed on the user's smartphone (e.g., "I understand your point of view, but I have a slightly different perspective")

[1879] Step 5:

[1880] The user reviews the improvement suggestions and, if necessary, submits a revised message.

[1881] Input: User's correction message after reviewing the improvement suggestions

[1882] Output: The corrected text message is finally sent.

[1883] In this way, users can get specific feedback to improve their expression in real time, which improves the quality of communication and ensures psychological safety.

[1884] In addition, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1885] "Example 1"

[1886] In one embodiment of the present invention, the system includes an emotion engine that recognizes the emotion of a user. The emotion engine infers the user's emotional state from the text when the user engages in text-based communication. For example, if the user enters the text "I am angry," the emotion engine infers that the user is in an angry emotional state.

[1887] "Example 2"

[1888] In one embodiment of the present invention, the emotion engine adjusts the remediation suggestions based on the estimated emotional state. For example, if the user is estimated to be in an angry emotional state, the emotion engine adjusts the remediation suggestions for "kind" text to appropriately express the anger, thereby respecting the user's emotional state and reducing aggressive or indifferent overtones.

[1889] "Example 3"

[1890] In one embodiment of the present invention, the emotion engine tracks a user's emotional state over time to help improve the user's communication skills. For example, if the user is estimated to be in an angry emotional state over a period of time, the emotion engine uses that information to understand what makes the user angry and suggests ways to address it, thereby helping the user improve their communication skills.

[1891] The process flow of each embodiment will be described below.

[1892] "Example 1"

[1893] Step 1: Users engage in text-based communication.

[1894] Step 2: The emotion engine infers the emotional state of the user from the text. For example, from the text "I am angry", it infers that the user is in an angry emotional state.

[1895] "Example 2"

[1896] Step 1: The emotion engine estimates the user's emotional state.

[1897] Step 2: The system adjusts the remediation suggestions based on the estimated emotional state. For example, if the user is estimated to be in an angry emotional state, the system will adjust the remediation suggestions for "kind" sentences to better express the anger.

[1898] "Example 3"

[1899] Step 1: The emotion engine estimates the user's emotional state and records that information.

[1900] Step 2: The emotion engine tracks the user's emotional state over time.

[1901] Step 3: Based on the long-term emotional state tracking results, the system helps users improve their communication skills. For example, if the user is estimated to be in an angry emotional state for a certain period of time, the system uses the information to understand the reasons why the user feels angry and suggests ways to deal with it.

[1902] Example 1

[1903] Next, a description will be given of Example 1 of the embodiment 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal".

[1904] In modern text-based communication, it is difficult to accurately understand the user's emotions and nuances and provide appropriate feedback and suggestions for improvement. In particular, messages that contain aggressive or indifferent nuances can reduce the quality of communication and undermine psychological safety. It is also difficult for users to accurately recognize and appropriately express their own emotional state, so there is a demand for smoother communication.

[1905] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1906] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for making suggestions for improving the text to be more "kind" based on the detected nuances, a means for receiving a text message input by a user, a means for analyzing the text message using natural language processing technology, a means for estimating the user's emotional state based on the analysis results, and a means for notifying the user of the estimation results. This makes it possible to accurately understand the user's emotions and nuances and provide appropriate feedback and suggestions for improvement.

[1907] "Text-based communication" refers to communication methods that use written information, including email and messenger apps.

[1908] "Aggressive or indifferent nuances" refers to expressions contained in text messages that convey negative or indifferent feelings or attitudes.

[1909] "Suggestions for improving considerate language" are suggestions for changing text messages that contain aggressive or insensitive nuances into more positive and considerate language.

[1910] The "means for receiving a text message input by a user" is a function that enables the server to receive a text message sent by a user through a terminal.

[1911] "Means for analyzing text messages using natural language processing techniques" refers to techniques for understanding the content of text messages and analyzing their meaning and context, and includes tokenization, part-of-speech tagging, context analysis, etc.

[1912] The "means for inferring the emotional state of the user based on the analysis results" is a function for inferring the emotional state of the user by using the analysis results obtained by natural language processing technology.

