system

Abstract

To provide a system in which collection, analysis and report generation of NPS comments are automated, and efficient operation of human resources and quick decision making are enabled.SOLUTION: A system comprising: means for collecting comments on a mobile network service provider from a website on the Internet; means for filtering and text preprocessing the collected comments; means for classifying the preprocessed comments into positive, negative, and neutral and calculating a net promoter score based on the preprocessed comments; means for calculating a distribution of positive and negative comments for each carrier and generating comparison information with other carriers; means for automatically generating a report including the generated information and a prediction result of a variation; and means for delivering and displaying the automatically generated report to a user 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] Previously, analyzing Net Promoter Score (NPS) comments about mobile network service providers was a time-consuming and labor-intensive process. This made it difficult to efficiently manage human resources and required quick decision-making. Furthermore, the comments were often similar, and there were many random responses, making it difficult to identify accurate areas for improvement. To improve this situation, it was necessary to automate the process from data collection to analysis and report generation to perform NPS analysis quickly and efficiently. [Means for solving the problem]

[0005] The present invention provides a system that includes: means for collecting comments about mobile network service providers from websites on the Internet; means for filtering and text preprocessing the collected comments; means for classifying the preprocessed comments into positive, negative, and neutral categories and calculating a Net Promoter Score (NPS). The system also includes means for calculating the distribution of positive and negative comments for each carrier and generating comparative information with other carriers; means for automatically generating a report including the generated information and predicted fluctuations; and means for delivering and displaying the automatically generated report to a user terminal. This system automates the collection, analysis, and report generation of NPS comments, enabling efficient use of human resources and rapid decision-making.

[0006] "Internet website" refers to any publicly accessible web page, including social media sites, online forums, review sites, etc.

[0007] "Mobile network service provider" is a general term for businesses that provide mobile communication services, such as MNOs (Mobile Network Operators) and MVNOs (Mobile Virtual Network Operators).

[0008] "Comments" means text messages or opinions posted by users on the Website.

[0009] "Collection methods" refers to the technical methods and tools used to obtain data from designated sources on the Internet.

[0010] "Filtering" refers to the process of extracting necessary information and removing unnecessary information.

[0011] "Text preprocessing" refers to the process of formatting text data, such as tokenization, normalization, and removal of stop words, which is carried out prior to natural language processing.

[0012] "Positive" is the sentiment category for comments that are judged to be favorable by AI analysis.

[0013] "Negative" is the sentiment category for comments that are determined to be negative by AI analysis.

[0014] "Neutral" is the emotional category of comments that AI analysis has determined to have little emotional bias.

[0015] The Net Promoter Score (NPS) is an index used to evaluate customer satisfaction and recommendation, and is rated on a scale of 0 to 10.

[0016] "Distribution of positive and negative comments" refers to the percentage of positive and negative comments received for a particular carrier.

[0017] "Comparative information with other carriers" refers to data used to compare NPS scores and comment ratings across multiple mobile network service providers.

[0018] "Predicted changes" refers to information that predicts future changes in NPS scores and comment ratings based on past data.

[0019] "Means for automated report generation" refers to technical methods and tools for automatically generating reports based on collected and analyzed data.

[0020] "User terminal" refers to a computer device, such as a PC, smartphone, tablet, etc., that a user uses to access the system.

[0021] A "dashboard" refers to an interface that allows users to visually view and manipulate data in real time. [Brief explanation of the drawings]

[0022] [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

[0023] 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.

[0024] First, the terms used in the following description will be explained.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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."

[0030] [First embodiment]

[0031] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0032] 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.

[0033] 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).

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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.

[0038] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0039] 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.

[0040] 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.

[0041] 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.

[0042] 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."

[0043] The present invention relates to a system that collects NPS comments about mobile network service providers from the Internet and generates analysis and result reports using AI. This system automates processes from data collection to sentiment analysis of comments, report generation, and distribution. Detailed descriptions of embodiments of the present invention are provided below.

[0044] Data collection from the internet

[0045] Processing performed by the server

[0046] The server periodically accesses designated websites (e.g., social networking sites, forums, review sites) on the Internet to collect comments about mobile network service providers. The server uses scraping tools and APIs to obtain text information and the following metadata for each comment:

[0047] timestamp

[0048] Source URL

[0049] Username (if available)

[0050] Applicable carrier name

[0051] As a specific example, the server uses a social media API to collect tweets containing hashtags such as "Career A Review" and store them in a database.

[0052] Data Preprocessing

[0053] Processing performed by the server

[0054] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization, normalization, and stop word removal is performed. This prepares the text data in a form suitable for analysis.

[0055] For example, a comment such as "This service is really great! Ad" is tokenized as "This service is really great" and unnecessary parts are removed.

[0056] Classification and scoring of NPS comments

[0057] Processing performed by the AI ​​module (on the server)

[0058] The pre-processed comments are sent to an AI sentiment analysis module. The server uses AI to analyze the sentiment of each comment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content of each comment. This also generates an overall NPS score for each carrier.

[0059] For example, the server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and sets the NPS score to 8.

[0060] Report Generation

[0061] Processing performed by the server

[0062] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each carrier. It also generates comparative information with other carriers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0063] As a concrete example, the server generates a report containing the information, "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan."

[0064] Report distribution and visualization

[0065] Processing performed by the user terminal

[0066] The generated reports are displayed in real time on a dashboard on the user's device. Users can log in to the dashboard to view the latest reports. Furthermore, they can use the filter and search functions to view detailed information for specific periods or by carrier.

[0067] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as reference for their next marketing initiative.

[0068] The above is a detailed description of the embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, enabling efficient use of human resources and rapid decision-making.

[0069] The processing flow will be explained below.

[0070] Step 1:

[0071] Processing performed by the server

[0072] The server periodically accesses designated websites on the Internet to collect comments about mobile network service providers. Using scraping tools and public APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, username (if available), and applicable carrier name.

[0073] Step 2:

[0074] Processing performed by the server

[0075] The collected comments are filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing is performed, including tokenization (splitting sentences into words), normalization (converting all to lowercase, removing special characters, etc.), and removal of stop words. This prepares the text data in a form suitable for analysis.

[0076] Step 3:

[0077] Processing performed by the server

[0078] The preprocessed comments are sent to the AI ​​sentiment analysis module, which uses an AI model to analyze the sentiment of each comment and classify it as positive, negative, or neutral. It then calculates an NPS score based on the content of each comment and stores it in a database.

[0079] Step 4:

[0080] Processing performed by the server

[0081] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, and generates comparative information with other carriers, including calculating each carrier's overall NPS score and analyzing trends over time.

[0082] Step 5:

[0083] Processing performed by the server

[0084] Based on the aggregated data and analysis results, the server automatically generates a report that includes the overall NPS score, the distribution of positive and negative comments, comparisons with other carriers, and predictions of future changes.

[0085] Step 6:

[0086] Processing performed by the user terminal

[0087] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to the dashboard to check the latest reports. They can also use the filter and search functions to view detailed information for specific periods or by carrier.

[0088] Step 7:

[0089] User actions

[0090] Planning and marketing staff conduct analysis based on the reports generated on the dashboard. This allows them to quickly evaluate the effectiveness of their initiatives and quickly consider and implement any necessary improvements. For example, they can evaluate the effectiveness of a new pricing plan based on information such as "Carrier A's NPS score has improved by 10 points since last month, and the number of positive comments has increased."

[0091] Example 1

[0092] 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."

[0093] It is inefficient to manually collect appropriate feedback in real time from a huge amount of user comments about mobile network service providers, perform sentiment analysis, and automatically generate comprehensive reports including competitive comparisons. There is a need for a system that solves this problem and automates and streamlines the process from data collection to report creation.

[0094] 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.

[0095] In this invention, the server includes a means for collecting comments about mobile network service providers from data sources on the Internet, a means for filtering the collected comments using regular expressions to remove unnecessary data such as spam and advertising links, and a text preprocessing means for tokenizing, normalizing, and removing stop words from the filtered comments. This enables useful information to be extracted from large amounts of data, enabling rapid and accurate sentiment analysis and NPS score evaluation. The server also includes a means for delivering automatically generated reports to user terminals and displaying them in real time, allowing users to make quick decisions based on the latest information.

[0096] "Data Source" refers to any place where information can be collected, such as any website, social media site, forum, review site, or other website accessible on the Internet.

[0097] "Comment" means text data such as feedback, opinions, reviews, etc. posted by a user to a mobile network service provider.

[0098] "Filtering" refers to the process of removing unnecessary information such as spam and advertising links from collected comment data and formatting it into data suitable for analysis.

[0099] "Tokenization" refers to the process of dividing text into the smallest meaningful units, such as words or phrases.

[0100] "Normalization" refers to the process of converting text data into a unified format to improve the accuracy of analysis.

[0101] "Stop words" refer to common words (e.g., wa, ga, no, etc.) that are of little importance and have no meaning in text analysis.

[0102] "Sentiment analysis" refers to the process of analyzing text data and identifying its emotional tendencies (positive, negative, neutral).

[0103] The Net Promoter Score (NPS) is an index that evaluates customer satisfaction and the degree of recommendation, and is scored on a scale of 0 to 10.

[0104] "Positive comments" refer to comments that contain positive or favorable feedback about a service or product.

[0105] "Negative Comments" refers to comments that contain negative or unsatisfactory feedback about a service or product.

[0106] "Report" means a document generated from collected and analyzed data that includes results, trends, comparative information, etc.

[0107] "Dashboard" refers to a web application or software that includes an interface that allows users to view, search, and analyze information in real time.

[0108] This invention relates to a system that collects comments about mobile network service providers from the Internet and uses AI to analyze those comments and generate a report of the results. This system automates the process from data collection to report generation and distribution through the following series of processes.

