System

Abstract

A system is provided.SOLUTION: This system is provided with a means for collecting contents transmitted by a user, a means for analyzing the collected contents, a means for checking contradiction with the past utterance based on the analyzed result, a means for checking whether or not any inappropriate contents are included in the utterance, and a means for notifying the user of the analyzed result.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] In recent years, opportunities for individuals and companies to disseminate information electronically through a variety of means, including social media, lectures, and books, have been increasing. In these circumstances, it is extremely important to verify whether the information disseminated is consistent and does not contradict previous statements. However, detecting inconsistencies within a vast amount of data and identifying inappropriate statements is extremely time-consuming, making it difficult for humans to do this efficiently by hand. To solve this problem, technology is needed to automatically detect inconsistencies and inappropriate content in disseminated information and provide feedback. [Means for solving the problem]

[0005] To solve the above problems, the present invention provides the following means. A system is provided that includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, and means for notifying the user of the analysis results. This system uses natural language processing technology to analyze the meaning of posts and identify inconsistencies with past posts. Furthermore, because the system collects data from a variety of information sources, such as social media posts, lecture records, and book data, it is possible to comprehensively monitor the content posted by users. This makes it easier for users to maintain consistency in their posts and avoid inappropriate comments.

[0006] "Means for collecting content posted by users" refers to a function that automatically acquires text and audio data posted by users through media such as social media, lectures, and books, and stores it in a database.

[0007] "Means for analyzing collected content" refers to algorithms and software that use natural language processing technology to analyze the content of collected text data and audio data and understand its context and meaning.

[0008] "Means for checking for inconsistencies with past statements based on the analysis results" is a function that compares the analyzed data with a past database and evaluates whether there are any inconsistencies in the content or context.

[0009] "Means for checking whether statements contain inappropriate content" refers to algorithms that scan the analyzed data to determine whether it contains inappropriate words or phrases, and if necessary, analyze the context to assess appropriateness.

[0010] "Means for notifying users of analysis results" refers to a function that notifies users of the results to their devices and provides detailed feedback when inconsistencies or inappropriate comments are identified. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0012] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0013] First, the terms used in the following description will be explained.

[0014] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0015] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0016] 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.

[0017] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0018] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0019] [First embodiment]

[0020] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0021] 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.

[0022] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0023] 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.

[0024] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0025] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0026] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0027] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0028] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0029] 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.

[0030] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0031] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0032] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, and notification.

[0033] Program processing

[0034] Data collection

[0035] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (application programming interface). The audio data from lectures is converted into text data using voice recognition technology, and all data is stored in a single database.

[0036] Text analytics

[0037] The server then analyzes the collected data, using natural language processing (NLP) techniques to analyze the text data and understand its context and meaning. This analysis includes tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[0038] Inconsistency check

[0039] Based on the analyzed data, the server compares current and past statements, using similarity calculations and conflict detection algorithms to identify inconsistencies. For example, a statement that "environmental protection is important" could be deemed a contradiction with a past statement that "environmental issues are not a concern."

[0040] Inappropriate remarks check

[0041] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for words and phrases that are deemed inappropriate and analyzes their context. For example, if the word "idiot" is used, it checks the context before and after it to determine whether it is truly inappropriate.

[0042] Notification of results

[0043] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are identified, a detailed report is sent to the user's device. The user receives a notification through their device and can view the report to identify the problems with their own comments.

[0044] Specific examples

[0045] scenario

[0046] User C posted on social media that "technological advances will make society better." However, in the past, the same User C stated in a lecture that "technological advances will have a negative impact on society."

[0047] Data collection

[0048] The server collects User C's latest SNS posts via API and stores them in a database. It also collects audio data from the lecture, converts it into text using speech recognition technology, and stores it.

[0049] Text analytics

[0050] The server analyzes the collected social media posts and speeches using an NLP engine to understand their meaning and context, and performs processes such as tokenization, keyword extraction, and sentiment analysis.

[0051] Inconsistency check

[0052] The server compares the current social media post, "Technological advances make society better," with the past speech, "Technological advances have a negative impact on society," and determines that there is a contradiction.

[0053] Inappropriate remarks check

[0054] In this example, there is no particularly inappropriate language, so we will skip this step.

[0055] Notification of results

[0056] The server generates a report summarizing the analysis results that identified the contradiction and notifies the terminal of User C. User C views this report and confirms that there is a contradiction between the past and present statements.

[0057] In this way, the system for implementing the present invention allows users to maintain consistency in the content they post and avoid making inappropriate comments, thereby preventing misunderstandings and trouble.

[0058] The processing flow will be explained below.

[0059] Step 1:

[0060] The server uses an API that connects to the user's social media account to periodically check whether a new post has been made. If a new post is detected, the text data is stored in a database.

[0061] Step 2:

[0062] The server collects the audio data of the lecture in real time. The collected audio data is converted into text data using speech recognition technology. The converted text data is stored in a database.

[0063] Step 3:

[0064] Users upload digital files of book data to the server, which then converts the uploaded book data into text format and stores it in a database.

[0065] Step 4:

[0066] The server performs preprocessing on the collected text data, including tokenization, stop word removal, and normalization, before passing the preprocessed text data to a natural language processing engine.

[0067] Step 5:

[0068] The server uses natural language processing techniques to semantically analyze the text data, including keyword extraction, sentiment analysis, sentence segmentation, and syntactic analysis, to understand the specific context and meaning of each utterance.

[0069] Step 6:

[0070] The server compares current and past utterance data. Using similarity calculations and contradiction detection algorithms, it identifies contradictions between utterances. For example, if a contradiction is detected between the utterances "Technological advances improve society" and "Technological advances have a negative impact on society," it will identify it.

[0071] Step 7:

[0072] The server scans the text data using a specialized dictionary containing inappropriate words and phrases, and analyzes the context of identified inappropriate words and phrases to determine whether they are in fact inappropriate.

[0073] Step 8:

[0074] The server generates a report summarizing the analysis and evaluation results, including specific areas of inconsistency and inappropriate comments, and sends the report to the user's device.

[0075] Step 9:

[0076] Users receive a notification from the server via their device, which includes a link to a detailed report that allows them to identify the issues with their comments.

[0077] The above is a concrete flow of a series of processing steps for collecting and analyzing user comments, checking for inconsistencies and inappropriate content, and notifying the results.

[0078] Example 1

[0079] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0080] Inconsistent content posted by users or inappropriate comments can lead to misunderstandings and problems. This problem poses a challenge, as it can undermine the user's credibility and lower social credibility. Another problem is that manually checking the consistency and appropriateness of comments by users is time-consuming and inefficient.

[0081] 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.

[0082] In this invention, the server includes means for collecting content posted by users, means for storing the collected content in a database, means for analyzing the stored content using natural language processing technology, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, and means for notifying the user of the analysis results. This enables users to maintain consistency in the content they post and avoid making inappropriate posts.

[0083] "User" refers to any individual or organization that transmits information or opinions.

[0084] "Content" refers to a collection of user-generated text, audio, or digital data.

[0085] "Means of collection" refers to methods for collecting user-generated content using APIs and data acquisition methods of various platforms.

[0086] "Database" refers to a data storage system for storing and managing collected data.

[0087] "Means for storage" refers to the method for storing and managing collected data in a database.

[0088] "Natural language processing technology" refers to technology that analyzes text data and allows it to understand its meaning and context.

[0089] "Means of analysis" refers to a method of analyzing collected text data using natural language processing technology to extract specific information and context.

[0090] "Means for checking for inconsistencies" refers to a method of comparing the latest statements with past statements based on analyzed data to determine whether there are any inconsistencies.

[0091] "Measures to check for inappropriate content" refers to methods that use specialized dictionaries and algorithms to determine whether a comment contains inappropriate words or phrases.

[0092] "Means for notifying" refers to a method for transmitting the analysis results to the user's terminal and notifying the user.

[0093] "Tokenization" refers to the process of dividing text data into words or phrases.

[0094] "Stop word removal" refers to the process of removing common words that are not necessary for analysis from the text to be analyzed.

[0095] "Keyword extraction" refers to the process of extracting important words and phrases from text data.

[0096] "Sentiment analysis" refers to the process of determining emotions and emotional direction from the context of text data.

[0097] "Sentence segmentation" refers to the process of dividing text data into sentences or segments.

[0098] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and provides feedback. The system mainly comprises a server, a terminal, and a user. A specific embodiment of this invention will be described below.

[0099] System Overview

[0100] The system involves a server collecting user posts, analyzing them using natural language processing, detecting inconsistencies and inappropriate content, and notifying the user of the results via their device. The server uses the SNS API, voice recognition technology, a database management system, and a natural language processing library.

[0101] Data collection and storage

[0102] The server collects user posts using the SNS API and uses the Google Cloud Speech-to-Text API to convert the voice data. The collected data is then stored in a database such as MySQL.

[0103] Text analytics

[0104] The server analyzes the collected text data using natural language processing techniques, such as tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation, using libraries such as SpaCy and NLTK.

[0105] Inconsistency check and inappropriate remark check

[0106] The server compares the most recent and past comments based on the analysis results to identify inconsistencies. Similarity calculations are performed using cosine similarity, etc. To check for inappropriate comments, a dedicated dictionary is used to scan for specific words and phrases, and the context is also analyzed.

[0107] Notification of results

[0108] The server compiles the results of the analysis of detected inconsistencies and inappropriate comments and generates a report. This report is sent to the user's device, allowing the user to check the problematic aspects of the comments. Notifications are sent to the user's device using an API.

[0109] Specific examples

[0110] A specific example of the operation of the system is shown below.

[0111] scenario

[0112] A user posted on social media that "technological advances will improve society." However, in the past, the same user stated in a lecture that "technological advances will have a negative impact on society."

[0113] Data collection

[0114] The server collects the latest posts via the SNS API and stores them in a database. It also collects audio data from lectures, converts it into text using the Google Cloud Speech-to-Text API, and stores it.

[0115] Text analytics

[0116] The server analyzes the collected social media posts and lecture comments using natural language processing technology (e.g., SpaCy) to understand the meaning and context of each.

[0117] Inconsistency check

[0118] The server compared the social media post, "Technological advances make society better," with the speech, "Technological advances have a negative impact on society," and determined that there was a contradiction.

[0119] Inappropriate remarks check

[0120] In this example, there is no particularly inappropriate language, so we will skip this step.

[0121] Notification of results

[0122] The server generates a report summarizing the analysis results that identified the inconsistencies and sends it to the user's device. The user can view this report and check the inconsistencies in the comments.

[0123] Prompt Sentence Examples

[0124] "Write a program that compares a user's social media posts with their past statements to identify inconsistencies."

[0125] This invention allows users to ensure consistency in the content they post and avoid making inappropriate comments, thereby preventing misunderstandings and trouble.

[0126] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0127] Step 1: Data collection

[0128] The server collects content posted by users on social media, audio data from lectures, and digital data from books. Specifically, it acquires data using various APIs (such as APIs for social networking services and voice recognition APIs).

[0129] Input: Social media posts, lecture audio files, digital book data

[0130] Data processing: Convert the audio data from the lecture into text using the Google Cloud Speech-to-Text API.

[0131] Output: Text data

[0132] Specific operation: The server uses the SNS API to collect the latest user posts, sends the audio file of the lecture called audio_file.wav to the Google Cloud Speech-to-Text API, and obtains the text data "Technological advances have a negative impact on society."

[0133] Step 2: Save your data

[0134] The server stores the collected data in a database, where the data is organized for each user and stored in a collated format.

[0135] Input: Collected text data

[0136] Data processing: structuring data

[0137] Output: A structured database

[0138] What happens: The server connects to the database and inserts new data, for example by executing an SQL query like INSERT INTO user_data (user_id, content, type) VALUES (123, 'Technological advances make society better', 'SNS').

[0139] Step 3: Text analysis

[0140] The server analyzes the stored text data using natural language processing (NLP) techniques, including tokenization, stop-word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[0141] Input: Saved text data

[0142] Data processing: tokenization, stop word removal, keyword extraction, sentiment analysis, sentence segmentation

[0143] Output: Analysis results

[0144] Specific operation: The server uses the SpaCy library to analyze text data. For example, it loads the model with spacy.load("en_core_web_sm") and analyzes the text with nlp("Technological advances make society better"). It extracts keywords from the analysis results and creates a list such as "technology", "advancement", "society", and "make it better".

[0145] Step 4: Check for inconsistencies

[0146] The server compares the latest and past comments based on the analysis results, and identifies inconsistencies using similarity calculations (cosine similarity) and conflict detection algorithms.

[0147] Input: Text analysis results

[0148] Data processing: Similarity calculation, conflict detection

[0149] Output: Contradiction judgment result

[0150] Specific operation: The server calculates the vectors of past and current statements and evaluates their similarity. For example, it calculates the similarity using cosine_similarity(vector_a, vector_b), and if the result is below a certain threshold, it determines that there is a contradiction.

[0151] Step 5: Check for inappropriate comments

[0152] The server uses a dedicated dictionary to check whether the comment contains inappropriate words or phrases, and also analyzes the context to determine whether the comment is truly inappropriate.

[0153] Input: Text analysis results

[0154] Data processing: dictionary matching, context analysis

[0155] Output: Inappropriate judgment result

[0156] What it does: The server splits the text into words and compares each word to a list of inappropriate words, e.g., if word in inappropriate_words: flag_as_inappropriate(word) to detect inappropriate words.

[0157] Step 6: Notification of results

[0158] The server compiles the results of the analysis of inconsistencies and inappropriate comments, generates a report, and sends the report to the user's device, where the user can check it.

[0159] Input: Conflict judgment result, inappropriate judgment result

[0160] Data Processing: Report Generation

[0161] Output: Notification messages, reports

[0162] Specific operation: The server creates a report based on the analysis results and notifies the user via the API on the user's device. For example, it calls the PUT / user_notifications API and sends the notification data {"user_id": 123, "message": "Past and current statements are inconsistent"}.

[0163] (Application example 1)

[0164] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0165] In physical stores, it is difficult for staff to provide consistent information to customers. Correcting statements made in the past can damage customer trust. Furthermore, there is no way to check past statements in real time, so there is a risk of repeating the same mistake. This raises concerns about lower customer satisfaction and worsening operational efficiency.

[0166] 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.

[0167] In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for notifying the user of the analysis results, and means for collecting and analyzing data in real time based on the user's posts. This allows customer service staff to maintain consistency in customer service, prevent the provision of incorrect information, and maintain customer trust.

[0168] "Means for collecting user-generated content" refers to processes and technologies that automatically collect data entered or transmitted by users in the form of information networks, oral presentations, documents, etc.

[0169] "Means for analyzing collected content" are the processes and techniques used to analyze collected data and identify its meaning and context.

[0170] "Means for checking for inconsistencies with past statements based on the analysis results" refers to the process and techniques for comparing the analysis results with previous statements to see if there are any inconsistencies.

[0171] "Measures to check speech for inappropriate content" are processes and techniques that scan speech for inappropriate words, phrases, or context.

[0172] "Means for notifying the user of the analysis results" refers to the process and technology for notifying the user of the results obtained by the analysis so that the user can check them.

[0173] "Means for collecting and analyzing data in real time based on user comments" refers to the process and technology for quickly collecting comments made by users on the spot and analyzing them immediately.

[0174] This invention is a system that analyzes the content of user comments in a physical store in real time and provides feedback on the results. This system includes the following main processes.

[0175] Data collection

[0176] The server acquires voice data to collect what the user is saying. The voice data is collected through a microphone in the smart glasses and converted into text data using voice recognition technology. This process uses the Google Cloud Speech-to-Text API. The collected data is then stored in a database.

[0177] Text analytics

[0178] The server analyzes the collected text data using natural language processing (NLP) techniques, using Python libraries such as NLTK and spaCy, and performs processes such as tokenization, stop word removal, keyword extraction, and sentiment analysis.

[0179] Inconsistency check

[0180] Based on the analyzed data, the server compares the current utterance with the past utterances, and uses similarity calculations and conflict detection algorithms to identify inconsistencies. In this process, the Python scikit-learn library is used, for example, to calculate cosine similarity.

[0181] Inappropriate remarks check

[0182] The server checks for inappropriate words and phrases. It uses a Python dictionary database to scan for potentially inappropriate words and phrases, and analyzes the context to determine whether they are truly inappropriate.

[0183] Notification of results

[0184] If any inconsistencies or inappropriate comments are identified, the server will summarize the analysis results and notify the user. The notification will be displayed in real time on the smart glasses display. The smart glasses application will be developed for Android or iOS.

[0185] Specific examples

[0186] Consider a case where a user tells a customer in a physical store, "The point card is valid for one year," but in the past has said, "The point card is only valid for six months."

[0187] Prompt Sentence Examples

[0188] "Check your customer interaction records and see if there are any contradictions between current and past statements. Past statement: 'The loyalty card is only valid for six months.' Current statement: 'The loyalty card is valid for one year.'"

[0189] This system allows users to maintain consistency in customer service and prevent the provision of inappropriate information, which is expected to improve customer satisfaction and operational efficiency.

[0190] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0191] Step 1:

[0192] Data collection

[0193] The server collects what the user says through a microphone installed in the smart glasses. The collected voice data is converted into text data using the Google Cloud Speech-to-Text API. The converted text data is then stored in a database.

[0194] Input: Audio data

[0195] Processing: Convert speech to text (using Google Cloud Speech-to-Text API)

[0196] Output: Text data (stored in database)

[0197] Step 2:

[0198] Text analytics

[0199] The server analyzes the collected text data using Python natural language processing libraries (such as NLTK or spaCy), specifically performing tokenization, stop word removal, keyword extraction, and sentiment analysis.

[0200] Input: Text data

[0201] Processing: Tokenization, stopword removal, keyword extraction, sentiment analysis (using NLTK and spaCy libraries)

[0202] Output: Parsed text data

[0203] Step 3:

[0204] Inconsistency check

[0205] The server compares current and past statements based on the parsed text data, using the Python scikit-learn library to identify inconsistencies using algorithms such as cosine similarity calculations.

[0206] Input: Analyzed text data, past text data

[0207] Processing: Similarity calculation (using the scikit-learn library)

[0208] Output: Conflicts

[0209] Step 4:

[0210] Inappropriate remarks check

[0211] The server then scans the parsed text data for inappropriate words and phrases, using a Python dictionary database and analyzing the context to make its decisions.

