system

Abstract

We provide the system. [Solution] A means of obtaining text entered by the user, A means for transmitting the acquired text to an analysis device, A means of analyzing the content of a document using an analysis device, A means of comparing past analysis databases with analysis results, A means of evaluating the risk of sending a document based on the comparison results, A means of displaying the risk assessment results to the user and confirming submission, A system that includes this.

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, which is performed by at least one processor, the method including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, 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] In recent years, with the spread of social networking services (SNS), the opportunities for individuals to easily transmit information have increased. Along with this, the number of posts that unintentionally cause an uproar has also increased, which has become a serious problem for individuals and companies. In particular, emotional or radical remarks are widely spread in a short period of time and often lead to unexpected uproars. Therefore, before a user posts on SNS, there is a need for means to evaluate the risk of the post content and suppress the transmission if necessary. Conventional manual checks are time-consuming and laborious, and depend on subjective judgments, so an objective and rapid risk assessment is required.

Means for Solving the Problems

[0005] To solve the above-mentioned problems, the present invention provides the following means: a system including means for acquiring text entered by a user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for comparing the analysis results with a past analysis database, means for evaluating the risk of transmitting the text based on the comparison results, and means for displaying the risk evaluation results to the user and confirming the transmission. This system allows users to automatically and quickly perform a risk evaluation before posting to social media and decide whether or not to transmit based on the results. This significantly reduces the risk of unintentionally causing a social media firestorm and helps maintain the social credibility of users and companies. Furthermore, if the analysis results are judged to be high risk, the system further includes means for prompting the user to confirm the transmission, thereby suppressing high-risk posts. In addition, if the risk evaluation results are judged to be low risk or medium risk, the system includes means for transmitting the text without user confirmation, thereby supporting rapid information dissemination.

[0006] A "user" refers to an individual or group that uses this system to post to social networking services (SNS).

[0007] "Device" refers to electronic devices such as computers, smartphones, and tablets used by users.

[0008] A "server" refers to a computer system that receives posted content sent from a terminal and performs processing such as analysis and comparison.

[0009] "Acquisition method" refers to the function by which the terminal detects text entered by the user and sends it to the server.

[0010] "Analysis device" refers to hardware and software used to automatically analyze the content of acquired text.

[0011] "Analysis means" refers to the process of using an analysis device to determine the content of a text and evaluate its characteristics and emotions.

[0012] A "database" refers to a collection of information that contains data on past controversies and related incidents.

[0013] "Comparison method" refers to the process of evaluating risk by comparing analysis results with past databases.

[0014] "Risk assessment tool" refers to a function that calculates the risk of a post going viral based on comparison results.

[0015] "Display means" refers to the interface and functions that visually present risk assessment results to the user.

[0016] "Transmission confirmation means" refers to a function that prompts the user to make a final confirmation of transmission when the risk assessment results indicate a high risk.

[0017] "Transmission method" refers to a function that automatically sends posts to social media without user confirmation in low-risk or medium-risk situations. [Brief explanation of the drawing]

[0018] [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] It 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] It shows an emotion map to which a plurality of emotions are mapped. [Figure 10] It shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Embodiments for Carrying Out the Invention

[0019] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described according to the accompanying drawings.

[0020] First, the language used in the following description will be explained.

[0021] In the following embodiments, a processor with a reference number (hereinafter simply referred to as "processor") may be one arithmetic unit or a combination of a plurality of arithmetic units. Also, the processor may be one type of arithmetic unit or a combination of a plurality of types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0022] 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.

[0023] 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.

[0024] 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).

[0025] 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."

[0026] [First Embodiment]

[0027] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0028] 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.

[0029] 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).

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] 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.

[0036] 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.

[0037] 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.

[0038] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0039] The present invention is a system that automatically analyzes the content of posts made by users on social media and assesses the risk of causing online outrage. The system includes means for acquiring text entered by the user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for comparing the analysis results with a past analysis database, means for evaluating the risk of sending the text based on the comparison results, and means for displaying the risk assessment results to the user and confirming the transmission.

[0040] System Overview

[0041] 1. User actions:

[0042] When a user enters text on the SNS post creation screen and presses the send button, the device acquires that input.

[0043] A concrete example is when a user types the sentence "This new product is completely useless" and presses the submit button.

[0044] 2. Retrieving and sending the posted content:

[0045] The terminal retrieves the text entered by the user and sends it to the server.

[0046] A concrete example of this would be when a device sends a message to a server stating, "This new product is completely unusable."

[0047] 3. Content analysis:

[0048] The server receives text from the terminal and passes it to the analysis device, which then analyzes its content.

[0049] The analysis device evaluates the tone, emotion, and keywords of the text.

[0050] For example, we can analyze how the phrase "completely useless" is extremely negative.

[0051] 4. Comparison and Risk Assessment:

[0052] The analysis system on the server compares the analysis results with past incidents of online firestorms.

[0053] The comparison method involves evaluating the similarity to past online controversies stored in the database.

[0054] For example, the phrase "completely useless" is found to have many points of agreement with past negative posts.

[0055] 5. Displaying results and confirming submission:

[0056] The server assesses the risk based on the comparison results and sends the results back to the terminal.

[0057] The terminal receives the risk assessment results from the server and displays them to the user. If the risk assessment indicates high risk, the terminal prompts the user to confirm submission.

[0058] A concrete example would be displaying a screen that asks the user, "This post has a high risk of causing a backlash. Do you really want to post it?"

[0059] The post will only be sent if the user confirms and answers "yes". If the user answers "no", the post will be canceled.

[0060] Specific example

[0061] The user enters the sentence, "This new product is completely useless."

[0062] The terminal sends that input to the server.

[0063] The server analyzes the sentence and determines that the phrase "completely useless" is extremely negative.

[0064] The server compares the data with a database of past online controversies and assesses it as high-risk due to numerous similarities.

[0065] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[0066] The post will only be sent if the user answers "yes".

[0067] Precautions

[0068] This system automatically performs a risk assessment before a user posts on social media and decides whether or not to post based on the results. This process significantly reduces the risk of unintentionally causing a social media firestorm. Furthermore, the system only requests user confirmation if it determines the post is high-risk; otherwise, it supports rapid information dissemination.

[0069] The following describes the processing flow.

[0070] Step 1:

[0071] The user enters the content of their post on the SNS post creation screen and presses the send button. This action retrieves the post content to the device. For example, they might enter the sentence, "This new product is completely useless."

[0072] Step 2:

[0073] The device detects when the user presses the send button and sends the entered content to the server. At this time, the string data of the posted content is sent as an API request.

[0074] Step 3:

[0075] The server receives the posted content sent from the terminal. To pass the received data to the analysis device, it first saves the posted content.

[0076] Step 4:

[0077] The analysis system on the server analyzes the received posts. During this process, it evaluates the tone of the text, keywords, and emotional tone. For example, the negative connotation of a phrase like "completely useless" is assessed.

[0078] Step 5:

[0079] The server compares the analysis results obtained by the analysis device with a database of past online controversies. It evaluates the similarity to the data in the database and determines the likelihood that the posted content will cause a controversy.

[0080] Step 6:

[0081] The server assesses the risk of the posted content based on the comparison results. For example, if the phrase "completely useless" matches many past controversial incidents, it will be assessed as high risk.

[0082] Step 7:

[0083] The server returns the risk assessment results to the terminal. The risk assessment results will be one of the following: "high risk," "medium risk," or "low risk."

[0084] Step 8:

[0085] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0086] Step 9:

[0087] Users respond to the displayed message with either "yes" or "no." If they answer "no," the post is canceled; if they answer "yes," the post is sent only.

[0088] Step 10:

[0089] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0090] By following these steps, users can be prevented from unconsciously making posts that carry a high risk of causing controversy.

[0091] (Example 1)

[0092] Next, we will describe 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."

[0093] There is a problem in that there are few means to detect posts that have a high risk of causing social backlash in advance and notify users. As a result, unintended backlash can occur, damaging the reputation of individuals and companies.

[0094] 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.

[0095] In this invention, the server includes means for acquiring text entered by the user, means for transmitting the acquired text to an information processing device, means for analyzing the content of the text using the information processing device, means for comparing the analysis results with past evaluation information, means for evaluating the risk of transmitting the text based on the comparison results, and means for displaying the risk evaluation results to the user and confirming the transmission. This makes it possible to prevent unintended online controversies by the user.

[0096] A "user" refers to a person who uses the system to post to social networking services (SNS).

[0097] "Text" refers to the text data that users input when posting on social media.

[0098] "Means of acquisition" refers to devices or programs used to collect text entered by the user.

[0099] "Information processing device" refers to a system for analyzing the content of text, including analysis devices such as servers.

[0100] "Means of transmission" refers to methods or devices for transferring acquired text to an information processing device.

[0101] "Means of analysis" refers to programs and algorithms used by information processing devices to analyze and evaluate the content of text.

[0102] "Past evaluation information" refers to analytical data stored in databases of past controversial incidents and posts.

[0103] "Means of comparison" refers to methods or programs for comparing analysis results with past evaluation information to assess similarities and risks.

[0104] "Means of risk assessment" refers to criteria or algorithms used to determine whether a piece of writing has the potential to cause a controversy or backlash.

[0105] "Risk assessment results" refer to the determination, based on the evaluation process, whether or not a post poses a risk of causing a social media firestorm.

[0106] "Means of display" refers to devices or programs that visually present risk assessment results to the user.

[0107] "Means for confirming submission" refers to methods or devices for notifying users of the risk assessment results and requesting final confirmation of the submission.

[0108] This invention is a system that analyzes the content of posts made by users on social media in advance and assesses the risk of causing online controversy. The system is configured as follows:

[0109] First, the user enters text on the social media post creation screen and presses the send button. The device retrieves this entered text and sends it to the server. A concrete example of this would be when a user enters the sentence "This new product is completely useless" and presses the send button.

[0110] Next, the server passes the text received from the terminal to the analysis device, which then analyzes its content. The analysis device evaluates the tone, emotion, and keywords of the text. For example, it evaluates the phrase "completely useless" as being very negative. The analysis device uses a high-performance server (for example, one equipped with an NVIDIA GPU), and natural language processing (NLP) algorithms such as Google's TENSORFLOW® and Facebook's PyTorch are used for analysis.

[0111] Subsequently, the analysis system on the server compares the analysis results with a database of past online controversies. The comparison method evaluates the similarity to past controversies stored in the database. For example, it is confirmed that the phrase "completely useless" has many points of agreement with past negative posts.

[0112] The server assesses the risk based on the comparison results and sends the results back to the terminal. The terminal receives the risk assessment results from the server and displays them to the user. If the risk assessment is high, the user is prompted to confirm submission. For example, a screen asking the user, "This post has a high risk of causing a controversy. Do you really want to submit it?" is displayed. The post is submitted only if the user confirms and answers "yes". If the user answers "no", the post is canceled.

[0113] In this way, the present invention automatically performs a risk assessment before a user makes a social media post and decides whether or not to post based on the results. This process makes it possible to significantly reduce the risk of unintentionally causing a social media firestorm. Furthermore, the system only requests user confirmation if it is determined to be high risk, and in other cases supports rapid information dissemination.

[0114] Specific example

[0115] When a user types "This new product is completely useless" on the posting screen and presses the send button, the device retrieves the text and sends it to the server. The server receives the text and passes it to the analysis device. The analysis device evaluates the phrase "completely useless" as extremely negative. The server compares the analysis results with a database of past online controversies and evaluates it as high risk due to many similarities. The device displays to the user, "This post has a high risk of causing a controversy. Do you really want to post it?" If the user answers "yes," the post is sent; if they answer "no," the post is canceled.

[0116] Example of a prompt

[0117] "I'm developing a text analysis system to assess the risk of social media posts about a new product going viral. This system requires prompts like the following:

[0118] For example, if a user enters the sentence, "This new product is completely useless," please explain how you would assess the risk of a backlash.

[0119] This prompt allows the generating AI model to understand the operation of a specific system and generate an appropriate risk assessment process.

[0120] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0121] Step 1:

[0122] The user enters text on the social media post creation screen and presses the send button.

[0123] Input: User's input text (e.g., "This new product is completely unusable")

[0124] Output: The user's input text is sent to the terminal.

[0125] Specific action: The user enters text on the posting screen and presses the submit button.

[0126] Step 2:

[0127] The terminal retrieves the user's input text and sends it to the server.

[0128] Input: User input text sent to the terminal

[0129] Output: User input text sent to the server

[0130] Specific operation: The terminal retrieves the entered text and sends it to the server. For example, the terminal sends the text "This new product is completely useless" to the server.

[0131] Step 3:

[0132] The server receives text from the terminal and passes it to the analysis device, which then analyzes its contents.

[0133] Input: User input text sent to the server

[0134] Output: User input text passed to the analysis device

[0135] Specific operation: The server passes the received text to the analysis device. The analysis device receives the text "This new product is completely unusable."

[0136] Step 4:

[0137] The analysis device evaluates the tone, emotion, and keywords of the text.

[0138] Input: User input text provided to the analysis device

[0139] Output: Text evaluation data (e.g., negative evaluation)

[0140] Specific operation: The analysis device uses natural language processing (NLP) techniques to analyze the text. For example, the phrase "completely useless" is evaluated as highly negative.

[0141] Step 5:

[0142] The server compares the analysis results with a database of past online controversies.

[0143] Input: Text evaluation data

[0144] Output: Risk assessment data

[0145] Specific operation: The server compares the analysis results with a database of past online controversies and evaluates similarities. For example, if the phrase "completely useless" is determined to have many points of agreement with past negative posts, it will be considered to have similarities.

[0146] Step 6:

[0147] The server sends the risk assessment results back to the terminal.

[0148] Input: Risk assessment data

[0149] Output: Risk assessment results sent to the terminal

[0150] Specific operation: The server generates a risk assessment result and sends it to the terminal. For example, the assessment result sent might say, "This post has a high risk of causing a controversy."

[0151] Step 7:

[0152] The terminal displays the risk assessment results to the user and confirms the transmission.

[0153] Input: Risk assessment results sent to the terminal

[0154] Output: Display and confirmation results for the user

[0155] Specific operation: The device displays the evaluation result to the user and asks the user to confirm submission. For example, it might ask the user, "This post has a high risk of causing controversy. Do you really want to submit it?" The post is submitted only if the user answers "yes," and canceled if they answer "no."

[0156] (Application Example 1)

[0157] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0158] In electronic payment services, it is essential to properly assess the risk of fraudulent use. However, conventional systems often fail to adequately assess the risk of fraudulent use that may occur when users conduct transactions, potentially leading to fraudulent transactions. This invention aims to solve these problems and provide a system that enables users to conduct electronic transactions securely.

[0159] 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.

[0160] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to an analysis device, means for the analysis device to analyze the content of the text, means for comparing the analysis results with a past analysis database, means for evaluating the risk of transmitting the text based on the comparison results, means for displaying the risk evaluation results to the user and confirming the transmission, means for acquiring the details of the electronic transaction and transmitting them to the analysis device, means for the analysis device to evaluate the risk of fraudulent use of the transaction, and means for notifying the user of the evaluation results and requesting confirmation. This enables the user to conduct electronic transactions securely.

[0161] A "user" is an individual or group that uses the system.

[0162] "Text" refers to the text data that a user enters into the system.

[0163] An "analysis device" is a combination of hardware and software used to analyze input text and details of electronic transactions.

[0164] A "past analysis database" is a data storage system that stores previously analyzed data and the results of those analyses.

[0165] "Transmission risk" is an assessment of the potential risks that the entered text or electronic transaction may pose.

[0166] "Risk assessment results" refer to the specific evaluation value or judgment result of transmission risk calculated by the analysis device.

[0167] "Electronic transactions" refer to transactions such as buying and selling property or services, remittances, and contracts conducted using telecommunications means.

[0168] "Fraudulent use risk" refers to the assessment of the risk of unauthorized or fraudulent activity in electronic transactions.

[0169] "Notification" refers to a means of communicating information to users regarding the results of a risk assessment.

[0170] The present invention is a system that automatically analyzes transaction details and assesses the risk of fraudulent use when a user conducts an electronic transaction. This system includes means for acquiring transaction details entered by the user, means for transmitting the acquired transaction details to an analysis device, means for analyzing the transaction details using the analysis device, means for comparing the analysis results with a past analysis database, means for evaluating the risk of fraudulent use of the transaction based on the comparison results, and means for displaying the risk assessment results to the user and confirming the transaction. Specific embodiments are described below.

[0171] The terminal retrieves transaction details and text entered by the user and sends them to the server. The server passes the retrieved transaction details and text to an analysis device, which then analyzes them. The analysis device uses data analysis software (e.g., Python or Scikit-learn) to evaluate the tone and keywords of the transaction details and text. Once the tone and keywords are evaluated, the analysis device compares the results to a database of past analyses. This database stores previously analyzed data, including past incidents of online controversies and instances of misuse.

[0172] The analysis device evaluates the risk of fraudulent use of a transaction based on the comparison results. The server receives these evaluation results and sends them back to the terminal. The terminal displays the risk evaluation results to the user, and if it is determined to be high risk, it prompts the user to confirm the transaction. For example, a message such as "This transaction is high risk. Do you really want to proceed?" is displayed. The transaction is executed only if the user confirms.

[0173] As a concrete example, consider a case where a user makes a transaction to "transfer 500,000 yen to purchase a car." The terminal sends this transaction details to the server, which then passes them to an analysis device. The analysis device evaluates the risk of fraudulent use of this transaction, and if it determines that the risk is high, the server sends the evaluation result back to the terminal. The terminal displays a message to the user saying, "This transaction is high risk. Do you really want to proceed?", and the transaction is executed only if the user confirms.

[0174] Example of a prompt:

[0175] I will send 500,000 yen to buy a car.

[0176] In this way, the present invention enables users to conduct electronic transactions securely. By preventing the risk of fraudulent use, a highly reliable payment service can be provided.

[0177] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0178] Step 1:

[0179] The terminal retrieves transaction details and text entered by the user. When the user enters transaction details and presses the send button, the terminal retrieves that input. The input is processed as text data. For example, the user enters the text "I will send 500,000 yen to buy a car." This input data is then sent to the next step.

[0180] Step 2:

[0181] The terminal sends the acquired transaction details and text to the server. Here, the text data entered by the user is sent to the server. For example, the input "I will send 500,000 yen to buy a car" is sent to the server.

[0182] Step 3:

[0183] The server passes transaction details and text received from the terminal to the analysis device, which then analyzes the content. The analysis device uses data analysis software (such as Python or Scikit-learn) to evaluate the tone and keywords of the transaction details and text. The input for the analysis is text data from the user, and the output is the analysis result. As a specific example, the sentence "I will send 500,000 yen to buy a car" is analyzed, and its tone and keywords are evaluated.

[0184] Step 4:

[0185] The server's analysis system compares the analysis results with a database of past analysis data. This database stores previously analyzed data and examples of fraudulent use. The analysis system compares the new analysis results with past data in the database and evaluates the similarity. The input for the comparison is the latest analysis result, and the output is the risk assessment result. As a specific example, the analyzed sentence, "I will send 500,000 yen to buy a car," is compared with past examples of fraudulent use.

[0186] Step 5:

[0187] The server assesses the risk of fraudulent use of a transaction based on the comparison results. The assessment result will be one of the following: high risk, medium risk, or low risk. This risk assessment result is then sent to the next step. For example, a transaction such as "transferring 500,000 yen to purchase a car" might be assessed as high risk.