[0109] Data collection

[0110] Processing performed by the server

[0111] The server periodically accesses designated websites (e.g., social networking sites, forums, review sites, etc.) on the Internet. Specifically, the server uses scraping tools and APIs (e.g., Twitter API) to collect targeted comments. The collected information includes text, timestamp, source URL, username (if available), and corresponding carrier name.

[0112] As a specific example, the server uses the Twitter API to collect tweets containing the hashtag "Career A Review" and store them in a database.

[0113] Data Preprocessing

[0114] Processing performed by the server

[0115] The collected comment data is first filtered using regular expressions to remove unwanted data such as spam and advertising links, then the comment text is tokenized, normalized, and stop words are removed.

[0116] For example, a comment like "This service is really great! Ad" would be tokenized as "This service is really great" and the spam part would be removed.

[0117] Classification and scoring of NPS comments

[0118] Processing performed by the server

[0119] The pre-processed comments are sent to an AI sentiment analysis module, which analyzes the sentiment of each comment and categorizes it as positive, negative, or neutral. It then calculates an NPS score based on the content of each comment and generates an overall NPS score for each carrier.

[0120] As a concrete example, the server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and sets the NPS score to 8.

[0121] Report Generation

[0122] Processing performed by the server

[0123] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, generates comparison information with other carriers, and automatically generates a comprehensive report with forecast results for important fluctuations.

[0124] As a specific example, the server compiles information such as "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan" and includes it in a report.

[0125] Report distribution and visualization

[0126] Processing performed by the user terminal

[0127] The generated reports are displayed in real time on a dashboard on the user's device, and users can log in to view the latest reports and use filters and search functions to view detailed information for specific periods or by carrier.

[0128] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as a reference for their next marketing initiative.

[0129] Examples of prompt statements

[0130] Examples of input prompts for generative AI models include:

[0131] "Please explain the process flow of a system that collects NPS comments about mobile network service providers, performs sentiment analysis and scoring using AI, and compiles the results into a report, including the specific processing steps."

[0132] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0133] Step 1: Data collection

[0134] The server periodically accesses designated data sources (e.g., social media sites, forums, review sites) on the Internet. The server uses scraping tools or social media APIs (e.g., Twitter API) to collect comments about mobile network service providers. As input, the server receives target comments and hashtags (e.g., "Carrier A Review"). As output, the server stores data in a database, including text, timestamp, source URL, username (if available), and the name of the corresponding carrier.

[0135] Specific operation: The server uses the Twitter API to collect tweets containing the hashtag "Career A Review" and stores that data in a database.

[0136] Step 2: Filtering the data

[0137] The server uses regular expressions to filter the collected comment data to remove unwanted data such as spam and advertising links. It receives the collected raw data as input and generates clean filtered data as output.

[0138] Specific operation: The comment "This service is really great! Ad" is converted into clean data using regular expressions to remove unnecessary parts such as "This service is really great."

[0139] Step 3: Tokenize and normalize the data

[0140] The filtered comments are then tokenized and normalized. The server receives the filtered text data as input, tokenizes it (splits the text into words and phrases), normalizes it (converts the text into a uniform format), and removes stop words. As output, it produces tokenized data suitable for analysis.

[0141] Specific behavior: Split the text "This service is really great!" into "This service is really great", removing spelling variations and extra words.

[0142] Step 4: Sentiment analysis and NPS score calculation

[0143] The preprocessed comments are sent to the AI ​​sentiment analysis module. The server receives the tokenized text data as input and performs sentiment analysis using an AI model. It classifies each comment as positive, negative, or neutral and calculates an NPS score based on its content. As output, it generates data containing the sentiment classification result and the NPS score.

[0144] Specific behavior: The server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and calculates an NPS score of 8.

[0145] Step 5: Generate a report

[0146] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, and automatically generates a comprehensive report including comparisons with other carriers and predictions of significant fluctuations. It receives the results of sentiment analysis and NPS scores as input and generates a comprehensive report as output.

[0147] Specific operation: Based on the information that "Carrier A's NPS score this month is 72, an improvement of 10 points from last month," the server compiles a report stating that 50% of users have given positive feedback about the new pricing plan.

[0148] Step 6: Report distribution and visualization

[0149] The user terminal displays the generated reports on a dashboard in real time. Users can log in and view the latest reports. They can also use filters and search functions to check detailed information for specific periods or by carrier. The system receives reports provided by the server as input and displays them on the dashboard as output.

[0150] Specific actions: The planner accesses the dashboard and checks the information that "Career A's NPS score this month is 76, and positive comments have increased by 50%," and uses this information as a reference for the next marketing initiative.

[0151] (Application example 1)

[0152] 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."

[0153] Content distribution services are required to collect and analyze viewer feedback in real time, and to efficiently recommend personalized content and propose improvements based on that feedback. Conventional methods have had the problem of being unable to provide appropriate content due to the difficulty of quickly reflecting viewer reactions.

[0154] 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.

[0155] In this invention, the server includes means for collecting comments about a mobile network service provider from websites on the Internet, means for filtering and text preprocessing the collected comments, means for classifying the preprocessed comments into positive, negative, and neutral categories and calculating a Net Promoter Score, means for calculating the distribution of positive and negative comments for each provider and generating comparative information with other providers, means for automatically generating a report including the generated information and predicted fluctuations, means for delivering and displaying the automatically generated report to a user terminal, and means for aggregating viewer responses in real time on the user terminal and recommending content and proposing improvements based on viewing trends, thereby enabling fast and efficient collection and analysis of viewer feedback and the provision of personalized content.

[0156] The "Internet" is a communications infrastructure that connects computer networks around the world and enables the exchange of information.

[0157] A "website" is a collection of information made up of multiple web pages published on the Internet.

[0158] A "mobile network service provider" is a company or organization that provides communications services to mobile phones and other mobile communication devices.

[0159] A "comment" is text information in which a user expresses an opinion or comment on a particular topic.

[0160] "Filtering" is the process of removing unnecessary information from collected data.

[0161] "Text preprocessing" refers to a series of processes used to format text data into a form suitable for analysis.

[0162] "Positive" refers to comments that express positive or favorable sentiment.

[0163] "Negative" refers to comments that express negative or unfavorable sentiment.

[0164] "Neutral" refers to comments that express a neutral sentiment, neither positive nor negative.

[0165] "Net Promoter Score" is a quantified indicator of the likelihood that a customer will recommend a service to others.

[0166] "Provider" refers to a company or organization that provides a particular service or content.

[0167] "Distribution" refers to the frequency or proportion of occurrence in each category of data.

[0168] "Comparative information" is information that shows the differences and similarities between two or more data sets.

[0169] A "report" is a document that summarizes analytical results and statistical data.

[0170] A "user terminal" is a device that a user uses to access the Internet or a system.

[0171] "Real-time" refers to the property that data and events are processed as they occur.

[0172] A "viewer" is a user who views video and audio content.

[0173] "Reactions" refer to the emotions and evaluations that viewers express toward content.

[0174] "Aggregation" is the process of statistically summarizing data.

[0175] "Viewing habits" refers to a viewer's general behavior and preference patterns regarding content.

[0176] "Content" refers to the media that viewers consume, such as television programs, movies, music, articles, etc.

[0177] A "recommendation" is the presentation of particular content to a viewer.

[0178] "Suggestions for improvement" refers to suggestions for improving the quality of content based on viewer reactions and feedback.

[0179] This invention provides a system that collects and analyzes viewer feedback in real time, and recommends personalized content and suggests improvements. Specific implementations of this system are described below.

[0180] Data collection method

[0181] The server collects viewers' comments from specific websites on the Internet (social networking sites, review sites, forums, etc.) using scraping tools and APIs to obtain text information and the following metadata for each comment:

[0182] timestamp

[0183] Source URL

[0184] Username (if available)

[0185] For example, the server uses a social media API to collect posts containing hashtags such as "Series X review" and store them in a database.

[0186] Data Preprocessing

[0187] The server filters the collected comment data and performs the following actions:

[0188] Remove unwanted data such as spam and advertising links (using regular expressions)

[0189] Text preprocessing, including tokenization, normalization, and stop word removal

[0190] This allows the comment to be formatted in a way that is suitable for sentiment analysis. For example, a comment like "This movie was great! Ad" is tokenized as "This movie was great" and unnecessary parts are removed.

[0191] Classification and scoring of NPS comments

[0192] The preprocessed comments are sent to an AI sentiment analysis module. The server uses AI (e.g., NLTK's VADER Sentiment Analyzer) to analyze the sentiment of each comment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content of each comment.

[0193] For example, a comment such as "This movie was amazing!" is considered positive and is assigned an NPS score of 9.

[0194] Report Generation and Delivery

[0195] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each provider, as well as comparisons with other providers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0196] For example, a report might be generated that includes information such as, "Provider A's NPS score this month is 50, up 15 points from last month. 30% of viewers have given positive feedback about new content."

[0197] The generated reports are displayed in real time on a dashboard on the user's device. Users can log in to this dashboard to view the latest reports. Furthermore, they can use the filter and search functions to view detailed information for specific periods or by provider.

[0198] Content recommendations and improvement suggestions

[0199] Viewer responses are collected in real time on the user's device, and content recommendations and suggestions for improvements are made based on viewing trends. For example, it is possible to recommend the next content to watch based on viewer feedback data. In addition, if there is a lot of negative feedback, it is possible to analyze the cause and suggest improvements.

[0200] The specific hardware and software used

[0201] Hardware: Servers, user devices (smartphones, tablets, PCs, etc.)

[0202] Software: Python, BeautifulSoup, requests library, NLTK, VADER Sentiment Analyzer, Matplotlib

[0203] Examples of prompt statements

[0204] Here is an example prompt:

[0205] Analyze the sentiment of this comment: "This movie was amazing! Great story and acting. Advertisement"

[0206] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0207] Step 1:

[0208] The server collects viewer comments from specific websites on the Internet. Using APIs and scraping tools, the server obtains metadata such as comment text information, timestamps, source URLs, and user names. The input is the URL of a specific website, and the output is a list of collected comments and their metadata.

