System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting posted comment data; means for analyzing the collected comment data with a natural language processing engine and calculating an aggression score; means for storing the calculated aggression score in a database; means for setting a filter strength by a user; means for filtering comments based on the filter strength; and means for sending a filtered comment list to the user's terminal.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Currently, many news sites face the challenge of posting a large number of offensive comments in their comment sections. This can make users feel uncomfortable and hinder healthy discussion. This problem can ruin the user experience, discourage them from using the site, and ultimately lead to a decrease in advertising revenue. Therefore, it is necessary to effectively filter offensive comments and provide an environment where users can safely view and participate in comments. [Means for solving the problem]

[0005] To solve this problem, the present invention provides the following means. First, a means for collecting posted comment data is provided. Next, a means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score is provided. Next, a means for saving the calculated offensiveness score in a database is provided, and a means for the user to set the filter strength is provided. Next, a means for filtering comments based on the filter strength is provided. Finally, a means for sending the filtered comment list to the user's terminal is provided. Furthermore, the present invention includes a means for dynamically updating the comment section displayed according to the set filter strength, and a means for recalculating the offensiveness score and updating the filtering results when a new comment is posted. This configuration improves the user experience and provides a healthy comment environment.

[0006] "Comment data" refers to the text information and metadata posted by users in response to news articles, etc.

[0007] A "natural language processing engine" refers to a software component that has the ability to analyze text data and understand or classify its content.

[0008] "Aggression score" refers to a numerical representation of the degree to which comment data is aggressive toward other users.

[0009] "Database" refers to the information management system for storing collected comment data, offensiveness scores, and related metadata.

[0010] "Filter strength" refers to the threshold set by the user for the offensiveness score of comments to be displayed.

[0011] "Filtering" refers to the process of deciding whether to display or hide a comment based on its offensiveness score.

[0012] "User Device" refers to the device (computer, smartphone, tablet, etc.) through which a User accesses the comments section of a news article.

[0013] "Dynamic updating" refers to the process of recalculating and reflecting new comments or changes to filter settings in real time or near real time to the displayed comment list. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0015] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0018] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0019] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0020] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0025] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0027] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0032] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0033] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0034] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0035] The present invention provides a system for filtering offensive comments in the comment sections of news sites, with the goal of creating an environment where users can safely view and participate in comments.

[0036] First, the server collects all posted comment data. This data collection includes metadata such as the comment content, the poster's user ID, and the posting date and time. For example, if user A posts a comment saying, "This news is ridiculous!", the content of that comment will be recorded by the server.

[0037] The server then passes the collected comment data to a natural language processing engine, which calculates an "offense score" for each comment. For example, the natural language processing engine calculates an offensive score of 70 for a comment such as "This news is ridiculous!" On the other hand, a comment such as "That was very interesting" would receive an offensive score of 5.

[0038] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[0039] Next, the user (visitor) sets the filter strength to their preference on the device. The filter strength (threshold of the offensive score) can be input or selected through the user interface. For example, if user B sets the filter strength to "20," this setting is confirmed on the device.

[0040] The server then filters the comments based on the filter strength set by the user. This process hides any comments with an offensiveness score above the set threshold. For example, if user B's filter strength is "20," any comments with an offensiveness score of 20 or higher will be hidden.

[0041] Finally, the server sends the filtered comment list to the user's device. This allows the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User B's device, a comment with an offensiveness score of 25, such as "This is a really awful story," will be hidden, while a comment with an offensiveness score of 10, such as "That was very interesting," will be displayed.

[0042] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[0043] This configuration allows users to freely adjust the filter strength and customize their comment viewing experience, promoting healthy discussion and protecting users from offensive comments, improving the overall user experience on news sites.

[0044] The processing flow will be explained below.

[0045] Step 1:

[0046] The server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. All posted comments are recorded in a database.

[0047] Example: When user A posts a comment saying "This news is ridiculous!", the server records the comment content and poster information in a database.

[0048] Step 2:

[0049] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment.

[0050] Example: For the comment "This news is ridiculous!", the natural language processing engine calculates an offensive score of "70".

[0051] Step 3:

[0052] The server associates the offensiveness score calculated by the natural language processing engine with the comment data and stores it in a database.

[0053] Example: Add an offensive score of 70 to the record of the comment "This news is stupid!" and save it in the database.

[0054] Step 4:

[0055] Users (visitors) can set their own preferred filter strength on their devices by inputting the filter strength (threshold of the offensiveness score) through the user interface.

[0056] Example: When user B sets the filter strength to "20", the setting is confirmed on the device.

[0057] Step 5:

[0058] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it refers to the offensiveness score of the comment and hides comments that exceed the standard.

[0059] Example: For user B who has set the filter strength to "20", comments with an offensive score of 20 or more will be hidden.

[0060] Step 6:

[0061] The server sends the filtered comment list to the user's terminal.

[0062] Example: Send a comment list to user B's device that includes only comments with an offensiveness score of less than 20.

[0063] Step 7:

[0064] The device updates the comment section based on the filtering results received from the server, allowing users to view only the filtered comments.

[0065] Example: On User B's device, comments with an offensiveness score of "25" are hidden, and comments with an offensiveness score of "10" are displayed.

[0066] Step 8:

[0067] When a new comment is posted or a user changes the filter strength, the server recalculates the offensiveness score and dynamically updates the filtering results.

[0068] Example: When a new comment "This is awful!" is posted, its offensiveness score is calculated and the list of comments displayed is instantly updated based on the existing filter strength settings.

[0069] Example 1

[0070] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0071] In the comment sections of conventional news sites, offensive comments are common, making it difficult for users to safely view and participate in the comments. It is also difficult to adapt to the different filter strength preferences of each user, so a system that dynamically updates the comment filtering and display has been required.

[0072] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0073] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine to calculate an offensiveness score, means for saving the calculated offensiveness score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for sending a filtered comment list to the user's terminal, and means for dynamically reflecting the filter strength setting on the server. This allows users to filter comments according to their preferences in real time, and provides a comment section that promotes healthy discussion.

[0074] "Comment data" refers to the text that users post to news sites and the metadata associated with those posts.

[0075] A "natural language processing engine" refers to software or algorithms that analyze human language to understand its meaning and sentiment and then perform specific tasks.

[0076] "Aggression score" refers to a numerical indicator that indicates whether a posted comment is offensive or not.

[0077] "Users" refer to people who visit news sites, post comments, and set filter strengths.

[0078] "Filter strength" refers to the threshold for the offensiveness score of comments set by the user, which determines the range of comments that will be displayed.

[0079] "Database" refers to an electronic storage system for storing and managing collected comment data and offensiveness scores.

[0080] "Device" refers to the electronic device (e.g., computer, smartphone, tablet, etc.) that a user uses to access a news site.

[0081] "Dynamic reflection" means that user settings and new comment posts are applied to the system in real time.

[0082] MODE FOR CARRYING OUT THE INVENTION

[0083] The present invention provides a system for filtering offensive comments in the comment section of a news site. The system aims to create an environment where users can safely view and participate in comments. A specific embodiment of the system is described below.

[0084] First, this system consists of multiple hardware and software components: a server, a natural language processing engine, a database, a user interface (UI), and a terminal.

[0085] 1. Collection of comment data

[0086] The server collects all comment data posted on the news site. This collected data includes metadata such as the comment content, the poster's user ID, and the posting date and time. For example, if user A posts a comment saying "This news is ridiculous!", the content and metadata are recorded by the server.

[0087] 2. Calculating offensive scores using a natural language processing engine

[0088] The server passes the collected comment data to a natural language processing engine, which calculates an "offense score" for each comment. This natural language processing engine uses, for example, Google NLP API or OpenAI's GPT-3. For example, the natural language processing engine calculates an offensiveness score of 70 for a comment such as "This news is ridiculous!". On the other hand, a comment such as "That was very interesting" would be assigned an offensiveness score of 5.

[0089] 3. Save the calculated aggression score to the database

[0090] The server associates the calculated offensiveness score with each comment data and stores it in a database. This database can be an SQL database or a NoSQL database. For example, a comment such as "This news is ridiculous!" is associated with an offensiveness score of "70" and stored in the database.

[0091] 4. User-defined filter strength

[0092] Users can set their preferred filter strength through the user interface on their device. The UI is provided via a web browser or mobile app. Users can input or select the filter strength (threshold of the aggressiveness score). For example, if user B sets the filter strength to "20," this setting is sent to the server in real time.

[0093] 5. Filtering process and sending results

[0094] The server filters comments based on the filter strength set by the user. This process hides comments with an offensiveness score above the set threshold. For example, if user B's filter strength is "20," comments with an offensiveness score of 20 or more will be hidden. The filtered comment list is sent from the server to the user's device, where it can be viewed by the user.

[0095] Examples of concrete examples and prompts

[0096] Examples:

[0097] 1. User A posts a comment saying, "This news is great!"

[0098] 2. The server collects these comments and passes them to a natural language processing engine.

[0099] 3. The natural language processing engine calculates an offensiveness score of "1".

[0100] 4. The server associates this score with the comment data and stores it in a database.

[0101] 5. User B sets the filter strength to "10".

[0102] 6. Based on the filter strength of "10", the server sends only comments with an offensiveness score of "less than 10" to the user's device.

[0103] Example prompt:

[0104] We want to make comments on this news site safer and friendlier. If you see any of the following comments posted, please rate their offensiveness:

[0105] Comment: "This news is ridiculous!"

[0106] Comment: "It was very interesting."

[0107] You can also filter comments based on the following settings to view the comments that apply to each setting:

[0108] Filter Strength: 20

[0109] With this setting, which comments will be visible and which will be hidden?

[0110] This allows users to freely adjust the filter strength and customize their comment viewing experience to suit their preferences. It also promotes healthy discussion and protects users from offensive comments, improving the overall user experience on news sites.

[0111] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0112] Program processing flow

[0113] Step 1:

[0114] The server collects all comment data posted on the news site. This includes metadata such as the comment content, the poster's user ID, and the posting date and time. The input is a newly posted comment on the news site, and the output is the comment data stored in the server. For example, if user A posts "This news is stupid!", the content and metadata are recorded by the server.

[0115] Step 2:

[0116] The server passes the collected comment data to a natural language processing engine. The input is the comment data stored in the server, and the output is an offensiveness score for each comment. Specifically, the comment "This news is stupid!" is sent to the natural language processing engine, and the engine returns an offensiveness score of "70" for this comment.

[0117] Step 3:

[0118] The server associates the calculated offensiveness score with each comment data and stores it in the database. The input is the comment data with the calculated offensiveness score, and the output is the comment data stored in the database and its offensiveness score. For example, the comment "This news is ridiculous!" is associated with an offensiveness score of "70" and stored in the database.

[0119] Step 4:

[0120] The user sets the filter strength to their preference on the device. The input is the user's filter strength setting value, and the output is the setting data sent to the server. Specifically, when user B sets the filter strength to "20," this setting is confirmed on the device and sent to the server.

[0121] Step 5:

[0122] The server filters comments based on the filter strength set by the user. The input is the user's filter strength setting and the comment data in the database, and the output is a filtered comment list. Specifically, based on a filter strength of "20," comments with an offensiveness score of 20 or more are excluded.

[0123] Step 6:

[0124] The server sends the filtered comment list to the user's device. The input is the filtered comment list, and the output is the comments displayed on the user's device. For example, on User B's device, a comment with an offensiveness score of "10" such as "That was very interesting," is displayed, while a comment with an offensiveness score of "25" such as "That's a really awful story," is hidden.

[0125] (Application example 1)

[0126] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0127] Comment sections on news sites and social media often contain offensive comments that cause discomfort and psychological stress to users. It is important to effectively filter these comments and provide an environment where users can safely view and participate in the comments. However, existing systems face challenges in filtering offensive comments in real time and dynamically updating the user display. Furthermore, they lack the ability for users to adjust the filter strength according to their preferences, making it difficult to display comments that meet the needs of individual users.

[0128] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0129] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score, means for saving the calculated offensiveness score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for transmitting a filtered comment list to the user's display device, means for adjusting the filter strength on the user's display device, and means for maintaining communication with the display device to dynamically perform the filtering process. This allows users to filter offensive comments in real time and comfortably view comments with a filter strength that suits their preferences.

[0130] "Means for collecting posted comment data" refers to a system for collecting comments posted on the Internet by users.

[0131] A "natural language processing engine" is software or algorithm that analyzes collected comment data and understands and evaluates its content.

[0132] The "offensiveness score" is a numerical indication of the degree of offensiveness of a comment, and is an index calculated by a natural language processing engine.

[0133] "Means for storing in a database" refers to a system for centrally managing and storing calculated offensive scores and comment data.

[0134] The "means for setting filter strength" is an interface that allows the user to set the threshold of the aggressiveness score according to their own preferences.

[0135] The "means for filtering comments" is a mechanism for selecting comments to be displayed based on the set filter strength.

[0136] "Means for transmitting to a user's display device" is a mechanism for transmitting filtered comments to a device used by the user (e.g., smart glasses).

[0137] The "means for adjusting the filter strength" is an adjustment function that allows the filter strength changed by the user in real time to be immediately reflected.

[0138] The "means for maintaining communication with the display device" is a system for constantly maintaining a connection with the user's display device in order to dynamically perform comment filtering processing.

[0139] "Dynamic update means" refers to a function that changes the comment section displayed in real time depending on the set filter strength and newly posted comments.

[0140] This invention is a system for filtering offensive comments in the comment sections of news sites and social media. To implement this system, the following steps and various hardware and software are required.

[0141] Hardware and software used

[0142] Server: Used to collect, analyze, score, and store comment data in a database.

[0143] Natural language processing engine: Used to analyze the offensiveness of comments and calculate an offensiveness score.

[0144] Database: Used to store calculated aggression scores and comment data.

[0145] User display device (e.g., smart glasses): Used to display filtered comments.

[0146] HTTP request library (e.g. requests for Python): Used to communicate between the server and various APIs.

[0147] Overview of the overall system processing flow

[0148] 1. Collection of comment data

[0149] The server collects comment data posted from news sites and social media, including the comment text, the poster's user ID, and the posting date and time.

[0150] 2. Calculating the Aggression Score

[0151] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment.

[0152] 3. Saving to the database

[0153] The calculated offensiveness score is stored in a database together with the corresponding comment data.

[0154] 4. Setting the filter strength

[0155] Users set the filter strength (threshold of offensiveness score) for comments via a display device such as smart glasses, which is done through a user interface.

[0156] 5. Comment Filtering

[0157] The server filters the comments based on the filter strength set by the user and transmits only the selected comments to the display device.

[0158] 6. Dynamic Updates

[0159] The server performs filtering processing in real time, even when a new comment is posted or when a user changes the filter strength, and dynamically updates the filtering results while maintaining communication with the display device.

[0160] Specific examples

[0161] For example, suppose user A posts "This news is stupid!" in the comments section of a news article. This comment is collected by the server and passed to a natural language processing engine. The server then calculates an offensiveness score of 70 for this comment and stores it in the database.

[0162] Later, when User B uses the smart glasses to view a news article, he sets the filter strength to "20." Based on this setting, the server hides comments with an offensiveness score of 20 or more. Therefore, a comment such as "This news is ridiculous!" will not be displayed. On the other hand, a comment with an offensiveness score of "5," such as "This was very interesting," will be displayed.

[0163] Prompt Sentence Examples

[0164] Example configuration:

[0165] article_id = 67890 Example of article ID setting

[0166] filter_threshold = 50 Example of filter strength setting

[0167] Sample prompt sentence:

[0168] Input prompt: "Please filter and display comments for article ID "67890" using filter strength "50"."

[0169] Example output:

[0170] Displayed comment: "This is very useful information." (Aggression score: 10)

[0171] Comment to be hidden: "This article is rubbish!" (Aggression score: 60)

[0172] This system allows users to filter out offensive comments in real time, allowing them to comfortably view comments with a filter strength that suits their preferences.

[0173] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0174] Step 1:

[0175] The server collects comment data posted from news sites and social media. This collected data includes the comment text, the poster's user ID, the posting date and time, etc. The input requires the URL of the news article or social media post to be retrieved, and the collected comment data is obtained as output.

[0176] Step 2:

[0177] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment. This process requires the collected comment data as input and obtains an offensiveness score corresponding to each comment as output. Specific operations include making an HTTP request to the natural language processing engine API.

[0178] Step 3:

[0179] The server stores the calculated offensiveness score in a database. This process requires the comment data and the corresponding offensiveness score as input, and obtains the database state as output. Specific operations include creating and updating records in the database.

[0180] Step 4:

[0181] The user sets the filter strength (threshold of the aggression score) through the interface of the smart glasses. This process requires the user to input the filter strength and saves the setting as output. Specific actions include entering a number or operating a slider on the user interface.

[0182] Step 5:

[0183] The server filters comments based on the filter strength set by the user. This process requires the comment data and offensiveness scores stored in the database and the filter strength set by the user as input, and obtains a filtered comment list as output. Specific operations include a process of comparing the offensiveness score and the filter strength.

[0184] Step 6:

[0185] The server sends the filtered comment list to the user's display device. The filtered comment list is required as input, and the comments displayed on the user's display device are obtained as output. Specific operations include a process of sending data to the display device.

[0186] Step 7:

[0187] When a user sets a new filter strength, the server performs the filtering process again and sends the result to the display device. The new filter strength, existing comment data, and offensiveness scores are required as input, and the latest filtering result is obtained as output. Specific operations include the re-filtering and sending process.