[0212] Input: Parsed text data

[0213] Processing: Scanning for inappropriate words and phrases (using dictionary database)

[0214] Output: Whether or not there is inappropriate remarks

[0215] Step 5:

[0216] Notification of results

[0217] Based on the results of the inconsistencies and inappropriate comments, the server compiles the analysis results and notifies the user's smart glasses in real time, which are displayed on the smart glasses' display.

[0218] Input: Whether there are contradictions or not, whether there are inappropriate comments or not

[0219] Processing: Creating and sending analysis results

[0220] Output: Feedback notification to the smart glasses display

[0221] This series of processes allows users to instantly correct any inconsistencies with past comments or inappropriate comments when dealing with customers in a physical store.

[0222] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0223] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and then analyzes the user's emotions using an emotion engine and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, emotion analysis, and notification.

[0224] Program processing

[0225] Data collection

[0226] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (application programming interface). The audio data from lectures is converted into text data using voice recognition technology, and all data is stored in a single database.

[0227] Text analytics

[0228] The server then analyzes the collected data, using natural language processing (NLP) techniques to analyze the text data and understand its context and meaning. This analysis includes tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[0229] Inconsistency check

[0230] Based on the analyzed data, the server compares current and past statements. It uses similarity calculations and conflict detection algorithms to identify contradictions. For example, a statement that "technological advances make society better" could be deemed a contradiction with a past statement that "technological advances have a negative impact on society."

[0231] Inappropriate remarks check

[0232] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for words and phrases that are deemed inappropriate and analyzes their context. For example, if the word "idiot" is used, it checks the context before and after it to determine whether it is truly inappropriate.

[0233] Emotion analysis

[0234] A distinctive feature of this invention is that the server uses an emotion engine to analyze the user's emotions from text data. This emotion engine uses NLP technology to identify emotions from the content of the user's speech. For example, it identifies emotions such as "very happy" or "very angry." This allows for a detailed understanding of the emotions expressed by the user's speech.

[0235] Notification of results

[0236] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are detected, a detailed report including the results of the sentiment analysis is sent to the user's device. The user receives the notification via their device and can view the report to identify problems with their own comments and fluctuations in sentiment.

[0237] Specific examples

[0238] scenario

[0239] User D posted on social media that "technological advances will make society better." However, in the past, the same User D stated at a lecture that "technological advances will have a negative impact on society." The social media post also expresses the emotion of "being very happy."

[0240] Data collection

[0241] The server collects the latest SNS posts from User D via API and stores them in a database. It also collects audio data from the lecture, converts it into text using speech recognition technology, and stores it.

[0242] Text analytics

[0243] The server analyzes the collected social media posts and speeches using an NLP engine to understand their meaning and context, and performs processes such as tokenization, keyword extraction, and sentiment analysis.

[0244] Inconsistency check

[0245] The server compares the current social media post, "Technological advances make society better," with the past speech, "Technological advances have a negative impact on society," and determines that there is a contradiction.

[0246] Inappropriate remarks check

[0247] In this example, there is no particularly inappropriate language, so we will skip this step.

[0248] Emotion analysis

[0249] The server analyzes the emotion "very happy" from the latest social media posts and adds it to the analysis results.

[0250] Notification of results

[0251] The server generates a report summarizing the analysis results that identified the contradictions and the results of the emotion analysis, and notifies the device of User D. User D views this report and confirms that there is a contradiction between his past and present statements, and that his latest post expresses the emotion of "very happy."

[0252] In this way, the system for implementing this invention not only allows users to maintain consistency in their speech and emotions and avoid inappropriate speech, but also allows users to accurately grasp emotional fluctuations, thereby preventing misunderstandings and troubles and deepening understanding of emotional states.

[0253] The processing flow will be explained below.

[0254] Step 1:

[0255] The server uses an API that connects to the user's social media account to periodically check whether a new post has been made. If a new post is detected, the text data is stored in a database.

[0256] Step 2:

[0257] The server collects the audio data of the lecture in real time. The collected audio data is converted into text data using speech recognition technology. The converted text data is stored in a database.

[0258] Step 3:

[0259] Users upload digital files of book data to the server, which then converts the uploaded book data into text format and stores it in a database.

[0260] Step 4:

[0261] The server performs preprocessing on the collected text data, including tokenization, stop word removal, and normalization, before passing the preprocessed text data to a natural language processing engine.

[0262] Step 5:

[0263] The server uses natural language processing techniques to semantically analyze the text data, including keyword extraction, sentiment analysis, sentence segmentation, and syntactic analysis, to understand the specific context and meaning of each utterance.

[0264] Step 6:

[0265] The server compares current and past utterance data. Using similarity calculations and contradiction detection algorithms, it identifies contradictions between utterances. For example, if a contradiction is detected between the utterances "Technological advances improve society" and "Technological advances have a negative impact on society," it will identify it.

[0266] Step 7:

[0267] The server scans the text data using a specialized dictionary containing inappropriate words and phrases, and analyzes the context of identified inappropriate words and phrases to determine whether they are in fact inappropriate.

[0268] Step 8:

[0269] The server uses an emotion engine to analyze the emotions in the user's comments. The emotion engine uses NLP technology to identify, for example, whether the comment contains the emotion "very happy."

[0270] Step 9:

[0271] The server generates a report summarizing the analysis and evaluation results, including identified inconsistencies, inappropriate comments, and sentiment analysis results, and sends the report to the user's device.

[0272] Step 10:

[0273] Users receive notifications from the server via their devices, which include a link to a detailed report that allows them to identify problems with their own comments and fluctuations in sentiment.

[0274] The above is a concrete flow of a series of processing steps that collects and analyzes user comments, checks for inconsistencies and inappropriate content, and provides feedback including the results and sentiment analysis.

[0275] Example 2

[0276] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0277] Conventional systems have difficulty detecting inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, etc., and lack a mechanism for analyzing users' emotions and providing feedback. This situation makes it difficult for users to maintain consistency in their posts and avoid inappropriate comments. Therefore, an objective of this invention is to provide a system that automatically and efficiently detects inconsistencies and inappropriate content in posts and analyzes emotions.

[0278] 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.

[0279] In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for analyzing the user's emotions using an emotion analysis engine, and means for notifying the user of the analysis results. This makes it possible to analyze the collected data and check the consistency and appropriateness of the posts, and further to analyze the user's emotions in detail and provide feedback.

[0280] "User" means any person or entity that uses the Software or System.

[0281] "Means of collection" refers to the ability to automatically or manually obtain data from sources such as social media posts, lecture notes, and book data.

[0282] "Means of analysis" refers to the function of using natural language processing technology to understand the context and meaning of the acquired data and extract the necessary information.

[0283] "Contradiction checkers" refer to algorithms or methods used to compare current statements with past statements to check for consistency and identify inconsistencies.

[0284] "Measures to check for inappropriate content" refers to a function that detects whether the collected data contains inappropriate words or phrases and prompts warnings or corrections as necessary.

[0285] "Sentiment analysis engine" refers to software or a system that uses natural language processing techniques and other sentiment analysis tools to extract user sentiment from text data and calculate a specific sentiment score.

[0286] "Notification means" refers to a method or system for notifying the user of the analysis results, and in particular refers to a function that sends information to the user's terminal so that the results can be viewed.

[0287] "Natural language processing technology" is a technology that enables computers to understand, interpret, and generate human language, and includes processes such as tokenization, stop word removal, and sentiment analysis.

[0288] "SNS post" refers to text or media content that a user publicly posts via a social networking service (SNS).

[0289] "Lecture recordings" refer to the content of lectures and seminars that have been audio-visually recorded and saved as digital data.

[0290] "Book data" refers to data that stores the contents of books and documents in digital format.

[0291] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and then analyzes the user's emotions using an emotion engine and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, emotion analysis, and notification.

[0292] Program processing

[0293] Data collection

[0294] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (e.g., Twitter API, Facebook Graph API). The audio data from lectures is converted into text data using the Google Cloud Speech-to-Text service, and all data is stored in a MySQL database.

[0295] Text analytics

[0296] The server analyzes the collected data using natural language processing (NLP) techniques using the Python libraries NLTK (Natural Language Toolkit) and Spacy, including tokenization, stop word removal, keyword extraction using TF-IDF (inverse document frequency), sentiment analysis, and sentence segmentation.

[0297] Inconsistency check

[0298] Based on the analyzed data, the server compares current and past statements using Python's difflib library to calculate similarities and identify inconsistencies. For example, if a user posts that "technological advances improve society" and then previously states that "technological advances have a negative impact on society," this will be detected as a contradiction.

[0299] Inappropriate remarks check

[0300] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for inappropriate words and phrases and analyzes their context. It uses dictionary data from NLTK and Spacy to determine whether a word like "idiot" is truly inappropriate in the context.

[0301] Emotion analysis

[0302] The server analyzes the user's emotions from the text data using an emotion engine. This emotion engine also uses NLP technology, such as VADER and TextBlob, to identify emotions from the speech content. For example, it calculates an emotion score such as "very happy" or "very angry."

[0303] Notification of results

[0304] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are detected, a detailed report including the results of the sentiment analysis is generated and sent to the user's device. The user receives the notification via their device and can view the report to identify problems with their own comments and fluctuations in sentiment.

[0305] Specific examples

[0306] User D posted on social media that "technological advances make society better," but in a past lecture he said that "technological advances have a negative impact on society." In addition, his latest social media post expresses the emotion of being "very happy."

[0307] The server collects User D's latest social media posts via the Twitter API and stores them in a MySQL database. It also collects audio data from lectures, converts it to text using Google Cloud Speech-to-Text, and stores it. The server then analyzes the data using the NLTK and Spacy NLP engines, and performs comparison and sentiment analysis. The server compiles the analysis results into a report and sends it to User D's device. User D can view this report to check for inconsistencies between past and present statements and the sentiment behind the latest posts.

[0308] Specific prompt examples

[0309] "Compare your past social media posts with your most recent ones to detect inconsistencies."

[0310] "Please analyze the sentiment of my social media posts and let me know the results."

[0311] "Please check whether what was said at the lecture matches what was posted on social media."

[0312] In this way, by inputting prompt sentences into the generative AI model, the system can operate effectively and provide the information the user is looking for.

[0313] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0314] Step 1: Data collection

[0315] The server collects content posted by users on various social networking sites, audio data from lectures, and digital data from books. Specifically, the server operates as follows:

[0316] Input: User ID or platform information to be collected

[0317] The server uses the Twitter API to retrieve post data related to the specified user ID in JSON format.

[0318] The server uses Google Cloud Speech-to-Text to convert the lecture's audio data (e.g., MP3 files) into text data.

[0319] The server parses the digital book data (e.g., PDF or EPUB file) and extracts the text content.

[0320] Output: Database records containing various collected text data

[0321] Step 2: Data Preprocessing

[0322] The server preprocesses the collected text data and prepares it in a format suitable for analysis.

[0323] Input: Raw text data

[0324] The server uses the Python library NLTK to tokenize (divide) the text data into words.

[0325] Remove stop words (e.g., frequently occurring words such as "wa" and "ga").

[0326] Calculate TF-IDF (inverse document frequency) and extract important keywords.

[0327] Output: Preprocessed text data

[0328] Step 3: Check for inconsistencies

[0329] The server analyzes the preprocessed text data and compares the current utterances with past utterances.

[0330] Input: Preprocessed text data

[0331] The server uses the Python difflib library to calculate the similarity scores of statements.

[0332] Identify when current statements contradict past statements above a certain threshold.

[0333] Output: Analysis results if inconsistencies are identified

[0334] Step 4: Check for inappropriate comments

[0335] The server checks the text data for inappropriate words or phrases.

[0336] Input: Preprocessed text data

[0337] The server uses dictionary data from NLTK and Spacy to scan for inappropriate words and phrases.

[0338] The context of any inappropriate words or phrases found is analyzed to determine whether they are truly inappropriate.

[0339] Output: Analysis results when inappropriate comments are identified

[0340] Step 5: Sentiment Analysis

[0341] The server utilizes an emotion engine that analyzes the user's emotions from the text data.

[0342] Input: Preprocessed text data

[0343] The server uses VADER or TextBlob to calculate the sentiment score for each piece of text.

[0344] Identify and score specific emotions (e.g., "happiness," "anger," "sadness," etc.).

[0345] Output: Sentiment analysis results

[0346] Step 6: Notification of results

[0347] The server compiles all the analysis results, generates a detailed report, and notifies the user.

[0348] Input: Results of inconsistency check, inappropriate comment check, and sentiment analysis

[0349] The server generates a report in JSON format and sends it to the user's device.

[0350] The user's terminal receives this report and displays and notifies the analysis results.

[0351] Output: User notification and detailed report

[0352] In this way, specific data processing and calculations are performed at each step, and the results of each process are passed on to the next step, allowing for a comprehensive analysis of the inconsistencies, inappropriateness, and emotional state of the user's comments.

[0353] (Application example 2)

[0354] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0355] Conventional systems have had difficulty detecting inappropriate expressions in user comments or inconsistencies with past comments. Furthermore, they lacked the functionality to understand users' emotional state through emotion analysis and assess security risks. This increased the risk of user comments causing misunderstandings and problems, hindering security.

[0356] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for analyzing the user's emotions using an emotion engine, and means for evaluating and notifying security risks based on the analysis results. This makes it possible to detect security risks early and respond effectively while maintaining the consistency and appropriateness of the content posted by users.

[0357] "User" refers to any individual or organizational representative who uses the system to make a statement.

[0358] "Content to be transmitted" refers to text data, audio data, and video data that users publish through social media, communication platforms, lectures, digital books, etc.

[0359] "Means of collection" refers to programs and APIs that automatically obtain user comments from various sources on the Internet.

[0360] "Means for analysis" refers to software or systems for analyzing the meaning of collected data using natural language processing techniques or other analytical algorithms.

[0361] "Means for checking for inconsistencies" refers to a program that compares current statements with past statements and detects inconsistencies when the content does not match.

[0362] "Measures to check for inappropriate content" refers to systems that use specialized dictionaries and contextual analysis to automatically detect inappropriate words and phrases.

[0363] "Means of notification" refers to the protocols and systems used to notify users of the results of the analysis, any detected inconsistencies, inappropriate content, and sentiment analysis.

[0364] "Emotion engine" refers to natural language processing technology used to analyze a user's emotional state from text data.

[0365] A means of assessing "security risk" refers to a system that determines whether a user's statements or actions pose a security risk based on the analysis results and notifies the user.

[0366] "Security risk" refers to any factor that could potentially threaten the security of a company or organization through user statements or actions.

[0367] The system of this invention collects content posted by users, analyzes it to detect inconsistencies and inappropriate content, evaluates security risks using an emotion engine, and provides feedback to users. A specific method for realizing this system is described below.

[0368] First, an API (Application Programming Interface) is used to collect content posted by users on social media, communication platforms, lectures, digital books, etc. A cloud-based server automatically retrieves content from these data sources and stores it in a database. Audio data from lectures is converted into text data using speech recognition technology.

[0369] The collected data is analyzed using natural language processing (NLP) techniques, including tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation, specifically using spaCy (spacy.io) and VaderSentiment (vaderSentiment).

[0370] The server then compares the current and past comments to check for inconsistencies. This is done using similarity calculations and conflict detection algorithms such as SequenceMatcher (difflib). The collected text data is also checked for inappropriate words and phrases. To detect inappropriate content, the server uses specialized dictionaries and context analysis.

[0371] Furthermore, an emotion engine is used to perform sentiment analysis. VaderSentiment is used to analyze the user's emotional state, such as "positive," "negative," or "neutral," from text data. This sentiment analysis is important for clarifying the emotions reflected in the user's comments.

[0372] The analysis results are sent to the user's device via a dedicated notification protocol or API, allowing the user to understand the problems and security risks of their own comments.

[0373] For example, if a user posts something like, "Today I received confidential information from a client," the system will immediately analyze it and detect security risks such as the leakage of confidential information. Inconsistencies, inappropriate content, and emotional states are analyzed, and an alert is sent to the user.

[0374] Examples of prompts include:

[0375] Analyze the latest SNS post by user ID "user123" titled "Today I received confidential information from a client." and check the following items:

[0376] 1. Are there any contradictions with previous posts?

[0377] 2. Does it contain inappropriate language?

[0378] 3. Analyze the emotional state and report the results.”

[0379] In this way, users can see in real time whether their speech is coherent, inappropriate, and reflects their emotional state, minimizing security risks and providing appropriate feedback.

[0380] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0381] Step 1:

[0382] The server collects content posted by users. In this case, data is obtained through the API of a social networking site or communication platform. The input is data from the API, and the output is collected text data or audio data. For example, the server collects the latest social networking post from "user123."

[0383] Step 2:

[0384] The server converts the collected voice data into text data using voice recognition technology. The input is voice data, and the output is the converted text data. Voice recognition software is used for this process. For example, "voice data of a lecture" is converted into "text of the lecture content."

[0385] Step 3:

[0386] The server analyzes the text data. Here, natural language processing (NLP) techniques are used to analyze the data and understand its context and meaning. The input is the text data, and the output is the analysis results. Specific operations include tokenization, stop word removal, and keyword extraction. For example, the server breaks down the "text of a lecture" into "individual words and phrases" and extracts important keywords.

[0387] Step 4:

[0388] The server uses the analysis results to check for inconsistencies between current and past comments. The input is the analyzed current text data and past text data, and the output is whether there are any inconsistencies. Similarity calculations and conflict detection algorithms are used. For example, by comparing the "current post" with the "content of a past lecture," inconsistencies are identified.

[0389] Step 5:

[0390] The server checks whether a post contains inappropriate content. The input is analyzed text data, and the output is whether or not there are any inappropriate comments. It uses a dedicated dictionary and context analysis. For example, it compares a "list of inappropriate words" with the "current post" to see if it contains any inappropriate words.

[0391] Step 6:

[0392] The server analyzes the user's emotions using an emotion engine. The input is the analyzed text data, and the output is the result of the emotion analysis. VaderSentiment is used to determine the user's emotional state from the text. For example, it assigns an emotion label of "positive," "negative," or "neutral" to the "current post."

[0393] Step 7:

[0394] The server evaluates security risks based on the analysis results and notifies the user. The input is the analysis results and the results of sentiment analysis, and the output is the notification to the user. The analysis results are compiled and sent as a report to the user's device via API. For example, the server may notify the user of information such as "the current post is contradictory," "it contains inappropriate words," or "it has strong negative sentiment."

