system

Abstract

To provide a system.SOLUTION: A system includes: means for collecting and accumulating flaming data of the past; means for collecting and accumulating data of effective words; means for analyzing using a natural language processing technique on the basis of the flaming data and the effective words; means for predicting economical effects of public information on the basis of an analysis result; and means for supporting at least either correction of creation of public information.SELECTED DRAWING: Figure 1

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot 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 an explanation of the chatbot's 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

Summary of the Invention

Problems to be Solved by the Invention

[0004] There is a risk that information disclosed by companies on SNS or websites will go viral and the corporate image will decline. In addition, there is a problem that the impact of the disclosed information is weak and the economic effect is low.

Means for Solving the Problems

[0005] Provided is a system that utilizes past viral data and effective word data and uses AI to predict the economic effect when these data are disclosed on SNS or websites. Furthermore, when proofreading, grasp the amount of loss when past viral data is used, and prevent the decline of the corporate image in advance.

Brief Description of the Drawings

[0006] [Figure 1]This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Embodiment 1 of Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1 of Form Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2 of Embodiment 2. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2 of Form Example 2. [Figure 15] This is a sequence diagram showing the processing flow of the data processing system in Embodiment 3 of Example 3. [Figure 16] This is a sequence diagram showing the processing flow of the data processing system in Application Example 3 of Form Example 3. [Figure 17] It is a sequence diagram showing the processing flow of the data processing system in Example 1 of Form Example 1 when combined with an emotion engine. [Figure 18] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1 of Form Example 1 when combined with an emotion engine. [Figure 19] It is a sequence diagram showing the processing flow of the data processing system in Example 2 of Form Example 2 when combined with an emotion engine. [Figure 20] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 of Form Example 2 when combined with an emotion engine. [Figure 21] It is a sequence diagram showing the processing flow of the data processing system in Example 3 of Form Example 3 when combined with an emotion engine. [Figure 22] It is a sequence diagram showing the processing flow of the data processing system in Application Example 3 of Form Example 3 when combined with an emotion engine.

Embodiments for Carrying Out the Invention

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

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

[0009] In the following embodiments, the signed processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Furthermore, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include CPU (Central Processing Unit), GPU (Graphics Processing Unit), GPGPU (General-Purpose computing on Graphics Processing Units), APU (Accelerated Processing Unit), or TPU (TENSOR PROCESSING UNIT®).

[0010] In the following embodiments, signed RAM (Random Access Memory) is a memory that temporarily stores information and is used as work memory by the processor.

[0011] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0012] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0013] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0014] [First Embodiment]

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

[0016] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0017] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0018] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0019] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

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

[0021] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

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

[0023] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0024] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

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

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

[0027] "Example of form 1"

[0028] One embodiment of this system is an AI-powered information disclosure support system for businesses. This system utilizes past data on online controversies and effective keywords to predict the economic impact of information published on social media and websites. Specifically, it stores past online controversy data in a database, which is then analyzed by AI. Similarly, it stores data on effective keywords, and uses this data to edit and create publicly available information.

[0029] "Example of form 2"

[0030] Furthermore, another embodiment of the present invention includes a function to determine the amount of loss incurred when using past online controversy data during the editing process. Specifically, the amount of loss incurred during a controversy is estimated from past controversy data and used as reference information for editing. This makes it possible to quantify the risk of online controversy and disclose information that takes that risk into account.

[0031] "Example of form 3"

[0032] Furthermore, a further embodiment of the present invention includes a function to proactively prevent damage to a company's image. Specifically, before information created or edited by the AI ​​is made public, it evaluates whether the information contains a risk of causing a public outcry and prompts for revisions as necessary. This makes it possible to proactively prevent damage to a company's image.

[0033] The following describes the processing flow for each example of the form.

[0034] "Example of form 1"

[0035] Step 1: Accumulate past online controversy data and effective keyword data in a database.

[0036] Step 2: The AI ​​analyzes this data and edits or creates publicly available information.

[0037] Step 3: Based on the economic impact predicted by the AI, determine the optimal timing and content for information disclosure.

[0038] "Example of form 2"

[0039] Step 1: Estimate the amount of losses incurred during a social media firestorm based on past data from similar incidents.

[0040] Step 2: Use the estimated loss amount as reference information for editing.

[0041] Step 3: Quantify the risk of online backlash and disclose information based on that risk.

[0042] "Example of form 3"

[0043] Step 1: Before the information created or edited by the AI ​​is published, evaluate whether it contains a risk of causing a public outcry.

[0044] Step 2: Based on the evaluation results, prompt for information correction as needed.

[0045] Step 3: Release the revised information to prevent damage to the company's image.

[0046] (Example 1)

[0047] Next, we will describe Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0048] When companies publish information on social media or their websites, they face the challenge of predicting the risk of online backlash and the resulting economic impact. In particular, there is a need for methods to mitigate the risks of publicly released information and maximize its economic benefits by utilizing past online controversies and effective wording. Furthermore, preventing damage to the company's image is also a crucial issue.

[0049] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 1 is realized by the following means. In this invention, the server includes means for collecting and storing past online controversy data, means for collecting and storing effective word data, means for analyzing this data using natural language processing technology, means for predicting the economic effects of publicly available information based on the analysis results, and means for supporting the editing and creation of publicly available information. This makes it possible for companies to reduce the risk of online controversy related to publicly available information and maximize economic effects. It also makes it possible to prevent damage to the company's image in advance.

[0050] "Past online controversy data" refers to information about past online controversies that have occurred on social media or websites, and specifically includes post content, comments, retweet counts, and user reactions.

[0051] "Effective word data" refers to information about keywords and phrases that elicit positive responses on social media and websites, specifically including keywords and phrases that have received favorable user reactions.

[0052] "Natural language processing technology" refers to the technology that enables computers to understand, analyze, and generate human language, and specifically includes sentiment analysis and keyword extraction.

[0053] "Predicting economic impact" refers to evaluating how much traffic publicly available information will attract on social media and websites, or the degree of risk of it going viral or causing a controversy.

[0054] "Proofreading and creation support for publicly available information" refers to improving the quality of user-created information by reducing the risk of online backlash and suggesting effective wording.

[0055] "Preventing damage to a company's image" refers to measures taken to prevent a company's reputation from deteriorating due to information it makes public.

[0056] Modes for carrying out the invention

[0057] This invention is an information disclosure support system designed to reduce the risk of online backlash and maximize economic benefits when companies publish information on social media and their websites. The system collects and stores past online backlash data and effective keyword data, and uses natural language processing technology to analyze this data. Furthermore, it predicts the economic impact of the published information based on the analysis results and supports the editing and creation of such information.

[0058] Data collection and storage

[0059] The server collects past online controversy data from social media and websites. Specifically, it uses social media APIs to search for posts containing specific keywords and hashtags, and retrieves data such as comments and retweet counts related to the controversies. It also uses marketing research tools to collect data on effective keywords. For example, it uses Google Analytics and SEMrush to analyze how much traffic specific keywords have attracted. The collected data is stored in databases such as MySQL and PostgreSQL.

[0060] Data analysis

[0061] The server analyzes accumulated online controversy data using TENSORFLOW® and PyTorch®. Specifically, it uses natural language processing (NLP) techniques to analyze the sentiment of posts and identify factors that cause online controversies. Similarly, it analyzes data on effective words to identify keywords and phrases that elicit positive responses. For example, it uses models such as Word2Vec and BERT to evaluate the relevance of keywords.

[0062] Predicting the economic impact

[0063] Based on the analysis results, the server predicts the economic impact of the information to be released. Specifically, it evaluates how much traffic the information scheduled for release will attract and the risk of it causing a public outcry. This prediction is made by combining historical data with an AI model.

[0064] Editing and creating publicly available information

[0065] Users create public information and send it from their device to the server. For example, they might input a new product description or campaign announcement for a company. The server evaluates the risk of online backlash based on the submitted information. Specifically, it analyzes the input text using NLP technology and assesses the risk by comparing it with past online backlash data. The server suggests effective words and revises the information. For example, it might make specific suggestions such as, "This expression has a high risk of causing a backlash, so please change it to this expression." Users revise the information based on the server's suggestions and create the final public information. They can also send the revised information back to the server for final confirmation.

[0066] Specific example

[0067] Example 1: Assessment of the risk of online backlash

[0068] A user creates a product description for a new product and sends it to the server from their device. The server evaluates the risk of this description causing a backlash based on past data on product controversies. For example, if the server determines that a phrase like "This product is superior to competitors' products" increases the risk of backlash, it suggests changing it to something like "This product is well-received by many users."

[0069] Example 2: Suggestions for effective words

[0070] The user creates a campaign announcement and sends it from their device to the server. The server, based on a database of effective words, suggests keywords that will elicit a positive response to the announcement. For example, if keywords such as "limited," "free," and "special offer" are deemed effective, the server suggests adding these keywords to the announcement.

[0071] Example of a prompt

[0072] "I have created a product description for our new product. Please assess the risk of this description causing a backlash if it is published on social media, and suggest revisions as needed."

[0073] "I've created a campaign announcement. Please suggest effective words to elicit a positive response."

[0074] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0075] Step 1: Data collection and storage

[0076] The server collects past online controversy data from social media and websites. Specifically, it uses social media APIs to search for posts containing specific keywords and hashtags, and retrieves data such as comments and retweet counts related to the controversies. The input is specific keywords and hashtags, and the output is the collected controversy data. The server stores the collected data in databases such as MySQL® or PostgreSQL®.

[0077] Step 2: Effective word data collection

[0078] The server uses marketing research tools to collect data on effective keywords. For example, it uses Google Analytics or SEMrush to analyze how much traffic specific keywords attract. The input is the specific keyword, and the output is data on effective keywords. The server stores the collected data in a database.

[0079] Step 3: Data Analysis

[0080] The server analyzes accumulated online controversy data using TensorFlow® and PyTorch®. Specifically, it uses natural language processing (NLP) techniques to analyze the sentiment of posts and identify the factors causing online controversies. The input is accumulated online controversy data, and the output is the identification of the factors causing the controversies. Similarly, it analyzes effective word data to identify keywords and phrases that elicit positive responses. The input is effective word data, and the output is keywords and phrases that elicit positive responses.

[0081] Step 4: Predicting the Economic Impact

[0082] The server predicts the economic impact of publicly released information based on the analysis results. Specifically, it evaluates how much traffic the information scheduled for release will attract and the risk of it causing a public outcry. The input is the analysis results, and the output is the predicted economic impact.

[0083] Step 5: Create and submit public information

[0084] Users create public information and send it from their terminal to the server. For example, they might input a company's new product introduction or campaign announcement. The input is the public information created by the user, and the output is the public information sent to the server.

[0085] Step 6: Evaluation and proposals for publicly available information

[0086] The server assesses the risk of online backlash based on the submitted information. Specifically, it analyzes the input text using NLP technology and evaluates the risk by comparing it with past online backlash data. The input is publicly available information created by the user, and the output is the assessment result of the online backlash risk. The server suggests effective words and edits the information. For example, it makes specific suggestions such as, "This expression has a high risk of causing online backlash, so please change it to this expression." The input is the assessment result of the online backlash risk, and the output is the suggested revisions.

[0087] Step 7: Correction and final confirmation of publicly available information

[0088] Users revise the information based on the server's suggestions to create the final published information. They can also resubmit the revised information to the server for final confirmation. The input is the server's revision suggestions, and the output is the revised published information.

[0089] (Application Example 1)

[0090] Next, we will describe Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server," and the smart device 14 will be referred to as a "terminal."

[0091] When companies publish advertisements on social media and their websites, they are required to predict the economic impact by utilizing past data on online controversies and effective keywords. However, there are no systems that effectively utilize this data to support the editing and creation of ad copy. Furthermore, there is a lack of means to proactively prevent damage to a company's image by predicting online controversy risk scores and effective keyword scores. As a result, companies are forced to operate their advertising campaigns while carrying inherent risks.

[0092] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0093] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including AI to predict the economic effects when published on social media and websites based on this data, means for assisting in editing and creating ad copy, and means for predicting online controversy risk scores and effective word scores. This enables companies to reduce risks when publishing advertisements and create effective ad copy.

[0094] "Past online controversy data" refers to information about past online controversies that have occurred on social media or websites.

[0095] "Effective word data" refers to information about words and phrases that are highly effective in advertising and information dissemination.

[0096] "AI that predicts economic impact" refers to artificial intelligence that predicts the economic impact of information published on social media and websites based on past data.

[0097] "Means of supporting the editing and creation of ad copy" refers to functions that assist companies in revising or creating new ad copy based on past data.

[0098] A "crisis risk score" refers to an index that quantifies the likelihood of a particular advertisement or piece of information causing a social media firestorm or backlash on a website.

[0099] "Effective word score" refers to a metric that quantifies the likelihood that a particular ad copy or piece of information will be highly effective.

[0100] "Means of preventing damage to a company's image in advance" refers to functions that prevent a company's image from being negatively affected by information it makes public.

[0101] The system for implementing this invention includes an AI that utilizes past online controversy data and effective word data to predict the economic impact of advertising copy published on social media and websites. Furthermore, it has functions to support the editing and creation of advertising copy and predict online controversy risk scores and effective word scores.

[0102] System Configuration

[0103] Hardware:

[0104] server

[0105] smartphone

[0106] software:

[0107] Python (registered trademark)

[0108] pandas

[0109] scikit-learn(registered trademark)

[0110] joblib

[0111] Data handling

[0112] The server stores historical online controversy data and effective keywords in a database. This data is saved as a CSV file and read using pandas. The read data is then converted into a numerical vector using TfidfVectorizer.

[0113] Model training and saving

[0114] The server trains a LogisticRegression model based on data converted into numerical vectors using TfidfVectorizer. The trained model is saved using joblib. This lays the foundation for predicting crisis risk scores and effective word scores.

[0115] Evaluation of ad copy

[0116] When a user enters ad copy using their smartphone, the server uses a stored model to predict the ad's risk of backlash and its effectiveness score. This allows users to understand the risks and effectiveness of their ad copy in advance.

[0117] Specific example

[0118] For example, if a user enters the following ad text:

[0119] To celebrate the launch of our new product, we're holding a special sale!

[0120] The server predicts a risk of backlash score and an effective word score for this ad copy. The predicted results show a risk of backlash score of 0.2 and an effective word score of 0.8.

[0121] Example of a prompt

[0122] Please predict the risk of backlash and the effectiveness score for the following ad copy.

[0123] Advertisement: "To celebrate the launch of our new product, we're holding a special sale!"

[0124] In this way, companies can reduce the risks involved in advertising and create effective ad copy.

[0125] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0126] Step 1:

[0127] The server reads past online controversy data and effective keywords from a database. This data is stored as a CSV file and read using pandas. The input is a CSV file, and the output is a DataFrame. Specifically, the pandas.read_csv function is used to read the data.

[0128] Step 2:

[0129] The server converts the loaded data into a numerical vector using TfidfVectorizer. The input is a data frame, and the output is a numerical vector. Specifically, the TfidfVectorizer's fit_transform method is used to convert text data into a numerical vector.

[0130] Step 3:

[0131] The server trains a LogisticRegression model based on data that has been converted into numerical vectors. The input is a numerical vector, and the output is the trained model. Specifically, the model is trained using the LogisticRegression's fit method.

[0132] Step 4:

[0133] The server saves the trained model and TfidfVectorizer using joblib. The input is the trained model and TfidfVectorizer, and the output is the saved model file. Specifically, the model and vectorizer are saved using the joblib.dump function.

[0134] Step 5:

[0135] The user enters the ad text using their smartphone. The input is the text of the ad, and the output is a request to the server. Specifically, the user enters the ad text through a smartphone application and sends it to the server.

[0136] Step 6:

[0137] The server converts the received ad copy into a numerical vector using TfidfVectorizer. The input is the text of the ad copy, and the output is a numerical vector. Specifically, the server uses the transform method of the saved TfidfVectorizer to convert the ad copy into a numerical vector.

[0138] Step 7:

[0139] The server predicts the risk of backlash and the effectiveness of words based on the ad copy converted into numerical vectors. The input is a numerical vector, and the output is the risk of backlash and the effectiveness of words. Specifically, the scores are predicted using the predict_proba method of the stored LogisticRegression model.

[0140] Step 8:

[0141] The server returns a predicted crisis risk score and an effective word score to the user. The input is the predicted score, and the output is the response to the user. Specifically, the score is displayed through a smartphone application.

[0142] (Example 2)

[0143] Next, we will describe Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0144] In today's information society, when companies and individuals publish information through social media and websites, they are required to use past online controversy data to predict risks and take appropriate action. However, conventional systems lack the means to effectively utilize past controversy data and concretely estimate the amount of loss, resulting in insufficient risk management. Furthermore, there is a lack of concrete means to prevent damage to a company's image in advance.

[0145] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0146] This invention provides a server that includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence that predicts the economic effects of information disclosure based on this data, means for analyzing past online controversy data and estimating the amount of loss, means for storing the estimated amount of loss in a database, and means for providing risk information, including the estimated amount of loss, when a user discloses information. This makes it possible to estimate specific amounts of loss based on past online controversy data, quantify the risk, and take appropriate action when disclosing information. It also provides specific means to prevent damage to a company's image in advance.

[0147] "Past online controversy data" refers to data on past online controversies that have occurred on online platforms such as social media and websites.

[0148] "Effective word data" refers to data on keywords and phrases that are effective when used when publishing information on social media or websites.

[0149] "Economic impact of information disclosure" refers to the expected economic effects and benefits that can be anticipated when information is made public on social media or a website.

[0150] Artificial intelligence is a system that uses machine learning and data analysis techniques to learn patterns from past data and make predictions and decisions about the future.

[0151] "Loss amount" refers to the estimated amount of economic loss based on past cases of online controversies.

[0152] A "database" is a system for efficiently storing, managing, and retrieving data.

[0153] "Risk information" refers to information about risks that should be considered when disclosing information, and specifically includes estimated loss amounts.

[0154] "Downturn in corporate image" refers to a decline in a company's reputation or brand value.

[0155] This invention is a system that utilizes past online controversy data to quantify the risks associated with information disclosure and enable appropriate responses. The following describes a specific embodiment of this system.

[0156] Server Role

[0157] The server retrieves past online fire incident data from a database, analyzes it, and estimates the amount of loss. Specifically, the server uses the following hardware and software:

[0158] Hardware: Server machine with a high-performance processor and sufficient memory.

[0159] Software: Relational database management systems (RDBMS) such as MySQL® and PostgreSQL®, and data analysis libraries such as pandas and scikit-learn® for Python®.

[0160] The server first executes an SQL query to retrieve past online fire incident data from the database. Next, it converts the retrieved data into a data frame using the Python® pandas library and performs preprocessing. Using the preprocessed data, it estimates the amount of loss using the scikit-learn® LinearRegression model. The estimated amount of loss is then saved back to the database.

[0161] Terminal role

[0162] The terminal retrieves risk information from the server when the user publishes information and displays it to the user. Specifically, the terminal uses the following hardware and software.

[0163] Hardware: Personal computers, tablets, smartphones, etc.

[0164] Software: Web browser, HTML, CSS, JavaScript (registered trademark)

[0165] The device sends an HTTP GET request to the server and receives risk information in JSON format. The received risk information is then displayed on a web page using HTML and JavaScript.

[0166] User roles

[0167] Users will check risk information through their devices and adjust the content of the information they disclose. Specifically, users will check the estimated loss amount displayed on the web page and take action such as changing the disclosed content if the risk is high.

[0168] Specific example

[0169] Specific Example 1: Acquisition and Analysis of Online Crisis Data

[0170] The server retrieves past social media controversy data from a MySQL® database. For example, it uses an SQL query like the following:

[0171] SQL

[0172] SELECT FROM flame_data WHERE date >= '2020-01-01';

[0173] The acquired data is analyzed using the Python pandas library to estimate the amount of loss incurred during a social media firestorm. For example, the following code is used:

[0174] Python (registered trademark)

[0175] import pandas as pd

[0176] from sklearn.linear_model import LinearRegression

[0177] Loading data

[0178] data = pd.read_sql('SELECT FROM flame_data WHERE date >= "2020-01-01"', con=database_connection)

[0179] Estimation of loss amount

[0180] model = LinearRegression()

[0181] model.fit(data[['followers', 'negative_comments']], data['loss_amount'])

[0182] predicted_loss = model.predict(new_data[['followers', 'negative_comments']])

[0183] Example 2: Displaying risk information

[0184] When a user publishes information, the device retrieves risk information from the server and displays it on an HTML page. For example, the following code is used:

[0185] html

[0186] <!DOCTYPE html>

[0187]

[0188]

[0189] <title> Risk Information< / title>

[0190]

[0191]

[0192] <h1> Risk Information< / h1>

[0193] Estimated loss: \

[0194] <script>

[0195] fetch('https: / / example.com / api / risk_info')

[0196] .then(response => response.json())

[0197] .then(data => {

[0198] document.getElementById('loss_amount').textContent = data.loss_amount;

[0199] });

[0200] < / script>

[0201]

[0202]

[0203] Example of a prompt

[0204] Examples of prompts to input into a generative AI model include the following:

[0205] "Please generate Python® code that estimates the amount of financial loss incurred during a social media firestorm, using past social media firestorm data."

[0206] By using this prompt statement, the generated AI model can produce appropriate Python® code.

[0207] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0208] Step 1:

[0209] The server retrieves past online controversy data from the database. It accepts database connection information and an SQL query as input, and generates a dataframe containing the controversy data as output. Specifically, the server executes the SQL query to extract past controversy data. For example, it might use the following SQL query:

[0210] SQL

[0211] SELECT FROM flame_data WHERE date >= '2020-01-01';

[0212] This query retrieves online controversy data for the specified period.

[0213] Step 2:

[0214] The server analyzes the acquired data on online controversies. It accepts a dataframe as input and generates a pre-processed dataframe as output. Specifically, the server imputes missing values ​​and removes unnecessary columns. For example, it preprocesses the data using the Python® pandas library.

[0215] Step 3:

[0216] The server estimates the loss amount using preprocessed data. It takes a preprocessed dataframe as input and generates an estimated loss amount as output. Specifically, the server uses the scikit-learn® LinearRegression model to estimate the loss amount. For example, the model is trained and the loss amount is predicted as follows:

[0217] Python (registered trademark)

[0218] model = LinearRegression()

[0219] model.fit(data[['followers', 'negative_comments']], data['loss_amount'])

[0220] predicted_loss = model.predict(new_data[['followers', 'negative_comments']])

[0221] This process estimates the amount of loss.

[0222] Step 4:

[0223] The server stores the estimated loss amount in a database. It takes the estimated loss amount and database connection information as input and generates the loss amount stored in the database as output. Specifically, the server executes an SQL query to insert the estimation results into the database. For example, it might use the following SQL query:

[0224] SQL

[0225] INSERT INTO risk_info (date, predicted_loss) VALUES ('2023-10-01', 500000);

[0226] This query will save the estimated loss amount to the database.

[0227] Step 5:

[0228] The device retrieves risk information from the server when the user publishes information. It receives an HTTP GET request as input and generates JSON data containing the risk information as output. Specifically, the device sends an HTTP GET request to the server and receives the risk information. For example, it uses JavaScript code like the following:

[0229] JavaScript

[0230] fetch('https: / / example.com / api / risk_info')

[0231] .then(response => response.json())

[0232] .then(data => {

[0233] document.getElementById('loss_amount').textContent = data.predicted_loss;

[0234] });

[0235] This process retrieves risk information.