[0188] Step 8:

[0189] When a new comment is posted, the server recalculates the offensiveness score of the comment, updates the database, and regenerates the filtering results and sends them to the display device. New comment data is required as input, and updated filtering results are obtained as output. Specific operations include automatic collection, score calculation, database update, and re-filtering.

[0190] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0191] The present invention provides a system that filters offensive comments in the comment section of a news site and provides an appropriate comment display environment that takes into account the user's feelings. The system aims to create an environment where users can view and participate in comments with peace of mind.

[0192] First, the server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. For example, if user A posts a comment saying, "This news is ridiculous!", the content of that comment is recorded by the server.

[0193] The server then passes the collected comment data to a natural language processing engine, which calculates an "aggression score" for each comment. For example, the natural language processing engine calculates an aggression score of 70 for a comment such as "This news is ridiculous!". The emotion engine also evaluates the poster's emotional state from the comment, and reflects the evaluation result in the aggression score. For example, a comment with a strong emotion of "anger" is assigned a higher aggression score.

[0194] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[0195] Next, the user (visitor) can set the filter strength to their preference on the device, or the emotion engine can recognize the user's current emotional state and automatically set the recommended filter strength. For example, User C can set the filter strength to "20," or the emotion engine can recognize User C's "irritation" and recommend the filter strength to "50."

[0196] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it references the offensiveness score of the comment and hides comments that exceed the standard. For example, for user C who sets the filter strength to "50", comments with an offensiveness score of 50 or more will be hidden.

[0197] The server then sends the filtered comment list to the user's device, allowing the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User C's device, comments with an offensiveness score of "70" are hidden, and comments with an offensiveness score of "30" are displayed.

[0198] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[0199] This configuration allows users to freely adjust the filter strength and enjoy a comment viewing experience that suits their preferences. It also improves the overall user experience of news sites by promoting healthy discussion and protecting users from offensive comments. Furthermore, the emotion engine provides appropriate filtering based on the user's emotional state, providing a more personalized viewing experience.

[0200] The processing flow will be explained below.

[0201] Step 1:

[0202] The server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. All posted comments are recorded in a database.

[0203] Example: When user A posts a comment saying "This news is ridiculous!", the server records the comment content and poster information in a database.

[0204] Step 2:

[0205] The server passes the collected comment data to a natural language processing engine, calculates an offensiveness score for each comment, and uses an emotion engine to evaluate the poster's emotional state based on the comment.

[0206] Example: In response to the comment "This news is ridiculous!", the natural language processing engine calculates an aggression score of "70" and the emotion engine evaluates it as "anger."

[0207] Step 3:

[0208] The server associates the aggression score calculated by the natural language processing engine and the emotional state evaluated by the emotion engine with the comment data and stores them in a database.

[0209] Example: Add an aggression score of 70 and emotion of anger to the record of the comment "This news is stupid!" and save it in the database.

[0210] Step 4:

[0211] Users (visitors) can set the filter strength to suit their preferences on their devices, or the emotion engine will recognize the user's current emotional state and automatically set the recommended filter strength.

[0212] Example: User C sets the filter strength to "20." Or, the emotion engine recognizes User C's "irritation" and recommends setting the filter strength to "50."

[0213] Step 5:

[0214] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it refers to the offensiveness score of the comment and hides comments that exceed the standard.

[0215] Example: For user C who has set the filter strength to "50", comments with an offensive score of 50 or more will be hidden.

[0216] Step 6:

[0217] The server sends the filtered comment list to the user's terminal.

[0218] Example: A comment list containing only comments with an offensiveness score of less than 50 is sent to user C's device.

[0219] Step 7:

[0220] The device updates the comment section based on the filtering results received from the server, allowing users to view only the filtered comments.

[0221] Example: On User C's device, comments with an aggression score of "70" are hidden, and comments with an aggression score of "30" are displayed.

[0222] Step 8:

[0223] When a new comment is posted or a user changes the filter strength, the server recalculates the offensiveness score and dynamically updates the filtering results.

[0224] Example: When a new comment "This is awful!" is posted, its offensiveness score is calculated and the list of comments displayed is instantly updated based on the existing filter strength settings.

[0225] Example 2

[0226] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0227] On traditional news sites, users are often exposed to offensive comments, making it difficult to ensure a comfortable viewing environment. In particular, the proliferation of offensive comments can hinder healthy discussion and increase the mental burden on users. Furthermore, traditional filtering systems are static and filter content based on uniform criteria without considering the user's emotional state, making it difficult to provide a personalized experience.

[0228] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0229] In this invention, the server includes a means for collecting posted comment data, a means for analyzing the collected comment data using a natural language processing engine to calculate an offensiveness score, and a means for analyzing the collected comment data using a sentiment analysis engine to evaluate the poster's emotional state. This enables filtering according to individual emotional states while preventing users from being exposed to offensive comments. The server also includes a means for storing the comment's offensiveness score and the emotional state evaluation results in a database, a means for the user to set the filter strength, and a means for recognizing the user's emotional state using the sentiment analysis engine and setting a recommended filter strength, thereby achieving flexible filtering according to individual user needs. The server also includes a means for transmitting a filtered comment list to the user's device and a means for recalculating the offensiveness score and updating the filtering results when a new comment is posted, providing a constantly up-to-date comment viewing environment. This improves the user experience across the news site and promotes healthy discussion.

[0230] "Posted comment data" refers to data that includes information such as the content of comments submitted by users of a news site, the poster's user ID, and the date and time of posting.

[0231] A "natural language processing engine" refers to a program or algorithm that analyzes text data and understands its content.

[0232] "Aggression score" refers to a numerical representation of the degree of aggression based on the content of a comment.

[0233] "Sentiment analysis engine" refers to a program or algorithm that identifies emotions contained in text data and evaluates that emotional state.

[0234] "User emotional state" refers to the current emotional state inferred from the user's input and actions.

[0235] "Filter strength" refers to the threshold of the aggressiveness score that serves as the basis for filtering.

[0236] A "database" refers to a system for systematically managing, storing, and searching information.

[0237] "Filtering" refers to the operation of selecting and displaying specific data based on pre-set criteria.

[0238] "User device" refers to a terminal, such as a PC, smartphone, or tablet, that receives and displays data sent from the server.

[0239] "Recalculation" refers to performing a new calculation process, and in this context specifically refers to recalculating the offensiveness score for newly posted comments.

[0240] "Dynamic update" refers to an operation in which information or data is automatically updated in real time or near real time.

[0241] "News site" refers to a website that provides news information on the Internet and allows users to post comments.

[0242] A "proper noun" is a name used to refer to a specific person, place, organization, etc., but in this context is written as a general name rather than a specific name.

[0243] The present invention is a system that filters offensive comments in the comment section of a news site and provides an appropriate comment display environment that takes into account the user's feelings. The purpose of this system is to create an environment where users can safely view and participate in comments.

[0244] First, the server collects all comment data posted on the news site. This collection includes information such as the comment content, the poster's user ID, and the posting date and time. For example, if User A posts a comment saying "This news is ridiculous!", the content of that comment is recorded by the server. For this purpose, the server uses a database management system (such as MySQL or PostgreSQL).

[0245] The server then passes the collected comment data to a natural language processing engine (such as Python's NLTK or spaCy) to calculate an offensiveness score for each comment. For example, the natural language processing engine calculates an offensiveness score of 70 for the comment "This news is ridiculous!". It also uses an emotion engine (such as Google Cloud Natural Language API) to evaluate the poster's emotional state from the comment, and reflects the evaluation result in the offensiveness score. For example, a comment with a strong emotional response of "anger" is assigned a higher offensiveness score.

[0246] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[0247] Next, the user (visitor) can set the filter strength to their preference on their device. Furthermore, the emotion engine can recognize the user's current emotional state and automatically set the recommended filter strength. For example, User C sets the filter strength to "20," or the emotion engine recognizes User C's "irritation" and recommends the filter strength to "50."

[0248] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it references the offensiveness score of the comment and hides comments that exceed the standard. For example, for user C who sets the filter strength to "50", comments with an offensiveness score of 50 or more will be hidden.

[0249] The server then sends the filtered comment list to the user's device, allowing the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User C's device, comments with an offensiveness score of "70" are hidden, and comments with an offensiveness score of "30" are displayed.

[0250] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[0251] Explanation with concrete examples

[0252] Prompt Sentence Examples

[0253] text

[0254] User A posts a comment saying, "This news is ridiculous!" Calculate the aggression score of this comment and evaluate the user's emotional state. For example, if the aggression score is 70 and the user's emotional state is recognized as "anger," determine whether the comment should be hidden as a result of filtering.

[0255] This configuration allows users to freely adjust the filter strength and enjoy a comment viewing experience that suits their preferences. It also improves the overall user experience of news sites by promoting healthy discussion and protecting users from offensive comments. Furthermore, the emotion engine provides appropriate filtering based on the user's emotional state, providing a more personalized viewing environment.

[0256] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0257] Step 1:

[0258] The server collects comment data posted on news sites.

[0259] Input: Comment content, poster's user ID, posting date and time

[0260] Specific operation: When user A posts a comment saying "This news is ridiculous!", the server records the comment content, user A's ID, and the posting date and time, and saves them in a database.

[0261] Output: Collected comment data (e.g., comment content "This news is ridiculous!", user ID "A", posting date and time "2023-10-01 12:34:56")

[0262] Step 2:

[0263] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score.

[0264] Input: Collected comment data

[0265] Specific behavior: The server inputs the comment "This news is stupid!" into a natural language processing engine (e.g., Python's NLTK or spaCy) for text analysis, resulting in an offensiveness score of "70."

[0266] Output: Aggression score (e.g., 70)

[0267] Step 3:

[0268] The server passes the collected comment data to a sentiment analysis engine to evaluate the poster's emotional state.

[0269] Input: Collected comment data

[0270] Specific operation: The server inputs the same comment into a sentiment analysis engine (e.g., Google Cloud Natural Language API) and detects that the sentiment "anger" is high.

[0271] Output: Emotional state rating (e.g., anger 60 / 100)

[0272] Step 4:

[0273] The server associates the calculated aggression score and the evaluation results of the emotional state with each comment data and stores them in a database.

[0274] Input: Aggression score and emotional state assessment results

[0275] Specific Action: Give the comment "This news is stupid!" an aggression score of "70" and an emotion rating of "Anger 60" and save it in the database.

[0276] Output: Comment data stored in the database and its rating (e.g., comment "This news is stupid!", aggression score "70", emotion "anger 60")

[0277] Step 5:

[0278] Users (visitors) can set the filter strength on their terminal to suit their preferences.

[0279] Input: User-input filter strength setting

[0280] Specific actions: User C sets the filter strength to "20" on their device, or the emotion analysis engine recognizes User C's "irritation" and recommends a filter strength of "50."

[0281] Output: User-defined filter strength (e.g., filter strength 20 or 50)

[0282] Step 6:

[0283] The server filters the comments based on the filter strength set by the user.

[0284] Input: User-defined filter strength and comment data

[0285] Specific operation: The server references user C's filter strength of "50" and retrieves the offensiveness score from the database. Comments with a score of "50" or higher are hidden.

[0286] Output: Filtered comment list (e.g. hide comments with a score of 70, show comments with a score of 30)

[0287] Step 7:

[0288] The server sends the filtered comment list to the user's terminal.

[0289] Input: filtered comment list

[0290] Specific operation: The server sends user C, who has a filter strength of "50," a list that hides comments with an offensive score of "70" and displays comments with a score of "30."

[0291] Output: A list of comments displayed on the user's device (e.g., comments with an offensiveness score of 70 are hidden, and comments with a score of 30 are displayed)

[0292] Step 8:

[0293] The server recalculates the offensiveness score when a new comment is posted and updates the filtering results.

[0294] Input: Newly posted comment data

[0295] Specific operation: When a new comment is posted, the server recalculates the score and emotional state using the natural language processing engine and sentiment analysis engine, and performs filtering.

[0296] Output: Updated filtered comment list (e.g., latest list with new comments)

[0297] Step 9:

[0298] If the user changes the filter strength, the server recalculates the displayed comment list and keeps it up to date.

[0299] Input: User changes to filter strength

[0300] What happens: When the user changes the filter strength, the server re-filters using the new criteria.

[0301] Output: Updated comment list (e.g., a list of comments based on the new filter strength)

[0302] (Application example 2)

[0303] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0304] In conventional content distribution services, some of the comments posted by users can be offensive, disrupting other users' comfortable viewing experiences. Furthermore, especially in environments using head-mounted displays, offensive comments can enter the user's field of vision, potentially causing psychological stress. It is necessary to provide an environment where users can enjoy content safely by automatically detecting and appropriately filtering these offensive comments.

[0305] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0306] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine and calculating an aggression score, means for saving the calculated aggression score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for transmitting a filtered comment list to the user's terminal, and means for filtering comments displayed on the head-mounted display. This allows users to filter aggressive comments according to their own emotional state and enjoy content with peace of mind.

[0307] "Posted comment data" refers to text information including ratings and opinions posted by users on the content distribution service.

[0308] A "natural language processing engine" is a software tool that analyzes text data to understand the meaning and sentiment of a sentence.

[0309] The "aggression score" is a numerical indicator of how aggressive the posted comment data is.

[0310] A "database" is a system for efficiently managing and storing collected data.

[0311] The "filter strength" is a reference value of the offensiveness score set by the user to determine whether or not to display.

[0312] "Filtering" is the process of filtering out harmful comments based on set criteria.

[0313] A "head-mounted display" is a display device that is worn on the user's head.

[0314] "Dynamic update" refers to a process of automatically changing the display content in response to the addition of new data or changes in filter strength.

[0315] "Real-time" means immediate, real-time processing or response.

[0316] This invention relates to a comment filtering system for content distribution services, and aims to improve the comment display environment when viewing using a head-mounted display (HMD). This system determines whether posted comments are offensive and filters them based on that, providing an environment in which users can enjoy content with peace of mind.

[0317] First, the server collects comment data posted by users. This data includes the comment text, the poster's identification information, and the posting date and time. Next, the collected comment data is analyzed using a natural language processing engine to calculate an offensiveness score. The natural language processing engine used is something like the VADER Sentiment Intensity Analyzer.

[0318] The calculated offensiveness score is stored in a database, which allows the offensiveness score of each comment to be managed. Users can set the filter strength via their own device or head-mounted display. This filter strength is the standard value that determines which comments are displayed. Based on the filter strength, comments with high offensive scores are filtered out, ensuring that only appropriate comments are displayed to the user.

[0319] The filtered comment list is sent to the user's device or head-mounted display. The system is dynamically updated, so the offensiveness score is recalculated every time a new comment is posted, and the filtering results are updated immediately. Furthermore, the filtering results are reflected in real time on the head-mounted display, allowing users to view the latest comments.

[0320] Specific examples of use

[0321] For example, if a user posts a comment such as "This video is not funny at all!", the server collects this comment and analyzes it using a natural language processing engine. The analysis results in an "offensiveness score of 70" which is then stored in the database. If the user sets the filter strength to "50," this comment will not be displayed, and only comments with an offensiveness score of 50 or less will be displayed.

[0322] Prompt Sentence Examples

[0323] The following prompt sentences are used by the system to collect user comments and calculate an offensiveness score:

[0324] Text format

[0325] comments_payload = {

[0326] "content": "This video is not funny at all!",

[0327] "user_id": "user123",

[0328] "post_date": "2023-10-01"

[0329] }

[0330] response = requests.post("http: / / localhost:5000 / add_comment", json=comments_payload)

[0331] print(response.json())

[0332] The prompts to set the filter strength and get the filtered comments are as follows:

[0333] Text format

[0334] filter_payload = {

[0335] "filter_strength": 50 Filter strength set by the user

[0336] }

[0337] response = requests.post("http: / / localhost:5000 / get_comments", json=filter_payload)

[0338] print(response.json())

[0339] This allows users to effectively filter out offensive comments and enjoy content comfortably, even when using content distribution services that use head-mounted displays.

[0340] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0341] Step 1:

[0342] The user puts on the head-mounted display, starts using the content distribution service, and posts comments.

[0343] Input: Comment text posted by the user, poster's identification information, posting date and time

[0344] Output: Comment data is sent to the server

[0345] Step 2:

[0346] The server receives the comment data sent by the user and temporarily stores it in a database.

[0347] Input: Comment data (text, identification information, posting date and time)

[0348] Output: Comment data is saved to the database.

[0349] Step 3:

[0350] The server passes the comment data collected from the database to a natural language processing engine and calculates the offensiveness score of the comment.

[0351] Input: Comment data stored in the database

[0352] Output: Aggression score is calculated

[0353] Step 4:

[0354] Based on the aggression score calculated by the natural language processing engine, the server links the aggression score to the comment data and stores it in a database.

[0355] Input: Aggression scores and comment data

[0356] Output: The linked data is saved in the database

[0357] Step 5:

[0358] The user sets the filter strength through the head-mounted display.

[0359] Input: Filter strength value set by the user

[0360] Output: Filter strength setting information is sent to the server

[0361] Step 6:

[0362] The server performs filtering based on the filter strength set by the user, and hides comments whose offensiveness score exceeds the filter strength.

[0363] Input: Comment data and user-defined filter strength

[0364] Output: Filtered comment list

[0365] Step 7:

[0366] The filtered comment list is displayed in real time on a head-mounted display.

[0367] Input: filtered comment list

[0368] Output: Comments displayed on the head-mounted display

[0369] Step 8:

[0370] When a new comment is posted, the server recalculates the offensiveness score and immediately updates the filtering results.

[0371] Input: New comment data

[0372] Output: The updated filtering results are reflected on the head-mounted display.