[1913] The "means for notifying the user of the estimation result" is a function for notifying the user of the result of the estimated emotional state.

[1914] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[1915] The present invention is a system for text-based communication that analyzes text messages entered by a user, estimates the user's emotional state, and provides appropriate feedback and suggestions for improvement. A specific embodiment of the system will be described below.

[1916] Hardware and Software Used

[1917] server

[1918] The server plays a central role in receiving and analyzing text messages sent by users. The following software is installed on the server:

[1919] Natural Language Processing (NLP) technology: Google's Natural Language API, Microsoft's Text Analytics API

[1920] Emotion engines: IBM's Watson Tone Analyzer, Amazon's Comprehend

[1921] Terminal

[1922] A terminal is a device that allows a user to input text messages and communicate with a server, such as a smartphone or PC.

[1923] Program Processing

[1924] User Actions

[1925] A user inputs a text message using a text-based communication method such as email or a messenger app. For example, the user inputs the message "I am angry."

[1926] Device operation

[1927] The device transmits the text message entered by the user to the server, which requires an Internet connection.

[1928] Server Operations

[1929] The server receives text messages sent from the terminals and temporarily stores the received messages in a database.

[1930] The server then analyzes the text messages using natural language processing (NLP) techniques, such as Google's Natural Language API and Microsoft's Text Analytics API, to tokenize the text, tag parts of speech, and perform contextual analysis.

[1931] Based on the analysis results, the server uses an emotion engine to estimate the user's emotional state. This emotion engine uses IBM's Watson Tone Analyzer and Amazon's Comprehend. For example, it estimates that the user is in an angry emotional state from the keyword "angry."

[1932] The estimation results are sent from the server to the terminal using a secure communication protocol (e.g., HTTPS).

[1933] Device operation

[1934] The terminal displays the estimated emotional state received from the server to the user. For example, if the user inputs "I am angry," the terminal displays "You are in an angry emotional state."

[1935] Examples

[1936] As a concrete example, consider a case where a user uses a messenger app to send a message saying "I'm angry."

[1937] 1. A user types "I'm angry" into a messenger app.

[1938] 2. The device sends this message to the server.

[1939] 3. The server receives the message and stores it in a database.

[1940] 4. The server uses Google's natural language API to extract the keyword "angry."

[1941] 5. The server uses IBM's Watson Tone Analyzer to estimate that the user is in an angry emotional state.

[1942] 6. The server sends the estimation results to the device.

[1943] 7. The device displays to the user, "You are in an angry emotional state."

[1944] Examples of prompts to be input to a generative AI model include:

[1945] A user sends a message saying "I am angry". Analyze this message and infer the user's emotional state.

[1946] By feeding this prompt into a generative AI model, the model can infer the user's emotional state and generate an appropriate response.

[1947] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1948] Step 1:

[1949] The user enters a text message.

[1950] A user enters a text message using a text-based means of communication such as email or a messenger app. For example, the user enters the message "I am angry." The entered text message is the output of this step.

[1951] Step 2:

[1952] The device sends a text message to the server.

[1953] The terminal transmits the text message entered by the user to the server. This transmission requires an Internet connection. The input is the text message entered by the user and the output is the text message transmitted to the server.

[1954] Step 3:

[1955] A server receives the text message.

[1956] The server receives text messages sent from the terminal. The received messages are temporarily stored in a database. The input is the text message sent from the terminal, and the output is the text message stored in the database.

[1957] Step 4:

[1958] The server analyzes the text message using natural language processing (NLP) techniques.

[1959] The server analyzes the received text messages using natural language processing (NLP) technology. Google's natural language API and Microsoft's Text Analytics API are used for this analysis. Specifically, text tokenization, part-of-speech tagging, context analysis, etc. are performed. The input is the text message stored in the database, and the output is the analysis result.

[1960] Step 5:

[1961] The server uses an emotion engine to estimate the user's emotional state.

[1962] The server uses an emotion engine based on the analysis results to estimate the user's emotional state. IBM's Watson Tone Analyzer and Amazon's Comprehend are used for this emotion engine. For example, it estimates that the user is in an angry emotional state based on the keyword "angry." The input is the analysis results, and the output is the estimated emotional state.