[0188] Step 6:

[0189] The server sends the risk assessment results back to the terminal. The output from the server is the risk assessment result, which the terminal receives. For example, a high-risk assessment result is sent to the terminal.

[0190] Step 7:

[0191] The terminal displays the risk assessment results from the server to the user and confirms the transaction. If the risk is high, a message is displayed prompting the user to confirm whether to proceed with the transaction. The output is the message to the user, and the input is the user's response. For example, the message "This transaction is high risk. Do you really want to proceed?" is displayed, and the user responds "Yes".

[0192] Step 8:

[0193] The transaction is executed only after the user confirms the transaction. When the user enters a response, the terminal sends that response to the server, instructing it to execute the transaction. The input is the user's confirmation response, and the output is the final transaction execution. For example, if the user responds "yes," a transfer of 500,000 yen is executed.

[0194] 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.

[0195] The present invention is a system that automatically analyzes the content of posts made by users on social media and evaluates the risk of online backlash and emotional impact. The system includes means for acquiring text entered by the user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for recognizing the emotions of the text using an emotion engine, means for comparing the analysis results with a past analysis database, means for evaluating the risk of sending the text based on the comparison results and emotion recognition results, and means for displaying the risk evaluation results to the user and confirming the transmission.

[0196] System Overview

[0197] 1. User actions:

[0198] The user enters text on the social media post creation screen and presses the send button. For example, they might enter the text "This new product is completely useless" and press the send button.

[0199] 2. Retrieving and sending the posted content:

[0200] The device retrieves the text entered by the user and sends it to the server. During this process, the string data of the posted content is sent as an API request.

[0201] 3. Content analysis and emotion recognition:

[0202] The server receives the posted content sent from the terminal. The received data is passed to the analysis device and emotion engine, which analyze and recognize its content.

[0203] The analysis device evaluates the tone and keywords of the text, while the emotion engine evaluates the emotions of the text. For example, it evaluates that the phrase "completely useless" is negative and contains pessimistic emotions.

[0204] 4. Comparison and Risk Assessment:

[0205] The server's analysis system compares the analysis results with a database of past online controversies. Simultaneously, it evaluates the results, including those from the sentiment engine. For example, if the phrase "completely useless" matches many past negative posts and contains pessimistic emotions, it is rated as high risk.

[0206] 5. Displaying results and confirming submission:

[0207] The server assesses the risk based on the comparison results and sentiment recognition results, and sends the results back to the terminal. The risk assessment result is one of the following: "high risk," "medium risk," or "low risk."

[0208] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0209] 6. User verification and submission:

[0210] Users respond to the displayed message with either "yes" or "no." If they answer "no," the post is canceled; if they answer "yes," the post is sent only.

[0211] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0212] Specific example

[0213] The user enters the sentence, "This new product is completely useless."

[0214] The terminal sends that input to the server.

[0215] The server analyzes the sentence and determines that the phrase "completely useless" is very negative, and the emotion engine recognizes it as a pessimistic emotion.

[0216] The server compares the analysis results and emotion recognition results with a database of past online controversies and assesses them as high-risk due to numerous similarities.

[0217] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[0218] The post will only be sent if the user answers "yes".

[0219] Precautions

[0220] This system automatically performs risk assessment and sentiment recognition before a user makes a social media post, and then decides whether or not to post based on the results. This process makes it possible to prevent users from unintentionally making posts that could cause controversy. Furthermore, the system only requests user confirmation if it determines that the post is high-risk; otherwise, it supports rapid information dissemination.

[0221] The following describes the processing flow.

[0222] Step 1:

[0223] The user enters the content of their post on the SNS post creation screen and presses the send button. This action retrieves the post content to the device. For example, they might enter the sentence, "This new product is completely useless."

[0224] Step 2:

[0225] The device detects when the user presses the send button and sends the entered content to the server. At this time, the string data of the posted content is sent to the server as an API request.

[0226] Step 3:

[0227] The server receives the content of the post sent from the terminal. It first saves the received data in order to pass it on to the analysis device and emotion engine.

[0228] Step 4:

[0229] The analysis system on the server analyzes the content of the received posts. During this process, the tone, keywords, and emotions of the writing are evaluated. For example, the phrase "completely useless" is evaluated as having a negative tone.

[0230] Step 5:

[0231] The emotion engine on the server recognizes emotions from the content of the received posts. In doing so, it judges the emotional nuance of the entire text and understands the user's emotional state. For example, it recognizes that the phrase "completely useless" indicates a pessimistic emotion.

[0232] Step 6:

[0233] The server compares the analysis results obtained by the analysis device with a database of past online controversies. At the same time, it also takes into account the results of the emotion engine to perform an overall risk assessment. For example, the phrase "completely useless" matches many past negative posts and contains pessimistic emotions, so it is assessed as high risk.

[0234] Step 7:

[0235] The server assesses the risk of the posted content based on the comparison results and sentiment recognition results. The risk assessment result will be one of the following: "high risk," "medium risk," or "low risk."

[0236] Step 8:

[0237] The server sends the risk assessment results back to the terminal. The risk assessment results are sent to the terminal.

[0238] Step 9:

[0239] The device displays the risk assessment results received from the server to the user. In particular, if the assessment is high risk, it displays a message asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0240] Step 10:

[0241] Users respond to the displayed message with either "Yes" or "No". If they respond with "No", the post is canceled. If they respond with "Yes", the post is sent.

[0242] Step 11:

[0243] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0244] By following these steps, users can be prevented from unconsciously making posts that pose a risk of causing controversy. Furthermore, by using an emotion engine, the emotional nuances of the text can be supplemented, enabling a more accurate risk assessment.

[0245] (Example 2)

[0246] Next, we will describe 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".

[0247] When users post on social media, there is a risk of online backlash due to inappropriate content or emotionally charged posts. It is necessary to assess this risk in advance and take preventative measures. However, manual verification is time-consuming and prone to errors. There is a need for automated methods to accurately assess risks and provide appropriate warnings and confirmations.

[0248] 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.

[0249] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to the server, means for the server to pass the received text to an analysis device and an emotion engine, means for the analysis device to analyze the tone and keywords of the text, means for the emotion engine to evaluate the emotion of the text, means for comparing the analysis results and emotion evaluation results with a database of past online controversies, means for evaluating the risk of transmitting the text based on the comparison results and evaluation results, and means for transmitting the risk evaluation results to the user terminal and confirming the transmission based on the risk evaluation results. This makes it possible to automatically analyze the content of a post before the user posts it and evaluate the risk of online controversy and the risk of emotional impact.

[0250] A "user" is an individual or group that uses an information system or application.

[0251] "Text" refers to the text data that users input to post on social media, etc.

[0252] A "terminal" refers to a device such as a computer or smartphone that is operated by a user.

[0253] A "server" is a central system that receives requests from terminals via a network and performs analysis and processing.

[0254] An "analysis device" refers to software or hardware within a server that analyzes the tone and keywords of a text.

[0255] An "emotion engine" is a technological element used to evaluate emotions from the content of a text.

[0256] The "Database of Past Online Controversies" is a data store that saves and manages information on posts and incidents that have caused online controversies in the past.

[0257] "Risk assessment" is the process of evaluating the likelihood of a post going viral or causing a backlash, based on results obtained from analytical devices and emotion engines.

[0258] "Confirmation of submission" is the act of notifying the user of the risk assessment results and leaving the final decision of whether or not to submit the post to the user.

[0259] This invention is a system that analyzes text posted on social media in advance and evaluates the risk of online backlash and emotional impact. Specific embodiments for implementing this invention are described in detail below.

[0260] 1. User actions

[0261] The user enters text on the post creation screen of the SNS application and presses the send button. The entered text is retrieved in real time by the device.

[0262] 2. Retrieving and sending the posted content

[0263] The terminal retrieves text entered by the user and sends it to the server. Specifically, it encodes the retrieved text data into JSON format and sends it to the server as an API request using the HTTPS protocol. For example, if the user enters "This new product is completely useless," that text data will be sent to the server.

[0264] 3. Content Analysis and Emotion Recognition

[0265] The server receives the content of posts sent from the terminal. The received data is passed to the analysis device and sentiment engine, which analyze and recognize its content. The analysis device uses text analysis libraries (e.g., NLTK and spaCy) to evaluate the tone and keywords of the text. The sentiment engine uses Azure Cognitive Services and Google NLP API, among others, to evaluate the sentiment of the text. For example, the phrase "completely useless" is evaluated as negative and contains pessimistic emotions.

[0266] 4. Comparison with historical data and risk assessment

[0267] The server's analysis system compares the obtained analysis results with a database of past online controversies. This database is a data store that stores and manages information on posts and incidents that have caused online controversies in the past. The analysis results and sentiment evaluation results are comprehensively evaluated to assess the risk of sending the post. For example, the phrase "completely useless" is similar to many past online controversies and is therefore rated as high risk.

[0268] 5. Return and display of risk results

[0269] The server sends the risk assessment results back to the terminal. The risk assessment results are one of the following: "high risk," "medium risk," or "low risk." The terminal displays the risk assessment results received from the server to the user. For example, if the risk assessment is determined to be high risk, a message will be displayed asking the user for confirmation, such as, "This post has a high risk of causing a social media firestorm. Do you really want to post it?"

[0270] 6. User verification and submission

[0271] The user responds to the displayed message with "yes" or "no". Only if the user responds with "yes" will the device send the post to social media. This will make the post actually public on social media. If the user responds with "no", the post will be canceled.

[0272] Specific example

[0273] This explains what happens when a user types "This new product is completely unusable" and presses the submit button.

[0274] 1. User input:

[0275] A user types "This new product is completely useless" on the social media app's posting screen and presses the send button.

[0276] 2. Retrieving and sending the posted content:

[0277] The device retrieves the entered text "This new product is completely unusable," encodes it in JSON format, and sends it to the API endpoint (e.g., https: / / api.example.com / evaluate).

[0278] 3. Content analysis and emotion recognition:

[0279] The server receives the API request and sends the text to the analysis device and emotion engine.

[0280] The analysis device recognizes "completely unusable" as a negative phrase, and the emotion engine evaluates a pessimistic emotion.

[0281] 4. Comparison with Past Data and Risk Assessment:

[0282] The server queries and matches the analysis result with the past database of arson case data.

[0283] Since many similarities are found, it is evaluated as "high risk".

[0284] 5. Return and Display of Risk Results:

[0285] The server returns the evaluation result of "high risk" to the terminal.

[0286] The terminal displays a message to the user: "This post has a high risk of going viral. Do you really want to post it?"

[0287] 6. User Confirmation and Submission of Post:

[0288] The user clicks "Yes".

[0289] The terminal sends the post content to the SNS API, and the post is published.

[0290] Prompt Example

[0291] Prompt text: "This system evaluates the risk of going viral and emotional risk before posting on SNS and warns the user. Please analyze the following text and conduct a risk assessment. 'This new product is completely unusable.'"

[0292] The flow of specific processing in Example 2 will be described using FIG. 13.

[0293] Flow of Program Processing

[0294] Step 1: User Input

[0295] The user inputs text on the SNS post creation screen and performs the operation of pressing the send button.

[0296] Input: User input on the SNS post creation screen (example: "This new product is completely useless")

[0297] Output: Text data based on the user input

[0298] Step 2: Acquisition of post content

[0299] The terminal acquires the text entered by the user in real time.

[0300] Input: Text data entered by the user

[0301] Output: Acquired text data (the text entered by the user)

[0302] Step 3: Sending of post content

[0303] The terminal sends the acquired post content to the server.

[0304] Input: Acquired text data

[0305] Data processing: Encoding the text data into JSON format

[0306] Output: Data in JSON format (example: "{"text": "This new product is completely useless"}")

[0307] Step 4: Content analysis and sentiment recognition

[0308] The server receives the post content sent from the terminal and passes it to the analysis device and the sentiment engine.

[0309] Input: Text data in JSON format

[0310] Data processing and data calculation:

[0311] The analysis device uses a text analysis library (e.g., NLTK or spaCy) to analyze the tone and keywords of the text.

[0312] The emotion engine uses emotion analysis APIs (e.g., Azure Cognitive Services or Google NLP API) to evaluate the sentiment of a text.

[0313] Output: Analysis results and emotion evaluation results (e.g., "Negative," "Pessimistic emotion")

[0314] Step 5: Comparison with historical data and risk assessment

[0315] The server's analysis device compares the obtained analysis results and sentiment evaluation results with a database of past online controversies to perform a risk assessment.

[0316] Input: Analysis results and sentiment evaluation results

[0317] Data processing and calculation: A process of comparing results obtained with a database of past online crises to perform risk assessment.

[0318] Output: Risk assessment results (e.g., "High Risk", "Medium Risk", "Low Risk")

[0319] Step 6: Return and display of risk results

[0320] The server sends the risk assessment results back to the terminal, which then displays them to the user.

[0321] Input: Risk assessment results

[0322] Data processing: Encode risk assessment results into JSON format.

[0323] Output: Risk assessment results sent to the terminal (e.g., "High Risk")

[0324] Step 7: User verification and submission

[0325] The user responds to the displayed message with either "yes" or "no".

[0326] Input: User confirmation input ("Yes" or "No")

[0327] Output: Results based on the user's "yes" or "no" answer.

[0328] Specific actions:

[0329] If the user answers "yes," the device sends the post content to the SNS API, and the post is made public.

[0330] If the user answers "No," the post will be canceled.

[0331] Through the above process, this system can automatically perform risk assessment and sentiment recognition before a user makes a social media post, and based on the results, decide whether or not to proceed with the post.

[0332] (Application Example 2)

[0333] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0334] In food delivery services, emotional and negative reviews and comments posted by users can lead to unfair evaluations of service providers, negative impacts on other users, and even a decline in the overall reputation of the service. To prevent such posts, a system is needed that analyzes review content before posting and appropriately assesses the risks involved.

[0335] 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.

[0336] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for recognizing the sentiment of the text using a sentiment engine, means for comparing the analysis results with a past analysis database, means for evaluating the risk of transmitting the text based on the comparison results and sentiment recognition results, means for displaying a warning message to the user if the risk is judged to be high based on the risk evaluation results, means for displaying the risk evaluation results to the user and confirming transmission, and means for transmitting the text only if the user performs a confirmation operation. This makes it possible to prevent users from unintentionally posting negative reviews and maintain trust between food delivery service providers and users.

[0337] "Method for obtaining user-entered text" refers to a system that automatically retrieves the content of reviews and comments that users enter for food delivery services.

[0338] "Means for transmitting acquired text to an analysis device" refers to a system that transfers the content of input reviews and comments to an analysis device in a specific format.

[0339] "Means of analyzing the content of text using an analysis device" refers to a system in which an analysis device processes the content of transmitted reviews and comments as data and understands its content and meaning.

[0340] "A means of recognizing the emotions of a text using an emotion engine" refers to an engine for evaluating and identifying the emotions contained in an analyzed text, and a mechanism for determining the emotional tone and content of the text.

[0341] "Means for comparing analysis results with past analysis databases" refers to a system that compares and evaluates the content and sentiment of analyzed texts against a database of past reviews and comments.

[0342] "Means for evaluating the risk of sending text based on comparison results and sentiment recognition results" refers to a mechanism that evaluates the potential risks that reviews and comments may pose based on the results of comparison and sentiment recognition.

[0343] "A means of displaying a warning message to the user when a high risk is determined based on the risk assessment results" refers to a mechanism for informing and warning the user of a risk when it is assessed as high risk.

[0344] "A means of displaying risk assessment results to the user and confirming submission" refers to a mechanism that shows the assessment results to the user and allows them to reconfirm the submission of reviews and comments.

[0345] "A method of sending text only after the user has performed a confirmation action" means that a review or comment is actually posted only if the user confirms that they have responded to a warning message.

[0346] This invention is a system that analyzes review text entered by users and evaluates the risk of online backlash and emotional impact when a review is posted for a food delivery service. This system is implemented by the following means:

[0347] 1. User actions:

[0348] The user enters text on the food delivery service review submission screen and clicks the submit button. For example, they might enter the text, "This restaurant's delivery was terrible. It was late and the food was cold," and then click the submit button.

[0349] 2. Retrieving and sending the posted content:

[0350] The device (smartphone) retrieves the text data of the review entered by the user and sends it to the server. At this time, the string data of the posted content is sent as an API request.

[0351] 3. Content analysis and emotion recognition:

[0352] The server receives the posted content sent from the terminal. The received data is passed to the analysis device and emotion engine, which analyze and recognize its content.

[0353] The analysis device uses Natural Language Processing (NLP) technology to evaluate the tone and keywords of the text. The emotion engine uses a generative AI model (e.g., BERT or GPT-3®) to evaluate the emotion of the text.

[0354] 4. Comparison and Risk Assessment:

[0355] The server's analysis system compares the analysis results with a database of past online controversies. Simultaneously, it evaluates the results, including those from the emotion engine. For example, phrases like "it was awful" and "they were cold" are considered highly negative, and the emotion engine recognizes them as indicating pessimistic emotions.

[0356] Based on the comparison results and sentiment recognition results, the risk of sending a message is classified into high-risk, medium-risk, and low-risk categories.

[0357] 5. Displaying results and confirming submission:

[0358] The server returns the risk assessment results to the terminal. The risk assessment results will be one of the following: "high risk," "medium risk," or "low risk."

[0359] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0360] 6. User verification and submission:

[0361] Users respond to the displayed message with either "Yes" or "No". If they respond with "No", the post is canceled. If they respond with "Yes", the post is sent.

[0362] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0363] For example, when a user enters a review stating, "This restaurant's delivery was terrible. It was late and the food was cold," the system retrieves the text and sends it to the server. The server analyzes the review using a sentiment engine (e.g., GPT-3) and a historical database, determines the rating to be "high risk," and displays a warning message on the user's device. The post is only actually sent if the user replies "yes."

[0364] Example of a prompt:

[0365] "Review content: 'This restaurant's delivery was terrible. It was late and the food was cold.' Please assess the risk of this review causing a social media firestorm and evaluate the emotional risk."

[0366] Thus, by using the system of the present invention, it is possible to check in advance whether user reviews posted on food delivery services are sound and to prevent misunderstandings and inappropriate posts.

[0367] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0368] Step 1:

[0369] The user enters the text of their review on the food delivery service's review submission screen and presses the submit button. The user enters the text, and the text data is temporarily saved on the user's device.

[0370] Step 2:

[0371] The terminal retrieves the text entered by the user and sends this string data to the server as an API request. Here, the entered text is converted into a request format and sent to the server.

[0372] Step 3:

[0373] The server receives the posted content sent from the terminal and passes it to the analysis device. Processing takes place from the time the input text data reaches the server until it is passed to the analysis device.

[0374] Step 4:

[0375] The analysis device on the server analyzes the received text using Natural Language Processing (NLP) technology. Specifically, it evaluates the tone, keywords, and context of the text and generates analysis results. The input is text data, and the output is data of the analysis results.

[0376] Step 5:

[0377] The emotion engine evaluates the sentiment of a text based on text data obtained from an analysis device. Using a generative AI model (e.g., GPT-3), it identifies the emotional tone within the text and outputs the evaluation result. The input is the analyzed text data, and the output is the sentiment recognition result.