[0373] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0374] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0375] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0376] [Second embodiment]

[0377] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0378] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0379] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0380] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0381] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0382] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0383] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0384] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0385] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0386] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0387] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0388] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0389] The present invention provides a system for filtering offensive comments in the comment sections of news sites, with the goal of creating an environment where users can safely view and participate in comments.

[0390] First, the server collects all posted comment data. This data collection includes metadata such as the comment content, the poster's user ID, and the posting date and time. For example, if user A posts a comment saying, "This news is ridiculous!", the content of that comment will be recorded by the server.

[0391] The server then passes the collected comment data to a natural language processing engine, which calculates an "offense score" for each comment. For example, the natural language processing engine calculates an offensive score of 70 for a comment such as "This news is ridiculous!" On the other hand, a comment such as "That was very interesting" would receive an offensive score of 5.

[0392] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[0393] Next, the user (visitor) sets the filter strength to their preference on the device. The filter strength (threshold of the offensive score) can be input or selected through the user interface. For example, if user B sets the filter strength to "20," this setting is confirmed on the device.

[0394] The server then filters the comments based on the filter strength set by the user. This process hides any comments with an offensiveness score above the set threshold. For example, if user B's filter strength is "20," any comments with an offensiveness score of 20 or higher will be hidden.

[0395] Finally, the server sends the filtered comment list to the user's device. This allows the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User B's device, a comment with an offensiveness score of 25, such as "This is a really awful story," will be hidden, while a comment with an offensiveness score of 10, such as "That was very interesting," will be displayed.

[0396] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[0397] This configuration allows users to freely adjust the filter strength and customize their comment viewing experience, promoting healthy discussion and protecting users from offensive comments, improving the overall user experience on news sites.

[0398] The processing flow will be explained below.

[0399] Step 1:

[0400] The server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. All posted comments are recorded in a database.

[0401] Example: When user A posts a comment saying "This news is ridiculous!", the server records the comment content and poster information in a database.

[0402] Step 2:

[0403] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment.

[0404] Example: For the comment "This news is ridiculous!", the natural language processing engine calculates an offensive score of "70".

[0405] Step 3:

[0406] The server associates the offensiveness score calculated by the natural language processing engine with the comment data and stores it in a database.

[0407] Example: Add an offensive score of 70 to the record of the comment "This news is stupid!" and save it in the database.

[0408] Step 4:

[0409] Users (visitors) can set their own preferred filter strength on their devices by inputting the filter strength (threshold of the offensiveness score) through the user interface.

[0410] Example: When user B sets the filter strength to "20", the setting is confirmed on the device.

[0411] Step 5:

[0412] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it refers to the offensiveness score of the comment and hides comments that exceed the standard.

[0413] Example: For user B who has set the filter strength to "20", comments with an offensive score of 20 or more will be hidden.

[0414] Step 6:

[0415] The server sends the filtered comment list to the user's terminal.

[0416] Example: Send a comment list to user B's device that includes only comments with an offensiveness score of less than 20.

[0417] Step 7:

[0418] The device updates the comment section based on the filtering results received from the server, allowing users to view only the filtered comments.

[0419] Example: On User B's device, comments with an offensiveness score of "25" are hidden, and comments with an offensiveness score of "10" are displayed.

[0420] Step 8:

[0421] When a new comment is posted or a user changes the filter strength, the server recalculates the offensiveness score and dynamically updates the filtering results.

[0422] Example: When a new comment "This is awful!" is posted, its offensiveness score is calculated and the list of comments displayed is instantly updated based on the existing filter strength settings.

[0423] Example 1

[0424] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0425] In the comment sections of conventional news sites, offensive comments are common, making it difficult for users to safely view and participate in the comments. It is also difficult to adapt to the different filter strength preferences of each user, so a system that dynamically updates the comment filtering and display has been required.

[0426] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0427] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine to calculate an offensiveness score, means for saving the calculated offensiveness score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for sending a filtered comment list to the user's terminal, and means for dynamically reflecting the filter strength setting on the server. This allows users to filter comments according to their preferences in real time, and provides a comment section that promotes healthy discussion.

[0428] "Comment data" refers to the text that users post to news sites and the metadata associated with those posts.

[0429] A "natural language processing engine" refers to software or algorithms that analyze human language to understand its meaning and sentiment and then perform specific tasks.

[0430] "Aggression score" refers to a numerical indicator that indicates whether a posted comment is offensive or not.

[0431] "Users" refer to people who visit news sites, post comments, and set filter strengths.

[0432] "Filter strength" refers to the threshold for the offensiveness score of comments set by the user, which determines the range of comments that will be displayed.

[0433] "Database" refers to an electronic storage system for storing and managing collected comment data and offensiveness scores.

[0434] "Device" refers to the electronic device (e.g., computer, smartphone, tablet, etc.) that a user uses to access a news site.

[0435] "Dynamic reflection" means that user settings and new comment posts are applied to the system in real time.

[0436] MODE FOR CARRYING OUT THE INVENTION

[0437] The present invention provides a system for filtering offensive comments in the comment section of a news site. The system aims to create an environment where users can safely view and participate in comments. A specific embodiment of the system is described below.

[0438] First, this system consists of multiple hardware and software components: a server, a natural language processing engine, a database, a user interface (UI), and a terminal.

[0439] 1. Collection of comment data

[0440] The server collects all comment data posted on the news site. This collected data includes metadata such as the comment content, the poster's user ID, and the posting date and time. For example, if user A posts a comment saying "This news is ridiculous!", the content and metadata are recorded by the server.

[0441] 2. Calculating offensive scores using a natural language processing engine

[0442] The server passes the collected comment data to a natural language processing engine, which calculates an "offense score" for each comment. This natural language processing engine uses, for example, Google NLP API or OpenAI's GPT-3. For example, the natural language processing engine calculates an offensiveness score of 70 for a comment such as "This news is ridiculous!". On the other hand, a comment such as "That was very interesting" would be assigned an offensiveness score of 5.

[0443] 3. Save the calculated aggression score to the database

[0444] The server associates the calculated offensiveness score with each comment data and stores it in a database. This database can be an SQL database or a NoSQL database. For example, a comment such as "This news is ridiculous!" is associated with an offensiveness score of "70" and stored in the database.

[0445] 4. User-defined filter strength

[0446] Users can set their preferred filter strength through the user interface on their device. The UI is provided via a web browser or mobile app. Users can input or select the filter strength (threshold of the aggressiveness score). For example, if user B sets the filter strength to "20," this setting is sent to the server in real time.

[0447] 5. Filtering process and sending results

[0448] The server filters comments based on the filter strength set by the user. This process hides comments with an offensiveness score above the set threshold. For example, if user B's filter strength is "20," comments with an offensiveness score of 20 or more will be hidden. The filtered comment list is sent from the server to the user's device, where it can be viewed by the user.

[0449] Examples of concrete examples and prompts

[0450] Examples:

[0451] 1. User A posts a comment saying, "This news is great!"

[0452] 2. The server collects these comments and passes them to a natural language processing engine.

[0453] 3. The natural language processing engine calculates an offensiveness score of "1".

[0454] 4. The server associates this score with the comment data and stores it in a database.

[0455] 5. User B sets the filter strength to "10".

[0456] 6. Based on the filter strength of "10", the server sends only comments with an offensiveness score of "less than 10" to the user's device.

[0457] Example prompt:

[0458] We want to make comments on this news site safer and friendlier. If you see any of the following comments posted, please rate their offensiveness:

[0459] Comment: "This news is ridiculous!"

[0460] Comment: "It was very interesting."

[0461] You can also filter comments based on the following settings to view the comments that apply to each setting:

[0462] Filter Strength: 20

[0463] With this setting, which comments will be visible and which will be hidden?

[0464] This allows users to freely adjust the filter strength and customize their comment viewing experience to suit their preferences. It also promotes healthy discussion and protects users from offensive comments, improving the overall user experience on news sites.

[0465] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0466] Program processing flow

[0467] Step 1:

[0468] The server collects all comment data posted on the news site. This includes metadata such as the comment content, the poster's user ID, and the posting date and time. The input is a newly posted comment on the news site, and the output is the comment data stored in the server. For example, if user A posts "This news is stupid!", the content and metadata are recorded by the server.

[0469] Step 2:

[0470] The server passes the collected comment data to a natural language processing engine. The input is the comment data stored in the server, and the output is an offensiveness score for each comment. Specifically, the comment "This news is stupid!" is sent to the natural language processing engine, and the engine returns an offensiveness score of "70" for this comment.

[0471] Step 3:

[0472] The server associates the calculated offensiveness score with each comment data and stores it in the database. The input is the comment data with the calculated offensiveness score, and the output is the comment data stored in the database and its offensiveness score. For example, the comment "This news is ridiculous!" is associated with an offensiveness score of "70" and stored in the database.

[0473] Step 4:

[0474] The user sets the filter strength to their preference on the device. The input is the user's filter strength setting value, and the output is the setting data sent to the server. Specifically, when user B sets the filter strength to "20," this setting is confirmed on the device and sent to the server.

[0475] Step 5:

[0476] The server filters comments based on the filter strength set by the user. The input is the user's filter strength setting and the comment data in the database, and the output is a filtered comment list. Specifically, based on a filter strength of "20," comments with an offensiveness score of 20 or more are excluded.

[0477] Step 6:

[0478] The server sends the filtered comment list to the user's device. The input is the filtered comment list, and the output is the comments displayed on the user's device. For example, on User B's device, a comment with an offensiveness score of "10" such as "That was very interesting," is displayed, while a comment with an offensiveness score of "25" such as "That's a really awful story," is hidden.

[0479] (Application example 1)

[0480] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0481] Comment sections on news sites and social media often contain offensive comments that cause discomfort and psychological stress to users. It is important to effectively filter these comments and provide an environment where users can safely view and participate in the comments. However, existing systems face challenges in filtering offensive comments in real time and dynamically updating the user display. Furthermore, they lack the ability for users to adjust the filter strength according to their preferences, making it difficult to display comments that meet the needs of individual users.

[0482] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0483] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score, means for saving the calculated offensiveness score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for transmitting a filtered comment list to the user's display device, means for adjusting the filter strength on the user's display device, and means for maintaining communication with the display device to dynamically perform the filtering process. This allows users to filter offensive comments in real time and comfortably view comments with a filter strength that suits their preferences.

[0484] "Means for collecting posted comment data" refers to a system for collecting comments posted on the Internet by users.

[0485] A "natural language processing engine" is software or algorithm that analyzes collected comment data and understands and evaluates its content.

[0486] The "offensiveness score" is a numerical indication of the degree of offensiveness of a comment, and is an index calculated by a natural language processing engine.

[0487] "Means for storing in a database" refers to a system for centrally managing and storing calculated offensive scores and comment data.

[0488] The "means for setting filter strength" is an interface that allows the user to set the threshold of the aggressiveness score according to their own preferences.

[0489] The "means for filtering comments" is a mechanism for selecting comments to be displayed based on the set filter strength.

[0490] "Means for transmitting to a user's display device" is a mechanism for transmitting filtered comments to a device used by the user (e.g., smart glasses).

[0491] The "means for adjusting the filter strength" is an adjustment function that allows the filter strength changed by the user in real time to be immediately reflected.

[0492] The "means for maintaining communication with the display device" is a system for constantly maintaining a connection with the user's display device in order to dynamically perform comment filtering processing.

[0493] "Dynamic update means" refers to a function that changes the comment section displayed in real time depending on the set filter strength and newly posted comments.

[0494] This invention is a system for filtering offensive comments in the comment sections of news sites and social media. To implement this system, the following steps and various hardware and software are required.

[0495] Hardware and software used

[0496] Server: Used to collect, analyze, score, and store comment data in a database.

[0497] Natural language processing engine: Used to analyze the offensiveness of comments and calculate an offensiveness score.

[0498] Database: Used to store calculated aggression scores and comment data.

[0499] User display device (e.g., smart glasses): Used to display filtered comments.

[0500] HTTP request library (e.g. requests for Python): Used to communicate between the server and various APIs.

[0501] Overview of the overall system processing flow

[0502] 1. Collection of comment data

[0503] The server collects comment data posted from news sites and social media, including the comment text, the poster's user ID, and the posting date and time.

[0504] 2. Calculating the Aggression Score

[0505] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment.

[0506] 3. Saving to the database

[0507] The calculated offensiveness score is stored in a database together with the corresponding comment data.

[0508] 4. Setting the filter strength

[0509] Users set the filter strength (threshold of offensiveness score) for comments via a display device such as smart glasses, which is done through a user interface.

[0510] 5. Comment Filtering

[0511] The server filters the comments based on the filter strength set by the user and transmits only the selected comments to the display device.

[0512] 6. Dynamic Updates

[0513] The server performs filtering processing in real time, even when a new comment is posted or when a user changes the filter strength, and dynamically updates the filtering results while maintaining communication with the display device.

[0514] Specific examples

[0515] For example, suppose user A posts "This news is stupid!" in the comments section of a news article. This comment is collected by the server and passed to a natural language processing engine. The server then calculates an offensiveness score of 70 for this comment and stores it in the database.

[0516] Later, when User B uses the smart glasses to view a news article, he sets the filter strength to "20." Based on this setting, the server hides comments with an offensiveness score of 20 or more. Therefore, a comment such as "This news is ridiculous!" will not be displayed. On the other hand, a comment with an offensiveness score of "5," such as "This was very interesting," will be displayed.

[0517] Prompt Sentence Examples

[0518] Example configuration:

[0519] article_id = 67890 Example of article ID setting

[0520] filter_threshold = 50 Example of filter strength setting

[0521] Sample prompt sentence:

[0522] Input prompt: "Please filter and display comments for article ID "67890" using filter strength "50"."

[0523] Example output:

[0524] Displayed comment: "This is very useful information." (Aggression score: 10)

[0525] Comment to be hidden: "This article is rubbish!" (Aggression score: 60)

[0526] This system allows users to filter out offensive comments in real time, allowing them to comfortably view comments with a filter strength that suits their preferences.

[0527] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0528] Step 1:

[0529] The server collects comment data posted from news sites and social media. This collected data includes the comment text, the poster's user ID, the posting date and time, etc. The input requires the URL of the news article or social media post to be retrieved, and the collected comment data is obtained as output.

[0530] Step 2:

[0531] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment. This process requires the collected comment data as input and obtains an offensiveness score corresponding to each comment as output. Specific operations include making an HTTP request to the natural language processing engine API.

[0532] Step 3:

[0533] The server stores the calculated offensiveness score in a database. This process requires the comment data and the corresponding offensiveness score as input, and obtains the database state as output. Specific operations include creating and updating records in the database.

[0534] Step 4:

[0535] The user sets the filter strength (threshold of the aggression score) through the interface of the smart glasses. This process requires the user to input the filter strength and saves the setting as output. Specific actions include entering a number or operating a slider on the user interface.

[0536] Step 5:

[0537] The server filters comments based on the filter strength set by the user. This process requires the comment data and offensiveness scores stored in the database and the filter strength set by the user as input, and obtains a filtered comment list as output. Specific operations include a process of comparing the offensiveness score and the filter strength.

[0538] Step 6:

[0539] The server sends the filtered comment list to the user's display device. The filtered comment list is required as input, and the comments displayed on the user's display device are obtained as output. Specific operations include a process of sending data to the display device.

[0540] Step 7:

[0541] When a user sets a new filter strength, the server performs the filtering process again and sends the result to the display device. The new filter strength, existing comment data, and offensiveness scores are required as input, and the latest filtering result is obtained as output. Specific operations include the re-filtering and sending process.

[0542] Step 8:

[0543] When a new comment is posted, the server recalculates the offensiveness score of the comment, updates the database, and regenerates the filtering results and sends them to the display device. New comment data is required as input, and updated filtering results are obtained as output. Specific operations include automatic collection, score calculation, database update, and re-filtering.

[0544] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0545] The present invention provides a system that filters offensive comments in the comment section of a news site and provides an appropriate comment display environment that takes into account the user's feelings. The system aims to create an environment where users can view and participate in comments with peace of mind.

[0546] First, the server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. For example, if user A posts a comment saying, "This news is ridiculous!", the content of that comment is recorded by the server.

[0547] The server then passes the collected comment data to a natural language processing engine, which calculates an "aggression score" for each comment. For example, the natural language processing engine calculates an aggression score of 70 for a comment such as "This news is ridiculous!". The emotion engine also evaluates the poster's emotional state from the comment, and reflects the evaluation result in the aggression score. For example, a comment with a strong emotion of "anger" is assigned a higher aggression score.

[0548] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[0549] Next, the user (visitor) can set the filter strength to their preference on the device, or the emotion engine can recognize the user's current emotional state and automatically set the recommended filter strength. For example, User C can set the filter strength to "20," or the emotion engine can recognize User C's "irritation" and recommend the filter strength to "50."

[0550] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it references the offensiveness score of the comment and hides comments that exceed the standard. For example, for user C who sets the filter strength to "50", comments with an offensiveness score of 50 or more will be hidden.

[0551] The server then sends the filtered comment list to the user's device, allowing the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User C's device, comments with an offensiveness score of "70" are hidden, and comments with an offensiveness score of "30" are displayed.

[0552] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[0553] This configuration allows users to freely adjust the filter strength and enjoy a comment viewing experience that suits their preferences. It also improves the overall user experience of news sites by promoting healthy discussion and protecting users from offensive comments. Furthermore, the emotion engine provides appropriate filtering based on the user's emotional state, providing a more personalized viewing experience.