[0395] Step 8:

[0396] The user receives and confirms the notification on the device. The input is the notification from the server, and the output is the user's confirmation result. The user can view the notification and take appropriate action. For example, they can take actions such as "correcting contradictory statements" or "correcting inappropriate expressions."

[0397] In this way, the entire system can monitor and analyze user comments in real time, minimizing security risks.

[0398] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0399] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0400] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0401] [Second embodiment]

[0402] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0403] 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.

[0404] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0405] 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.

[0406] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0407] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0408] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0409] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0410] 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.

[0411] 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.

[0412] 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.

[0413] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0414] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, and notification.

[0415] Program processing

[0416] Data collection

[0417] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (application programming interface). The audio data from lectures is converted into text data using voice recognition technology, and all data is stored in a single database.

[0418] Text analytics

[0419] The server then analyzes the collected data, using natural language processing (NLP) techniques to analyze the text data and understand its context and meaning. This analysis includes tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[0420] Inconsistency check

[0421] Based on the analyzed data, the server compares current and past statements, using similarity calculations and conflict detection algorithms to identify inconsistencies. For example, a statement that "environmental protection is important" could be deemed a contradiction with a past statement that "environmental issues are not a concern."

[0422] Inappropriate remarks check

[0423] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for words and phrases that are deemed inappropriate and analyzes their context. For example, if the word "idiot" is used, it checks the context before and after it to determine whether it is truly inappropriate.

[0424] Notification of results

[0425] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are identified, a detailed report is sent to the user's device. The user receives a notification through their device and can view the report to identify the problems with their own comments.

[0426] Specific examples

[0427] scenario

[0428] User C posted on social media that "technological advances will make society better." However, in the past, the same User C stated in a lecture that "technological advances will have a negative impact on society."

[0429] Data collection

[0430] The server collects User C's latest SNS posts via API and stores them in a database. It also collects audio data from the lecture, converts it into text using speech recognition technology, and stores it.

[0431] Text analytics

[0432] The server analyzes the collected social media posts and speeches using an NLP engine to understand their meaning and context, and performs processes such as tokenization, keyword extraction, and sentiment analysis.

[0433] Inconsistency check

[0434] The server compares the current social media post, "Technological advances make society better," with the past speech, "Technological advances have a negative impact on society," and determines that there is a contradiction.

[0435] Inappropriate remarks check

[0436] In this example, there is no particularly inappropriate language, so we will skip this step.

[0437] Notification of results

[0438] The server generates a report summarizing the analysis results that identified the contradiction and notifies the terminal of User C. User C views this report and confirms that there is a contradiction between the past and present statements.

[0439] In this way, the system for implementing the present invention allows users to maintain consistency in the content they post and avoid making inappropriate comments, thereby preventing misunderstandings and trouble.

[0440] The processing flow will be explained below.

[0441] Step 1:

[0442] The server uses an API that connects to the user's social media account to periodically check whether a new post has been made. If a new post is detected, the text data is stored in a database.

[0443] Step 2:

[0444] The server collects the audio data of the lecture in real time. The collected audio data is converted into text data using speech recognition technology. The converted text data is stored in a database.

[0445] Step 3:

[0446] Users upload digital files of book data to the server, which then converts the uploaded book data into text format and stores it in a database.

[0447] Step 4:

[0448] The server performs preprocessing on the collected text data, including tokenization, stop word removal, and normalization, before passing the preprocessed text data to a natural language processing engine.

[0449] Step 5:

[0450] The server uses natural language processing techniques to semantically analyze the text data, including keyword extraction, sentiment analysis, sentence segmentation, and syntactic analysis, to understand the specific context and meaning of each utterance.

[0451] Step 6:

[0452] The server compares current and past utterance data. Using similarity calculations and contradiction detection algorithms, it identifies contradictions between utterances. For example, if a contradiction is detected between the utterances "Technological advances improve society" and "Technological advances have a negative impact on society," it will identify it.

[0453] Step 7:

[0454] The server scans the text data using a specialized dictionary containing inappropriate words and phrases, and analyzes the context of identified inappropriate words and phrases to determine whether they are in fact inappropriate.

[0455] Step 8:

[0456] The server generates a report summarizing the analysis and evaluation results, including specific areas of inconsistency and inappropriate comments, and sends the report to the user's device.

[0457] Step 9:

[0458] Users receive a notification from the server via their device, which includes a link to a detailed report that allows them to identify the issues with their comments.

[0459] The above is a concrete flow of a series of processing steps for collecting and analyzing user comments, checking for inconsistencies and inappropriate content, and notifying the results.

[0460] Example 1

[0461] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0462] Inconsistent content posted by users or inappropriate comments can lead to misunderstandings and problems. This problem poses a challenge, as it can undermine the user's credibility and lower social credibility. Another problem is that manually checking the consistency and appropriateness of comments by users is time-consuming and inefficient.

[0463] 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.

[0464] In this invention, the server includes means for collecting content posted by users, means for storing the collected content in a database, means for analyzing the stored content using natural language processing technology, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, and means for notifying the user of the analysis results. This enables users to maintain consistency in the content they post and avoid making inappropriate posts.

[0465] "User" refers to any individual or organization that transmits information or opinions.

[0466] "Content" refers to a collection of user-generated text, audio, or digital data.

[0467] "Means of collection" refers to methods for collecting user-generated content using APIs and data acquisition methods of various platforms.

[0468] "Database" refers to a data storage system for storing and managing collected data.

[0469] "Means for storage" refers to the method for storing and managing collected data in a database.

[0470] "Natural language processing technology" refers to technology that analyzes text data and allows it to understand its meaning and context.

[0471] "Means of analysis" refers to a method of analyzing collected text data using natural language processing technology to extract specific information and context.

[0472] "Means for checking for inconsistencies" refers to a method of comparing the latest statements with past statements based on analyzed data to determine whether there are any inconsistencies.

[0473] "Measures to check for inappropriate content" refers to methods that use specialized dictionaries and algorithms to determine whether a comment contains inappropriate words or phrases.

[0474] "Means for notifying" refers to a method for transmitting the analysis results to the user's terminal and notifying the user.

[0475] "Tokenization" refers to the process of dividing text data into words or phrases.

[0476] "Stop word removal" refers to the process of removing common words that are not necessary for analysis from the text to be analyzed.

[0477] "Keyword extraction" refers to the process of extracting important words and phrases from text data.

[0478] "Sentiment analysis" refers to the process of determining emotions and emotional direction from the context of text data.

[0479] "Sentence segmentation" refers to the process of dividing text data into sentences or segments.

[0480] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and provides feedback. The system mainly comprises a server, a terminal, and a user. A specific embodiment of this invention will be described below.

[0481] System Overview

[0482] The system involves a server collecting user posts, analyzing them using natural language processing, detecting inconsistencies and inappropriate content, and notifying the user of the results via their device. The server uses the SNS API, voice recognition technology, a database management system, and a natural language processing library.

[0483] Data collection and storage

[0484] The server collects user posts using the SNS API and uses the Google Cloud Speech-to-Text API to convert the voice data. The collected data is then stored in a database such as MySQL.

[0485] Text analytics

[0486] The server analyzes the collected text data using natural language processing techniques, such as tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation, using libraries such as SpaCy and NLTK.

[0487] Inconsistency check and inappropriate remark check

[0488] The server compares the most recent and past comments based on the analysis results to identify inconsistencies. Similarity calculations are performed using cosine similarity, etc. To check for inappropriate comments, a dedicated dictionary is used to scan for specific words and phrases, and the context is also analyzed.

[0489] Notification of results

[0490] The server compiles the results of the analysis of detected inconsistencies and inappropriate comments and generates a report. This report is sent to the user's device, allowing the user to check the problematic aspects of the comments. Notifications are sent to the user's device using an API.

[0491] Specific examples

[0492] A specific example of the operation of the system is shown below.

[0493] scenario

[0494] A user posted on social media that "technological advances will improve society." However, in the past, the same user stated in a lecture that "technological advances will have a negative impact on society."

[0495] Data collection

[0496] The server collects the latest posts via the SNS API and stores them in a database. It also collects audio data from lectures, converts it into text using the Google Cloud Speech-to-Text API, and stores it.

[0497] Text analytics

[0498] The server analyzes the collected social media posts and lecture comments using natural language processing technology (e.g., SpaCy) to understand the meaning and context of each.

[0499] Inconsistency check

[0500] The server compared the social media post, "Technological advances make society better," with the speech, "Technological advances have a negative impact on society," and determined that there was a contradiction.

[0501] Inappropriate remarks check

[0502] In this example, there is no particularly inappropriate language, so we will skip this step.

[0503] Notification of results

[0504] The server generates a report summarizing the analysis results that identified the inconsistencies and sends it to the user's device. The user can view this report and check the inconsistencies in the comments.

[0505] Prompt Sentence Examples

[0506] "Write a program that compares a user's social media posts with their past statements to identify inconsistencies."

[0507] This invention allows users to ensure consistency in the content they post and avoid making inappropriate comments, thereby preventing misunderstandings and trouble.

[0508] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0509] Step 1: Data collection

[0510] The server collects content posted by users on social media, audio data from lectures, and digital data from books. Specifically, it acquires data using various APIs (such as APIs for social networking services and voice recognition APIs).

[0511] Input: Social media posts, lecture audio files, digital book data

[0512] Data processing: Convert the audio data from the lecture into text using the Google Cloud Speech-to-Text API.

[0513] Output: Text data

[0514] Specific operation: The server uses the SNS API to collect the latest user posts, sends the audio file of the lecture called audio_file.wav to the Google Cloud Speech-to-Text API, and obtains the text data "Technological advances have a negative impact on society."

[0515] Step 2: Save your data

[0516] The server stores the collected data in a database, where the data is organized for each user and stored in a collated format.

[0517] Input: Collected text data

[0518] Data processing: structuring data

[0519] Output: A structured database

[0520] What happens: The server connects to the database and inserts new data, for example by executing an SQL query like INSERT INTO user_data (user_id, content, type) VALUES (123, 'Technological advances make society better', 'SNS').

[0521] Step 3: Text analysis

[0522] The server analyzes the stored text data using natural language processing (NLP) techniques, including tokenization, stop-word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[0523] Input: Saved text data

[0524] Data processing: tokenization, stop word removal, keyword extraction, sentiment analysis, sentence segmentation

[0525] Output: Analysis results

[0526] Specific operation: The server uses the SpaCy library to analyze text data. For example, it loads the model with spacy.load("en_core_web_sm") and analyzes the text with nlp("Technological advances make society better"). It extracts keywords from the analysis results and creates a list such as "technology", "advancement", "society", and "make it better".

[0527] Step 4: Check for inconsistencies

[0528] The server compares the latest and past comments based on the analysis results, and identifies inconsistencies using similarity calculations (cosine similarity) and conflict detection algorithms.

[0529] Input: Text analysis results

[0530] Data processing: Similarity calculation, conflict detection

[0531] Output: Contradiction judgment result

[0532] Specific operation: The server calculates the vectors of past and current statements and evaluates their similarity. For example, it calculates the similarity using cosine_similarity(vector_a, vector_b), and if the result is below a certain threshold, it determines that there is a contradiction.

[0533] Step 5: Check for inappropriate comments

[0534] The server uses a dedicated dictionary to check whether the comment contains inappropriate words or phrases, and also analyzes the context to determine whether the comment is truly inappropriate.

[0535] Input: Text analysis results

[0536] Data processing: dictionary matching, context analysis

[0537] Output: Inappropriate judgment result

[0538] What it does: The server splits the text into words and compares each word to a list of inappropriate words, e.g., if word in inappropriate_words: flag_as_inappropriate(word) to detect inappropriate words.

[0539] Step 6: Notification of results

[0540] The server compiles the results of the analysis of inconsistencies and inappropriate comments, generates a report, and sends the report to the user's device, where the user can check it.

[0541] Input: Conflict judgment result, inappropriate judgment result

[0542] Data Processing: Report Generation

[0543] Output: Notification messages, reports

[0544] Specific operation: The server creates a report based on the analysis results and notifies the user via the API on the user's device. For example, it calls the PUT / user_notifications API and sends the notification data {"user_id": 123, "message": "Past and current statements are inconsistent"}.

[0545] (Application example 1)

[0546] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0547] In physical stores, it is difficult for staff to provide consistent information to customers. Correcting statements made in the past can damage customer trust. Furthermore, there is no way to check past statements in real time, so there is a risk of repeating the same mistake. This raises concerns about lower customer satisfaction and worsening operational efficiency.

[0548] 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.

[0549] In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for notifying the user of the analysis results, and means for collecting and analyzing data in real time based on the user's posts. This allows customer service staff to maintain consistency in customer service, prevent the provision of incorrect information, and maintain customer trust.

[0550] "Means for collecting user-generated content" refers to processes and technologies that automatically collect data entered or transmitted by users in the form of information networks, oral presentations, documents, etc.

[0551] "Means for analyzing collected content" are the processes and techniques used to analyze collected data and identify its meaning and context.

[0552] "Means for checking for inconsistencies with past statements based on the analysis results" refers to the process and techniques for comparing the analysis results with previous statements to see if there are any inconsistencies.

[0553] "Measures to check speech for inappropriate content" are processes and techniques that scan speech for inappropriate words, phrases, or context.

[0554] "Means for notifying the user of the analysis results" refers to the process and technology for notifying the user of the results obtained by the analysis so that the user can check them.

[0555] "Means for collecting and analyzing data in real time based on user comments" refers to the process and technology for quickly collecting comments made by users on the spot and analyzing them immediately.

[0556] This invention is a system that analyzes the content of user comments in a physical store in real time and provides feedback on the results. This system includes the following main processes.

[0557] Data collection

[0558] The server acquires voice data to collect what the user is saying. The voice data is collected through a microphone in the smart glasses and converted into text data using voice recognition technology. This process uses the Google Cloud Speech-to-Text API. The collected data is then stored in a database.

[0559] Text analytics

[0560] The server analyzes the collected text data using natural language processing (NLP) techniques, using Python libraries such as NLTK and spaCy, and performs processes such as tokenization, stop word removal, keyword extraction, and sentiment analysis.

[0561] Inconsistency check

[0562] Based on the analyzed data, the server compares the current utterance with the past utterances, and uses similarity calculations and conflict detection algorithms to identify inconsistencies. In this process, the Python scikit-learn library is used, for example, to calculate cosine similarity.

[0563] Inappropriate remarks check

[0564] The server checks for inappropriate words and phrases. It uses a Python dictionary database to scan for potentially inappropriate words and phrases, and analyzes the context to determine whether they are truly inappropriate.

[0565] Notification of results

[0566] If any inconsistencies or inappropriate comments are identified, the server will summarize the analysis results and notify the user. The notification will be displayed in real time on the smart glasses display. The smart glasses application will be developed for Android or iOS.

[0567] Specific examples

[0568] Consider a case where a user tells a customer in a physical store, "The point card is valid for one year," but in the past has said, "The point card is only valid for six months."

[0569] Prompt Sentence Examples

[0570] "Check your customer interaction records and see if there are any contradictions between current and past statements. Past statement: 'The loyalty card is only valid for six months.' Current statement: 'The loyalty card is valid for one year.'"

[0571] This system allows users to maintain consistency in customer service and prevent the provision of inappropriate information, which is expected to improve customer satisfaction and operational efficiency.

[0572] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0573] Step 1:

[0574] Data collection

[0575] The server collects what the user says through a microphone installed in the smart glasses. The collected voice data is converted into text data using the Google Cloud Speech-to-Text API. The converted text data is then stored in a database.

[0576] Input: Audio data

[0577] Processing: Convert speech to text (using Google Cloud Speech-to-Text API)

[0578] Output: Text data (stored in database)

[0579] Step 2:

[0580] Text analytics

[0581] The server analyzes the collected text data using Python natural language processing libraries (such as NLTK or spaCy), specifically performing tokenization, stop word removal, keyword extraction, and sentiment analysis.

[0582] Input: Text data

[0583] Processing: Tokenization, stopword removal, keyword extraction, sentiment analysis (using NLTK and spaCy libraries)

[0584] Output: Parsed text data

[0585] Step 3:

[0586] Inconsistency check

[0587] The server compares current and past statements based on the parsed text data, using the Python scikit-learn library to identify inconsistencies using algorithms such as cosine similarity calculations.

[0588] Input: Analyzed text data, past text data

[0589] Processing: Similarity calculation (using the scikit-learn library)

[0590] Output: Conflicts

[0591] Step 4:

[0592] Inappropriate remarks check

[0593] The server then scans the parsed text data for inappropriate words and phrases, using a Python dictionary database and analyzing the context to make its decisions.

[0594] Input: Parsed text data

[0595] Processing: Scanning for inappropriate words and phrases (using dictionary database)

[0596] Output: Whether or not there is inappropriate remarks

[0597] Step 5:

[0598] Notification of results

[0599] Based on the results of the inconsistencies and inappropriate comments, the server compiles the analysis results and notifies the user's smart glasses in real time, which are displayed on the smart glasses' display.

[0600] Input: Whether there are contradictions or not, whether there are inappropriate comments or not

[0601] Processing: Creating and sending analysis results

[0602] Output: Feedback notification to the smart glasses display

[0603] This series of processes allows users to instantly correct any inconsistencies with past comments or inappropriate comments when dealing with customers in a physical store.

[0604] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0605] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and then analyzes the user's emotions using an emotion engine and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, emotion analysis, and notification.

[0606] Program processing

[0607] Data collection

[0608] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (application programming interface). The audio data from lectures is converted into text data using voice recognition technology, and all data is stored in a single database.

[0609] Text analytics

[0610] The server then analyzes the collected data, using natural language processing (NLP) techniques to analyze the text data and understand its context and meaning. This analysis includes tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[0611] Inconsistency check

[0612] Based on the analyzed data, the server compares current and past statements. It uses similarity calculations and conflict detection algorithms to identify contradictions. For example, a statement that "technological advances make society better" could be deemed a contradiction with a past statement that "technological advances have a negative impact on society."

[0613] Inappropriate remarks check

[0614] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for words and phrases that are deemed inappropriate and analyzes their context. For example, if the word "idiot" is used, it checks the context before and after it to determine whether it is truly inappropriate.