[0378] Step 6:

[0379] The server's analysis system compares the analysis results and sentiment recognition results with a database of past online controversies to perform a risk assessment. It compares the analysis results with past data and determines the risk level based on similarities and matches. The input is the analysis results and sentiment recognition results, and the output is the risk assessment result.

[0380] Step 7:

[0381] The server sends the risk assessment results back to the terminal. If the assessment result is "high risk," "medium risk," or "low risk," the assessment result data is sent to the terminal.

[0382] Step 8:

[0383] The terminal displays the risk assessment results received from the server to the user. In particular, in cases of high risk, it displays a message such as, "This post has a high risk of causing a controversy. Do you really want to post it?" The input is the risk assessment result, and the output is the display of the warning message.

[0384] Step 9:

[0385] The user responds to the displayed message with "yes" or "no." Based on this input, it is decided whether or not to send the post. The input is the user's response, and the output is an instruction to send the post.

[0386] Step 10:

[0387] If the user answers "yes," the device sends the content of the post to the food delivery service's review system, and the review is published. The input is the user's "yes" response, and the output is the actual posted review content.

[0388] Through the above processing steps, it is ensured that user reviews are properly evaluated and submitted under low-risk conditions.

[0389] 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.

[0390] Data generation model 58 is a 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> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">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.

[0391] 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.

[0392] [Second Embodiment]

[0393] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0394] 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.

[0395] 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).

[0396] 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.

[0397] 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.

[0398] 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).

[0399] 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.

[0400] 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.

[0401] 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.

[0402] 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.

[0403] 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.

[0404] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. 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".

[0405] The present invention is a system that automatically analyzes the content of posts made by users on social media and assesses the risk of causing online outrage. The system includes means for acquiring text entered by the user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for comparing the analysis results with a past analysis database, means for evaluating the risk of sending the text based on the comparison results, and means for displaying the risk assessment results to the user and confirming the transmission.

[0406] System Overview

[0407] 1. User actions:

[0408] When a user enters text on the SNS post creation screen and presses the send button, the device acquires that input.

[0409] A concrete example is when a user types the sentence "This new product is completely useless" and presses the submit button.

[0410] 2. Retrieving and sending the posted content:

[0411] The terminal retrieves the text entered by the user and sends it to the server.

[0412] A concrete example of this would be when a device sends a message to a server stating, "This new product is completely unusable."

[0413] 3. Content analysis:

[0414] The server receives text from the terminal and passes it to the analysis device, which then analyzes its content.

[0415] The analysis device evaluates the tone, emotion, and keywords of the text.

[0416] For example, we can analyze how the phrase "completely useless" is extremely negative.

[0417] 4. Comparison and Risk Assessment:

[0418] The analysis system on the server compares the analysis results with past incidents of online firestorms.

[0419] The comparison method involves evaluating the similarity to past online controversies stored in the database.

[0420] For example, the phrase "completely useless" is found to have many points of agreement with past negative posts.

[0421] 5. Displaying results and confirming submission:

[0422] The server assesses the risk based on the comparison results and sends the results back to the terminal.

[0423] The terminal receives the risk assessment results from the server and displays them to the user. If the risk assessment indicates high risk, the terminal prompts the user to confirm submission.

[0424] A concrete example would be displaying a screen that asks the user, "This post has a high risk of causing a backlash. Do you really want to post it?"

[0425] The post will only be sent if the user confirms and answers "yes". If the user answers "no", the post will be canceled.

[0426] Specific example

[0427] The user enters the sentence, "This new product is completely useless."

[0428] The terminal sends that input to the server.

[0429] The server analyzes the sentence and determines that the phrase "completely useless" is extremely negative.

[0430] The server compares the data with a database of past online controversies and assesses it as high-risk due to numerous similarities.

[0431] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[0432] The post will only be sent if the user answers "yes".

[0433] Precautions

[0434] This system automatically performs a risk assessment before a user posts on social media and decides whether or not to post based on the results. This process significantly reduces the risk of unintentionally causing a social media firestorm. Furthermore, the system only requests user confirmation if it determines the post is high-risk; otherwise, it supports rapid information dissemination.

[0435] The following describes the processing flow.

[0436] Step 1:

[0437] The user enters the content of their post on the SNS post creation screen and presses the send button. This action retrieves the post content to the device. For example, they might enter the sentence, "This new product is completely useless."

[0438] Step 2:

[0439] The device detects when the user presses the send button and sends the entered content to the server. At this time, the string data of the posted content is sent as an API request.

[0440] Step 3:

[0441] The server receives the posted content sent from the terminal. To pass the received data to the analysis device, it first saves the posted content.

[0442] Step 4:

[0443] The analysis system on the server analyzes the received posts. During this process, it evaluates the tone of the text, keywords, and emotional tone. For example, the negative connotation of a phrase like "completely useless" is assessed.

[0444] Step 5:

[0445] The server compares the analysis results obtained by the analysis device with a database of past online controversies. It evaluates the similarity to the data in the database and determines the likelihood that the posted content will cause a controversy.

[0446] Step 6:

[0447] The server assesses the risk of the posted content based on the comparison results. For example, if the phrase "completely useless" matches many past controversial incidents, it will be assessed as high risk.

[0448] Step 7:

[0449] The server returns the risk assessment results to the terminal. The risk assessment results will be one of the following: "high risk," "medium risk," or "low risk."

[0450] Step 8:

[0451] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0452] Step 9:

[0453] Users respond to the displayed message with either "yes" or "no." If they respond with "no," the post is canceled; if they respond with "yes," the post is sent only.

[0454] Step 10:

[0455] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0456] By following these steps, users can be prevented from unconsciously making posts that carry a high risk of causing controversy.

[0457] (Example 1)

[0458] Next, we will describe 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".

[0459] There is a problem in that there are few means to detect posts that have a high risk of causing social backlash in advance and notify users. As a result, unintended backlash can occur, damaging the reputation of individuals and 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] In this invention, the server includes means for acquiring text entered by the user, means for transmitting the acquired text to an information processing device, means for analyzing the content of the text using the information processing device, means for comparing the analysis results with past evaluation information, means for evaluating the risk of transmitting the text based on the comparison results, and means for displaying the risk evaluation results to the user and confirming the transmission. This makes it possible to prevent unintended online controversies by the user.

[0462] A "user" refers to a person who uses the system to post to social networking services (SNS).

[0463] "Text" refers to the text data that users input when posting on social media.

[0464] "Means of acquisition" refers to devices or programs used to collect text entered by the user.

[0465] "Information processing device" refers to a system for analyzing the content of text, including analysis devices such as servers.

[0466] "Means of transmission" refers to methods or devices for transferring acquired text to an information processing device.

[0467] "Means of analysis" refers to programs and algorithms used by information processing devices to analyze and evaluate the content of text.

[0468] "Past evaluation information" refers to analytical data stored in databases of past controversial incidents and posts.

[0469] "Means of comparison" refers to methods or programs for comparing analysis results with past evaluation information to assess similarities and risks.

[0470] "Means of risk assessment" refers to criteria or algorithms used to determine whether a piece of writing has the potential to cause a controversy or backlash.

[0471] "Risk assessment results" refer to the determination, based on the evaluation process, whether or not a post poses a risk of causing a social media firestorm.

[0472] "Means of display" refers to devices or programs that visually present risk assessment results to the user.

[0473] "Means for confirming submission" refers to methods or devices for notifying users of the risk assessment results and requesting final confirmation of the submission.

[0474] This invention is a system that analyzes the content of posts made by users on social media in advance and assesses the risk of causing online controversy. The system is configured as follows:

[0475] First, the user enters text on the social media post creation screen and presses the send button. The device retrieves this entered text and sends it to the server. A concrete example of this would be when a user enters the sentence "This new product is completely useless" and presses the send button.

[0476] Next, the server passes the text received from the terminal to the analysis device, which then analyzes its content. The analysis device evaluates the tone, emotion, and keywords of the text. For example, it might evaluate the phrase "completely useless" as being highly negative. The analysis device uses a high-performance server (for example, one equipped with an NVIDIA GPU), and natural language processing (NLP) algorithms such as Google's TensorFlow or Facebook's PyTorch are used for the analysis.

[0477] Subsequently, the analysis system on the server compares the analysis results with a database of past online controversies. The comparison method evaluates the similarity to past controversies stored in the database. For example, it is confirmed that the phrase "completely useless" has many points of agreement with past negative posts.

[0478] The server assesses the risk based on the comparison results and sends the results back to the terminal. The terminal receives the risk assessment results from the server and displays them to the user. If the risk assessment is high, the user is prompted to confirm submission. For example, a screen asking the user, "This post has a high risk of causing a controversy. Do you really want to submit it?" is displayed. The post is submitted only if the user confirms and answers "yes". If the user answers "no", the post is canceled.

[0479] In this way, the present invention automatically performs a risk assessment before a user makes a social media post and decides whether or not to post based on the results. This process makes it possible to significantly reduce the risk of unintentionally causing a social media firestorm. Furthermore, the system only requests user confirmation if it is determined to be high risk, and in other cases supports rapid information dissemination.

[0480] Specific example

[0481] When a user types "This new product is completely useless" on the posting screen and presses the send button, the device retrieves the text and sends it to the server. The server receives the text and passes it to the analysis device. The analysis device evaluates the phrase "completely useless" as extremely negative. The server compares the analysis results with a database of past online controversies and evaluates it as high risk due to many similarities. The device displays to the user, "This post has a high risk of causing a controversy. Do you really want to post it?" If the user answers "yes," the post is sent; if they answer "no," the post is canceled.

[0482] Example of a prompt

[0483] "I'm developing a text analysis system to assess the risk of social media posts about a new product going viral. This system requires prompts like the following:

[0484] For example, if a user enters the sentence, "This new product is completely useless," please explain how you would assess the risk of a backlash.

[0485] This prompt allows the generating AI model to understand the operation of a specific system and generate an appropriate risk assessment process.

[0486] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0487] Step 1:

[0488] The user enters text on the social media post creation screen and presses the send button.

[0489] Input: User's input text (e.g., "This new product is completely unusable")

[0490] Output: The user's input text is sent to the terminal.

[0491] Specific action: The user enters text on the posting screen and presses the submit button.

[0492] Step 2:

[0493] The terminal retrieves the user's input text and sends it to the server.

[0494] Input: User input text sent to the terminal

[0495] Output: User input text sent to the server

[0496] Specific operation: The terminal retrieves the entered text and sends it to the server. For example, the terminal sends the text "This new product is completely useless" to the server.

[0497] Step 3:

[0498] The server receives text from the terminal and passes it to the analysis device, which then analyzes its contents.

[0499] Input: User input text sent to the server

[0500] Output: User input text passed to the analysis device

[0501] Specific operation: The server passes the received text to the analysis device. The analysis device receives the text "This new product is completely unusable."

[0502] Step 4:

[0503] The analysis device evaluates the tone, emotion, and keywords of the text.

[0504] Input: User input text provided to the analysis device

[0505] Output: Text evaluation data (e.g., negative evaluation)

[0506] Specific operation: The analysis device uses natural language processing (NLP) techniques to analyze the text. For example, the phrase "completely useless" is evaluated as being very negative.

[0507] Step 5:

[0508] The server compares the analysis results with a database of past online controversies.

[0509] Input: Text evaluation data

[0510] Output: Risk assessment data

[0511] Specific operation: The server compares the analysis results with a database of past online controversies and evaluates similarities. For example, if the phrase "completely useless" is determined to have many points of agreement with past negative posts, it will be considered to have similarities.

[0512] Step 6:

[0513] The server sends the risk assessment results back to the terminal.

[0514] Input: Risk assessment data

[0515] Output: Risk assessment results sent to the terminal

[0516] Specific operation: The server generates a risk assessment result and sends it to the terminal. For example, the assessment result sent might say, "This post has a high risk of causing a controversy."

[0517] Step 7:

[0518] The terminal displays the risk assessment results to the user and confirms the transmission.

[0519] Input: Risk assessment results sent to the terminal

[0520] Output: Display and confirmation results for the user

[0521] Specific operation: The device displays the evaluation result to the user and asks the user to confirm submission. For example, it might ask the user, "This post has a high risk of causing controversy. Do you really want to submit it?" The post is submitted only if the user answers "yes," and canceled if they answer "no."

[0522] (Application Example 1)

[0523] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0524] In electronic payment services, it is essential to properly assess the risk of fraudulent use. However, conventional systems often fail to adequately assess the risk of fraudulent use that may occur when users conduct transactions, potentially leading to fraudulent transactions. This invention aims to solve these problems and provide a system that enables users to conduct electronic transactions securely.

[0525] 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.

[0526] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to an analysis device, means for the analysis device to analyze the content of the text, means for comparing the analysis results with a past analysis database, means for evaluating the risk of transmitting the text based on the comparison results, means for displaying the risk evaluation results to the user and confirming the transmission, means for acquiring details of the electronic transaction and transmitting them to the analysis device, means for the analysis device to evaluate the risk of fraudulent use of the transaction, and means for notifying the user of the evaluation results and requesting confirmation. This enables the user to conduct electronic transactions securely.

[0527] A "user" is an individual or group that uses the system.

[0528] "Text" refers to the text data that a user enters into the system.

[0529] An "analysis device" is a combination of hardware and software used to analyze input text and details of electronic transactions.

[0530] A "past analysis database" is a data storage system that stores previously analyzed data and the results of those analyses.

[0531] "Transmission risk" is an assessment of the potential risks that the entered text or electronic transaction may pose.

[0532] "Risk assessment results" refer to the specific evaluation value or judgment result of transmission risk calculated by the analysis device.

[0533] "Electronic transactions" refer to transactions such as buying and selling property or services, remittances, and contracts conducted using telecommunications means.

[0534] "Fraudulent use risk" refers to the assessment of the risk of unauthorized or fraudulent activity in electronic transactions.

[0535] "Notification" refers to a means of communicating information to users regarding the results of a risk assessment.

[0536] The present invention is a system that automatically analyzes transaction details and assesses the risk of fraudulent use when a user conducts an electronic transaction. This system includes means for acquiring transaction details entered by the user, means for transmitting the acquired transaction details to an analysis device, means for analyzing the transaction details using the analysis device, means for comparing the analysis results with a past analysis database, means for evaluating the risk of fraudulent use of the transaction based on the comparison results, and means for displaying the risk assessment results to the user and confirming the transaction. Specific embodiments are described below.

[0537] The terminal retrieves transaction details and text entered by the user and sends them to the server. The server passes the retrieved transaction details and text to an analysis device, which then analyzes them. The analysis device uses data analysis software (e.g., Python or Scikit-learn) to evaluate the tone and keywords of the transaction details and text. Once the tone and keywords are evaluated, the analysis device compares the results to a database of past analyses. This database stores previously analyzed data, including past incidents of online controversies and instances of misuse.

[0538] The analysis device evaluates the risk of fraudulent use of a transaction based on the comparison results. The server receives these evaluation results and sends them back to the terminal. The terminal displays the risk evaluation results to the user, and if it is determined to be high risk, it prompts the user to confirm the transaction. For example, a message such as "This transaction is high risk. Do you really want to proceed?" is displayed. The transaction is executed only if the user confirms.

[0539] As a concrete example, consider a case where a user makes a transaction to "transfer 500,000 yen to purchase a car." The terminal sends this transaction details to the server, which then passes them to an analysis device. The analysis device evaluates the risk of fraudulent use of this transaction, and if it determines that the risk is high, the server sends the evaluation result back to the terminal. The terminal displays a message to the user saying, "This transaction is high risk. Do you really want to proceed?", and the transaction is executed only if the user confirms.

[0540] Example of a prompt:

[0541] I will send 500,000 yen to buy a car.

[0542] In this way, the present invention enables users to conduct electronic transactions securely. By preventing the risk of fraudulent use, a highly reliable payment service can be provided.

[0543] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0544] Step 1:

[0545] The terminal retrieves transaction details and text entered by the user. When the user enters transaction details and presses the send button, the terminal retrieves that input. The input is processed as text data. For example, the user enters the text "I will send 500,000 yen to buy a car." This input data is then sent to the next step.

[0546] Step 2:

[0547] The terminal sends the acquired transaction details and text to the server. Here, the text data entered by the user is sent to the server. For example, the input "I will send 500,000 yen to buy a car" is sent to the server.

[0548] Step 3:

[0549] The server passes transaction details and text received from the terminal to the analysis device, which then analyzes the content. The analysis device uses data analysis software (such as Python or Scikit-learn) to evaluate the tone and keywords of the transaction details and text. The input for the analysis is text data from the user, and the output is the analysis result. As a specific example, the sentence "I will send 500,000 yen to buy a car" is analyzed, and its tone and keywords are evaluated.

[0550] Step 4:

[0551] The server's analysis system compares the analysis results with a database of past analysis data. This database stores previously analyzed data and examples of fraudulent use. The analysis system compares the new analysis results with past data in the database and evaluates the similarity. The input for the comparison is the latest analysis result, and the output is the risk assessment result. As a specific example, the analyzed sentence, "I will send 500,000 yen to buy a car," is compared with past examples of fraudulent use.

[0552] Step 5:

[0553] The server evaluates the risk of fraudulent use of a transaction based on the comparison results. The evaluation result will be one of the following: high risk, medium risk, or low risk. This risk evaluation result is then sent to the next step. For example, a transaction such as "transferring 500,000 yen to purchase a car" might be evaluated as high risk.

[0554] Step 6:

[0555] The server sends the risk assessment results back to the terminal. The output from the server is the risk assessment result, which the terminal receives. For example, a high-risk assessment result is sent to the terminal.

[0556] Step 7:

[0557] The terminal displays the risk assessment results from the server to the user and confirms the transaction. If the risk is high, a message is displayed prompting the user to confirm whether to proceed with the transaction. The output is the message to the user, and the input is the user's response. For example, the message "This transaction is high risk. Do you really want to proceed?" is displayed, and the user responds "Yes".

[0558] Step 8:

[0559] The transaction is executed only after the user confirms the transaction. When the user enters a response, the terminal sends that response to the server, instructing it to execute the transaction. The input is the user's confirmation response, and the output is the final transaction execution. For example, if the user responds "yes," a transfer of 500,000 yen is executed.

[0560] 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.

[0561] The present invention is a system that automatically analyzes the content of posts made by users on social media and evaluates the risk of online backlash and emotional impact. The system includes means for acquiring text entered by the user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for recognizing the emotions of the text using an emotion engine, means for comparing the analysis results with a past analysis database, means for evaluating the risk of sending the text based on the comparison results and emotion recognition results, and means for displaying the risk evaluation results to the user and confirming the transmission.

[0562] System Overview

[0563] 1. User actions:

[0564] The user enters text on the social media post creation screen and presses the send button. For example, they might enter the text "This new product is completely useless" and press the send button.

[0565] 2. Retrieving and sending the posted content:

[0566] The device retrieves the text entered by the user and sends it to the server. During this process, the string data of the posted content is sent as an API request.

[0567] 3. Content analysis and emotion recognition:

[0568] The server receives the posted content sent from the terminal. The received data is passed to the analysis device and emotion engine, which analyze and recognize its content.

[0569] The analysis device evaluates the tone and keywords of the text, while the emotion engine evaluates the emotions of the text. For example, it evaluates that the phrase "completely useless" is negative and contains pessimistic emotions.