[0554] The processing flow will be explained below.

[0555] Step 1:

[0556] The server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. All posted comments are recorded in a database.

[0557] Example: When user A posts a comment saying "This news is ridiculous!", the server records the comment content and poster information in a database.

[0558] Step 2:

[0559] The server passes the collected comment data to a natural language processing engine, calculates an offensiveness score for each comment, and uses an emotion engine to evaluate the poster's emotional state based on the comment.

[0560] Example: In response to the comment "This news is ridiculous!", the natural language processing engine calculates an aggression score of "70" and the emotion engine evaluates it as "anger."

[0561] Step 3:

[0562] The server associates the aggression score calculated by the natural language processing engine and the emotional state evaluated by the emotion engine with the comment data and stores them in a database.

[0563] Example: Add an aggression score of 70 and emotion of anger to the record of the comment "This news is stupid!" and save it in the database.

[0564] Step 4:

[0565] Users (visitors) can set the filter strength to suit their preferences on their devices, or the emotion engine will recognize the user's current emotional state and automatically set the recommended filter strength.

[0566] Example: User C sets the filter strength to "20." Or, the emotion engine recognizes User C's "irritation" and recommends setting the filter strength to "50."

[0567] Step 5:

[0568] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it refers to the offensiveness score of the comment and hides comments that exceed the standard.

[0569] Example: For user C who has set the filter strength to "50", comments with an offensive score of 50 or more will be hidden.

[0570] Step 6:

[0571] The server sends the filtered comment list to the user's terminal.

[0572] Example: A comment list containing only comments with an offensiveness score of less than 50 is sent to user C's device.

[0573] Step 7:

[0574] The device updates the comment section based on the filtering results received from the server, allowing users to view only the filtered comments.

[0575] Example: On User C's device, comments with an aggression score of "70" are hidden, and comments with an aggression score of "30" are displayed.

[0576] Step 8:

[0577] When a new comment is posted or a user changes the filter strength, the server recalculates the offensiveness score and dynamically updates the filtering results.

[0578] Example: When a new comment "This is awful!" is posted, its offensiveness score is calculated and the list of comments displayed is instantly updated based on the existing filter strength settings.

[0579] Example 2

[0580] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0581] On traditional news sites, users are often exposed to offensive comments, making it difficult to ensure a comfortable viewing environment. In particular, the proliferation of offensive comments can hinder healthy discussion and increase the mental burden on users. Furthermore, traditional filtering systems are static and filter content based on uniform criteria without considering the user's emotional state, making it difficult to provide a personalized experience.

[0582] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0583] In this invention, the server includes a means for collecting posted comment data, a means for analyzing the collected comment data using a natural language processing engine to calculate an offensiveness score, and a means for analyzing the collected comment data using a sentiment analysis engine to evaluate the poster's emotional state. This enables filtering according to individual emotional states while preventing users from being exposed to offensive comments. The server also includes a means for storing the comment's offensiveness score and the emotional state evaluation results in a database, a means for the user to set the filter strength, and a means for recognizing the user's emotional state using the sentiment analysis engine and setting a recommended filter strength, thereby achieving flexible filtering according to individual user needs. The server also includes a means for transmitting a filtered comment list to the user's device and a means for recalculating the offensiveness score and updating the filtering results when a new comment is posted, providing a constantly up-to-date comment viewing environment. This improves the user experience across the news site and promotes healthy discussion.

[0584] "Posted comment data" refers to data that includes information such as the content of comments submitted by users of a news site, the poster's user ID, and the date and time of posting.

[0585] A "natural language processing engine" refers to a program or algorithm that analyzes text data and understands its content.

[0586] "Aggression score" refers to a numerical representation of the degree of aggression based on the content of a comment.

[0587] "Sentiment analysis engine" refers to a program or algorithm that identifies emotions contained in text data and evaluates that emotional state.

[0588] "User emotional state" refers to the current emotional state inferred from the user's input and actions.

[0589] "Filter strength" refers to the threshold of the aggressiveness score that serves as the basis for filtering.

[0590] A "database" refers to a system for systematically managing, storing, and searching information.

[0591] "Filtering" refers to the operation of selecting and displaying specific data based on pre-set criteria.

[0592] "User device" refers to a terminal, such as a PC, smartphone, or tablet, that receives and displays data sent from the server.

[0593] "Recalculation" refers to performing a new calculation process, and in this context specifically refers to recalculating the offensiveness score for newly posted comments.

[0594] "Dynamic update" refers to an operation in which information or data is automatically updated in real time or near real time.

[0595] "News site" refers to a website that provides news information on the Internet and allows users to post comments.

[0596] A "proper noun" is a name used to refer to a specific person, place, organization, etc., but in this context is written as a general name rather than a specific name.

[0597] The present invention is a system that filters offensive comments in the comment section of a news site and provides an appropriate comment display environment that takes into account the user's feelings. The purpose of this system is to create an environment where users can safely view and participate in comments.

[0598] First, the server collects all comment data posted on the news site. This collection includes information such as the comment content, the poster's user ID, and the posting date and time. For example, if User A posts a comment saying "This news is ridiculous!", the content of that comment is recorded by the server. For this purpose, the server uses a database management system (such as MySQL or PostgreSQL).

[0599] The server then passes the collected comment data to a natural language processing engine (such as Python's NLTK or spaCy) to calculate an offensiveness score for each comment. For example, the natural language processing engine calculates an offensiveness score of 70 for the comment "This news is ridiculous!". It also uses an emotion engine (such as Google Cloud Natural Language API) to evaluate the poster's emotional state from the comment, and reflects the evaluation result in the offensiveness score. For example, a comment with a strong emotional response of "anger" is assigned a higher offensiveness score.

[0600] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[0601] Next, the user (visitor) can set the filter strength to their preference on their device. Furthermore, the emotion engine can recognize the user's current emotional state and automatically set the recommended filter strength. For example, User C sets the filter strength to "20," or the emotion engine recognizes User C's "irritation" and recommends the filter strength to "50."

[0602] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it references the offensiveness score of the comment and hides comments that exceed the standard. For example, for user C who sets the filter strength to "50", comments with an offensiveness score of 50 or more will be hidden.

[0603] The server then sends the filtered comment list to the user's device, allowing the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User C's device, comments with an offensiveness score of "70" are hidden, and comments with an offensiveness score of "30" are displayed.

[0604] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[0605] Explanation with concrete examples

[0606] Prompt Sentence Examples

[0607] text

[0608] User A posts a comment saying, "This news is ridiculous!" Calculate the aggression score of this comment and evaluate the user's emotional state. For example, if the aggression score is 70 and the user's emotional state is recognized as "anger," determine whether the comment should be hidden as a result of filtering.

[0609] This configuration allows users to freely adjust the filter strength and enjoy a comment viewing experience that suits their preferences. It also improves the overall user experience of news sites by promoting healthy discussion and protecting users from offensive comments. Furthermore, the emotion engine provides appropriate filtering based on the user's emotional state, providing a more personalized viewing environment.

[0610] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0611] Step 1:

[0612] The server collects comment data posted on news sites.

[0613] Input: Comment content, poster's user ID, posting date and time

[0614] Specific operation: When user A posts a comment saying "This news is ridiculous!", the server records the comment content, user A's ID, and the posting date and time, and saves them in a database.

[0615] Output: Collected comment data (e.g., comment content "This news is ridiculous!", user ID "A", posting date and time "2023-10-01 12:34:56")

[0616] Step 2:

[0617] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score.

[0618] Input: Collected comment data

[0619] Specific behavior: The server inputs the comment "This news is stupid!" into a natural language processing engine (e.g., Python's NLTK or spaCy) for text analysis, resulting in an offensiveness score of "70."

[0620] Output: Aggression score (e.g., 70)

[0621] Step 3:

[0622] The server passes the collected comment data to a sentiment analysis engine to evaluate the poster's emotional state.

[0623] Input: Collected comment data

[0624] Specific operation: The server inputs the same comment into a sentiment analysis engine (e.g., Google Cloud Natural Language API) and detects that the sentiment "anger" is high.

[0625] Output: Emotional state rating (e.g., anger 60 / 100)

[0626] Step 4:

[0627] The server associates the calculated aggression score and the evaluation results of the emotional state with each comment data and stores them in a database.

[0628] Input: Aggression score and emotional state assessment results

[0629] Specific Action: Give the comment "This news is stupid!" an aggression score of "70" and an emotion rating of "Anger 60" and save it in the database.

[0630] Output: Comment data stored in the database and its rating (e.g., comment "This news is stupid!", aggression score "70", emotion "anger 60")

[0631] Step 5:

[0632] Users (visitors) can set the filter strength on their terminal to suit their preferences.

[0633] Input: User-input filter strength setting

[0634] Specific actions: User C sets the filter strength to "20" on their device, or the emotion analysis engine recognizes User C's "irritation" and recommends a filter strength of "50."

[0635] Output: User-defined filter strength (e.g., filter strength 20 or 50)

[0636] Step 6:

[0637] The server filters the comments based on the filter strength set by the user.

[0638] Input: User-defined filter strength and comment data

[0639] Specific operation: The server references user C's filter strength of "50" and retrieves the offensiveness score from the database. Comments with a score of "50" or higher are hidden.

[0640] Output: Filtered comment list (e.g. hide comments with a score of 70, show comments with a score of 30)

[0641] Step 7:

[0642] The server sends the filtered comment list to the user's terminal.

[0643] Input: filtered comment list

[0644] Specific operation: The server sends user C, who has a filter strength of "50," a list that hides comments with an offensive score of "70" and displays comments with a score of "30."

[0645] Output: A list of comments displayed on the user's device (e.g., comments with an offensiveness score of 70 are hidden, and comments with a score of 30 are displayed)

[0646] Step 8:

[0647] The server recalculates the offensiveness score when a new comment is posted and updates the filtering results.

[0648] Input: Newly posted comment data

[0649] Specific operation: When a new comment is posted, the server recalculates the score and emotional state using the natural language processing engine and sentiment analysis engine, and performs filtering.

[0650] Output: Updated filtered comment list (e.g., latest list with new comments)

[0651] Step 9:

[0652] If the user changes the filter strength, the server recalculates the displayed comment list and keeps it up to date.

[0653] Input: User changes to filter strength

[0654] What happens: When the user changes the filter strength, the server re-filters using the new criteria.

[0655] Output: Updated comment list (e.g., a list of comments based on the new filter strength)

[0656] (Application example 2)

[0657] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0658] In conventional content distribution services, some of the comments posted by users can be offensive, disrupting other users' comfortable viewing experiences. Furthermore, especially in environments using head-mounted displays, offensive comments can enter the user's field of vision, potentially causing psychological stress. It is necessary to provide an environment where users can enjoy content safely by automatically detecting and appropriately filtering these offensive comments.

[0659] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0660] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine and calculating an aggression score, means for saving the calculated aggression score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for transmitting a filtered comment list to the user's terminal, and means for filtering comments displayed on the head-mounted display. This allows users to filter aggressive comments according to their own emotional state and enjoy content with peace of mind.

[0661] "Posted comment data" refers to text information including ratings and opinions posted by users on the content distribution service.

[0662] A "natural language processing engine" is a software tool that analyzes text data to understand the meaning and sentiment of a sentence.

[0663] The "aggression score" is a numerical indicator of how aggressive the posted comment data is.

[0664] A "database" is a system for efficiently managing and storing collected data.

[0665] The "filter strength" is a reference value of the offensiveness score set by the user to determine whether or not to display.

[0666] "Filtering" is the process of filtering out harmful comments based on set criteria.

[0667] A "head-mounted display" is a display device that is worn on the user's head.

[0668] "Dynamic update" refers to a process of automatically changing the display content in response to the addition of new data or changes in filter strength.

[0669] "Real-time" means immediate, real-time processing or response.

[0670] This invention relates to a comment filtering system for content distribution services, and aims to improve the comment display environment when viewing using a head-mounted display (HMD). This system determines whether posted comments are offensive and filters them based on that, providing an environment in which users can enjoy content with peace of mind.

[0671] First, the server collects comment data posted by users. This data includes the comment text, the poster's identification information, and the posting date and time. Next, the collected comment data is analyzed using a natural language processing engine to calculate an offensiveness score. The natural language processing engine used is something like the VADER Sentiment Intensity Analyzer.

[0672] The calculated offensiveness score is stored in a database, which allows the offensiveness score of each comment to be managed. Users can set the filter strength via their own device or head-mounted display. This filter strength is the standard value that determines which comments are displayed. Based on the filter strength, comments with high offensive scores are filtered out, ensuring that only appropriate comments are displayed to the user.

[0673] The filtered comment list is sent to the user's device or head-mounted display. The system is dynamically updated, so the offensiveness score is recalculated every time a new comment is posted, and the filtering results are updated immediately. Furthermore, the filtering results are reflected in real time on the head-mounted display, allowing users to view the latest comments.

[0674] Specific examples of use

[0675] For example, if a user posts a comment such as "This video is not funny at all!", the server collects this comment and analyzes it using a natural language processing engine. The analysis results in an "offensiveness score of 70" which is then stored in the database. If the user sets the filter strength to "50," this comment will not be displayed, and only comments with an offensiveness score of 50 or less will be displayed.

[0676] Prompt Sentence Examples

[0677] The following prompt sentences are used by the system to collect user comments and calculate an offensiveness score:

[0678] Text format

[0679] comments_payload = {

[0680] "content": "This video is not funny at all!",

[0681] "user_id": "user123",

[0682] "post_date": "2023-10-01"

[0683] }

[0684] response = requests.post("http: / / localhost:5000 / add_comment", json=comments_payload)

[0685] print(response.json())

[0686] The prompts to set the filter strength and get the filtered comments are as follows:

[0687] Text format

[0688] filter_payload = {

[0689] "filter_strength": 50 Filter strength set by the user

[0690] }

[0691] response = requests.post("http: / / localhost:5000 / get_comments", json=filter_payload)

[0692] print(response.json())

[0693] This allows users to effectively filter out offensive comments and enjoy content comfortably, even when using content distribution services that use head-mounted displays.

[0694] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0695] Step 1:

[0696] The user puts on the head-mounted display, starts using the content distribution service, and posts comments.

[0697] Input: Comment text posted by the user, poster's identification information, posting date and time

[0698] Output: Comment data is sent to the server

[0699] Step 2:

[0700] The server receives the comment data sent by the user and temporarily stores it in a database.

[0701] Input: Comment data (text, identification information, posting date and time)

[0702] Output: Comment data is saved to the database.

[0703] Step 3:

[0704] The server passes the comment data collected from the database to a natural language processing engine and calculates the offensiveness score of the comment.

[0705] Input: Comment data stored in the database

[0706] Output: Aggression score is calculated

[0707] Step 4:

[0708] Based on the aggression score calculated by the natural language processing engine, the server links the aggression score to the comment data and stores it in a database.

[0709] Input: Aggression scores and comment data

[0710] Output: The linked data is saved in the database

[0711] Step 5:

[0712] The user sets the filter strength through the head-mounted display.

[0713] Input: Filter strength value set by the user

[0714] Output: Filter strength setting information is sent to the server

[0715] Step 6:

[0716] The server performs filtering based on the filter strength set by the user, and hides comments whose offensiveness score exceeds the filter strength.

[0717] Input: Comment data and user-defined filter strength

[0718] Output: Filtered comment list

[0719] Step 7:

[0720] The filtered comment list is displayed in real time on a head-mounted display.

[0721] Input: filtered comment list

[0722] Output: Comments displayed on the head-mounted display

[0723] Step 8:

[0724] When a new comment is posted, the server recalculates the offensiveness score and immediately updates the filtering results.

[0725] Input: New comment data

[0726] Output: The updated filtering results are reflected on the head-mounted display.

[0727] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0728] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0729] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0730] [Third embodiment]

[0731] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0732] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0733] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0734] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0735] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0736] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0737] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0738] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0739] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0740] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0741] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0742] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0743] The present invention provides a system for filtering offensive comments in the comment sections of news sites, with the goal of creating an environment where users can safely view and participate in comments.

[0744] First, the server collects all posted comment data. This data collection includes metadata such as the comment content, the poster's user ID, and the posting date and time. For example, if user A posts a comment saying, "This news is ridiculous!", the content of that comment will be recorded by the server.

[0745] The server then passes the collected comment data to a natural language processing engine, which calculates an "offense score" for each comment. For example, the natural language processing engine calculates an offensive score of 70 for a comment such as "This news is ridiculous!" On the other hand, a comment such as "That was very interesting" would receive an offensive score of 5.

[0746] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[0747] Next, the user (visitor) sets the filter strength to their preference on the device. The filter strength (threshold of the offensive score) can be input or selected through the user interface. For example, if user B sets the filter strength to "20," this setting is confirmed on the device.

[0748] The server then filters the comments based on the filter strength set by the user. This process hides any comments with an offensiveness score above the set threshold. For example, if user B's filter strength is "20," any comments with an offensiveness score of 20 or higher will be hidden.

[0749] Finally, the server sends the filtered comment list to the user's device. This allows the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User B's device, a comment with an offensiveness score of 25, such as "This is a really awful story," will be hidden, while a comment with an offensiveness score of 10, such as "That was very interesting," will be displayed.

[0750] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[0751] This configuration allows users to freely adjust the filter strength and customize their comment viewing experience, promoting healthy discussion and protecting users from offensive comments, improving the overall user experience on news sites.

[0752] The processing flow will be explained below.

[0753] Step 1:

[0754] The server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. All posted comments are recorded in a database.

[0755] Example: When user A posts a comment saying "This news is ridiculous!", the server records the comment content and poster information in a database.