[0615] Emotion analysis

[0616] A distinctive feature of this invention is that the server uses an emotion engine to analyze the user's emotions from text data. This emotion engine uses NLP technology to identify emotions from the content of the user's speech. For example, it identifies emotions such as "very happy" or "very angry." This allows for a detailed understanding of the emotions expressed by the user's speech.

[0617] Notification of results

[0618] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are detected, a detailed report including the results of the sentiment analysis is sent to the user's device. The user receives the notification via their device and can view the report to identify problems with their own comments and fluctuations in sentiment.

[0619] Specific examples

[0620] scenario

[0621] User D posted on social media that "technological advances will make society better." However, in the past, the same User D stated at a lecture that "technological advances will have a negative impact on society." The social media post also expresses the emotion of "being very happy."

[0622] Data collection

[0623] The server collects the latest SNS posts from User D via API and stores them in a database. It also collects audio data from the lecture, converts it into text using speech recognition technology, and stores it.

[0624] Text analytics

[0625] The server analyzes the collected social media posts and speeches using an NLP engine to understand their meaning and context, and performs processes such as tokenization, keyword extraction, and sentiment analysis.

[0626] Inconsistency check

[0627] The server compares the current social media post, "Technological advances make society better," with the past speech, "Technological advances have a negative impact on society," and determines that there is a contradiction.

[0628] Inappropriate remarks check

[0629] In this example, there is no particularly inappropriate language, so we will skip this step.

[0630] Emotion analysis

[0631] The server analyzes the emotion "very happy" from the latest social media posts and adds it to the analysis results.

[0632] Notification of results

[0633] The server generates a report summarizing the analysis results that identified the contradictions and the results of the emotion analysis, and notifies the device of User D. User D views this report and confirms that there is a contradiction between his past and present statements, and that his latest post expresses the emotion of "very happy."

[0634] In this way, the system for implementing this invention not only allows users to maintain consistency in their speech and emotions and avoid inappropriate speech, but also allows users to accurately grasp emotional fluctuations, thereby preventing misunderstandings and troubles and deepening understanding of emotional states.

[0635] The processing flow will be explained below.

[0636] Step 1:

[0637] The server uses an API that connects to the user's social media account to periodically check whether a new post has been made. If a new post is detected, the text data is stored in a database.

[0638] Step 2:

[0639] The server collects the audio data of the lecture in real time. The collected audio data is converted into text data using speech recognition technology. The converted text data is stored in a database.

[0640] Step 3:

[0641] Users upload digital files of book data to the server, which then converts the uploaded book data into text format and stores it in a database.

[0642] Step 4:

[0643] The server performs preprocessing on the collected text data, including tokenization, stop word removal, and normalization, before passing the preprocessed text data to a natural language processing engine.

[0644] Step 5:

[0645] The server uses natural language processing techniques to semantically analyze the text data, including keyword extraction, sentiment analysis, sentence segmentation, and syntactic analysis, to understand the specific context and meaning of each utterance.

[0646] Step 6:

[0647] The server compares current and past utterance data. Using similarity calculations and contradiction detection algorithms, it identifies contradictions between utterances. For example, if a contradiction is detected between the utterances "Technological advances improve society" and "Technological advances have a negative impact on society," it will identify it.

[0648] Step 7:

[0649] The server scans the text data using a specialized dictionary containing inappropriate words and phrases, and analyzes the context of identified inappropriate words and phrases to determine whether they are in fact inappropriate.

[0650] Step 8:

[0651] The server uses an emotion engine to analyze the emotions in the user's comments. The emotion engine uses NLP technology to identify, for example, whether the comment contains the emotion "very happy."

[0652] Step 9:

[0653] The server generates a report summarizing the analysis and evaluation results, including identified inconsistencies, inappropriate comments, and sentiment analysis results, and sends the report to the user's device.

[0654] Step 10:

[0655] Users receive notifications from the server via their devices, which include a link to a detailed report that allows them to identify problems with their own comments and fluctuations in sentiment.

[0656] The above is a concrete flow of a series of processing steps that collects and analyzes user comments, checks for inconsistencies and inappropriate content, and provides feedback including the results and sentiment analysis.

[0657] Example 2

[0658] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0659] Conventional systems have difficulty detecting inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, etc., and lack a mechanism for analyzing users' emotions and providing feedback. This situation makes it difficult for users to maintain consistency in their posts and avoid inappropriate comments. Therefore, an objective of this invention is to provide a system that automatically and efficiently detects inconsistencies and inappropriate content in posts and analyzes emotions.

[0660] 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.

[0661] In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for analyzing the user's emotions using an emotion analysis engine, and means for notifying the user of the analysis results. This makes it possible to analyze the collected data and check the consistency and appropriateness of the posts, and further to analyze the user's emotions in detail and provide feedback.

[0662] "User" means any person or entity that uses the Software or System.

[0663] "Means of collection" refers to the ability to automatically or manually obtain data from sources such as social media posts, lecture notes, and book data.

[0664] "Means of analysis" refers to the function of using natural language processing technology to understand the context and meaning of the acquired data and extract the necessary information.

[0665] "Contradiction checkers" refer to algorithms or methods used to compare current statements with past statements to check for consistency and identify inconsistencies.

[0666] "Measures to check for inappropriate content" refers to a function that detects whether the collected data contains inappropriate words or phrases and prompts warnings or corrections as necessary.

[0667] "Sentiment analysis engine" refers to software or a system that uses natural language processing techniques and other sentiment analysis tools to extract user sentiment from text data and calculate a specific sentiment score.

[0668] "Notification means" refers to a method or system for notifying the user of the analysis results, and in particular refers to a function that sends information to the user's terminal so that the results can be viewed.

[0669] "Natural language processing technology" is a technology that enables computers to understand, interpret, and generate human language, and includes processes such as tokenization, stop word removal, and sentiment analysis.

[0670] "SNS post" refers to text or media content that a user publicly posts via a social networking service (SNS).

[0671] "Lecture recordings" refer to the content of lectures and seminars that have been audio-visually recorded and saved as digital data.

[0672] "Book data" refers to data that stores the contents of books and documents in digital format.

[0673] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and then analyzes the user's emotions using an emotion engine and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, emotion analysis, and notification.

[0674] Program processing

[0675] Data collection

[0676] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (e.g., Twitter API, Facebook Graph API). The audio data from lectures is converted into text data using the Google Cloud Speech-to-Text service, and all data is stored in a MySQL database.

[0677] Text analytics

[0678] The server analyzes the collected data using natural language processing (NLP) techniques using the Python libraries NLTK (Natural Language Toolkit) and Spacy, including tokenization, stop word removal, keyword extraction using TF-IDF (inverse document frequency), sentiment analysis, and sentence segmentation.

[0679] Inconsistency check

[0680] Based on the analyzed data, the server compares current and past statements using Python's difflib library to calculate similarities and identify inconsistencies. For example, if a user posts that "technological advances improve society" and then previously states that "technological advances have a negative impact on society," this will be detected as a contradiction.

[0681] Inappropriate remarks check

[0682] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for inappropriate words and phrases and analyzes their context. It uses dictionary data from NLTK and Spacy to determine whether a word like "idiot" is truly inappropriate in the context.

[0683] Emotion analysis

[0684] The server analyzes the user's emotions from the text data using an emotion engine. This emotion engine also uses NLP technology, such as VADER and TextBlob, to identify emotions from the speech content. For example, it calculates an emotion score such as "very happy" or "very angry."

[0685] Notification of results

[0686] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are detected, a detailed report including the results of the sentiment analysis is generated and sent to the user's device. The user receives the notification via their device and can view the report to identify problems with their own comments and fluctuations in sentiment.

[0687] Specific examples

[0688] User D posted on social media that "technological advances make society better," but in a past lecture he said that "technological advances have a negative impact on society." In addition, his latest social media post expresses the emotion of being "very happy."

[0689] The server collects User D's latest social media posts via the Twitter API and stores them in a MySQL database. It also collects audio data from lectures, converts it to text using Google Cloud Speech-to-Text, and stores it. The server then analyzes the data using the NLTK and Spacy NLP engines, and performs comparison and sentiment analysis. The server compiles the analysis results into a report and sends it to User D's device. User D can view this report to check for inconsistencies between past and present statements and the sentiment behind the latest posts.

[0690] Specific prompt examples

[0691] "Compare your past social media posts with your most recent ones to detect inconsistencies."

[0692] "Please analyze the sentiment of my social media posts and let me know the results."

[0693] "Please check whether what was said at the lecture matches what was posted on social media."

[0694] In this way, by inputting prompt sentences into the generative AI model, the system can operate effectively and provide the information the user is looking for.

[0695] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0696] Step 1: Data collection

[0697] The server collects content posted by users on various social networking sites, audio data from lectures, and digital data from books. Specifically, the server operates as follows:

[0698] Input: User ID or platform information to be collected

[0699] The server uses the Twitter API to retrieve post data related to the specified user ID in JSON format.

[0700] The server uses Google Cloud Speech-to-Text to convert the lecture's audio data (e.g., MP3 files) into text data.

[0701] The server parses the digital book data (e.g., PDF or EPUB file) and extracts the text content.

[0702] Output: Database records containing various collected text data

[0703] Step 2: Data Preprocessing

[0704] The server preprocesses the collected text data and prepares it in a format suitable for analysis.

[0705] Input: Raw text data

[0706] The server uses the Python library NLTK to tokenize (divide) the text data into words.

[0707] Remove stop words (e.g., frequently occurring words such as "wa" and "ga").

[0708] Calculate TF-IDF (inverse document frequency) and extract important keywords.

[0709] Output: Preprocessed text data

[0710] Step 3: Check for inconsistencies

[0711] The server analyzes the preprocessed text data and compares the current utterances with past utterances.

[0712] Input: Preprocessed text data

[0713] The server uses the Python difflib library to calculate the similarity scores of statements.

[0714] Identify when current statements contradict past statements above a certain threshold.

[0715] Output: Analysis results if inconsistencies are identified

[0716] Step 4: Check for inappropriate comments

[0717] The server checks the text data for inappropriate words or phrases.

[0718] Input: Preprocessed text data

[0719] The server uses dictionary data from NLTK and Spacy to scan for inappropriate words and phrases.

[0720] The context of any inappropriate words or phrases found is analyzed to determine whether they are truly inappropriate.

[0721] Output: Analysis results when inappropriate comments are identified

[0722] Step 5: Sentiment Analysis

[0723] The server utilizes an emotion engine that analyzes the user's emotions from the text data.

[0724] Input: Preprocessed text data

[0725] The server uses VADER or TextBlob to calculate the sentiment score for each piece of text.

[0726] Identify and score specific emotions (e.g., "happiness," "anger," "sadness," etc.).

[0727] Output: Sentiment analysis results

[0728] Step 6: Notification of results

[0729] The server compiles all the analysis results, generates a detailed report, and notifies the user.

[0730] Input: Results of inconsistency check, inappropriate comment check, and sentiment analysis

[0731] The server generates a report in JSON format and sends it to the user's device.

[0732] The user's terminal receives this report and displays and notifies the analysis results.

[0733] Output: User notification and detailed report

[0734] In this way, specific data processing and calculations are performed at each step, and the results of each process are passed on to the next step, allowing for a comprehensive analysis of the inconsistencies, inappropriateness, and emotional state of the user's comments.

[0735] (Application example 2)

[0736] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0737] Conventional systems have had difficulty detecting inappropriate expressions in user comments or inconsistencies with past comments. Furthermore, they lacked the functionality to understand users' emotional state through emotion analysis and assess security risks. This increased the risk of user comments causing misunderstandings and problems, hindering security.

[0738] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for analyzing the user's emotions using an emotion engine, and means for evaluating and notifying security risks based on the analysis results. This makes it possible to detect security risks early and respond effectively while maintaining the consistency and appropriateness of the content posted by users.

[0739] "User" refers to any individual or organizational representative who uses the system to make a statement.

[0740] "Content to be transmitted" refers to text data, audio data, and video data that users publish through social media, communication platforms, lectures, digital books, etc.

[0741] "Means of collection" refers to programs and APIs that automatically obtain user comments from various sources on the Internet.

[0742] "Means for analysis" refers to software or systems for analyzing the meaning of collected data using natural language processing techniques or other analytical algorithms.

[0743] "Means for checking for inconsistencies" refers to a program that compares current statements with past statements and detects inconsistencies when the content does not match.

[0744] "Measures to check for inappropriate content" refers to systems that use specialized dictionaries and contextual analysis to automatically detect inappropriate words and phrases.

[0745] "Means of notification" refers to the protocols and systems used to notify users of the results of the analysis, any detected inconsistencies, inappropriate content, and sentiment analysis.

[0746] "Emotion engine" refers to natural language processing technology used to analyze a user's emotional state from text data.

[0747] A means of assessing "security risk" refers to a system that determines whether a user's statements or actions pose a security risk based on the analysis results and notifies the user.

[0748] "Security risk" refers to any factor that could potentially threaten the security of a company or organization through user statements or actions.

[0749] The system of this invention collects content posted by users, analyzes it to detect inconsistencies and inappropriate content, evaluates security risks using an emotion engine, and provides feedback to users. A specific method for realizing this system is described below.

[0750] First, an API (Application Programming Interface) is used to collect content posted by users on social media, communication platforms, lectures, digital books, etc. A cloud-based server automatically retrieves content from these data sources and stores it in a database. Audio data from lectures is converted into text data using speech recognition technology.

[0751] The collected data is analyzed using natural language processing (NLP) techniques, including tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation, specifically using spaCy (spacy.io) and VaderSentiment (vaderSentiment).

[0752] The server then compares the current and past comments to check for inconsistencies. This is done using similarity calculations and conflict detection algorithms such as SequenceMatcher (difflib). The collected text data is also checked for inappropriate words and phrases. To detect inappropriate content, the server uses specialized dictionaries and context analysis.

[0753] Furthermore, an emotion engine is used to perform sentiment analysis. VaderSentiment is used to analyze the user's emotional state, such as "positive," "negative," or "neutral," from text data. This sentiment analysis is important for clarifying the emotions reflected in the user's comments.

[0754] The analysis results are sent to the user's device via a dedicated notification protocol or API, allowing the user to understand the problems and security risks of their own comments.

[0755] For example, if a user posts something like, "Today I received confidential information from a client," the system will immediately analyze it and detect security risks such as the leakage of confidential information. Inconsistencies, inappropriate content, and emotional states are analyzed, and an alert is sent to the user.

[0756] Examples of prompts include:

[0757] Analyze the latest SNS post by user ID "user123" titled "Today I received confidential information from a client." and check the following items:

[0758] 1. Are there any contradictions with previous posts?

[0759] 2. Does it contain inappropriate language?

[0760] 3. Analyze the emotional state and report the results.”

[0761] In this way, users can see in real time whether their speech is coherent, inappropriate, and reflects their emotional state, minimizing security risks and providing appropriate feedback.

[0762] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0763] Step 1:

[0764] The server collects content posted by users. In this case, data is obtained through the API of a social networking site or communication platform. The input is data from the API, and the output is collected text data or audio data. For example, the server collects the latest social networking post from "user123."

[0765] Step 2:

[0766] The server converts the collected voice data into text data using voice recognition technology. The input is voice data, and the output is the converted text data. Voice recognition software is used for this process. For example, "voice data of a lecture" is converted into "text of the lecture content."

[0767] Step 3:

[0768] The server analyzes the text data. Here, natural language processing (NLP) techniques are used to analyze the data and understand its context and meaning. The input is the text data, and the output is the analysis results. Specific operations include tokenization, stop word removal, and keyword extraction. For example, the server breaks down the "text of a lecture" into "individual words and phrases" and extracts important keywords.

[0769] Step 4:

[0770] The server uses the analysis results to check for inconsistencies between current and past comments. The input is the analyzed current text data and past text data, and the output is whether there are any inconsistencies. Similarity calculations and conflict detection algorithms are used. For example, by comparing the "current post" with the "content of a past lecture," inconsistencies are identified.

[0771] Step 5:

[0772] The server checks whether a post contains inappropriate content. The input is analyzed text data, and the output is whether or not there are any inappropriate comments. It uses a dedicated dictionary and context analysis. For example, it compares a "list of inappropriate words" with the "current post" to see if it contains any inappropriate words.

[0773] Step 6:

[0774] The server analyzes the user's emotions using an emotion engine. The input is the analyzed text data, and the output is the result of the emotion analysis. VaderSentiment is used to determine the user's emotional state from the text. For example, it assigns an emotion label of "positive," "negative," or "neutral" to the "current post."

[0775] Step 7:

[0776] The server evaluates security risks based on the analysis results and notifies the user. The input is the analysis results and the results of sentiment analysis, and the output is the notification to the user. The analysis results are compiled and sent as a report to the user's device via API. For example, the server may notify the user of information such as "the current post is contradictory," "it contains inappropriate words," or "it has strong negative sentiment."

[0777] Step 8:

[0778] The user receives and confirms the notification on the device. The input is the notification from the server, and the output is the user's confirmation result. The user can view the notification and take appropriate action. For example, they can take actions such as "correcting contradictory statements" or "correcting inappropriate expressions."

[0779] In this way, the entire system can monitor and analyze user comments in real time, minimizing security risks.

[0780] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0781] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0782] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0783] [Third embodiment]

[0784] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0785] 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.

[0786] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0787] 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.

[0788] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0789] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0790] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0791] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0792] 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.

[0793] 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.

[0794] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0795] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0796] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, and notification.

[0797] Program processing

[0798] Data collection

[0799] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (application programming interface). The audio data from lectures is converted into text data using voice recognition technology, and all data is stored in a single database.

[0800] Text analytics

[0801] The server then analyzes the collected data, using natural language processing (NLP) techniques to analyze the text data and understand its context and meaning. This analysis includes tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[0802] Inconsistency check

[0803] Based on the analyzed data, the server compares current and past statements, using similarity calculations and conflict detection algorithms to identify inconsistencies. For example, a statement that "environmental protection is important" could be deemed a contradiction with a past statement that "environmental issues are not a concern."

[0804] Inappropriate remarks check

[0805] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for words and phrases that are deemed inappropriate and analyzes their context. For example, if the word "idiot" is used, it checks the context before and after it to determine whether it is truly inappropriate.

[0806] Notification of results

[0807] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are identified, a detailed report is sent to the user's device. The user receives a notification through their device and can view the report to identify the problems with their own comments.

[0808] Specific examples

[0809] scenario

[0810] User C posted on social media that "technological advances will make society better." However, in the past, the same User C stated in a lecture that "technological advances will have a negative impact on society."