[0570] 4. Comparison and Risk Assessment:

[0571] The server's analysis system compares the analysis results with a database of past online controversies. Simultaneously, it evaluates the results, including those from the sentiment engine. For example, if the phrase "completely useless" matches many past negative posts and contains pessimistic emotions, it is rated as high risk.

[0572] 5. Displaying results and confirming submission:

[0573] The server assesses the risk based on the comparison results and sentiment recognition results, and sends the results back to the terminal. The risk assessment result is one of the following: "high risk," "medium risk," or "low risk."

[0574] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0575] 6. User verification and submission:

[0576] Users respond to the displayed message with either "yes" or "no." If they respond with "no," the post is canceled; if they respond with "yes," the post is sent only.

[0577] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0578] Specific example

[0579] The user enters the sentence, "This new product is completely useless."

[0580] The terminal sends that input to the server.

[0581] The server analyzes the sentence and determines that the phrase "completely useless" is very negative, and the emotion engine recognizes it as a pessimistic emotion.

[0582] The server compares the analysis results and emotion recognition results with a database of past online controversies and assesses them as high-risk due to numerous similarities.

[0583] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[0584] The post will only be sent if the user answers "yes".

[0585] Precautions

[0586] This system automatically performs risk assessment and sentiment recognition before a user makes a social media post, and then decides whether or not to post based on the results. This process makes it possible to prevent users from unintentionally making posts that could cause controversy. Furthermore, the system only requests user confirmation if it determines that the post is high-risk; otherwise, it supports rapid information dissemination.

[0587] The following describes the processing flow.

[0588] Step 1:

[0589] The user enters the content of their post on the SNS post creation screen and presses the send button. This action retrieves the post content to the device. For example, they might enter the sentence, "This new product is completely useless."

[0590] Step 2:

[0591] The device detects when the user presses the send button and sends the entered content to the server. At this time, the string data of the posted content is sent to the server as an API request.

[0592] Step 3:

[0593] The server receives the content of the post sent from the terminal. It first saves the received data in order to pass it on to the analysis device and emotion engine.

[0594] Step 4:

[0595] The analysis system on the server analyzes the content of the received posts. During this process, the tone, keywords, and emotions of the writing are evaluated. For example, the phrase "completely useless" is evaluated as having a negative tone.

[0596] Step 5:

[0597] The emotion engine on the server recognizes emotions from the content of the received posts. In doing so, it judges the emotional nuance of the entire text and understands the user's emotional state. For example, it recognizes that the phrase "completely useless" indicates a pessimistic emotion.

[0598] Step 6:

[0599] The server compares the analysis results obtained by the analysis device with a database of past online controversies. At the same time, it also takes into account the results of the emotion engine to perform an overall risk assessment. For example, the phrase "completely useless" matches many past negative posts and contains pessimistic emotions, so it is assessed as high risk.

[0600] Step 7:

[0601] The server assesses the risk of the posted content based on the comparison results and sentiment recognition results. The risk assessment result will be one of the following: "high risk," "medium risk," or "low risk."

[0602] Step 8:

[0603] The server sends the risk assessment results back to the terminal. The risk assessment results are sent to the terminal.

[0604] Step 9:

[0605] The device displays the risk assessment results received from the server to the user. In particular, if the assessment is high risk, it displays a message asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0606] Step 10:

[0607] Users respond to the displayed message with either "Yes" or "No". If they respond with "No", the post is canceled. If they respond with "Yes", the post is sent.

[0608] Step 11:

[0609] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0610] By following these steps, users can be prevented from unconsciously making posts that pose a risk of causing controversy. Furthermore, by using an emotion engine, the emotional nuances of the text can be supplemented, enabling a more accurate risk assessment.

[0611] (Example 2)

[0612] Next, we will describe 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".

[0613] When users post on social media, there is a risk of online backlash due to inappropriate content or emotionally charged posts. It is necessary to assess this risk in advance and take preventative measures. However, manual verification is time-consuming and prone to errors. There is a need for automated methods to accurately assess risks and provide appropriate warnings and confirmations.

[0614] 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.

[0615] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to the server, means for the server to pass the received text to an analysis device and an emotion engine, means for the analysis device to analyze the tone and keywords of the text, means for the emotion engine to evaluate the emotion of the text, means for comparing the analysis results and emotion evaluation results with a database of past online controversies, means for evaluating the risk of transmitting the text based on the comparison results and evaluation results, and means for transmitting the risk evaluation results to the user terminal and confirming the transmission based on the risk evaluation results. This makes it possible to automatically analyze the content of a post before the user posts it and evaluate the risk of online controversy and the risk of emotional impact.

[0616] A "user" is an individual or group that uses an information system or application.

[0617] "Text" refers to the text data that users input to post on social media, etc.

[0618] A "terminal" refers to a device such as a computer or smartphone that is operated by a user.

[0619] A "server" is a central system that receives requests from terminals via a network and performs analysis and processing.

[0620] An "analysis device" refers to software or hardware within a server that analyzes the tone and keywords of a text.

[0621] An "emotion engine" is a technological element used to evaluate emotions from the content of a text.

[0622] The "Database of Past Online Controversies" is a data store that saves and manages information on posts and incidents that have caused online controversies in the past.

[0623] "Risk assessment" is the process of evaluating the likelihood of a post going viral or causing a backlash, based on results obtained from analytical devices and emotion engines.

[0624] "Confirmation of submission" is the act of notifying the user of the risk assessment results and leaving the final decision of whether or not to submit the post to the user.

[0625] This invention is a system that analyzes text posted on social media in advance and evaluates the risk of online backlash and emotional impact. Specific embodiments for implementing this invention are described in detail below.

[0626] 1. User actions

[0627] The user enters text on the post creation screen of the SNS application and presses the send button. The entered text is retrieved in real time by the device.

[0628] 2. Retrieving and sending the posted content

[0629] The terminal retrieves text entered by the user and sends it to the server. Specifically, it encodes the retrieved text data into JSON format and sends it to the server as an API request using the HTTPS protocol. For example, if the user enters "This new product is completely useless," that text data will be sent to the server.

[0630] 3. Content Analysis and Emotion Recognition

[0631] The server receives the content of posts sent from the terminal. The received data is passed to the analysis device and sentiment engine, which analyze and recognize its content. The analysis device uses text analysis libraries (e.g., NLTK or spaCy) to evaluate the tone and keywords of the text. The sentiment engine uses Azure Cognitive Services or Google NLP API to evaluate the sentiment of the text. For example, the phrase "completely useless" is evaluated as negative and contains pessimistic emotions.

[0632] 4. Comparison with historical data and risk assessment

[0633] The server's analysis system compares the obtained analysis results with a database of past online controversies. This database is a data store that stores and manages information on posts and incidents that have caused online controversies in the past. The analysis results and sentiment evaluation results are comprehensively evaluated to assess the risk of sending the post. For example, the phrase "completely useless" is similar to many past online controversies and is therefore rated as high risk.

[0634] 5. Return and display of risk results

[0635] The server sends the risk assessment results back to the terminal. The risk assessment results are one of the following: "high risk," "medium risk," or "low risk." The terminal displays the risk assessment results received from the server to the user. For example, if the risk assessment is determined to be high risk, a message will be displayed asking the user for confirmation, such as, "This post has a high risk of causing a social media firestorm. Do you really want to post it?"

[0636] 6. User verification and submission

[0637] The user responds to the displayed message with "yes" or "no". Only if the user responds with "yes" will the device send the post to social media. This will make the post actually public on social media. If the user responds with "no", the post will be canceled.

[0638] Specific example

[0639] This explains what happens when a user types "This new product is completely unusable" and presses the submit button.

[0640] 1. User input:

[0641] A user types "This new product is completely useless" on the social media app's posting screen and presses the send button.

[0642] 2. Retrieving and sending the posted content:

[0643] The device retrieves the entered text "This new product is completely unusable," encodes it in JSON format, and sends it to the API endpoint (e.g., https: / / api.example.com / evaluate).

[0644] 3. Content analysis and emotion recognition:

[0645] The server receives the API request and sends the text to the analysis device and emotion engine.

[0646] The analysis device recognizes "completely useless" as a negative phrase, and the emotion engine evaluates it as a pessimistic emotion.

[0647] 4. Comparison with historical data and risk assessment:

[0648] The server queries a database of past online controversies to compare the analysis results.

[0649] Due to the numerous similarities observed, it is rated as "high risk."

[0650] 5. Return and display of risk results:

[0651] The server returns a "high risk" assessment result to the terminal.

[0652] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[0653] 6. User verification and submission:

[0654] The user clicks "Yes".

[0655] The device sends the content of the post to the SNS API, and the post is made public.

[0656] Example prompt

[0657] Prompt: "This system evaluates the risk of controversy and emotional impact before posting a message on social media and warns the user. Please analyze the following message and perform a risk assessment: 'This new product is completely useless.'"

[0658] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0659] Program processing flow

[0660] Step 1: User input

[0661] The user enters text on the social media post creation screen and presses the send button.

[0662] Input: User input on the SNS post creation screen (e.g., "This new product is completely unusable").

[0663] Output: Text data based on user input

[0664] Step 2: Retrieve the posted content

[0665] The device retrieves the text entered by the user in real time.

[0666] Input: Text data entered by the user

[0667] Output: Acquired text data (text entered by the user)

[0668] Step 3: Submit your post

[0669] The device sends the retrieved post content to the server.

[0670] Input: Acquired text data

[0671] Data processing: Encoding text data into JSON format

[0672] Output: Data in JSON format (Example: "{"text": "This new product is completely unusable"}")

[0673] Step 4: Content Analysis and Sentiment Recognition

[0674] The server receives the posted content sent from the terminal and passes it on to the analysis device and emotion engine.

[0675] Input: Text data in JSON format

[0676] Data processing and data computation:

[0677] The analysis device uses a text analysis library (e.g., NLTK or spaCy) to analyze the tone and keywords of the text.

[0678] The emotion engine uses emotion analysis APIs (e.g., Azure Cognitive Services or Google NLP API) to evaluate the sentiment of a text.

[0679] Output: Analysis results and emotion evaluation results (e.g., "Negative," "Pessimistic emotion")

[0680] Step 5: Comparison with historical data and risk assessment

[0681] The server's analysis device compares the obtained analysis results and sentiment evaluation results with a database of past online controversies to perform a risk assessment.

[0682] Input: Analysis results and sentiment evaluation results

[0683] Data processing and calculation: A process of comparing results obtained with a database of past online crises to perform risk assessment.

[0684] Output: Risk assessment results (e.g., "High Risk", "Medium Risk", "Low Risk")

[0685] Step 6: Return and display of risk results

[0686] The server sends the risk assessment results back to the terminal, which then displays them to the user.

[0687] Input: Risk assessment results

[0688] Data processing: Encode risk assessment results into JSON format.

[0689] Output: Risk assessment results sent to the terminal (e.g., "High Risk")

[0690] Step 7: User verification and submission

[0691] The user responds to the displayed message with either "yes" or "no".

[0692] Input: User confirmation input ("Yes" or "No")

[0693] Output: Results based on the user's "yes" or "no" answer.

[0694] Specific actions:

[0695] If the user answers "yes," the device sends the post content to the SNS API, and the post is made public.

[0696] If the user answers "No," the post will be canceled.

[0697] Through the above process, this system can automatically perform risk assessment and sentiment recognition before a user makes a social media post, and based on the results, decide whether or not to proceed with the post.

[0698] (Application Example 2)

[0699] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0700] In food delivery services, emotional and negative reviews and comments posted by users can lead to unfair evaluations of service providers, negative impacts on other users, and even a decline in the overall reputation of the service. To prevent such posts, a system is needed that analyzes review content before posting and appropriately assesses the risks involved.

[0701] 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.

[0702] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for recognizing the sentiment of the text using a sentiment engine, means for comparing the analysis results with a past analysis database, means for evaluating the risk of transmitting the text based on the comparison results and sentiment recognition results, means for displaying a warning message to the user if the risk is judged to be high based on the risk evaluation results, means for displaying the risk evaluation results to the user and confirming transmission, and means for transmitting the text only if the user performs a confirmation operation. This makes it possible to prevent the posting of unintended negative reviews by users and maintain trust between food delivery service providers and users.

[0703] "Method for obtaining user-entered text" refers to a system that automatically retrieves the content of reviews and comments that users enter for food delivery services.

[0704] "Means for transmitting acquired text to an analysis device" refers to a system that transfers the content of input reviews and comments to an analysis device in a specific format.

[0705] "Means of analyzing the content of text using an analysis device" refers to a system in which an analysis device processes the content of transmitted reviews and comments as data and understands its content and meaning.

[0706] "A means of recognizing the emotions of a text using an emotion engine" refers to an engine for evaluating and identifying the emotions contained in an analyzed text, and a mechanism for determining the emotional tone and content of the text.

[0707] "Means for comparing analysis results with past analysis databases" refers to a system that compares and evaluates the content and sentiment of analyzed texts against a database of past reviews and comments.

[0708] "Means for evaluating the risk of sending text based on comparison results and sentiment recognition results" refers to a mechanism that evaluates the potential risks that reviews and comments may pose based on the results of comparison and sentiment recognition.

[0709] "A means of displaying a warning message to the user when a high risk is determined based on the risk assessment results" refers to a mechanism for informing and warning the user of a risk when it is assessed as high risk.

[0710] "A means of displaying risk assessment results to the user and confirming submission" refers to a mechanism that shows the assessment results to the user and allows them to reconfirm the submission of reviews and comments.

[0711] "A method of sending text only after the user has performed a confirmation action" means that a review or comment is actually posted only if the user confirms that they have responded to a warning message.

[0712] This invention is a system that analyzes review text entered by users and evaluates the risk of online backlash and emotional impact when a review is posted for a food delivery service. This system is implemented by the following means:

[0713] 1. User actions:

[0714] The user enters text on the food delivery service review submission screen and clicks the submit button. For example, they might enter the text, "This restaurant's delivery was terrible. It was late and the food was cold," and then click the submit button.

[0715] 2. Retrieving and sending the posted content:

[0716] The device (smartphone) retrieves the text data of the review entered by the user and sends it to the server. At this time, the string data of the posted content is sent as an API request.

[0717] 3. Content analysis and emotion recognition:

[0718] The server receives the posted content sent from the terminal. The received data is passed to the analysis device and emotion engine, which analyze and recognize its content.

[0719] The analysis device uses Natural Language Processing (NLP) technology to evaluate the tone and keywords of the text. The sentiment engine uses a generative AI model (e.g., BERT or GPT-3) to evaluate the sentiment of the text.

[0720] 4. Comparison and Risk Assessment:

[0721] The server's analysis system compares the analysis results with a database of past online controversies. Simultaneously, it evaluates the results, including those from the emotion engine. For example, phrases like "it was the worst" and "they were cold" are considered highly negative, and the emotion engine recognizes them as indicating pessimistic emotions.

[0722] Based on the comparison results and sentiment recognition results, the risk of sending a message is classified into high-risk, medium-risk, and low-risk categories.

[0723] 5. Displaying results and confirming submission:

[0724] The server returns the risk assessment results to the terminal. The risk assessment results will be one of the following: "high risk," "medium risk," or "low risk."

[0725] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0726] 6. User verification and submission:

[0727] Users respond to the displayed message with either "Yes" or "No". If they respond with "No", the post is canceled. If they respond with "Yes", the post is sent.

[0728] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0729] For example, when a user enters a review stating, "This restaurant's delivery was terrible. It was late and the food was cold," the system retrieves the text and sends it to the server. The server analyzes the review using a sentiment engine (e.g., GPT-3) and a historical database, determines the rating to be "high risk," and displays a warning message on the user's device. The post is only actually sent if the user replies "yes."

[0730] Example of a prompt:

[0731] "Review content: 'This restaurant's delivery was terrible. It was late and the food was cold.' Please assess the risk of this review causing a social media firestorm and evaluate the emotional risk."

[0732] Thus, by using the system of the present invention, it is possible to check in advance whether user reviews posted on food delivery services are sound and to prevent misunderstandings and inappropriate posts.

[0733] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0734] Step 1:

[0735] The user enters the text of their review on the food delivery service's review submission screen and presses the submit button. The user enters the text, and the text data is temporarily saved on the user's device.

[0736] Step 2:

[0737] The terminal retrieves the text entered by the user and sends this string data to the server as an API request. Here, the entered text is converted into a request format and sent to the server.

[0738] Step 3:

[0739] The server receives the posted content sent from the terminal and passes it to the analysis device. Processing takes place from the time the input text data reaches the server until it is passed to the analysis device.

[0740] Step 4:

[0741] The analysis device on the server analyzes the received text using Natural Language Processing (NLP) technology. Specifically, it evaluates the tone, keywords, and context of the text and generates analysis results. The input is text data, and the output is data of the analysis results.

[0742] Step 5:

[0743] The emotion engine evaluates the sentiment of a text based on text data obtained from an analysis device. Using a generative AI model (e.g., GPT-3), it identifies the emotional tone within the text and outputs the evaluation result. The input is the analyzed text data, and the output is the sentiment recognition result.

[0744] Step 6:

[0745] The server's analysis system compares the analysis results and sentiment recognition results with a database of past online controversies to perform a risk assessment. It compares the analysis results with past data and determines the risk level based on similarities and matches. The input is the analysis results and sentiment recognition results, and the output is the risk assessment result.

[0746] Step 7:

[0747] The server sends the risk assessment results back to the terminal. If the assessment result is "high risk," "medium risk," or "low risk," the assessment result data is sent to the terminal.

[0748] Step 8:

[0749] The terminal displays the risk assessment results received from the server to the user. In particular, in cases of high risk, it displays a message such as, "This post has a high risk of causing a controversy. Do you really want to post it?" The input is the risk assessment result, and the output is the display of the warning message.

[0750] Step 9:

[0751] The user responds to the displayed message with "yes" or "no." Based on this input, it is decided whether or not to send the post. The input is the user's response, and the output is an instruction to send the post.

[0752] Step 10:

[0753] If the user answers "yes," the device sends the content of the post to the food delivery service's review system, and the review is published. The input is the user's "yes" response, and the output is the actual posted review content.

[0754] Through the above processing steps, it is ensured that user reviews are properly evaluated and submitted under low-risk conditions.

[0755] 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.

[0756] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">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.

[0757] 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.

[0758] [Third Embodiment]

[0759] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0760] 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.

[0761] 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).

[0762] 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.

[0763] 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.

[0764] 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).

[0765] 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.

[0766] 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.

[0767] 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.

[0768] 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.

[0769] 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.

[0770] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0771] The present invention is a system that automatically analyzes the content of posts made by users on social media and assesses the risk of causing online outrage. The system includes means for acquiring text entered by the user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for comparing the analysis results with a past analysis database, means for evaluating the risk of sending the text based on the comparison results, and means for displaying the risk assessment results to the user and confirming the transmission.

[0772] System Overview

[0773] 1. User actions:

[0774] When a user enters text on the SNS post creation screen and presses the send button, the device acquires that input.

[0775] A concrete example is when a user types the sentence "This new product is completely useless" and presses the submit button.

[0776] 2. Retrieving and sending the posted content:

[0777] The terminal retrieves the text entered by the user and sends it to the server.

[0778] A concrete example of this would be when a device sends a message to a server stating, "This new product is completely unusable."

[0779] 3. Content analysis:

[0780] The server receives text from the terminal and passes it to the analysis device, which then analyzes its content.

[0781] The analysis device evaluates the tone, emotion, and keywords of the text.