[0209] Step 2:

[0210] The server filters the collected comment data and removes inappropriate data (such as spam or advertising links). It uses regular expressions to detect and remove unnecessary data. The input is the comment data collected in the previous step, and the output is the clean comment data after filtering.

[0211] Step 3:

[0212] The server performs text preprocessing such as tokenization, normalization, and stop word removal. Tokenization splits comments into words, normalization converts uppercase to lowercase, and stop word removal removes common words such as "の" and "は." The input is the filtered comment data, and the output is the preprocessed text data.

[0213] Step 4:

[0214] The server sends the preprocessed comment data to the AI ​​sentiment analysis module for sentiment analysis. It uses NLTK's VADER Sentiment Analyzer to classify each comment's sentiment as positive, negative, or neutral. It then calculates the Net Promoter Score (NPS score) for each comment. The input is the preprocessed text data, and the output is the sentiment classification and NPS score for each comment.

[0215] Step 5:

[0216] The server calculates the distribution of positive and negative comments for each provider and generates comparative information with other providers. It also automatically generates a comprehensive report including predictions of significant fluctuations. The input is the comment data after sentiment analysis and the NPS score, and the output is a report including information for each provider and comparisons.

[0217] Step 6:

[0218] The server delivers the automatically generated reports to the user terminal and displays them. The user terminal displays the reports in real time on a dashboard, allowing the user to search for detailed information for a specific period or by provider. The input is the generated report, and the output is the dashboard display on the user terminal.

[0219] Step 7:

[0220] The user terminal aggregates viewer responses in real time and makes content recommendations and suggestions for improvements based on viewing trends. The input is viewer feedback data, and the output is content recommendations and suggestions for improvements to the viewer.

[0221] 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.

[0222] The present invention relates to a system that collects NPS comments about mobile network service providers from the Internet and generates analysis and result reports using AI and an emotion engine. This system automates processes from data collection to emotion analysis of comments, report generation, and distribution. Furthermore, by combining this system with an emotion engine that can analyze user emotions in real time, highly accurate NPS scoring is achieved. Detailed descriptions of embodiments of the present invention are provided below.

[0223] Data collection from the internet

[0224] Processing performed by the server

[0225] The server periodically accesses designated websites (social media sites, forums, review sites, etc.) on the Internet to collect comments about mobile network service providers. Using scraping tools and APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, username (if available), and applicable carrier name.

[0226] As a specific example, the server uses a social media API to collect tweets containing hashtags such as "Career A Review" and store them in a database.

[0227] Data Preprocessing

[0228] Processing performed by the server

[0229] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization and normalization is performed. This prepares the text data in a form suitable for analysis.

[0230] For example, a comment such as "This service is really great! Ad" is tokenized as "This service is really great" and unnecessary parts are removed.

[0231] Classification and scoring of NPS comments

[0232] Processing performed by the AI ​​module and emotion engine (on the server)

[0233] The preprocessed comments are sent to an AI sentiment analysis module and sentiment engine. The server uses the sentiment engine to analyze user sentiment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content and sentiment data of each comment, which also generates an overall NPS score for each carrier.

[0234] For example, the server might send a comment like "This service is really great" to the sentiment engine, which would then consider it positive and set the NPS score to 8.

[0235] Report Generation

[0236] Processing performed by the server

[0237] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each carrier. It also generates comparative information with other carriers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0238] As a concrete example, the server generates a report containing the information, "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan."

[0239] Report distribution and visualization

[0240] Processing performed by the user terminal

[0241] The user device displays reports delivered from the server on a dashboard in real time. Users can log in to this dashboard to check the latest reports. They can also use the filter and search functions to check detailed information for specific periods or by carrier.

[0242] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as reference for their next marketing initiative.

[0243] The above is a detailed description of an embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, enabling efficient use of human resources and rapid decision-making. Furthermore, the use of an emotion engine makes it possible to more accurately grasp user emotions and achieve highly accurate NPS scoring.

[0244] The processing flow will be explained below.

[0245] Step 1:

[0246] Processing performed by the server

[0247] The server periodically accesses designated websites on the Internet and uses scraping tools or public APIs to collect comments about mobile network service providers, along with metadata for each comment (timestamp, source URL, user name, and corresponding carrier name), which are then stored in a database.

[0248] Step 2:

[0249] Processing performed by the server

[0250] The collected comment data is filtered, using regular expressions to remove unwanted data such as spam and advertising links, and the comments are tokenized and subjected to text preprocessing, which includes splitting sentences into words, normalization, and stop word removal.

[0251] Step 3:

[0252] Processing performed by the server

[0253] The preprocessed comments are sent to the sentiment engine and AI sentiment analysis module. The sentiment engine analyzes the user sentiment of each comment and classifies it as positive, negative, or neutral. The AI ​​sentiment analysis module calculates an NPS score for each comment based on the sentiment data, which then derives an overall NPS score for each carrier.

[0254] Step 4:

[0255] Processing performed by the server

[0256] The server periodically compiles NPS data, calculates the distribution of positive and negative comments for each carrier, and generates comparative information with other carriers, including calculating each carrier's overall NPS score and analyzing trends over time.

[0257] Step 5:

[0258] Processing performed by the server

[0259] Based on the aggregated data and analysis results, the server automatically generates a report that includes the overall NPS score, the distribution of positive and negative comments, comparison information with other carriers, and a forecast of future fluctuations, allowing users to grasp the necessary information at a glance.

[0260] Step 6:

[0261] Processing performed by the user terminal

[0262] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to the dashboard to check the latest reports. They can also use the filter and search functions to view detailed information for specific periods or by carrier.

[0263] Step 7:

[0264] User actions

[0265] Planning and marketing staff conduct analysis based on the reports generated on the dashboard. This allows them to quickly evaluate the effectiveness of their initiatives and quickly consider and implement any necessary improvements. For example, they can evaluate the effectiveness of a new pricing plan based on information such as "Carrier A's NPS score has improved by 10 points since last month, and the number of positive comments has increased."

[0266] Example 2

[0267] 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."

[0268] The vast amount of comment data available online varies in content and format, making it difficult to manually collect and analyze, and therefore difficult to efficiently calculate NPS scores. Furthermore, accurate understanding of user sentiment and reflecting it in marketing and service improvements requires advanced analysis. There is a need for a system that can solve these problems, achieve highly accurate NPS scoring, and generate reports quickly.

[0269] 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.

[0270] In this invention, the server includes means for collecting evaluation comments about a communication service provider from Internet sources, including designated websites and social networks, means for filtering and text preprocessing the collected comments, means for classifying the preprocessed comments into positive, negative, and neutral categories and calculating a Net Promoter Score, means for calculating the distribution of positive and negative comments for each communication service provider and generating comparative information with other communication service providers, means for automatically generating a report including the generated information and predicted fluctuations, and means for delivering and displaying the automatically generated report to a user device. This allows for efficient calculation of NPS scores while accurately grasping user sentiment, and enables rapid report generation and delivery.

[0271] "Designated websites and social networks" are pre-selected internet websites, social media platforms, and other online platforms used to collect information.

[0272] A "communications service provider" is a company or organization that provides communications services such as mobile communications and internet access.

[0273] "Evaluation comments" are written evaluations such as feedback, reviews, opinions, etc., given by users to communication service providers.

[0274] "Filtering" is the process of removing unnecessary data to extract only useful data.

[0275] "Text preprocessing" refers to the process of preparing data for later analysis, such as tokenizing and normalizing comment data.

[0276] "Classifying into positive, negative, and neutral" is the process of classifying emotions and evaluations into three categories based on the content of comments.

[0277] "Net Promoter Score" is an index that scores the likelihood that users will recommend a service to others.

[0278] "Distribution of positive and negative comments" is data that shows the percentage of positive and negative comments for a particular period or career.

[0279] "Comparative information" is information for comparing the ratings and performance of different communication service providers.

[0280] A "report including predicted change results" is a report that predicts NPS scores and evaluation data that change over time and summarizes the results.

[0281] A "user device" is a device such as a computer, smartphone, or tablet that a user uses to receive and display reports.

[0282] The present invention relates to a system that collects evaluation comments about telecommunications service providers from the Internet and generates analysis and result reports using AI and an emotion engine. This system automates processes from data collection to emotion analysis of comments, report generation, and distribution. Furthermore, by combining it with an emotion engine that can analyze user emotions in real time, highly accurate Net Promoter Scoring is achieved. The following describes in detail an embodiment of the present invention.

[0283] Data collection from the internet

[0284] The server accesses designated websites and social networks (e.g., social networking sites, forums, review sites) to periodically collect comments about telecommunications service providers. Using scraping tools (e.g., the BeautifulSoup library) or APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, user name, and the name of the applicable carrier.

[0285] Data Preprocessing

[0286] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization and normalization is performed. This prepares the text data in a form suitable for analysis.

[0287] Classification and scoring of NPS comments

[0288] The preprocessed comments are sent to an AI sentiment analysis module and sentiment engine. The server uses the sentiment engine to analyze user sentiment and categorize it as positive, negative, or neutral. It then calculates an NPS score based on the content and sentiment data of each comment, which then generates an overall NPS score for each telecommunications service provider.

[0289] Report Generation

[0290] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each telecommunications service provider. It also generates comparative information with other telecommunications service providers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0291] Report distribution and visualization

[0292] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to this dashboard and check the latest reports. They can also use the filter and search functions to check detailed information for specific periods or by communication service provider.

[0293] As a concrete example of a prompt sentence, the following input is given to the generative AI model:

[0294] "Analyze the sentiment of NPS comments using the data below and generate a report.