[0756] Step 2:

[0757] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment.

[0758] Example: For the comment "This news is ridiculous!", the natural language processing engine calculates an offensive score of "70".

[0759] Step 3:

[0760] The server associates the offensiveness score calculated by the natural language processing engine with the comment data and stores it in a database.

[0761] Example: Add an offensive score of 70 to the record of the comment "This news is stupid!" and save it in the database.

[0762] Step 4:

[0763] Users (visitors) can set their own preferred filter strength on their devices by inputting the filter strength (threshold of the offensiveness score) through the user interface.

[0764] Example: When user B sets the filter strength to "20", the setting is confirmed on the device.

[0765] Step 5:

[0766] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it refers to the offensiveness score of the comment and hides comments that exceed the standard.

[0767] Example: For user B who has set the filter strength to "20", comments with an offensive score of 20 or more will be hidden.

[0768] Step 6:

[0769] The server sends the filtered comment list to the user's terminal.

[0770] Example: Send a comment list to user B's device that includes only comments with an offensiveness score of less than 20.

[0771] Step 7:

[0772] The device updates the comment section based on the filtering results received from the server, allowing users to view only the filtered comments.

[0773] Example: On User B's device, comments with an offensiveness score of "25" are hidden, and comments with an offensiveness score of "10" are displayed.

[0774] Step 8:

[0775] When a new comment is posted or a user changes the filter strength, the server recalculates the offensiveness score and dynamically updates the filtering results.

[0776] Example: When a new comment "This is awful!" is posted, its offensiveness score is calculated and the list of comments displayed is instantly updated based on the existing filter strength settings.

[0777] Example 1

[0778] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0779] In the comment sections of conventional news sites, offensive comments are common, making it difficult for users to safely view and participate in the comments. It is also difficult to adapt to the different filter strength preferences of each user, so a system that dynamically updates the comment filtering and display has been required.

[0780] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0781] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine to calculate an offensiveness score, means for saving the calculated offensiveness score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for sending a filtered comment list to the user's terminal, and means for dynamically reflecting the filter strength setting on the server. This allows users to filter comments according to their preferences in real time, and provides a comment section that promotes healthy discussion.

[0782] "Comment data" refers to the text that users post to news sites and the metadata associated with those posts.

[0783] A "natural language processing engine" refers to software or algorithms that analyze human language to understand its meaning and sentiment and then perform specific tasks.

[0784] "Aggression score" refers to a numerical indicator that indicates whether a posted comment is offensive or not.

[0785] "Users" refer to people who visit news sites, post comments, and set filter strengths.

[0786] "Filter strength" refers to the threshold for the offensiveness score of comments set by the user, which determines the range of comments that will be displayed.

[0787] "Database" refers to an electronic storage system for storing and managing collected comment data and offensiveness scores.

[0788] "Device" refers to the electronic device (e.g., computer, smartphone, tablet, etc.) that a user uses to access a news site.

[0789] "Dynamic reflection" means that user settings and new comment posts are applied to the system in real time.

[0790] MODE FOR CARRYING OUT THE INVENTION

[0791] The present invention provides a system for filtering offensive comments in the comment section of a news site. The system aims to create an environment where users can safely view and participate in comments. A specific embodiment of the system is described below.

[0792] First, this system consists of multiple hardware and software components: a server, a natural language processing engine, a database, a user interface (UI), and a terminal.

[0793] 1. Collection of comment data

[0794] The server collects all comment data posted on the news site. This collected data includes metadata such as the comment content, the poster's user ID, and the posting date and time. For example, if user A posts a comment saying "This news is ridiculous!", the content and metadata are recorded by the server.

[0795] 2. Calculating offensive scores using a natural language processing engine

[0796] The server passes the collected comment data to a natural language processing engine, which calculates an "offense score" for each comment. This natural language processing engine uses, for example, Google NLP API or OpenAI's GPT-3. For example, the natural language processing engine calculates an offensiveness score of 70 for a comment such as "This news is ridiculous!". On the other hand, a comment such as "That was very interesting" would be assigned an offensiveness score of 5.

[0797] 3. Save the calculated aggression score to the database

[0798] The server associates the calculated offensiveness score with each comment data and stores it in a database. This database can be an SQL database or a NoSQL database. For example, a comment such as "This news is ridiculous!" is associated with an offensiveness score of "70" and stored in the database.

[0799] 4. User-defined filter strength

[0800] Users can set their preferred filter strength through the user interface on their device. The UI is provided via a web browser or mobile app. Users can input or select the filter strength (threshold of the aggressiveness score). For example, if user B sets the filter strength to "20," this setting is sent to the server in real time.

[0801] 5. Filtering process and sending results

[0802] The server filters comments based on the filter strength set by the user. This process hides comments with an offensiveness score above the set threshold. For example, if user B's filter strength is "20," comments with an offensiveness score of 20 or more will be hidden. The filtered comment list is sent from the server to the user's device, where it can be viewed by the user.

[0803] Examples of concrete examples and prompts

[0804] Examples:

[0805] 1. User A posts a comment saying, "This news is great!"

[0806] 2. The server collects these comments and passes them to a natural language processing engine.

[0807] 3. The natural language processing engine calculates an offensiveness score of "1".

[0808] 4. The server associates this score with the comment data and stores it in a database.

[0809] 5. User B sets the filter strength to "10".

[0810] 6. Based on the filter strength of "10", the server sends only comments with an offensiveness score of "less than 10" to the user's device.

[0811] Example prompt:

[0812] We want to make comments on this news site safer and friendlier. If you see any of the following comments posted, please rate their offensiveness:

[0813] Comment: "This news is ridiculous!"

[0814] Comment: "It was very interesting."

[0815] You can also filter comments based on the following settings to view the comments that apply to each setting:

[0816] Filter Strength: 20

[0817] With this setting, which comments will be visible and which will be hidden?

[0818] This allows users to freely adjust the filter strength and customize their comment viewing experience to suit their preferences. It also promotes healthy discussion and protects users from offensive comments, improving the overall user experience on news sites.

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

[0820] Program processing flow

[0821] Step 1:

[0822] The server collects all comment data posted on the news site. This includes metadata such as the comment content, the poster's user ID, and the posting date and time. The input is a newly posted comment on the news site, and the output is the comment data stored in the server. For example, if user A posts "This news is stupid!", the content and metadata are recorded by the server.

[0823] Step 2:

[0824] The server passes the collected comment data to a natural language processing engine. The input is the comment data stored in the server, and the output is an offensiveness score for each comment. Specifically, the comment "This news is stupid!" is sent to the natural language processing engine, and the engine returns an offensiveness score of "70" for this comment.

[0825] Step 3:

[0826] The server associates the calculated offensiveness score with each comment data and stores it in the database. The input is the comment data with the calculated offensiveness score, and the output is the comment data stored in the database and its offensiveness score. For example, the comment "This news is ridiculous!" is associated with an offensiveness score of "70" and stored in the database.

[0827] Step 4:

[0828] The user sets the filter strength to their preference on the device. The input is the user's filter strength setting value, and the output is the setting data sent to the server. Specifically, when user B sets the filter strength to "20," this setting is confirmed on the device and sent to the server.

[0829] Step 5:

[0830] The server filters comments based on the filter strength set by the user. The input is the user's filter strength setting and the comment data in the database, and the output is a filtered comment list. Specifically, based on a filter strength of "20," comments with an offensiveness score of 20 or more are excluded.

[0831] Step 6:

[0832] The server sends the filtered comment list to the user's device. The input is the filtered comment list, and the output is the comments displayed on the user's device. For example, on User B's device, a comment with an offensiveness score of "10" such as "That was very interesting," is displayed, while a comment with an offensiveness score of "25" such as "That's a really awful story," is hidden.

[0833] (Application example 1)

[0834] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0835] Comment sections on news sites and social media often contain offensive comments that cause discomfort and psychological stress to users. It is important to effectively filter these comments and provide an environment where users can safely view and participate in the comments. However, existing systems face challenges in filtering offensive comments in real time and dynamically updating the user display. Furthermore, they lack the ability for users to adjust the filter strength according to their preferences, making it difficult to display comments that meet the needs of individual users.

[0836] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0837] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score, means for saving the calculated offensiveness score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for transmitting a filtered comment list to the user's display device, means for adjusting the filter strength on the user's display device, and means for maintaining communication with the display device to dynamically perform the filtering process. This allows users to filter offensive comments in real time and comfortably view comments with a filter strength that suits their preferences.

[0838] "Means for collecting posted comment data" refers to a system for collecting comments posted on the Internet by users.

[0839] A "natural language processing engine" is software or algorithm that analyzes collected comment data and understands and evaluates its content.

[0840] The "offensiveness score" is a numerical indication of the degree of offensiveness of a comment, and is an index calculated by a natural language processing engine.

[0841] "Means for storing in a database" refers to a system for centrally managing and storing calculated offensive scores and comment data.

[0842] The "means for setting filter strength" is an interface that allows the user to set the threshold of the aggressiveness score according to their own preferences.

[0843] The "means for filtering comments" is a mechanism for selecting comments to be displayed based on the set filter strength.

[0844] "Means for transmitting to a user's display device" is a mechanism for transmitting filtered comments to a device used by the user (e.g., smart glasses).

[0845] The "means for adjusting the filter strength" is an adjustment function that allows the filter strength changed by the user in real time to be immediately reflected.

[0846] The "means for maintaining communication with the display device" is a system for constantly maintaining a connection with the user's display device in order to dynamically perform comment filtering processing.

[0847] "Dynamic update means" refers to a function that changes the comment section displayed in real time depending on the set filter strength and newly posted comments.

[0848] This invention is a system for filtering offensive comments in the comment sections of news sites and social media. To implement this system, the following steps and various hardware and software are required.

[0849] Hardware and software used

[0850] Server: Used to collect, analyze, score, and store comment data in a database.

[0851] Natural language processing engine: Used to analyze the offensiveness of comments and calculate an offensiveness score.

[0852] Database: Used to store calculated aggression scores and comment data.

[0853] User display device (e.g., smart glasses): Used to display filtered comments.

[0854] HTTP request library (e.g. requests for Python): Used to communicate between the server and various APIs.

[0855] Overview of the overall system processing flow

[0856] 1. Collection of comment data

[0857] The server collects comment data posted from news sites and social media, including the comment text, the poster's user ID, and the posting date and time.

[0858] 2. Calculating the Aggression Score

[0859] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment.

[0860] 3. Saving to the database

[0861] The calculated offensiveness score is stored in a database together with the corresponding comment data.

[0862] 4. Setting the filter strength

[0863] Users set the filter strength (threshold of offensiveness score) for comments via a display device such as smart glasses, which is done through a user interface.

[0864] 5. Comment Filtering

[0865] The server filters the comments based on the filter strength set by the user and transmits only the selected comments to the display device.

[0866] 6. Dynamic Updates

[0867] The server performs filtering processing in real time, even when a new comment is posted or when a user changes the filter strength, and dynamically updates the filtering results while maintaining communication with the display device.

[0868] Specific examples

[0869] For example, suppose user A posts "This news is stupid!" in the comments section of a news article. This comment is collected by the server and passed to a natural language processing engine. The server then calculates an offensiveness score of 70 for this comment and stores it in the database.

[0870] Later, when User B uses the smart glasses to view a news article, he sets the filter strength to "20." Based on this setting, the server hides comments with an offensiveness score of 20 or more. Therefore, a comment such as "This news is ridiculous!" will not be displayed. On the other hand, a comment with an offensiveness score of "5," such as "This was very interesting," will be displayed.

[0871] Prompt Sentence Examples

[0872] Example configuration:

[0873] article_id = 67890 Example of article ID setting

[0874] filter_threshold = 50 Example of filter strength setting

[0875] Sample prompt sentence:

[0876] Input prompt: "Please filter and display comments for article ID "67890" using filter strength "50"."

[0877] Example output:

[0878] Displayed comment: "This is very useful information." (Aggression score: 10)

[0879] Comment to be hidden: "This article is rubbish!" (Aggression score: 60)

[0880] This system allows users to filter out offensive comments in real time, allowing them to comfortably view comments with a filter strength that suits their preferences.

[0881] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0882] Step 1:

[0883] The server collects comment data posted from news sites and social media. This collected data includes the comment text, the poster's user ID, the posting date and time, etc. The input requires the URL of the news article or social media post to be retrieved, and the collected comment data is obtained as output.

[0884] Step 2:

[0885] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment. This process requires the collected comment data as input and obtains an offensiveness score corresponding to each comment as output. Specific operations include making an HTTP request to the natural language processing engine API.

[0886] Step 3:

[0887] The server stores the calculated offensiveness score in a database. This process requires the comment data and the corresponding offensiveness score as input, and obtains the database state as output. Specific operations include creating and updating records in the database.

[0888] Step 4:

[0889] The user sets the filter strength (threshold of the aggression score) through the interface of the smart glasses. This process requires the user to input the filter strength and saves the setting as output. Specific actions include entering a number or operating a slider on the user interface.

[0890] Step 5:

[0891] The server filters comments based on the filter strength set by the user. This process requires the comment data and offensiveness scores stored in the database and the filter strength set by the user as input, and obtains a filtered comment list as output. Specific operations include a process of comparing the offensiveness score and the filter strength.

[0892] Step 6:

[0893] The server sends the filtered comment list to the user's display device. The filtered comment list is required as input, and the comments displayed on the user's display device are obtained as output. Specific operations include a process of sending data to the display device.

[0894] Step 7:

[0895] When a user sets a new filter strength, the server performs the filtering process again and sends the result to the display device. The new filter strength, existing comment data, and offensiveness scores are required as input, and the latest filtering result is obtained as output. Specific operations include the re-filtering and sending process.

[0896] Step 8:

[0897] When a new comment is posted, the server recalculates the offensiveness score of the comment, updates the database, and regenerates the filtering results and sends them to the display device. New comment data is required as input, and updated filtering results are obtained as output. Specific operations include automatic collection, score calculation, database update, and re-filtering.

[0898] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0899] The present invention provides a system that filters offensive comments in the comment section of a news site and provides an appropriate comment display environment that takes into account the user's feelings. The system aims to create an environment where users can view and participate in comments with peace of mind.

[0900] First, the server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. For example, if user A posts a comment saying, "This news is ridiculous!", the content of that comment is recorded by the server.

[0901] The server then passes the collected comment data to a natural language processing engine, which calculates an "aggression score" for each comment. For example, the natural language processing engine calculates an aggression score of 70 for a comment such as "This news is ridiculous!". The emotion engine also evaluates the poster's emotional state from the comment, and reflects the evaluation result in the aggression score. For example, a comment with a strong emotion of "anger" is assigned a higher aggression score.

[0902] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[0903] Next, the user (visitor) can set the filter strength to their preference on the device, or the emotion engine can recognize the user's current emotional state and automatically set the recommended filter strength. For example, User C can set the filter strength to "20," or the emotion engine can recognize User C's "irritation" and recommend the filter strength to "50."

[0904] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it references the offensiveness score of the comment and hides comments that exceed the standard. For example, for user C who sets the filter strength to "50", comments with an offensiveness score of 50 or more will be hidden.

[0905] The server then sends the filtered comment list to the user's device, allowing the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User C's device, comments with an offensiveness score of "70" are hidden, and comments with an offensiveness score of "30" are displayed.

[0906] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[0907] This configuration allows users to freely adjust the filter strength and enjoy a comment viewing experience that suits their preferences. It also improves the overall user experience of news sites by promoting healthy discussion and protecting users from offensive comments. Furthermore, the emotion engine provides appropriate filtering based on the user's emotional state, providing a more personalized viewing experience.

[0908] The processing flow will be explained below.

[0909] Step 1:

[0910] The server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. All posted comments are recorded in a database.

[0911] Example: When user A posts a comment saying "This news is ridiculous!", the server records the comment content and poster information in a database.

[0912] Step 2:

[0913] The server passes the collected comment data to a natural language processing engine, calculates an offensiveness score for each comment, and uses an emotion engine to evaluate the poster's emotional state based on the comment.

[0914] Example: In response to the comment "This news is ridiculous!", the natural language processing engine calculates an aggression score of "70" and the emotion engine evaluates it as "anger."

[0915] Step 3:

[0916] The server associates the aggression score calculated by the natural language processing engine and the emotional state evaluated by the emotion engine with the comment data and stores them in a database.

[0917] Example: Add an aggression score of 70 and emotion of anger to the record of the comment "This news is stupid!" and save it in the database.

[0918] Step 4:

[0919] Users (visitors) can set the filter strength to suit their preferences on their devices, or the emotion engine will recognize the user's current emotional state and automatically set the recommended filter strength.

[0920] Example: User C sets the filter strength to "20." Or, the emotion engine recognizes User C's "irritation" and recommends setting the filter strength to "50."

[0921] Step 5:

[0922] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it refers to the offensiveness score of the comment and hides comments that exceed the standard.

[0923] Example: For user C who has set the filter strength to "50", comments with an offensive score of 50 or more will be hidden.

[0924] Step 6:

[0925] The server sends the filtered comment list to the user's terminal.

[0926] Example: A comment list containing only comments with an offensiveness score of less than 50 is sent to user C's device.

[0927] Step 7:

[0928] The device updates the comment section based on the filtering results received from the server, allowing users to view only the filtered comments.

[0929] Example: On User C's device, comments with an aggression score of "70" are hidden, and comments with an aggression score of "30" are displayed.

[0930] Step 8:

[0931] When a new comment is posted or a user changes the filter strength, the server recalculates the offensiveness score and dynamically updates the filtering results.