[0782] For example, we can analyze how the phrase "completely useless" is extremely negative.

[0783] 4. Comparison and Risk Assessment:

[0784] The analysis system on the server compares the analysis results with past incidents of online firestorms.

[0785] The comparison method involves evaluating the similarity to past online controversies stored in the database.

[0786] For example, the phrase "completely useless" is found to have many points of agreement with past negative posts.

[0787] 5. Displaying results and confirming submission:

[0788] The server assesses the risk based on the comparison results and sends the results back to the terminal.

[0789] The terminal receives the risk assessment results from the server and displays them to the user. If the risk assessment indicates high risk, the terminal prompts the user to confirm submission.

[0790] A concrete example would be displaying a screen that asks the user, "This post has a high risk of causing a backlash. Do you really want to post it?"

[0791] The post will only be sent if the user confirms and answers "yes". If the user answers "no", the post will be canceled.

[0792] Specific example

[0793] The user enters the sentence, "This new product is completely useless."

[0794] The terminal sends that input to the server.

[0795] The server analyzes the sentence and determines that the phrase "completely useless" is extremely negative.

[0796] The server compares the data with a database of past online controversies and assesses it as high-risk due to numerous similarities.

[0797] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[0798] The post will only be sent if the user answers "yes".

[0799] Precautions

[0800] This system automatically performs a risk assessment before a user posts on social media and decides whether or not to post based on the results. This process significantly reduces the risk of unintentionally causing a social media firestorm. Furthermore, the system only requests user confirmation if it determines the post is high-risk; otherwise, it supports rapid information dissemination.

[0801] The following describes the processing flow.

[0802] Step 1:

[0803] The user enters the content of their post on the SNS post creation screen and presses the send button. This action retrieves the post content to the device. For example, they might enter the sentence, "This new product is completely useless."

[0804] Step 2:

[0805] The device detects when the user presses the send button and sends the entered content to the server. At this time, the string data of the posted content is sent as an API request.

[0806] Step 3:

[0807] The server receives the posted content sent from the terminal. To pass the received data to the analysis device, it first saves the posted content.

[0808] Step 4:

[0809] The analysis system on the server analyzes the received posts. During this process, it evaluates the tone of the text, keywords, and emotional tone. For example, the negative connotation of a phrase like "completely useless" is assessed.

[0810] Step 5:

[0811] The server compares the analysis results obtained by the analysis device with a database of past online controversies. It evaluates the similarity to the data in the database and determines the likelihood that the posted content will cause a controversy.

[0812] Step 6:

[0813] The server assesses the risk of the posted content based on the comparison results. For example, if the phrase "completely useless" matches many past controversial incidents, it will be assessed as high risk.

[0814] Step 7:

[0815] The server returns the risk assessment results to the terminal. The risk assessment results will be one of the following: "high risk," "medium risk," or "low risk."

[0816] Step 8:

[0817] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0818] Step 9:

[0819] Users respond to the displayed message with either "yes" or "no." If they respond with "no," the post is canceled; if they respond with "yes," the post is sent only.

[0820] Step 10:

[0821] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0822] By following these steps, users can be prevented from unconsciously making posts that carry a high risk of causing controversy.

[0823] (Example 1)

[0824] Next, we will describe 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."

[0825] There is a problem in that there are few means to detect posts that have a high risk of causing social backlash in advance and notify users. As a result, unintended backlash can occur, damaging the reputation of individuals and companies.

[0826] 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.

[0827] In this invention, the server includes means for acquiring text entered by the user, means for transmitting the acquired text to an information processing device, means for analyzing the content of the text using the information processing device, means for comparing the analysis results with past evaluation information, means for evaluating the risk of transmitting the text based on the comparison results, and means for displaying the risk evaluation results to the user and confirming the transmission. This makes it possible to prevent unintended online controversies by the user.

[0828] A "user" refers to a person who uses the system to post to social networking services (SNS).

[0829] "Text" refers to the text data that users input when posting on social media.

[0830] "Means of acquisition" refers to devices or programs used to collect text entered by the user.

[0831] "Information processing device" refers to a system for analyzing the content of text, including analysis devices such as servers.

[0832] "Means of transmission" refers to methods or devices for transferring acquired text to an information processing device.

[0833] "Means of analysis" refers to programs and algorithms used by information processing devices to analyze and evaluate the content of text.

[0834] "Past evaluation information" refers to analytical data stored in databases of past controversial incidents and posts.

[0835] "Means of comparison" refers to methods or programs for comparing analysis results with past evaluation information to assess similarities and risks.

[0836] "Means of risk assessment" refers to criteria or algorithms used to determine whether a piece of writing has the potential to cause a controversy or backlash.

[0837] "Risk assessment results" refer to the determination, based on the evaluation process, whether or not a post poses a risk of causing a social media firestorm.

[0838] "Means of display" refers to devices or programs that visually present risk assessment results to the user.

[0839] "Means for confirming submission" refers to methods or devices for notifying users of the risk assessment results and requesting final confirmation of the submission.

[0840] This invention is a system that analyzes the content of posts made by users on social media in advance and assesses the risk of causing online controversy. The system is configured as follows:

[0841] First, the user enters text on the social media post creation screen and presses the send button. The device retrieves this entered text and sends it to the server. A concrete example of this would be when a user enters the sentence "This new product is completely useless" and presses the send button.

[0842] Next, the server passes the text received from the terminal to the analysis device, which then analyzes its content. The analysis device evaluates the tone, emotion, and keywords of the text. For example, it might evaluate the phrase "completely useless" as being highly negative. The analysis device uses a high-performance server (for example, one equipped with an NVIDIA GPU), and natural language processing (NLP) algorithms such as Google's TensorFlow or Facebook's PyTorch are used for the analysis.

[0843] Subsequently, the analysis system on the server compares the analysis results with a database of past online controversies. The comparison method evaluates the similarity to past controversies stored in the database. For example, it is confirmed that the phrase "completely useless" has many points of agreement with past negative posts.

[0844] The server assesses the risk based on the comparison results and sends the results back to the terminal. The terminal receives the risk assessment results from the server and displays them to the user. If the risk assessment is high, the user is prompted to confirm submission. For example, a screen asking the user, "This post has a high risk of causing a controversy. Do you really want to submit it?" is displayed. The post is submitted only if the user confirms and answers "yes". If the user answers "no", the post is canceled.

[0845] In this way, the present invention automatically performs a risk assessment before a user makes a social media post and decides whether or not to post based on the results. This process makes it possible to significantly reduce the risk of unintentionally causing a social media firestorm. Furthermore, the system only requests user confirmation if it is determined to be high risk, and in other cases supports rapid information dissemination.

[0846] Specific example

[0847] When a user types "This new product is completely useless" on the posting screen and presses the send button, the device retrieves the text and sends it to the server. The server receives the text and passes it to the analysis device. The analysis device evaluates the phrase "completely useless" as extremely negative. The server compares the analysis results with a database of past online controversies and evaluates it as high risk due to many similarities. The device displays to the user, "This post has a high risk of causing a controversy. Do you really want to post it?" If the user answers "yes," the post is sent; if they answer "no," the post is canceled.

[0848] Example of a prompt

[0849] "I'm developing a text analysis system to assess the risk of social media posts about a new product going viral. This system requires prompts like the following:

[0850] For example, if a user enters the sentence, "This new product is completely useless," please explain how you would assess the risk of a backlash.

[0851] This prompt allows the generating AI model to understand the operation of a specific system and generate an appropriate risk assessment process.

[0852] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0853] Step 1:

[0854] The user enters text on the social media post creation screen and presses the send button.

[0855] Input: User's input text (e.g., "This new product is completely unusable")

[0856] Output: The user's input text is sent to the terminal.

[0857] Specific action: The user enters text on the posting screen and presses the submit button.

[0858] Step 2:

[0859] The terminal retrieves the user's input text and sends it to the server.

[0860] Input: User input text sent to the terminal

[0861] Output: User input text sent to the server

[0862] Specific operation: The terminal retrieves the entered text and sends it to the server. For example, the terminal sends the text "This new product is completely useless" to the server.

[0863] Step 3:

[0864] The server receives text from the terminal and passes it to the analysis device, which then analyzes its contents.

[0865] Input: User input text sent to the server

[0866] Output: User input text passed to the analysis device

[0867] Specific operation: The server passes the received text to the analysis device. The analysis device receives the text "This new product is completely unusable."

[0868] Step 4:

[0869] The analysis device evaluates the tone, emotion, and keywords of the text.

[0870] Input: User input text provided to the analysis device

[0871] Output: Text evaluation data (e.g., negative evaluation)

[0872] Specific operation: The analysis device uses natural language processing (NLP) techniques to analyze the text. For example, the phrase "completely useless" is evaluated as being very negative.

[0873] Step 5:

[0874] The server compares the analysis results with a database of past online controversies.

[0875] Input: Text evaluation data

[0876] Output: Risk assessment data

[0877] Specific operation: The server compares the analysis results with a database of past online controversies and evaluates similarities. For example, if the phrase "completely useless" is determined to have many points of agreement with past negative posts, it will be considered to have similarities.

[0878] Step 6:

[0879] The server sends the risk assessment results back to the terminal.

[0880] Input: Risk assessment data

[0881] Output: Risk assessment results sent to the terminal

[0882] Specific operation: The server generates a risk assessment result and sends it to the terminal. For example, the assessment result sent might say, "This post has a high risk of causing a controversy."

[0883] Step 7:

[0884] The terminal displays the risk assessment results to the user and confirms the transmission.

[0885] Input: Risk assessment results sent to the terminal

[0886] Output: Display and confirmation results for the user

[0887] Specific operation: The device displays the evaluation result to the user and asks the user to confirm submission. For example, it might ask the user, "This post has a high risk of causing controversy. Do you really want to submit it?" The post is submitted only if the user answers "yes," and canceled if they answer "no."

[0888] (Application Example 1)

[0889] Next, we will explain Application Example 1. In the following explanation, 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."

[0890] In electronic payment services, it is essential to properly assess the risk of fraudulent use. However, conventional systems often fail to adequately assess the risk of fraudulent use that may occur when users conduct transactions, potentially leading to fraudulent transactions. This invention aims to solve these problems and provide a system that enables users to conduct electronic transactions securely.

[0891] 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.

[0892] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to an analysis device, means for the analysis device to analyze the content of the text, means for comparing the analysis results with a past analysis database, means for evaluating the risk of transmitting the text based on the comparison results, means for displaying the risk evaluation results to the user and confirming the transmission, means for acquiring details of the electronic transaction and transmitting them to the analysis device, means for the analysis device to evaluate the risk of fraudulent use of the transaction, and means for notifying the user of the evaluation results and requesting confirmation. This enables the user to conduct electronic transactions securely.

[0893] A "user" is an individual or group that uses the system.

[0894] "Text" refers to the text data that a user enters into the system.

[0895] An "analysis device" is a combination of hardware and software used to analyze input text and details of electronic transactions.

[0896] A "past analysis database" is a data storage system that stores previously analyzed data and the results of those analyses.

[0897] "Transmission risk" is an assessment of the potential risks that the entered text or electronic transaction may pose.

[0898] "Risk assessment results" refer to the specific evaluation value or judgment result of transmission risk calculated by the analysis device.

[0899] "Electronic transactions" refer to transactions such as buying and selling property or services, remittances, and contracts conducted using telecommunications means.

[0900] "Fraudulent use risk" refers to the assessment of the risk of unauthorized or fraudulent activity in electronic transactions.

[0901] "Notification" refers to a means of communicating information to users regarding the results of a risk assessment.

[0902] The present invention is a system that automatically analyzes transaction details and assesses the risk of fraudulent use when a user conducts an electronic transaction. This system includes means for acquiring transaction details entered by the user, means for transmitting the acquired transaction details to an analysis device, means for analyzing the transaction details using the analysis device, means for comparing the analysis results with a past analysis database, means for evaluating the risk of fraudulent use of the transaction based on the comparison results, and means for displaying the risk assessment results to the user and confirming the transaction. Specific embodiments are described below.

[0903] The terminal retrieves transaction details and text entered by the user and sends them to the server. The server passes the retrieved transaction details and text to an analysis device, which then analyzes them. The analysis device uses data analysis software (e.g., Python or Scikit-learn) to evaluate the tone and keywords of the transaction details and text. Once the tone and keywords are evaluated, the analysis device compares the results to a database of past analyses. This database stores previously analyzed data, including past incidents of online controversies and instances of misuse.

[0904] The analysis device evaluates the risk of fraudulent use of a transaction based on the comparison results. The server receives these evaluation results and sends them back to the terminal. The terminal displays the risk evaluation results to the user, and if it is determined to be high risk, it prompts the user to confirm the transaction. For example, a message such as "This transaction is high risk. Do you really want to proceed?" is displayed. The transaction is executed only if the user confirms.

[0905] As a concrete example, consider a case where a user makes a transaction to "transfer 500,000 yen to purchase a car." The terminal sends this transaction details to the server, which then passes them to an analysis device. The analysis device evaluates the risk of fraudulent use of this transaction, and if it determines that the risk is high, the server sends the evaluation result back to the terminal. The terminal displays a message to the user saying, "This transaction is high risk. Do you really want to proceed?", and the transaction is executed only if the user confirms.

[0906] Example of a prompt:

[0907] I will send 500,000 yen to buy a car.

[0908] In this way, the present invention enables users to conduct electronic transactions securely. By preventing the risk of fraudulent use, a highly reliable payment service can be provided.

[0909] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0910] Step 1:

[0911] The terminal retrieves transaction details and text entered by the user. When the user enters transaction details and presses the send button, the terminal retrieves that input. The input is processed as text data. For example, the user enters the text "I will send 500,000 yen to buy a car." This input data is then sent to the next step.

[0912] Step 2:

[0913] The terminal sends the acquired transaction details and text to the server. Here, the text data entered by the user is sent to the server. For example, the input "I will send 500,000 yen to buy a car" is sent to the server.

[0914] Step 3:

[0915] The server passes transaction details and text received from the terminal to the analysis device, which then analyzes the content. The analysis device uses data analysis software (such as Python or Scikit-learn) to evaluate the tone and keywords of the transaction details and text. The input for the analysis is text data from the user, and the output is the analysis result. As a specific example, the sentence "I will send 500,000 yen to buy a car" is analyzed, and its tone and keywords are evaluated.

[0916] Step 4:

[0917] The server's analysis system compares the analysis results with a database of past analysis data. This database stores previously analyzed data and examples of fraudulent use. The analysis system compares the new analysis results with past data in the database and evaluates the similarity. The input for the comparison is the latest analysis result, and the output is the risk assessment result. As a specific example, the analyzed sentence, "I will send 500,000 yen to buy a car," is compared with past examples of fraudulent use.

[0918] Step 5:

[0919] The server evaluates the risk of fraudulent use of a transaction based on the comparison results. The evaluation result will be one of the following: high risk, medium risk, or low risk. This risk evaluation result is then sent to the next step. For example, a transaction such as "transferring 500,000 yen to purchase a car" might be evaluated as high risk.

[0920] Step 6:

[0921] The server sends the risk assessment results back to the terminal. The output from the server is the risk assessment result, which the terminal receives. For example, a high-risk assessment result is sent to the terminal.

[0922] Step 7:

[0923] The terminal displays the risk assessment results from the server to the user and confirms the transaction. If the risk is high, a message is displayed prompting the user to confirm whether to proceed with the transaction. The output is the message to the user, and the input is the user's response. For example, the message "This transaction is high risk. Do you really want to proceed?" is displayed, and the user responds "Yes".

[0924] Step 8:

[0925] The transaction is executed only after the user confirms the transaction. When the user enters a response, the terminal sends that response to the server, instructing it to execute the transaction. The input is the user's confirmation response, and the output is the final transaction execution. For example, if the user responds "yes," a transfer of 500,000 yen is executed.

[0926] 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.

[0927] The present invention is a system that automatically analyzes the content of posts made by users on social media and evaluates the risk of online backlash and emotional impact. The system includes means for acquiring text entered by the user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for recognizing the emotions of the text using an emotion engine, means for comparing the analysis results with a past analysis database, means for evaluating the risk of sending the text based on the comparison results and emotion recognition results, and means for displaying the risk evaluation results to the user and confirming the transmission.

[0928] System Overview

[0929] 1. User actions:

[0930] The user enters text on the social media post creation screen and presses the send button. For example, they might enter the text "This new product is completely useless" and press the send button.

[0931] 2. Retrieving and sending the posted content:

[0932] The device retrieves the text entered by the user and sends it to the server. During this process, the string data of the posted content is sent as an API request.

[0933] 3. Content analysis and emotion recognition:

[0934] The server receives the posted content sent from the terminal. The received data is passed to the analysis device and emotion engine, which analyze and recognize its content.

[0935] The analysis device evaluates the tone and keywords of the text, while the emotion engine evaluates the emotions of the text. For example, it evaluates that the phrase "completely useless" is negative and contains pessimistic emotions.

[0936] 4. Comparison and Risk Assessment:

[0937] The server's analysis system compares the analysis results with a database of past online controversies. Simultaneously, it evaluates the results, including those from the sentiment engine. For example, if the phrase "completely useless" matches many past negative posts and contains pessimistic emotions, it is rated as high risk.

[0938] 5. Displaying results and confirming submission:

[0939] The server assesses the risk based on the comparison results and sentiment recognition results, and sends the results back to the terminal. The risk assessment result is one of the following: "high risk," "medium risk," or "low risk."

[0940] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0941] 6. User verification and submission:

[0942] Users respond to the displayed message with either "yes" or "no." If they respond with "no," the post is canceled; if they respond with "yes," the post is sent only.

[0943] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0944] Specific example

[0945] The user enters the sentence, "This new product is completely useless."

[0946] The terminal sends that input to the server.

[0947] The server analyzes the sentence and determines that the phrase "completely useless" is very negative, and the emotion engine recognizes it as a pessimistic emotion.

[0948] The server compares the analysis results and emotion recognition results with a database of past online controversies and assesses them as high-risk due to numerous similarities.

[0949] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[0950] The post will only be sent if the user answers "yes".

[0951] Precautions

[0952] This system automatically performs risk assessment and sentiment recognition before a user makes a social media post, and then decides whether or not to post based on the results. This process makes it possible to prevent users from unintentionally making posts that could cause controversy. Furthermore, the system only requests user confirmation if it determines that the post is high-risk; otherwise, it supports rapid information dissemination.

[0953] The following describes the processing flow.

[0954] Step 1:

[0955] The user enters the content of their post on the SNS post creation screen and presses the send button. This action retrieves the post content to the device. For example, they might enter the sentence, "This new product is completely useless."

[0956] Step 2:

[0957] The device detects when the user presses the send button and sends the entered content to the server. At this time, the string data of the posted content is sent to the server as an API request.

[0958] Step 3:

[0959] The server receives the content of the post sent from the terminal. It first saves the received data in order to pass it on to the analysis device and emotion engine.

[0960] Step 4:

[0961] The analysis system on the server analyzes the content of the received posts. During this process, the tone, keywords, and emotions of the writing are evaluated. For example, the phrase "completely useless" is evaluated as having a negative tone.

[0962] Step 5:

[0963] The emotion engine on the server recognizes emotions from the content of the received posts. In doing so, it judges the emotional nuance of the entire text and understands the user's emotional state. For example, it recognizes that the phrase "completely useless" indicates a pessimistic emotion.