[0236] Step 6:

[0237] The device displays the acquired risk information to the user. It receives JSON data containing risk information as input and generates an HTML page displaying the risk information as output. Specifically, the device uses HTML and JavaScript to display the risk information on the web page. For example, it can be implemented as follows:

[0238] html

[0239] <!DOCTYPE html>

[0240]

[0241]

[0242] <title> Risk Information< / title>

[0243]

[0244]

[0245] <h1> Risk Information< / h1>

[0246] Estimated loss: \

[0247] <script>

[0248] fetch('https: / / example.com / api / risk_info')

[0249] .then(response => response.json())

[0250] .then(data => {

[0251] document.getElementById('loss_amount').textContent = data.predicted_loss;

[0252] });

[0253] < / script>

[0254]

[0255]

[0256] This process displays risk information to the user.

[0257] Step 7:

[0258] Users check risk information through their devices. They receive the displayed risk information as input and adjust the content of the information disclosure as output. Specifically, users check the estimated loss amount displayed on the webpage and take action, such as changing the disclosed content, if the risk is high.

[0259] (Application Example 2)

[0260] Next, we will describe application example 2 of form example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0261] In advertising campaigns, there is a need for a system that can quantify the risk of online backlash by utilizing past data and predict the amount of loss. However, current systems have the challenge of not being able to specifically quantify the risk of online backlash and provide information disclosure and concrete advice for risk reduction based on that risk.

[0262] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0263] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including a generative AI model that predicts the economic effect when published on social media or websites based on this data, means for inputting the content of an advertising campaign and quantifying the risk of online controversy based on past controversy data, means for estimating the amount of loss based on the risk of online controversy, and means for providing specific advice to reduce the risk. This makes it possible to quantify the risk of online controversy in an advertising campaign and provide information disclosure and specific advice to reduce the risk based on that risk.

[0264] "Past online controversy data" refers to data on negative reactions, comments, and criticisms that have occurred on social media or websites in the past.

[0265] "Effective word data" refers to data on keywords and phrases that are considered effective in eliciting positive responses on social media and websites.

[0266] A "generative AI model" refers to an algorithm or system that uses machine learning or artificial intelligence technology to generate a specific output from input data.

[0267] "Methods for quantifying the risk of online backlash" refer to methods or systems that express the likelihood of a particular advertising campaign going viral in a numerical way, based on past online backlash data.

[0268] "Means for estimating the amount of loss" refers to methods or systems for calculating predicted economic losses based on the risk of online backlash.

[0269] "Means of providing specific advice" refers to methods and systems that propose specific actions and measures to reduce the risk of online backlash.

[0270] The system for implementing this invention operates through the coordinated efforts of a server, a terminal, and a user. A specific embodiment is described below.

[0271] Server Role

[0272] The server stores historical data on online controversies and effective keywords, and uses a generative AI model based on this data to predict economic impact. The server uses the following software and hardware:

[0273] Software: Python (registered trademark), Pandas, Scikit-learn (registered trademark)

[0274] Hardware: Servers equipped with high-performance processors and large amounts of memory.

[0275] The server reads past online controversy data and converts it into a data frame. Next, it sets features (campaign length, audience size, number of negative comments) and a target (amount of loss) and trains a linear regression model. This allows the system to quantify the risk of online controversy and predict the amount of loss when given the content of an advertising campaign.

[0276] Role of the Terminal

[0277] The terminal provides an interface for the user to input the content of an advertising campaign. The terminal uses the following software and hardware.

[0278] Software: Web browser, mobile application

[0279] Hardware: Smartphone, tablet, personal computer

[0280] When the user inputs the details of an advertising campaign through the terminal, the data is sent to the server. Based on the received data, the server quantifies the risk of a blaze-up and predicts the amount of loss. The prediction result is sent back to the terminal and displayed to the user.

[0281] Role of the User

[0282] The user uses the terminal to input the content of an advertising campaign and checks the prediction results of the risk of a blaze-up and the amount of loss. Furthermore, the user receives specific advice for risk reduction provided by the server and takes appropriate measures.

[0283] Specific Example

[0284] For example, when the user inputs that the length of an advertising campaign is 30 days, the size of the audience is 100,000 people, and the number of negative comments is 50, the server quantifies the risk of a blaze-up based on this data and calculates the predicted amount of loss. As a result, the predicted amount of loss is displayed, and specific advice for risk reduction is provided.

[0285] Example of a Prompt Sentence

[0286] "When the length of an advertising campaign is 30 days, the size of the audience is 100,000 people, and the number of negative comments is 50, please tell me the predicted amount of loss."

[0287] In this way, it becomes possible to specifically quantify the risk of a social media firestorm in an advertising campaign and provide specific advice for information disclosure and risk reduction based on that risk.

[0288] The flow of the specific process in Application Example 2 will be described using FIG. 14.

[0289] Step 1:

[0290] The user inputs the details of the advertising campaign using the terminal.

[0291] Input: Length of the advertising campaign, size of the audience, number of negative comments

[0292] Output: The input data is sent to the server.

[0293] Specific operation: The user inputs the detailed information of the advertising campaign through the interface of a smartphone or a personal computer and clicks the "Send" button.

[0294] Step 2:

[0295] Based on the data received by the server, the server reads the past social media firestorm data.

[0296] Input: Detailed data of the advertising campaign sent by the user

[0297] Output: The past social media firestorm data is converted into a data frame.

[0298] Specific operation: The server uses the Pandas library of Python (registered trademark) to read the past social media firestorm data from a CSV file and convert it into a data frame.

[0299] Step 3:

[0300] The server sets features and a target and trains a linear regression model.

[0301] Input: Past arson data converted into a data frame

[0302] Output: Trained linear regression model

[0303] Specific operation: The server uses the Scikit-learn (registered trademark) library to set features (campaign length, audience size, number of negative comments) and target (loss amount), and trains a linear regression model.

[0304] Step 4:

[0305] Based on the detailed data of the advertising campaign received by the server from the user, the arson risk is quantified and the loss amount is predicted.

[0306] Input: Detailed data of the advertising campaign received from the user, trained linear regression model

[0307] Output: Numerical value of arson risk and predicted loss amount

[0308] Specific operation: The server inputs the detailed data of the received advertising campaign as features, and uses the trained linear regression model to quantify the arson risk and predict the loss amount.

[0309] Step 5:

[0310] [[ID=�8]] The server returns the prediction results of the arson risk and the loss amount to the terminal.

[0311] Input: Numerical value of arson risk and predicted loss amount

[0312] Output: The prediction results are displayed on the terminal

[0313] Specific operation: The server sends the prediction results to the terminal in JSON format, and the terminal receives this and displays it to the user.

[0314] Step 6:

[0315] The server generates specific advice to mitigate risk and sends it to the terminal.

[0316] Input: Crisis risk and estimated loss amount

[0317] Output: Specific advice for risk reduction

[0318] Specific operation: The server uses a generated AI model to generate specific advice to reduce the risk of online firestorms and sends it to the terminal. The terminal receives this and displays it to the user.

[0319] In this way, it becomes possible to quantify the risk of online backlash in advertising campaigns and provide specific advice on information disclosure and risk reduction based on that risk.

[0320] (Example 3)

[0321] Next, we will describe Embodiment 3 of Embodiment Example 3. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0322] In today's information society, companies face an increasing risk of online backlash when publishing information through social media and websites. This risk can damage a company's image and lead to financial losses, making it crucial to assess the risk beforehand and modify information as needed. However, current systems fail to adequately assess this risk and modify information, posing a significant challenge for businesses.

[0323] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 3 is realized by the following means.

[0324] This invention includes a server that utilizes past online controversy data, a server that utilizes effective word data, a server that uses a generative AI model to generate information based on this data, a server that evaluates the risk of online controversy before the generated information is made public, a server that prompts correction of the information if there is a risk of online controversy, and a server that predicts the economic effects of publishing the information on social media or a website based on these means. This enables companies to evaluate the risk of online controversy before disclosing information and correct the information as needed, thereby preventing damage to their corporate image.

[0325] "Past online controversy data" refers to data on past online controversies caused by information published on social media, websites, etc.

[0326] "Effective word data" refers to data on words and phrases that are considered effective in eliciting positive responses when publishing information on social media or websites.

[0327] A "generative AI model" is an artificial intelligence model that generates information based on prompt text entered by the user.

[0328] "Methods for assessing the risk of online backlash" refer to methods for analyzing and evaluating the likelihood of online backlash when generated information is made public.

[0329] "Methods to encourage information correction" refer to methods used to prompt users to correct information that has been assessed as having a risk of causing a public outcry.

[0330] "Methods for predicting economic effects" refer to methods for predicting the economic impact that information published on social media and websites has on companies.

[0331] Modes for carrying out the invention

[0332] This invention is a system that prevents damage to a company's image by pre-assessing the risk of online backlash when a company publishes information on social media or its website, and by modifying the information as needed. This system consists of three main elements: a server, a terminal, and a user.

[0333] 1. Program generation

[0334] The server generates a program to proactively prevent damage to the company's image. This program uses a generative AI model to create or edit information and has the function of evaluating the risk of online backlash before the information is made public. Specifically, it uses OpenAI's GPT-4® as the generative AI model.

[0335] 2. Program Processing

[0336] The server uses a generative AI model (e.g., OpenAI's GPT-4®) to generate information based on the prompt text entered by the user. The generated information is then sent to a crisis risk assessment module. This module uses natural language processing (NLP) techniques to analyze the content of the information and assess whether there is a crisis risk. Specifically, it uses a risk assessment algorithm (e.g., a BERT-based classification model) to determine whether the information contains a crisis risk.

[0337] The terminal sends the prompt text entered by the user to the server and receives generated information and risk assessment results from the server. The user reviews the risk assessment results on the terminal and corrects the information as needed.

[0338] 3. Specific Examples and Examples of Prompt Statements

[0339] As a concrete example, consider a case where a user is creating a press release for a new product. The user enters the following prompt into the terminal:

[0340] Example of a prompt:

[0341] "Please create a press release for our new product. The product name is 'EcoSmart,' and please emphasize its environmentally friendly features."

[0342] The server inputs this prompt into the AI ​​model that generates the following information:

[0343] Examples of generated information:

[0344] "EcoSmart is a new product that utilizes the latest environmental technologies. It is energy-efficient and uses recyclable materials."

[0345] Next, the server sends the generated information to a crisis risk assessment module for risk assessment. If the assessment result is determined to be "risky," the server sends a message to the user prompting them to make corrections. The user can then review the proposed corrections on their device and modify the information as needed to prevent damage to the company's image.

[0346] This system allows companies to assess the risk of public backlash before disclosing information and, if necessary, modify the information to prevent damage to their corporate image. The flow of the specific processing in Example 3 will be explained using Figure 15.

[0347] Step 1:

[0348] The user enters a prompt message.

[0349] The user enters a prompt message into the terminal's input field. For example, they might enter, "Please create a press release for a new product. The product name is 'EcoSmart,' and please emphasize its environmentally friendly features." The entered prompt message is temporarily stored in the terminal's memory.

[0350] Step 2:

[0351] The terminal sends a prompt message to the server.

[0352] The terminal converts the prompt text entered by the user into the appropriate format and sends a request to the server's API endpoint. Specifically, it sends the prompt text to the server using an HTTP POST request. The input is the prompt text, and the output is the request sent to the server.

[0353] Step 3:

[0354] The server generates information using an AI model.

[0355] The server inputs the received prompt message into a generating AI model (e.g., OpenAI's GPT-4®) and generates information. The generating AI model analyzes the prompt message and generates information based on the appropriate context. The input is the prompt message, and the output is the generated information. For example, information such as "EcoSmart is a new product that utilizes the latest environmental technologies. It is energy-efficient and uses recyclable materials." might be generated.

[0356] Step 4:

[0357] The server sends the generated information to the fire risk assessment module.

[0358] The server sends the generated information to a crisis risk assessment module for risk assessment. This module uses natural language processing (NLP) techniques to analyze the content of the information and assess whether there is a risk of crisis. Specifically, it uses a risk assessment algorithm (e.g., a BERT-based classification model) to determine whether the information contains a risk of crisis. The input is the generated information, and the output is the risk assessment result.

[0359] Step 5:

[0360] The server generates the risk assessment results and sends them to the terminal.

[0361] The server receives the assessment results from the fire risk assessment module and sends them to the terminal. The assessment results include judgments such as "no risk" or "risk present." The input is the risk assessment result, and the output is the transmission to the terminal.

[0362] Step 6:

[0363] The device displays the risk assessment results to the user.

[0364] The terminal displays the risk assessment results received from the server to the user. For example, a message such as "This information has a risk of causing a public outcry. We recommend correcting it." might be displayed. The input is the risk assessment results, and the output is what is displayed to the user.

[0365] Step 7:

[0366] Users can modify the information as needed.

[0367] The user reviews the risk assessment results displayed on their device and corrects the information as needed. The corrected information is sent back to the server and, after going through the same process, is finally published as safe information. The input is the corrected information, and the output is the re-evaluated information.

[0368] (Application Example 3)

[0369] Next, we will describe application example 3 of form example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0370] When companies publish information on social media or their websites, they are required to assess the risk of online backlash beforehand and prevent damage to their corporate image. However, traditional methods have not been able to fully utilize past backlash data or effective keyword data, and predictions of economic effects have been insufficient. Furthermore, they lacked the function of providing concrete corrective measures, making it difficult for companies to respond quickly and appropriately. As a result, companies often took on unnecessary risks.

[0371] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means.

[0372] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic effects of disclosing information based on this data, means for evaluating the risk of information becoming controversial using a generative AI model, and means for presenting specific corrective measures if there is a risk of controversy. This enables companies to evaluate the risk of controversy before disclosing information and obtain specific corrective measures as needed.

[0373] "Past online controversy data" refers to information about past online controversies that have occurred on social media, websites, etc. Specifically, it includes data such as the causes and impact of the controversies, and related keywords.

[0374] "Effective word data" refers to data on words and phrases used on social media, websites, and other platforms that are considered to be particularly effective in eliciting positive responses.

[0375] "Artificial intelligence that predicts the economic effects of disclosing information" refers to artificial intelligence technology used to predict the impact that disclosed information will have on a company's economic activities, and it makes predictions based on past data and trends.

[0376] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate text, and has the ability to generate sentences based on specific prompts.

[0377] "Methods for assessing the risk of online backlash" refer to methods for evaluating the likelihood of information being made public causing online backlash, and these methods use past online backlash data and generative AI models to determine the risk.

[0378] "Means of presenting specific revision proposals" refers to methods for presenting specific revision proposals to mitigate the risk of information being deemed to have a high risk of causing a public outcry.

[0379] The system for implementing this invention operates through the coordinated efforts of a server, a terminal, and a user. A specific embodiment is described below.

[0380] System Configuration

[0381] The server includes means of utilizing past online controversy data, means of utilizing effective word data, means of including artificial intelligence to predict the economic effects of disclosing information, means of using generative AI models to assess the risk of information going viral, and means of suggesting specific corrective measures if there is a risk of going viral.

[0382] A terminal is a device used by users to input information and communicate with a server. Specifically, this includes smartphones and personal computers.

[0383] Users are typically corporate public relations or marketing personnel who input information and receive feedback from the server.

[0384] Program processing

[0385] The server first receives information entered by the user. This information includes advertisements and campaign details that are planned to be published on social media and websites.

[0386] Next, the server assesses the information's risk of causing a controversy based on past controversy data and effective word data. Specifically, it uses a generative AI model to determine whether the input information contains a risk of causing a controversy.

[0387] If the evaluation determines that there is a risk of a social media firestorm, the server will generate specific revision proposals. These revision proposals use a generation AI model to suggest specific wording and phrases to mitigate the risk.

[0388] Finally, the server sends the evaluation results and suggested corrections to the user's terminal. The user receives this and corrects the information as needed.

[0389] Hardware and software to be used

[0390] Hardware: Servers, smartphones, personal computers

[0391] Software: OpenAI API, Python (registered trademark)

[0392] Specific example

[0393] For example, suppose a user enters the following ad text.

[0394] This product is superior to other products on the market. Buy it now!

[0395] The server receives this ad text and generates the following prompt.

[0396] Please evaluate the following ad copy and determine if it poses a risk of causing a social media firestorm.

[0397] Advertisement: This product is superior to other products. Buy now!

[0398] If there is a risk of backlash, please provide specific proposed solutions.

[0399] The generative AI model evaluates this prompt and proposes the following revisions.

[0400] This advertisement contains disparaging language towards competitors' products and carries the risk of backlash. We recommend revising it as follows:

[0401] Revised version: "This product has satisfied many customers. Buy it now and experience the difference!"

[0402] In this way, users can receive feedback from the server and correct the information, thereby reducing the risk of online backlash and preventing damage to the company's image.

[0403] The flow of the specific processing in Application Example 3 will be explained using Figure 16.

[0404] Step 1:

[0405] Users use their devices to input information such as ad copy and campaign details.

[0406] Input: Ad copy and campaign details

[0407] Output: The entered information is sent to the server.

[0408] Specific operation: The user enters the advertisement text into an input form on their smartphone or computer and presses the submit button. The device sends this information to the server.

[0409] Step 2:

[0410] The server compares the information it receives with past online controversy data and effective keyword data.

[0411] Input: User-submitted information, past online controversy data, effective keyword data

[0412] Output: Matching results (data necessary for evaluating the risk of online backlash)

[0413] Specific operation: The server retrieves past online controversy data and effective keyword data from the database and compares it with the information received from the user.

[0414] Step 3:

[0415] The server uses a generated AI model to assess the risk of information going viral.

[0416] Input: Matching results, generated AI model

[0417] Output: Evaluation results of the risk of online backlash

[0418] Specific operation: The server inputs the matching results into the generating AI model, generates a prompt message, and performs an evaluation. The generating AI model determines the risk of a firestorm and returns the result to the server.

[0419] Step 4:

[0420] If a server is at risk of crashing, it will generate specific corrective action plans.

[0421] Input: Results of the assessment of the risk of online backlash

[0422] Output: Revision proposal

[0423] Specific operation: The server uses the generative AI model again to generate specific corrective measures to reduce the risk of a crisis. The generative AI model generates corrective measures based on the prompt and returns the results to the server.

[0424] Step 5:

[0425] The server sends the evaluation results and suggested corrections to the user's terminal.

[0426] Input: Evaluation results, proposed revisions

[0427] Output: Evaluation results and proposed revisions sent to the user's device.

[0428] Specific operation: The server compiles the evaluation results and suggested revisions and sends them to the user's device. The user reviews this information on their device and modifies the ad copy as needed.

[0429] Step 6:

[0430] The user corrects the information based on the proposed revisions and resends it to the server.

[0431] Input: Information corrected based on the proposed amendments

[0432] Output: The corrected information is sent to the server.

[0433] Specific operation: The user modifies the ad copy based on the suggested revisions received from the server and resubmits it to the server. The server receives the revised information and performs a re-evaluation.

[0434] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0435] "Example of form 1"

[0436] One example of an invention incorporating an emotion engine is a system that combines an emotion engine that recognizes user emotions with an AI that predicts the economic impact of publishing content on social media and websites based on past online controversy data and effective word data. Specifically, it analyzes emotions from text data posted by users and uses the results to edit or create published information. For example, if a user posts text expressing joy, that information can be used to publish information that reinforces a positive image of the company's products or services.

[0437] "Example of form 2"

[0438] As an example of an invention combining an emotion engine, the system described in claim 2 incorporates an emotion engine that recognizes the user's emotions as a means for determining the amount of loss incurred when past online controversy data is used during editing. Specifically, the system analyzes the user's emotions and evaluates the risk of online controversy based on the results. For example, if a user posts text indicating anger, that information is used to avoid disclosing information that would increase the risk of online controversy.

[0439] "Example of form 3"

[0440] As an example of a third form of the invention incorporating an emotion engine, the system described in claim 3 incorporates an emotion engine that recognizes user emotions as a means to prevent damage to the company's image in advance. Specifically, it analyzes the user's emotions and, based on the results, discloses information to prevent damage to the company's image. For example, if a user posts text indicating sadness, that information is used to disclose information about the company's response and service improvements.

[0441] The following describes the processing flow for each example of the form.

[0442] "Example of form 1"

[0443] Step 1: The user posts text data to social media or a website.

[0444] Step 2: The emotion engine analyzes the user's emotions from the text data.

[0445] Step 3: The AI ​​predicts the economic impact of the posted text data based on past online controversy data and effective word data.

[0446] Step 4: Based on the analysis results of the emotion engine and the AI's prediction results, edit or create publicly available information.

[0447] "Example of form 2"

[0448] Step 1: The user posts text data to social media or a website.

[0449] Step 2: The emotion engine analyzes the user's emotions from the text data.

[0450] Step 3: The AI ​​uses past online controversy data to determine the potential loss of the posted text data.

[0451] Step 4: Based on the analysis results of the emotion engine and the AI's loss estimation results, assess the risk of online backlash.

[0452] "Example of form 3"

[0453] Step 1: The user posts text data to social media or a website.

[0454] Step 2: The emotion engine analyzes the user's emotions from the text data.

[0455] Step 3: The AI ​​predicts the impact of the posted text data on the company's image, based on past data on online controversies and effective keywords.

[0456] Step 4: Based on the analysis results of the emotion engine and the AI's prediction results, disclose information to prevent damage to the company's image.

[0457] (Example 1)

[0458] Next, we will describe Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0459] When companies publish information on social media and websites, they are required to predict economic impact by utilizing data on past online controversies and effective keywords. However, there is a lack of systems to effectively utilize this data, making it difficult to prevent damage to corporate image and economic losses. Furthermore, there is a lack of means to analyze user sentiment and revise or create information based on that analysis, resulting in ineffective information disclosure strategies for companies.

[0460] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0461] This invention includes a server that collects and stores past online controversy data, collects and stores effective word data, uses artificial intelligence to predict the economic effects of publishing information on social media and websites based on this data, analyzes emotions from text data posted by users, and edits or creates public information based on the results of the emotion analysis. This enables companies to predict economic effects by utilizing past online controversy cases and effective word data, and to publish information based on user emotions.

[0462] "Past online controversy data" refers to detailed information about past online controversies that have occurred on social media and the internet.

[0463] "Effective word data" refers to information about words and phrases that are considered effective in marketing and advertising.

[0464] "SNS" is an abbreviation for Social Networking Service, which refers to an online platform for users to share information and interact with each other.

[0465] A "website" is a collection of information published on the internet, and refers to an online page used by companies and individuals to disseminate information.

[0466] "Economic effect" refers to the impact that a particular action or event has on the economy, and specifically includes increases or decreases in sales and profits.

[0467] "Artificial intelligence" refers to technologies that use machine learning and data analysis to mimic human intelligence.

[0468] "User" refers to an individual or group that uses a system or service.

[0469] "Text data" refers to digital data that includes character information.

[0470] "Emotion analysis" refers to a technology that identifies emotions from text data and evaluates the type and intensity of those emotions.

[0471] "Public information" refers to information that is made publicly available through social media and websites.

[0472] "Editing" refers to the act of correcting existing text and revising it to make it more appropriate.

[0473] "Creation" refers to the act of generating new text or information.

[0474] Modes for carrying out the invention

[0475] This invention is an AI-powered information disclosure support system for businesses that utilizes past data on online controversies and effective keywords to predict the economic impact of information published on social media and websites. It also includes a function to analyze user sentiment and revise or create published information based on the results.

[0476] Data collection and storage

[0477] The server collects past incidents of online controversies that have occurred on social media and the internet. Specifically, it uses web scraping tools (e.g., Beautiful Soup) to obtain data. The collected data is stored in a database (e.g., MySQL®, PostgreSQL®). Similarly, data on effective keywords is also collected and stored in the same database.

[0478] Data analysis

[0479] The server analyzes accumulated online controversy data using AI models (e.g., TensorFlow®, PyTorch®). Specifically, it uses natural language processing (NLP) techniques to extract the causes and patterns of online controversies. Similarly, data on effective words is also analyzed by AI models to identify which words are effective in which situations. These analysis results are stored in a database.