[1963] Step 6:

[1964] The server transmits the estimation result to the terminal.

[1965] The server transmits the result of the emotional state estimated by the emotion engine to the terminal. A secure communication protocol (e.g., HTTPS) is used for this transmission. The input is the estimated result of the emotional state, and the output is the estimation result sent to the terminal.

[1966] Step 7:

[1967] The terminal displays the estimation results to the user.

[1968] The terminal displays the estimation result of the emotional state received from the server to the user. For example, if the user inputs "I am angry," the terminal displays "You are in an angry emotional state." The input is the estimation result sent from the server, and the output is the estimation result of the emotional state displayed to the user.

[1969] (Application example 1)

[1970] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server", and the headset type terminal 314 will be referred to as a "terminal".

[1971] In modern text-based communication, it is necessary to properly recognize the user's emotional state and respond quickly when an abnormal emotional state is detected. However, conventional systems often have difficulty accurately grasping the user's emotional state, and appropriate responses are often delayed. They also lack the ability to detect aggressive or indifferent nuances and suggest improvements to make the text more considerate. This may undermine the user's psychological safety.

[1972] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1973] In this invention, the server includes a means for performing text-based communication, a means for detecting aggressive or indifferent nuances in the text-based communication, a means for suggesting improvements to make the text more "kind" based on the detected nuances, a means for estimating the emotional state of the user, and a means for issuing an alert if the emotional state is abnormal. This makes it possible to monitor the emotional state of the user in real time and quickly respond when an abnormality is detected.

[1974] "Text-based communication" refers to the exchange of information using written text, including email and messenger apps.

[1975] "Aggressive or indifferent nuances" are words or expressions that show hostility or indifference toward the other person, and can have a negative impact on communication.

[1976] "Suggestions for improving language to be more considerate" refers to making suggestions to change text that contains aggressive or insensitive nuances into more friendly and considerate language.

[1977] "Natural language processing technology" is a technology that enables computers to understand, analyze, and generate human language, and is used to understand the meaning of text.

[1978] "Means for inferring a user's emotional state" refers to technologies or algorithms for analyzing a user's emotions from text-based communications and inferring that emotional state.

[1979] "Means of sending an alert" refers to a function for sending a notification to an administrator or security officer if a user's emotional state is determined to be abnormal.

[1980] As an embodiment of the present invention, the following system can be constructed.

[1981] First, the server provides a means for text-based communication. This means can be in the form of email, messenger apps, etc. When a user communicates using these means, the server receives the text and uses natural language processing technology to understand the meaning of the text.

[1982] Next, the server has a means for detecting offensive or insensitive nuances in the text-based communication, which is realized using a natural language processing library (e.g., the nltk library), and specifically, a SentimentIntensityAnalyzer is used to analyze the sentiment score of the text.

[1983] In addition, the server provides a means to suggest improvements to make sentences more "kind" based on the detected nuances. This means uses a generative AI model to generate appropriate suggestions. For example, it generates prompt sentences to convert aggressive expressions into friendly expressions.

[1984] The server also includes a means for estimating the user's emotional state, which includes techniques and algorithms for analyzing the user's emotions from text-based communications and estimating the emotional state. Specifically, the server uses a SentimentIntensityAnalyzer to analyze the user's emotion score.

[1985] Finally, the server provides a means for sending an alert if the emotional state is abnormal, including functionality for sending a notification to an administrator or security officer if the emotional score falls below a certain threshold, by sending an email using the smtplib library.

[1986] For example, if a user types "I'm so angry," the emotion engine will recognize this text as a negative emotion and send an alert to the administrator, with the following example prompt:

[1987] If a user types "I am so angry", the sentiment engine should recognize this text as a negative sentiment and send an alert to an admin.

[1988] In this way, emotion-aware security apps can monitor a user's emotional state in real time and respond quickly if anomalies are detected.

[1989] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1990] Step 1:

[1991] The users engage in text-based communication.

[1992] Input: Text typed by the user (e.g. "I'm so angry")

[1993] Output: The text data entered

[1994] Specific action: A user types and sends text via email or a messenger app.