[0964] Step 6:

[0965] The server compares the analysis results obtained by the analysis device with a database of past online controversies. At the same time, it also takes into account the results of the emotion engine to perform an overall risk assessment. For example, the phrase "completely useless" matches many past negative posts and contains pessimistic emotions, so it is assessed as high risk.

[0966] Step 7:

[0967] The server assesses the risk of the posted content based on the comparison results and sentiment recognition results. The risk assessment result will be one of the following: "high risk," "medium risk," or "low risk."

[0968] Step 8:

[0969] The server sends the risk assessment results back to the terminal. The risk assessment results are sent to the terminal.

[0970] Step 9:

[0971] The device displays the risk assessment results received from the server to the user. In particular, if the assessment is high risk, it displays a message asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[0972] Step 10:

[0973] Users respond to the displayed message with either "Yes" or "No". If they respond with "No", the post is canceled. If they respond with "Yes", the post is sent.

[0974] Step 11:

[0975] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[0976] By following these steps, users can be prevented from unconsciously making posts that pose a risk of causing controversy. Furthermore, by using an emotion engine, the emotional nuances of the text can be supplemented, enabling a more accurate risk assessment.

[0977] (Example 2)

[0978] Next, we will describe 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."

[0979] When users post on social media, there is a risk of online backlash due to inappropriate content or emotionally charged posts. It is necessary to assess this risk in advance and take preventative measures. However, manual verification is time-consuming and prone to errors. There is a need for automated methods to accurately assess risks and provide appropriate warnings and confirmations.

[0980] 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.

[0981] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to the server, means for the server to pass the received text to an analysis device and an emotion engine, means for the analysis device to analyze the tone and keywords of the text, means for the emotion engine to evaluate the emotion of the text, means for comparing the analysis results and emotion evaluation results with a database of past online controversies, means for evaluating the risk of transmitting the text based on the comparison results and evaluation results, and means for transmitting the risk evaluation results to the user terminal and confirming the transmission based on the risk evaluation results. This makes it possible to automatically analyze the content of a post before the user posts it and evaluate the risk of online controversy and the risk of emotional impact.

[0982] A "user" is an individual or group that uses an information system or application.

[0983] "Text" refers to the text data that users input to post on social media, etc.

[0984] A "terminal" refers to a device such as a computer or smartphone that is operated by a user.

[0985] A "server" is a central system that receives requests from terminals via a network and performs analysis and processing.

[0986] An "analysis device" refers to software or hardware within a server that analyzes the tone and keywords of a text.

[0987] An "emotion engine" is a technological element used to evaluate emotions from the content of a text.

[0988] The "Database of Past Online Controversies" is a data store that saves and manages information on posts and incidents that have caused online controversies in the past.

[0989] "Risk assessment" is the process of evaluating the likelihood of a post going viral or causing a backlash, based on results obtained from analytical devices and emotion engines.

[0990] "Confirmation of submission" is the act of notifying the user of the risk assessment results and leaving the final decision of whether or not to submit the post to the user.

[0991] This invention is a system that analyzes text posted on social media in advance and evaluates the risk of online backlash and emotional impact. Specific embodiments for implementing this invention are described in detail below.

[0992] 1. User actions

[0993] The user enters text on the post creation screen of the SNS application and presses the send button. The entered text is retrieved in real time by the device.

[0994] 2. Retrieving and sending the posted content

[0995] The terminal retrieves text entered by the user and sends it to the server. Specifically, it encodes the retrieved text data into JSON format and sends it to the server as an API request using the HTTPS protocol. For example, if the user enters "This new product is completely useless," that text data will be sent to the server.

[0996] 3. Content Analysis and Emotion Recognition

[0997] The server receives the content of posts sent from the terminal. The received data is passed to the analysis device and sentiment engine, which analyze and recognize its content. The analysis device uses text analysis libraries (e.g., NLTK or spaCy) to evaluate the tone and keywords of the text. The sentiment engine uses Azure Cognitive Services or Google NLP API to evaluate the sentiment of the text. For example, the phrase "completely useless" is evaluated as negative and contains pessimistic emotions.

[0998] 4. Comparison with historical data and risk assessment

[0999] The server's analysis system compares the obtained analysis results with a database of past online controversies. This database is a data store that stores and manages information on posts and incidents that have caused online controversies in the past. The analysis results and sentiment evaluation results are comprehensively evaluated to assess the risk of sending the post. For example, the phrase "completely useless" is similar to many past online controversies and is therefore rated as high risk.

[1000] 5. Return and display of risk results

[1001] The server sends the risk assessment results back to the terminal. The risk assessment results are one of the following: "high risk," "medium risk," or "low risk." The terminal displays the risk assessment results received from the server to the user. For example, if the risk assessment is determined to be high risk, a message will be displayed asking the user for confirmation, such as, "This post has a high risk of causing a social media firestorm. Do you really want to post it?"

[1002] 6. User verification and submission

[1003] The user responds to the displayed message with "yes" or "no". Only if the user responds with "yes" will the device send the post to social media. This will make the post actually public on social media. If the user responds with "no", the post will be canceled.

[1004] Specific example

[1005] This explains what happens when a user types "This new product is completely unusable" and presses the submit button.

[1006] 1. User input:

[1007] A user types "This new product is completely useless" on the social media app's posting screen and presses the send button.

[1008] 2. Retrieving and sending the posted content:

[1009] The device retrieves the entered text "This new product is completely unusable," encodes it in JSON format, and sends it to the API endpoint (e.g., https: / / api.example.com / evaluate).

[1010] 3. Content analysis and emotion recognition:

[1011] The server receives the API request and sends the text to the analysis device and emotion engine.

[1012] The analysis device recognizes "completely useless" as a negative phrase, and the emotion engine evaluates it as a pessimistic emotion.

[1013] 4. Comparison with historical data and risk assessment:

[1014] The server queries a database of past online controversies to compare the analysis results.

[1015] Due to the numerous similarities observed, it is rated as "high risk."

[1016] 5. Return and display of risk results:

[1017] The server returns a "high risk" assessment result to the terminal.

[1018] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[1019] 6. User verification and submission:

[1020] The user clicks "Yes".

[1021] The device sends the content of the post to the SNS API, and the post is made public.

[1022] Example prompt

[1023] Prompt: "This system evaluates the risk of controversy and emotional impact before posting a message on social media and warns the user. Please analyze the following message and perform a risk assessment: 'This new product is completely useless.'"

[1024] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1025] Program processing flow

[1026] Step 1: User input

[1027] The user enters text on the social media post creation screen and presses the send button.

[1028] Input: User input on the SNS post creation screen (e.g., "This new product is completely unusable").

[1029] Output: Text data based on user input

[1030] Step 2: Retrieve the posted content

[1031] The device retrieves the text entered by the user in real time.

[1032] Input: Text data entered by the user

[1033] Output: Acquired text data (text entered by the user)

[1034] Step 3: Submit your post

[1035] The device sends the retrieved post content to the server.

[1036] Input: Acquired text data

[1037] Data processing: Encoding text data into JSON format

[1038] Output: Data in JSON format (Example: "{"text": "This new product is completely unusable"}")

[1039] Step 4: Content Analysis and Sentiment Recognition

[1040] The server receives the posted content sent from the terminal and passes it on to the analysis device and emotion engine.

[1041] Input: Text data in JSON format

[1042] Data processing and data computation:

[1043] The analysis device uses a text analysis library (e.g., NLTK or spaCy) to analyze the tone and keywords of the text.

[1044] The emotion engine uses emotion analysis APIs (e.g., Azure Cognitive Services or Google NLP API) to evaluate the sentiment of a text.

[1045] Output: Analysis results and emotion evaluation results (e.g., "Negative," "Pessimistic emotion")

[1046] Step 5: Comparison with historical data and risk assessment

[1047] The server's analysis device compares the obtained analysis results and sentiment evaluation results with a database of past online controversies to perform a risk assessment.

[1048] Input: Analysis results and sentiment evaluation results

[1049] Data processing and calculation: A process of comparing results obtained with a database of past online crises to perform risk assessment.

[1050] Output: Risk assessment results (e.g., "High Risk", "Medium Risk", "Low Risk")

[1051] Step 6: Return and display of risk results

[1052] The server sends the risk assessment results back to the terminal, which then displays them to the user.

[1053] Input: Risk assessment results

[1054] Data processing: Encode risk assessment results into JSON format.

[1055] Output: Risk assessment results sent to the terminal (e.g., "High Risk")

[1056] Step 7: User verification and submission

[1057] The user responds to the displayed message with either "yes" or "no".

[1058] Input: User confirmation input ("Yes" or "No")

[1059] Output: Results based on the user's "yes" or "no" answer.

[1060] Specific actions:

[1061] If the user answers "yes," the device sends the post content to the SNS API, and the post is made public.

[1062] If the user answers "No," the post will be canceled.

[1063] Through the above process, this system can automatically perform risk assessment and sentiment recognition before a user makes a social media post, and based on the results, decide whether or not to proceed with the post.

[1064] (Application Example 2)

[1065] Next, we will explain application example 2. In the following explanation, 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."

[1066] In food delivery services, emotional and negative reviews and comments posted by users can lead to unfair evaluations of service providers, negative impacts on other users, and even a decline in the overall reputation of the service. To prevent such posts, a system is needed that analyzes review content before posting and appropriately assesses the risks involved.

[1067] 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.

[1068] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for recognizing the sentiment of the text using a sentiment engine, means for comparing the analysis results with a past analysis database, means for evaluating the risk of transmitting the text based on the comparison results and sentiment recognition results, means for displaying a warning message to the user if the risk is judged to be high based on the risk evaluation results, means for displaying the risk evaluation results to the user and confirming transmission, and means for transmitting the text only if the user performs a confirmation operation. This makes it possible to prevent the posting of unintended negative reviews by users and maintain trust between food delivery service providers and users.

[1069] "Method for obtaining user-entered text" refers to a system that automatically retrieves the content of reviews and comments that users enter for food delivery services.

[1070] "Means for transmitting acquired text to an analysis device" refers to a system that transfers the content of input reviews and comments to an analysis device in a specific format.

[1071] "Means of analyzing the content of text using an analysis device" refers to a system in which an analysis device processes the content of transmitted reviews and comments as data and understands its content and meaning.

[1072] "A means of recognizing the emotions of a text using an emotion engine" refers to an engine for evaluating and identifying the emotions contained in an analyzed text, and a mechanism for determining the emotional tone and content of the text.

[1073] "Means for comparing analysis results with past analysis databases" refers to a system that compares and evaluates the content and sentiment of analyzed texts against a database of past reviews and comments.

[1074] "Means for evaluating the risk of sending text based on comparison results and sentiment recognition results" refers to a mechanism that evaluates the potential risks that reviews and comments may pose based on the results of comparison and sentiment recognition.

[1075] "A means of displaying a warning message to the user when a high risk is determined based on the risk assessment results" refers to a mechanism for informing and warning the user of a risk when it is assessed as high risk.

[1076] "A means of displaying risk assessment results to the user and confirming submission" refers to a mechanism that shows the assessment results to the user and allows them to reconfirm the submission of reviews and comments.

[1077] "A method of sending text only after the user has performed a confirmation action" means that a review or comment is actually posted only if the user confirms that they have responded to a warning message.

[1078] This invention is a system that analyzes review text entered by users and evaluates the risk of online backlash and emotional impact when a review is posted for a food delivery service. This system is implemented by the following means:

[1079] 1. User actions:

[1080] The user enters text on the food delivery service review submission screen and clicks the submit button. For example, they might enter the text, "This restaurant's delivery was terrible. It was late and the food was cold," and then click the submit button.

[1081] 2. Retrieving and sending the posted content:

[1082] The device (smartphone) retrieves the text data of the review entered by the user and sends it to the server. At this time, the string data of the posted content is sent as an API request.

[1083] 3. Content analysis and emotion recognition:

[1084] The server receives the posted content sent from the terminal. The received data is passed to the analysis device and emotion engine, which analyze and recognize its content.

[1085] The analysis device uses Natural Language Processing (NLP) technology to evaluate the tone and keywords of the text. The sentiment engine uses a generative AI model (e.g., BERT or GPT-3) to evaluate the sentiment of the text.

[1086] 4. Comparison and Risk Assessment:

[1087] The server's analysis system compares the analysis results with a database of past online controversies. Simultaneously, it evaluates the results, including those from the emotion engine. For example, phrases like "it was the worst" and "they were cold" are considered highly negative, and the emotion engine recognizes them as indicating pessimistic emotions.

[1088] Based on the comparison results and sentiment recognition results, the risk of sending a message is classified into high-risk, medium-risk, and low-risk categories.

[1089] 5. Displaying results and confirming submission:

[1090] The server returns the risk assessment results to the terminal. The risk assessment results will be one of the following: "high risk," "medium risk," or "low risk."

[1091] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[1092] 6. User verification and submission:

[1093] Users respond to the displayed message with either "Yes" or "No". If they respond with "No", the post is canceled. If they respond with "Yes", the post is sent.

[1094] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[1095] For example, when a user enters a review stating, "This restaurant's delivery was terrible. It was late and the food was cold," the system retrieves the text and sends it to the server. The server analyzes the review using a sentiment engine (e.g., GPT-3) and a historical database, determines the rating to be "high risk," and displays a warning message on the user's device. The post is only actually sent if the user replies "yes."

[1096] Example of a prompt:

[1097] "Review content: 'This restaurant's delivery was terrible. It was late and the food was cold.' Please assess the risk of this review causing a social media firestorm and evaluate the emotional risk."

[1098] Thus, by using the system of the present invention, it is possible to check in advance whether user reviews posted on food delivery services are sound and to prevent misunderstandings and inappropriate posts.

[1099] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1100] Step 1:

[1101] The user enters the text of their review on the food delivery service's review submission screen and presses the submit button. The user enters the text, and the text data is temporarily saved on the user's device.

[1102] Step 2:

[1103] The terminal retrieves the text entered by the user and sends this string data to the server as an API request. Here, the entered text is converted into a request format and sent to the server.

[1104] Step 3:

[1105] The server receives the posted content sent from the terminal and passes it to the analysis device. Processing takes place from the time the input text data reaches the server until it is passed to the analysis device.

[1106] Step 4:

[1107] The analysis device on the server analyzes the received text using Natural Language Processing (NLP) technology. Specifically, it evaluates the tone, keywords, and context of the text and generates analysis results. The input is text data, and the output is data of the analysis results.

[1108] Step 5:

[1109] The emotion engine evaluates the sentiment of a text based on text data obtained from an analysis device. Using a generative AI model (e.g., GPT-3), it identifies the emotional tone within the text and outputs the evaluation result. The input is the analyzed text data, and the output is the sentiment recognition result.

[1110] Step 6:

[1111] The server's analysis system compares the analysis results and sentiment recognition results with a database of past online controversies to perform a risk assessment. It compares the analysis results with past data and determines the risk level based on similarities and matches. The input is the analysis results and sentiment recognition results, and the output is the risk assessment result.

[1112] Step 7:

[1113] The server sends the risk assessment results back to the terminal. If the assessment result is "high risk," "medium risk," or "low risk," the assessment result data is sent to the terminal.

[1114] Step 8:

[1115] The terminal displays the risk assessment results received from the server to the user. In particular, in cases of high risk, it displays a message such as, "This post has a high risk of causing a controversy. Do you really want to post it?" The input is the risk assessment result, and the output is the display of the warning message.

[1116] Step 9:

[1117] The user responds to the displayed message with "yes" or "no." Based on this input, it is decided whether or not to send the post. The input is the user's response, and the output is an instruction to send the post.

[1118] Step 10:

[1119] If the user answers "yes," the device sends the content of the post to the food delivery service's review system, and the review is published. The input is the user's "yes" response, and the output is the actual posted review content.

[1120] Through the above processing steps, it is ensured that user reviews are properly evaluated and submitted under low-risk conditions.

[1121] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 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.

[1122] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">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.

[1123] 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 specific processing may also be performed by the headset terminal 314.

[1124] [Fourth Embodiment]

[1125] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1126] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1127] 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).

[1128] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. 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 controlled object 443 are also connected to the bus 52.

[1129] 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.

[1130] 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).

[1131] 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.

[1132] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1133] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1134] 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.

[1135] 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.

[1136] In robot 414, 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.

[1137] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1138] The present invention is a system that automatically analyzes the content of posts made by users on social media and assesses the risk of causing online outrage. The system includes means for acquiring text entered by the user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for comparing the analysis results with a past analysis database, means for evaluating the risk of sending the text based on the comparison results, and means for displaying the risk assessment results to the user and confirming the transmission.

[1139] System Overview

[1140] 1. User actions:

[1141] When a user enters text on the SNS post creation screen and presses the send button, the device acquires that input.

[1142] A concrete example is when a user types the sentence "This new product is completely useless" and presses the submit button.

[1143] 2. Retrieving and sending the posted content:

[1144] The terminal retrieves the text entered by the user and sends it to the server.

[1145] A concrete example of this would be when a device sends a message to a server stating, "This new product is completely unusable."

[1146] 3. Content analysis:

[1147] The server receives text from the terminal and passes it to the analysis device, which then analyzes its content.

[1148] The analysis device evaluates the tone, emotion, and keywords of the text.

[1149] For example, we can analyze how the phrase "completely useless" is extremely negative.

[1150] 4. Comparison and Risk Assessment:

[1151] The analysis system on the server compares the analysis results with past incidents of online firestorms.

[1152] The comparison method involves evaluating the similarity to past online controversies stored in the database.

[1153] For example, the phrase "completely useless" is found to have many points of agreement with past negative posts.

[1154] 5. Displaying results and confirming submission:

[1155] The server assesses the risk based on the comparison results and sends the results back to the terminal.

[1156] The terminal receives the risk assessment results from the server and displays them to the user. If the risk assessment indicates high risk, the terminal prompts the user to confirm submission.

[1157] A concrete example would be displaying a screen that asks the user, "This post has a high risk of causing a backlash. Do you really want to post it?"

[1158] The post will only be sent if the user confirms and answers "yes". If the user answers "no", the post will be canceled.

[1159] Specific example

[1160] The user enters the sentence, "This new product is completely useless."

[1161] The terminal sends that input to the server.

[1162] The server analyzes the sentence and determines that the phrase "completely useless" is extremely negative.

[1163] The server compares the data with a database of past online controversies and assesses it as high-risk due to numerous similarities.

[1164] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[1165] The post will only be sent if the user answers "yes".

[1166] Precautions

[1167] This system automatically performs a risk assessment before a user posts on social media and decides whether or not to post based on the results. This process significantly reduces the risk of unintentionally causing a social media firestorm. Furthermore, the system only requests user confirmation if it determines the post is high-risk; otherwise, it supports rapid information dissemination.

[1168] The following describes the processing flow.

[1169] Step 1:

[1170] The user enters the content of their post on the SNS post creation screen and presses the send button. This action retrieves the post content to the device. For example, they might enter the sentence, "This new product is completely useless."

[1171] Step 2:

[1172] The device detects when the user presses the send button and sends the entered content to the server. At this time, the string data of the posted content is sent as an API request.

[1173] Step 3:

[1174] The server receives the posted content sent from the terminal. To pass the received data to the analysis device, it first saves the posted content.

[1175] Step 4:

[1176] The analysis system on the server analyzes the received posts. During this process, it evaluates the tone of the text, keywords, and emotional tone. For example, the negative connotation of a phrase like "completely useless" is assessed.