[0480] Predicting the economic impact

[0481] The server predicts the economic impact of information published on social media and websites, based on past data on online controversies and effective keywords. Specifically, it uses machine learning models to simulate the level of response the published information will generate. The prediction results are stored in a database.

[0482] Combination of emotional engines

[0483] The terminal sends text data posted by the user to an emotion engine (e.g., IBM Watson®, Microsoft® Azure® emotion analysis API). The server receives the analysis results returned by the emotion engine and stores them in a database. Based on the emotion analysis results, the server edits or creates public information. Specifically, it generates information that reinforces the positive image of a company's products and services based on text that shows positive emotions. The generated information is published on social media and websites.

[0484] Specific example

[0485] For example, if a user posts the text "I really love the new product!", the device sends this text to the emotion engine. The server receives the emotion analysis result for "joy" from the emotion engine, and based on this analysis, generates positive information such as "many users like the new product" and publishes it on social media.

[0486] Example of a prompt

[0487] Examples of prompts to input into a generative AI model include the following:

[0488] "Based on past data on online controversies and effective keywords, predict the economic impact of information published on social media and websites. Also, analyze user sentiment and revise the published information based on the results."

[0489] The above describes the modes for carrying out the invention.

[0490] The flow of the specific processing in Example 1 will be explained using Figure 17.

[0491] Step 1:

[0492] The server collects past incidents of online controversies that have occurred on social media and the internet. Specifically, it uses web scraping tools (e.g., Beautiful Soup) to obtain data. The input is text data from social media and websites, and the output is the collected data on these controversies. This data is stored in a database (e.g., MySQL®, PostgreSQL®).

[0493] Step 2:

[0494] The server collects data on words considered effective in marketing and advertising. This utilizes existing marketing databases and expert knowledge. The input is word data from the marketing database, and the output is the collected data on effective words. This data is also stored in the same database.

[0495] Step 3:

[0496] The server analyzes accumulated online controversy data using AI models (e.g., TensorFlow®, PyTorch®). Specifically, it uses natural language processing (NLP) techniques to extract the causes and patterns of online controversies. The input is online controversy data stored in a database, and the output is the analysis results of the causes and patterns of online controversies. These analysis results are stored in the database.

[0497] Step 4:

[0498] The server analyzes effective word data using an AI model to identify which words are effective in which situations. The input is effective word data stored in the database, and the output is the analysis results of effective words. These analysis results are also stored in the database.

[0499] Step 5:

[0500] The server predicts the economic impact of information published on social media and websites, based on past online controversy data and effective keyword data. Specifically, it uses a machine learning model to simulate the level of impact the published information will generate. The input is the analysis results of online controversy data and effective keyword data, and the output is the predicted economic impact. The prediction results are stored in a database.

[0501] Step 6:

[0502] The terminal sends the text data posted by the user to an emotion engine (e.g., IBM Watson, Microsoft Azure's emotion analysis API). The input is the text data posted by the user, and the output is the emotion analysis result from the emotion engine.

[0503] Step 7:

[0504] The server receives the analysis results returned from the emotion engine and stores them in the database. The input is the emotion analysis result, and the output is the emotion analysis result stored in the database.

[0505] Step 8:

[0506] The server edits and creates publicly available information based on the results of sentiment analysis. Specifically, it generates information that reinforces a positive image of a company's products and services based on text that expresses positive emotions. The input is the sentiment analysis results, and the output is the generated publicly available information. The generated information is published on social media and websites.

[0507] Step 9:

[0508] For example, if a user posts the text "I really love the new product!", the device sends this text to the emotion engine. The server receives the emotion analysis result for "joy" from the emotion engine, and based on this analysis, generates positive information such as "many users like the new product" and publishes it on social media.

[0509] (Application Example 1)

[0510] Next, we will describe Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server," and the smart device 14 will be referred to as a "terminal."

[0511] When companies publish information on social media and their websites, it is important to predict economic impact by utilizing past data on online controversies and effective keywords. However, conventional systems lack the ability to analyze user emotions and generate optimal ad copy based on that analysis, making it difficult to improve corporate image and maximize economic impact.

[0512] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means. In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including an AI that predicts the economic effect when published on social media or websites based on this data, means for analyzing user emotions, means for generating ad copy based on the emotion analysis results, and means for predicting the economic effect of the generated ad copy. This enables companies to generate optimal ad copy based on user emotions and maximize economic effect.

[0513] "Past online controversy data" refers to data on negative reactions, critical comments, and posts that have occurred on social media or websites in the past.

[0514] "Effective word data" refers to data on keywords and phrases that have generated positive responses or high economic impact on social media and websites.

[0515] "AI that predicts economic impact" is artificial intelligence that uses past data to predict the extent of economic impact that information published on social media and websites will have.

[0516] "Methods for analyzing user emotions" refers to technologies that analyze the emotions expressed in text data posted by users and identify emotional states such as positive, negative, and neutral.

[0517] "Methods for generating ad copy based on sentiment analysis results" refers to technologies that automatically generate optimal ad copy using the results of user sentiment analysis.

[0518] "Methods for predicting the economic impact of generated ad copy" refers to technologies that predict the extent of economic impact that generated ad copy will have.

[0519] The following system configuration will be described as an embodiment for carrying out this invention.

[0520] The server includes means of utilizing past online controversy data, means of utilizing effective word data, means including AI to predict the economic impact when published on social media and websites based on this data, means of analyzing user sentiment, means of generating ad copy based on sentiment analysis results, and means of predicting the economic impact of the generated ad copy.

[0521] Hardware and software to be used

[0522] Hardware: Servers, smartphones

[0523] Software: Python (registered trademark), pandas, scikit-learn (registered trademark), TextBlob, OpenAI API

[0524] Data processing and data calculation

[0525] 1. Data Loading: The server loads past online controversy data and effective keywords from a CSV file.

[0526] 2. Vectorization: The server uses TfidfVectorizer to vectorize the text data.

[0527] 3. Model Training: The server trains the economic impact prediction model using LinearRegression.

[0528] 4. Sentiment Analysis: The server uses TextBlob to analyze the sentiment of the user's posted text.

[0529] 5. Ad copy generation: The server uses the OpenAI API to generate sentiment-based ad copy.

[0530] 6. Economic Impact Prediction: The server uses a trained model to predict the economic impact of the generated ad copy.

[0531] Specific example

[0532] When a user posts the text "I want to create a new smartphone ad" using their smartphone, the server performs the following actions:

[0533] 1. The server uses TextBlob to perform sentiment analysis on the user's posted text and detects positive sentiment.

[0534] 2. The server uses the OpenAI API to generate the following prompt text and create the ad text.

[0535] Example prompt: "Create a positive advertisement text for the following product: I want to create an advertisement for the new smartphone."

[0536] 3. The server predicts the economic impact based on the generated ad copy.

[0537] In this way, companies can generate optimal ad copy based on user emotions and maximize economic impact.

[0538] The flow of a specific process in Application Example 1 will be explained using Figure 18.

[0539] Step 1:

[0540] The server reads historical online controversy data and effective keywords from a CSV file. It accepts a CSV file containing historical controversy data and effective keywords as input, and outputs this data in memory as a DataFrame. Specifically, it uses the pandas library to read the CSV file and convert it to a DataFrame.

[0541] Step 2:

[0542] The server uses TfidfVectorizer to vectorize text data. It takes a text string of past flame war data as input and obtains a vector representation of the text data as output. Specifically, it initializes TfidfVectorizer and vectorizes the text data using the fit_transform method.

[0543] Step 3:

[0544] The server trains an economic impact prediction model using LinearRegression. It takes vectorized text data and corresponding economic impact data as input and outputs a trained economic impact prediction model. Specifically, it initializes the LinearRegression model and trains it using the fit method.

[0545] Step 4:

[0546] The server uses TextBlob to analyze the sentiment of user-submitted text. It receives user-submitted text data as input and outputs a sentiment analysis result (positive, negative, or neutral). Specifically, it creates a TextBlob object and uses its `sentiment` property to analyze the sentiment.

[0547] Step 5:

[0548] The server uses the OpenAI API to generate sentiment-based ad copy. It receives sentiment analysis results and user-submitted text as input, and outputs the generated ad copy. Specifically, it generates a prompt based on the sentiment analysis results and sends a request to the OpenAI API to generate the ad copy.

[0549] Step 6:

[0550] The server predicts the economic impact of generated ad copy using a trained model. It receives generated ad copy as input and outputs a predicted economic impact value. Specifically, it vectorizes the generated ad copy and uses the trained model's predict method to predict the economic impact.

[0551] (Example 2)

[0552] Next, we will describe Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0553] In today's information society, it is crucial for companies and individuals to properly assess the risk of online backlash and prevent economic losses when publishing information on social media and websites. However, conventional systems failed to adequately utilize past backlash data and user sentiment, resulting in insufficient assessment of backlash risk. Furthermore, there was a lack of means to proactively prevent damage to a company's image. As a result, risk management during information publication was difficult, and the likelihood of economic losses and damage to a company's image was high.

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

[0555] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic impact of information disclosure based on this data, means for analyzing user input text in real time and recognizing emotions, and means for evaluating the risk of online controversy based on the emotion analysis results. This makes it possible to quantify the risk of online controversy concretely before information disclosure, thereby preventing economic losses and damage to the company's image.

[0556] "Past online controversy data" refers to information about past online controversies that have occurred on social media or websites, and specifically includes data such as the content of posts, comments, and related financial losses.

[0557] "Effective word data" refers to data on keywords and phrases that are considered to have a positive impact when information is released.

[0558] "Artificial intelligence that predicts the economic impact of information disclosure" refers to machine learning models and algorithms that predict the economic impact of information disclosure based on past data.

[0559] "A means of analyzing user input text in real time and recognizing emotions" refers to technology that instantly analyzes text entered by a user and identifies the emotions contained within that text.

[0560] "Methods for evaluating the risk of online backlash based on sentiment analysis results" refers to technologies that use the results of sentiment analysis to assess the likelihood that input text will cause online backlash.

[0561] "Methods for determining the amount of loss" refers to techniques for identifying and quantifying the economic losses incurred during a social media firestorm, based on past data from similar incidents.

[0562] "Methods to prevent damage to a company's image in advance" refer to techniques for assessing the risk of a public outcry before information is released, and for preventing a decline in a company's reputation and image.

[0563] Modes for carrying out the invention

[0564] This invention is a system that utilizes past online controversy data to analyze user input text in real time, recognize emotions, and assess the risk of online controversy. A specific embodiment of this system is described below.

[0565] Server Processing

[0566] The server first collects data on past online controversies. Using SQL queries, it retrieves data such as posts, comments, and associated losses related to past controversies from the database. Next, it trains a model to predict losses based on the collected data. The data is preprocessed using the Pandas library in Python, and a regression model is built using Scikit-learn. The trained model is saved as a file using the Pickle library for later use.

[0567] Terminal processing

[0568] The device analyzes user-entered text in real time and calls an emotion analysis API to recognize emotions. Specifically, it receives user text input and uses the Google Cloud Natural Language API to analyze the emotion of the text. The analysis results are displayed to the user on the device. For example, if the emotion "anger" is recognized, the user will be notified accordingly.

[0569] User actions

[0570] Users input text into the system, and the system evaluates the risk of that text causing a social media firestorm. For example, users can input content for social media posts or blog articles. The system evaluates the risk of the input text causing a firestorm based on past firestorm data and sentiment analysis results. The evaluation results are displayed to the user, and the user can modify the content of their post based on those results.

[0571] Specific example

[0572] Examples of prompts to input into a generative AI model

[0573] "Design a system that uses past online controversy data to assess the risk of specific text going viral. Include features that analyze user-entered text in real time, recognize emotions using an emotion engine, and then assess the risk based on those results."

[0574] This system allows users to quantify and check the risk of controversy before posting, enabling them to publish information while taking risks into account.

[0575] The flow of the specific processing in Example 2 will be explained using Figure 19.

[0576] System program processing flow

[0577] Server Processing

[0578] Step 1: Collecting data on past online controversies.

[0579] The server collects data on past online controversies from the database. Specifically, it uses SQL queries to retrieve data such as post content, comments, and associated losses.

[0580] Input: Database connection information, SQL query

[0581] Output: Past online controversy dataset

[0582] Specific operation: The server executes the SQL query "SELECT FROM EnjoData WHERE Date BETWEEN(registered trademark) '2018-01-01' AND '2023-01-01';" to retrieve the necessary data from the database.

[0583] Step 2: Data preprocessing

[0584] The server preprocesses the collected data. Specifically, it uses the Python® Pandas library to impute missing values ​​and remove outliers.

[0585] Input: Past online controversy dataset

[0586] Output: Preprocessed dataset

[0587] Specific operation: The server uses Pandas to perform operations such as "df.fillna(0)" and "df.dropna()" to clean up the data.

[0588] Step 3: Training the loss prediction model

[0589] The server trains a model to predict the amount of loss based on preprocessed data. Specifically, it builds a regression model using Scikit-learn®.

[0590] Input: Preprocessed dataset

[0591] Output: Trained loss prediction model

[0592] Specific operation: The server executes code such as "from sklearn.linear_model import LinearRegression" and "model = LinearRegression().fit(X_train, y_train)" to train the model.

[0593] Step 4: Save the model

[0594] The server saves the trained model as a file for later use. The Pickle library is used to save the model.

[0595] Input: Trained loss prediction model

[0596] Output: Saved model file

[0597] Specific operation: The server executes the code "import pickle" and "with open('loss_prediction_model.pkl', 'wb') as file: pickle.dump(model, file)" to save the model.

[0598] Terminal processing

[0599] Step 5: Receiving User Input

[0600] The device receives text entered by the user. For example, it retrieves input from a web form or a text box in a mobile app.

[0601] Input: User's text input

[0602] Output: Received text data

[0603] Specific action: The terminal responds when "the user enters 'This product is completely unusable' into a web form."

[0604] Step 6: Call the Sentiment Analysis API

[0605] The device sends the entered text to a sentiment analysis API to recognize its sentiment. The Google Cloud Natural Language API is used to analyze the sentiment of the text.

[0606] Input: Received text data

[0607] Output: Emotion analysis results

[0608] Specific operation: The terminal executes the code "response = client.analyze_sentiment(document=document)" and "sentiment = response.document_sentiment" to analyze the sentiment.

[0609] Step 7: Displaying the analysis results

[0610] The device displays the results of the emotion analysis to the user. For example, if the emotion "anger" is detected, the user will be notified accordingly.

[0611] Input: Sentiment analysis results

[0612] Output: Notification message to the user

[0613] Specific action: The terminal displays the message "This text contains angry emotions" to the user.

[0614] User actions

[0615] Step 8: Enter text

[0616] Users input text into the system, such as social media posts or blog articles.

[0617] Input: Text content

[0618] Output: Send text data to the system

[0619] Specific action: The user enters "This service is completely unusable" and sends it to the system.

[0620] Step 9: Check the risk of a public outcry.

[0621] The user reviews the system's assessment of online crisis risk. The system assesses the risk based on past online crisis data and sentiment analysis results.

[0622] Input: Evaluation results from the system

[0623] Output: Display of evaluation results to the user

[0624] Specific operation: The system displays an evaluation result stating, "This post has a high risk of causing a controversy," and the user confirms this.

[0625] In this way, users can quantify and confirm the risk of controversy before posting, enabling them to publish information while taking risks into account.

[0626] (Application Example 2)

[0627] Next, we will describe application example 2 of form example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0628] In today's information society, companies and individuals are required to use past online controversy data to assess the risk of online backlash and prevent economic losses when publishing information on social media and websites. However, conventional systems lack the means to analyze user sentiment and assess the risk of online backlash, making it difficult to warn users in advance about high-risk posts.

[0629] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0630] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence that predicts the economic effects when published on social media or websites based on this data, means for analyzing user sentiment, means for evaluating the risk of online controversy, and means for providing advance warnings about high-risk posts. This makes it possible to evaluate the risk of online controversy while considering user sentiment and to provide advance warnings about high-risk posts.

[0631] "Past online controversy data" refers to data on online controversies that occurred in the past due to information published on social media, websites, etc.

[0632] "Effective word data" refers to data on words and phrases that are likely to elicit a positive response when publishing information on social media or websites.

[0633] "Artificial intelligence" is a technology in which computer systems imitate human intelligence to learn, reason, and self-correct.

[0634] "Methods for analyzing user emotions" refers to technologies that identify emotions from text and comments posted by users and evaluate the type and intensity of those emotions.

[0635] "Methods for evaluating the risk of online backlash" refer to technologies that quantify and evaluate the likelihood of information disclosure causing a backlash, based on past online backlash data and user sentiment analysis results.

[0636] "A means of warning users in advance about high-risk posts" refers to a technology that issues advance warnings to users about posts that have been assessed as having a high risk of causing a social media firestorm.

[0637] The system for implementing this invention consists of three main elements: a server, a terminal, and a user. The server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic effects when content is published on social media or websites based on this data, means for analyzing user sentiment, means for evaluating the risk of online controversy, and means for warning users in advance about high-risk posts.

[0638] The server stores past online controversy data and effective word data in a database, and uses artificial intelligence to predict economic impact based on this data. Specifically, the server uses a text analysis engine (e.g., TextBlob) to analyze user posts and identify sentiment. Furthermore, it uses a machine learning model (e.g., LinearRegression) to assess the risk of online controversy based on past controversy data and sentiment analysis results.

[0639] When a user posts to social media or a website, the device receives a crisis risk assessment result from the server and displays a warning for high-risk posts. This allows users to reconsider their posts and reduce the risk of a social media firestorm.

[0640] For example, if a user tries to post "This product is completely useless!", the device sends this text to the server. The server uses a text analysis engine to analyze the sentiment and identify anger. Then, a machine learning model is used to assess the risk of the post causing a firestorm, and if a high risk is detected, a warning is displayed on the device.

[0641] Example of a prompt:

[0642] Please enter the user's comment: This product is completely useless!

[0643] This system allows users to check the risk of controversy before posting and avoid high-risk posts, thereby preventing financial losses for companies and individuals.

[0644] The flow of a specific process in Application Example 2 will be explained using Figure 20.

[0645] Step 1:

[0646] The user enters text to post to social media or a website using their device. The entered text is sent from the device to the server. The input data is the user's post content.

[0647] Step 2:

[0648] The server passes the received text to a text analysis engine (e.g., TextBlob) for sentiment analysis. The text analysis engine identifies the sentiment in the text and evaluates the type of sentiment (e.g., anger, joy, sadness) and its intensity. The input data is the user's posted content, and the output data is the type and intensity of the sentiment.

[0649] Step 3:

[0650] The server uses a machine learning model (e.g., LinearRegression) to assess the risk of online backlash based on the sentiment analysis results. The machine learning model takes historical backlash data and sentiment analysis results as input and quantifies the risk of online backlash. The input data consists of the type and intensity of the emotion, and historical backlash data, while the output data is a numerical value of the risk of online backlash.

[0651] Step 4:

[0652] The server sends the results of the crisis risk assessment to the terminal. Based on the received crisis risk value, the terminal displays a warning to the user if the risk is high. The input data is the crisis risk value, and the output data is the warning message.

[0653] Step 5:

[0654] The user reviews the warning message displayed on their device and reconsiders their post. If necessary, they revise the post and resubmit it to the server. The input data is the warning message, and the output data is the revised post.

[0655] This process allows users to check the risk of controversy before posting and avoid high-risk posts, thereby preventing financial losses for companies and individuals.

[0656] (Example 3)

[0657] Next, we will describe Embodiment 3 of Embodiment Example 3. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0658] Companies are required to proactively mitigate the risk of information going viral and damaging their corporate image. However, conventional systems struggle to consistently handle information generation, editing, risk assessment, and user sentiment analysis, often resulting in ineffective countermeasures. Furthermore, there is a lack of means to predict economic impact by utilizing past data on online controversies and effective keywords. As a result, companies face significant risks when disclosing information.

[0659] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 3 is realized by the following means. In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means for using a generative AI model that generates information based on this data, means for editing the generated information, means for evaluating the online controversy risk of the generated or edited information, means for prompting corrections when the online controversy risk is high, means for using an emotion engine that analyzes user emotions, means for publishing information based on the analysis results, and means for predicting the economic effects when the information is published on social media or a website based on this data. This makes it possible to evaluate the online controversy risk of information published by a company in advance and prompt corrections as necessary, thereby preventing damage to the company's image.

[0660] "Past online controversy data" refers to data on past online controversies caused by information published on social media, websites, etc.

[0661] "Effective word data" refers to data on words and phrases that are considered effective in eliciting positive responses when information is released.

[0662] A "generative AI model" is an artificial intelligence model that generates natural language text based on an input prompt.

[0663] "Editing" is the process of checking the grammatical and phrasing appropriateness of generated information and correcting it as needed.

[0664] "Assessing the risk of online backlash" is the process of analyzing the likelihood of online backlash occurring when generated or edited information is published, and then determining the risk.

[0665] "Methods to encourage correction" refer to methods of notifying users that information deemed to have a high risk of causing a public outcry needs to be corrected.

[0666] An "emotion engine" is a technology that analyzes text posted by users and identifies the emotions (joy, sadness, anger, etc.) contained within that text.

[0667] "Information disclosure" is the process of publishing information on social media, websites, etc., based on analysis results, in order to prevent damage to the company's image.

[0668] "Economic impact forecasting" is the process of predicting the economic impact that publicly available information will have on a company.

[0669] Modes for carrying out the invention

[0670] This invention is a system that prevents damage to a company's image by pre-assessing the risk of information going viral for companies and prompting corrections as necessary. A specific embodiment of this system is described below.

[0671] Hardware and software to be used

[0672] Hardware: Servers, terminals

[0673] Software: Generative AI models, emotion engines, database management systems (DBMS)

[0674] System Configuration

[0675] 1. The server has the means to store past online controversy data in a database and utilize it.

[0676] 2. The server has the means to store effective word data in a database and utilize it.

[0677] 3. The server has the means to generate information using a generative AI model.

[0678] 4. The server has the means to edit the generated information.

[0679] 5. The server has means to assess the risk of the generated or edited information becoming controversial.

[0680] 6. The server has mechanisms to prompt corrections when there is a high risk of a system failure.

[0681] 7. The server has the means to use an emotion engine to analyze the user's emotions.

[0682] 8. The server has the means to publish information based on the analysis results.

[0683] 9. The server has the means to predict the economic impact of publishing this data on social media and websites.

[0684] System operation

[0685] 1. The server uses a generation AI model to generate the information that the company intends to publish. For example, the prompt "Generate an introductory text describing the features of the new product" is input to the generation AI model, and an introductory text for the new product is generated.

[0686] 2. The server uses AI to proofread the generated information. Specifically, it checks grammar and the appropriateness of expressions, and makes corrections as needed.

[0687] 3. The server analyzes the edited information and assesses the risk of online backlash. Natural language processing (NLP) techniques are used to detect risk factors within the text.

[0688] 4. If a server is deemed to be at high risk of becoming unsafe, a notification prompting correction will be sent to the user's device. The user will receive the notification and correct the information as necessary.

[0689] 5. The server uses an emotion engine to analyze the user's emotions. For example, if a user posts "The price of the new product is too high," the server analyzes the text and detects the emotion of sadness.

[0690] 6. Based on the analysis results, the server will publish information to prevent damage to the company's image. For example, it will publish information regarding pricing and service improvements.

[0691] Specific example

[0692] For example, consider a case where a company uses this system when announcing a new product.

[0693] 1. The server inputs the prompt message "Generate an introductory text describing the features of the new product" into the AI ​​model.

[0694] 2. The AI ​​model generates an introductory text such as, "This new product features high performance and a user-friendly design."

[0695] 3. The AI ​​grammatically checks the introductory text generated by the server and corrects it to "It features high performance and a user-friendly design."