[0295] Example data:

[0296] Comment: "This service is absolutely amazing!"

[0297] Timestamp: "2023-10-13 12:00:00"

[0298] Source URL: "https: / / example.com / review / 12345"

[0299] Username: "user123"

[0300] Applicable carrier name: "Carrier A"

[0301] This concludes the details of the embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, making it possible to accurately grasp user sentiment and efficiently calculate NPS scores.

[0302] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0303] Processing Steps:

[0304] Step 1: Collecting data from the internet

[0305] Processing performed by the server

[0306] Input: The URL or API endpoint of a specified website or social network (social network, forum, review site, etc.).

[0307] Specific operation: Uses Python's BeautifulSoup library to scrape and retrieve comments containing specified keywords (e.g., "Telecommunications Service Provider A Review"). If necessary, uses the social media API to collect data in real time.

[0308] Output: Collected comment data (comment text, timestamp, source URL, username, carrier name).

[0309] Step 2: Preprocessing the data

[0310] Processing performed by the server

[0311] Input: Collected comment data.

[0312] What it does: It uses regular expressions to filter out unwanted data like spam and advertising links, then uses the NLTK library to tokenize and normalize the comment text.

[0313] Output: Filtered and preprocessed comment data.

[0314] Step 3: Categorize and score NPS comments

[0315] Processing performed by the server

[0316] Input: Preprocessed comment data.

[0317] What it does: It uses HuggingFace's Transformers library to run a sentiment analysis model to classify comments as positive, negative, or neutral, and then calculates an NPS score based on the classification (8 for positive, 2 for negative, 5 for neutral).

[0318] Output: Comment data with sentiment classification results and NPS scores.

[0319] Step 4: Generate a report

[0320] Processing performed by the server

[0321] Input: Comment data with sentiment classification results and NPS scores.

[0322] What it does: Aggregates the distribution of positive, negative, and neutral comments for each telecommunications service provider, calculates an overall NPS score, performs comparative analysis with other telecommunications service providers, and automatically generates a comprehensive report with predicted results of important fluctuations.

[0323] Output: An automatically generated report containing the aggregated and analyzed results.

[0324] Step 5: Report distribution and visualization

[0325] Processing performed by the user terminal

[0326] Input: Report delivered from the server.

[0327] Specific operation: The report is reflected in a dashboard that displays the report in real time on the user's device. On the dashboard, users can search and filter for detailed information for specific periods or by communication service provider.

[0328] Output: The latest report and details displayed to the user.

[0329] This flow of processing steps automates the entire system, allowing users to efficiently check the analysis results of NPS comments.

[0330] (Application example 2)

[0331] 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."

[0332] Currently, systems for collecting and analyzing feedback on the efficiency and performance of factory robots are inadequate, preventing effective use of this information. This makes it difficult to quickly and accurately develop specific measures to improve factory operational efficiency and robot performance. Furthermore, when the process from collecting feedback to analyzing, generating, and distributing reports is carried out manually, it results in wasted human resources and makes it difficult to reflect information in a timely manner.

[0333] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for collecting comments about objects from websites on the Internet; means for filtering and text preprocessing the collected comments; means for classifying the comments into positive, negative, and neutral based on the preprocessed comments and calculating a performance score; means for calculating the distribution of positive and negative comments for each object and generating comparative information with other objects; means for automatically generating a report including the generated information and predicted fluctuations; means for delivering and displaying the automatically generated report to a user terminal; means for a user to view the report in real time via a dashboard and check detailed information; and means for using AI and an emotion engine to collect and analyze feedback comments about the efficiency and performance of objects. This makes it possible to quickly and automatically collect and analyze feedback about the efficiency and performance of factory robots and generate and distribute highly accurate performance reports.

[0334] 1. "Internet website" refers to an information site on the Internet that is accessible to users.

[0335] 2. "Object" refers to the factory robot or other equipment that is the subject of evaluation in this invention.

[0336] 3. "Comments" refers to the written opinions and feedback of users regarding the subject matter.

[0337] 4. "Collection methods" refers to the technologies and methods used to extract comments from websites on the Internet.

[0338] 5. "Filtering" refers to the process of removing unnecessary information and noise from acquired comment data.

[0339] 6. "Text preprocessing" refers to procedures that transform comment text into a form suitable for analysis, such as tokenization and normalization.

[0340] 7. "Positive" refers to comments that have a favorable opinion or reputation about the subject.

[0341] 8. "Negative" refers to comments that have a negative opinion or reputation about the subject.

[0342] 9. "Neutral" refers to comments that have a neutral opinion or do not evoke any particular strong emotion toward the subject.

[0343] 10. "Performance Score" refers to a numerical rating that indicates the efficiency or performance of an object.

[0344] 11. "Means for calculating distribution" refers to the techniques and methods for calculating the ratio of positive and negative comments from the collected comment data.

[0345] 12. "Comparative information" refers to data and statistics that allow the evaluation of an object to be compared with other objects.

[0346] 13. "Means for automatically generating reports" refers to technologies and methods for periodically generating reports based on analysis results.

[0347] 14. "User terminal" refers to a device used to receive and display reports, such as a personal computer or smartphone.

[0348] 15. "Means of delivery" refers to the technology or method for sending automatically generated reports to user terminals.

[0349] 16. "Means of display" refers to the technology or method for displaying the delivered report in an easy-to-read format on the user's device.

[0350] 17. "Dashboard" refers to an interface for aggregating and visually displaying various data and reports.

[0351] 18. "Real-time viewing means" refers to technologies and methods that allow users to instantly view the latest information and reports.

[0352] 19. "Means for checking detailed information" refers to technologies and methods that allow users to search and view detailed data for specific periods or subjects.

[0353] 20. "Feedback comments" refers to user evaluations and opinions regarding an object.

[0354] 21. “AI” refers to artificial intelligence technologies, including machine learning models and data analysis algorithms.

[0355] 22. “Sentiment Engine” refers to technology or software used to analyze comment data to identify user sentiment.

[0356] This invention relates to a system that collects feedback on the efficiency and performance of factory robots from websites on the internet, analyzes the feedback, and generates and distributes reports. The system uses AI and an emotion engine to analyze the sentiment of comments and calculate a performance score.

[0357] Program generation and processing steps

[0358] Server-side processing

[0359] 1. Data Collection:

[0360] The server uses web scraping tools such as Scrapy to collect comments about factory robots from social media, review sites, and forums on the Internet, such as comments containing the hashtag "factory robot performance."

[0361] 2. Data preprocessing:

[0362] Collected comments are filtered, tokenized, and normalized using Python, which includes removing spam and ads, and eliminating unnecessary data.

[0363] 3. Sentiment Analysis and Scoring:

[0364] The preprocessed comments are fed into a sentiment analysis model powered by TENSORFLOW® / Keras, where they are classified as positive, negative, or neutral and a performance score is calculated.

[0365] 4. Report Generation:

[0366] Based on the aggregated data, a report is automatically generated using pandas and matplotlib, including the distribution of positive and negative comments, as well as a comparison with other robots.

[0367] 5. Report Delivery:

[0368] The reports are delivered to the user's device and displayed on a dashboard in real time.

[0369] Processing on the user terminal side

[0370] 1. Dashboard view:

[0371] The delivered reports are displayed in real time on the user's device using JavaScript (registered trademark) (especially React.js), and users can view detailed information for specific periods or for each factory robot through this dashboard.

[0372] Examples of concrete examples and prompts

[0373] Examples:

[0374] When a user accesses the dashboard, the "Performance Report for Factory Robot A this Month" will show "Performance score increased by 45% and positive comments increased by 60%." Users can also use the filter function to specify a specific period and view performance data for that period in more detail.

[0375] Example prompt sentence:

[0376] The prompt sentence is an example of input to the generative AI model, and looks like this:

[0377] Generate a monthly report for a factory robot performance feedback system. Separate positive, negative, and neutral comments, calculate a score for each, and output the report.

[0378] This completes the description of an embodiment of the present invention, which enables the rapid and automatic collection and analysis of feedback on the efficiency and performance of factory robots, and the generation and distribution of highly accurate performance reports.

[0379] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0380] Step 1: Data collection

[0381] The server uses a web scraping tool such as Scrapy to collect comments about factory robots from social media, review sites, and forums on the Internet. For example, it collects comments containing the hashtag "factory robot performance." The input is the comment data obtained from social media and review sites, and the output is a dataset containing these comments.

[0382] Step 2: Filtering the data

[0383] The server filters the collected comment data. It uses regular expressions in Python to remove unnecessary data, such as spam and advertisements. Specifically, it detects comments containing "advertisements" or "spam links" and does not send them to the preprocessing step. The input is the initial comment data, and the output is the cleaned comment data.

[0384] Step 3: Preprocessing the text

[0385] The server preprocesses the filtered comments, performing tokenization and normalization to make the text suitable for analysis. Specifically, it splits sentences into tokens and removes unnecessary symbols and special characters. For example, it converts a comment such as "This factory robot is great" into "This factory robot is great." The input is filtered comment data, and the output is preprocessed text data.

[0386] Step 4: Sentiment analysis

[0387] The preprocessed comments are sent to a sentiment analysis model (generative AI model) using TensorFlow / Keras. The server performs data calculations to classify comments into positive, negative, or neutral. Specifically, the model predicts the sentiment of each comment and classifies it as a category. For example, a comment saying "great" is classified as positive. The input is the preprocessed text data, and the output is the sentiment category (positive, negative, neutral) and its confidence score.

[0388] Step 5: Calculating the Performance Score

[0389] The server calculates a performance score based on the results of the sentiment analysis. It aggregates the percentage of positive and negative comments and converts them into a score. Specifically, it calculates the performance score from the ratio of positive comments and generates an overall score from, for example, "60% positive, 40% negative." The input is sentiment category data, and the output is a performance score.