[0932] Example: When a new comment "This is awful!" is posted, its offensiveness score is calculated and the list of comments displayed is instantly updated based on the existing filter strength settings.

[0933] Example 2

[0934] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0935] On traditional news sites, users are often exposed to offensive comments, making it difficult to ensure a comfortable viewing environment. In particular, the proliferation of offensive comments can hinder healthy discussion and increase the mental burden on users. Furthermore, traditional filtering systems are static and filter content based on uniform criteria without considering the user's emotional state, making it difficult to provide a personalized experience.

[0936] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0937] In this invention, the server includes a means for collecting posted comment data, a means for analyzing the collected comment data using a natural language processing engine to calculate an offensiveness score, and a means for analyzing the collected comment data using a sentiment analysis engine to evaluate the poster's emotional state. This enables filtering according to individual emotional states while preventing users from being exposed to offensive comments. The server also includes a means for storing the comment's offensiveness score and the emotional state evaluation results in a database, a means for the user to set the filter strength, and a means for recognizing the user's emotional state using the sentiment analysis engine and setting a recommended filter strength, thereby achieving flexible filtering according to individual user needs. The server also includes a means for transmitting a filtered comment list to the user's device and a means for recalculating the offensiveness score and updating the filtering results when a new comment is posted, providing a constantly up-to-date comment viewing environment. This improves the user experience across the news site and promotes healthy discussion.

[0938] "Posted comment data" refers to data that includes information such as the content of comments submitted by users of a news site, the poster's user ID, and the date and time of posting.

[0939] A "natural language processing engine" refers to a program or algorithm that analyzes text data and understands its content.

[0940] "Aggression score" refers to a numerical representation of the degree of aggression based on the content of a comment.

[0941] "Sentiment analysis engine" refers to a program or algorithm that identifies emotions contained in text data and evaluates that emotional state.

[0942] "User emotional state" refers to the current emotional state inferred from the user's input and actions.

[0943] "Filter strength" refers to the threshold of the aggressiveness score that serves as the basis for filtering.

[0944] A "database" refers to a system for systematically managing, storing, and searching information.

[0945] "Filtering" refers to the operation of selecting and displaying specific data based on pre-set criteria.

[0946] "User device" refers to a terminal, such as a PC, smartphone, or tablet, that receives and displays data sent from the server.

[0947] "Recalculation" refers to performing a new calculation process, and in this context specifically refers to recalculating the offensiveness score for newly posted comments.

[0948] "Dynamic update" refers to an operation in which information or data is automatically updated in real time or near real time.

[0949] "News site" refers to a website that provides news information on the Internet and allows users to post comments.

[0950] A "proper noun" is a name used to refer to a specific person, place, organization, etc., but in this context is written as a general name rather than a specific name.

[0951] The present invention is a system that filters offensive comments in the comment section of a news site and provides an appropriate comment display environment that takes into account the user's feelings. The purpose of this system is to create an environment where users can safely view and participate in comments.

[0952] First, the server collects all comment data posted on the news site. This collection includes information such as the comment content, the poster's user ID, and the posting date and time. For example, if User A posts a comment saying "This news is ridiculous!", the content of that comment is recorded by the server. For this purpose, the server uses a database management system (such as MySQL or PostgreSQL).

[0953] The server then passes the collected comment data to a natural language processing engine (such as Python's NLTK or spaCy) to calculate an offensiveness score for each comment. For example, the natural language processing engine calculates an offensiveness score of 70 for the comment "This news is ridiculous!". It also uses an emotion engine (such as Google Cloud Natural Language API) to evaluate the poster's emotional state from the comment, and reflects the evaluation result in the offensiveness score. For example, a comment with a strong emotional response of "anger" is assigned a higher offensiveness score.

[0954] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[0955] Next, the user (visitor) can set the filter strength to their preference on their device. Furthermore, the emotion engine can recognize the user's current emotional state and automatically set the recommended filter strength. For example, User C sets the filter strength to "20," or the emotion engine recognizes User C's "irritation" and recommends the filter strength to "50."

[0956] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it references the offensiveness score of the comment and hides comments that exceed the standard. For example, for user C who sets the filter strength to "50", comments with an offensiveness score of 50 or more will be hidden.

[0957] The server then sends the filtered comment list to the user's device, allowing the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User C's device, comments with an offensiveness score of "70" are hidden, and comments with an offensiveness score of "30" are displayed.

[0958] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[0959] Explanation with concrete examples

[0960] Prompt Sentence Examples

[0961] text

[0962] User A posts a comment saying, "This news is ridiculous!" Calculate the aggression score of this comment and evaluate the user's emotional state. For example, if the aggression score is 70 and the user's emotional state is recognized as "anger," determine whether the comment should be hidden as a result of filtering.

[0963] This configuration allows users to freely adjust the filter strength and enjoy a comment viewing experience that suits their preferences. It also improves the overall user experience of news sites by promoting healthy discussion and protecting users from offensive comments. Furthermore, the emotion engine provides appropriate filtering based on the user's emotional state, providing a more personalized viewing environment.

[0964] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0965] Step 1:

[0966] The server collects comment data posted on news sites.

[0967] Input: Comment content, poster's user ID, posting date and time

[0968] Specific operation: When user A posts a comment saying "This news is ridiculous!", the server records the comment content, user A's ID, and the posting date and time, and saves them in a database.

[0969] Output: Collected comment data (e.g., comment content "This news is ridiculous!", user ID "A", posting date and time "2023-10-01 12:34:56")

[0970] Step 2:

[0971] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score.

[0972] Input: Collected comment data

[0973] Specific behavior: The server inputs the comment "This news is stupid!" into a natural language processing engine (e.g., Python's NLTK or spaCy) for text analysis, resulting in an offensiveness score of "70."

[0974] Output: Aggression score (e.g., 70)

[0975] Step 3:

[0976] The server passes the collected comment data to a sentiment analysis engine to evaluate the poster's emotional state.

[0977] Input: Collected comment data

[0978] Specific operation: The server inputs the same comment into a sentiment analysis engine (e.g., Google Cloud Natural Language API) and detects that the sentiment "anger" is high.

[0979] Output: Emotional state rating (e.g., anger 60 / 100)

[0980] Step 4:

[0981] The server associates the calculated aggression score and the evaluation results of the emotional state with each comment data and stores them in a database.

[0982] Input: Aggression score and emotional state assessment results

[0983] Specific Action: Give the comment "This news is stupid!" an aggression score of "70" and an emotion rating of "Anger 60" and save it in the database.

[0984] Output: Comment data stored in the database and its rating (e.g., comment "This news is stupid!", aggression score "70", emotion "anger 60")

[0985] Step 5:

[0986] Users (visitors) can set the filter strength on their terminal to suit their preferences.

[0987] Input: User-input filter strength setting

[0988] Specific actions: User C sets the filter strength to "20" on their device, or the emotion analysis engine recognizes User C's "irritation" and recommends a filter strength of "50."

[0989] Output: User-defined filter strength (e.g., filter strength 20 or 50)

[0990] Step 6:

[0991] The server filters the comments based on the filter strength set by the user.

[0992] Input: User-defined filter strength and comment data

[0993] Specific operation: The server references user C's filter strength of "50" and retrieves the offensiveness score from the database. Comments with a score of "50" or higher are hidden.

[0994] Output: Filtered comment list (e.g. hide comments with a score of 70, show comments with a score of 30)

[0995] Step 7:

[0996] The server sends the filtered comment list to the user's terminal.

[0997] Input: filtered comment list

[0998] Specific operation: The server sends user C, who has a filter strength of "50," a list that hides comments with an offensive score of "70" and displays comments with a score of "30."

[0999] Output: A list of comments displayed on the user's device (e.g., comments with an offensiveness score of 70 are hidden, and comments with a score of 30 are displayed)

[1000] Step 8:

[1001] The server recalculates the offensiveness score when a new comment is posted and updates the filtering results.

[1002] Input: Newly posted comment data

[1003] Specific operation: When a new comment is posted, the server recalculates the score and emotional state using the natural language processing engine and sentiment analysis engine, and performs filtering.

[1004] Output: Updated filtered comment list (e.g., latest list with new comments)

[1005] Step 9:

[1006] If the user changes the filter strength, the server recalculates the displayed comment list and keeps it up to date.

[1007] Input: User changes to filter strength

[1008] What happens: When the user changes the filter strength, the server re-filters using the new criteria.

[1009] Output: Updated comment list (e.g., a list of comments based on the new filter strength)

[1010] (Application example 2)

[1011] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1012] In conventional content distribution services, some of the comments posted by users can be offensive, disrupting other users' comfortable viewing experiences. Furthermore, especially in environments using head-mounted displays, offensive comments can enter the user's field of vision, potentially causing psychological stress. It is necessary to provide an environment where users can enjoy content safely by automatically detecting and appropriately filtering these offensive comments.

[1013] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1014] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine and calculating an aggression score, means for saving the calculated aggression score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for transmitting a filtered comment list to the user's terminal, and means for filtering comments displayed on the head-mounted display. This allows users to filter aggressive comments according to their own emotional state and enjoy content with peace of mind.

[1015] "Posted comment data" refers to text information including ratings and opinions posted by users on the content distribution service.

[1016] A "natural language processing engine" is a software tool that analyzes text data to understand the meaning and sentiment of a sentence.

[1017] The "aggression score" is a numerical indicator of how aggressive the posted comment data is.

[1018] A "database" is a system for efficiently managing and storing collected data.

[1019] The "filter strength" is a reference value of the offensiveness score set by the user to determine whether or not to display.

[1020] "Filtering" is the process of filtering out harmful comments based on set criteria.

[1021] A "head-mounted display" is a display device that is worn on the user's head.

[1022] "Dynamic update" refers to a process of automatically changing the display content in response to the addition of new data or changes in filter strength.

[1023] "Real-time" means immediate, real-time processing or response.

[1024] This invention relates to a comment filtering system for content distribution services, and aims to improve the comment display environment when viewing using a head-mounted display (HMD). This system determines whether posted comments are offensive and filters them based on that, providing an environment in which users can enjoy content with peace of mind.

[1025] First, the server collects comment data posted by users. This data includes the comment text, the poster's identification information, and the posting date and time. Next, the collected comment data is analyzed using a natural language processing engine to calculate an offensiveness score. The natural language processing engine used is something like the VADER Sentiment Intensity Analyzer.

[1026] The calculated offensiveness score is stored in a database, which allows the offensiveness score of each comment to be managed. Users can set the filter strength via their own device or head-mounted display. This filter strength is the standard value that determines which comments are displayed. Based on the filter strength, comments with high offensive scores are filtered out, ensuring that only appropriate comments are displayed to the user.

[1027] The filtered comment list is sent to the user's device or head-mounted display. The system is dynamically updated, so the offensiveness score is recalculated every time a new comment is posted, and the filtering results are updated immediately. Furthermore, the filtering results are reflected in real time on the head-mounted display, allowing users to view the latest comments.

[1028] Specific examples of use

[1029] For example, if a user posts a comment such as "This video is not funny at all!", the server collects this comment and analyzes it using a natural language processing engine. The analysis results in an "offensiveness score of 70" which is then stored in the database. If the user sets the filter strength to "50," this comment will not be displayed, and only comments with an offensiveness score of 50 or less will be displayed.

[1030] Prompt Sentence Examples

[1031] The following prompt sentences are used by the system to collect user comments and calculate an offensiveness score:

[1032] Text format

[1033] comments_payload = {

[1034] "content": "This video is not funny at all!",

[1035] "user_id": "user123",

[1036] "post_date": "2023-10-01"

[1037] }

[1038] response = requests.post("http: / / localhost:5000 / add_comment", json=comments_payload)

[1039] print(response.json())

[1040] The prompts to set the filter strength and get the filtered comments are as follows:

[1041] Text format

[1042] filter_payload = {

[1043] "filter_strength": 50 Filter strength set by the user

[1044] }

[1045] response = requests.post("http: / / localhost:5000 / get_comments", json=filter_payload)

[1046] print(response.json())

[1047] This allows users to effectively filter out offensive comments and enjoy content comfortably, even when using content distribution services that use head-mounted displays.

[1048] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1049] Step 1:

[1050] The user puts on the head-mounted display, starts using the content distribution service, and posts comments.

[1051] Input: Comment text posted by the user, poster's identification information, posting date and time

[1052] Output: Comment data is sent to the server

[1053] Step 2:

[1054] The server receives the comment data sent by the user and temporarily stores it in a database.

[1055] Input: Comment data (text, identification information, posting date and time)

[1056] Output: Comment data is saved to the database.

[1057] Step 3:

[1058] The server passes the comment data collected from the database to a natural language processing engine and calculates the offensiveness score of the comment.

[1059] Input: Comment data stored in the database

[1060] Output: Aggression score is calculated

[1061] Step 4:

[1062] Based on the aggression score calculated by the natural language processing engine, the server links the aggression score to the comment data and stores it in a database.

[1063] Input: Aggression scores and comment data

[1064] Output: The linked data is saved in the database

[1065] Step 5:

[1066] The user sets the filter strength through the head-mounted display.

[1067] Input: Filter strength value set by the user

[1068] Output: Filter strength setting information is sent to the server

[1069] Step 6:

[1070] The server performs filtering based on the filter strength set by the user, and hides comments whose offensiveness score exceeds the filter strength.

[1071] Input: Comment data and user-defined filter strength

[1072] Output: Filtered comment list

[1073] Step 7:

[1074] The filtered comment list is displayed in real time on a head-mounted display.

[1075] Input: filtered comment list

[1076] Output: Comments displayed on the head-mounted display

[1077] Step 8:

[1078] When a new comment is posted, the server recalculates the offensiveness score and immediately updates the filtering results.

[1079] Input: New comment data

[1080] Output: The updated filtering results are reflected on the head-mounted display.

[1081] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1082] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1083] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1084] [Fourth embodiment]

[1085] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1086] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1087] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1088] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1089] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1090] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1091] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1092] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1093] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1094] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1095] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1096] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1097] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1098] The present invention provides a system for filtering offensive comments in the comment sections of news sites, with the goal of creating an environment where users can safely view and participate in comments.

[1099] First, the server collects all posted comment data. This data collection includes metadata such as the comment content, the poster's user ID, and the posting date and time. For example, if user A posts a comment saying, "This news is ridiculous!", the content of that comment will be recorded by the server.

[1100] The server then passes the collected comment data to a natural language processing engine, which calculates an "offense score" for each comment. For example, the natural language processing engine calculates an offensive score of 70 for a comment such as "This news is ridiculous!" On the other hand, a comment such as "That was very interesting" would receive an offensive score of 5.

[1101] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[1102] Next, the user (visitor) sets the filter strength to their preference on the device. The filter strength (threshold of the offensive score) can be input or selected through the user interface. For example, if user B sets the filter strength to "20," this setting is confirmed on the device.

[1103] The server then filters the comments based on the filter strength set by the user. This process hides any comments with an offensiveness score above the set threshold. For example, if user B's filter strength is "20," any comments with an offensiveness score of 20 or higher will be hidden.

[1104] Finally, the server sends the filtered comment list to the user's device. This allows the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User B's device, a comment with an offensiveness score of 25, such as "This is a really awful story," will be hidden, while a comment with an offensiveness score of 10, such as "That was very interesting," will be displayed.

[1105] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[1106] This configuration allows users to freely adjust the filter strength and customize their comment viewing experience, promoting healthy discussion and protecting users from offensive comments, improving the overall user experience on news sites.

[1107] The processing flow will be explained below.

[1108] Step 1:

[1109] The server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. All posted comments are recorded in a database.

[1110] Example: When user A posts a comment saying "This news is ridiculous!", the server records the comment content and poster information in a database.

[1111] Step 2:

[1112] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment.

[1113] Example: For the comment "This news is ridiculous!", the natural language processing engine calculates an offensive score of "70".

[1114] Step 3:

[1115] The server associates the offensiveness score calculated by the natural language processing engine with the comment data and stores it in a database.

[1116] Example: Add an offensive score of 70 to the record of the comment "This news is stupid!" and save it in the database.

[1117] Step 4:

[1118] Users (visitors) can set their own preferred filter strength on their devices by inputting the filter strength (threshold of the offensiveness score) through the user interface.

[1119] Example: When user B sets the filter strength to "20", the setting is confirmed on the device.

[1120] Step 5:

[1121] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it refers to the offensiveness score of the comment and hides comments that exceed the standard.

[1122] Example: For user B who has set the filter strength to "20", comments with an offensive score of 20 or more will be hidden.

[1123] Step 6:

[1124] The server sends the filtered comment list to the user's terminal.

[1125] Example: Send a comment list to user B's device that includes only comments with an offensiveness score of less than 20.

[1126] Step 7:

[1127] The device updates the comment section based on the filtering results received from the server, allowing users to view only the filtered comments.

[1128] Example: On User B's device, comments with an offensiveness score of "25" are hidden, and comments with an offensiveness score of "10" are displayed.

[1129] Step 8:

[1130] When a new comment is posted or a user changes the filter strength, the server recalculates the offensiveness score and dynamically updates the filtering results.

[1131] Example: When a new comment "This is awful!" is posted, its offensiveness score is calculated and the list of comments displayed is instantly updated based on the existing filter strength settings.

[1132] Example 1

[1133] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1134] In the comment sections of conventional news sites, offensive comments are common, making it difficult for users to safely view and participate in the comments. It is also difficult to adapt to the different filter strength preferences of each user, so a system that dynamically updates the comment filtering and display has been required.