[0696] 4. The server analyzes the revised introductory text using NLP technology and determines that the expression "high performance" may implicitly criticize competitors' products.

[0697] 5. The server sends a notification to the device prompting a correction, suggesting that the expression "high performance" be changed to "multifunctional."

[0698] 6. The user receives the notification and corrects it to "It features a multi-functional and user-friendly design."

[0699] 7. The server uses an emotion engine to analyze the text posted by the user, "The price of the new product is too high," and detects the emotion of sadness.

[0700] 8. Based on the analysis results, the server publishes an "explanation regarding the price of the new product," stating that "this new product uses high-quality materials, and therefore the price has been set higher."

[0701] In this way, a system that proactively prevents damage to a company's image functions effectively. The flow of the specific processing in Example 3 will be explained using Figure 21.

[0702] Step 1:

[0703] The server inputs a prompt message into the AI ​​model that generates information. Specifically, it inputs the prompt message "Generate an introductory text describing the features of the new product." Based on this prompt message, the AI ​​model generates an introductory text such as "This new product is high-performance and features a user-friendly design." The input is the prompt message, and the output is the generated introductory text.

[0704] Step 2:

[0705] The server uses AI to proofread the generated information. Specifically, it checks the grammar of the generated introductory text and verifies the appropriateness of the expression. For example, it corrects the sentence "It features high performance and a user-friendly design" to "It features high performance and a user-friendly design." The input is the generated introductory text, and the output is the proofread introductory text.

[0706] Step 3:

[0707] The server analyzes the edited information and assesses the risk of online backlash. Specifically, it uses natural language processing (NLP) techniques to detect risk factors within the text. For example, it might determine that the phrase "high performance" implicitly criticizes a competitor's product. The input is the edited introductory text, and the output is the assessment result of the online backlash risk.

[0708] Step 4:

[0709] If a server is deemed to be at high risk of becoming a "firestorm," a notification prompting corrections will be sent to the user's device. Specifically, the notification will specify the parts that need correction and the reasons why. For example, it might suggest changing the expression "high performance" to "multifunctional." The input is the firestorm risk assessment result, and the output is the notification prompting corrections.

[0710] Step 5:

[0711] The user receives a notification and corrects the information as needed. Specifically, they revise the introductory text based on the notification. For example, they might change the sentence "It features high performance and an easy-to-use design" to "It features multi-functionality and an easy-to-use design." The input is a notification prompting correction, and the output is the corrected introductory text.

[0712] Step 6:

[0713] The server uses an emotion engine to analyze the user's emotions. Specifically, it analyzes the text posted by the user to identify the type of emotion (joy, sadness, anger, etc.). For example, if a user posts "The price of the new product is too high," the server analyzes that text and detects the emotion of sadness. The input is the user's posted text, and the output is the result of the emotion analysis.

[0714] Step 7:

[0715] The server publishes information based on the analysis results to prevent damage to the company's image. Specifically, it provides explanations regarding service improvements and pricing based on the analysis results. For example, it might publish an explanation regarding the pricing of a new product, stating, "This new product uses high-quality materials, and therefore the price is set higher." The input is the result of the sentiment analysis, and the output is the information that is published.

[0716] (Application Example 3)

[0717] Next, we will describe application example 3 of form example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0718] In today's information society, the risk of information released by companies causing a public backlash is increasing, potentially leading to damage to the company's image and economic losses. Furthermore, information releases that disregard user sentiment can worsen the relationship between companies and their users. To address these issues, it is necessary to assess the risk of public backlash before information is released and to optimize information while considering user sentiment.

[0719] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means. In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic effect when the information is made public based on this data, means for analyzing user sentiment, and means for optimizing information based on the analysis results. This makes it possible to evaluate the risk of online controversy before information is made public and to optimize information while taking user sentiment into consideration.

[0720] "Past online controversy data" refers to data on online controversies that occurred due to information that was publicly released in the past.

[0721] "Effective word data" refers to data on words and phrases that are expected to elicit a positive response when used in information disclosure.

[0722] "Artificial intelligence that predicts the economic impact of disclosure" refers to artificial intelligence used to calculate the expected economic impact when information is made public.

[0723] "Means for analyzing user emotions" refers to technologies for analyzing the emotions expressed by users from sources such as text and audio.

[0724] "Means of optimizing information" refers to technologies for modifying and adjusting information to be made public in the most optimal form, based on analysis results and predictive data.

[0725] The system for implementing this invention consists of a server and a user terminal. The server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic effects when published, means for analyzing user sentiment, and means for optimizing information based on the analysis results.

[0726] Hardware and software to be used

[0727] Hardware:

[0728] Server (cloud-based)

[0729] User terminals (smartphones, personal computers)

[0730] software:

[0731] Sentiment analysis engines (e.g., IBM Watson, Google Cloud Natural Language API)

[0732] AI models for assessing online controversy risk (e.g., OpenAI GPT-4®)

[0733] AI models for generating advertisements (e.g., OpenAI GPT-4®)

[0734] Data processing and data calculation

[0735] 1. Ad generation:

[0736] The user's device sends basic information about the advertising campaign (target audience, product information, etc.) to the server.

[0737] The server uses an AI model to generate ad copy.

[0738] 2. Assessment of the risk of online backlash:

[0739] The server inputs the generated ad copy into an AI model for assessing online backlash risk and performs a risk evaluation.

[0740] If the risk is high, the server will present a suggested solution to the user's terminal.

[0741] 3. Emotion analysis:

[0742] The server collects user sentiment data (e.g., social media posts, feedback, etc.).

[0743] The emotion analysis engine is used to analyze the user's emotions.

[0744] Based on the analysis results, the server optimizes the ad content.

[0745] Specific example

[0746] For example, when running an advertising campaign for a new smartphone, the user's device sends a prompt message to the server like the following:

[0747] Example of a prompt

[0748] Target audience: Young people (18-25 years old)

[0749] Product Information: New smartphone "XPhone" with high-performance camera, long battery life, and latest OS.

[0750] Ad copy tone: Casual and approachable

[0751] Based on this prompt, the server uses a generative AI model to generate ad copy, and a crisis risk assessment AI model evaluates the risk. If the risk is high, it suggests revisions, and finally, the sentiment analysis engine optimizes the ad content based on the user's emotions.

[0752] The above describes the embodiments for carrying out this invention.

[0753] The flow of the specific processing in Application Example 3 will be explained using Figure 22.

[0754] Step 1:

[0755] The user enters basic information about the advertising campaign (target audience, product information, ad copy tone, etc.) into their device. The entered information is then sent to the server.

[0756] Step 2:

[0757] The server generates ad copy using an ad generation AI model based on the received basic information. The generated ad copy is temporarily stored on the server.

[0758] Step 3:

[0759] The server inputs the generated ad copy into an AI model for assessing online backlash risk and performs a risk evaluation. The results of the risk evaluation are output, and if the risk is high, the server generates revised suggestions.

[0760] Step 4:

[0761] The server sends the risk assessment results and proposed revisions to the user's terminal. The user reviews the proposed revisions and modifies the ad copy as needed.

[0762] Step 5:

[0763] The server collects user sentiment data (e.g., social media posts, feedback, etc.). The collected sentiment data is then input into a sentiment analysis engine.

[0764] Step 6:

[0765] The server uses an emotion analysis engine to analyze the user's emotions. The analysis results are output and stored on the server.

[0766] Step 7:

[0767] The server optimizes the ad content based on the analysis results. The optimized ad copy is stored on the server.

[0768] Step 8:

[0769] The server sends the optimized ad copy to the user's device. The user reviews the optimized ad copy and publishes it as the final ad copy.

[0770] The above is the processing flow of this program.

[0771] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0772] Data generation model 58 is a form of so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0773] Other examples of generative AI include Gemini® (registered trademark) (Internet search). <url: https: gemini.google.com ?hl="ja">) are some examples.

[0774] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0775] [Second Embodiment]

[0776] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0777] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0778] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0779] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0780] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0781] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0782] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0783] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0784] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0785] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0786] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0787] Next, the identification process performed by the identification processing unit 290 of the data processing device 12 will be described.

[0788] "Example of form 1"

[0789] One embodiment of this system is an AI-powered information disclosure support system for businesses. This system utilizes past data on online controversies and effective keywords to predict the economic impact of information published on social media and websites. Specifically, it stores past online controversy data in a database, which is then analyzed by AI. Similarly, it stores data on effective keywords, and uses this data to edit and create publicly available information.

[0790] "Example of form 2"

[0791] Furthermore, another embodiment of the present invention includes a function to determine the amount of loss incurred when using past online controversy data during the editing process. Specifically, the amount of loss incurred during a controversy is estimated from past controversy data and used as reference information for editing. This makes it possible to quantify the risk of online controversy and disclose information that takes that risk into account.

[0792] "Example of form 3"

[0793] Furthermore, a further embodiment of the present invention includes a function to proactively prevent damage to a company's image. Specifically, before information created or edited by the AI ​​is made public, it evaluates whether the information contains a risk of causing a public outcry and prompts for revisions as necessary. This makes it possible to proactively prevent damage to a company's image.

[0794] The following describes the processing flow for each example of the form.

[0795] "Example of form 1"

[0796] Step 1: Accumulate past online controversy data and effective keyword data in a database.

[0797] Step 2: The AI ​​analyzes this data and edits or creates publicly available information.

[0798] Step 3: Based on the economic impact predicted by the AI, determine the optimal timing and content for information disclosure.

[0799] "Example of form 2"

[0800] Step 1: Estimate the amount of losses incurred during a social media firestorm based on past data from similar incidents.

[0801] Step 2: Use the estimated loss amount as reference information for editing.

[0802] Step 3: Quantify the risk of online backlash and disclose information based on that risk.

[0803] "Example of form 3"

[0804] Step 1: Before the information created or edited by the AI ​​is published, evaluate whether it contains a risk of causing a public outcry.

[0805] Step 2: Based on the evaluation results, prompt for information correction as needed.

[0806] Step 3: Release the revised information to prevent damage to the company's image.

[0807] (Example 1)

[0808] Next, we will describe Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0809] When companies publish information on social media or their websites, they face the challenge of predicting the risk of online backlash and the resulting economic impact. In particular, there is a need for methods to mitigate the risks of publicly released information and maximize its economic benefits by utilizing past online controversies and effective wording. Furthermore, preventing damage to the company's image is also a crucial issue.

[0810] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 1 is realized by the following means. In this invention, the server includes means for collecting and storing past online controversy data, means for collecting and storing effective word data, means for analyzing this data using natural language processing technology, means for predicting the economic effects of publicly available information based on the analysis results, and means for supporting the editing and creation of publicly available information. This makes it possible for companies to reduce the risk of online controversy related to publicly available information and maximize economic effects. It also makes it possible to prevent damage to the company's image in advance.

[0811] "Past online controversy data" refers to information about past online controversies that have occurred on social media or websites, and specifically includes post content, comments, retweet counts, and user reactions.

[0812] "Effective word data" refers to information about keywords and phrases that elicit positive responses on social media and websites, specifically including keywords and phrases that have received favorable user reactions.

[0813] "Natural language processing technology" refers to the technology that enables computers to understand, analyze, and generate human language, and specifically includes sentiment analysis and keyword extraction.

[0814] "Predicting economic impact" refers to evaluating how much traffic publicly available information will attract on social media and websites, or the degree of risk of it going viral or causing a controversy.

[0815] "Proofreading and creation support for publicly available information" refers to improving the quality of user-created information by reducing the risk of online backlash and suggesting effective wording.

[0816] "Preventing damage to a company's image" refers to measures taken to prevent a company's reputation from deteriorating due to information it makes public.

[0817] Modes for carrying out the invention

[0818] This invention is an information disclosure support system designed to reduce the risk of online backlash and maximize economic benefits when companies publish information on social media and their websites. The system collects and stores past online backlash data and effective keyword data, and uses natural language processing technology to analyze this data. Furthermore, it predicts the economic impact of the published information based on the analysis results and supports the editing and creation of such information.

[0819] Data collection and storage

[0820] The server collects past online controversy data from social media and websites. Specifically, it uses social media APIs to search for posts containing specific keywords and hashtags, and retrieves data such as comments and retweet counts related to the controversies. It also uses marketing research tools to collect data on effective keywords. For example, it uses Google Analytics and SEMrush to analyze how much traffic specific keywords have attracted. The collected data is stored in databases such as MySQL® and PostgreSQL®.

[0821] Data analysis

[0822] The server analyzes accumulated online controversy data using TensorFlow® and PyTorch®. Specifically, it uses natural language processing (NLP) techniques to analyze the sentiment of posts and identify factors that cause online controversies. Similarly, it analyzes data on effective words to identify keywords and phrases that elicit positive responses. For example, it uses models such as Word2Vec and BERT to evaluate the relevance of keywords.

[0823] Predicting the economic impact

[0824] Based on the analysis results, the server predicts the economic impact of the information to be released. Specifically, it evaluates how much traffic the information scheduled for release will attract and the risk of it causing a public outcry. This prediction is made by combining historical data with an AI model.

[0825] Editing and creating publicly available information

[0826] Users create public information and send it from their device to the server. For example, they might input a new product description or campaign announcement for a company. The server evaluates the risk of online backlash based on the submitted information. Specifically, it analyzes the input text using NLP technology and assesses the risk by comparing it with past online backlash data. The server suggests effective words and revises the information. For example, it might make specific suggestions such as, "This expression has a high risk of causing a backlash, so please change it to this expression." Users revise the information based on the server's suggestions and create the final public information. They can also send the revised information back to the server for final confirmation.

[0827] Specific example

[0828] Example 1: Assessment of the risk of online backlash

[0829] A user creates a product description for a new product and sends it to the server from their device. The server evaluates the risk of this description causing a backlash based on past data on product controversies. For example, if the server determines that a phrase like "This product is superior to competitors' products" increases the risk of backlash, it suggests changing it to something like "This product is well-received by many users."

[0830] Example 2: Suggestions for effective words

[0831] The user creates a campaign announcement and sends it from their device to the server. The server, based on a database of effective words, suggests keywords that will elicit a positive response to the announcement. For example, if keywords such as "limited," "free," and "special offer" are deemed effective, the server suggests adding these keywords to the announcement.

[0832] Example of a prompt

[0833] "I have created a product description for our new product. Please assess the risk of this description causing a backlash if it is published on social media, and suggest revisions as needed."

[0834] "I've created a campaign announcement. Please suggest effective words to elicit a positive response."

[0835] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0836] Step 1: Data collection and storage

[0837] The server collects past online controversy data from social media and websites. Specifically, it uses social media APIs to search for posts containing specific keywords and hashtags, and retrieves data such as comments and retweet counts related to the controversies. The input is specific keywords and hashtags, and the output is the collected controversy data. The server stores the collected data in databases such as MySQL® or PostgreSQL®.

[0838] Step 2: Effective word data collection

[0839] The server uses marketing research tools to collect data on effective keywords. For example, it uses Google Analytics or SEMrush to analyze how much traffic specific keywords attract. The input is the specific keyword, and the output is data on effective keywords. The server stores the collected data in a database.

[0840] Step 3: Data Analysis

[0841] The server analyzes accumulated online controversy data using TensorFlow® and PyTorch®. Specifically, it uses natural language processing (NLP) techniques to analyze the sentiment of posts and identify the factors causing online controversies. The input is accumulated online controversy data, and the output is the identification of the factors causing the controversies. Similarly, it analyzes effective word data to identify keywords and phrases that elicit positive responses. The input is effective word data, and the output is keywords and phrases that elicit positive responses.

[0842] Step 4: Predicting the Economic Impact

[0843] The server predicts the economic impact of publicly released information based on the analysis results. Specifically, it evaluates how much traffic the information scheduled for release will attract and the risk of it causing a public outcry. The input is the analysis results, and the output is the predicted economic impact.

[0844] Step 5: Create and submit public information

[0845] Users create public information and send it from their terminal to the server. For example, they might input a company's new product introduction or campaign announcement. The input is the public information created by the user, and the output is the public information sent to the server.

[0846] Step 6: Evaluation and proposals for publicly available information

[0847] The server assesses the risk of online backlash based on the submitted information. Specifically, it analyzes the input text using NLP technology and evaluates the risk by comparing it with past online backlash data. The input is publicly available information created by the user, and the output is the assessment result of the online backlash risk. The server suggests effective words and edits the information. For example, it makes specific suggestions such as, "This expression has a high risk of causing online backlash, so please change it to this expression." The input is the assessment result of the online backlash risk, and the output is the suggested revisions.

[0848] Step 7: Correction and final confirmation of publicly available information

[0849] Users revise the information based on the server's suggestions to create the final published information. They can also resubmit the revised information to the server for final confirmation. The input is the server's revision suggestions, and the output is the revised published information.

[0850] (Application Example 1)

[0851] Next, we will describe Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0852] When companies publish advertisements on social media and their websites, they are required to predict the economic impact by utilizing past data on online controversies and effective keywords. However, there are no systems that effectively utilize this data to support the editing and creation of ad copy. Furthermore, there is a lack of means to proactively prevent damage to a company's image by predicting online controversy risk scores and effective keyword scores. As a result, companies are forced to operate their advertising campaigns while carrying inherent risks.

[0853] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0854] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including AI to predict the economic effects when published on social media and websites based on this data, means for assisting in editing and creating ad copy, and means for predicting online controversy risk scores and effective word scores. This enables companies to reduce risks when publishing advertisements and create effective ad copy.

[0855] "Past online controversy data" refers to information about past online controversies that have occurred on social media or websites.

[0856] "Effective word data" refers to information about words and phrases that are highly effective in advertising and information dissemination.

[0857] "AI that predicts economic impact" refers to artificial intelligence that predicts the economic impact of information published on social media and websites based on past data.

[0858] "Means of supporting the editing and creation of ad copy" refers to functions that assist companies in revising or creating new ad copy based on past data.

[0859] A "crisis risk score" refers to an index that quantifies the likelihood of a particular advertisement or piece of information causing a social media firestorm or backlash on a website.

[0860] "Effective word score" refers to a metric that quantifies the likelihood that a particular ad copy or piece of information will be highly effective.

[0861] "Means of preventing damage to a company's image in advance" refers to functions that prevent a company's image from being negatively affected by information it makes public.

[0862] The system for implementing this invention includes an AI that utilizes past online controversy data and effective word data to predict the economic impact of advertising copy published on social media and websites. Furthermore, it has functions to support the editing and creation of advertising copy and predict online controversy risk scores and effective word scores.

[0863] System Configuration

[0864] Hardware:

[0865] server

[0866] smartphone

[0867] software:

[0868] Python (registered trademark)

[0869] pandas

[0870] scikit-learn(registered trademark)

[0871] joblib

[0872] Data handling

[0873] The server stores historical online controversy data and effective keywords in a database. This data is saved as a CSV file and read using pandas. The read data is then converted into a numerical vector using TfidfVectorizer.

[0874] Model training and saving

[0875] The server trains a LogisticRegression model based on data converted into numerical vectors using TfidfVectorizer. The trained model is saved using joblib. This lays the foundation for predicting crisis risk scores and effective word scores.

[0876] Evaluation of ad copy

[0877] When a user enters ad copy using their smartphone, the server uses a stored model to predict the ad's risk of backlash and its effectiveness score. This allows users to understand the risks and effectiveness of their ad copy in advance.

[0878] Specific example

[0879] For example, if a user enters the following ad text:

[0880] To celebrate the launch of our new product, we're holding a special sale!

[0881] The server predicts a risk of backlash score and an effective word score for this ad copy. The predicted results show a risk of backlash score of 0.2 and an effective word score of 0.8.

[0882] Example of a prompt

[0883] Please predict the risk of backlash and the effectiveness score for the following ad copy.

[0884] Advertisement: "To celebrate the launch of our new product, we're holding a special sale!"

[0885] In this way, companies can reduce the risks involved in advertising and create effective ad copy.

[0886] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0887] Step 1:

[0888] The server reads past online controversy data and effective keywords from a database. This data is stored as a CSV file and read using pandas. The input is a CSV file, and the output is a DataFrame. Specifically, the pandas.read_csv function is used to read the data.

[0889] Step 2:

[0890] The server converts the loaded data into a numerical vector using TfidfVectorizer. The input is a data frame, and the output is a numerical vector. Specifically, the TfidfVectorizer's fit_transform method is used to convert text data into a numerical vector.

[0891] Step 3:

[0892] The server trains a LogisticRegression model based on data that has been converted into numerical vectors. The input is a numerical vector, and the output is the trained model. Specifically, the model is trained using the LogisticRegression's fit method.

[0893] Step 4:

[0894] The server saves the trained model and TfidfVectorizer using joblib. The input is the trained model and TfidfVectorizer, and the output is the saved model file. Specifically, the model and vectorizer are saved using the joblib.dump function.

[0895] Step 5:

[0896] The user enters the ad text using their smartphone. The input is the text of the ad, and the output is a request to the server. Specifically, the user enters the ad text through a smartphone application and sends it to the server.

[0897] Step 6:

[0898] The server converts the received ad copy into a numerical vector using TfidfVectorizer. The input is the text of the ad copy, and the output is a numerical vector. Specifically, the server uses the transform method of the saved TfidfVectorizer to convert the ad copy into a numerical vector.

[0899] Step 7:

[0900] The server predicts the risk of backlash and the effectiveness of words based on the ad copy converted into numerical vectors. The input is a numerical vector, and the output is the risk of backlash and the effectiveness of words. Specifically, the scores are predicted using the predict_proba method of the stored LogisticRegression model.

[0901] Step 8:

[0902] The server returns a predicted crisis risk score and an effective word score to the user. The input is the predicted score, and the output is the response to the user. Specifically, the score is displayed through a smartphone application.

[0903] (Example 2)

[0904] Next, we will describe Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0905] In today's information society, when companies and individuals publish information through social media and websites, they are required to use past online controversy data to predict risks and take appropriate action. However, conventional systems lack the means to effectively utilize past controversy data and concretely estimate the amount of loss, resulting in insufficient risk management. Furthermore, there is a lack of concrete means to prevent damage to a company's image in advance.

[0906] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0907] This invention provides a server that includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence that predicts the economic effects of information disclosure based on this data, means for analyzing past online controversy data and estimating the amount of loss, means for storing the estimated amount of loss in a database, and means for providing risk information, including the estimated amount of loss, when a user discloses information. This makes it possible to estimate specific amounts of loss based on past online controversy data, quantify the risk, and take appropriate action when disclosing information. It also provides specific means to prevent damage to a company's image in advance.

[0908] "Past online controversy data" refers to data on past online controversies that have occurred on online platforms such as social media and websites.

[0909] "Effective word data" refers to data on keywords and phrases that are effective when used when publishing information on social media or websites.

[0910] "Economic impact of information disclosure" refers to the expected economic effects and benefits that can be anticipated when information is made public on social media or a website.

[0911] Artificial intelligence is a system that uses machine learning and data analysis techniques to learn patterns from past data and make predictions and decisions about the future.

[0912] "Loss amount" refers to the estimated amount of economic loss based on past cases of online controversies.

[0913] A "database" is a system for efficiently storing, managing, and retrieving data.

[0914] "Risk information" refers to information about risks that should be considered when disclosing information, and specifically includes estimated loss amounts.

[0915] "Downturn in corporate image" refers to a decline in a company's reputation or brand value.

[0916] This invention is a system that utilizes past online controversy data to quantify the risks associated with information disclosure and enable appropriate responses. The following describes a specific embodiment of this system.

[0917] Server Role

[0918] The server retrieves past online fire incident data from a database, analyzes it, and estimates the amount of loss. Specifically, the server uses the following hardware and software:

[0919] Hardware: Server machine with a high-performance processor and sufficient memory.

[0920] Software: Relational database management systems (RDBMS) such as MySQL® and PostgreSQL®, and data analysis libraries such as pandas and scikit-learn® for Python®.