[0811] Data collection

[0812] The server collects User C's latest SNS posts via API and stores them in a database. It also collects audio data from the lecture, converts it into text using speech recognition technology, and stores it.

[0813] Text analytics

[0814] The server analyzes the collected social media posts and speeches using an NLP engine to understand their meaning and context, and performs processes such as tokenization, keyword extraction, and sentiment analysis.

[0815] Inconsistency check

[0816] The server compares the current social media post, "Technological advances make society better," with the past speech, "Technological advances have a negative impact on society," and determines that there is a contradiction.

[0817] Inappropriate remarks check

[0818] In this example, there is no particularly inappropriate language, so we will skip this step.

[0819] Notification of results

[0820] The server generates a report summarizing the analysis results that identified the contradiction and notifies the terminal of User C. User C views this report and confirms that there is a contradiction between the past and present statements.

[0821] In this way, the system for implementing the present invention allows users to maintain consistency in the content they post and avoid making inappropriate comments, thereby preventing misunderstandings and trouble.

[0822] The processing flow will be explained below.

[0823] Step 1:

[0824] The server uses an API that connects to the user's social media account to periodically check whether a new post has been made. If a new post is detected, the text data is stored in a database.

[0825] Step 2:

[0826] The server collects the audio data of the lecture in real time. The collected audio data is converted into text data using speech recognition technology. The converted text data is stored in a database.

[0827] Step 3:

[0828] Users upload digital files of book data to the server, which then converts the uploaded book data into text format and stores it in a database.

[0829] Step 4:

[0830] The server performs preprocessing on the collected text data, including tokenization, stop word removal, and normalization, before passing the preprocessed text data to a natural language processing engine.

[0831] Step 5:

[0832] The server uses natural language processing techniques to semantically analyze the text data, including keyword extraction, sentiment analysis, sentence segmentation, and syntactic analysis, to understand the specific context and meaning of each utterance.

[0833] Step 6:

[0834] The server compares current and past utterance data. Using similarity calculations and contradiction detection algorithms, it identifies contradictions between utterances. For example, if a contradiction is detected between the utterances "Technological advances improve society" and "Technological advances have a negative impact on society," it will identify it.

[0835] Step 7:

[0836] The server scans the text data using a specialized dictionary containing inappropriate words and phrases, and analyzes the context of identified inappropriate words and phrases to determine whether they are in fact inappropriate.

[0837] Step 8:

[0838] The server generates a report summarizing the analysis and evaluation results, including specific areas of inconsistency and inappropriate comments, and sends the report to the user's device.

[0839] Step 9:

[0840] Users receive a notification from the server via their device, which includes a link to a detailed report that allows them to identify the issues with their comments.

[0841] The above is a concrete flow of a series of processing steps for collecting and analyzing user comments, checking for inconsistencies and inappropriate content, and notifying the results.

[0842] Example 1

[0843] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0844] Inconsistent content posted by users or inappropriate comments can lead to misunderstandings and problems. This problem poses a challenge, as it can undermine the user's credibility and lower social credibility. Another problem is that manually checking the consistency and appropriateness of comments by users is time-consuming and inefficient.

[0845] 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.

[0846] In this invention, the server includes means for collecting content posted by users, means for storing the collected content in a database, means for analyzing the stored content using natural language processing technology, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, and means for notifying the user of the analysis results. This enables users to maintain consistency in the content they post and avoid making inappropriate posts.

[0847] "User" refers to any individual or organization that transmits information or opinions.

[0848] "Content" refers to a collection of user-generated text, audio, or digital data.

[0849] "Means of collection" refers to methods for collecting user-generated content using APIs and data acquisition methods of various platforms.

[0850] "Database" refers to a data storage system for storing and managing collected data.

[0851] "Means for storage" refers to the method for storing and managing collected data in a database.

[0852] "Natural language processing technology" refers to technology that analyzes text data and allows it to understand its meaning and context.

[0853] "Means of analysis" refers to a method of analyzing collected text data using natural language processing technology to extract specific information and context.

[0854] "Means for checking for inconsistencies" refers to a method of comparing the latest statements with past statements based on analyzed data to determine whether there are any inconsistencies.

[0855] "Measures to check for inappropriate content" refers to methods that use specialized dictionaries and algorithms to determine whether a comment contains inappropriate words or phrases.

[0856] "Means for notifying" refers to a method for transmitting the analysis results to the user's terminal and notifying the user.

[0857] "Tokenization" refers to the process of dividing text data into words or phrases.

[0858] "Stop word removal" refers to the process of removing common words that are not necessary for analysis from the text to be analyzed.

[0859] "Keyword extraction" refers to the process of extracting important words and phrases from text data.

[0860] "Sentiment analysis" refers to the process of determining emotions and emotional direction from the context of text data.

[0861] "Sentence segmentation" refers to the process of dividing text data into sentences or segments.

[0862] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and provides feedback. The system mainly comprises a server, a terminal, and a user. A specific embodiment of this invention will be described below.

[0863] System Overview

[0864] The system involves a server collecting user posts, analyzing them using natural language processing, detecting inconsistencies and inappropriate content, and notifying the user of the results via their device. The server uses the SNS API, voice recognition technology, a database management system, and a natural language processing library.

[0865] Data collection and storage

[0866] The server collects user posts using the SNS API and uses the Google Cloud Speech-to-Text API to convert the voice data. The collected data is then stored in a database such as MySQL.

[0867] Text analytics

[0868] The server analyzes the collected text data using natural language processing techniques, such as tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation, using libraries such as SpaCy and NLTK.

[0869] Inconsistency check and inappropriate remark check

[0870] The server compares the most recent and past comments based on the analysis results to identify inconsistencies. Similarity calculations are performed using cosine similarity, etc. To check for inappropriate comments, a dedicated dictionary is used to scan for specific words and phrases, and the context is also analyzed.

[0871] Notification of results

[0872] The server compiles the results of the analysis of detected inconsistencies and inappropriate comments and generates a report. This report is sent to the user's device, allowing the user to check the problematic aspects of the comments. Notifications are sent to the user's device using an API.

[0873] Specific examples

[0874] A specific example of the operation of the system is shown below.

[0875] scenario

[0876] A user posted on social media that "technological advances will improve society." However, in the past, the same user stated in a lecture that "technological advances will have a negative impact on society."

[0877] Data collection

[0878] The server collects the latest posts via the SNS API and stores them in a database. It also collects audio data from lectures, converts it into text using the Google Cloud Speech-to-Text API, and stores it.

[0879] Text analytics

[0880] The server analyzes the collected social media posts and lecture comments using natural language processing technology (e.g., SpaCy) to understand the meaning and context of each.

[0881] Inconsistency check

[0882] The server compared the social media post, "Technological advances make society better," with the speech, "Technological advances have a negative impact on society," and determined that there was a contradiction.

[0883] Inappropriate remarks check

[0884] In this example, there is no particularly inappropriate language, so we will skip this step.

[0885] Notification of results

[0886] The server generates a report summarizing the analysis results that identified the inconsistencies and sends it to the user's device. The user can view this report and check the inconsistencies in the comments.

[0887] Prompt Sentence Examples

[0888] "Write a program that compares a user's social media posts with their past statements to identify inconsistencies."

[0889] This invention allows users to ensure consistency in the content they post and avoid making inappropriate comments, thereby preventing misunderstandings and trouble.

[0890] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0891] Step 1: Data collection

[0892] The server collects content posted by users on social media, audio data from lectures, and digital data from books. Specifically, it acquires data using various APIs (such as APIs for social networking services and voice recognition APIs).

[0893] Input: Social media posts, lecture audio files, digital book data

[0894] Data processing: Convert the audio data from the lecture into text using the Google Cloud Speech-to-Text API.

[0895] Output: Text data

[0896] Specific operation: The server uses the SNS API to collect the latest user posts, sends the audio file of the lecture called audio_file.wav to the Google Cloud Speech-to-Text API, and obtains the text data "Technological advances have a negative impact on society."

[0897] Step 2: Save your data

[0898] The server stores the collected data in a database, where the data is organized for each user and stored in a collated format.

[0899] Input: Collected text data

[0900] Data processing: structuring data

[0901] Output: A structured database

[0902] What happens: The server connects to the database and inserts new data, for example by executing an SQL query like INSERT INTO user_data (user_id, content, type) VALUES (123, 'Technological advances make society better', 'SNS').

[0903] Step 3: Text analysis

[0904] The server analyzes the stored text data using natural language processing (NLP) techniques, including tokenization, stop-word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[0905] Input: Saved text data

[0906] Data processing: tokenization, stop word removal, keyword extraction, sentiment analysis, sentence segmentation

[0907] Output: Analysis results

[0908] Specific operation: The server uses the SpaCy library to analyze text data. For example, it loads the model with spacy.load("en_core_web_sm") and analyzes the text with nlp("Technological advances make society better"). It extracts keywords from the analysis results and creates a list such as "technology", "advancement", "society", and "make it better".

[0909] Step 4: Check for inconsistencies

[0910] The server compares the latest and past comments based on the analysis results, and identifies inconsistencies using similarity calculations (cosine similarity) and conflict detection algorithms.

[0911] Input: Text analysis results

[0912] Data processing: Similarity calculation, conflict detection

[0913] Output: Contradiction judgment result

[0914] Specific operation: The server calculates the vectors of past and current statements and evaluates their similarity. For example, it calculates the similarity using cosine_similarity(vector_a, vector_b), and if the result is below a certain threshold, it determines that there is a contradiction.

[0915] Step 5: Check for inappropriate comments

[0916] The server uses a dedicated dictionary to check whether the comment contains inappropriate words or phrases, and also analyzes the context to determine whether the comment is truly inappropriate.

[0917] Input: Text analysis results

[0918] Data processing: dictionary matching, context analysis

[0919] Output: Inappropriate judgment result

[0920] What it does: The server splits the text into words and compares each word to a list of inappropriate words, e.g., if word in inappropriate_words: flag_as_inappropriate(word) to detect inappropriate words.

[0921] Step 6: Notification of results

[0922] The server compiles the results of the analysis of inconsistencies and inappropriate comments, generates a report, and sends the report to the user's device, where the user can check it.

[0923] Input: Conflict judgment result, inappropriate judgment result

[0924] Data Processing: Report Generation

[0925] Output: Notification messages, reports

[0926] Specific operation: The server creates a report based on the analysis results and notifies the user via the API on the user's device. For example, it calls the PUT / user_notifications API and sends the notification data {"user_id": 123, "message": "Past and current statements are inconsistent"}.

[0927] (Application example 1)

[0928] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0929] In physical stores, it is difficult for staff to provide consistent information to customers. Correcting statements made in the past can damage customer trust. Furthermore, there is no way to check past statements in real time, so there is a risk of repeating the same mistake. This raises concerns about lower customer satisfaction and worsening operational efficiency.

[0930] 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.

[0931] In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for notifying the user of the analysis results, and means for collecting and analyzing data in real time based on the user's posts. This allows customer service staff to maintain consistency in customer service, prevent the provision of incorrect information, and maintain customer trust.

[0932] "Means for collecting user-generated content" refers to processes and technologies that automatically collect data entered or transmitted by users in the form of information networks, oral presentations, documents, etc.

[0933] "Means for analyzing collected content" are the processes and techniques used to analyze collected data and identify its meaning and context.

[0934] "Means for checking for inconsistencies with past statements based on the analysis results" refers to the process and techniques for comparing the analysis results with previous statements to see if there are any inconsistencies.

[0935] "Measures to check speech for inappropriate content" are processes and techniques that scan speech for inappropriate words, phrases, or context.

[0936] "Means for notifying the user of the analysis results" refers to the process and technology for notifying the user of the results obtained by the analysis so that the user can check them.

[0937] "Means for collecting and analyzing data in real time based on user comments" refers to the process and technology for quickly collecting comments made by users on the spot and analyzing them immediately.

[0938] This invention is a system that analyzes the content of user comments in a physical store in real time and provides feedback on the results. This system includes the following main processes.

[0939] Data collection

[0940] The server acquires voice data to collect what the user is saying. The voice data is collected through a microphone in the smart glasses and converted into text data using voice recognition technology. This process uses the Google Cloud Speech-to-Text API. The collected data is then stored in a database.

[0941] Text analytics

[0942] The server analyzes the collected text data using natural language processing (NLP) techniques, using Python libraries such as NLTK and spaCy, and performs processes such as tokenization, stop word removal, keyword extraction, and sentiment analysis.

[0943] Inconsistency check

[0944] Based on the analyzed data, the server compares the current utterance with the past utterances, and uses similarity calculations and conflict detection algorithms to identify inconsistencies. In this process, the Python scikit-learn library is used, for example, to calculate cosine similarity.

[0945] Inappropriate remarks check

[0946] The server checks for inappropriate words and phrases. It uses a Python dictionary database to scan for potentially inappropriate words and phrases, and analyzes the context to determine whether they are truly inappropriate.

[0947] Notification of results

[0948] If any inconsistencies or inappropriate comments are identified, the server will summarize the analysis results and notify the user. The notification will be displayed in real time on the smart glasses display. The smart glasses application will be developed for Android or iOS.

[0949] Specific examples

[0950] Consider a case where a user tells a customer in a physical store, "The point card is valid for one year," but in the past has said, "The point card is only valid for six months."

[0951] Prompt Sentence Examples

[0952] "Check your customer interaction records and see if there are any contradictions between current and past statements. Past statement: 'The loyalty card is only valid for six months.' Current statement: 'The loyalty card is valid for one year.'"

[0953] This system allows users to maintain consistency in customer service and prevent the provision of inappropriate information, which is expected to improve customer satisfaction and operational efficiency.

[0954] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0955] Step 1:

[0956] Data collection

[0957] The server collects what the user says through a microphone installed in the smart glasses. The collected voice data is converted into text data using the Google Cloud Speech-to-Text API. The converted text data is then stored in a database.

[0958] Input: Audio data

[0959] Processing: Convert speech to text (using Google Cloud Speech-to-Text API)

[0960] Output: Text data (stored in database)

[0961] Step 2:

[0962] Text analytics

[0963] The server analyzes the collected text data using Python natural language processing libraries (such as NLTK or spaCy), specifically performing tokenization, stop word removal, keyword extraction, and sentiment analysis.

[0964] Input: Text data

[0965] Processing: Tokenization, stopword removal, keyword extraction, sentiment analysis (using NLTK and spaCy libraries)

[0966] Output: Parsed text data

[0967] Step 3:

[0968] Inconsistency check

[0969] The server compares current and past statements based on the parsed text data, using the Python scikit-learn library to identify inconsistencies using algorithms such as cosine similarity calculations.

[0970] Input: Analyzed text data, past text data

[0971] Processing: Similarity calculation (using the scikit-learn library)

[0972] Output: Conflicts

[0973] Step 4:

[0974] Inappropriate remarks check

[0975] The server then scans the parsed text data for inappropriate words and phrases, using a Python dictionary database and analyzing the context to make its decisions.

[0976] Input: Parsed text data

[0977] Processing: Scanning for inappropriate words and phrases (using dictionary database)

[0978] Output: Whether or not there is inappropriate remarks

[0979] Step 5:

[0980] Notification of results

[0981] Based on the results of the inconsistencies and inappropriate comments, the server compiles the analysis results and notifies the user's smart glasses in real time, which are displayed on the smart glasses' display.

[0982] Input: Whether there are contradictions or not, whether there are inappropriate comments or not

[0983] Processing: Creating and sending analysis results

[0984] Output: Feedback notification to the smart glasses display

[0985] This series of processes allows users to instantly correct any inconsistencies with past comments or inappropriate comments when dealing with customers in a physical store.

[0986] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0987] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and then analyzes the user's emotions using an emotion engine and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, emotion analysis, and notification.

[0988] Program processing

[0989] Data collection

[0990] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (application programming interface). The audio data from lectures is converted into text data using voice recognition technology, and all data is stored in a single database.

[0991] Text analytics

[0992] The server then analyzes the collected data, using natural language processing (NLP) techniques to analyze the text data and understand its context and meaning. This analysis includes tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[0993] Inconsistency check

[0994] Based on the analyzed data, the server compares current and past statements. It uses similarity calculations and conflict detection algorithms to identify contradictions. For example, a statement that "technological advances make society better" could be deemed a contradiction with a past statement that "technological advances have a negative impact on society."

[0995] Inappropriate remarks check

[0996] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for words and phrases that are deemed inappropriate and analyzes their context. For example, if the word "idiot" is used, it checks the context before and after it to determine whether it is truly inappropriate.

[0997] Emotion analysis

[0998] A distinctive feature of this invention is that the server uses an emotion engine to analyze the user's emotions from text data. This emotion engine uses NLP technology to identify emotions from the content of the user's speech. For example, it identifies emotions such as "very happy" or "very angry." This allows for a detailed understanding of the emotions expressed by the user's speech.

[0999] Notification of results

[1000] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are detected, a detailed report including the results of the sentiment analysis is sent to the user's device. The user receives the notification via their device and can view the report to identify problems with their own comments and fluctuations in sentiment.

[1001] Specific examples

[1002] scenario

[1003] User D posted on social media that "technological advances will make society better." However, in the past, the same User D stated at a lecture that "technological advances will have a negative impact on society." The social media post also expresses the emotion of "being very happy."

[1004] Data collection

[1005] The server collects the latest SNS posts from User D via API and stores them in a database. It also collects audio data from the lecture, converts it into text using speech recognition technology, and stores it.

[1006] Text analytics

[1007] The server analyzes the collected social media posts and speeches using an NLP engine to understand their meaning and context, and performs processes such as tokenization, keyword extraction, and sentiment analysis.

[1008] Inconsistency check

[1009] The server compares the current social media post, "Technological advances make society better," with the past speech, "Technological advances have a negative impact on society," and determines that there is a contradiction.

[1010] Inappropriate remarks check

[1011] In this example, there is no particularly inappropriate language, so we will skip this step.

[1012] Emotion analysis

[1013] The server analyzes the emotion "very happy" from the latest social media posts and adds it to the analysis results.

[1014] Notification of results

[1015] The server generates a report summarizing the analysis results that identified the contradictions and the results of the emotion analysis, and notifies the device of User D. User D views this report and confirms that there is a contradiction between his past and present statements, and that his latest post expresses the emotion of "very happy."