[1135] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1136] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine to calculate an offensiveness score, means for saving the calculated offensiveness score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for sending a filtered comment list to the user's terminal, and means for dynamically reflecting the filter strength setting on the server. This allows users to filter comments according to their preferences in real time, and provides a comment section that promotes healthy discussion.

[1137] "Comment data" refers to the text that users post to news sites and the metadata associated with those posts.

[1138] A "natural language processing engine" refers to software or algorithms that analyze human language to understand its meaning and sentiment and then perform specific tasks.

[1139] "Aggression score" refers to a numerical indicator that indicates whether a posted comment is offensive or not.

[1140] "Users" refer to people who visit news sites, post comments, and set filter strengths.

[1141] "Filter strength" refers to the threshold for the offensiveness score of comments set by the user, which determines the range of comments that will be displayed.

[1142] "Database" refers to an electronic storage system for storing and managing collected comment data and offensiveness scores.

[1143] "Device" refers to the electronic device (e.g., computer, smartphone, tablet, etc.) that a user uses to access a news site.

[1144] "Dynamic reflection" means that user settings and new comment posts are applied to the system in real time.

[1145] MODE FOR CARRYING OUT THE INVENTION

[1146] The present invention provides a system for filtering offensive comments in the comment section of a news site. The system aims to create an environment where users can safely view and participate in comments. A specific embodiment of the system is described below.

[1147] First, this system consists of multiple hardware and software components: a server, a natural language processing engine, a database, a user interface (UI), and a terminal.

[1148] 1. Collection of comment data

[1149] The server collects all comment data posted on the news site. This collected data includes metadata such as the comment content, the poster's user ID, and the posting date and time. For example, if user A posts a comment saying "This news is ridiculous!", the content and metadata are recorded by the server.

[1150] 2. Calculating offensive scores using a natural language processing engine

[1151] The server passes the collected comment data to a natural language processing engine, which calculates an "offense score" for each comment. This natural language processing engine uses, for example, Google NLP API or OpenAI's GPT-3. For example, the natural language processing engine calculates an offensiveness score of 70 for a comment such as "This news is ridiculous!". On the other hand, a comment such as "That was very interesting" would be assigned an offensiveness score of 5.

[1152] 3. Save the calculated aggression score to the database

[1153] The server associates the calculated offensiveness score with each comment data and stores it in a database. This database can be an SQL database or a NoSQL database. For example, a comment such as "This news is ridiculous!" is associated with an offensiveness score of "70" and stored in the database.

[1154] 4. User-defined filter strength

[1155] Users can set their preferred filter strength through the user interface on their device. The UI is provided via a web browser or mobile app. Users can input or select the filter strength (threshold of the aggressiveness score). For example, if user B sets the filter strength to "20," this setting is sent to the server in real time.

[1156] 5. Filtering process and sending results

[1157] The server filters comments based on the filter strength set by the user. This process hides comments with an offensiveness score above the set threshold. For example, if user B's filter strength is "20," comments with an offensiveness score of 20 or more will be hidden. The filtered comment list is sent from the server to the user's device, where it can be viewed by the user.

[1158] Examples of concrete examples and prompts

[1159] Examples:

[1160] 1. User A posts a comment saying, "This news is great!"

[1161] 2. The server collects these comments and passes them to a natural language processing engine.

[1162] 3. The natural language processing engine calculates an offensiveness score of "1".

[1163] 4. The server associates this score with the comment data and stores it in a database.

[1164] 5. User B sets the filter strength to "10".

[1165] 6. Based on the filter strength of "10", the server sends only comments with an offensiveness score of "less than 10" to the user's device.

[1166] Example prompt:

[1167] We want to make comments on this news site safer and friendlier. If you see any of the following comments posted, please rate their offensiveness:

[1168] Comment: "This news is ridiculous!"

[1169] Comment: "It was very interesting."

[1170] You can also filter comments based on the following settings to view the comments that apply to each setting:

[1171] Filter Strength: 20

[1172] With this setting, which comments will be visible and which will be hidden?

[1173] This allows users to freely adjust the filter strength and customize their comment viewing experience to suit their preferences. It also promotes healthy discussion and protects users from offensive comments, improving the overall user experience on news sites.

[1174] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1175] Program processing flow

[1176] Step 1:

[1177] The server collects all comment data posted on the news site. This includes metadata such as the comment content, the poster's user ID, and the posting date and time. The input is a newly posted comment on the news site, and the output is the comment data stored in the server. For example, if user A posts "This news is stupid!", the content and metadata are recorded by the server.

[1178] Step 2:

[1179] The server passes the collected comment data to a natural language processing engine. The input is the comment data stored in the server, and the output is an offensiveness score for each comment. Specifically, the comment "This news is stupid!" is sent to the natural language processing engine, and the engine returns an offensiveness score of "70" for this comment.

[1180] Step 3:

[1181] The server associates the calculated offensiveness score with each comment data and stores it in the database. The input is the comment data with the calculated offensiveness score, and the output is the comment data stored in the database and its offensiveness score. For example, the comment "This news is ridiculous!" is associated with an offensiveness score of "70" and stored in the database.

[1182] Step 4:

[1183] The user sets the filter strength to their preference on the device. The input is the user's filter strength setting value, and the output is the setting data sent to the server. Specifically, when user B sets the filter strength to "20," this setting is confirmed on the device and sent to the server.

[1184] Step 5:

[1185] The server filters comments based on the filter strength set by the user. The input is the user's filter strength setting and the comment data in the database, and the output is a filtered comment list. Specifically, based on a filter strength of "20," comments with an offensiveness score of 20 or more are excluded.

[1186] Step 6:

[1187] The server sends the filtered comment list to the user's device. The input is the filtered comment list, and the output is the comments displayed on the user's device. For example, on User B's device, a comment with an offensiveness score of "10" such as "That was very interesting," is displayed, while a comment with an offensiveness score of "25" such as "That's a really awful story," is hidden.

[1188] (Application example 1)

[1189] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1190] Comment sections on news sites and social media often contain offensive comments that cause discomfort and psychological stress to users. It is important to effectively filter these comments and provide an environment where users can safely view and participate in the comments. However, existing systems face challenges in filtering offensive comments in real time and dynamically updating the user display. Furthermore, they lack the ability for users to adjust the filter strength according to their preferences, making it difficult to display comments that meet the needs of individual users.

[1191] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1192] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score, means for saving the calculated offensiveness score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for transmitting a filtered comment list to the user's display device, means for adjusting the filter strength on the user's display device, and means for maintaining communication with the display device to dynamically perform the filtering process. This allows users to filter offensive comments in real time and comfortably view comments with a filter strength that suits their preferences.

[1193] "Means for collecting posted comment data" refers to a system for collecting comments posted on the Internet by users.

[1194] A "natural language processing engine" is software or algorithm that analyzes collected comment data and understands and evaluates its content.

[1195] The "offensiveness score" is a numerical indication of the degree of offensiveness of a comment, and is an index calculated by a natural language processing engine.

[1196] "Means for storing in a database" refers to a system for centrally managing and storing calculated offensive scores and comment data.

[1197] The "means for setting filter strength" is an interface that allows the user to set the threshold of the aggressiveness score according to their own preferences.

[1198] The "means for filtering comments" is a mechanism for selecting comments to be displayed based on the set filter strength.

[1199] "Means for transmitting to a user's display device" is a mechanism for transmitting filtered comments to a device used by the user (e.g., smart glasses).

[1200] The "means for adjusting the filter strength" is an adjustment function that allows the filter strength changed by the user in real time to be immediately reflected.

[1201] The "means for maintaining communication with the display device" is a system for constantly maintaining a connection with the user's display device in order to dynamically perform comment filtering processing.

[1202] "Dynamic update means" refers to a function that changes the comment section displayed in real time depending on the set filter strength and newly posted comments.

[1203] This invention is a system for filtering offensive comments in the comment sections of news sites and social media. To implement this system, the following steps and various hardware and software are required.

[1204] Hardware and software used

[1205] Server: Used to collect, analyze, score, and store comment data in a database.

[1206] Natural language processing engine: Used to analyze the offensiveness of comments and calculate an offensiveness score.

[1207] Database: Used to store calculated aggression scores and comment data.

[1208] User display device (e.g., smart glasses): Used to display filtered comments.

[1209] HTTP request library (e.g. requests for Python): Used to communicate between the server and various APIs.

[1210] Overview of the overall system processing flow

[1211] 1. Collection of comment data

[1212] The server collects comment data posted from news sites and social media, including the comment text, the poster's user ID, and the posting date and time.

[1213] 2. Calculating the Aggression Score

[1214] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment.

[1215] 3. Saving to the database

[1216] The calculated offensiveness score is stored in a database together with the corresponding comment data.

[1217] 4. Setting the filter strength

[1218] Users set the filter strength (threshold of offensiveness score) for comments via a display device such as smart glasses, which is done through a user interface.

[1219] 5. Comment Filtering

[1220] The server filters the comments based on the filter strength set by the user and transmits only the selected comments to the display device.

[1221] 6. Dynamic Updates

[1222] The server performs filtering processing in real time, even when a new comment is posted or when a user changes the filter strength, and dynamically updates the filtering results while maintaining communication with the display device.

[1223] Specific examples

[1224] For example, suppose user A posts "This news is stupid!" in the comments section of a news article. This comment is collected by the server and passed to a natural language processing engine. The server then calculates an offensiveness score of 70 for this comment and stores it in the database.

[1225] Later, when User B uses the smart glasses to view a news article, he sets the filter strength to "20." Based on this setting, the server hides comments with an offensiveness score of 20 or more. Therefore, a comment such as "This news is ridiculous!" will not be displayed. On the other hand, a comment with an offensiveness score of "5," such as "This was very interesting," will be displayed.

[1226] Prompt Sentence Examples

[1227] Example configuration:

[1228] article_id = 67890 Example of article ID setting

[1229] filter_threshold = 50 Example of filter strength setting

[1230] Sample prompt sentence:

[1231] Input prompt: "Please filter and display comments for article ID "67890" using filter strength "50"."

[1232] Example output:

[1233] Displayed comment: "This is very useful information." (Aggression score: 10)

[1234] Comment to be hidden: "This article is rubbish!" (Aggression score: 60)

[1235] This system allows users to filter out offensive comments in real time, allowing them to comfortably view comments with a filter strength that suits their preferences.

[1236] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1237] Step 1:

[1238] The server collects comment data posted from news sites and social media. This collected data includes the comment text, the poster's user ID, the posting date and time, etc. The input requires the URL of the news article or social media post to be retrieved, and the collected comment data is obtained as output.

[1239] Step 2:

[1240] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score for each comment. This process requires the collected comment data as input and obtains an offensiveness score corresponding to each comment as output. Specific operations include making an HTTP request to the natural language processing engine API.

[1241] Step 3:

[1242] The server stores the calculated offensiveness score in a database. This process requires the comment data and the corresponding offensiveness score as input, and obtains the database state as output. Specific operations include creating and updating records in the database.

[1243] Step 4:

[1244] The user sets the filter strength (threshold of the aggression score) through the interface of the smart glasses. This process requires the user to input the filter strength and saves the setting as output. Specific actions include entering a number or operating a slider on the user interface.

[1245] Step 5:

[1246] The server filters comments based on the filter strength set by the user. This process requires the comment data and offensiveness scores stored in the database and the filter strength set by the user as input, and obtains a filtered comment list as output. Specific operations include a process of comparing the offensiveness score and the filter strength.

[1247] Step 6:

[1248] The server sends the filtered comment list to the user's display device. The filtered comment list is required as input, and the comments displayed on the user's display device are obtained as output. Specific operations include a process of sending data to the display device.

[1249] Step 7:

[1250] When a user sets a new filter strength, the server performs the filtering process again and sends the result to the display device. The new filter strength, existing comment data, and offensiveness scores are required as input, and the latest filtering result is obtained as output. Specific operations include the re-filtering and sending process.

[1251] Step 8:

[1252] When a new comment is posted, the server recalculates the offensiveness score of the comment, updates the database, and regenerates the filtering results and sends them to the display device. New comment data is required as input, and updated filtering results are obtained as output. Specific operations include automatic collection, score calculation, database update, and re-filtering.

[1253] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1254] The present invention provides a system that filters offensive comments in the comment section of a news site and provides an appropriate comment display environment that takes into account the user's feelings. The system aims to create an environment where users can view and participate in comments with peace of mind.

[1255] First, the server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. For example, if user A posts a comment saying, "This news is ridiculous!", the content of that comment is recorded by the server.

[1256] The server then passes the collected comment data to a natural language processing engine, which calculates an "aggression score" for each comment. For example, the natural language processing engine calculates an aggression score of 70 for a comment such as "This news is ridiculous!". The emotion engine also evaluates the poster's emotional state from the comment, and reflects the evaluation result in the aggression score. For example, a comment with a strong emotion of "anger" is assigned a higher aggression score.

[1257] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[1258] Next, the user (visitor) can set the filter strength to their preference on the device, or the emotion engine can recognize the user's current emotional state and automatically set the recommended filter strength. For example, User C can set the filter strength to "20," or the emotion engine can recognize User C's "irritation" and recommend the filter strength to "50."

[1259] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it references the offensiveness score of the comment and hides comments that exceed the standard. For example, for user C who sets the filter strength to "50", comments with an offensiveness score of 50 or more will be hidden.

[1260] The server then sends the filtered comment list to the user's device, allowing the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User C's device, comments with an offensiveness score of "70" are hidden, and comments with an offensiveness score of "30" are displayed.

[1261] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[1262] This configuration allows users to freely adjust the filter strength and enjoy a comment viewing experience that suits their preferences. It also improves the overall user experience of news sites by promoting healthy discussion and protecting users from offensive comments. Furthermore, the emotion engine provides appropriate filtering based on the user's emotional state, providing a more personalized viewing experience.

[1263] The processing flow will be explained below.

[1264] Step 1:

[1265] The server collects all posted comment data, including the content of the comment, the poster's user ID, the posting date and time, etc. All posted comments are recorded in a database.

[1266] Example: When user A posts a comment saying "This news is ridiculous!", the server records the comment content and poster information in a database.

[1267] Step 2:

[1268] The server passes the collected comment data to a natural language processing engine, calculates an offensiveness score for each comment, and uses an emotion engine to evaluate the poster's emotional state based on the comment.

[1269] Example: In response to the comment "This news is ridiculous!", the natural language processing engine calculates an aggression score of "70" and the emotion engine evaluates it as "anger."

[1270] Step 3:

[1271] The server associates the aggression score calculated by the natural language processing engine and the emotional state evaluated by the emotion engine with the comment data and stores them in a database.

[1272] Example: Add an aggression score of 70 and emotion of anger to the record of the comment "This news is stupid!" and save it in the database.

[1273] Step 4:

[1274] Users (visitors) can set the filter strength to suit their preferences on their devices, or the emotion engine will recognize the user's current emotional state and automatically set the recommended filter strength.

[1275] Example: User C sets the filter strength to "20." Or, the emotion engine recognizes User C's "irritation" and recommends setting the filter strength to "50."

[1276] Step 5:

[1277] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it refers to the offensiveness score of the comment and hides comments that exceed the standard.

[1278] Example: For user C who has set the filter strength to "50", comments with an offensive score of 50 or more will be hidden.

[1279] Step 6:

[1280] The server sends the filtered comment list to the user's terminal.

[1281] Example: A comment list containing only comments with an offensiveness score of less than 50 is sent to user C's device.

[1282] Step 7:

[1283] The device updates the comment section based on the filtering results received from the server, allowing users to view only the filtered comments.

[1284] Example: On User C's device, comments with an aggression score of "70" are hidden, and comments with an aggression score of "30" are displayed.

[1285] Step 8:

[1286] When a new comment is posted or a user changes the filter strength, the server recalculates the offensiveness score and dynamically updates the filtering results.

[1287] Example: When a new comment "This is awful!" is posted, its offensiveness score is calculated and the list of comments displayed is instantly updated based on the existing filter strength settings.

[1288] Example 2

[1289] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1290] On traditional news sites, users are often exposed to offensive comments, making it difficult to ensure a comfortable viewing environment. In particular, the proliferation of offensive comments can hinder healthy discussion and increase the mental burden on users. Furthermore, traditional filtering systems are static and filter content based on uniform criteria without considering the user's emotional state, making it difficult to provide a personalized experience.

[1291] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1292] In this invention, the server includes a means for collecting posted comment data, a means for analyzing the collected comment data using a natural language processing engine to calculate an offensiveness score, and a means for analyzing the collected comment data using a sentiment analysis engine to evaluate the poster's emotional state. This enables filtering according to individual emotional states while preventing users from being exposed to offensive comments. The server also includes a means for storing the comment's offensiveness score and the emotional state evaluation results in a database, a means for the user to set the filter strength, and a means for recognizing the user's emotional state using the sentiment analysis engine and setting a recommended filter strength, thereby achieving flexible filtering according to individual user needs. The server also includes a means for transmitting a filtered comment list to the user's device and a means for recalculating the offensiveness score and updating the filtering results when a new comment is posted, providing a constantly up-to-date comment viewing environment. This improves the user experience across the news site and promotes healthy discussion.

[1293] "Posted comment data" refers to data that includes information such as the content of comments submitted by users of a news site, the poster's user ID, and the date and time of posting.

[1294] A "natural language processing engine" refers to a program or algorithm that analyzes text data and understands its content.

[1295] "Aggression score" refers to a numerical representation of the degree of aggression based on the content of a comment.

[1296] "Sentiment analysis engine" refers to a program or algorithm that identifies emotions contained in text data and evaluates that emotional state.