[0921] The server first executes an SQL query to retrieve past online fire incident data from the database. Next, it converts the retrieved data into a data frame using the Python® pandas library and performs preprocessing. Using the preprocessed data, it estimates the amount of loss using the scikit-learn® LinearRegression model. The estimated amount of loss is then saved back to the database.

[0922] Terminal role

[0923] The terminal retrieves risk information from the server when the user publishes information and displays it to the user. Specifically, the terminal uses the following hardware and software.

[0924] Hardware: Personal computers, tablets, smartphones, etc.

[0925] Software: Web browser, HTML, CSS, JavaScript

[0926] The device sends an HTTP GET request to the server and receives risk information in JSON format. The received risk information is then displayed on a web page using HTML and JavaScript.

[0927] User roles

[0928] Users will check risk information through their devices and adjust the content of the information they disclose. Specifically, users will check the estimated loss amount displayed on the web page and take action such as changing the disclosed content if the risk is high.

[0929] Specific example

[0930] Specific Example 1: Acquisition and Analysis of Online Crisis Data

[0931] The server retrieves past social media controversy data from a MySQL® database. For example, it uses an SQL query like the following:

[0932] SQL

[0933] SELECT FROM flame_data WHERE date >= '2020-01-01';

[0934] The acquired data is analyzed using the Python pandas library to estimate the amount of loss incurred during a social media firestorm. For example, the following code is used:

[0935] Python (registered trademark)

[0936] import pandas as pd

[0937] from sklearn.linear_model import LinearRegression

[0938] Loading data

[0939] data = pd.read_sql('SELECT FROM flame_data WHERE date >= "2020-01-01"', con=database_connection)

[0940] Estimation of loss amount

[0941] model = LinearRegression()

[0942] model.fit(data[['followers', 'negative_comments']], data['loss_amount'])

[0943] predicted_loss = model.predict(new_data[['followers', 'negative_comments']])

[0944] Example 2: Displaying risk information

[0945] When a user publishes information, the device retrieves risk information from the server and displays it on an HTML page. For example, the following code is used:

[0946] html

[0947] <!DOCTYPE html>

[0948]

[0949]

[0950] <title> Risk Information< / title>

[0951]

[0952]

[0953] <h1> Risk Information< / h1>

[0954] Estimated loss: \

[0955] <script>

[0956] fetch('https: / / example.com / api / risk_info')

[0957] .then(response => response.json())

[0958] .then(data => {

[0959] document.getElementById('loss_amount').textContent = data.loss_amount;

[0960] });

[0961] < / script>

[0962]

[0963]

[0964] Example of a prompt

[0965] Examples of prompts to input into a generative AI model include the following:

[0966] "Please generate Python® code that estimates the amount of financial loss incurred during a social media firestorm, using past social media firestorm data."

[0967] By using this prompt statement, the generated AI model can produce appropriate Python® code.

[0968] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0969] Step 1:

[0970] The server retrieves past online controversy data from the database. It accepts database connection information and an SQL query as input, and generates a dataframe containing the controversy data as output. Specifically, the server executes the SQL query to extract past controversy data. For example, it might use the following SQL query:

[0971] SQL

[0972] SELECT FROM flame_data WHERE date >= '2020-01-01';

[0973] This query retrieves online controversy data for the specified period.

[0974] Step 2:

[0975] The server analyzes the acquired data on online controversies. It accepts a dataframe as input and generates a pre-processed dataframe as output. Specifically, the server imputes missing values ​​and removes unnecessary columns. For example, it preprocesses the data using the Python® pandas library.

[0976] Step 3:

[0977] The server estimates the loss amount using preprocessed data. It takes a preprocessed dataframe as input and generates an estimated loss amount as output. Specifically, the server uses the scikit-learn® LinearRegression model to estimate the loss amount. For example, the model is trained and the loss amount is predicted as follows:

[0978] Python (registered trademark)

[0979] model = LinearRegression()

[0980] model.fit(data[['followers', 'negative_comments']], data['loss_amount'])

[0981] predicted_loss = model.predict(new_data[['followers', 'negative_comments']])

[0982] This process estimates the amount of loss.

[0983] Step 4:

[0984] The server stores the estimated loss amount in a database. It takes the estimated loss amount and database connection information as input and generates the loss amount stored in the database as output. Specifically, the server executes an SQL query to insert the estimation results into the database. For example, it might use the following SQL query:

[0985] SQL

[0986] INSERT INTO risk_info (date, predicted_loss) VALUES ('2023-10-01', 500000);

[0987] This query will save the estimated loss amount to the database.

[0988] Step 5:

[0989] The device retrieves risk information from the server when the user publishes information. It receives an HTTP GET request as input and generates JSON data containing the risk information as output. Specifically, the device sends an HTTP GET request to the server and receives the risk information. For example, it uses JavaScript code like the following:

[0990] JavaScript

[0991] fetch('https: / / example.com / api / risk_info')

[0992] .then(response => response.json())

[0993] .then(data => {

[0994] document.getElementById('loss_amount').textContent = data.predicted_loss;

[0995] });

[0996] This process retrieves risk information.

[0997] Step 6:

[0998] The device displays the acquired risk information to the user. It receives JSON data containing risk information as input and generates an HTML page displaying the risk information as output. Specifically, the device uses HTML and JavaScript to display the risk information on the web page. For example, it can be implemented as follows:

[0999] html

[1000] <!DOCTYPE html>

[1001]

[1002]

[1003] <title> Risk Information< / title>

[1004]

[1005]

[1006] <h1> Risk Information< / h1>

[1007] Estimated loss: \

[1008] <script>

[1009] fetch('https: / / example.com / api / risk_info')

[1010] .then(response => response.json())

[1011] .then(data => {

[1012] document.getElementById('loss_amount').textContent = data.predicted_loss;

[1013] });

[1014] < / script>

[1015]

[1016]

[1017] This process displays risk information to the user.

[1018] Step 7:

[1019] Users check risk information through their devices. They receive the displayed risk information as input and adjust the content of the information disclosure as output. Specifically, users check the estimated loss amount displayed on the webpage and take action, such as changing the disclosed content, if the risk is high.

[1020] (Application Example 2)

[1021] Next, we will describe application example 2 of form example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 as the "terminal".

[1022] In advertising campaigns, there is a need for a system that can quantify the risk of online backlash by utilizing past data and predict the amount of loss. However, current systems have the challenge of not being able to specifically quantify the risk of online backlash and provide information disclosure and concrete advice for risk reduction based on that risk.

[1023] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1024] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including a generative AI model that predicts the economic effect when published on social media or websites based on this data, means for inputting the content of an advertising campaign and quantifying the risk of online controversy based on past controversy data, means for estimating the amount of loss based on the risk of online controversy, and means for providing specific advice to reduce the risk. This makes it possible to quantify the risk of online controversy in an advertising campaign and provide information disclosure and specific advice to reduce the risk based on that risk.

[1025] "Past online controversy data" refers to data on negative reactions, comments, and criticisms that have occurred on social media or websites in the past.

[1026] "Effective word data" refers to data on keywords and phrases that are considered effective in eliciting positive responses on social media and websites.

[1027] A "generative AI model" refers to an algorithm or system that uses machine learning or artificial intelligence technology to generate a specific output from input data.

[1028] "Methods for quantifying the risk of online backlash" refer to methods or systems that express the likelihood of a particular advertising campaign going viral in a numerical way, based on past online backlash data.

[1029] "Means for estimating the amount of loss" refers to methods or systems for calculating predicted economic losses based on the risk of online backlash.

[1030] "Means of providing specific advice" refers to methods and systems that propose specific actions and measures to reduce the risk of online backlash.

[1031] The system for implementing this invention operates through the coordinated efforts of a server, a terminal, and a user. A specific embodiment is described below.

[1032] Server Role

[1033] The server stores historical data on online controversies and effective keywords, and uses a generative AI model based on this data to predict economic impact. The server uses the following software and hardware:

[1034] Software: Python (registered trademark), Pandas, Scikit-learn (registered trademark)

[1035] Hardware: Servers equipped with high-performance processors and large amounts of memory.

[1036] The server reads past online controversy data and converts it into a data frame. Next, it sets features (campaign length, audience size, number of negative comments) and a target (amount of loss) and trains a linear regression model. This allows the system to quantify the risk of online controversy and predict the amount of loss when given the content of an advertising campaign.

[1037] Terminal role

[1038] The device provides an interface for users to input details about advertising campaigns. The device uses the following software and hardware:

[1039] Software: Web browsers, mobile applications

[1040] Hardware: Smartphones, tablets, PCs

[1041] When a user enters details of an advertising campaign through their device, that data is sent to a server. Based on the received data, the server quantifies the risk of a social media firestorm and predicts the amount of potential loss. The prediction results are sent back to the device and displayed to the user.

[1042] User roles

[1043] Users enter the details of their advertising campaign using their device and view predictions of potential backlash and potential losses. They also receive specific risk mitigation advice from the server and take appropriate measures.

[1044] Specific example

[1045] For example, if a user enters an ad campaign length of 30 days, an audience size of 100,000 people, and 50 negative comments, the server uses this data to quantify the risk of a social media firestorm and calculate the estimated loss. As a result, the estimated loss is displayed, and specific advice for risk reduction is provided.

[1046] Example of a prompt

[1047] "If an ad campaign lasts 30 days, has an audience size of 100,000, and receives 50 negative comments, what is the estimated loss?"

[1048] In this way, it becomes possible to quantify the risk of online backlash in advertising campaigns and provide specific advice on information disclosure and risk reduction based on that risk.

[1049] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1050] Step 1:

[1051] The user uses their device to enter details about the advertising campaign.

[1052] Input: ad campaign length, audience size, number of negative comments

[1053] Output: Input data is sent to the server.

[1054] Specific operation: The user enters details of the advertising campaign through a smartphone or computer interface and clicks the "Submit" button.

[1055] Step 2:

[1056] Based on the data received by the server, past online controversies are read.

[1057] Input: Detailed data of advertising campaigns submitted by users

[1058] Output: Past online controversy data is converted into a data frame.

[1059] Specific operation: The server uses the Python® Pandas library to read past online controversy data from a CSV file and convert it into a DataFrame.

[1060] Step 3:

[1061] The server sets the features and target, and then trains a linear regression model.

[1062] Input: Past online controversy data converted into a data frame

[1063] Output: Trained linear regression model

[1064] Specific operation: The server uses the Scikit-learn® library to set features (campaign length, audience size, number of negative comments) and target (loss amount), and then trains a linear regression model.

[1065] Step 4:

[1066] Based on detailed advertising campaign data received by the server from users, the risk of a social media firestorm is quantified, and the amount of potential loss is predicted.

[1067] Input: Detailed data on advertising campaigns received from users, a trained linear regression model.

[1068] Output: Crisis risk and estimated loss amount

[1069] Specific operation: The server takes detailed data of received advertising campaigns as features, uses a trained linear regression model to quantify the risk of backlash, and predicts the amount of loss.

[1070] Step 5:

[1071] The server sends back the predicted risk of a firestorm and the estimated loss amount to the terminal.

[1072] Input: Crisis risk and estimated loss amount

[1073] Output: The prediction results are displayed on the device.

[1074] Specific operation: The server sends the prediction results to the terminal in JSON format, and the terminal receives and displays them to the user.

[1075] Step 6:

[1076] The server generates specific advice to mitigate risk and sends it to the terminal.

[1077] Input: Crisis risk and estimated loss amount

[1078] Output: Specific advice for risk reduction

[1079] Specific operation: The server uses a generated AI model to generate specific advice to reduce the risk of online firestorms and sends it to the terminal. The terminal receives this and displays it to the user.

[1080] In this way, it becomes possible to quantify the risk of online backlash in advertising campaigns and provide specific advice on information disclosure and risk reduction based on that risk.

[1081] (Example 3)

[1082] Next, we will describe Embodiment 3 of Embodiment Example 3. 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".

[1083] In today's information society, companies face an increasing risk of online backlash when publishing information through social media and websites. This risk can damage a company's image and lead to financial losses, making it crucial to assess the risk beforehand and modify information as needed. However, current systems fail to adequately assess this risk and modify information, posing a significant challenge for businesses.

[1084] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 3 is realized by the following means.

[1085] This invention includes a server that utilizes past online controversy data, a server that utilizes effective word data, a server that uses a generative AI model to generate information based on this data, a server that evaluates the risk of online controversy before the generated information is made public, a server that prompts correction of the information if there is a risk of online controversy, and a server that predicts the economic effects of publishing the information on social media or a website based on these means. This enables companies to evaluate the risk of online controversy before disclosing information and correct the information as needed, thereby preventing damage to their corporate image.

[1086] "Past online controversy data" refers to data on past online controversies caused by information published on social media, websites, etc.

[1087] "Effective word data" refers to data on words and phrases that are considered effective in eliciting positive responses when publishing information on social media or websites.

[1088] A "generative AI model" is an artificial intelligence model that generates information based on prompt text entered by the user.

[1089] "Methods for assessing the risk of online backlash" refer to methods for analyzing and evaluating the likelihood of online backlash when generated information is made public.

[1090] "Methods to encourage information correction" refer to methods used to prompt users to correct information that has been assessed as having a risk of causing a public outcry.

[1091] "Methods for predicting economic effects" refer to methods for predicting the economic impact that information published on social media and websites has on companies.

[1092] Modes for carrying out the invention

[1093] This invention is a system that prevents damage to a company's image by pre-assessing the risk of online backlash when a company publishes information on social media or its website, and by modifying the information as needed. This system consists of three main elements: a server, a terminal, and a user.

[1094] 1. Program generation

[1095] The server generates a program to proactively prevent damage to the company's image. This program uses a generative AI model to create or edit information and has the function of evaluating the risk of online backlash before the information is made public. Specifically, it uses OpenAI's GPT-4 (registered trademark) as the generative AI model.

[1096] 2. Program Processing

[1097] The server uses a generative AI model (e.g., OpenAI's GPT-4®) to generate information based on the prompt text entered by the user. The generated information is then sent to a crisis risk assessment module. This module uses natural language processing (NLP) techniques to analyze the content of the information and assess whether there is a crisis risk. Specifically, it uses a risk assessment algorithm (e.g., a BERT-based classification model) to determine whether the information contains a crisis risk.

[1098] The terminal sends the prompt text entered by the user to the server and receives generated information and risk assessment results from the server. The user reviews the risk assessment results on the terminal and corrects the information as needed.

[1099] 3. Specific Examples and Examples of Prompt Statements

[1100] As a concrete example, consider a case where a user is creating a press release for a new product. The user enters the following prompt into the terminal:

[1101] Example of a prompt:

[1102] "Please create a press release for our new product. The product name is 'EcoSmart,' and please emphasize its environmentally friendly features."

[1103] The server inputs this prompt into the AI ​​model that generates the following information:

[1104] Examples of generated information:

[1105] "EcoSmart is a new product that utilizes the latest environmental technologies. It is energy-efficient and uses recyclable materials."

[1106] Next, the server sends the generated information to a crisis risk assessment module for risk assessment. If the assessment result is determined to be "risky," the server sends a message to the user prompting them to make corrections. The user can then review the proposed corrections on their device and modify the information as needed to prevent damage to the company's image.

[1107] This system allows companies to assess the risk of public backlash before disclosing information and, if necessary, modify the information to prevent damage to their corporate image. The flow of the specific processing in Example 3 will be explained using Figure 15.

[1108] Step 1:

[1109] The user enters a prompt message.

[1110] The user enters a prompt message into the terminal's input field. For example, they might enter, "Please create a press release for a new product. The product name is 'EcoSmart,' and please emphasize its environmentally friendly features." The entered prompt message is temporarily stored in the terminal's memory.

[1111] Step 2:

[1112] The terminal sends a prompt message to the server.

[1113] The terminal converts the prompt text entered by the user into the appropriate format and sends a request to the server's API endpoint. Specifically, it sends the prompt text to the server using an HTTP POST request. The input is the prompt text, and the output is the request sent to the server.

[1114] Step 3:

[1115] The server generates information using an AI model.

[1116] The server inputs the received prompt message into a generating AI model (e.g., OpenAI's GPT-4®) and generates information. The generating AI model analyzes the prompt message and generates information based on the appropriate context. The input is the prompt message, and the output is the generated information. For example, information such as "EcoSmart is a new product that utilizes the latest environmental technologies. It is energy-efficient and uses recyclable materials." might be generated.

[1117] Step 4:

[1118] The server sends the generated information to the fire risk assessment module.

[1119] The server sends the generated information to a crisis risk assessment module for risk assessment. This module uses natural language processing (NLP) techniques to analyze the content of the information and assess whether there is a risk of crisis. Specifically, it uses a risk assessment algorithm (e.g., a BERT-based classification model) to determine whether the information contains a risk of crisis. The input is the generated information, and the output is the risk assessment result.

[1120] Step 5:

[1121] The server generates the risk assessment results and sends them to the terminal.

[1122] The server receives the assessment results from the fire risk assessment module and sends them to the terminal. The assessment results include judgments such as "no risk" or "risk present." The input is the risk assessment results, and the output is the transmission to the terminal.

[1123] Step 6:

[1124] The device displays the risk assessment results to the user.

[1125] The terminal displays the risk assessment results received from the server to the user. For example, a message such as "This information has a risk of causing a public outcry. We recommend correcting it." might be displayed. The input is the risk assessment results, and the output is what is displayed to the user.

[1126] Step 7:

[1127] Users can modify the information as needed.

[1128] The user reviews the risk assessment results displayed on their device and corrects the information as needed. The corrected information is sent back to the server and, after going through the same process, is finally published as safe information. The input is the corrected information, and the output is the re-evaluated information.

[1129] (Application Example 3)

[1130] Next, we will describe application example 3 of form example 3. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 as the "terminal".

[1131] When companies publish information on social media or their websites, they are required to assess the risk of online backlash beforehand and prevent damage to their corporate image. However, traditional methods have not been able to fully utilize past backlash data or effective keyword data, and predictions of economic effects have been insufficient. Furthermore, they lacked the function of providing concrete corrective measures, making it difficult for companies to respond quickly and appropriately. As a result, companies often took on unnecessary risks.

[1132] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means.

[1133] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic effects of disclosing information based on this data, means for evaluating the risk of information becoming controversial using a generative AI model, and means for presenting specific corrective measures if there is a risk of controversy. This enables companies to evaluate the risk of controversy before disclosing information and obtain specific corrective measures as needed.

[1134] "Past online controversy data" refers to information about past online controversies that have occurred on social media, websites, etc. Specifically, it includes data such as the causes and impact of the controversies, and related keywords.

[1135] "Effective word data" refers to data on words and phrases used on social media, websites, and other platforms that are considered to be particularly effective in eliciting positive responses.

[1136] "Artificial intelligence that predicts the economic effects of disclosing information" refers to artificial intelligence technology used to predict the impact that disclosed information will have on a company's economic activities, and it makes predictions based on past data and trends.

[1137] A "generative AI model" is an artificial intelligence model that uses natural language processing technology to generate text, and has the ability to generate sentences based on specific prompts.

[1138] "Methods for assessing the risk of online backlash" refer to methods for evaluating the likelihood of information being made public causing online backlash, and these methods use past online backlash data and generative AI models to determine the risk.

[1139] "Means of presenting specific revision proposals" refers to methods for presenting specific revision proposals to mitigate the risk of information being deemed to have a high risk of causing a public outcry.

[1140] The system for implementing this invention operates through the coordinated efforts of a server, a terminal, and a user. A specific embodiment is described below.

[1141] System Configuration

[1142] The server includes means of utilizing past online controversy data, means of utilizing effective word data, means of including artificial intelligence to predict the economic effects of disclosing information, means of using generative AI models to assess the risk of information going viral, and means of presenting specific corrective measures if there is a risk of going viral.

[1143] A terminal is a device used by users to input information and communicate with a server. Specifically, this includes smartphones and personal computers.

[1144] Users are typically corporate public relations or marketing personnel who input information and receive feedback from the server.

[1145] Program processing

[1146] The server first receives information entered by the user. This information includes advertisements and campaign details that are planned to be published on social media and websites.

[1147] Next, the server assesses the information's risk of causing a controversy based on past controversy data and effective word data. Specifically, it uses a generative AI model to determine whether the input information contains a risk of causing a controversy.

[1148] If the evaluation determines that there is a risk of a social media firestorm, the server will generate specific revision proposals. These revision proposals use a generation AI model to suggest specific wording and phrases to mitigate the risk.

[1149] Finally, the server sends the evaluation results and suggested corrections to the user's terminal. The user receives this and corrects the information as needed.

[1150] Hardware and software to be used

[1151] Hardware: Servers, smartphones, personal computers

[1152] Software: OpenAI API, Python (registered trademark)

[1153] Specific example

[1154] For example, suppose a user enters the following ad text.

[1155] This product is superior to other products on the market. Buy it now!

[1156] The server receives this ad text and generates the following prompt.

[1157] Please evaluate the following ad copy and determine if it poses a risk of causing a social media firestorm.

[1158] Advertisement: This product is superior to other products. Buy now!

[1159] If there is a risk of backlash, please provide specific proposed solutions.

[1160] The generative AI model evaluates this prompt and proposes the following revisions.

[1161] This advertisement contains disparaging language towards competitors' products and carries the risk of backlash. We recommend revising it as follows:

[1162] Revised version: "This product has satisfied many customers. Buy it now and experience the difference!"

[1163] In this way, users can receive feedback from the server and correct the information, thereby reducing the risk of online backlash and preventing damage to the company's image.

[1164] The flow of the specific processing in Application Example 3 will be explained using Figure 16.

[1165] Step 1:

[1166] Users use their devices to input information such as ad copy and campaign details.

[1167] Input: Ad copy and campaign details

[1168] Output: The entered information is sent to the server.

[1169] Specific operation: The user enters the advertisement text into an input form on their smartphone or computer and presses the submit button. The device sends this information to the server.

[1170] Step 2:

[1171] The server compares the information it receives with past online controversy data and effective keyword data.

[1172] Input: User-submitted information, past online controversy data, effective keyword data

[1173] Output: Matching results (data necessary for evaluating the risk of online backlash)

[1174] Specific operation: The server retrieves past online controversy data and effective keyword data from the database and compares it with the information received from the user.

[1175] Step 3:

[1176] The server uses a generated AI model to assess the risk of information going viral.

[1177] Input: Matching results, generated AI model

[1178] Output: Evaluation results of the risk of online backlash

[1179] Specific operation: The server inputs the matching results into the generating AI model, generates a prompt message, and performs an evaluation. The generating AI model determines the risk of a firestorm and returns the result to the server.

[1180] Step 4:

[1181] If a server is at risk of crashing, it will generate specific corrective action plans.

[1182] Input: Results of the assessment of the risk of online backlash

[1183] Output: Revision proposal

[1184] Specific operation: The server uses the generative AI model again to generate specific corrective measures to reduce the risk of a crisis. The generative AI model generates corrective measures based on the prompt and returns the results to the server.

[1185] Step 5:

[1186] The server sends the evaluation results and suggested corrections to the user's terminal.

[1187] Input: Evaluation results, proposed revisions

[1188] Output: Evaluation results and proposed revisions sent to the user's device.

[1189] Specific operation: The server compiles the evaluation results and suggested revisions and sends them to the user's device. The user reviews this information on their device and modifies the ad copy as needed.

[1190] Step 6:

[1191] The user corrects the information based on the proposed revisions and resends it to the server.

[1192] Input: Information corrected based on the proposed amendments

[1193] Output: The corrected information is sent to the server.

[1194] Specific operation: The user modifies the ad copy based on the suggested revisions received from the server and resubmits it to the server. The server receives the revised information and performs a re-evaluation.