[1016] In this way, the system for implementing this invention not only allows users to maintain consistency in their speech and emotions and avoid inappropriate speech, but also allows users to accurately grasp emotional fluctuations, thereby preventing misunderstandings and troubles and deepening understanding of emotional states.

[1017] The processing flow will be explained below.

[1018] Step 1:

[1019] The server uses an API that connects to the user's social media account to periodically check whether a new post has been made. If a new post is detected, the text data is stored in a database.

[1020] Step 2:

[1021] The server collects the audio data of the lecture in real time. The collected audio data is converted into text data using speech recognition technology. The converted text data is stored in a database.

[1022] Step 3:

[1023] Users upload digital files of book data to the server, which then converts the uploaded book data into text format and stores it in a database.

[1024] Step 4:

[1025] The server performs preprocessing on the collected text data, including tokenization, stop word removal, and normalization, before passing the preprocessed text data to a natural language processing engine.

[1026] Step 5:

[1027] The server uses natural language processing techniques to semantically analyze the text data, including keyword extraction, sentiment analysis, sentence segmentation, and syntactic analysis, to understand the specific context and meaning of each utterance.

[1028] Step 6:

[1029] The server compares current and past utterance data. Using similarity calculations and contradiction detection algorithms, it identifies contradictions between utterances. For example, if a contradiction is detected between the utterances "Technological advances improve society" and "Technological advances have a negative impact on society," it will identify it.

[1030] Step 7:

[1031] The server scans the text data using a specialized dictionary containing inappropriate words and phrases, and analyzes the context of identified inappropriate words and phrases to determine whether they are in fact inappropriate.

[1032] Step 8:

[1033] The server uses an emotion engine to analyze the emotions in the user's comments. The emotion engine uses NLP technology to identify, for example, whether the comment contains the emotion "very happy."

[1034] Step 9:

[1035] The server generates a report summarizing the analysis and evaluation results, including identified inconsistencies, inappropriate comments, and sentiment analysis results, and sends the report to the user's device.

[1036] Step 10:

[1037] Users receive notifications from the server via their devices, which include a link to a detailed report that allows them to identify problems with their own comments and fluctuations in sentiment.

[1038] The above is a concrete flow of a series of processing steps that collects and analyzes user comments, checks for inconsistencies and inappropriate content, and provides feedback including the results and sentiment analysis.

[1039] Example 2

[1040] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1041] Conventional systems have difficulty detecting inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, etc., and lack a mechanism for analyzing users' emotions and providing feedback. This situation makes it difficult for users to maintain consistency in their posts and avoid inappropriate comments. Therefore, an objective of this invention is to provide a system that automatically and efficiently detects inconsistencies and inappropriate content in posts and analyzes emotions.

[1042] 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.

[1043] In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for analyzing the user's emotions using an emotion analysis engine, and means for notifying the user of the analysis results. This makes it possible to analyze the collected data and check the consistency and appropriateness of the posts, and further to analyze the user's emotions in detail and provide feedback.

[1044] "User" means any person or entity that uses the Software or System.

[1045] "Means of collection" refers to the ability to automatically or manually obtain data from sources such as social media posts, lecture notes, and book data.

[1046] "Means of analysis" refers to the function of using natural language processing technology to understand the context and meaning of the acquired data and extract the necessary information.

[1047] "Contradiction checkers" refer to algorithms or methods used to compare current statements with past statements to check for consistency and identify inconsistencies.

[1048] "Measures to check for inappropriate content" refers to a function that detects whether the collected data contains inappropriate words or phrases and prompts warnings or corrections as necessary.

[1049] "Sentiment analysis engine" refers to software or a system that uses natural language processing techniques and other sentiment analysis tools to extract user sentiment from text data and calculate a specific sentiment score.

[1050] "Notification means" refers to a method or system for notifying the user of the analysis results, and in particular refers to a function that sends information to the user's terminal so that the results can be viewed.

[1051] "Natural language processing technology" is a technology that enables computers to understand, interpret, and generate human language, and includes processes such as tokenization, stop word removal, and sentiment analysis.

[1052] "SNS post" refers to text or media content that a user publicly posts via a social networking service (SNS).

[1053] "Lecture recordings" refer to the content of lectures and seminars that have been audio-visually recorded and saved as digital data.

[1054] "Book data" refers to data that stores the contents of books and documents in digital format.

[1055] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and then analyzes the user's emotions using an emotion engine and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, emotion analysis, and notification.

[1056] Program processing

[1057] Data collection

[1058] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (e.g., Twitter API, Facebook Graph API). The audio data from lectures is converted into text data using the Google Cloud Speech-to-Text service, and all data is stored in a MySQL database.

[1059] Text analytics

[1060] The server analyzes the collected data using natural language processing (NLP) techniques using the Python libraries NLTK (Natural Language Toolkit) and Spacy, including tokenization, stop word removal, keyword extraction using TF-IDF (inverse document frequency), sentiment analysis, and sentence segmentation.

[1061] Inconsistency check

[1062] Based on the analyzed data, the server compares current and past statements using Python's difflib library to calculate similarities and identify inconsistencies. For example, if a user posts that "technological advances improve society" and then previously states that "technological advances have a negative impact on society," this will be detected as a contradiction.

[1063] Inappropriate remarks check

[1064] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for inappropriate words and phrases and analyzes their context. It uses dictionary data from NLTK and Spacy to determine whether a word like "idiot" is truly inappropriate in the context.

[1065] Emotion analysis

[1066] The server analyzes the user's emotions from the text data using an emotion engine. This emotion engine also uses NLP technology, such as VADER and TextBlob, to identify emotions from the speech content. For example, it calculates an emotion score such as "very happy" or "very angry."

[1067] Notification of results

[1068] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are detected, a detailed report including the results of the sentiment analysis is generated and sent to the user's device. The user receives the notification via their device and can view the report to identify problems with their own comments and fluctuations in sentiment.

[1069] Specific examples

[1070] User D posted on social media that "technological advances make society better," but in a past lecture he said that "technological advances have a negative impact on society." In addition, his latest social media post expresses the emotion of being "very happy."

[1071] The server collects User D's latest social media posts via the Twitter API and stores them in a MySQL database. It also collects audio data from lectures, converts it to text using Google Cloud Speech-to-Text, and stores it. The server then analyzes the data using the NLTK and Spacy NLP engines, and performs comparison and sentiment analysis. The server compiles the analysis results into a report and sends it to User D's device. User D can view this report to check for inconsistencies between past and present statements and the sentiment behind the latest posts.

[1072] Specific prompt examples

[1073] "Compare your past social media posts with your most recent ones to detect inconsistencies."

[1074] "Please analyze the sentiment of my social media posts and let me know the results."

[1075] "Please check whether what was said at the lecture matches what was posted on social media."

[1076] In this way, by inputting prompt sentences into the generative AI model, the system can operate effectively and provide the information the user is looking for.

[1077] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1078] Step 1: Data collection

[1079] The server collects content posted by users on various social networking sites, audio data from lectures, and digital data from books. Specifically, the server operates as follows:

[1080] Input: User ID or platform information to be collected

[1081] The server uses the Twitter API to retrieve post data related to the specified user ID in JSON format.

[1082] The server uses Google Cloud Speech-to-Text to convert the lecture's audio data (e.g., MP3 files) into text data.

[1083] The server parses the digital book data (e.g., PDF or EPUB file) and extracts the text content.

[1084] Output: Database records containing various collected text data

[1085] Step 2: Data Preprocessing

[1086] The server preprocesses the collected text data and prepares it in a format suitable for analysis.

[1087] Input: Raw text data

[1088] The server uses the Python library NLTK to tokenize (divide) the text data into words.

[1089] Remove stop words (e.g., frequently occurring words such as "wa" and "ga").

[1090] Calculate TF-IDF (inverse document frequency) and extract important keywords.

[1091] Output: Preprocessed text data

[1092] Step 3: Check for inconsistencies

[1093] The server analyzes the preprocessed text data and compares the current utterances with past utterances.

[1094] Input: Preprocessed text data

[1095] The server uses the Python difflib library to calculate the similarity scores of statements.

[1096] Identify when current statements contradict past statements above a certain threshold.

[1097] Output: Analysis results if inconsistencies are identified

[1098] Step 4: Check for inappropriate comments

[1099] The server checks the text data for inappropriate words or phrases.

[1100] Input: Preprocessed text data

[1101] The server uses dictionary data from NLTK and Spacy to scan for inappropriate words and phrases.

[1102] The context of any inappropriate words or phrases found is analyzed to determine whether they are truly inappropriate.

[1103] Output: Analysis results when inappropriate comments are identified

[1104] Step 5: Sentiment Analysis

[1105] The server utilizes an emotion engine that analyzes the user's emotions from the text data.

[1106] Input: Preprocessed text data

[1107] The server uses VADER or TextBlob to calculate the sentiment score for each piece of text.

[1108] Identify and score specific emotions (e.g., "happiness," "anger," "sadness," etc.).

[1109] Output: Sentiment analysis results

[1110] Step 6: Notification of results

[1111] The server compiles all the analysis results, generates a detailed report, and notifies the user.

[1112] Input: Results of inconsistency check, inappropriate comment check, and sentiment analysis

[1113] The server generates a report in JSON format and sends it to the user's device.

[1114] The user's terminal receives this report and displays and notifies the analysis results.

[1115] Output: User notification and detailed report

[1116] In this way, specific data processing and calculations are performed at each step, and the results of each process are passed on to the next step, allowing for a comprehensive analysis of the inconsistencies, inappropriateness, and emotional state of the user's comments.

[1117] (Application example 2)

[1118] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1119] Conventional systems have had difficulty detecting inappropriate expressions in user comments or inconsistencies with past comments. Furthermore, they lacked the functionality to understand users' emotional state through emotion analysis and assess security risks. This increased the risk of user comments causing misunderstandings and problems, hindering security.

[1120] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for analyzing the user's emotions using an emotion engine, and means for evaluating and notifying security risks based on the analysis results. This makes it possible to detect security risks early and respond effectively while maintaining the consistency and appropriateness of the content posted by users.

[1121] "User" refers to any individual or organizational representative who uses the system to make a statement.

[1122] "Content to be transmitted" refers to text data, audio data, and video data that users publish through social media, communication platforms, lectures, digital books, etc.

[1123] "Means of collection" refers to programs and APIs that automatically obtain user comments from various sources on the Internet.

[1124] "Means for analysis" refers to software or systems for analyzing the meaning of collected data using natural language processing techniques or other analytical algorithms.

[1125] "Means for checking for inconsistencies" refers to a program that compares current statements with past statements and detects inconsistencies when the content does not match.

[1126] "Measures to check for inappropriate content" refers to systems that use specialized dictionaries and contextual analysis to automatically detect inappropriate words and phrases.

[1127] "Means of notification" refers to the protocols and systems used to notify users of the results of the analysis, any detected inconsistencies, inappropriate content, and sentiment analysis.

[1128] "Emotion engine" refers to natural language processing technology used to analyze a user's emotional state from text data.

[1129] A means of assessing "security risk" refers to a system that determines whether a user's statements or actions pose a security risk based on the analysis results and notifies the user.

[1130] "Security risk" refers to any factor that could potentially threaten the security of a company or organization through user statements or actions.

[1131] The system of this invention collects content posted by users, analyzes it to detect inconsistencies and inappropriate content, evaluates security risks using an emotion engine, and provides feedback to users. A specific method for realizing this system is described below.

[1132] First, an API (Application Programming Interface) is used to collect content posted by users on social media, communication platforms, lectures, digital books, etc. A cloud-based server automatically retrieves content from these data sources and stores it in a database. Audio data from lectures is converted into text data using speech recognition technology.

[1133] The collected data is analyzed using natural language processing (NLP) techniques, including tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation, specifically using spaCy (spacy.io) and VaderSentiment (vaderSentiment).

[1134] The server then compares the current and past comments to check for inconsistencies. This is done using similarity calculations and conflict detection algorithms such as SequenceMatcher (difflib). The collected text data is also checked for inappropriate words and phrases. To detect inappropriate content, the server uses specialized dictionaries and context analysis.

[1135] Furthermore, an emotion engine is used to perform sentiment analysis. VaderSentiment is used to analyze the user's emotional state, such as "positive," "negative," or "neutral," from text data. This sentiment analysis is important for clarifying the emotions reflected in the user's comments.

[1136] The analysis results are sent to the user's device via a dedicated notification protocol or API, allowing the user to understand the problems and security risks of their own comments.

[1137] For example, if a user posts something like, "Today I received confidential information from a client," the system will immediately analyze it and detect security risks such as the leakage of confidential information. Inconsistencies, inappropriate content, and emotional states are analyzed, and an alert is sent to the user.

[1138] Examples of prompts include:

[1139] Analyze the latest SNS post by user ID "user123" titled "Today I received confidential information from a client." and check the following items:

[1140] 1. Are there any contradictions with previous posts?

[1141] 2. Does it contain inappropriate language?

[1142] 3. Analyze the emotional state and report the results.”

[1143] In this way, users can see in real time whether their speech is coherent, inappropriate, and reflects their emotional state, minimizing security risks and providing appropriate feedback.

[1144] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1145] Step 1:

[1146] The server collects content posted by users. In this case, data is obtained through the API of a social networking site or communication platform. The input is data from the API, and the output is collected text data or audio data. For example, the server collects the latest social networking post from "user123."

[1147] Step 2:

[1148] The server converts the collected voice data into text data using voice recognition technology. The input is voice data, and the output is the converted text data. Voice recognition software is used for this process. For example, "voice data of a lecture" is converted into "text of the lecture content."

[1149] Step 3:

[1150] The server analyzes the text data. Here, natural language processing (NLP) techniques are used to analyze the data and understand its context and meaning. The input is the text data, and the output is the analysis results. Specific operations include tokenization, stop word removal, and keyword extraction. For example, the server breaks down the "text of a lecture" into "individual words and phrases" and extracts important keywords.

[1151] Step 4:

[1152] The server uses the analysis results to check for inconsistencies between current and past comments. The input is the analyzed current text data and past text data, and the output is whether there are any inconsistencies. Similarity calculations and conflict detection algorithms are used. For example, by comparing the "current post" with the "content of a past lecture," inconsistencies are identified.

[1153] Step 5:

[1154] The server checks whether a post contains inappropriate content. The input is analyzed text data, and the output is whether or not there are any inappropriate comments. It uses a dedicated dictionary and context analysis. For example, it compares a "list of inappropriate words" with the "current post" to see if it contains any inappropriate words.

[1155] Step 6:

[1156] The server analyzes the user's emotions using an emotion engine. The input is the analyzed text data, and the output is the result of the emotion analysis. VaderSentiment is used to determine the user's emotional state from the text. For example, it assigns an emotion label of "positive," "negative," or "neutral" to the "current post."

[1157] Step 7:

[1158] The server evaluates security risks based on the analysis results and notifies the user. The input is the analysis results and the results of sentiment analysis, and the output is the notification to the user. The analysis results are compiled and sent as a report to the user's device via API. For example, the server may notify the user of information such as "the current post is contradictory," "it contains inappropriate words," or "it has strong negative sentiment."

[1159] Step 8:

[1160] The user receives and confirms the notification on the device. The input is the notification from the server, and the output is the user's confirmation result. The user can view the notification and take appropriate action. For example, they can take actions such as "correcting contradictory statements" or "correcting inappropriate expressions."

[1161] In this way, the entire system can monitor and analyze user comments in real time, minimizing security risks.

[1162] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1163] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1164] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1165] [Fourth embodiment]

[1166] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1167] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1168] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1169] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1170] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1171] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1172] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1173] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1174] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1175] 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.

[1176] 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.

[1177] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1178] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1179] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, and notification.

[1180] Program processing

[1181] Data collection

[1182] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (application programming interface). The audio data from lectures is converted into text data using voice recognition technology, and all data is stored in a single database.

[1183] Text analytics

[1184] The server then analyzes the collected data, using natural language processing (NLP) techniques to analyze the text data and understand its context and meaning. This analysis includes tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[1185] Inconsistency check

[1186] Based on the analyzed data, the server compares current and past statements, using similarity calculations and conflict detection algorithms to identify inconsistencies. For example, a statement that "environmental protection is important" could be deemed a contradiction with a past statement that "environmental issues are not a concern."

[1187] Inappropriate remarks check

[1188] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for words and phrases that are deemed inappropriate and analyzes their context. For example, if the word "idiot" is used, it checks the context before and after it to determine whether it is truly inappropriate.

[1189] Notification of results

[1190] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are identified, a detailed report is sent to the user's device. The user receives a notification through their device and can view the report to identify the problems with their own comments.

[1191] Specific examples

[1192] scenario

[1193] User C posted on social media that "technological advances will make society better." However, in the past, the same User C stated in a lecture that "technological advances will have a negative impact on society."

[1194] Data collection

[1195] The server collects User C's latest SNS posts via API and stores them in a database. It also collects audio data from the lecture, converts it into text using speech recognition technology, and stores it.

[1196] Text analytics

[1197] The server analyzes the collected social media posts and speeches using an NLP engine to understand their meaning and context, and performs processes such as tokenization, keyword extraction, and sentiment analysis.

[1198] Inconsistency check

[1199] The server compares the current social media post, "Technological advances make society better," with the past speech, "Technological advances have a negative impact on society," and determines that there is a contradiction.

[1200] Inappropriate remarks check

[1201] In this example, there is no particularly inappropriate language, so we will skip this step.

[1202] Notification of results

[1203] The server generates a report summarizing the analysis results that identified the contradiction and notifies the terminal of User C. User C views this report and confirms that there is a contradiction between the past and present statements.

[1204] In this way, the system for implementing the present invention allows users to maintain consistency in the content they post and avoid making inappropriate comments, thereby preventing misunderstandings and trouble.

[1205] The processing flow will be explained below.

[1206] Step 1:

[1207] The server uses an API that connects to the user's social media account to periodically check whether a new post has been made. If a new post is detected, the text data is stored in a database.

[1208] Step 2:

[1209] The server collects the audio data of the lecture in real time. The collected audio data is converted into text data using speech recognition technology. The converted text data is stored in a database.

[1210] Step 3:

[1211] Users upload digital files of book data to the server, which then converts the uploaded book data into text format and stores it in a database.

[1212] Step 4:

[1213] The server performs preprocessing on the collected text data, including tokenization, stop word removal, and normalization, before passing the preprocessed text data to a natural language processing engine.