[0390] Step 6: Generate reports

[0391] The server automatically generates a report using pandas and matplotlib based on the aggregated data. This report includes the distribution of positive and negative comments and comparison information with other robots. Specifically, it creates a graph of the progress of scores and distribution, and compiles it into a report document. The input is the aggregated data, and the output is a report document.

[0392] Step 7: Report Delivery

[0393] The server delivers the generated report to the user's device. The report is automatically sent via email or a cloud platform. Specifically, the report file is delivered to the user using an email sending API or cloud storage service. The input is the report document, and the output is a notification that the report has been sent.

[0394] Step 8: View the dashboard

[0395] On the user's device, JavaScript (particularly React.js) is used to display the delivered reports in real time on a dashboard. Through this dashboard, users can check detailed information for a specific period or for each factory robot. Specifically, specific data can be displayed using date filters and search functions. The input is the report data, and the output is a visualized report display.

[0396] These are the specific processing steps of the system program that realizes this application example, which allows feedback on the efficiency and performance of factory robots to be collected and analyzed sequentially, with the results immediately available.

[0397] 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.

[0398] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (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.

[0399] 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.

[0400] [Second embodiment]

[0401] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0402] 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.

[0403] 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).

[0404] 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.

[0405] 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.

[0406] 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).

[0407] 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.

[0408] 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.

[0409] 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.

[0410] 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.

[0411] 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.

[0412] 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."

[0413] The present invention relates to a system that collects NPS comments about mobile network service providers from the Internet and generates analysis and result reports using AI. This system automates processes from data collection to sentiment analysis of comments, report generation, and distribution. Detailed descriptions of embodiments of the present invention are provided below.

[0414] Data collection from the internet

[0415] Processing performed by the server

[0416] The server periodically accesses designated websites (e.g., social networking sites, forums, review sites) on the Internet to collect comments about mobile network service providers. The server uses scraping tools and APIs to obtain text information and the following metadata for each comment:

[0417] timestamp

[0418] Source URL

[0419] Username (if available)

[0420] Applicable carrier name

[0421] As a specific example, the server uses a social media API to collect tweets containing hashtags such as "Career A Review" and store them in a database.

[0422] Data Preprocessing

[0423] Processing performed by the server

[0424] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization, normalization, and stop word removal is performed. This prepares the text data in a form suitable for analysis.

[0425] For example, a comment such as "This service is really great! Ad" is tokenized as "This service is really great" and unnecessary parts are removed.

[0426] Classification and scoring of NPS comments

[0427] Processing performed by the AI ​​module (on the server)

[0428] The pre-processed comments are sent to an AI sentiment analysis module. The server uses AI to analyze the sentiment of each comment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content of each comment. This also generates an overall NPS score for each carrier.

[0429] For example, the server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and sets the NPS score to 8.

[0430] Report Generation

[0431] Processing performed by the server

[0432] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each carrier. It also generates comparative information with other carriers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0433] As a concrete example, the server generates a report containing the information, "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan."

[0434] Report distribution and visualization

[0435] Processing performed by the user terminal

[0436] The generated reports are displayed in real time on a dashboard on the user's device. Users can log in to the dashboard to view the latest reports. Furthermore, they can use the filter and search functions to view detailed information for specific periods or by carrier.

[0437] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as reference for their next marketing initiative.

[0438] The above is a detailed description of the embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, enabling efficient use of human resources and rapid decision-making.

[0439] The processing flow will be explained below.

[0440] Step 1:

[0441] Processing performed by the server

[0442] The server periodically accesses designated websites on the Internet to collect comments about mobile network service providers. Using scraping tools and public APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, username (if available), and applicable carrier name.

[0443] Step 2:

[0444] Processing performed by the server

[0445] The collected comments are filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing is performed, including tokenization (splitting sentences into words), normalization (converting all to lowercase, removing special characters, etc.), and removal of stop words. This prepares the text data in a form suitable for analysis.

[0446] Step 3:

[0447] Processing performed by the server

[0448] The preprocessed comments are sent to the AI ​​sentiment analysis module, which uses an AI model to analyze the sentiment of each comment and classify it as positive, negative, or neutral. It then calculates an NPS score based on the content of each comment and stores it in a database.

[0449] Step 4:

[0450] Processing performed by the server

[0451] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, and generates comparative information with other carriers, including calculating each carrier's overall NPS score and analyzing trends over time.

[0452] Step 5:

[0453] Processing performed by the server

[0454] Based on the aggregated data and analysis results, the server automatically generates a report that includes the overall NPS score, the distribution of positive and negative comments, comparisons with other carriers, and predictions of future changes.

[0455] Step 6:

[0456] Processing performed by the user terminal

[0457] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to the dashboard to check the latest reports. They can also use the filter and search functions to view detailed information for specific periods or by carrier.

[0458] Step 7:

[0459] User actions

[0460] Planning and marketing staff conduct analysis based on the reports generated on the dashboard. This allows them to quickly evaluate the effectiveness of their initiatives and quickly consider and implement any necessary improvements. For example, they can evaluate the effectiveness of a new pricing plan based on information such as "Carrier A's NPS score has improved by 10 points since last month, and the number of positive comments has increased."

[0461] Example 1

[0462] 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."

[0463] It is inefficient to manually collect appropriate feedback in real time from a huge amount of user comments about mobile network service providers, perform sentiment analysis, and automatically generate comprehensive reports including competitive comparisons. There is a need for a system that solves this problem and automates and streamlines the process from data collection to report creation.

[0464] 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.

[0465] In this invention, the server includes a means for collecting comments about mobile network service providers from data sources on the Internet, a means for filtering the collected comments using regular expressions to remove unnecessary data such as spam and advertising links, and a text preprocessing means for tokenizing, normalizing, and removing stop words from the filtered comments. This enables useful information to be extracted from large amounts of data, enabling rapid and accurate sentiment analysis and NPS score evaluation. The server also includes a means for delivering automatically generated reports to user terminals and displaying them in real time, allowing users to make quick decisions based on the latest information.

[0466] "Data Source" refers to any place where information can be collected, such as any website, social media site, forum, review site, or other website accessible on the Internet.

[0467] "Comment" means text data such as feedback, opinions, reviews, etc. posted by a user to a mobile network service provider.

[0468] "Filtering" refers to the process of removing unnecessary information such as spam and advertising links from collected comment data and formatting it into data suitable for analysis.

[0469] "Tokenization" refers to the process of dividing text into the smallest meaningful units, such as words or phrases.

[0470] "Normalization" refers to the process of converting text data into a unified format to improve the accuracy of analysis.

[0471] "Stop words" refer to common words (e.g., wa, ga, no, etc.) that are of little importance and have no meaning in text analysis.

[0472] "Sentiment analysis" refers to the process of analyzing text data and identifying its emotional tendencies (positive, negative, neutral).

[0473] The Net Promoter Score (NPS) is an index that evaluates customer satisfaction and the degree of recommendation, and is scored on a scale of 0 to 10.

[0474] "Positive comments" refer to comments that contain positive or favorable feedback about a service or product.

[0475] "Negative Comments" refers to comments that contain negative or unsatisfactory feedback about a service or product.

[0476] "Report" means a document generated from collected and analyzed data that includes results, trends, comparative information, etc.

[0477] "Dashboard" refers to a web application or software that includes an interface that allows users to view, search, and analyze information in real time.

[0478] This invention relates to a system that collects comments about mobile network service providers from the Internet and uses AI to analyze those comments and generate a report of the results. This system automates the process from data collection to report generation and distribution through the following series of processes.

[0479] Data collection

[0480] Processing performed by the server

[0481] The server periodically accesses designated websites (e.g., social networking sites, forums, review sites, etc.) on the Internet. Specifically, the server uses scraping tools and APIs (e.g., Twitter API) to collect targeted comments. The collected information includes text, timestamp, source URL, username (if available), and corresponding carrier name.

[0482] As a specific example, the server uses the Twitter API to collect tweets containing the hashtag "Career A Review" and store them in a database.

[0483] Data Preprocessing

[0484] Processing performed by the server

[0485] The collected comment data is first filtered using regular expressions to remove unwanted data such as spam and advertising links, then the comment text is tokenized, normalized, and stop words are removed.

[0486] For example, a comment like "This service is really great! Ad" would be tokenized as "This service is really great" and the spam part would be removed.

[0487] Classification and scoring of NPS comments

[0488] Processing performed by the server

[0489] The pre-processed comments are sent to an AI sentiment analysis module, which analyzes the sentiment of each comment and categorizes it as positive, negative, or neutral. It then calculates an NPS score based on the content of each comment and generates an overall NPS score for each carrier.

[0490] As a concrete example, the server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and sets the NPS score to 8.

[0491] Report Generation

[0492] Processing performed by the server

[0493] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, generates comparison information with other carriers, and automatically generates a comprehensive report with forecast results for important fluctuations.

[0494] As a specific example, the server compiles information such as "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan" and includes it in a report.

[0495] Report distribution and visualization

[0496] Processing performed by the user terminal

[0497] The generated reports are displayed in real time on a dashboard on the user's device, and users can log in to view the latest reports and use filters and search functions to view detailed information for specific periods or by carrier.

[0498] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as a reference for their next marketing initiative.

[0499] Examples of prompt statements

[0500] Examples of input prompts for generative AI models include:

[0501] "Please explain the process flow of a system that collects NPS comments about mobile network service providers, performs sentiment analysis and scoring using AI, and compiles the results into a report, including the specific processing steps."

[0502] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0503] Step 1: Data collection

[0504] The server periodically accesses designated data sources (e.g., social media sites, forums, review sites) on the Internet. The server uses scraping tools or social media APIs (e.g., Twitter API) to collect comments about mobile network service providers. As input, the server receives target comments and hashtags (e.g., "Carrier A Review"). As output, the server stores data in a database, including text, timestamp, source URL, username (if available), and the name of the corresponding carrier.