[1177] Step 5:

[1178] The server compares the analysis results obtained by the analysis device with a database of past online controversies. It evaluates the similarity to the data in the database and determines the likelihood that the posted content will cause a controversy.

[1179] Step 6:

[1180] The server assesses the risk of the posted content based on the comparison results. For example, if the phrase "completely useless" matches many past controversial incidents, it will be assessed as high risk.

[1181] Step 7:

[1182] The server returns the risk assessment results to the terminal. The risk assessment results will be one of the following: "high risk," "medium risk," or "low risk."

[1183] Step 8:

[1184] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[1185] Step 9:

[1186] Users respond to the displayed message with either "yes" or "no." If they respond with "no," the post is canceled; if they respond with "yes," the post is sent only.

[1187] Step 10:

[1188] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[1189] By following these steps, users can be prevented from unconsciously making posts that carry a high risk of causing controversy.

[1190] (Example 1)

[1191] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1192] There is a problem in that there are few means to detect posts that have a high risk of causing social backlash in advance and notify users. As a result, unintended backlash can occur, damaging the reputation of individuals and companies.

[1193] 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.

[1194] In this invention, the server includes means for acquiring text entered by the user, means for transmitting the acquired text to an information processing device, means for analyzing the content of the text using the information processing device, means for comparing the analysis results with past evaluation information, means for evaluating the risk of transmitting the text based on the comparison results, and means for displaying the risk evaluation results to the user and confirming the transmission. This makes it possible to prevent unintended online controversies by the user.

[1195] A "user" refers to a person who uses the system to post to social networking services (SNS).

[1196] "Text" refers to the text data that users input when posting on social media.

[1197] "Means of acquisition" refers to devices or programs used to collect text entered by the user.

[1198] "Information processing device" refers to a system for analyzing the content of text, including analysis devices such as servers.

[1199] "Means of transmission" refers to methods or devices for transferring acquired text to an information processing device.

[1200] "Means of analysis" refers to programs and algorithms used by information processing devices to analyze and evaluate the content of text.

[1201] "Past evaluation information" refers to analytical data stored in databases of past controversial incidents and posts.

[1202] "Means of comparison" refers to methods or programs for comparing analysis results with past evaluation information to assess similarities and risks.

[1203] "Means of risk assessment" refers to criteria or algorithms used to determine whether a piece of writing has the potential to cause a controversy or backlash.

[1204] "Risk assessment results" refer to the determination, based on the evaluation process, whether or not a post poses a risk of causing a social media firestorm.

[1205] "Means of display" refers to devices or programs that visually present risk assessment results to the user.

[1206] "Means for confirming submission" refers to methods or devices for notifying users of the risk assessment results and requesting final confirmation of the submission.

[1207] This invention is a system that analyzes the content of posts made by users on social media in advance and assesses the risk of causing online controversy. The system is configured as follows:

[1208] First, the user enters text on the social media post creation screen and presses the send button. The device retrieves this entered text and sends it to the server. A concrete example of this would be when a user enters the sentence "This new product is completely useless" and presses the send button.

[1209] Next, the server passes the text received from the terminal to the analysis device, which then analyzes its content. The analysis device evaluates the tone, emotion, and keywords of the text. For example, it might evaluate the phrase "completely useless" as being highly negative. The analysis device uses a high-performance server (for example, one equipped with an NVIDIA GPU), and natural language processing (NLP) algorithms such as Google's TensorFlow or Facebook's PyTorch are used for the analysis.

[1210] Subsequently, the analysis system on the server compares the analysis results with a database of past online controversies. The comparison method evaluates the similarity to past controversies stored in the database. For example, it is confirmed that the phrase "completely useless" has many points of agreement with past negative posts.

[1211] The server assesses the risk based on the comparison results and sends the results back to the terminal. The terminal receives the risk assessment results from the server and displays them to the user. If the risk assessment is high, the user is prompted to confirm submission. For example, a screen asking the user, "This post has a high risk of causing a controversy. Do you really want to submit it?" is displayed. The post is submitted only if the user confirms and answers "yes". If the user answers "no", the post is canceled.

[1212] In this way, the present invention automatically performs a risk assessment before a user makes a social media post and decides whether or not to post based on the results. This process makes it possible to significantly reduce the risk of unintentionally causing a social media firestorm. Furthermore, the system only requests user confirmation if it is determined to be high risk, and in other cases supports rapid information dissemination.

[1213] Specific example

[1214] When a user types "This new product is completely useless" on the posting screen and presses the send button, the device retrieves the text and sends it to the server. The server receives the text and passes it to the analysis device. The analysis device evaluates the phrase "completely useless" as extremely negative. The server compares the analysis results with a database of past online controversies and evaluates it as high risk due to many similarities. The device displays to the user, "This post has a high risk of causing a controversy. Do you really want to post it?" If the user answers "yes," the post is sent; if they answer "no," the post is canceled.

[1215] Example of a prompt

[1216] "I'm developing a text analysis system to assess the risk of social media posts about a new product going viral. This system requires prompts like the following:

[1217] For example, if a user enters the sentence, "This new product is completely useless," please explain how you would assess the risk of a backlash.

[1218] This prompt allows the generating AI model to understand the operation of a specific system and generate an appropriate risk assessment process.

[1219] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1220] Step 1:

[1221] The user enters text on the social media post creation screen and presses the send button.

[1222] Input: User's input text (e.g., "This new product is completely unusable")

[1223] Output: The user's input text is sent to the terminal.

[1224] Specific action: The user enters text on the posting screen and presses the submit button.

[1225] Step 2:

[1226] The terminal retrieves the user's input text and sends it to the server.

[1227] Input: User input text sent to the terminal

[1228] Output: User input text sent to the server

[1229] Specific operation: The terminal retrieves the entered text and sends it to the server. For example, the terminal sends the text "This new product is completely useless" to the server.

[1230] Step 3:

[1231] The server receives text from the terminal and passes it to the analysis device, which then analyzes its contents.

[1232] Input: User input text sent to the server

[1233] Output: User input text passed to the analysis device

[1234] Specific operation: The server passes the received text to the analysis device. The analysis device receives the text "This new product is completely unusable."

[1235] Step 4:

[1236] The analysis device evaluates the tone, emotion, and keywords of the text.

[1237] Input: User input text provided to the analysis device

[1238] Output: Text evaluation data (e.g., negative evaluation)

[1239] Specific operation: The analysis device uses natural language processing (NLP) techniques to analyze the text. For example, the phrase "completely useless" is evaluated as being very negative.

[1240] Step 5:

[1241] The server compares the analysis results with a database of past online controversies.

[1242] Input: Text evaluation data

[1243] Output: Risk assessment data

[1244] Specific operation: The server compares the analysis results with a database of past online controversies and evaluates similarities. For example, if the phrase "completely useless" is determined to have many points of agreement with past negative posts, it will be considered to have similarities.

[1245] Step 6:

[1246] The server sends the risk assessment results back to the terminal.

[1247] Input: Risk assessment data

[1248] Output: Risk assessment results sent to the terminal

[1249] Specific operation: The server generates a risk assessment result and sends it to the terminal. For example, the assessment result sent might say, "This post has a high risk of causing a controversy."

[1250] Step 7:

[1251] The terminal displays the risk assessment results to the user and confirms the transmission.

[1252] Input: Risk assessment results sent to the terminal

[1253] Output: Display and confirmation results for the user

[1254] Specific operation: The device displays the evaluation result to the user and asks the user to confirm submission. For example, it might ask the user, "This post has a high risk of causing controversy. Do you really want to submit it?" The post is submitted only if the user answers "yes," and canceled if they answer "no."

[1255] (Application Example 1)

[1256] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1257] In electronic payment services, it is essential to properly assess the risk of fraudulent use. However, conventional systems often fail to adequately assess the risk of fraudulent use that may occur when users conduct transactions, potentially leading to fraudulent transactions. This invention aims to solve these problems and provide a system that enables users to conduct electronic transactions securely.

[1258] 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.

[1259] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to an analysis device, means for the analysis device to analyze the content of the text, means for comparing the analysis results with a past analysis database, means for evaluating the risk of transmitting the text based on the comparison results, means for displaying the risk evaluation results to the user and confirming the transmission, means for acquiring details of the electronic transaction and transmitting them to the analysis device, means for the analysis device to evaluate the risk of fraudulent use of the transaction, and means for notifying the user of the evaluation results and requesting confirmation. This enables the user to conduct electronic transactions securely.

[1260] A "user" is an individual or group that uses the system.

[1261] "Text" refers to the text data that a user enters into the system.

[1262] An "analysis device" is a combination of hardware and software used to analyze input text and details of electronic transactions.

[1263] A "past analysis database" is a data storage system that stores previously analyzed data and the results of those analyses.

[1264] "Transmission risk" is an assessment of the potential risks that the entered text or electronic transaction may pose.

[1265] "Risk assessment results" refer to the specific evaluation value or judgment result of transmission risk calculated by the analysis device.

[1266] "Electronic transactions" refer to transactions such as buying and selling property or services, remittances, and contracts conducted using telecommunications means.

[1267] "Fraudulent use risk" refers to the assessment of the risk of unauthorized or fraudulent activity in electronic transactions.

[1268] "Notification" refers to a means of communicating information to users regarding the results of a risk assessment.

[1269] The present invention is a system that automatically analyzes transaction details and assesses the risk of fraudulent use when a user conducts an electronic transaction. This system includes means for acquiring transaction details entered by the user, means for transmitting the acquired transaction details to an analysis device, means for analyzing the transaction details using the analysis device, means for comparing the analysis results with a past analysis database, means for evaluating the risk of fraudulent use of the transaction based on the comparison results, and means for displaying the risk assessment results to the user and confirming the transaction. Specific embodiments are described below.

[1270] The terminal retrieves transaction details and text entered by the user and sends them to the server. The server passes the retrieved transaction details and text to an analysis device, which then analyzes them. The analysis device uses data analysis software (e.g., Python or Scikit-learn) to evaluate the tone and keywords of the transaction details and text. Once the tone and keywords are evaluated, the analysis device compares the results to a database of past analyses. This database stores previously analyzed data, including past incidents of online controversies and instances of misuse.

[1271] The analysis device evaluates the risk of fraudulent use of a transaction based on the comparison results. The server receives these evaluation results and sends them back to the terminal. The terminal displays the risk evaluation results to the user, and if it is determined to be high risk, it prompts the user to confirm the transaction. For example, a message such as "This transaction is high risk. Do you really want to proceed?" is displayed. The transaction is executed only if the user confirms.

[1272] As a concrete example, consider a case where a user makes a transaction to "transfer 500,000 yen to purchase a car." The terminal sends this transaction details to the server, which then passes them to an analysis device. The analysis device evaluates the risk of fraudulent use of this transaction, and if it determines that the risk is high, the server sends the evaluation result back to the terminal. The terminal displays a message to the user saying, "This transaction is high risk. Do you really want to proceed?", and the transaction is executed only if the user confirms.

[1273] Example of a prompt:

[1274] I will send 500,000 yen to buy a car.

[1275] In this way, the present invention enables users to conduct electronic transactions securely. By preventing the risk of fraudulent use, a highly reliable payment service can be provided.

[1276] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1277] Step 1:

[1278] The terminal retrieves transaction details and text entered by the user. When the user enters transaction details and presses the send button, the terminal retrieves that input. The input is processed as text data. For example, the user enters the text "I will send 500,000 yen to buy a car." This input data is then sent to the next step.

[1279] Step 2:

[1280] The terminal sends the acquired transaction details and text to the server. Here, the text data entered by the user is sent to the server. For example, the input "I will send 500,000 yen to buy a car" is sent to the server.

[1281] Step 3:

[1282] The server passes transaction details and text received from the terminal to the analysis device, which then analyzes the content. The analysis device uses data analysis software (such as Python or Scikit-learn) to evaluate the tone and keywords of the transaction details and text. The input for the analysis is text data from the user, and the output is the analysis result. As a specific example, the sentence "I will send 500,000 yen to buy a car" is analyzed, and its tone and keywords are evaluated.

[1283] Step 4:

[1284] The server's analysis system compares the analysis results with a database of past analysis data. This database stores previously analyzed data and examples of fraudulent use. The analysis system compares the new analysis results with past data in the database and evaluates the similarity. The input for the comparison is the latest analysis result, and the output is the risk assessment result. As a specific example, the analyzed sentence, "I will send 500,000 yen to buy a car," is compared with past examples of fraudulent use.

[1285] Step 5:

[1286] The server evaluates the risk of fraudulent use of a transaction based on the comparison results. The evaluation result will be one of the following: high risk, medium risk, or low risk. This risk evaluation result is then sent to the next step. For example, a transaction such as "transferring 500,000 yen to purchase a car" might be evaluated as high risk.

[1287] Step 6:

[1288] The server sends the risk assessment results back to the terminal. The output from the server is the risk assessment result, which the terminal receives. For example, a high-risk assessment result is sent to the terminal.

[1289] Step 7:

[1290] The terminal displays the risk assessment results from the server to the user and confirms the transaction. If the risk is high, a message is displayed prompting the user to confirm whether to proceed with the transaction. The output is the message to the user, and the input is the user's response. For example, the message "This transaction is high risk. Do you really want to proceed?" is displayed, and the user responds "Yes".

[1291] Step 8:

[1292] The transaction is executed only after the user confirms the transaction. When the user enters a response, the terminal sends that response to the server, instructing it to execute the transaction. The input is the user's confirmation response, and the output is the final transaction execution. For example, if the user responds "yes," a transfer of 500,000 yen is executed.

[1293] 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.

[1294] The present invention is a system that automatically analyzes the content of posts made by users on social media and evaluates the risk of online backlash and emotional impact. The system includes means for acquiring text entered by the user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for recognizing the emotions of the text using an emotion engine, means for comparing the analysis results with a past analysis database, means for evaluating the risk of sending the text based on the comparison results and emotion recognition results, and means for displaying the risk evaluation results to the user and confirming the transmission.

[1295] System Overview

[1296] 1. User actions:

[1297] The user enters text on the social media post creation screen and presses the send button. For example, they might enter the text "This new product is completely useless" and press the send button.

[1298] 2. Retrieving and sending the posted content:

[1299] The device retrieves the text entered by the user and sends it to the server. During this process, the string data of the posted content is sent as an API request.

[1300] 3. Content analysis and emotion recognition:

[1301] The server receives the posted content sent from the terminal. The received data is passed to the analysis device and emotion engine, which analyze and recognize its content.

[1302] The analysis device evaluates the tone and keywords of the text, while the emotion engine evaluates the emotions of the text. For example, it evaluates that the phrase "completely useless" is negative and contains pessimistic emotions.

[1303] 4. Comparison and Risk Assessment:

[1304] The server's analysis system compares the analysis results with a database of past online controversies. Simultaneously, it evaluates the results, including those from the sentiment engine. For example, if the phrase "completely useless" matches many past negative posts and contains pessimistic emotions, it is rated as high risk.

[1305] 5. Displaying results and confirming submission:

[1306] The server assesses the risk based on the comparison results and sentiment recognition results, and sends the results back to the terminal. The risk assessment result is one of the following: "high risk," "medium risk," or "low risk."

[1307] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[1308] 6. User verification and submission:

[1309] Users respond to the displayed message with either "yes" or "no." If they respond with "no," the post is canceled; if they respond with "yes," the post is sent only.

[1310] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[1311] Specific example

[1312] The user enters the sentence, "This new product is completely useless."

[1313] The terminal sends that input to the server.

[1314] The server analyzes the sentence and determines that the phrase "completely useless" is very negative, and the emotion engine recognizes it as a pessimistic emotion.

[1315] The server compares the analysis results and emotion recognition results with a database of past online controversies and assesses them as high-risk due to numerous similarities.

[1316] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[1317] The post will only be sent if the user answers "yes".

[1318] Precautions

[1319] This system automatically performs risk assessment and sentiment recognition before a user makes a social media post, and then decides whether or not to post based on the results. This process makes it possible to prevent users from unintentionally making posts that could cause controversy. Furthermore, the system only requests user confirmation if it determines that the post is high-risk; otherwise, it supports rapid information dissemination.

[1320] The following describes the processing flow.

[1321] Step 1:

[1322] The user enters the content of their post on the SNS post creation screen and presses the send button. This action retrieves the post content to the device. For example, they might enter the sentence, "This new product is completely useless."

[1323] Step 2:

[1324] The device detects when the user presses the send button and sends the entered content to the server. At this time, the string data of the posted content is sent to the server as an API request.

[1325] Step 3:

[1326] The server receives the content of the post sent from the terminal. It first saves the received data in order to pass it on to the analysis device and emotion engine.

[1327] Step 4:

[1328] The analysis system on the server analyzes the content of the received posts. During this process, the tone, keywords, and emotions of the writing are evaluated. For example, the phrase "completely useless" is evaluated as having a negative tone.

[1329] Step 5:

[1330] The emotion engine on the server recognizes emotions from the content of the received posts. In doing so, it judges the emotional nuance of the entire text and understands the user's emotional state. For example, it recognizes that the phrase "completely useless" indicates a pessimistic emotion.

[1331] Step 6:

[1332] The server compares the analysis results obtained by the analysis device with a database of past online controversies. At the same time, it also takes into account the results of the emotion engine to perform an overall risk assessment. For example, the phrase "completely useless" matches many past negative posts and contains pessimistic emotions, so it is assessed as high risk.

[1333] Step 7:

[1334] The server assesses the risk of the posted content based on the comparison results and sentiment recognition results. The risk assessment result will be one of the following: "high risk," "medium risk," or "low risk."

[1335] Step 8:

[1336] The server sends the risk assessment results back to the terminal. The risk assessment results are sent to the terminal.

[1337] Step 9:

[1338] The device displays the risk assessment results received from the server to the user. In particular, if the assessment is high risk, it displays a message asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[1339] Step 10:

[1340] Users respond to the displayed message with either "Yes" or "No". If they respond with "No", the post is canceled. If they respond with "Yes", the post is sent.

[1341] Step 11:

[1342] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[1343] By following these steps, users can be prevented from unconsciously making posts that pose a risk of causing controversy. Furthermore, by using an emotion engine, the emotional nuances of the text can be supplemented, enabling a more accurate risk assessment.

[1344] (Example 2)

[1345] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1346] When users post on social media, there is a risk of online backlash due to inappropriate content or emotionally charged posts. It is necessary to assess this risk in advance and take preventative measures. However, manual verification is time-consuming and prone to errors. There is a need for automated methods to accurately assess risks and provide appropriate warnings and confirmations.

[1347] 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.

[1348] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to the server, means for the server to pass the received text to an analysis device and an emotion engine, means for the analysis device to analyze the tone and keywords of the text, means for the emotion engine to evaluate the emotion of the text, means for comparing the analysis results and emotion evaluation results with a database of past online controversies, means for evaluating the risk of transmitting the text based on the comparison results and evaluation results, and means for transmitting the risk evaluation results to the user terminal and confirming the transmission based on the risk evaluation results. This makes it possible to automatically analyze the content of a post before the user posts it and evaluate the risk of online controversy and the risk of emotional impact.

[1349] A "user" is an individual or group that uses an information system or application.

[1350] "Text" refers to the text data that users input to post on social media, etc.

[1351] A "terminal" refers to a device such as a computer or smartphone that is operated by a user.