[1214] Step 5:

[1215] The server uses natural language processing techniques to semantically analyze the text data, including keyword extraction, sentiment analysis, sentence segmentation, and syntactic analysis, to understand the specific context and meaning of each utterance.

[1216] Step 6:

[1217] The server compares current and past utterance data. Using similarity calculations and contradiction detection algorithms, it identifies contradictions between utterances. For example, if a contradiction is detected between the utterances "Technological advances improve society" and "Technological advances have a negative impact on society," it will identify it.

[1218] Step 7:

[1219] The server scans the text data using a specialized dictionary containing inappropriate words and phrases, and analyzes the context of identified inappropriate words and phrases to determine whether they are in fact inappropriate.

[1220] Step 8:

[1221] The server generates a report summarizing the analysis and evaluation results, including specific areas of inconsistency and inappropriate comments, and sends the report to the user's device.

[1222] Step 9:

[1223] Users receive a notification from the server via their device, which includes a link to a detailed report that allows them to identify the issues with their comments.

[1224] The above is a concrete flow of a series of processing steps for collecting and analyzing user comments, checking for inconsistencies and inappropriate content, and notifying the results.

[1225] Example 1

[1226] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1227] Inconsistent content posted by users or inappropriate comments can lead to misunderstandings and problems. This problem poses a challenge, as it can undermine the user's credibility and lower social credibility. Another problem is that manually checking the consistency and appropriateness of comments by users is time-consuming and inefficient.

[1228] 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.

[1229] In this invention, the server includes means for collecting content posted by users, means for storing the collected content in a database, means for analyzing the stored content using natural language processing technology, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, and means for notifying the user of the analysis results. This enables users to maintain consistency in the content they post and avoid making inappropriate posts.

[1230] "User" refers to any individual or organization that transmits information or opinions.

[1231] "Content" refers to a collection of user-generated text, audio, or digital data.

[1232] "Means of collection" refers to methods for collecting user-generated content using APIs and data acquisition methods of various platforms.

[1233] "Database" refers to a data storage system for storing and managing collected data.

[1234] "Means for storage" refers to the method for storing and managing collected data in a database.

[1235] "Natural language processing technology" refers to technology that analyzes text data and allows it to understand its meaning and context.

[1236] "Means of analysis" refers to a method of analyzing collected text data using natural language processing technology to extract specific information and context.

[1237] "Means for checking for inconsistencies" refers to a method of comparing the latest statements with past statements based on analyzed data to determine whether there are any inconsistencies.

[1238] "Measures to check for inappropriate content" refers to methods that use specialized dictionaries and algorithms to determine whether a comment contains inappropriate words or phrases.

[1239] "Means for notifying" refers to a method for transmitting the analysis results to the user's terminal and notifying the user.

[1240] "Tokenization" refers to the process of dividing text data into words or phrases.

[1241] "Stop word removal" refers to the process of removing common words that are not necessary for analysis from the text to be analyzed.

[1242] "Keyword extraction" refers to the process of extracting important words and phrases from text data.

[1243] "Sentiment analysis" refers to the process of determining emotions and emotional direction from the context of text data.

[1244] "Sentence segmentation" refers to the process of dividing text data into sentences or segments.

[1245] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and provides feedback. The system mainly comprises a server, a terminal, and a user. A specific embodiment of this invention will be described below.

[1246] System Overview

[1247] The system involves a server collecting user posts, analyzing them using natural language processing, detecting inconsistencies and inappropriate content, and notifying the user of the results via their device. The server uses the SNS API, voice recognition technology, a database management system, and a natural language processing library.

[1248] Data collection and storage

[1249] The server collects user posts using the SNS API and uses the Google Cloud Speech-to-Text API to convert the voice data. The collected data is then stored in a database such as MySQL.

[1250] Text analytics

[1251] The server analyzes the collected text data using natural language processing techniques, such as tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation, using libraries such as SpaCy and NLTK.

[1252] Inconsistency check and inappropriate remark check

[1253] The server compares the most recent and past comments based on the analysis results to identify inconsistencies. Similarity calculations are performed using cosine similarity, etc. To check for inappropriate comments, a dedicated dictionary is used to scan for specific words and phrases, and the context is also analyzed.

[1254] Notification of results

[1255] The server compiles the results of the analysis of detected inconsistencies and inappropriate comments and generates a report. This report is sent to the user's device, allowing the user to check the problematic aspects of the comments. Notifications are sent to the user's device using an API.

[1256] Specific examples

[1257] A specific example of the operation of the system is shown below.

[1258] scenario

[1259] A user posted on social media that "technological advances will improve society." However, in the past, the same user stated in a lecture that "technological advances will have a negative impact on society."

[1260] Data collection

[1261] The server collects the latest posts via the SNS API and stores them in a database. It also collects audio data from lectures, converts it into text using the Google Cloud Speech-to-Text API, and stores it.

[1262] Text analytics

[1263] The server analyzes the collected social media posts and lecture comments using natural language processing technology (e.g., SpaCy) to understand the meaning and context of each.

[1264] Inconsistency check

[1265] The server compared the social media post, "Technological advances make society better," with the speech, "Technological advances have a negative impact on society," and determined that there was a contradiction.

[1266] Inappropriate remarks check

[1267] In this example, there is no particularly inappropriate language, so we will skip this step.

[1268] Notification of results

[1269] The server generates a report summarizing the analysis results that identified the inconsistencies and sends it to the user's device. The user can view this report and check the inconsistencies in the comments.

[1270] Prompt Sentence Examples

[1271] "Write a program that compares a user's social media posts with their past statements to identify inconsistencies."

[1272] This invention allows users to ensure consistency in the content they post and avoid making inappropriate comments, thereby preventing misunderstandings and trouble.

[1273] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1274] Step 1: Data collection

[1275] The server collects content posted by users on social media, audio data from lectures, and digital data from books. Specifically, it acquires data using various APIs (such as APIs for social networking services and voice recognition APIs).

[1276] Input: Social media posts, lecture audio files, digital book data

[1277] Data processing: Convert the audio data from the lecture into text using the Google Cloud Speech-to-Text API.

[1278] Output: Text data

[1279] Specific operation: The server uses the SNS API to collect the latest user posts, sends the audio file of the lecture called audio_file.wav to the Google Cloud Speech-to-Text API, and obtains the text data "Technological advances have a negative impact on society."

[1280] Step 2: Save your data

[1281] The server stores the collected data in a database, where the data is organized for each user and stored in a collated format.

[1282] Input: Collected text data

[1283] Data processing: structuring data

[1284] Output: A structured database

[1285] What happens: The server connects to the database and inserts new data, for example by executing an SQL query like INSERT INTO user_data (user_id, content, type) VALUES (123, 'Technological advances make society better', 'SNS').

[1286] Step 3: Text analysis

[1287] The server analyzes the stored text data using natural language processing (NLP) techniques, including tokenization, stop-word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[1288] Input: Saved text data

[1289] Data processing: tokenization, stop word removal, keyword extraction, sentiment analysis, sentence segmentation

[1290] Output: Analysis results

[1291] Specific operation: The server uses the SpaCy library to analyze text data. For example, it loads the model with spacy.load("en_core_web_sm") and analyzes the text with nlp("Technological advances make society better"). It extracts keywords from the analysis results and creates a list such as "technology", "advancement", "society", and "make it better".

[1292] Step 4: Check for inconsistencies

[1293] The server compares the latest and past comments based on the analysis results, and identifies inconsistencies using similarity calculations (cosine similarity) and conflict detection algorithms.

[1294] Input: Text analysis results

[1295] Data processing: Similarity calculation, conflict detection

[1296] Output: Contradiction judgment result

[1297] Specific operation: The server calculates the vectors of past and current statements and evaluates their similarity. For example, it calculates the similarity using cosine_similarity(vector_a, vector_b), and if the result is below a certain threshold, it determines that there is a contradiction.

[1298] Step 5: Check for inappropriate comments

[1299] The server uses a dedicated dictionary to check whether the comment contains inappropriate words or phrases, and also analyzes the context to determine whether the comment is truly inappropriate.

[1300] Input: Text analysis results

[1301] Data processing: dictionary matching, context analysis

[1302] Output: Inappropriate judgment result

[1303] What it does: The server splits the text into words and compares each word to a list of inappropriate words, e.g., if word in inappropriate_words: flag_as_inappropriate(word) to detect inappropriate words.

[1304] Step 6: Notification of results

[1305] The server compiles the results of the analysis of inconsistencies and inappropriate comments, generates a report, and sends the report to the user's device, where the user can check it.

[1306] Input: Conflict judgment result, inappropriate judgment result

[1307] Data Processing: Report Generation

[1308] Output: Notification messages, reports

[1309] Specific operation: The server creates a report based on the analysis results and notifies the user via the API on the user's device. For example, it calls the PUT / user_notifications API and sends the notification data {"user_id": 123, "message": "Past and current statements are inconsistent"}.

[1310] (Application example 1)

[1311] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1312] In physical stores, it is difficult for staff to provide consistent information to customers. Correcting statements made in the past can damage customer trust. Furthermore, there is no way to check past statements in real time, so there is a risk of repeating the same mistake. This raises concerns about lower customer satisfaction and worsening operational efficiency.

[1313] 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.

[1314] In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for notifying the user of the analysis results, and means for collecting and analyzing data in real time based on the user's posts. This allows customer service staff to maintain consistency in customer service, prevent the provision of incorrect information, and maintain customer trust.

[1315] "Means for collecting user-generated content" refers to processes and technologies that automatically collect data entered or transmitted by users in the form of information networks, oral presentations, documents, etc.

[1316] "Means for analyzing collected content" are the processes and techniques used to analyze collected data and identify its meaning and context.

[1317] "Means for checking for inconsistencies with past statements based on the analysis results" refers to the process and techniques for comparing the analysis results with previous statements to see if there are any inconsistencies.

[1318] "Measures to check speech for inappropriate content" are processes and techniques that scan speech for inappropriate words, phrases, or context.

[1319] "Means for notifying the user of the analysis results" refers to the process and technology for notifying the user of the results obtained by the analysis so that the user can check them.

[1320] "Means for collecting and analyzing data in real time based on user comments" refers to the process and technology for quickly collecting comments made by users on the spot and analyzing them immediately.

[1321] This invention is a system that analyzes the content of user comments in a physical store in real time and provides feedback on the results. This system includes the following main processes.

[1322] Data collection

[1323] The server acquires voice data to collect what the user is saying. The voice data is collected through a microphone in the smart glasses and converted into text data using voice recognition technology. This process uses the Google Cloud Speech-to-Text API. The collected data is then stored in a database.

[1324] Text analytics

[1325] The server analyzes the collected text data using natural language processing (NLP) techniques, using Python libraries such as NLTK and spaCy, and performs processes such as tokenization, stop word removal, keyword extraction, and sentiment analysis.

[1326] Inconsistency check

[1327] Based on the analyzed data, the server compares the current utterance with the past utterances, and uses similarity calculations and conflict detection algorithms to identify inconsistencies. In this process, the Python scikit-learn library is used, for example, to calculate cosine similarity.

[1328] Inappropriate remarks check

[1329] The server checks for inappropriate words and phrases. It uses a Python dictionary database to scan for potentially inappropriate words and phrases, and analyzes the context to determine whether they are truly inappropriate.

[1330] Notification of results

[1331] If any inconsistencies or inappropriate comments are identified, the server will summarize the analysis results and notify the user. The notification will be displayed in real time on the smart glasses display. The smart glasses application will be developed for Android or iOS.

[1332] Specific examples

[1333] Consider a case where a user tells a customer in a physical store, "The point card is valid for one year," but in the past has said, "The point card is only valid for six months."

[1334] Prompt Sentence Examples

[1335] "Check your customer interaction records and see if there are any contradictions between current and past statements. Past statement: 'The loyalty card is only valid for six months.' Current statement: 'The loyalty card is valid for one year.'"

[1336] This system allows users to maintain consistency in customer service and prevent the provision of inappropriate information, which is expected to improve customer satisfaction and operational efficiency.

[1337] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1338] Step 1:

[1339] Data collection

[1340] The server collects what the user says through a microphone installed in the smart glasses. The collected voice data is converted into text data using the Google Cloud Speech-to-Text API. The converted text data is then stored in a database.

[1341] Input: Audio data

[1342] Processing: Convert speech to text (using Google Cloud Speech-to-Text API)

[1343] Output: Text data (stored in database)

[1344] Step 2:

[1345] Text analytics

[1346] The server analyzes the collected text data using Python natural language processing libraries (such as NLTK or spaCy), specifically performing tokenization, stop word removal, keyword extraction, and sentiment analysis.

[1347] Input: Text data

[1348] Processing: Tokenization, stopword removal, keyword extraction, sentiment analysis (using NLTK and spaCy libraries)

[1349] Output: Parsed text data

[1350] Step 3:

[1351] Inconsistency check

[1352] The server compares current and past statements based on the parsed text data, using the Python scikit-learn library to identify inconsistencies using algorithms such as cosine similarity calculations.

[1353] Input: Analyzed text data, past text data

[1354] Processing: Similarity calculation (using the scikit-learn library)

[1355] Output: Conflicts

[1356] Step 4:

[1357] Inappropriate remarks check

[1358] The server then scans the parsed text data for inappropriate words and phrases, using a Python dictionary database and analyzing the context to make its decisions.

[1359] Input: Parsed text data

[1360] Processing: Scanning for inappropriate words and phrases (using dictionary database)

[1361] Output: Whether or not there is inappropriate remarks

[1362] Step 5:

[1363] Notification of results

[1364] Based on the results of the inconsistencies and inappropriate comments, the server compiles the analysis results and notifies the user's smart glasses in real time, which are displayed on the smart glasses' display.

[1365] Input: Whether there are contradictions or not, whether there are inappropriate comments or not

[1366] Processing: Creating and sending analysis results

[1367] Output: Feedback notification to the smart glasses display

[1368] This series of processes allows users to instantly correct any inconsistencies with past comments or inappropriate comments when dealing with customers in a physical store.

[1369] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1370] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and then analyzes the user's emotions using an emotion engine and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, emotion analysis, and notification.

[1371] Program processing

[1372] Data collection

[1373] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (application programming interface). The audio data from lectures is converted into text data using voice recognition technology, and all data is stored in a single database.

[1374] Text analytics

[1375] The server then analyzes the collected data, using natural language processing (NLP) techniques to analyze the text data and understand its context and meaning. This analysis includes tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[1376] Inconsistency check

[1377] Based on the analyzed data, the server compares current and past statements. It uses similarity calculations and conflict detection algorithms to identify contradictions. For example, a statement that "technological advances make society better" could be deemed a contradiction with a past statement that "technological advances have a negative impact on society."

[1378] Inappropriate remarks check

[1379] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for words and phrases that are deemed inappropriate and analyzes their context. For example, if the word "idiot" is used, it checks the context before and after it to determine whether it is truly inappropriate.

[1380] Emotion analysis

[1381] A distinctive feature of this invention is that the server uses an emotion engine to analyze the user's emotions from text data. This emotion engine uses NLP technology to identify emotions from the content of the user's speech. For example, it identifies emotions such as "very happy" or "very angry." This allows for a detailed understanding of the emotions expressed by the user's speech.

[1382] Notification of results

[1383] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are detected, a detailed report including the results of the sentiment analysis is sent to the user's device. The user receives the notification via their device and can view the report to identify problems with their own comments and fluctuations in sentiment.

[1384] Specific examples

[1385] scenario

[1386] User D posted on social media that "technological advances will make society better." However, in the past, the same User D stated at a lecture that "technological advances will have a negative impact on society." The social media post also expresses the emotion of "being very happy."

[1387] Data collection

[1388] The server collects the latest SNS posts from User D via API and stores them in a database. It also collects audio data from the lecture, converts it into text using speech recognition technology, and stores it.

[1389] Text analytics

[1390] The server analyzes the collected social media posts and speeches using an NLP engine to understand their meaning and context, and performs processes such as tokenization, keyword extraction, and sentiment analysis.

[1391] Inconsistency check

[1392] The server compares the current social media post, "Technological advances make society better," with the past speech, "Technological advances have a negative impact on society," and determines that there is a contradiction.

[1393] Inappropriate remarks check

[1394] In this example, there is no particularly inappropriate language, so we will skip this step.

[1395] Emotion analysis

[1396] The server analyzes the emotion "very happy" from the latest social media posts and adds it to the analysis results.

[1397] Notification of results

[1398] The server generates a report summarizing the analysis results that identified the contradictions and the results of the emotion analysis, and notifies the device of User D. User D views this report and confirms that there is a contradiction between his past and present statements, and that his latest post expresses the emotion of "very happy."

[1399] In this way, the system for implementing this invention not only allows users to maintain consistency in their speech and emotions and avoid inappropriate speech, but also allows users to accurately grasp emotional fluctuations, thereby preventing misunderstandings and troubles and deepening understanding of emotional states.

[1400] The processing flow will be explained below.

[1401] Step 1:

[1402] The server uses an API that connects to the user's social media account to periodically check whether a new post has been made. If a new post is detected, the text data is stored in a database.

[1403] Step 2:

[1404] The server collects the audio data of the lecture in real time. The collected audio data is converted into text data using speech recognition technology. The converted text data is stored in a database.

[1405] Step 3:

[1406] Users upload digital files of book data to the server, which then converts the uploaded book data into text format and stores it in a database.

[1407] Step 4:

[1408] The server performs preprocessing on the collected text data, including tokenization, stop word removal, and normalization, before passing the preprocessed text data to a natural language processing engine.

[1409] Step 5:

[1410] The server uses natural language processing techniques to semantically analyze the text data, including keyword extraction, sentiment analysis, sentence segmentation, and syntactic analysis, to understand the specific context and meaning of each utterance.

[1411] Step 6:

[1412] The server compares current and past utterance data. Using similarity calculations and contradiction detection algorithms, it identifies contradictions between utterances. For example, if a contradiction is detected between the utterances "Technological advances improve society" and "Technological advances have a negative impact on society," it will identify it.

[1413] Step 7:

[1414] The server scans the text data using a specialized dictionary containing inappropriate words and phrases, and analyzes the context of identified inappropriate words and phrases to determine whether they are in fact inappropriate.

[1415] Step 8:

[1416] The server uses an emotion engine to analyze the emotions in the user's comments. The emotion engine uses NLP technology to identify, for example, whether the comment contains the emotion "very happy."