[0505] Specific operation: The server uses the Twitter API to collect tweets containing the hashtag "Career A Review" and stores that data in a database.

[0506] Step 2: Filtering the data

[0507] The server uses regular expressions to filter the collected comment data to remove unwanted data such as spam and advertising links. It receives the collected raw data as input and generates clean filtered data as output.

[0508] Specific operation: The comment "This service is really great! Ad" is converted into clean data using regular expressions to remove unnecessary parts such as "This service is really great."

[0509] Step 3: Tokenize and normalize the data

[0510] The filtered comments are then tokenized and normalized. The server receives the filtered text data as input, tokenizes it (splits the text into words and phrases), normalizes it (converts the text into a uniform format), and removes stop words. As output, it produces tokenized data suitable for analysis.

[0511] Specific behavior: Split the text "This service is really great!" into "This service is really great", removing spelling variations and extra words.

[0512] Step 4: Sentiment analysis and NPS score calculation

[0513] The preprocessed comments are sent to the AI ​​sentiment analysis module. The server receives the tokenized text data as input and performs sentiment analysis using an AI model. It classifies each comment as positive, negative, or neutral and calculates an NPS score based on its content. As output, it generates data containing the sentiment classification result and the NPS score.

[0514] Specific behavior: The server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and calculates an NPS score of 8.

[0515] Step 5: Generate a report

[0516] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, and automatically generates a comprehensive report including comparisons with other carriers and predictions of significant fluctuations. It receives the results of sentiment analysis and NPS scores as input and generates a comprehensive report as output.

[0517] Specific operation: Based on the information that "Carrier A's NPS score this month is 72, an improvement of 10 points from last month," the server compiles a report stating that 50% of users have given positive feedback about the new pricing plan.

[0518] Step 6: Report distribution and visualization

[0519] The user terminal displays the generated reports on a dashboard in real time. Users can log in and view the latest reports. They can also use filters and search functions to check detailed information for specific periods or by carrier. The system receives reports provided by the server as input and displays them on the dashboard as output.

[0520] Specific actions: The planner accesses the dashboard and checks the information that "Career A's NPS score this month is 76, and positive comments have increased by 50%," and uses this information as a reference for the next marketing initiative.

[0521] (Application example 1)

[0522] 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."

[0523] Content distribution services are required to collect and analyze viewer feedback in real time, and to efficiently recommend personalized content and propose improvements based on that feedback. Conventional methods have had the problem of being unable to provide appropriate content due to the difficulty of quickly reflecting viewer reactions.

[0524] 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.

[0525] In this invention, the server includes means for collecting comments about a mobile network service provider from websites on the Internet, means for filtering and text preprocessing the collected comments, means for classifying the preprocessed comments into positive, negative, and neutral categories and calculating a Net Promoter Score, means for calculating the distribution of positive and negative comments for each provider and generating comparative information with other providers, means for automatically generating a report including the generated information and predicted fluctuations, means for delivering and displaying the automatically generated report to a user terminal, and means for aggregating viewer responses in real time on the user terminal and recommending content and proposing improvements based on viewing trends, thereby enabling fast and efficient collection and analysis of viewer feedback and the provision of personalized content.

[0526] The "Internet" is a communications infrastructure that connects computer networks around the world and enables the exchange of information.

[0527] A "website" is a collection of information made up of multiple web pages published on the Internet.

[0528] A "mobile network service provider" is a company or organization that provides communications services to mobile phones and other mobile communication devices.

[0529] A "comment" is text information in which a user expresses an opinion or comment on a particular topic.

[0530] "Filtering" is the process of removing unnecessary information from collected data.

[0531] "Text preprocessing" refers to a series of processes used to format text data into a form suitable for analysis.

[0532] "Positive" refers to comments that express positive or favorable sentiment.

[0533] "Negative" refers to comments that express negative or unfavorable sentiment.

[0534] "Neutral" refers to comments that express a neutral sentiment, neither positive nor negative.

[0535] "Net Promoter Score" is a quantified indicator of the likelihood that a customer will recommend a service to others.

[0536] "Provider" refers to a company or organization that provides a particular service or content.

[0537] "Distribution" refers to the frequency or proportion of occurrence in each category of data.

[0538] "Comparative information" is information that shows the differences and similarities between two or more data sets.

[0539] A "report" is a document that summarizes analytical results and statistical data.

[0540] A "user terminal" is a device that a user uses to access the Internet or a system.

[0541] "Real-time" refers to the property that data and events are processed as they occur.

[0542] A "viewer" is a user who views video and audio content.

[0543] "Reactions" refer to the emotions and evaluations that viewers express toward content.

[0544] "Aggregation" is the process of statistically summarizing data.

[0545] "Viewing habits" refers to a viewer's general behavior and preference patterns regarding content.

[0546] "Content" refers to the media that viewers consume, such as television programs, movies, music, articles, etc.

[0547] A "recommendation" is the presentation of particular content to a viewer.

[0548] "Suggestions for improvement" refers to suggestions for improving the quality of content based on viewer reactions and feedback.

[0549] This invention provides a system that collects and analyzes viewer feedback in real time, and recommends personalized content and suggests improvements. Specific implementations of this system are described below.

[0550] Data collection method

[0551] The server collects viewers' comments from specific websites on the Internet (social networking sites, review sites, forums, etc.) using scraping tools and APIs to obtain text information and the following metadata for each comment:

[0552] timestamp

[0553] Source URL

[0554] Username (if available)

[0555] For example, the server uses a social media API to collect posts containing hashtags such as "Series X review" and store them in a database.

[0556] Data Preprocessing

[0557] The server filters the collected comment data and performs the following actions:

[0558] Remove unwanted data such as spam and advertising links (using regular expressions)

[0559] Text preprocessing, including tokenization, normalization, and stop word removal

[0560] This allows the comment to be formatted in a way that is suitable for sentiment analysis. For example, a comment like "This movie was great! Ad" is tokenized as "This movie was great" and unnecessary parts are removed.

[0561] Classification and scoring of NPS comments

[0562] The preprocessed comments are sent to an AI sentiment analysis module. The server uses AI (e.g., NLTK's VADER Sentiment Analyzer) to analyze the sentiment of each comment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content of each comment.

[0563] For example, a comment such as "This movie was amazing!" is considered positive and is assigned an NPS score of 9.

[0564] Report Generation and Delivery

[0565] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each provider, as well as comparisons with other providers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0566] For example, a report might be generated that includes information such as, "Provider A's NPS score this month is 50, up 15 points from last month. 30% of viewers have given positive feedback about new content."

[0567] The generated reports are displayed in real time on a dashboard on the user's device. Users can log in to this dashboard to view the latest reports. Furthermore, they can use the filter and search functions to view detailed information for specific periods or by provider.

[0568] Content recommendations and improvement suggestions

[0569] Viewer responses are collected in real time on the user's device, and content recommendations and suggestions for improvements are made based on viewing trends. For example, it is possible to recommend the next content to watch based on viewer feedback data. In addition, if there is a lot of negative feedback, it is possible to analyze the cause and suggest improvements.

[0570] The specific hardware and software used

[0571] Hardware: Servers, user devices (smartphones, tablets, PCs, etc.)

[0572] Software: Python, BeautifulSoup, requests library, NLTK, VADER Sentiment Analyzer, Matplotlib

[0573] Examples of prompt statements

[0574] Here is an example prompt:

[0575] Analyze the sentiment of this comment: "This movie was amazing! Great story and acting. Advertisement"

[0576] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0577] Step 1:

[0578] The server collects viewer comments from specific websites on the Internet. Using APIs and scraping tools, the server obtains metadata such as comment text information, timestamps, source URLs, and user names. The input is the URL of a specific website, and the output is a list of collected comments and their metadata.

[0579] Step 2:

[0580] The server filters the collected comment data and removes inappropriate data (such as spam or advertising links). It uses regular expressions to detect and remove unnecessary data. The input is the comment data collected in the previous step, and the output is the clean comment data after filtering.

[0581] Step 3:

[0582] The server performs text preprocessing such as tokenization, normalization, and stop word removal. Tokenization splits comments into words, normalization converts uppercase to lowercase, and stop word removal removes common words such as "の" and "は." The input is the filtered comment data, and the output is the preprocessed text data.

[0583] Step 4:

[0584] The server sends the preprocessed comment data to the AI ​​sentiment analysis module for sentiment analysis. It uses NLTK's VADER Sentiment Analyzer to classify each comment's sentiment as positive, negative, or neutral. It then calculates the Net Promoter Score (NPS score) for each comment. The input is the preprocessed text data, and the output is the sentiment classification and NPS score for each comment.

[0585] Step 5:

[0586] The server calculates the distribution of positive and negative comments for each provider and generates comparative information with other providers. It also automatically generates a comprehensive report including predictions of significant fluctuations. The input is the comment data after sentiment analysis and the NPS score, and the output is a report including information for each provider and comparisons.

[0587] Step 6:

[0588] The server delivers the automatically generated reports to the user terminal and displays them. The user terminal displays the reports in real time on a dashboard, allowing the user to search for detailed information for a specific period or by provider. The input is the generated report, and the output is the dashboard display on the user terminal.

[0589] Step 7:

[0590] The user terminal aggregates viewer responses in real time and makes content recommendations and suggestions for improvements based on viewing trends. The input is viewer feedback data, and the output is content recommendations and suggestions for improvements to the viewer.

[0591] 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.

[0592] The present invention relates to a system that collects NPS comments about mobile network service providers from the Internet and generates analysis and result reports using AI and an emotion engine. This system automates processes from data collection to emotion analysis of comments, report generation, and distribution. Furthermore, by combining this system with an emotion engine that can analyze user emotions in real time, highly accurate NPS scoring is achieved. Detailed descriptions of embodiments of the present invention are provided below.