[1195] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1196] "Example of form 1"

[1197] One example of an invention incorporating an emotion engine is a system that combines an emotion engine that recognizes user emotions with an AI that predicts the economic impact of publishing content on social media and websites based on past online controversy data and effective word data. Specifically, it analyzes emotions from text data posted by users and uses the results to edit or create published information. For example, if a user posts text expressing joy, that information can be used to publish information that reinforces a positive image of the company's products or services.

[1198] "Example of form 2"

[1199] As an example of an invention combining an emotion engine, the system described in claim 2 incorporates an emotion engine that recognizes the user's emotions as a means for determining the amount of loss incurred when past online controversy data is used during editing. Specifically, the system analyzes the user's emotions and evaluates the risk of online controversy based on the results. For example, if a user posts text indicating anger, that information is used to avoid disclosing information that would increase the risk of online controversy.

[1200] "Example of form 3"

[1201] As an example of a third form of the invention incorporating an emotion engine, the system described in claim 3 incorporates an emotion engine that recognizes user emotions as a means to prevent damage to the company's image in advance. Specifically, it analyzes the user's emotions and, based on the results, discloses information to prevent damage to the company's image. For example, if a user posts text indicating sadness, that information is used to disclose information about the company's response and service improvements.

[1202] The following describes the processing flow for each example of the form.

[1203] "Example of form 1"

[1204] Step 1: The user posts text data to social media or a website.

[1205] Step 2: The emotion engine analyzes the user's emotions from the text data.

[1206] Step 3: The AI ​​predicts the economic impact of the posted text data based on past online controversy data and effective word data.

[1207] Step 4: Based on the analysis results of the emotion engine and the AI's prediction results, edit or create publicly available information.

[1208] "Example of form 2"

[1209] Step 1: The user posts text data to social media or a website.

[1210] Step 2: The emotion engine analyzes the user's emotions from the text data.

[1211] Step 3: The AI ​​uses past online controversy data to determine the potential loss of the posted text data.

[1212] Step 4: Based on the analysis results of the emotion engine and the AI's loss estimation results, assess the risk of online backlash.

[1213] "Example of form 3"

[1214] Step 1: The user posts text data to social media or a website.

[1215] Step 2: The emotion engine analyzes the user's emotions from the text data.

[1216] Step 3: The AI ​​predicts the impact of the posted text data on the company's image, based on past data on online controversies and effective keywords.

[1217] Step 4: Based on the analysis results of the emotion engine and the AI's prediction results, disclose information to prevent damage to the company's image.

[1218] (Example 1)

[1219] Next, we will describe Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[1220] When companies publish information on social media and websites, they are required to predict economic impact by utilizing data on past online controversies and effective keywords. However, there is a lack of systems to effectively utilize this data, making it difficult to prevent damage to corporate image and economic losses. Furthermore, there is a lack of means to analyze user sentiment and revise or create information based on that analysis, resulting in ineffective information disclosure strategies for companies.

[1221] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1222] This invention includes a server that collects and stores past online controversy data, collects and stores effective word data, uses artificial intelligence to predict the economic effects of publishing information on social media and websites based on this data, analyzes emotions from text data posted by users, and edits or creates public information based on the results of the emotion analysis. This enables companies to predict economic effects by utilizing past online controversy cases and effective word data, and to publish information based on user emotions.

[1223] "Past online controversy data" refers to detailed information about past online controversies that have occurred on social media and the internet.

[1224] "Effective word data" refers to information about words and phrases that are considered effective in marketing and advertising.

[1225] "SNS" is an abbreviation for Social Networking Service, which refers to an online platform for users to share information and interact with each other.

[1226] A "website" is a collection of information published on the internet, and refers to an online page used by companies and individuals to disseminate information.

[1227] "Economic effect" refers to the impact that a particular action or event has on the economy, and specifically includes increases or decreases in sales and profits.

[1228] "Artificial intelligence" refers to technologies that use machine learning and data analysis to mimic human intelligence.

[1229] "User" refers to an individual or group that uses a system or service.

[1230] "Text data" refers to digital data that includes character information.

[1231] "Sentiment analysis" refers to a technology that identifies emotions from text data and evaluates the type and intensity of those emotions.

[1232] "Public information" refers to information that is made publicly available through social media and websites.

[1233] "Editing" refers to the act of correcting existing text and revising it to make it more appropriate.

[1234] "Creation" refers to the act of generating new text or information.

[1235] Modes for carrying out the invention

[1236] This invention is an AI-powered information disclosure support system for businesses that utilizes past data on online controversies and effective keywords to predict the economic impact of information published on social media and websites. It also includes a function to analyze user sentiment and revise or create published information based on the results.

[1237] Data collection and storage

[1238] The server collects past incidents of online controversies that have occurred on social media and the internet. Specifically, it uses web scraping tools (e.g., Beautiful Soup) to obtain data. The collected data is stored in a database (e.g., MySQL®, PostgreSQL®). Similarly, data on effective keywords is also collected and stored in the same database.

[1239] Data analysis

[1240] The server analyzes accumulated online controversy data using AI models (e.g., TensorFlow®, PyTorch®). Specifically, it uses natural language processing (NLP) techniques to extract the causes and patterns of online controversies. Similarly, data on effective words is also analyzed by AI models to identify which words are effective in which situations. These analysis results are stored in a database.

[1241] Predicting the economic impact

[1242] The server predicts the economic impact of information published on social media and websites, based on past data on online controversies and effective keywords. Specifically, it uses machine learning models to simulate the level of response the published information will generate. The prediction results are stored in a database.

[1243] Combination of emotional engines

[1244] The terminal sends text data posted by the user to an emotion engine (e.g., IBM Watson, Microsoft Azure's emotion analysis API). The server receives the analysis results returned by the emotion engine and stores them in a database. Based on the emotion analysis results, the server edits or creates public information. Specifically, it generates information that reinforces the positive image of a company's products and services based on text that shows positive emotions. The generated information is then published on social media and websites.

[1245] Specific example

[1246] For example, if a user posts the text "I really love the new product!", the device sends this text to the emotion engine. The server receives the emotion analysis result for "joy" from the emotion engine, and based on this analysis, generates positive information such as "many users like the new product" and publishes it on social media.

[1247] Example of a prompt

[1248] Examples of prompts to input into a generative AI model include the following:

[1249] "Based on past data on online controversies and effective keywords, predict the economic impact of information published on social media and websites. Also, analyze user sentiment and revise the published information based on the results."

[1250] The above describes the modes for carrying out the invention.

[1251] The flow of the specific processing in Example 1 will be explained using Figure 17.

[1252] Step 1:

[1253] The server collects past incidents of online controversies and online firestorms from social media and the internet. Specifically, it uses web scraping tools (e.g., Beautiful Soup) to obtain data. The input is text data from social media and websites, and the output is the collected data on these controversies. This data is stored in a database (e.g., MySQL®, PostgreSQL®).

[1254] Step 2:

[1255] The server collects data on words considered effective in marketing and advertising. This utilizes existing marketing databases and expert knowledge. The input is word data from the marketing database, and the output is the collected data on effective words. This data is also stored in the same database.

[1256] Step 3:

[1257] The server analyzes accumulated online controversy data using AI models (e.g., TensorFlow®, PyTorch®). Specifically, it uses natural language processing (NLP) techniques to extract the causes and patterns of online controversies. The input is online controversy data stored in a database, and the output is the analysis results of the causes and patterns of online controversies. These analysis results are stored in the database.

[1258] Step 4:

[1259] The server analyzes effective word data using an AI model to identify which words are effective in which situations. The input is effective word data stored in the database, and the output is the analysis results of effective words. These analysis results are also stored in the database.

[1260] Step 5:

[1261] The server predicts the economic impact of information published on social media and websites, based on past online controversy data and effective keyword data. Specifically, it uses a machine learning model to simulate the level of impact the published information will generate. The input is the analysis results of online controversy data and effective keyword data, and the output is the predicted economic impact. The prediction results are stored in a database.

[1262] Step 6:

[1263] The terminal sends the text data posted by the user to an emotion engine (e.g., IBM Watson, Microsoft Azure's emotion analysis API). The input is the text data posted by the user, and the output is the emotion analysis result from the emotion engine.

[1264] Step 7:

[1265] The server receives the analysis results returned from the emotion engine and stores them in the database. The input is the emotion analysis result, and the output is the emotion analysis result stored in the database.

[1266] Step 8:

[1267] The server edits and creates publicly available information based on the results of sentiment analysis. Specifically, it generates information that reinforces a positive image of a company's products and services based on text that expresses positive emotions. The input is the sentiment analysis results, and the output is the generated publicly available information. The generated information is published on social media and websites.

[1268] Step 9:

[1269] For example, if a user posts the text "I really love the new product!", the device sends this text to the emotion engine. The server receives the emotion analysis result for "joy" from the emotion engine, and based on this analysis, generates positive information such as "many users like the new product" and publishes it on social media.

[1270] (Application Example 1)

[1271] Next, we will describe Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[1272] When companies publish information on social media and their websites, it is important to predict economic impact by utilizing past data on online controversies and effective keywords. However, conventional systems lack the ability to analyze user emotions and generate optimal ad copy based on that analysis, making it difficult to improve corporate image and maximize economic impact.

[1273] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means. In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including an AI that predicts the economic effect when published on social media or websites based on this data, means for analyzing user emotions, means for generating ad copy based on the emotion analysis results, and means for predicting the economic effect of the generated ad copy. This enables companies to generate optimal ad copy based on user emotions and maximize economic effect.

[1274] "Past online controversy data" refers to data on negative reactions, critical comments, and posts that have occurred on social media or websites in the past.

[1275] "Effective word data" refers to data on keywords and phrases that have generated positive responses or high economic impact on social media and websites.

[1276] "AI that predicts economic impact" is artificial intelligence that uses past data to predict the extent of economic impact that information published on social media and websites will have.

[1277] "Methods for analyzing user emotions" refers to technologies that analyze the emotions expressed in text data posted by users and identify emotional states such as positive, negative, and neutral.

[1278] "A method for generating ad copy based on sentiment analysis results" refers to a technology that automatically generates optimal ad copy using the results of user sentiment analysis.

[1279] "Methods for predicting the economic impact of generated ad copy" refers to technologies that predict the extent of economic impact that generated ad copy will have.

[1280] The following system configuration will be described as an embodiment for carrying out this invention.

[1281] The server includes means of utilizing past online controversy data, means of utilizing effective word data, means including AI to predict the economic impact when published on social media and websites based on this data, means of analyzing user sentiment, means of generating ad copy based on sentiment analysis results, and means of predicting the economic impact of the generated ad copy.

[1282] Hardware and software to be used

[1283] Hardware: Servers, smartphones

[1284] Software: Python (registered trademark), pandas, scikit-learn (registered trademark), TextBlob, OpenAI API

[1285] Data processing and data calculation

[1286] 1. Data Loading: The server loads past online controversy data and effective keywords from a CSV file.

[1287] 2. Vectorization: The server uses TfidfVectorizer to vectorize the text data.

[1288] 3. Model Training: The server trains the economic impact prediction model using LinearRegression.

[1289] 4. Sentiment Analysis: The server uses TextBlob to analyze the sentiment of the user's posted text.

[1290] 5. Ad copy generation: The server uses the OpenAI API to generate sentiment-based ad copy.

[1291] 6. Economic Impact Prediction: The server uses a trained model to predict the economic impact of the generated ad copy.

[1292] Specific example

[1293] When a user posts the text "I want to create a new smartphone ad" using their smartphone, the server performs the following actions:

[1294] 1. The server uses TextBlob to perform sentiment analysis on the user's posted text and detects positive sentiment.

[1295] 2. The server uses the OpenAI API to generate the following prompt text and create the ad text.

[1296] Example prompt: "Create a positive advertisement text for the following product: I want to create an advertisement for the new smartphone."

[1297] 3. The server predicts the economic impact based on the generated ad copy.

[1298] In this way, companies can generate optimal ad copy based on user emotions and maximize economic impact.

[1299] The flow of a specific process in Application Example 1 will be explained using Figure 18.

[1300] Step 1:

[1301] The server reads historical online controversy data and effective keywords from a CSV file. It accepts a CSV file containing historical controversy data and effective keywords as input, and outputs this data in memory as a DataFrame. Specifically, it uses the pandas library to read the CSV file and convert it to a DataFrame.

[1302] Step 2:

[1303] The server uses TfidfVectorizer to vectorize text data. It takes a text string of past flame war data as input and obtains a vector representation of the text data as output. Specifically, it initializes TfidfVectorizer and vectorizes the text data using the fit_transform method.

[1304] Step 3:

[1305] The server trains an economic impact prediction model using LinearRegression. It takes vectorized text data and corresponding economic impact data as input and outputs a trained economic impact prediction model. Specifically, it initializes the LinearRegression model and trains it using the fit method.

[1306] Step 4:

[1307] The server uses TextBlob to analyze the sentiment of user-submitted text. It receives user-submitted text data as input and outputs a sentiment analysis result (positive, negative, or neutral). Specifically, it creates a TextBlob object and uses its `sentiment` property to analyze the sentiment.

[1308] Step 5:

[1309] The server uses the OpenAI API to generate sentiment-based ad copy. It receives sentiment analysis results and user-submitted text as input, and outputs the generated ad copy. Specifically, it generates a prompt based on the sentiment analysis results and sends a request to the OpenAI API to generate the ad copy.

[1310] Step 6:

[1311] The server predicts the economic impact of generated ad copy using a trained model. It receives generated ad copy as input and outputs a predicted economic impact value. Specifically, it vectorizes the generated ad copy and uses the trained model's predict method to predict the economic impact.

[1312] (Example 2)

[1313] Next, we will describe Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[1314] In today's information society, it is crucial for companies and individuals to properly assess the risk of online backlash and prevent economic losses when publishing information on social media and websites. However, conventional systems failed to adequately utilize past backlash data and user sentiment, resulting in insufficient assessment of backlash risk. Furthermore, there was a lack of means to proactively prevent damage to a company's image. As a result, risk management during information publication was difficult, and the likelihood of economic losses and damage to a company's image was high.

[1315] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1316] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic impact of information disclosure based on this data, means for analyzing user input text in real time and recognizing emotions, and means for evaluating the risk of online controversy based on the emotion analysis results. This makes it possible to quantify the risk of online controversy concretely before information disclosure, thereby preventing economic losses and damage to the company's image.

[1317] "Past online controversy data" refers to information about past online controversies that have occurred on social media or websites, and specifically includes data such as the content of posts, comments, and related financial losses.

[1318] "Effective word data" refers to data on keywords and phrases that are considered to have a positive impact when information is released.

[1319] "Artificial intelligence that predicts the economic impact of information disclosure" refers to machine learning models and algorithms that predict the economic impact of information disclosure based on past data.

[1320] "A means of analyzing user input text in real time and recognizing emotions" refers to technology that instantly analyzes text entered by a user and identifies the emotions contained within that text.

[1321] "Methods for evaluating the risk of online backlash based on sentiment analysis results" refers to technologies that use the results of sentiment analysis to assess the likelihood that input text will cause online backlash.

[1322] "Methods for determining the amount of loss" refers to techniques for identifying and quantifying the economic losses incurred during a social media firestorm, based on past data from similar incidents.

[1323] "Methods to prevent damage to a company's image in advance" refer to techniques for assessing the risk of a public outcry before information is released, and for preventing a decline in a company's reputation and image.

[1324] Modes for carrying out the invention

[1325] This invention is a system that utilizes past online controversy data to analyze user input text in real time, recognize emotions, and assess the risk of online controversy. A specific embodiment of this system is described below.

[1326] Server Processing

[1327] The server first collects data on past online controversies. Using SQL queries, it retrieves data such as posts, comments, and associated losses related to past controversies from the database. Next, it trains a model to predict losses based on the collected data. The data is preprocessed using the Pandas library in Python, and a regression model is built using Scikit-learn. The trained model is saved as a file using the Pickle library for later use.

[1328] Terminal processing

[1329] The device analyzes user-entered text in real time and calls an emotion analysis API to recognize emotions. Specifically, it receives user text input and uses the Google Cloud Natural Language API to analyze the emotion of the text. The analysis results are displayed to the user on the device. For example, if the emotion "anger" is recognized, the user will be notified accordingly.

[1330] User actions

[1331] Users input text into the system, and the system evaluates the risk of that text causing a social media firestorm. For example, users can input content for social media posts or blog articles. The system evaluates the risk of the input text causing a firestorm based on past firestorm data and sentiment analysis results. The evaluation results are displayed to the user, and the user can modify the content of their post based on those results.

[1332] Specific example

[1333] Examples of prompts to input into a generative AI model

[1334] "Design a system that uses past online controversy data to assess the risk of specific text going viral. Include features that analyze user-entered text in real time, recognize emotions using an emotion engine, and then assess the risk based on those results."

[1335] This system allows users to quantify and check the risk of controversy before posting, enabling them to publish information while taking risks into account.

[1336] The flow of the specific processing in Example 2 will be explained using Figure 19.

[1337] System program processing flow

[1338] Server Processing

[1339] Step 1: Collecting data on past online controversies.

[1340] The server collects data on past online controversies from the database. Specifically, it uses SQL queries to retrieve data such as post content, comments, and associated losses.

[1341] Input: Database connection information, SQL query

[1342] Output: Past online controversy dataset

[1343] Specific operation: The server executes the SQL query "SELECT FROM EnjoData WHERE Date BETWEEN '2018-01-01' AND '2023-01-01';" to retrieve the necessary data from the database.

[1344] Step 2: Data preprocessing

[1345] The server preprocesses the collected data. Specifically, it uses the Python® Pandas library to impute missing values ​​and remove outliers.

[1346] Input: Past online controversy dataset

[1347] Output: Preprocessed dataset

[1348] Specific operation: The server uses Pandas to perform operations such as "df.fillna(0)" and "df.dropna()" to clean up the data.

[1349] Step 3: Training the loss prediction model

[1350] The server trains a model to predict the amount of loss based on preprocessed data. Specifically, it builds a regression model using Scikit-learn®.

[1351] Input: Preprocessed dataset

[1352] Output: Trained loss prediction model

[1353] Specific operation: The server executes code such as "from sklearn.linear_model import LinearRegression" and "model = LinearRegression().fit(X_train, y_train)" to train the model.

[1354] Step 4: Save the model

[1355] The server saves the trained model as a file for later use. The Pickle library is used to save the model.

[1356] Input: Trained loss prediction model

[1357] Output: Saved model file

[1358] Specific operation: The server executes the code "import pickle" and "with open('loss_prediction_model.pkl', 'wb') as file: pickle.dump(model, file)" to save the model.

[1359] Terminal processing

[1360] Step 5: Receiving User Input

[1361] The device receives text entered by the user. For example, it retrieves input from a web form or a text box in a mobile app.

[1362] Input: User's text input

[1363] Output: Received text data

[1364] Specific action: The terminal responds when "the user enters 'This product is completely unusable' into a web form."

[1365] Step 6: Call the Sentiment Analysis API

[1366] The device sends the entered text to a sentiment analysis API to recognize its sentiment. The Google Cloud Natural Language API is used to analyze the sentiment of the text.

[1367] Input: Received text data

[1368] Output: Emotion analysis results

[1369] Specific operation: The terminal executes the code "response = client.analyze_sentiment(document=document)" and "sentiment = response.document_sentiment" to analyze the sentiment.

[1370] Step 7: Displaying the analysis results

[1371] The device displays the results of the emotion analysis to the user. For example, if the emotion "anger" is detected, the user will be notified accordingly.

[1372] Input: Sentiment analysis results

[1373] Output: Notification message to the user

[1374] Specific action: The terminal displays the message "This text contains angry emotions" to the user.

[1375] User actions

[1376] Step 8: Enter text

[1377] Users input text into the system, such as social media posts or blog articles.

[1378] Input: Text content

[1379] Output: Send text data to the system

[1380] Specific action: The user enters "This service is completely unusable" and sends it to the system.

[1381] Step 9: Check the risk of a public outcry.

[1382] The user reviews the system's assessment of online crisis risk. The system assesses the risk based on past online crisis data and sentiment analysis results.

[1383] Input: Evaluation results from the system

[1384] Output: Display of evaluation results to the user

[1385] Specific operation: The system displays an evaluation result stating, "This post has a high risk of causing a controversy," and the user confirms this.

[1386] In this way, users can quantify and confirm the risk of controversy before posting, enabling them to publish information while taking risks into account.

[1387] (Application Example 2)

[1388] Next, we will describe application example 2 of form example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 as the "terminal".

[1389] In today's information society, companies and individuals are required to use past online controversy data to assess the risk of online backlash and prevent economic losses when publishing information on social media and websites. However, conventional systems lack the means to analyze user sentiment and assess the risk of online backlash, making it difficult to warn users in advance about high-risk posts.

[1390] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1391] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence that predicts the economic effects when published on social media or websites based on this data, means for analyzing user sentiment, means for evaluating the risk of online controversy, and means for providing advance warnings about high-risk posts. This makes it possible to evaluate the risk of online controversy while considering user sentiment and to provide advance warnings about high-risk posts.

[1392] "Past online controversy data" refers to data on online controversies that occurred in the past due to information published on social media, websites, etc.

[1393] "Effective word data" refers to data on words and phrases that are likely to elicit a positive response when publishing information on social media or websites.

[1394] "Artificial intelligence" is a technology in which computer systems imitate human intelligence to learn, reason, and self-correct.

[1395] "Methods for analyzing user emotions" refers to technologies that identify emotions from text and comments posted by users and evaluate the type and intensity of those emotions.

[1396] "Methods for evaluating the risk of online backlash" refer to technologies that quantify and evaluate the likelihood of information disclosure causing a backlash, based on past online backlash data and user sentiment analysis results.

[1397] "A means of warning users in advance about high-risk posts" refers to a technology that issues advance warnings to users about posts that have been assessed as having a high risk of causing a social media firestorm.

[1398] The system for implementing this invention consists of three main elements: a server, a terminal, and a user. The server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic effects when content is published on social media or websites based on this data, means for analyzing user sentiment, means for evaluating the risk of online controversy, and means for warning users in advance about high-risk posts.

[1399] The server stores past online controversy data and effective word data in a database, and uses artificial intelligence to predict economic impact based on this data. Specifically, the server uses a text analysis engine (e.g., TextBlob) to analyze user posts and identify sentiment. Furthermore, it uses a machine learning model (e.g., LinearRegression) to assess the risk of online controversy based on past controversy data and sentiment analysis results.

[1400] When a user posts to social media or a website, the device receives a crisis risk assessment result from the server and displays a warning for high-risk posts. This allows users to reconsider their posts and reduce the risk of a social media firestorm.

[1401] For example, if a user tries to post "This product is completely useless!", the device sends this text to the server. The server uses a text analysis engine to analyze the sentiment and identify anger. Then, a machine learning model is used to assess the risk of the post causing a firestorm, and if a high risk is detected, a warning is displayed on the device.

[1402] Example of a prompt:

[1403] Please enter the user's comment: This product is completely useless!

[1404] This system allows users to check the risk of controversy before posting and avoid high-risk posts, thereby preventing financial losses for companies and individuals.

[1405] The flow of a specific process in Application Example 2 will be explained using Figure 20.

[1406] Step 1:

[1407] The user enters text to post to social media or a website using their device. The entered text is sent from the device to the server. The input data is the user's post content.

[1408] Step 2:

[1409] The server passes the received text to a text analysis engine (e.g., TextBlob) for sentiment analysis. The text analysis engine identifies the sentiment in the text and evaluates the type of sentiment (e.g., anger, joy, sadness) and its intensity. The input data is the user's posted content, and the output data is the type and intensity of the sentiment.