[1417] Step 9:

[1418] The server generates a report summarizing the analysis and evaluation results, including identified inconsistencies, inappropriate comments, and sentiment analysis results, and sends the report to the user's device.

[1419] Step 10:

[1420] Users receive notifications from the server via their devices, which include a link to a detailed report that allows them to identify problems with their own comments and fluctuations in sentiment.

[1421] The above is a concrete flow of a series of processing steps that collects and analyzes user comments, checks for inconsistencies and inappropriate content, and provides feedback including the results and sentiment analysis.

[1422] Example 2

[1423] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1424] Conventional systems have difficulty detecting inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, etc., and lack a mechanism for analyzing users' emotions and providing feedback. This situation makes it difficult for users to maintain consistency in their posts and avoid inappropriate comments. Therefore, an objective of this invention is to provide a system that automatically and efficiently detects inconsistencies and inappropriate content in posts and analyzes emotions.

[1425] 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.

[1426] In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for analyzing the user's emotions using an emotion analysis engine, and means for notifying the user of the analysis results. This makes it possible to analyze the collected data and check the consistency and appropriateness of the posts, and further to analyze the user's emotions in detail and provide feedback.

[1427] "User" means any person or entity that uses the Software or System.

[1428] "Means of collection" refers to the ability to automatically or manually obtain data from sources such as social media posts, lecture notes, and book data.

[1429] "Means of analysis" refers to the function of using natural language processing technology to understand the context and meaning of the acquired data and extract the necessary information.

[1430] "Contradiction checkers" refer to algorithms or methods used to compare current statements with past statements to check for consistency and identify inconsistencies.

[1431] "Measures to check for inappropriate content" refers to a function that detects whether the collected data contains inappropriate words or phrases and prompts warnings or corrections as necessary.

[1432] "Sentiment analysis engine" refers to software or a system that uses natural language processing techniques and other sentiment analysis tools to extract user sentiment from text data and calculate a specific sentiment score.

[1433] "Notification means" refers to a method or system for notifying the user of the analysis results, and in particular refers to a function that sends information to the user's terminal so that the results can be viewed.

[1434] "Natural language processing technology" is a technology that enables computers to understand, interpret, and generate human language, and includes processes such as tokenization, stop word removal, and sentiment analysis.

[1435] "SNS post" refers to text or media content that a user publicly posts via a social networking service (SNS).

[1436] "Lecture recordings" refer to the content of lectures and seminars that have been audio-visually recorded and saved as digital data.

[1437] "Book data" refers to data that stores the contents of books and documents in digital format.

[1438] This invention relates to a system that automatically detects inconsistencies with past posts or inappropriate content in posts posted by users on social media, at lectures, in books, etc., and then analyzes the user's emotions using an emotion engine and provides feedback. This system is composed of a program that performs a series of processes: data collection, analysis, evaluation, emotion analysis, and notification.

[1439] Program processing

[1440] Data collection

[1441] The server collects content posted by users on social media, audio from lectures, and digital data from books. To do this, the server can automatically obtain data using each platform's API (e.g., Twitter API, Facebook Graph API). The audio data from lectures is converted into text data using the Google Cloud Speech-to-Text service, and all data is stored in a MySQL database.

[1442] Text analytics

[1443] The server analyzes the collected data using natural language processing (NLP) techniques using the Python libraries NLTK (Natural Language Toolkit) and Spacy, including tokenization, stop word removal, keyword extraction using TF-IDF (inverse document frequency), sentiment analysis, and sentence segmentation.

[1444] Inconsistency check

[1445] Based on the analyzed data, the server compares current and past statements using Python's difflib library to calculate similarities and identify inconsistencies. For example, if a user posts that "technological advances improve society" and then previously states that "technological advances have a negative impact on society," this will be detected as a contradiction.

[1446] Inappropriate remarks check

[1447] The server checks for inappropriate words and phrases. It uses a dedicated dictionary to scan for inappropriate words and phrases and analyzes their context. It uses dictionary data from NLTK and Spacy to determine whether a word like "idiot" is truly inappropriate in the context.

[1448] Emotion analysis

[1449] The server analyzes the user's emotions from the text data using an emotion engine. This emotion engine also uses NLP technology, such as VADER and TextBlob, to identify emotions from the speech content. For example, it calculates an emotion score such as "very happy" or "very angry."

[1450] Notification of results

[1451] Finally, the server summarizes the analysis results and notifies the user. If any inconsistencies or inappropriate comments are detected, a detailed report including the results of the sentiment analysis is generated and sent to the user's device. The user receives the notification via their device and can view the report to identify problems with their own comments and fluctuations in sentiment.

[1452] Specific examples

[1453] User D posted on social media that "technological advances make society better," but in a past lecture he said that "technological advances have a negative impact on society." In addition, his latest social media post expresses the emotion of being "very happy."

[1454] The server collects User D's latest social media posts via the Twitter API and stores them in a MySQL database. It also collects audio data from lectures, converts it to text using Google Cloud Speech-to-Text, and stores it. The server then analyzes the data using the NLTK and Spacy NLP engines, and performs comparison and sentiment analysis. The server compiles the analysis results into a report and sends it to User D's device. User D can view this report to check for inconsistencies between past and present statements and the sentiment behind the latest posts.

[1455] Specific prompt examples

[1456] "Compare your past social media posts with your most recent ones to detect inconsistencies."

[1457] "Please analyze the sentiment of my social media posts and let me know the results."

[1458] "Please check whether what was said at the lecture matches what was posted on social media."

[1459] In this way, by inputting prompt sentences into the generative AI model, the system can operate effectively and provide the information the user is looking for.

[1460] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1461] Step 1: Data collection

[1462] The server collects content posted by users on various social networking sites, audio data from lectures, and digital data from books. Specifically, the server operates as follows:

[1463] Input: User ID or platform information to be collected

[1464] The server uses the Twitter API to retrieve post data related to the specified user ID in JSON format.

[1465] The server uses Google Cloud Speech-to-Text to convert the lecture's audio data (e.g., MP3 files) into text data.

[1466] The server parses the digital book data (e.g., PDF or EPUB file) and extracts the text content.

[1467] Output: Database records containing various collected text data

[1468] Step 2: Data Preprocessing

[1469] The server preprocesses the collected text data and prepares it in a format suitable for analysis.

[1470] Input: Raw text data

[1471] The server uses the Python library NLTK to tokenize (divide) the text data into words.

[1472] Remove stop words (e.g., frequently occurring words such as "wa" and "ga").

[1473] Calculate TF-IDF (inverse document frequency) and extract important keywords.

[1474] Output: Preprocessed text data

[1475] Step 3: Check for inconsistencies

[1476] The server analyzes the preprocessed text data and compares the current utterances with past utterances.

[1477] Input: Preprocessed text data

[1478] The server uses the Python difflib library to calculate the similarity scores of statements.

[1479] Identify when current statements contradict past statements above a certain threshold.

[1480] Output: Analysis results if inconsistencies are identified

[1481] Step 4: Check for inappropriate comments

[1482] The server checks the text data for inappropriate words or phrases.

[1483] Input: Preprocessed text data

[1484] The server uses dictionary data from NLTK and Spacy to scan for inappropriate words and phrases.

[1485] The context of any inappropriate words or phrases found is analyzed to determine whether they are truly inappropriate.

[1486] Output: Analysis results when inappropriate comments are identified

[1487] Step 5: Sentiment Analysis

[1488] The server utilizes an emotion engine that analyzes the user's emotions from the text data.

[1489] Input: Preprocessed text data

[1490] The server uses VADER or TextBlob to calculate the sentiment score for each piece of text.

[1491] Identify and score specific emotions (e.g., "happiness," "anger," "sadness," etc.).

[1492] Output: Sentiment analysis results

[1493] Step 6: Notification of results

[1494] The server compiles all the analysis results, generates a detailed report, and notifies the user.

[1495] Input: Results of inconsistency check, inappropriate comment check, and sentiment analysis

[1496] The server generates a report in JSON format and sends it to the user's device.

[1497] The user's terminal receives this report and displays and notifies the analysis results.

[1498] Output: User notification and detailed report

[1499] In this way, specific data processing and calculations are performed at each step, and the results of each process are passed on to the next step, allowing for a comprehensive analysis of the inconsistencies, inappropriateness, and emotional state of the user's comments.

[1500] (Application example 2)

[1501] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1502] Conventional systems have had difficulty detecting inappropriate expressions in user comments or inconsistencies with past comments. Furthermore, they lacked the functionality to understand users' emotional state through emotion analysis and assess security risks. This increased the risk of user comments causing misunderstandings and problems, hindering security.

[1503] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting content posted by users, means for analyzing the collected content, means for checking for inconsistencies with past posts based on the analysis results, means for checking whether the posts contain inappropriate content, means for analyzing the user's emotions using an emotion engine, and means for evaluating and notifying security risks based on the analysis results. This makes it possible to detect security risks early and respond effectively while maintaining the consistency and appropriateness of the content posted by users.

[1504] "User" refers to any individual or organizational representative who uses the system to make a statement.

[1505] "Content to be transmitted" refers to text data, audio data, and video data that users publish through social media, communication platforms, lectures, digital books, etc.

[1506] "Means of collection" refers to programs and APIs that automatically obtain user comments from various sources on the Internet.

[1507] "Means for analysis" refers to software or systems for analyzing the meaning of collected data using natural language processing techniques or other analytical algorithms.

[1508] "Means for checking for inconsistencies" refers to a program that compares current statements with past statements and detects inconsistencies when the content does not match.

[1509] "Measures to check for inappropriate content" refers to systems that use specialized dictionaries and contextual analysis to automatically detect inappropriate words and phrases.

[1510] "Means of notification" refers to the protocols and systems used to notify users of the results of the analysis, any detected inconsistencies, inappropriate content, and sentiment analysis.

[1511] "Emotion engine" refers to natural language processing technology used to analyze a user's emotional state from text data.

[1512] A means of assessing "security risk" refers to a system that determines whether a user's statements or actions pose a security risk based on the analysis results and notifies the user.

[1513] "Security risk" refers to any factor that could potentially threaten the security of a company or organization through user statements or actions.

[1514] The system of this invention collects content posted by users, analyzes it to detect inconsistencies and inappropriate content, evaluates security risks using an emotion engine, and provides feedback to users. A specific method for realizing this system is described below.

[1515] First, an API (Application Programming Interface) is used to collect content posted by users on social media, communication platforms, lectures, digital books, etc. A cloud-based server automatically retrieves content from these data sources and stores it in a database. Audio data from lectures is converted into text data using speech recognition technology.

[1516] The collected data is analyzed using natural language processing (NLP) techniques, including tokenization, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation, specifically using spaCy (spacy.io) and VaderSentiment (vaderSentiment).

[1517] The server then compares the current and past comments to check for inconsistencies. This is done using similarity calculations and conflict detection algorithms such as SequenceMatcher (difflib). The collected text data is also checked for inappropriate words and phrases. To detect inappropriate content, the server uses specialized dictionaries and context analysis.

[1518] Furthermore, an emotion engine is used to perform sentiment analysis. VaderSentiment is used to analyze the user's emotional state, such as "positive," "negative," or "neutral," from text data. This sentiment analysis is important for clarifying the emotions reflected in the user's comments.

[1519] The analysis results are sent to the user's device via a dedicated notification protocol or API, allowing the user to understand the problems and security risks of their own comments.

[1520] For example, if a user posts something like, "Today I received confidential information from a client," the system will immediately analyze it and detect security risks such as the leakage of confidential information. Inconsistencies, inappropriate content, and emotional states are analyzed, and an alert is sent to the user.

[1521] Examples of prompts include:

[1522] Analyze the latest SNS post by user ID "user123" titled "Today I received confidential information from a client." and check the following items:

[1523] 1. Are there any contradictions with previous posts?

[1524] 2. Does it contain inappropriate language?

[1525] 3. Analyze the emotional state and report the results.”

[1526] In this way, users can see in real time whether their speech is coherent, inappropriate, and reflects their emotional state, minimizing security risks and providing appropriate feedback.

[1527] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1528] Step 1:

[1529] The server collects content posted by users. In this case, data is obtained through the API of a social networking site or communication platform. The input is data from the API, and the output is collected text data or audio data. For example, the server collects the latest social networking post from "user123."

[1530] Step 2:

[1531] The server converts the collected voice data into text data using voice recognition technology. The input is voice data, and the output is the converted text data. Voice recognition software is used for this process. For example, "voice data of a lecture" is converted into "text of the lecture content."

[1532] Step 3:

[1533] The server analyzes the text data. Here, natural language processing (NLP) techniques are used to analyze the data and understand its context and meaning. The input is the text data, and the output is the analysis results. Specific operations include tokenization, stop word removal, and keyword extraction. For example, the server breaks down the "text of a lecture" into "individual words and phrases" and extracts important keywords.

[1534] Step 4:

[1535] The server uses the analysis results to check for inconsistencies between current and past comments. The input is the analyzed current text data and past text data, and the output is whether there are any inconsistencies. Similarity calculations and conflict detection algorithms are used. For example, by comparing the "current post" with the "content of a past lecture," inconsistencies are identified.

[1536] Step 5:

[1537] The server checks whether a post contains inappropriate content. The input is analyzed text data, and the output is whether or not there are any inappropriate comments. It uses a dedicated dictionary and context analysis. For example, it compares a "list of inappropriate words" with the "current post" to see if it contains any inappropriate words.

[1538] Step 6:

[1539] The server analyzes the user's emotions using an emotion engine. The input is the analyzed text data, and the output is the result of the emotion analysis. VaderSentiment is used to determine the user's emotional state from the text. For example, it assigns an emotion label of "positive," "negative," or "neutral" to the "current post."

[1540] Step 7:

[1541] The server evaluates security risks based on the analysis results and notifies the user. The input is the analysis results and the results of sentiment analysis, and the output is the notification to the user. The analysis results are compiled and sent as a report to the user's device via API. For example, the server may notify the user of information such as "the current post is contradictory," "it contains inappropriate words," or "it has strong negative sentiment."

[1542] Step 8:

[1543] The user receives and confirms the notification on the device. The input is the notification from the server, and the output is the user's confirmation result. The user can view the notification and take appropriate action. For example, they can take actions such as "correcting contradictory statements" or "correcting inappropriate expressions."

[1544] In this way, the entire system can monitor and analyze user comments in real time, minimizing security risks.

[1545] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1546] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1547] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1548] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1549] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1550] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1551] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1552] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1553] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1554] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1555] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1556] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1557] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1558] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1559] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1560] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1561] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1562] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1563] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1564] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1565] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1566] The following is further disclosed regarding the above embodiment.

[1567] (Claim 1)

[1568] A means for collecting user-generated content;

[1569] a means for analyzing the collected content; and

[1570] Based on the analysis results, there is a means to check for contradictions with past statements,

[1571] A means of checking whether comments contain inappropriate content;

[1572] The system includes a means for notifying the user of the analysis results.

[1573] (Claim 2)

[1574] The system of claim 1, wherein the analyzing means utilizes natural language processing techniques.

[1575] (Claim 3)

[1576] The system according to claim 1, wherein the collection means collects data from social media posts, lecture records, and book data.

[1577] "Example 1"

[1578] (Claim 1)

[1579] A means for collecting user-generated content;

[1580] a means for storing the collected content in a database;

[1581] A means for analyzing the stored content using natural language processing technology;

[1582] Based on the analysis results, there is a means to check for contradictions with past statements,

[1583] A means of checking whether comments contain inappropriate content;

[1584] The system includes a means for notifying the user of the analysis results.

[1585] (Claim 2)

[1586] 2. The system of claim 1, wherein the analysis means performs processing including tokenization of text data, stop word removal, keyword extraction, sentiment analysis, and sentence segmentation.

[1587] (Claim 3)

[1588] 2. The system according to claim 1, wherein the collecting means collects data from posts on social networking services, audio data from lectures, and data from books.

[1589] "Application Example 1"

[1590] (Claim 1)

[1591] A means for collecting user-generated content;

[1592] a means for analyzing the collected content; and

[1593] Based on the analysis results, there is a means to check for contradictions with past statements,

[1594] A means of checking whether comments contain inappropriate content;

[1595] a means for notifying the user of the analysis results;

[1596] A system that includes a means for collecting and analyzing data in real time based on user utterances.

[1597] (Claim 2)

[1598] The system of claim 1, wherein the analyzing means utilizes natural language processing techniques.

[1599] (Claim 3)

[1600] 10. The system of claim 1, wherein the collecting means collects data from information network posts, oral presentation records, and document data.

[1601] "Example 2: Combining Emotion Engines"

[1602] (Claim 1)

[1603] A means for collecting user-generated content;

[1604] a means for analyzing the collected content; and

[1605] Based on the analysis results, there is a means to check for contradictions with past statements,

[1606] A means of checking whether comments contain inappropriate content;

[1607] A means for analyzing user emotions using an emotion analysis engine;

[1608] The system includes a means for notifying the user of the analysis results.

[1609] (Claim 2)

[1610] The system of claim 1, wherein the analyzing means utilizes natural language processing techniques.

[1611] (Claim 3)

[1612] The system according to claim 1, wherein the collection means collects data from social media posts, lecture records, and book data.

[1613] "Application example 2 when combining emotion engines"

[1614] (Claim 1)

[1615] A means for collecting user-generated content;

[1616] a means for analyzing the collected content; and

[1617] Based on the analysis results, there is a means to check for contradictions with past statements,

[1618] A means of checking whether comments contain inappropriate content;

[1619] a means for notifying the user of the analysis results;

[1620] means for analyzing a user's emotions using an emotion engine;

[1621] A means for assessing and notifying security risks based on the analysis results;

[1622] A system including:

[1623] (Claim 2)

[1624] The system of claim 1, wherein the analyzing means utilizes natural language processing techniques.

[1625] (Claim 3)

[1626] 10. The system of claim 1, wherein the collecting means collects data from communication platform posts, lecture recordings, and digital book data. [Explanation of symbols]

[1627] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for collecting user-generated content; a means for analyzing the collected content; and Based on the analysis results, there is a means to check for contradictions with past statements, A means of checking whether comments contain inappropriate content; The system includes a means for notifying the user of the analysis results.

2. The system of claim 1 , wherein the analyzing means utilizes natural language processing techniques.

3. The system according to claim 1 , wherein the collection means collects data from social media posts, lecture records, and book data.

Images