[0593] Data collection from the internet

[0594] Processing performed by the server

[0595] The server periodically accesses designated websites (social media sites, forums, review sites, etc.) on the Internet to collect comments about mobile network service providers. Using scraping tools and APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, username (if available), and applicable carrier name.

[0596] As a specific example, the server uses a social media API to collect tweets containing hashtags such as "Career A Review" and store them in a database.

[0597] Data Preprocessing

[0598] Processing performed by the server

[0599] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization and normalization is performed. This prepares the text data in a form suitable for analysis.

[0600] For example, a comment such as "This service is really great! Ad" is tokenized as "This service is really great" and unnecessary parts are removed.

[0601] Classification and scoring of NPS comments

[0602] Processing performed by the AI ​​module and emotion engine (on the server)

[0603] The preprocessed comments are sent to an AI sentiment analysis module and sentiment engine. The server uses the sentiment engine to analyze user sentiment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content and sentiment data of each comment, which also generates an overall NPS score for each carrier.

[0604] For example, the server might send a comment like "This service is really great" to the sentiment engine, which would then consider it positive and set the NPS score to 8.

[0605] Report Generation

[0606] Processing performed by the server

[0607] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each carrier. It also generates comparative information with other carriers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0608] As a concrete example, the server generates a report containing the information, "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan."

[0609] Report distribution and visualization

[0610] Processing performed by the user terminal

[0611] The user device displays reports delivered from the server on a dashboard in real time. Users can log in to this dashboard to check the latest reports. They can also use the filter and search functions to check detailed information for specific periods or by carrier.

[0612] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as reference for their next marketing initiative.

[0613] The above is a detailed description of an embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, enabling efficient use of human resources and rapid decision-making. Furthermore, the use of an emotion engine makes it possible to more accurately grasp user emotions and achieve highly accurate NPS scoring.

[0614] The processing flow will be explained below.

[0615] Step 1:

[0616] Processing performed by the server

[0617] The server periodically accesses designated websites on the Internet and uses scraping tools or public APIs to collect comments about mobile network service providers, along with metadata for each comment (timestamp, source URL, user name, and corresponding carrier name), which are then stored in a database.

[0618] Step 2:

[0619] Processing performed by the server

[0620] The collected comment data is filtered, using regular expressions to remove unwanted data such as spam and advertising links, and the comments are tokenized and subjected to text preprocessing, which includes splitting sentences into words, normalization, and stop word removal.

[0621] Step 3:

[0622] Processing performed by the server

[0623] The preprocessed comments are sent to the sentiment engine and AI sentiment analysis module. The sentiment engine analyzes the user sentiment of each comment and classifies it as positive, negative, or neutral. The AI ​​sentiment analysis module calculates an NPS score for each comment based on the sentiment data, which then derives an overall NPS score for each carrier.

[0624] Step 4:

[0625] Processing performed by the server

[0626] The server periodically compiles NPS data, calculates the distribution of positive and negative comments for each carrier, and generates comparative information with other carriers, including calculating each carrier's overall NPS score and analyzing trends over time.

[0627] Step 5:

[0628] Processing performed by the server

[0629] Based on the aggregated data and analysis results, the server automatically generates a report that includes the overall NPS score, the distribution of positive and negative comments, comparison information with other carriers, and a forecast of future fluctuations, allowing users to grasp the necessary information at a glance.

[0630] Step 6:

[0631] Processing performed by the user terminal

[0632] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to the dashboard to check the latest reports. They can also use the filter and search functions to view detailed information for specific periods or by carrier.

[0633] Step 7:

[0634] User actions

[0635] Planning and marketing staff conduct analysis based on the reports generated on the dashboard. This allows them to quickly evaluate the effectiveness of their initiatives and quickly consider and implement any necessary improvements. For example, they can evaluate the effectiveness of a new pricing plan based on information such as "Carrier A's NPS score has improved by 10 points since last month, and the number of positive comments has increased."

[0636] Example 2

[0637] 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."

[0638] The vast amount of comment data available online varies in content and format, making it difficult to manually collect and analyze, and therefore difficult to efficiently calculate NPS scores. Furthermore, accurate understanding of user sentiment and reflecting it in marketing and service improvements requires advanced analysis. There is a need for a system that can solve these problems, achieve highly accurate NPS scoring, and generate reports quickly.

[0639] 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.

[0640] In this invention, the server includes means for collecting evaluation comments about a communication service provider from Internet sources, including designated websites and social networks, means for filtering and text preprocessing the collected comments, means for classifying the preprocessed comments into positive, negative, and neutral categories and calculating a Net Promoter Score, means for calculating the distribution of positive and negative comments for each communication service provider and generating comparative information with other communication service providers, means for automatically generating a report including the generated information and predicted fluctuations, and means for delivering and displaying the automatically generated report to a user device. This allows for efficient calculation of NPS scores while accurately grasping user sentiment, and enables rapid report generation and delivery.

[0641] "Designated websites and social networks" are pre-selected internet websites, social media platforms, and other online platforms used to collect information.

[0642] A "communications service provider" is a company or organization that provides communications services such as mobile communications and internet access.

[0643] "Evaluation comments" are written evaluations such as feedback, reviews, opinions, etc., given by users to communication service providers.

[0644] "Filtering" is the process of removing unnecessary data to extract only useful data.

[0645] "Text preprocessing" refers to the process of preparing data for later analysis, such as tokenizing and normalizing comment data.

[0646] "Classifying into positive, negative, and neutral" is the process of classifying emotions and evaluations into three categories based on the content of comments.

[0647] "Net Promoter Score" is an index that scores the likelihood that users will recommend a service to others.

[0648] "Distribution of positive and negative comments" is data that shows the percentage of positive and negative comments for a particular period or career.

[0649] "Comparative information" is information for comparing the ratings and performance of different communication service providers.

[0650] A "report including predicted change results" is a report that predicts NPS scores and evaluation data that change over time and summarizes the results.

[0651] A "user device" is a device such as a computer, smartphone, or tablet that a user uses to receive and display reports.

[0652] The present invention relates to a system that collects evaluation comments about telecommunications service providers from the Internet and generates analysis and result reports using AI and an emotion engine. This system automates processes from data collection to emotion analysis of comments, report generation, and distribution. Furthermore, by combining it with an emotion engine that can analyze user emotions in real time, highly accurate Net Promoter Scoring is achieved. The following describes in detail an embodiment of the present invention.

[0653] Data collection from the internet

[0654] The server accesses designated websites and social networks (e.g., social networking sites, forums, review sites) to periodically collect comments about telecommunications service providers. Using scraping tools (e.g., the BeautifulSoup library) or APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, user name, and the name of the applicable carrier.

[0655] Data Preprocessing

[0656] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization and normalization is performed. This prepares the text data in a form suitable for analysis.

[0657] Classification and scoring of NPS comments

[0658] The preprocessed comments are sent to an AI sentiment analysis module and sentiment engine. The server uses the sentiment engine to analyze user sentiment and categorize it as positive, negative, or neutral. It then calculates an NPS score based on the content and sentiment data of each comment, which then generates an overall NPS score for each telecommunications service provider.

[0659] Report Generation

[0660] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each telecommunications service provider. It also generates comparative information with other telecommunications service providers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0661] Report distribution and visualization

[0662] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to this dashboard and check the latest reports. They can also use the filter and search functions to check detailed information for specific periods or by communication service provider.

[0663] As a concrete example of a prompt sentence, the following input is given to the generative AI model:

[0664] "Analyze the sentiment of NPS comments using the data below and generate a report.

[0665] Example data:

[0666] Comment: "This service is absolutely amazing!"

[0667] Timestamp: "2023-10-13 12:00:00"

[0668] Source URL: "https: / / example.com / review / 12345"

[0669] Username: "user123"

[0670] Applicable carrier name: "Carrier A"

[0671] This concludes the details of the embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, making it possible to accurately grasp user sentiment and efficiently calculate NPS scores.

[0672] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0673] Processing Steps:

[0674] Step 1: Collecting data from the internet

[0675] Processing performed by the server

[0676] Input: The URL or API endpoint of a specified website or social network (social network, forum, review site, etc.).

[0677] Specific operation: Uses Python's BeautifulSoup library to scrape and retrieve comments containing specified keywords (e.g., "Telecommunications Service Provider A Review"). If necessary, uses the social media API to collect data in real time.

[0678] Output: Collected comment data (comment text, timestamp, source URL, username, carrier name).

[0679] Step 2: Preprocessing the data

[0680] Processing performed by the server

[0681] Input: Collected comment data.

[0682] What it does: It uses regular expressions to filter out unwanted data like spam and advertising links, then uses the NLTK library to tokenize and normalize the comment text.

[0683] Output: Filtered and preprocessed comment data.

[0684] Step 3: Categorize and score NPS comments

[0685] Processing performed by the server

[0686] Input: Preprocessed comment data.

[0687] What it does: It uses HuggingFace's Transformers library to run a sentiment analysis model to classify comments as positive, negative, or neutral, and then calculates an NPS score based on the classification (8 for positive, 2 for negative, 5 for neutral).

[0688] Output: Comment data with sentiment classification results and NPS scores.

[0689] Step 4: Generate a report

[0690] Processing performed by the server

[0691] Input: Comment data with sentiment classification results and NPS scores.

[0692] What it does: Aggregates the distribution of positive, negative, and neutral comments for each telecommunications service provider, calculates an overall NPS score, performs comparative analysis with other telecommunications service providers, and automatically generates a comprehensive report with predicted results of important fluctuations.

[0693] Output: An automatically generated report containing the aggregated and analyzed results.

[0694] Step 5: Report distribution and visualization

[0695] Processing performed by the user terminal

[0696] Input: Report delivered from the server.