[1995] Step 2:

[1996] The server receives the entered text.

[1997] Input: Text data submitted by the user

[1998] Output: Received text data

[1999] Specific operation: The server receives text data through a mail server or messenger server.

[2000] Step 3:

[2001] The server uses natural language processing techniques to analyze the meaning of the text.

[2002] Input: Received text data

[2003] Output: Sentiment score of the text (e.g. {'neg': 0.8, 'neu': 0.1, 'pos': 0.1, 'compound': -0.7})

[2004] Specific behavior: The server uses the SentimentIntensityAnalyzer from the nltk library to calculate the sentiment score of the text.

[2005] Step 4:

[2006] The server detects nuances in the text and determines whether it contains offensive or insensitive content.

[2007] Input: Sentiment score of the text

[2008] Output: Nuance detection result (e.g. aggressive)

[2009] Specific operation: The server analyzes the emotion scores and detects aggressive nuances when the negative scores are high.

[2010] Step 5:

[2011] Based on the nuances detected by the server, suggestions for improving the text to be more "considerate" are generated.

[2012] Input: Nuance detection results, original text data

[2013] Output: Text of the improvement suggestion (e.g. "I am a bit dissatisfied")

[2014] How it works: The server uses a generative AI model to generate prompts that convert aggressive expressions into friendly ones.

[2015] Step 6:

[2016] The server estimates the user's emotional state.

[2017] Input: Sentiment score of the text

[2018] Output: Estimated emotional state (e.g. anger)

[2019] Specific operation: The server estimates the user's emotional state based on the emotion score.

[2020] Step 7:

[2021] The server issues an alert if the emotional state is abnormal.

[2022] Input: Emotion state estimation result

[2023] Output: Alert notification (e.g. email notification to administrator)

[2024] Specific behavior: If the server determines that the emotional state is abnormal (e.g., very angry), it uses the smtplib library to send an alert via email to the administrator.

[2025] In this way, emotion-aware security apps can monitor a user's emotional state in real time and respond quickly if anomalies are detected.

[2026] Example 2

[2027] Next, a second embodiment of the second form will be described. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal".

[2028] In text-based communication, aggressive or indifferent nuances may be included, which may reduce the quality of communication. Also, if users cannot express their emotional state appropriately, misunderstandings and conflicts may occur. This may worsen relationships between users and damage psychological safety.

[2029] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for acquiring a text-based message created by a user, a means for recognizing the user's emotion from the message using an emotion engine, and a means for making an improvement proposal for the message using a prompt sentence for instructing to suggest an improvement of the message based on the recognized emotion of the user when the message contains an aggressive nuance, and a generative AI model. This makes it possible to improve the quality of communication by mitigating aggressive or indifferent nuances while respecting the emotional state of the user. In addition, the means for making an improvement proposal for the message in the server determines whether the message created by the user contains an aggressive nuance based on an exchange between the message created by the user and a message created by the other party, and makes an improvement proposal for the message using a prompt sentence for instructing to suggest an improvement of the message based on the recognized emotion of the user when the aggressive nuance is included, and a generative AI model.

[2030] "Text-based communication" refers to the exchange of information using written information.

[2031] "Aggressive or indifferent nuances" refer to words and expressions that show hostility or indifference toward the other person.

[2032] "Thoughtful writing" refers to words and expressions that take into consideration the other person's feelings and situation and show consideration and kindness.

[2033] "Sentiment analysis technology" refers to technology that analyzes words, phrases, and context in text and infers emotional states.

[2034] "Words, phrases, and context" refer to the basic linguistic elements that make up a text and their relationships.

[2035] "Emotional state" refers to the type and intensity of emotion expressed through text.

[2036] "Improvement suggestions" refer to suggesting more appropriate expressions or words based on detected nuances.

[2037] "Natural language processing technology" refers to technology that enables computers to understand and generate human language.

[2038] A "generative AI model" is a model for generating text using artificial intelligence.

[2039] "Inference" refers to predicting a state or outcome based on given data.

[2040] This invention is a system that detects aggressive or indifferent nuances in text-based communication and suggests ways to improve the text to be more "kind" based on the emotional state. This system works in cooperation with a server, a terminal, and a user.