[1352] A "server" is a central system that receives requests from terminals via a network and performs analysis and processing.

[1353] An "analysis device" refers to software or hardware within a server that analyzes the tone and keywords of a text.

[1354] An "emotion engine" is a technological element used to evaluate emotions from the content of a text.

[1355] The "Database of Past Online Controversies" is a data store that saves and manages information on posts and incidents that have caused online controversies in the past.

[1356] "Risk assessment" is the process of evaluating the likelihood of a post going viral or causing a backlash, based on results obtained from analytical devices and emotion engines.

[1357] "Confirmation of submission" is the act of notifying the user of the risk assessment results and leaving the final decision of whether or not to submit the post to the user.

[1358] This invention is a system that analyzes text posted on social media in advance and evaluates the risk of online backlash and emotional impact. Specific embodiments for implementing this invention are described in detail below.

[1359] 1. User actions

[1360] The user enters text on the post creation screen of the SNS application and presses the send button. The entered text is retrieved in real time by the device.

[1361] 2. Retrieving and sending the posted content

[1362] The terminal retrieves text entered by the user and sends it to the server. Specifically, it encodes the retrieved text data into JSON format and sends it to the server as an API request using the HTTPS protocol. For example, if the user enters "This new product is completely useless," that text data will be sent to the server.

[1363] 3. Content Analysis and Emotion Recognition

[1364] The server receives the content of posts sent from the terminal. The received data is passed to the analysis device and sentiment engine, which analyze and recognize its content. The analysis device uses text analysis libraries (e.g., NLTK or spaCy) to evaluate the tone and keywords of the text. The sentiment engine uses Azure Cognitive Services or Google NLP API to evaluate the sentiment of the text. For example, the phrase "completely useless" is evaluated as negative and contains pessimistic emotions.

[1365] 4. Comparison with historical data and risk assessment

[1366] The server's analysis system compares the obtained analysis results with a database of past online controversies. This database is a data store that stores and manages information on posts and incidents that have caused online controversies in the past. The analysis results and sentiment evaluation results are comprehensively evaluated to assess the risk of sending the post. For example, the phrase "completely useless" is similar to many past online controversies and is therefore rated as high risk.

[1367] 5. Return and display of risk results

[1368] The server sends the risk assessment results back to the terminal. The risk assessment results are one of the following: "high risk," "medium risk," or "low risk." The terminal displays the risk assessment results received from the server to the user. For example, if the risk assessment is determined to be high risk, a message will be displayed asking the user for confirmation, such as, "This post has a high risk of causing a social media firestorm. Do you really want to post it?"

[1369] 6. User verification and submission

[1370] The user responds to the displayed message with "yes" or "no". Only if the user responds with "yes" will the device send the post to social media. This will make the post actually public on social media. If the user responds with "no", the post will be canceled.

[1371] Specific example

[1372] This explains what happens when a user types "This new product is completely unusable" and presses the submit button.

[1373] 1. User input:

[1374] A user types "This new product is completely useless" on the social media app's posting screen and presses the send button.

[1375] 2. Retrieving and sending the posted content:

[1376] The device retrieves the entered text "This new product is completely unusable," encodes it in JSON format, and sends it to the API endpoint (e.g., https: / / api.example.com / evaluate).

[1377] 3. Content analysis and emotion recognition:

[1378] The server receives the API request and sends the text to the analysis device and emotion engine.

[1379] The analysis device recognizes "completely useless" as a negative phrase, and the emotion engine evaluates it as a pessimistic emotion.

[1380] 4. Comparison with historical data and risk assessment:

[1381] The server queries a database of past online controversies to compare the analysis results.

[1382] Due to the numerous similarities observed, it is rated as "high risk."

[1383] 5. Return and display of risk results:

[1384] The server returns a "high risk" assessment result to the terminal.

[1385] The device displays a message to the user saying, "This post has a high risk of causing controversy. Do you really want to post it?"

[1386] 6. User verification and submission:

[1387] The user clicks "Yes".

[1388] The device sends the content of the post to the SNS API, and the post is made public.

[1389] Example prompt

[1390] Prompt: "This system evaluates the risk of controversy and emotional impact before posting a message on social media and warns the user. Please analyze the following message and perform a risk assessment: 'This new product is completely useless.'"

[1391] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1392] Program processing flow

[1393] Step 1: User input

[1394] The user enters text on the social media post creation screen and presses the send button.

[1395] Input: User input on the SNS post creation screen (e.g., "This new product is completely unusable").

[1396] Output: Text data based on user input

[1397] Step 2: Retrieve the posted content

[1398] The device retrieves the text entered by the user in real time.

[1399] Input: Text data entered by the user

[1400] Output: Acquired text data (text entered by the user)

[1401] Step 3: Submit your post

[1402] The device sends the retrieved post content to the server.

[1403] Input: Acquired text data

[1404] Data processing: Encoding text data into JSON format

[1405] Output: Data in JSON format (Example: "{"text": "This new product is completely unusable"}")

[1406] Step 4: Content Analysis and Sentiment Recognition

[1407] The server receives the posted content sent from the terminal and passes it on to the analysis device and emotion engine.

[1408] Input: Text data in JSON format

[1409] Data processing and data computation:

[1410] The analysis device uses a text analysis library (e.g., NLTK or spaCy) to analyze the tone and keywords of the text.

[1411] The emotion engine uses emotion analysis APIs (e.g., Azure Cognitive Services or Google NLP API) to evaluate the sentiment of a text.

[1412] Output: Analysis results and emotion evaluation results (e.g., "Negative," "Pessimistic emotion")

[1413] Step 5: Comparison with historical data and risk assessment

[1414] The server's analysis device compares the obtained analysis results and sentiment evaluation results with a database of past online controversies to perform a risk assessment.

[1415] Input: Analysis results and sentiment evaluation results

[1416] Data processing and calculation: A process of comparing results obtained with a database of past online crises to perform risk assessment.

[1417] Output: Risk assessment results (e.g., "High Risk", "Medium Risk", "Low Risk")

[1418] Step 6: Return and display of risk results

[1419] The server sends the risk assessment results back to the terminal, which then displays them to the user.

[1420] Input: Risk assessment results

[1421] Data processing: Encode risk assessment results into JSON format.

[1422] Output: Risk assessment results sent to the terminal (e.g., "High Risk")

[1423] Step 7: User verification and submission

[1424] The user responds to the displayed message with either "yes" or "no".

[1425] Input: User confirmation input ("Yes" or "No")

[1426] Output: Results based on the user's "yes" or "no" answer.

[1427] Specific actions:

[1428] If the user answers "yes," the device sends the post content to the SNS API, and the post is made public.

[1429] If the user answers "No," the post will be canceled.

[1430] Through the above process, this system can automatically perform risk assessment and sentiment recognition before a user makes a social media post, and based on the results, decide whether or not to proceed with the post.

[1431] (Application Example 2)

[1432] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1433] In food delivery services, emotional and negative reviews and comments posted by users can lead to unfair evaluations of service providers, negative impacts on other users, and even a decline in the overall reputation of the service. To prevent such posts, a system is needed that analyzes review content before posting and appropriately assesses the risks involved.

[1434] 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.

[1435] In this invention, the server includes means for acquiring text entered by a user, means for transmitting the acquired text to an analysis device, means for analyzing the content of the text using the analysis device, means for recognizing the sentiment of the text using a sentiment engine, means for comparing the analysis results with a past analysis database, means for evaluating the risk of transmitting the text based on the comparison results and sentiment recognition results, means for displaying a warning message to the user if the risk is judged to be high based on the risk evaluation results, means for displaying the risk evaluation results to the user and confirming transmission, and means for transmitting the text only if the user performs a confirmation operation. This makes it possible to prevent the posting of unintended negative reviews by users and maintain trust between food delivery service providers and users.

[1436] "Method for obtaining user-entered text" refers to a system that automatically retrieves the content of reviews and comments that users enter for food delivery services.

[1437] "Means for transmitting acquired text to an analysis device" refers to a system that transfers the content of input reviews and comments to an analysis device in a specific format.

[1438] "Means of analyzing the content of text using an analysis device" refers to a system in which an analysis device processes the content of transmitted reviews and comments as data and understands its content and meaning.

[1439] "A means of recognizing the emotions of a text using an emotion engine" refers to an engine for evaluating and identifying the emotions contained in an analyzed text, and a mechanism for determining the emotional tone and content of the text.

[1440] "Means for comparing analysis results with past analysis databases" refers to a system that compares and evaluates the content and sentiment of analyzed texts against a database of past reviews and comments.

[1441] "Means for evaluating the risk of sending text based on comparison results and sentiment recognition results" refers to a mechanism that evaluates the potential risks that reviews and comments may pose based on the results of comparison and sentiment recognition.

[1442] "A means of displaying a warning message to the user when a high risk is determined based on the risk assessment results" refers to a mechanism for informing and warning the user of a risk when it is assessed as high risk.

[1443] "A means of displaying risk assessment results to the user and confirming submission" refers to a mechanism that shows the assessment results to the user and allows them to reconfirm the submission of reviews and comments.

[1444] "A method of sending text only after the user has performed a confirmation action" means that a review or comment is actually posted only if the user confirms that they have responded to a warning message.

[1445] This invention is a system that analyzes review text entered by users and evaluates the risk of online backlash and emotional impact when a review is posted for a food delivery service. This system is implemented by the following means:

[1446] 1. User actions:

[1447] The user enters text on the food delivery service review submission screen and clicks the submit button. For example, they might enter the text, "This restaurant's delivery was terrible. It was late and the food was cold," and then click the submit button.

[1448] 2. Retrieving and sending the posted content:

[1449] The device (smartphone) retrieves the text data of the review entered by the user and sends it to the server. At this time, the string data of the posted content is sent as an API request.

[1450] 3. Content analysis and emotion recognition:

[1451] The server receives the posted content sent from the terminal. The received data is passed to the analysis device and emotion engine, which analyze and recognize its content.

[1452] The analysis device uses Natural Language Processing (NLP) technology to evaluate the tone and keywords of the text. The sentiment engine uses a generative AI model (e.g., BERT or GPT-3) to evaluate the sentiment of the text.

[1453] 4. Comparison and Risk Assessment:

[1454] The server's analysis system compares the analysis results with a database of past online controversies. Simultaneously, it evaluates the results, including those from the emotion engine. For example, phrases like "it was the worst" and "they were cold" are considered highly negative, and the emotion engine recognizes them as indicating pessimistic emotions.

[1455] Based on the comparison results and sentiment recognition results, the risk of sending a message is classified into high-risk, medium-risk, and low-risk categories.

[1456] 5. Displaying results and confirming submission:

[1457] The server returns the risk assessment results to the terminal. The risk assessment results will be one of the following: "high risk," "medium risk," or "low risk."

[1458] The device displays the risk assessment results received from the server to the user. If the assessment is high risk, a message is displayed asking the user for confirmation, such as, "This post has a high risk of causing a controversy. Do you really want to post it?"

[1459] 6. User verification and submission:

[1460] Users respond to the displayed message with either "Yes" or "No". If they respond with "No", the post is canceled. If they respond with "Yes", the post is sent.

[1461] If the user answers "yes," the device sends the post content to social media. This makes the post actually public.

[1462] For example, when a user enters a review stating, "This restaurant's delivery was terrible. It was late and the food was cold," the system retrieves the text and sends it to the server. The server analyzes the review using a sentiment engine (e.g., GPT-3) and a historical database, determines the rating to be "high risk," and displays a warning message on the user's device. The post is only actually sent if the user replies "yes."

[1463] Example of a prompt:

[1464] "Review content: 'This restaurant's delivery was terrible. It was late and the food was cold.' Please assess the risk of this review causing a social media firestorm and evaluate the emotional risk."

[1465] Thus, by using the system of the present invention, it is possible to check in advance whether user reviews posted on food delivery services are sound and to prevent misunderstandings and inappropriate posts.

[1466] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1467] Step 1:

[1468] The user enters the text of their review on the food delivery service's review submission screen and presses the submit button. The user enters the text, and the text data is temporarily saved on the user's device.

[1469] Step 2:

[1470] The terminal retrieves the text entered by the user and sends this string data to the server as an API request. Here, the entered text is converted into a request format and sent to the server.

[1471] Step 3:

[1472] The server receives the posted content sent from the terminal and passes it to the analysis device. Processing takes place from the time the input text data reaches the server until it is passed to the analysis device.

[1473] Step 4:

[1474] The analysis device on the server analyzes the received text using Natural Language Processing (NLP) technology. Specifically, it evaluates the tone, keywords, and context of the text and generates analysis results. The input is text data, and the output is data of the analysis results.

[1475] Step 5:

[1476] The emotion engine evaluates the sentiment of a text based on text data obtained from an analysis device. Using a generative AI model (e.g., GPT-3), it identifies the emotional tone within the text and outputs the evaluation result. The input is the analyzed text data, and the output is the sentiment recognition result.

[1477] Step 6:

[1478] The server's analysis system compares the analysis results and sentiment recognition results with a database of past online controversies to perform a risk assessment. It compares the analysis results with past data and determines the risk level based on similarities and matches. The input is the analysis results and sentiment recognition results, and the output is the risk assessment result.

[1479] Step 7:

[1480] The server sends the risk assessment results back to the terminal. If the assessment result is "high risk," "medium risk," or "low risk," the assessment result data is sent to the terminal.

[1481] Step 8:

[1482] The terminal displays the risk assessment results received from the server to the user. In particular, in cases of high risk, it displays a message such as, "This post has a high risk of causing a controversy. Do you really want to post it?" The input is the risk assessment result, and the output is the display of the warning message.

[1483] Step 9:

[1484] The user responds to the displayed message with "yes" or "no." Based on this input, it is decided whether or not to send the post. The input is the user's response, and the output is an instruction to send the post.

[1485] Step 10:

[1486] If the user answers "yes," the device sends the content of the post to the food delivery service's review system, and the review is published. The input is the user's "yes" response, and the output is the actual posted review content.

[1487] Through the above processing steps, it is ensured that user reviews are properly evaluated and submitted under low-risk conditions.

[1488] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 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.

[1489] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">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.

[1490] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1491] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1492] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1493] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1494] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1495] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1496] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1497] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1498] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1499] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1500] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1501] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1502] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1503] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1504] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1505] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1506] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1507] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1508] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[1509] The following is further disclosed regarding the embodiments described above.

[1510] (Claim 1)

[1511] A means of obtaining text entered by the user,

[1512] A means for transmitting the acquired text to an analysis device,

[1513] A means of analyzing the content of a document using an analysis device,

[1514] A means of comparing past analysis databases with analysis results,

[1515] A means of evaluating the risk of sending a document based on the comparison results,

[1516] A means of displaying the risk assessment results to the user and confirming submission,

[1517] A system that includes this.

[1518] (Claim 2)

[1519] The system according to claim 1, further comprising means for prompting the user to confirm transmission if the analysis results are determined to be high-risk.

[1520] (Claim 3)

[1521] The system according to claim 1, further comprising means for sending a message without user confirmation if the risk assessment result is determined to be low-risk or medium-risk.

[1522] "Example 1"

[1523] (Claim 1)

[1524] A means of obtaining text entered by the user,

[1525] A means for transmitting the acquired text to an information processing device,

[1526] A means of analyzing the content of a document using an information processing device,

[1527] A means of comparing past evaluation information with analysis results,

[1528] A means of evaluating the risk of sending a document based on the comparison results,

[1529] A means of displaying the risk assessment results to the user and confirming submission,

[1530] A system that includes this.

[1531] (Claim 2)

[1532] The system according to claim 1, further comprising means for prompting the user to confirm transmission if the analysis results are determined to be high-risk.

[1533] (Claim 3)

[1534] The system according to claim 1, further comprising means for sending a message without user confirmation if the risk assessment result is determined to be low risk or medium risk.

[1535] "Application Example 1"

[1536] (Claim 1)

[1537] A means of obtaining the text entered by the user,

[1538] Means for transmitting the acquired text to the analysis device,

[1539] A means of analyzing the content of a sentence using an analysis device,

[1540] A means of comparing past analysis databases with analysis results,

[1541] A means of evaluating the risk of sending a message based on the comparison results,

[1542] A means of displaying the risk assessment results to the user and confirming submission,

[1543] A means for obtaining details of electronic transactions and transmitting them to an analysis device,

[1544] The analytical device provides a means for evaluating the risk of fraudulent use of transactions,

[1545] A means of notifying the user of the evaluation results and requesting their confirmation,

[1546] A system that includes this.

[1547] (Claim 2)

[1548] The system according to claim 1, further comprising means for prompting the user to confirm transmission if the analysis results are determined to be high-risk.

[1549] (Claim 3)

[1550] The system according to claim 1, further comprising means for sending a message without user confirmation if the risk assessment result is determined to be low-risk or medium-risk.

[1551] "Example 2 of combining an emotion engine"

[1552] (Claim 1)

[1553] A means of obtaining text entered by the user,

[1554] A means of sending the acquired text to the server,

[1555] A means for passing the text received by the server to an analysis device and an emotion engine,

[1556] A means of analyzing the tone and keywords of a text using an analysis device,

[1557] A means of evaluating the emotion of a text using an emotion engine,

[1558] A means of comparing the analysis results and sentiment evaluation results with a database of past online controversies,

[1559] A means for evaluating the risk of sending a document based on comparison results and evaluation results,

[1560] A means for sending risk assessment results to the user terminal and confirming transmission based on the risk assessment results,

[1561] A system that includes this.

[1562] (Claim 2)

[1563] The system according to claim 1, further comprising means for prompting the user to confirm submission if the analysis results and sentiment evaluation results are determined to be high risk.

[1564] (Claim 3)

[1565] The system according to claim 1, further comprising means for sending a message without requiring user confirmation if the risk assessment result is determined to be low-risk or medium-risk.

[1566] "Application example 2 when combining with an emotional engine"

[1567] (Claim 1)

[1568] A means of obtaining text entered by the user,

[1569] A means for transmitting the acquired text to an analysis device,

[1570] A means of analyzing the content of a document using an analysis device,

[1571] A means of recognizing the emotions in text using an emotion engine,

[1572] A means of comparing past analysis databases with analysis results,

[1573] A means for evaluating the risk of sending a message based on comparison results and sentiment recognition results,

[1574] A means of displaying the risk assessment results to the user and confirming submission,

[1575] If a risk assessment determines that the risk is high, a means of displaying a warning message to the user is provided.

[1576] The means of sending a message is to be used only when the user performs a confirmation action.

[1577] A system that includes this.

[1578] (Claim 2)

[1579] The system according to claim 1, further comprising means for prompting the user to confirm transmission if the analysis results are determined to be high-risk.

[1580] (Claim 3)

[1581] The system according to claim 1, further comprising means for sending a message without user confirmation if the risk assessment result is determined to be low-risk or medium-risk. [Explanation of Symbols]

[1582] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of obtaining text entered by the user, A means for transmitting the acquired text to an analysis device, A means of analyzing the content of a document using an analysis device, A means of comparing past analysis databases with analysis results, A means of evaluating the risk of sending a document based on the comparison results, A means of displaying the risk assessment results to the user and confirming submission, A system that includes this.

2. The system according to claim 1, further comprising means for prompting the user to confirm transmission if the analysis results are determined to be high-risk.

3. The system according to claim 1, further comprising means for sending a message without user confirmation if the risk assessment result is determined to be low risk or medium risk.

Images