[1410] Step 3:

[1411] The server uses a machine learning model (e.g., LinearRegression) to assess the risk of online backlash based on the sentiment analysis results. The machine learning model takes historical backlash data and sentiment analysis results as input and quantifies the risk of online backlash. The input data consists of the type and intensity of the emotion, and historical backlash data, while the output data is a numerical value of the risk of online backlash.

[1412] Step 4:

[1413] The server sends the results of the crisis risk assessment to the terminal. Based on the received crisis risk value, the terminal displays a warning to the user if the risk is high. The input data is the crisis risk value, and the output data is the warning message.

[1414] Step 5:

[1415] The user reviews the warning message displayed on their device and reconsiders their post. If necessary, they revise the post and resubmit it to the server. The input data is the warning message, and the output data is the revised post.

[1416] This process allows users to check the risk of controversy before posting and avoid high-risk posts, thereby preventing financial losses for companies and individuals.

[1417] (Example 3)

[1418] Next, we will describe Embodiment 3 of Embodiment Example 3. 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".

[1419] Companies are required to proactively mitigate the risk of information going viral and damaging their corporate image. However, conventional systems struggle to consistently handle information generation, editing, risk assessment, and user sentiment analysis, often resulting in ineffective countermeasures. Furthermore, there is a lack of means to predict economic impact by utilizing past data on online controversies and effective keywords. As a result, companies face significant risks when disclosing information.

[1420] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 3 is realized by the following means. In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means for using a generative AI model that generates information based on this data, means for editing the generated information, means for evaluating the online controversy risk of the generated or edited information, means for prompting corrections when the online controversy risk is high, means for using an emotion engine that analyzes user emotions, means for publishing information based on the analysis results, and means for predicting the economic effects when the information is published on social media or a website based on this data. This makes it possible to evaluate the online controversy risk of information published by a company in advance and prompt corrections as necessary, thereby preventing damage to the company's image.

[1421] "Past online controversy data" refers to data on past online controversies caused by information published on social media, websites, etc.

[1422] "Effective word data" refers to data on words and phrases that are considered effective in eliciting positive responses when information is released.

[1423] A "generative AI model" is an artificial intelligence model that generates natural language text based on an input prompt.

[1424] "Editing" is the process of checking the grammatical and phrasing appropriateness of generated information and correcting it as needed.

[1425] "Assessing the risk of online backlash" is the process of analyzing the likelihood of online backlash occurring when generated or edited information is published, and then determining the risk.

[1426] "Methods to encourage correction" refer to methods of notifying users that information deemed to have a high risk of causing a public outcry needs to be corrected.

[1427] An "emotion engine" is a technology that analyzes text posted by users and identifies the emotions (joy, sadness, anger, etc.) contained within that text.

[1428] "Information disclosure" is the process of publishing information on social media, websites, etc., based on analysis results, in order to prevent damage to the company's image.

[1429] "Economic impact forecasting" is the process of predicting the economic impact that publicly available information will have on a company.

[1430] Modes for carrying out the invention

[1431] This invention is a system that prevents damage to a company's image by pre-assessing the risk of information going viral for companies and prompting corrections as necessary. A specific embodiment of this system is described below.

[1432] Hardware and software to be used

[1433] Hardware: Servers, terminals

[1434] Software: Generative AI models, emotion engines, database management systems (DBMS)

[1435] System Configuration

[1436] 1. The server has the means to store past online controversy data in a database and utilize it.

[1437] 2. The server has the means to store effective word data in a database and utilize it.

[1438] 3. The server has the means to generate information using a generative AI model.

[1439] 4. The server has the means to edit the generated information.

[1440] 5. The server has means to assess the risk of the generated or edited information becoming controversial.

[1441] 6. The server has mechanisms to prompt corrections when there is a high risk of a system failure.

[1442] 7. The server has the means to use an emotion engine to analyze the user's emotions.

[1443] 8. The server has the means to publish information based on the analysis results.

[1444] 9. The server has the means to predict the economic impact of publishing this data on social media and websites.

[1445] System operation

[1446] 1. The server uses a generation AI model to generate the information that the company intends to publish. For example, the prompt "Generate an introductory text describing the features of the new product" is input to the generation AI model, and an introductory text for the new product is generated.

[1447] 2. The server uses AI to proofread the generated information. Specifically, it checks grammar and the appropriateness of expressions, and makes corrections as needed.

[1448] 3. The server analyzes the edited information and assesses the risk of online backlash. Natural language processing (NLP) techniques are used to detect risk factors within the text.

[1449] 4. If a server is deemed to be at high risk of becoming unsafe, a notification prompting correction will be sent to the user's device. The user will receive the notification and correct the information as necessary.

[1450] 5. The server uses an emotion engine to analyze the user's emotions. For example, if a user posts "The price of the new product is too high," the server analyzes the text and detects the emotion of sadness.

[1451] 6. Based on the analysis results, the server will publish information to prevent damage to the company's image. For example, it will publish information regarding pricing and service improvements.

[1452] Specific example

[1453] For example, consider a case where a company uses this system when announcing a new product.

[1454] 1. The server inputs the prompt message "Generate an introductory text describing the features of the new product" into the AI ​​model.

[1455] 2. The AI ​​model generates an introductory text such as, "This new product features high performance and a user-friendly design."

[1456] 3. The AI ​​grammatically checks the introductory text generated by the server and corrects it to "It features high performance and a user-friendly design."

[1457] 4. The server analyzes the revised introductory text using NLP technology and determines that the expression "high performance" may implicitly criticize competitors' products.

[1458] 5. The server sends a notification to the device prompting a correction, suggesting that the expression "high performance" be changed to "multifunctional."

[1459] 6. The user receives the notification and corrects it to "It features a multi-functional and user-friendly design."

[1460] 7. The server uses an emotion engine to analyze the text posted by the user, "The price of the new product is too high," and detects the emotion of sadness.

[1461] 8. Based on the analysis results, the server publishes an "explanation regarding the price of the new product," stating that "this new product uses high-quality materials, and therefore the price has been set higher."

[1462] In this way, a system that proactively prevents damage to a company's image functions effectively. The flow of the specific processing in Example 3 will be explained using Figure 21.

[1463] Step 1:

[1464] The server inputs a prompt message into the AI ​​model that generates information. Specifically, it inputs the prompt message "Generate an introductory text describing the features of the new product." Based on this prompt message, the AI ​​model generates an introductory text such as "This new product is high-performance and features a user-friendly design." The input is the prompt message, and the output is the generated introductory text.

[1465] Step 2:

[1466] The server uses AI to proofread the generated information. Specifically, it checks the grammar of the generated introductory text and verifies the appropriateness of the expression. For example, it corrects the sentence "It features high performance and a user-friendly design" to "It features high performance and a user-friendly design." The input is the generated introductory text, and the output is the proofread introductory text.

[1467] Step 3:

[1468] The server analyzes the edited information and assesses the risk of online backlash. Specifically, it uses natural language processing (NLP) techniques to detect risk factors within the text. For example, it might determine that the phrase "high performance" implicitly criticizes a competitor's product. The input is the edited introductory text, and the output is the assessment result of the online backlash risk.

[1469] Step 4:

[1470] If a server is deemed to be at high risk of becoming a "firestorm," a notification prompting corrections will be sent to the user's device. Specifically, the notification will specify the parts that need correction and the reasons why. For example, it might suggest changing the expression "high performance" to "multifunctional." The input is the firestorm risk assessment result, and the output is the notification prompting corrections.

[1471] Step 5:

[1472] The user receives a notification and corrects the information as needed. Specifically, they revise the introductory text based on the notification. For example, they might change the sentence "It features high performance and an easy-to-use design" to "It features multi-functionality and an easy-to-use design." The input is a notification prompting correction, and the output is the corrected introductory text.

[1473] Step 6:

[1474] The server uses an emotion engine to analyze the user's emotions. Specifically, it analyzes the text posted by the user to identify the type of emotion (joy, sadness, anger, etc.). For example, if a user posts "The price of the new product is too high," the server analyzes that text and detects the emotion of sadness. The input is the user's posted text, and the output is the result of the emotion analysis.

[1475] Step 7:

[1476] The server publishes information based on the analysis results to prevent damage to the company's image. Specifically, it provides explanations regarding service improvements and pricing based on the analysis results. For example, it might publish an explanation regarding the pricing of a new product, stating, "This new product uses high-quality materials, and therefore the price is set higher." The input is the result of the sentiment analysis, and the output is the information that is published.

[1477] (Application Example 3)

[1478] Next, we will describe application example 3 of form example 3. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 as the "terminal".

[1479] In today's information society, the risk of information released by companies causing a public backlash is increasing, potentially leading to damage to the company's image and economic losses. Furthermore, information releases that disregard user sentiment can worsen the relationship between companies and their users. To address these issues, it is necessary to assess the risk of public backlash before information is released and to optimize information while considering user sentiment.

[1480] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means. In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic effect when the information is made public based on this data, means for analyzing user sentiment, and means for optimizing information based on the analysis results. This makes it possible to evaluate the risk of online controversy before information is made public and to optimize information while taking user sentiment into consideration.

[1481] "Past online controversy data" refers to data on online controversies that occurred due to information that was publicly released in the past.

[1482] "Effective word data" refers to data on words and phrases that are expected to elicit a positive response when used in information disclosure.

[1483] "Artificial intelligence that predicts the economic impact of disclosure" refers to artificial intelligence used to calculate the expected economic impact when information is made public.

[1484] "Means for analyzing user emotions" refers to technologies for analyzing the emotions expressed by users from sources such as text and audio.

[1485] "Means of optimizing information" refers to technologies for modifying and adjusting information to be made public in the most optimal form, based on analysis results and predictive data.

[1486] The system for implementing this invention consists of a server and a user terminal. The server includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence to predict the economic effects when published, means for analyzing user sentiment, and means for optimizing information based on the analysis results.

[1487] Hardware and software to be used

[1488] Hardware:

[1489] Server (cloud-based)

[1490] User terminals (smartphones, personal computers)

[1491] software:

[1492] Sentiment analysis engines (e.g., IBM Watson, Google Cloud Natural Language API)

[1493] AI models for assessing online controversy risk (e.g., OpenAI GPT-4®)

[1494] AI models for generating advertisements (e.g., OpenAI GPT-4®)

[1495] Data processing and data calculation

[1496] 1. Ad generation:

[1497] The user's device sends basic information about the advertising campaign (target audience, product information, etc.) to the server.

[1498] The server uses an AI model to generate ad copy.

[1499] 2. Assessment of the risk of online backlash:

[1500] The server inputs the generated ad copy into an AI model for assessing online backlash risk and performs a risk evaluation.

[1501] If the risk is high, the server will present a suggested solution to the user's terminal.

[1502] 3. Emotion analysis:

[1503] The server collects user sentiment data (e.g., social media posts, feedback, etc.).

[1504] The emotion analysis engine is used to analyze the user's emotions.

[1505] Based on the analysis results, the server optimizes the ad content.

[1506] Specific example

[1507] For example, when running an advertising campaign for a new smartphone, the user's device sends a prompt message to the server like the following:

[1508] Example of a prompt

[1509] Target audience: Young people (18-25 years old)

[1510] Product Information: New smartphone "XPhone" with high-performance camera, long battery life, and latest OS.

[1511] Ad copy tone: Casual and approachable

[1512] Based on this prompt, the server uses a generative AI model to generate ad copy, and a crisis risk assessment AI model evaluates the risk. If the risk is high, it suggests revisions, and finally, the sentiment analysis engine optimizes the ad content based on the user's emotions.

[1513] The above describes the embodiments for carrying out this invention.

[1514] The flow of the specific processing in Application Example 3 will be explained using Figure 22.

[1515] Step 1:

[1516] The user enters basic information about the advertising campaign (target audience, product information, ad copy tone, etc.) into their device. The entered information is then sent to the server.

[1517] Step 2:

[1518] The server generates ad copy using an ad generation AI model based on the received basic information. The generated ad copy is temporarily stored on the server.

[1519] Step 3:

[1520] The server inputs the generated ad copy into an AI model for assessing online backlash risk and performs a risk evaluation. The results of the risk evaluation are output, and if the risk is high, the server generates revised suggestions.

[1521] Step 4:

[1522] The server sends the risk assessment results and proposed revisions to the user's terminal. The user reviews the proposed revisions and modifies the ad copy as needed.

[1523] Step 5:

[1524] The server collects user sentiment data (e.g., social media posts, feedback, etc.). The collected sentiment data is then input into a sentiment analysis engine.

[1525] Step 6:

[1526] The server uses an emotion analysis engine to analyze the user's emotions. The analysis results are output and stored on the server.

[1527] Step 7:

[1528] The server optimizes the ad content based on the analysis results. The optimized ad copy is stored on the server.

[1529] Step 8:

[1530] The server sends the optimized ad copy to the user's device. The user reviews the optimized ad copy and publishes it as the final ad copy.

[1531] The above is the processing flow of this program.

[1532] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1533] The data generation model 58 is a form of so-called generative AI (Artificial Intelligence). One example of the data generation model 58 is ChatGPT (Internet Search).<URL: https: / / openai.com / blog / chatgpt> Examples of generative AI include the following. Data generation model 58 is

[1534] This is obtained by performing deep learning on a neural network. The data generation model 58 receives prompts containing instructions, as well as inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1535] Other examples of generative AI include Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) are some examples.

[1536] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[1537] [Third Embodiment]

[1538] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[1539] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1540] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1541] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[1542] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1543] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1544] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1545] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1546] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1547] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1548] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1549] Next, the identification process performed by the identification processing unit 290 of the data processing device 12 will be described.

[1550] "Example of form 1"

[1551] One embodiment of this system is an AI-powered information disclosure support system for businesses. This system utilizes past data on online controversies and effective keywords to predict the economic impact of information published on social media and websites. Specifically, it stores past online controversy data in a database, which is then analyzed by AI. Similarly, it stores data on effective keywords, and uses this data to edit and create publicly available information.

[1552] "Example of form 2"

[1553] Furthermore, another embodiment of the present invention includes a function to determine the amount of loss incurred when using past online controversy data during the editing process. Specifically, the amount of loss incurred during a controversy is estimated from past controversy data and used as reference information for editing. This makes it possible to quantify the risk of online controversy and disclose information that takes that risk into account.

[1554] "Example of form 3"

[1555] Furthermore, a further embodiment of the present invention includes a function to proactively prevent damage to a company's image. Specifically, before information created or edited by the AI ​​is made public, it evaluates whether the information contains a risk of causing a public outcry and prompts for revisions as necessary. This makes it possible to proactively prevent damage to a company's image.

[1556] The following describes the processing flow for each example of the form.

[1557] "Example of form 1"

[1558] Step 1: Accumulate past online controversy data and effective keyword data in a database.

[1559] Step 2: The AI ​​analyzes this data and edits or creates publicly available information.

[1560] Step 3: Based on the economic impact predicted by the AI, determine the optimal timing and content for information disclosure.

[1561] do.

[1562] "Example of form 2"

[1563] Step 1: Estimate the amount of losses incurred during a social media firestorm based on past data from similar incidents.

[1564] Step 2: Use the estimated loss amount as reference information for editing.

[1565] Step 3: Quantify the risk of online backlash and disclose information based on that risk.

[1566] "Example of form 3"

[1567] Step 1: Before the information created or edited by the AI ​​is published, evaluate whether it contains a risk of causing a public outcry.

[1568] Step 2: Based on the evaluation results, prompt for information correction as needed.

[1569] Step 3: Release the revised information to prevent damage to the company's image.

[1570] (Example 1)

[1571] Next, we will describe Embodiment 1 of Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1572] When companies publish information on social media or their websites, they face the challenge of predicting the risk of online backlash and the resulting economic impact. In particular, there is a need for methods to mitigate the risks of publicly released information and maximize its economic benefits by utilizing past online controversies and effective wording. Furthermore, preventing damage to the company's image is also a crucial issue.

[1573] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 1 is realized by the following means. In this invention, the server includes means for collecting and storing past online controversy data, means for collecting and storing effective word data, means for analyzing this data using natural language processing technology, means for predicting the economic effects of publicly available information based on the analysis results, and means for supporting the editing and creation of publicly available information. This makes it possible for companies to reduce the risk of online controversy related to publicly available information and maximize economic effects. It also makes it possible to prevent damage to the company's image in advance.

[1574] "Past online controversy data" refers to information about past online controversies that have occurred on social media or websites, and specifically includes post content, comments, retweet counts, and user reactions.

[1575] "Effective word data" refers to information about keywords and phrases that elicit positive responses on social media and websites, specifically including keywords and phrases that have received favorable user reactions.

[1576] "Natural language processing technology" refers to the technology that enables computers to understand, analyze, and generate human language, and specifically includes sentiment analysis and keyword extraction.

[1577] "Predicting economic impact" refers to evaluating how much traffic publicly available information will attract on social media and websites, or the degree of risk of it going viral or causing a controversy.

[1578] "Proofreading and creation support for publicly available information" refers to improving the quality of user-created information by reducing the risk of online backlash and suggesting effective wording.

[1579] "Preventing damage to a company's image" refers to measures taken to prevent a company's reputation from deteriorating due to information it makes public.

[1580] Modes for carrying out the invention

[1581] This invention is an information disclosure support system designed to reduce the risk of online backlash and maximize economic benefits when companies publish information on social media and their websites. The system collects and stores past online backlash data and effective keyword data, and uses natural language processing technology to analyze this data. Furthermore, it predicts the economic impact of the published information based on the analysis results and supports the editing and creation of such information.

[1582] Data collection and storage

[1583] The server collects past online controversy data from social media and websites. Specifically, it uses social media APIs to search for posts containing specific keywords and hashtags, and retrieves data such as comments and retweet counts related to the controversies. It also uses marketing research tools to collect data on effective keywords. For example, it uses Google Analytics and SEMrush to analyze how much traffic specific keywords have attracted. The collected data is stored in databases such as MySQL® and PostgreSQL®.

[1584] Data analysis

[1585] The server analyzes accumulated online controversy data using TensorFlow® and PyTorch®. Specifically, it uses natural language processing (NLP) techniques to analyze the sentiment of posts and identify factors that cause online controversies. Similarly, it analyzes data on effective words to identify keywords and phrases that elicit positive responses. For example, it uses models such as Word2Vec and BERT to evaluate the relevance of keywords.

[1586] Predicting the economic impact

[1587] Based on the analysis results, the server predicts the economic impact of the information to be released. Specifically, it evaluates how much traffic the information scheduled for release will attract and the risk of it causing a public outcry. This prediction is made by combining historical data with an AI model.

[1588] Editing and creating publicly available information

[1589] Users create public information and send it from their device to the server. For example, they might input a new product description or campaign announcement for a company. The server evaluates the risk of online backlash based on the submitted information. Specifically, it analyzes the input text using NLP technology and assesses the risk by comparing it with past online backlash data. The server suggests effective words and revises the information. For example, it might make specific suggestions such as, "This expression has a high risk of causing a backlash, so please change it to this expression." Users revise the information based on the server's suggestions and create the final public information. They can also send the revised information back to the server for final confirmation.

[1590] Specific example

[1591] Example 1: Assessment of the risk of online backlash

[1592] A user creates a product description for a new product and sends it to the server from their device. The server evaluates the risk of this description causing a backlash based on past data on product controversies. For example, if the server determines that a phrase like "This product is superior to competitors' products" increases the risk of backlash, it suggests changing it to something like "This product is well-received by many users."

[1593] Example 2: Suggestions for effective words

[1594] The user creates a campaign announcement and sends it from their device to the server. The server, based on a database of effective words, suggests keywords that will elicit a positive response to the announcement. For example, if keywords such as "limited," "free," and "special offer" are deemed effective, the server suggests adding these keywords to the announcement.

[1595] Example of a prompt

[1596] "I have created a product description for our new product. Please assess the risk of this description causing a backlash if it is published on social media, and suggest revisions as needed."

[1597] "I've created a campaign announcement. Please suggest effective words to elicit a positive response."

[1598] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1599] Step 1: Data collection and storage

[1600] The server collects past online controversy data from social media and websites. Specifically, it uses social media APIs to search for posts containing specific keywords and hashtags, and retrieves data such as comments and retweet counts related to the controversies. The input is specific keywords and hashtags, and the output is the collected controversy data. The server stores the collected data in databases such as MySQL® or PostgreSQL®.

[1601] Step 2: Effective word data collection

[1602] The server uses marketing research tools to collect data on effective keywords. For example, it uses Google Analytics or SEMrush to analyze how much traffic specific keywords attract. The input is the specific keyword, and the output is data on effective keywords. The server stores the collected data in a database.

[1603] Step 3: Data Analysis

[1604] The server analyzes accumulated online controversy data using TensorFlow® and PyTorch®. Specifically, it uses natural language processing (NLP) techniques to analyze the sentiment of posts and identify the factors causing online controversies. The input is accumulated online controversy data, and the output is the identification of the factors causing the controversies. Similarly, it analyzes effective word data to identify keywords and phrases that elicit positive responses. The input is effective word data, and the output is keywords and phrases that elicit positive responses.

[1605] Step 4: Predicting the Economic Impact

[1606] The server predicts the economic impact of publicly released information based on the analysis results. Specifically, it evaluates how much traffic the information scheduled for release will attract and the risk of it causing a public outcry. The input is the analysis results, and the output is the predicted economic impact.

[1607] Step 5: Create and submit public information

[1608] Users create public information and send it from their terminal to the server. For example, they might input a company's new product introduction or campaign announcement. The input is the public information created by the user, and the output is the public information sent to the server.

[1609] Step 6: Evaluation and proposals for publicly available information

[1610] The server assesses the risk of online backlash based on the submitted information. Specifically, it analyzes the input text using NLP technology and evaluates the risk by comparing it with past online backlash data. The input is publicly available information created by the user, and the output is the assessment result of the online backlash risk. The server suggests effective words and edits the information. For example, it makes specific suggestions such as, "This expression has a high risk of causing online backlash, so please change it to this expression." The input is the assessment result of the online backlash risk, and the output is the suggested revisions.

[1611] Step 7: Correction and final confirmation of publicly available information

[1612] Users revise the information based on the server's suggestions to create the final published information. They can also resubmit the revised information to the server for final confirmation. The input is the server's revision suggestions, and the output is the revised published information.

[1613] (Application Example 1)

[1614] Next, we will describe Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1615] When companies publish advertisements on social media and their websites, they are required to predict the economic impact by utilizing past data on online controversies and effective keywords. However, there are no systems that effectively utilize this data to support the editing and creation of ad copy. Furthermore, there is a lack of means to proactively prevent damage to a company's image by predicting online controversy risk scores and effective keyword scores. As a result, companies are forced to operate their advertising campaigns while carrying inherent risks.

[1616] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1617] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including AI to predict the economic effects when published on social media and websites based on this data, means for assisting in editing and creating ad copy, and means for predicting online controversy risk scores and effective word scores. This enables companies to reduce risks when publishing advertisements and create effective ad copy.

[1618] "Past online controversy data" refers to information about past online controversies that have occurred on social media or websites.

[1619] "Effective word data" refers to information about words and phrases that are highly effective in advertising and information dissemination.

[1620] "AI that predicts economic impact" refers to artificial intelligence that predicts the economic impact of information published on social media and websites based on past data.

[1621] "Means of supporting the editing and creation of ad copy" refers to functions that assist companies in revising or creating new ad copy based on past data.

[1622] A "crisis risk score" refers to an index that quantifies the likelihood of a particular advertisement or piece of information causing a social media firestorm or backlash on a website.

[1623] "Effective word score" refers to a metric that quantifies the likelihood that a particular ad copy or piece of information will be highly effective.

[1624] "Means of preventing damage to a company's image in advance" refers to functions that prevent a company's image from being negatively affected by information it makes public.

[1625] The system for implementing this invention includes an AI that utilizes past online controversy data and effective word data to predict the economic impact of advertising copy published on social media and websites. Furthermore, it has functions to support the editing and creation of advertising copy and predict online controversy risk scores and effective word scores.