[1297] "User emotional state" refers to the current emotional state inferred from the user's input and actions.

[1298] "Filter strength" refers to the threshold of the aggressiveness score that serves as the basis for filtering.

[1299] A "database" refers to a system for systematically managing, storing, and searching information.

[1300] "Filtering" refers to the operation of selecting and displaying specific data based on pre-set criteria.

[1301] "User device" refers to a terminal, such as a PC, smartphone, or tablet, that receives and displays data sent from the server.

[1302] "Recalculation" refers to performing a new calculation process, and in this context specifically refers to recalculating the offensiveness score for newly posted comments.

[1303] "Dynamic update" refers to an operation in which information or data is automatically updated in real time or near real time.

[1304] "News site" refers to a website that provides news information on the Internet and allows users to post comments.

[1305] A "proper noun" is a name used to refer to a specific person, place, organization, etc., but in this context is written as a general name rather than a specific name.

[1306] The present invention is a system that filters offensive comments in the comment section of a news site and provides an appropriate comment display environment that takes into account the user's feelings. The purpose of this system is to create an environment where users can safely view and participate in comments.

[1307] First, the server collects all comment data posted on the news site. This collection includes information such as the comment content, the poster's user ID, and the posting date and time. For example, if User A posts a comment saying "This news is ridiculous!", the content of that comment is recorded by the server. For this purpose, the server uses a database management system (such as MySQL or PostgreSQL).

[1308] The server then passes the collected comment data to a natural language processing engine (such as Python's NLTK or spaCy) to calculate an offensiveness score for each comment. For example, the natural language processing engine calculates an offensiveness score of 70 for the comment "This news is ridiculous!". It also uses an emotion engine (such as Google Cloud Natural Language API) to evaluate the poster's emotional state from the comment, and reflects the evaluation result in the offensiveness score. For example, a comment with a strong emotional response of "anger" is assigned a higher offensiveness score.

[1309] The server then associates the calculated offensiveness score with each comment and stores it in the database. This same comment is given an offensiveness score of 70 and stored in the database.

[1310] Next, the user (visitor) can set the filter strength to their preference on their device. Furthermore, the emotion engine can recognize the user's current emotional state and automatically set the recommended filter strength. For example, User C sets the filter strength to "20," or the emotion engine recognizes User C's "irritation" and recommends the filter strength to "50."

[1311] The server filters comments based on the filter strength set by the user. Depending on the filter strength, it references the offensiveness score of the comment and hides comments that exceed the standard. For example, for user C who sets the filter strength to "50", comments with an offensiveness score of 50 or more will be hidden.

[1312] The server then sends the filtered comment list to the user's device, allowing the user to view only the comments selected according to the filter strength. For example, in a news article displayed on User C's device, comments with an offensiveness score of "70" are hidden, and comments with an offensiveness score of "30" are displayed.

[1313] The system is also dynamically updated. Every time a new comment is posted, the server recalculates the offensiveness score and updates the filtering results accordingly. Furthermore, when a user changes the filter strength, the displayed comment list is instantly recalculated and kept up to date.

[1314] Explanation with concrete examples

[1315] Prompt Sentence Examples

[1316] text

[1317] User A posts a comment saying, "This news is ridiculous!" Calculate the aggression score of this comment and evaluate the user's emotional state. For example, if the aggression score is 70 and the user's emotional state is recognized as "anger," determine whether the comment should be hidden as a result of filtering.

[1318] This configuration allows users to freely adjust the filter strength and enjoy a comment viewing experience that suits their preferences. It also improves the overall user experience of news sites by promoting healthy discussion and protecting users from offensive comments. Furthermore, the emotion engine provides appropriate filtering based on the user's emotional state, providing a more personalized viewing environment.

[1319] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1320] Step 1:

[1321] The server collects comment data posted on news sites.

[1322] Input: Comment content, poster's user ID, posting date and time

[1323] Specific operation: When user A posts a comment saying "This news is ridiculous!", the server records the comment content, user A's ID, and the posting date and time, and saves them in a database.

[1324] Output: Collected comment data (e.g., comment content "This news is ridiculous!", user ID "A", posting date and time "2023-10-01 12:34:56")

[1325] Step 2:

[1326] The server passes the collected comment data to a natural language processing engine and calculates an offensiveness score.

[1327] Input: Collected comment data

[1328] Specific behavior: The server inputs the comment "This news is stupid!" into a natural language processing engine (e.g., Python's NLTK or spaCy) for text analysis, resulting in an offensiveness score of "70."

[1329] Output: Aggression score (e.g., 70)

[1330] Step 3:

[1331] The server passes the collected comment data to a sentiment analysis engine to evaluate the poster's emotional state.

[1332] Input: Collected comment data

[1333] Specific operation: The server inputs the same comment into a sentiment analysis engine (e.g., Google Cloud Natural Language API) and detects that the sentiment "anger" is high.

[1334] Output: Emotional state rating (e.g., anger 60 / 100)

[1335] Step 4:

[1336] The server associates the calculated aggression score and the evaluation results of the emotional state with each comment data and stores them in a database.

[1337] Input: Aggression score and emotional state assessment results

[1338] Specific Action: Give the comment "This news is stupid!" an aggression score of "70" and an emotion rating of "Anger 60" and save it in the database.

[1339] Output: Comment data stored in the database and its rating (e.g., comment "This news is stupid!", aggression score "70", emotion "anger 60")

[1340] Step 5:

[1341] Users (visitors) can set the filter strength on their terminal to suit their preferences.

[1342] Input: User-input filter strength setting

[1343] Specific actions: User C sets the filter strength to "20" on their device, or the emotion analysis engine recognizes User C's "irritation" and recommends a filter strength of "50."

[1344] Output: User-defined filter strength (e.g., filter strength 20 or 50)

[1345] Step 6:

[1346] The server filters the comments based on the filter strength set by the user.

[1347] Input: User-defined filter strength and comment data

[1348] Specific operation: The server references user C's filter strength of "50" and retrieves the offensiveness score from the database. Comments with a score of "50" or higher are hidden.

[1349] Output: Filtered comment list (e.g. hide comments with a score of 70, show comments with a score of 30)

[1350] Step 7:

[1351] The server sends the filtered comment list to the user's terminal.

[1352] Input: filtered comment list

[1353] Specific operation: The server sends user C, who has a filter strength of "50," a list that hides comments with an offensive score of "70" and displays comments with a score of "30."

[1354] Output: A list of comments displayed on the user's device (e.g., comments with an offensiveness score of 70 are hidden, and comments with a score of 30 are displayed)

[1355] Step 8:

[1356] The server recalculates the offensiveness score when a new comment is posted and updates the filtering results.

[1357] Input: Newly posted comment data

[1358] Specific operation: When a new comment is posted, the server recalculates the score and emotional state using the natural language processing engine and sentiment analysis engine, and performs filtering.

[1359] Output: Updated filtered comment list (e.g., latest list with new comments)

[1360] Step 9:

[1361] If the user changes the filter strength, the server recalculates the displayed comment list and keeps it up to date.

[1362] Input: User changes to filter strength

[1363] What happens: When the user changes the filter strength, the server re-filters using the new criteria.

[1364] Output: Updated comment list (e.g., a list of comments based on the new filter strength)

[1365] (Application example 2)

[1366] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1367] In conventional content distribution services, some of the comments posted by users can be offensive, disrupting other users' comfortable viewing experiences. Furthermore, especially in environments using head-mounted displays, offensive comments can enter the user's field of vision, potentially causing psychological stress. It is necessary to provide an environment where users can enjoy content safely by automatically detecting and appropriately filtering these offensive comments.

[1368] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1369] In this invention, the server includes means for collecting posted comment data, means for analyzing the collected comment data using a natural language processing engine and calculating an aggression score, means for saving the calculated aggression score in a database, means for a user to set a filter strength, means for filtering comments based on the filter strength, means for transmitting a filtered comment list to the user's terminal, and means for filtering comments displayed on the head-mounted display. This allows users to filter aggressive comments according to their own emotional state and enjoy content with peace of mind.

[1370] "Posted comment data" refers to text information including ratings and opinions posted by users on the content distribution service.

[1371] A "natural language processing engine" is a software tool that analyzes text data to understand the meaning and sentiment of a sentence.

[1372] The "aggression score" is a numerical indicator of how aggressive the posted comment data is.

[1373] A "database" is a system for efficiently managing and storing collected data.

[1374] The "filter strength" is a reference value of the offensiveness score set by the user to determine whether or not to display.

[1375] "Filtering" is the process of filtering out harmful comments based on set criteria.

[1376] A "head-mounted display" is a display device that is worn on the user's head.

[1377] "Dynamic update" refers to a process of automatically changing the display content in response to the addition of new data or changes in filter strength.

[1378] "Real-time" means immediate, real-time processing or response.

[1379] This invention relates to a comment filtering system for content distribution services, and aims to improve the comment display environment when viewing using a head-mounted display (HMD). This system determines whether posted comments are offensive and filters them based on that, providing an environment in which users can enjoy content with peace of mind.

[1380] First, the server collects comment data posted by users. This data includes the comment text, the poster's identification information, and the posting date and time. Next, the collected comment data is analyzed using a natural language processing engine to calculate an offensiveness score. The natural language processing engine used is something like the VADER Sentiment Intensity Analyzer.

[1381] The calculated offensiveness score is stored in a database, which allows the offensiveness score of each comment to be managed. Users can set the filter strength via their own device or head-mounted display. This filter strength is the standard value that determines which comments are displayed. Based on the filter strength, comments with high offensive scores are filtered out, ensuring that only appropriate comments are displayed to the user.

[1382] The filtered comment list is sent to the user's device or head-mounted display. The system is dynamically updated, so the offensiveness score is recalculated every time a new comment is posted, and the filtering results are updated immediately. Furthermore, the filtering results are reflected in real time on the head-mounted display, allowing users to view the latest comments.

[1383] Specific examples of use

[1384] For example, if a user posts a comment such as "This video is not funny at all!", the server collects this comment and analyzes it using a natural language processing engine. The analysis results in an "offensiveness score of 70" which is then stored in the database. If the user sets the filter strength to "50," this comment will not be displayed, and only comments with an offensiveness score of 50 or less will be displayed.

[1385] Prompt Sentence Examples

[1386] The following prompt sentences are used by the system to collect user comments and calculate an offensiveness score:

[1387] Text format

[1388] comments_payload = {

[1389] "content": "This video is not funny at all!",

[1390] "user_id": "user123",

[1391] "post_date": "2023-10-01"

[1392] }

[1393] response = requests.post("http: / / localhost:5000 / add_comment", json=comments_payload)

[1394] print(response.json())

[1395] The prompts to set the filter strength and get the filtered comments are as follows:

[1396] Text format

[1397] filter_payload = {

[1398] "filter_strength": 50 Filter strength set by the user

[1399] }

[1400] response = requests.post("http: / / localhost:5000 / get_comments", json=filter_payload)

[1401] print(response.json())

[1402] This allows users to effectively filter out offensive comments and enjoy content comfortably, even when using content distribution services that use head-mounted displays.

[1403] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1404] Step 1:

[1405] The user puts on the head-mounted display, starts using the content distribution service, and posts comments.

[1406] Input: Comment text posted by the user, poster's identification information, posting date and time

[1407] Output: Comment data is sent to the server

[1408] Step 2:

[1409] The server receives the comment data sent by the user and temporarily stores it in a database.

[1410] Input: Comment data (text, identification information, posting date and time)

[1411] Output: Comment data is saved to the database.

[1412] Step 3:

[1413] The server passes the comment data collected from the database to a natural language processing engine and calculates the offensiveness score of the comment.

[1414] Input: Comment data stored in the database

[1415] Output: Aggression score is calculated

[1416] Step 4:

[1417] Based on the aggression score calculated by the natural language processing engine, the server links the aggression score to the comment data and stores it in a database.

[1418] Input: Aggression scores and comment data

[1419] Output: The linked data is saved in the database

[1420] Step 5:

[1421] The user sets the filter strength through the head-mounted display.

[1422] Input: Filter strength value set by the user

[1423] Output: Filter strength setting information is sent to the server

[1424] Step 6:

[1425] The server performs filtering based on the filter strength set by the user, and hides comments whose offensiveness score exceeds the filter strength.

[1426] Input: Comment data and user-defined filter strength

[1427] Output: Filtered comment list

[1428] Step 7:

[1429] The filtered comment list is displayed in real time on a head-mounted display.

[1430] Input: filtered comment list

[1431] Output: Comments displayed on the head-mounted display

[1432] Step 8:

[1433] When a new comment is posted, the server recalculates the offensiveness score and immediately updates the filtering results.

[1434] Input: New comment data

[1435] Output: The updated filtering results are reflected on the head-mounted display.

[1436] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1437] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1438] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1439] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1440] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1441] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1442] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1443] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1444] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1445] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1446] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1447] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1448] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1449] 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.

[1450] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1451] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1452] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1453] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1454] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1455] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1456] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1457] The following is further disclosed regarding the above embodiment.

[1458] (Claim 1)

[1459] A means for collecting posted comment data;

[1460] A means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score;

[1461] a means for storing the calculated aggression score in a database;

[1462] a means for the user to set the filter strength;

[1463] means for filtering comments based on a filter strength;

[1464] a means for transmitting the filtered comment list to a user's terminal;

[1465] A system including:

[1466] (Claim 2)

[1467] 10. The system of claim 1, further comprising: means for dynamically updating a displayed comment section according to a set filter strength.

[1468] (Claim 3)

[1469] 10. The system of claim 1, further comprising: means for recalculating the offensiveness score and updating the filtering results when a new comment is posted.

[1470] "Example 1"

[1471] (Claim 1)

[1472] A means for collecting posted comment data;

[1473] A means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score;

[1474] a means for storing the calculated aggression score in a database;

[1475] a means for the user to set the filter strength;

[1476] means for filtering comments based on a filter strength;

[1477] a means for transmitting the filtered comment list to a user's terminal;

[1478] A means for dynamically reflecting the filter strength setting on the server;

[1479] A system including:

[1480] (Claim 2)

[1481] 10. The system of claim 1, further comprising: means for dynamically updating a displayed comment section according to a set filter strength.

[1482] (Claim 3)

[1483] 10. The system of claim 1, further comprising: means for recalculating the offensiveness score and updating the filtering results when a new comment is posted.

[1484] "Application Example 1"

[1485] (Claim 1)

[1486] A means for collecting posted comment data;

[1487] A means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score;

[1488] a means for storing the calculated aggression score in a database;

[1489] a means for the user to set the filter strength;

[1490] means for filtering comments based on a filter strength;

[1491] means for transmitting the filtered comment list to the user's display device;

[1492] means for allowing a user to adjust the filter strength on their display device;

[1493] means for maintaining communication with the display device to dynamically perform the filtering process;

[1494] A system including:

[1495] (Claim 2)

[1496] 10. The system of claim 1, further comprising: means for dynamically updating a displayed comment section according to a set filter strength; and means for changing the filter strength in real time through a user interface.

[1497] (Claim 3)

[1498] The system of claim 1, further comprising: means for recalculating the offensiveness score and updating the filtering results when a new comment is posted; and means for transmitting a re-filtered comment list based on the user's settings to the display device.

[1499] "Example 2: Combining Emotion Engines"

[1500] (Claim 1)

[1501] A means for collecting posted comment data;

[1502] A means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score;

[1503] A means for analyzing the collected comment data using an emotion analysis engine and evaluating the emotional state of the poster;

[1504] a means for storing the calculated aggression score and the evaluation result of the emotional state in a database;

[1505] a means for the user to set the filter strength;

[1506] a means for recognizing a user's emotional state through an emotion analysis engine and setting a recommended filter strength;

[1507] means for filtering comments based on a filter strength;

[1508] means for transmitting the filtered comment list to a user's device;

[1509] A means to recalculate the offensiveness score and update the filtering results when a new comment is posted;

[1510] A system including:

[1511] (Claim 2)

[1512] 10. The system of claim 1, wherein the displayed comment section is dynamically updated according to a set filter strength.

[1513] (Claim 3)

[1514] 2. The system of claim 1, wherein the displayed comment list is recalculated and kept up to date when the user changes the filter strength.

[1515] "Application example 2 when combining emotion engines"

[1516] (Claim 1)

[1517] A means for collecting posted comment data;

[1518] A means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score;

[1519] a means for storing the calculated aggression score in a database;

[1520] a means for the user to set the filter strength;

[1521] means for filtering comments based on a filter strength;

[1522] a means for transmitting the filtered comment list to a user's terminal;

[1523] a means for filtering comments displayed on the head-mounted display;

[1524] A system including:

[1525] (Claim 2)

[1526] 10. The system of claim 1, further comprising: means for dynamically updating a displayed comment section according to a set filter strength.

[1527] (Claim 3)

[1528] The system of claim 1, further comprising: means for recalculating the offensiveness score and updating the filtering results when a new comment is posted; and means for reflecting the results on the screen of the head-mounted display in real time. [Explanation of symbols]

[1529] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for collecting posted comment data; A means for analyzing the collected comment data using a natural language processing engine and calculating an offensiveness score; a means for storing the calculated aggression score in a database; a means for the user to set the filter strength; means for filtering comments based on a filter strength; a means for transmitting the filtered comment list to a user's terminal; A system including:

2. The system of claim 1 , further comprising: means for dynamically updating the displayed comment section in response to a set filter strength.

3. The system of claim 1 , further comprising: means for recalculating the offensiveness score and updating the filtering results when a new comment is posted.

Images