[2041] The server uses sentiment analysis techniques to analyze words, phrases, and context in the text to infer the emotional state. Specifically, it uses a sentiment analysis engine built using Python and TENSORFLOW. The sentiment analysis engine uses pre-trained generative AI models (e.g., BERT and GPT-3) to analyze the text.

[2042] The device is responsible for sending the text entered by the user to the server. When the user enters text into the input field of the device, the device sends the text in JSON format to the server. The server analyzes the received text and estimates the emotional state.

[2043] For example, if a user enters "I don't understand your opinion at all," the server passes this text to a sentiment analysis engine. The sentiment analysis engine analyzes the text and infers that the user is in an angry emotional state. Next, the server uses a prompt sentence that instructs the server to suggest improvements to the message based on the recognized user's emotions if the message contains aggressive nuances, and a generative AI model to make suggestions for improving the message, generating suggestions for improving the message to be more "kind." For example, the server generates a suggestion such as, "Could you please explain your opinion in more detail?"

[2044] Here is an example of a prompt to be input to the generative AI model:

[2045] "Input text: "I don't understand your opinion at all" Emotional state: "Anger" Determine whether the input text contains offensive nuances, and if so, suggest improvements to the message based on the perceived emotional state of the user."

[2046] Example of a suggestion for improvement: "Can you explain your opinion in more detail?"

[2047] The server sends the generated improvement proposal in JSON format to the terminal. The terminal displays the received improvement proposal to the user. The user can check the displayed improvement proposal and modify the text if necessary.

[2048] In this way, it is possible to improve the quality of communication by respecting the user's emotional state while mitigating aggressive or indifferent nuances. When making an improvement suggestion, it is preferable to further consider the interaction between the user's message and the other party's message. In this case, the server judges whether the message created by the user contains an aggressive nuance based on the interaction between the message created by the user and the message created by the other party, and if an aggressive nuance is included, makes an improvement suggestion for the message using a prompt sentence instructing to suggest an improvement of the message based on the recognized emotion of the user and a generative AI model. Here is an example of a prompt to be input to the generative AI model:

[2049] After entering the exchange between your message and the other person's message, "Input text: "I don't understand your opinion at all" Emotional state: "Anger" Given the messaging conversation, determine whether the input text contains offensive nuances and, if so, suggest improvements to the message based on the perceived emotional state of the user.”

[2050] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2051] Step 1:

[2052] A user inputs text into an input field on a terminal. For example, the user inputs "I don't understand your opinion at all." This input text becomes the input data for the next processing step.

[2053] Step 2:

[2054] The terminal converts the text entered by the user into JSON format and sends it to the server as an HTTP request. At this time, the input text is sent to the server as JSON format data.

[2055] Step 3:

[2056] The server parses the JSON data received from the terminal and extracts the text part. Specifically, it extracts the "text" field from the received JSON data and passes its contents to the next processing step. This extracted text becomes the input data for the next processing step.

[2057] Step 4:

[2058] The server passes the extracted text to a sentiment analysis engine, which is built using Python and TensorFlow and analyzes the text using pre-trained generative AI models (e.g., BERT and GPT-3). The sentiment analysis engine considers words, phrases, and context in the text to infer an emotional state. This inferred emotional state becomes the input data for the next processing step.

[2059] Step 5:

[2060] The server gen...

Claims

1. means for obtaining a text-based message created by a user; means for recognizing a user's emotion from the message using an emotion engine; A means for making a suggestion to improve the message using a prompt sentence that instructs the system to suggest an improvement to the message based on the recognized emotion of the user when the message includes an aggressive nuance, and a generative AI model; A system including:

2. The system of claim 1 , further comprising means for issuing an alert if the recognized emotion is anomalous.

3. The system according to claim 1, wherein the means for making improvement suggestions for the message determines whether the message created by the user contains an aggressive nuance based on an exchange between the message created by the user and a message created by the other party, and if the aggressive nuance is included, makes an improvement suggestion for the message using a prompt sentence instructing to suggest an improvement to the message based on the recognized emotions of the user and a generative AI model.

Images