[1626] System Configuration

[1627] Hardware:

[1628] server

[1629] smartphone

[1630] software:

[1631] Python (registered trademark)

[1632] pandas

[1633] scikit-learn(registered trademark)

[1634] joblib

[1635] Data handling

[1636] The server stores historical online controversy data and effective keywords in a database. This data is saved as a CSV file and read using pandas. The read data is then converted into a numerical vector using TfidfVectorizer.

[1637] Model training and saving

[1638] The server trains a LogisticRegression model based on data converted into numerical vectors using TfidfVectorizer. The trained model is saved using joblib. This lays the foundation for predicting crisis risk scores and effective word scores.

[1639] Evaluation of ad copy

[1640] When a user enters ad copy using their smartphone, the server uses a stored model to predict the ad's risk of backlash and its effectiveness score. This allows users to understand the risks and effectiveness of their ad copy in advance.

[1641] Specific example

[1642] For example, if a user enters the following ad text:

[1643] To celebrate the launch of our new product, we're holding a special sale!

[1644] The server predicts a risk of backlash score and an effective word score for this ad copy. The predicted results show a risk of backlash score of 0.2 and an effective word score of 0.8.

[1645] Example of a prompt

[1646] Please predict the risk of backlash and the effectiveness score for the following ad copy.

[1647] Advertisement: "To celebrate the launch of our new product, we're holding a special sale!"

[1648] In this way, companies can reduce the risks involved in advertising and create effective ad copy.

[1649] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1650] Step 1:

[1651] The server reads past online controversy data and effective keywords from a database. This data is stored as a CSV file and read using pandas. The input is a CSV file, and the output is a DataFrame. Specifically, the pandas.read_csv function is used to read the data.

[1652] Step 2:

[1653] The server converts the loaded data into a numerical vector using TfidfVectorizer. The input is a data frame, and the output is a numerical vector. Specifically, the TfidfVectorizer's fit_transform method is used to convert text data into a numerical vector.

[1654] Step 3:

[1655] The server trains a LogisticRegression model based on data that has been converted into numerical vectors. The input is a numerical vector, and the output is the trained model. Specifically, the model is trained using the LogisticRegression's fit method.

[1656] Step 4:

[1657] The server saves the trained model and TfidfVectorizer using joblib. The input is the trained model and TfidfVectorizer, and the output is the saved model file. Specifically, the model and vectorizer are saved using the joblib.dump function.

[1658] Step 5:

[1659] The user enters the ad text using their smartphone. The input is the text of the ad, and the output is a request to the server. Specifically, the user enters the ad text through a smartphone application and sends it to the server.

[1660] Step 6:

[1661] The server converts the received ad copy into a numerical vector using TfidfVectorizer. The input is the text of the ad copy, and the output is a numerical vector. Specifically, the server uses the transform method of the saved TfidfVectorizer to convert the ad copy into a numerical vector.

[1662] Step 7:

[1663] The server predicts the risk of backlash and the effectiveness of words based on the ad copy converted into numerical vectors. The input is a numerical vector, and the output is the risk of backlash and the effectiveness of words. Specifically, the scores are predicted using the predict_proba method of the stored LogisticRegression model.

[1664] Step 8:

[1665] The server returns a predicted crisis risk score and an effective word score to the user. The input is the predicted score, and the output is the response to the user. Specifically, the score is displayed through a smartphone application.

[1666] (Example 2)

[1667] Next, we will describe Example 2 of the Form Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1668] In today's information society, when companies and individuals publish information through social media and websites, they are required to use past online controversy data to predict risks and take appropriate action. However, conventional systems lack the means to effectively utilize past controversy data and concretely estimate the amount of loss, resulting in insufficient risk management. Furthermore, there is a lack of concrete means to prevent damage to a company's image in advance.

[1669] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1670] This invention provides a server that includes means for utilizing past online controversy data, means for utilizing effective word data, means including artificial intelligence that predicts the economic effects of information disclosure based on this data, means for analyzing past online controversy data and estimating the amount of loss, means for storing the estimated amount of loss in a database, and means for providing risk information, including the estimated amount of loss, when a user discloses information. This makes it possible to estimate specific amounts of loss based on past online controversy data, quantify the risk, and take appropriate action when disclosing information. It also provides specific means to prevent damage to a company's image in advance.

[1671] "Past online controversy data" refers to data on past online controversies that have occurred on online platforms such as social media and websites.

[1672] "Effective word data" refers to data on keywords and phrases that are effective when used when publishing information on social media or websites.

[1673] "Economic impact of information disclosure" refers to the expected economic effects and benefits that can be anticipated when information is made public on social media or a website.

[1674] Artificial intelligence is a system that uses machine learning and data analysis techniques to learn patterns from past data and make predictions and decisions about the future.

[1675] "Loss amount" refers to the estimated amount of economic loss based on past cases of online controversies.

[1676] A "database" is a system for efficiently storing, managing, and retrieving data.

[1677] "Risk information" refers to information about risks that should be considered when disclosing information, and specifically includes estimated loss amounts.

[1678] "Downturn in corporate image" refers to a decline in a company's reputation or brand value.

[1679] This invention is a system that utilizes past online controversy data to quantify the risks associated with information disclosure and enable appropriate responses. The following describes a specific embodiment of this system.

[1680] Server Role

[1681] The server retrieves past online fire incident data from a database, analyzes it, and estimates the amount of loss. Specifically, the server uses the following hardware and software:

[1682] Hardware: Server machine with a high-performance processor and sufficient memory.

[1683] Software: Relational database management systems (RDBMS) such as MySQL® and PostgreSQL®, and data analysis libraries such as pandas and scikit-learn® for Python®.

[1684] The server first executes an SQL query to retrieve past online fire incident data from the database. Next, it converts the retrieved data into a data frame using the Python® pandas library and performs preprocessing. Using the preprocessed data, it estimates the amount of loss using the scikit-learn® LinearRegression model. The estimated amount of loss is then saved back to the database.

[1685] Terminal role

[1686] The terminal retrieves risk information from the server when the user publishes information and displays it to the user. Specifically, the terminal uses the following hardware and software.

[1687] Hardware: Personal computers, tablets, smartphones, etc.

[1688] Software: Web browser, HTML, CSS, JavaScript

[1689] The device sends an HTTP GET request to the server and receives risk information in JSON format. The received risk information is then displayed on a web page using HTML and JavaScript.

[1690] User roles

[1691] Users will check risk information through their devices and adjust the content of the information they disclose. Specifically, users will check the estimated loss amount displayed on the web page and take action such as changing the disclosed content if the risk is high.

[1692] Specific example

[1693] Specific Example 1: Acquisition and Analysis of Online Crisis Data

[1694] The server retrieves past social media controversy data from a MySQL® database. For example, it uses an SQL query like the following:

[1695] SQL

[1696] SELECT FROM flame_data WHERE date >= '2020-01-01';

[1697] The acquired data is analyzed using the Python pandas library to estimate the amount of loss incurred during a social media firestorm. For example, the following code is used:

[1698] Python (registered trademark)

[1699] import pandas as pd

[1700] from sklearn.linear_model import LinearRegression

[1701] Loading data

[1702] data = pd.read_sql('SELECT FROM flame_data WHERE date >= "2020-01-01"', con=database_connection)

[1703] Estimation of loss amount

[1704] model = LinearRegression()

[1705] model.fit(data[['followers', 'negative_comments']], data['loss_amount'])

[1706] predicted_loss = model.predict(new_data[['followers', 'negative_comments']])

[1707] Example 2: Displaying risk information

[1708] When a user publishes information, the device retrieves risk information from the server and displays it on an HTML page. For example, the following code is used:

[1709] html

[1710] <!DOCTYPE html>

[1711]

[1712]

[1713] <title> Risk Information< / title>

[1714]

[1715]

[1716] <h1> Risk Information< / h1>

[1717] Estimated loss: \

[1718] <script>

[1719] fetch('https: / / example.com / api / risk_info')

[1720] .then(response => response.json())

[1721] .then(data => {

[1722] document.getElementById('loss_amount').textContent = data.loss_amount;

[1723] });

[1724] < / script>

[1725]

[1726]

[1727] Example of a prompt

[1728] Examples of prompts to input into a generative AI model include the following:

[1729] "Please generate Python® code that estimates the amount of financial loss incurred during a social media firestorm, using past social media firestorm data."

[1730] By using this prompt statement, the generated AI model can produce appropriate Python® code.

[1731] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1732] Step 1:

[1733] The server retrieves past online controversy data from the database. It accepts database connection information and an SQL query as input, and generates a dataframe containing the controversy data as output. Specifically, the server executes the SQL query to extract past controversy data. For example, it might use the following SQL query:

[1734] SQL

[1735] SELECT FROM flame_data WHERE date >= '2020-01-01';

[1736] This query retrieves online controversy data for the specified period.

[1737] Step 2:

[1738] The server analyzes the acquired data on online controversies. It accepts a dataframe as input and generates a pre-processed dataframe as output. Specifically, the server imputes missing values ​​and removes unnecessary columns. For example, it preprocesses the data using the Python® pandas library.

[1739] Step 3:

[1740] The server estimates the loss amount using preprocessed data. It takes a preprocessed dataframe as input and generates an estimated loss amount as output. Specifically, the server uses the scikit-learn® LinearRegression model to estimate the loss amount. For example, the model is trained and the loss amount is predicted as follows:

[1741] Python (registered trademark)

[1742] model = LinearRegression()

[1743] model.fit(data[['followers', 'negative_comments']], data['loss_amount'])

[1744] predicted_loss = model.predict(new_data[['followers', 'negative_comments']])

[1745] This process estimates the amount of loss.

[1746] Step 4:

[1747] The server stores the estimated loss amount in a database. It takes the estimated loss amount and database connection information as input and generates the loss amount stored in the database as output. Specifically, the server executes an SQL query to insert the estimation results into the database. For example, it might use the following SQL query:

[1748] SQL

[1749] INSERT INTO risk_info (date, predicted_loss) VALUES ('2023-10-01', 500000);

[1750] This query will save the estimated loss amount to the database.

[1751] Step 5:

[1752] The device retrieves risk information from the server when the user publishes information. It receives an HTTP GET request as input and generates JSON data containing the risk information as output. Specifically, the device sends an HTTP GET request to the server and receives the risk information. For example, it uses JavaScript code like the following:

[1753] JavaScript

[1754] fetch('https: / / example.com / api / risk_info')

[1755] .then(response => response.json())

[1756] .then(data => {

[1757] document.getElementById('loss_amount').textContent = data.predicted_loss;

[1758] });

[1759] This process retrieves risk information.

[1760] Step 6:

[1761] The device displays the acquired risk information to the user. It receives JSON data containing risk information as input and generates an HTML page displaying the risk information as output. Specifically, the device uses HTML and JavaScript to display the risk information on the web page. For example, it can be implemented as follows:

[1762] html

[1763] <!DOCTYPE html>

[1764]

[1765]

[1766] <title> Risk Information< / title>

[1767]

[1768]

[1769] <h1> Risk Information< / h1>

[1770] Estimated loss: \

[1771] <script>

[1772] fetch('https: / / example.com / api / risk_info')

[1773] .then(response => response.json())

[1774] .then(data => {

[1775] document.getElementById('loss_amount').textContent = data.predicted_loss;

[1776] });

[1777] < / script>

[1778]

[1779]

[1780] This process displays risk information to the user.

[1781] Step 7:

[1782] Users check risk information through their devices. They receive the displayed risk information as input and adjust the content of the information disclosure as output. Specifically, users check the estimated loss amount displayed on the webpage and take action, such as changing the disclosed content, if the risk is high.

[1783] (Application Example 2)

[1784] Next, we will describe application example 2 of form example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1785] In advertising campaigns, there is a need for a system that can quantify the risk of online backlash by utilizing past data and predict the amount of loss. However, current systems have the challenge of not being able to specifically quantify the risk of online backlash and provide information disclosure and concrete advice for risk reduction based on that risk.

[1786] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1787] In this invention, the server includes means for utilizing past online controversy data, means for utilizing effective word data, means including a generative AI model that predicts the economic effect when published on social media or websites based on this data, means for inputting the content of an advertising campaign and quantifying the risk of online controversy based on past controversy data, means for estimating the amount of loss based on the risk of online controversy, and means for providing specific advice to reduce the risk. This makes it possible to quantify the risk of online controversy in an advertising campaign and provide information disclosure and specific advice to reduce the risk based on that risk.

[1788] "Past online controversy data" refers to data on negative reactions, comments, and criticisms that have occurred on social media or websites in the past.

[1789] "Effective word data" refers to data on keywords and phrases that are considered effective in eliciting positive responses on social media and websites.

[1790] A "generative AI model" refers to an algorithm or system that uses machine learning or artificial intelligence technology to generate a specific output from input data.

[1791] "Methods for quantifying the risk of online backlash" refer to methods or systems that express the likelihood of a particular advertising campaign going viral in a numerical way, based on past online backlash data.

[1792] "Means for estimating the amount of loss" refers to methods or systems for calculating predicted economic losses based on the risk of online backlash.

[1793] "Means of providing specific advice" refers to methods and systems that propose specific actions and measures to reduce the risk of online backlash.

[1794] The system for implementing this invention operates through the coordinated efforts of a server, a terminal, and a user. A specific embodiment is described below.

[1795] Server Role

[1796] The server stores historical data on online controversies and effective keywords, and uses a generative AI model based on this data to predict economic impact. The server uses the following software and hardware:

[1797] Software: Python (registered trademark), Pandas, Scikit-learn (registered trademark)

[1798] Hardware: Servers equipped with high-performance processors and large amounts of memory.

[1799] The server reads past online controversy data and converts it into a data frame. Next, it sets features (campaign length, audience size, number of negative comments) and a target (amount of loss) and trains a linear regression model. This allows the system to quantify the risk of online controversy and predict the amount of loss when given the content of an advertising campaign.

[1800] Terminal role

[1801] The device provides an interface for users to input details about advertising campaigns. The device uses the following software and hardware:

[1802] Software: Web browsers, mobile applications

[1803] Hardware: Smartphones, tablets, PCs

[1804] When a user enters details of an advertising campaign through their device, that data is sent to a server. Based on the received data, the server quantifies the risk of a social media firestorm and predicts the amount of potential loss. The prediction results are sent back to the device and displayed to the user.

[1805] User roles

[1806] Users enter the details of their advertising campaign using their device and view predictions of potential backlash and potential losses. They also receive specific risk mitigation advice from the server and take appropriate measures.

[1807] Specific example

[1808] For example, if a user enters an ad campaign length of 30 days, an audience size of 100,000 people, and 50 negative comments, the server uses this data to quantify the risk of a social media firestorm and calculate the estimated loss. As a result, the estimated loss is displayed, and specific advice for risk reduction is provided.

[1809] Example of a prompt

[1810] "If an ad campaign lasts 30 days, has an audience size of 100,000, and receives 50 negative comments, what is the estimated loss?"

[1811] In this way, it becomes possible to quantify the risk of online backlash in advertising campaigns and provide specific advice on information disclosure and risk reduction based on that risk.

[1812] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1813] Step 1:

[1814] The user uses their device to enter details about the advertising campaign.

[1815] Input: ad campaign length, audience size, number of negative comments

[1816] Output: Input data is sent to the server.

[1817] Specific operation: The user enters details of the advertising campaign through a smartphone or computer interface and clicks the "Submit" button.

[1818] Step 2:

[1819] Based on the data received by the server, past online controversies are read.

[1820] Input: Detailed data of advertising campaigns submitted by users

[1821] Output: Past online controversy data is converted into a data frame.

[1822] Specific operation: The server uses the Python® Pandas library to read past online controversy data from a CSV file and convert it into a DataFrame.

[1823] Step 3:

[1824] The server sets the features and target, and then trains a linear regression model.

[1825] Input: Past online controversy data converted into a data frame

[1826] Output: Trained linear regression model

[1827] Specific operation: The server uses the Scikit-learn® library to set features (campaign length, audience size, number of negative comments) and target (loss amount), and then trains a linear regression model.

[1828] Step 4:

[1829] Based on detailed advertising campaign data received by the server from users, the risk of a social media firestorm is quantified, and the amount of potential loss is predicted.

[1830] Input: Detailed data on advertising campaigns received from users, a trained linear regression model.

[1831] Output: Crisis risk and estimated loss amount

[1832] Specific operation: The server takes detailed data of received advertising campaigns as features, uses a trained linear regression model to quantify the risk of backlash, and predicts the amount of loss.

[1833] Step 5:

[1834] The server sends back the predicted risk of a firestorm and the estimated loss amount to the terminal.

[1835] Input: Crisis risk and estimated loss amount

[1836] Output: The prediction results are displayed on the device.

[1837] Specific operation: The server sends the prediction results to the terminal in JSON format, and the terminal receives and displays them to the user.

[1838] Step 6:

[1839] The server generates specific advice to mitigate risk and sends it to the terminal.

[1840] Input: Crisis risk and estimated loss amount

[1841] Output: Specific advice for risk reduction

[1842] Specific operation: The server uses a generated AI model to generate specific advice to reduce the risk of online firestorms and sends it to the terminal. The terminal receives this and displays it to the user.

[1843] In this way, it becomes possible to quantify the risk of online backlash in advertising campaigns and provide specific advice on information disclosure and risk reduction based on that risk.

[1844] (Example 3)

[1845] Next, we will describe Embodiment 3 of Embodiment Example 3. 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."

[1846] In today's information society, companies face an increasing risk of online backlash when publishing information through social media and websites. This risk can damage a company's image and lead to financial losses, making it crucial to assess the risk beforehand and modify information as needed. However, current systems fail to adequately assess this risk and modify information, posing a significant challenge for businesses.

[1847] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 3 is realized by the following means.

[1848] This invention includes a server that utilizes past online controversy data, a server that utilizes effective word data, a server that uses a generative AI model to generate information based on this data, a server that evaluates the risk of online controversy before the generated information is made public, a server that prompts correction of the information if there is a risk of online controversy, and a server that predicts the economic effects of publishing the information on social media or a website based on these means. This enables companies to evaluate the risk of online controversy before disclosing information and correct the information as needed, thereby preventing damage to their corporate image.

[1849] "Past online controversy data" refers to data on past online controversies caused by information published on social media, websites, etc.

[1850] "Effective word data" refers to data on words and phrases that are considered effective in eliciting positive responses when publishing information on social media or websites.

[1851] A "generative AI model" is an artificial intelligence model that generates information based on prompt text entered by the user.

[1852] "Methods for assessing the risk of online backlash" refer to methods for analyzing and evaluating the likelihood of online backlash when generated information is made public.

[1853] "Methods to encourage information correction" refer to methods used to prompt users to correct information that has been assessed as having a risk of causing a public outcry.

[1854] "Methods for predicting economic effects" refer to methods for predicting the economic impact that information published on social media and websites has on companies.

[1855] Modes for carrying out the invention

[1856] This invention is a system that prevents damage to a company's image by pre-assessing the risk of online backlash when a company publishes information on social media or its website, and by modifying the information as needed. This system consists of three main elements: a server, a terminal, and a user.

[1857] 1. Program generation

[1858] The server generates a program to proactively prevent damage to the company's image. This program uses a generative AI model to create or edit information and has the function of evaluating the risk of online backlash before the information is made public. Specifically, it uses OpenAI's GPT-4 (registered trademark) as the generative AI model.

[1859] 2. Program Processing

[1860] The server uses a generative AI model (e.g., OpenAI's GPT-4®) to generate information based on the prompt text entered by the user. The generated information is then sent to a crisis risk assessment module. This module uses natural language processing (NLP) techniques to analyze the content of the information and assess whether there is a crisis risk. Specifically, it uses a risk assessment algorithm (e.g., a BERT-based classification model) to determine whether the information contains a crisis risk.

[1861] The terminal sends the prompt text entered by the user to the server and receives generated information and risk assessment results from the server. The user reviews the risk assessment results on the terminal and corrects the information as needed.

[1862] 3. Specific Examples and Examples of Prompt Statements

[1863] As a concrete example, consider a case where a user is creating a press release for a new product. The user enters the following prompt into the terminal:

[1864] Example of a prompt:

[1865] "Please create a press release for our new product. The product name is 'EcoSmart,' and please emphasize its environmentally friendly features."

[1866] The server inputs this prompt into the AI ​​model that generates the following information:

[1867] Examples of generated information:

[1868] "EcoSmart is a new product that utilizes the latest environmental technologies. It is energy-efficient and uses recyclable materials."

[1869] Next, the server sends the generated information to a crisis risk assessment module for risk assessment. If the assessment result is determined to be "risky," the server sends a message to the user prompting them to make corrections. The user can then review the proposed corrections on their device and modify the information as needed to prevent damage to the company's image.

[1870] This system allows companies to assess the risk of public backlash before disclosing information and, if necessary, modify the information to prevent damage to their corporate image. The flow of the specific processing in Example 3 will be explained using Figure 15.

[1871] Step 1:

[1872] The user enters a prompt message.

[1873] The user enters a prompt message into the terminal's input field. For example, they might enter, "Please create a press release for a new product. The product name is 'EcoSmart,' and please emphasize its environmentally friendly features." The entered prompt message is temporarily stored in the terminal's memory.

[1874] Step 2:

[1875] The terminal sends a prompt message to the server.

[1876] The terminal converts the prompt text entered by the user into the appropriate format and sends a request to the server's API endpoint. Specifically, it sends the prompt text to the server using an HTTP POST request. The input is the prompt text, and the output is the request sent to the server.

[1877] Step 3:

[1878] The server generates information using an AI model.

[1879] The server inputs the received prompt message into a generating AI model (e.g., OpenAI's GPT-4®) and generates information. The generating AI model analyzes the prompt message and generates information based on the appropriate context. The input is the prompt message, and the output is the generated information. For example, information such as "EcoSmart is a new product that utilizes the latest environmental technologies. It is energy-efficient and uses recyclable materials." might be generated.

[1880] Step 4:

[1881] The server sends the generated information to the fire risk assessment module.

[1882] The server sends the generated information to a crisis risk assessment module for risk assessment. This module uses natural language processing (NLP) techniques to analyze the content of the information and assess whether there is a risk of crisis. Specifically, it uses a risk assessment algorithm (e.g., a BERT-based classification model) to determine whether the information contains a risk of crisis. The input is the generated information, and the output is the risk assessment result.

[1883] Step 5:

[1884] The server generates the risk assessment results and sends them to the terminal.

[1885] The server receives the assessment results from the fire risk assessment module and sends them to the terminal. The assessment results include judgments such as "no risk" or "risk present." The input is the risk assessment results, and the output is the transmission to the terminal.

[1886] Step 6:

[1887] The device displays the risk assessment results to the user.

[1888] The terminal displays the risk assessment results received from the server to th...

Claims

[Claim 1] A means of collecting and accumulating past online controversy data, A means of collecting and accumulating data on effective words that are considered effective in eliciting positive responses on social media or websites, A means for identifying factors causing online outrage by performing sentiment analysis on the content of online outrage data using natural language processing technology, and for identifying keywords that elicit positive responses on social media or websites by evaluating the relevance of effective words, A means of generating ad copy using a generative AI model based on information about the advertising campaign, including the target audience, product information, and the tone of the ad copy. A method for predicting the risk of the aforementioned advertisement becoming controversial if it is published based on the analysis results, A means for evaluating the aforementioned advertisement copy in comparison with the aforementioned past online controversy data, and if the risk is high, sending a revised proposal to the user terminal that changes specific expressions in the advertisement copy to different expressions and adds the aforementioned keywords, A means for predicting the amount of loss using a linear regression model trained on past online controversy data, based on input of the content of the aforementioned advertising campaign, Means for providing risk information, including the amount of loss, to the user terminal, A system that includes this.

Images