[0697] Specific operation: The report is reflected in a dashboard that displays the report in real time on the user's device. On the dashboard, users can search and filter for detailed information for specific periods or by communication service provider.

[0698] Output: The latest report and details displayed to the user.

[0699] This flow of processing steps automates the entire system, allowing users to efficiently check the analysis results of NPS comments.

[0700] (Application example 2)

[0701] 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."

[0702] Currently, systems for collecting and analyzing feedback on the efficiency and performance of factory robots are inadequate, preventing effective use of this information. This makes it difficult to quickly and accurately develop specific measures to improve factory operational efficiency and robot performance. Furthermore, when the process from collecting feedback to analyzing, generating, and distributing reports is carried out manually, it results in wasted human resources and makes it difficult to reflect information in a timely manner.

[0703] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for collecting comments about objects from websites on the Internet; means for filtering and text preprocessing the collected comments; means for classifying the comments into positive, negative, and neutral based on the preprocessed comments and calculating a performance score; means for calculating the distribution of positive and negative comments for each object and generating comparative information with other objects; means for automatically generating a report including the generated information and predicted fluctuations; means for delivering and displaying the automatically generated report to a user terminal; means for a user to view the report in real time via a dashboard and check detailed information; and means for using AI and an emotion engine to collect and analyze feedback comments about the efficiency and performance of objects. This makes it possible to quickly and automatically collect and analyze feedback about the efficiency and performance of factory robots and generate and distribute highly accurate performance reports.

[0704] 1. "Internet website" refers to an information site on the Internet that is accessible to users.

[0705] 2. "Object" refers to the factory robot or other equipment that is the subject of evaluation in this invention.

[0706] 3. "Comments" refers to the written opinions and feedback of users regarding the subject matter.

[0707] 4. "Collection methods" refers to the technologies and methods used to extract comments from websites on the Internet.

[0708] 5. "Filtering" refers to the process of removing unnecessary information and noise from acquired comment data.

[0709] 6. "Text preprocessing" refers to procedures that transform comment text into a form suitable for analysis, such as tokenization and normalization.

[0710] 7. "Positive" refers to comments that have a favorable opinion or reputation about the subject.

[0711] 8. "Negative" refers to comments that have a negative opinion or reputation about the subject.

[0712] 9. "Neutral" refers to comments that have a neutral opinion or do not evoke any particular strong emotion toward the subject.

[0713] 10. "Performance Score" refers to a numerical rating that indicates the efficiency or performance of an object.

[0714] 11. "Means for calculating distribution" refers to the techniques and methods for calculating the ratio of positive and negative comments from the collected comment data.

[0715] 12. "Comparative information" refers to data and statistics that allow the evaluation of an object to be compared with other objects.

[0716] 13. "Means for automatically generating reports" refers to technologies and methods for periodically generating reports based on analysis results.

[0717] 14. "User terminal" refers to a device used to receive and display reports, such as a personal computer or smartphone.

[0718] 15. "Means of delivery" refers to the technology or method for sending automatically generated reports to user terminals.

[0719] 16. "Means of display" refers to the technology or method for displaying the delivered report in an easy-to-read format on the user's device.

[0720] 17. "Dashboard" refers to an interface for aggregating and visually displaying various data and reports.

[0721] 18. "Real-time viewing means" refers to technologies and methods that allow users to instantly view the latest information and reports.

[0722] 19. "Means for checking detailed information" refers to technologies and methods that allow users to search and view detailed data for specific periods or subjects.

[0723] 20. "Feedback comments" refers to user evaluations and opinions regarding an object.

[0724] 21. “AI” refers to artificial intelligence technologies, including machine learning models and data analysis algorithms.

[0725] 22. “Sentiment Engine” refers to technology or software used to analyze comment data to identify user sentiment.

[0726] This invention relates to a system that collects feedback on the efficiency and performance of factory robots from websites on the internet, analyzes the feedback, and generates and distributes reports. The system uses AI and an emotion engine to analyze the sentiment of comments and calculate a performance score.

[0727] Program generation and processing steps

[0728] Server-side processing

[0729] 1. Data Collection:

[0730] The server uses web scraping tools such as Scrapy to collect comments about factory robots from social media, review sites, and forums on the Internet, such as comments containing the hashtag "factory robot performance."

[0731] 2. Data preprocessing:

[0732] Collected comments are filtered, tokenized, and normalized using Python, which includes removing spam and ads, and eliminating unnecessary data.

[0733] 3. Sentiment Analysis and Scoring:

[0734] The preprocessed comments are fed into a sentiment analysis model powered by TensorFlow / Keras, where they are classified as positive, negative, or neutral and a performance score is calculated.

[0735] 4. Report Generation:

[0736] Based on the aggregated data, a report is automatically generated using pandas and matplotlib, including the distribution of positive and negative comments, as well as a comparison with other robots.

[0737] 5. Report Delivery:

[0738] The reports are delivered to the user's device and displayed on a dashboard in real time.

[0739] Processing on the user terminal side

[0740] 1. Dashboard view:

[0741] The delivered reports are displayed in real time on the user's device using JavaScript (especially React.js), and users can view detailed information for specific periods or for each factory robot through this dashboard.

[0742] Examples of concrete examples and prompts

[0743] Examples:

[0744] When a user accesses the dashboard, the "Performance Report for Factory Robot A this Month" will show "Performance score increased by 45% and positive comments increased by 60%." Users can also use the filter function to specify a specific period and view performance data for that period in more detail.

[0745] Example prompt sentence:

[0746] The prompt sentence is an example of input to the generative AI model, and looks like this:

[0747] Generate a monthly report for a factory robot performance feedback system. Separate positive, negative, and neutral comments, calculate a score for each, and output the report.

[0748] This completes the description of an embodiment of the present invention, which enables the rapid and automatic collection and analysis of feedback on the efficiency and performance of factory robots, and the generation and distribution of highly accurate performance reports.

[0749] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0750] Step 1: Data collection

[0751] The server uses a web scraping tool such as Scrapy to collect comments about factory robots from social media, review sites, and forums on the Internet. For example, it collects comments containing the hashtag "factory robot performance." The input is the comment data obtained from social media and review sites, and the output is a dataset containing these comments.

[0752] Step 2: Filtering the data

[0753] The server filters the collected comment data. It uses regular expressions in Python to remove unnecessary data, such as spam and advertisements. Specifically, it detects comments containing "advertisements" or "spam links" and does not send them to the preprocessing step. The input is the initial comment data, and the output is the cleaned comment data.

[0754] Step 3: Preprocessing the text

[0755] The server preprocesses the filtered comments, performing tokenization and normalization to make the text suitable for analysis. Specifically, it splits sentences into tokens and removes unnecessary symbols and special characters. For example, it converts a comment such as "This factory robot is great" into "This factory robot is great." The input is filtered comment data, and the output is preprocessed text data.

[0756] Step 4: Sentiment analysis

[0757] The preprocessed comments are sent to a sentiment analysis model (generative AI model) using TensorFlow / Keras. The server performs data calculations to classify comments into positive, negative, or neutral. Specifically, the model predicts the sentiment of each comment and classifies it as a category. For example, a comment saying "great" is classified as positive. The input is the preprocessed text data, and the output is the sentiment category (positive, negative, neutral) and its confidence score.

[0758] Step 5: Calculating the Performance Score

[0759] The server calculates a performance score based on the results of the sentiment analysis. It aggregates the percentage of positive and negative comments and converts them into a score. Specifically, it calculates the performance score from the ratio of positive comments and generates an overall score from, for example, "60% positive, 40% negative." The input is sentiment category data, and the output is a performance score.

[0760] Step 6: Generate reports

[0761] The server automatically generates a report using pandas and matplotlib based on the aggregated data. This report includes the distribution of positive and negative comments and comparison information with other robots. Specifically, it creates a graph of the progress of scores and distribution, and compiles it into a report document. The input is the aggregated data, and the output is a report document.

[0762] Step 7: Report Delivery

[0763] The server delivers the generated report to the user's device. The report is automatically sent via email or a cloud platform. Specifically, the report file is delivered to the user using an email sending API or cloud storage service. The input is the report document, and the output is a notification that the report has been sent.

[0764] Step 8: View the dashboard

[0765] On the user's device, JavaScript (particularly React.js) is used to display the delivered reports in real time on a dashboard. Through this dashboard, users can check detailed information for a specific period or for each factory robot. Specifically, specific data can be displayed using date filters and search functions. The input is the report data, and the output is a visualized report display.

[0766] These are the specific processing steps of the system program that realizes this application example, which allows feedback on the efficiency and performance of factory robots to be collected and analyzed sequentially, with the results immediately available.

[0767] 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.

[0768] 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.

[0769] 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.

[0770] [Third embodiment]

[0771] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0772] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0773] 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).

[0774] 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.

[0775] 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.

[0776] 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).

[0777] 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.

[0778] 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.

[0779] 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.

[0780] 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.

[0781] 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.

[0782] 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."

[0783] The present invention relates to a system that collects NPS comments about mobile network service providers from the Internet and generates analysis and result reports using AI. This system automates processes from data collection to sentiment analysis of comments, report generation, and distribution. Detailed descriptions of embodiments of the present invention are provided below.

[0784] Data collection from the internet

[0785] Processing performed by the server

[0786] The server periodically accesses designated websites (e.g., social networking sites, forums, review sites) on the Internet to collect comments about mobile network service providers. The server uses scraping tools and APIs to obtain text information and the following metadata for each comment:

[0787] timestamp

[0788] Source URL

[0789] Username (if available)

[0790] Applicable carrier name

[0791] As a specific example, the server uses a social media API to collect tweets containing hashtags such as "Career A Review" and store them in a database.

[0792] Data Preprocessing

[0793] Processing performed by the server

[0794] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization, normalization, and stop word removal is performed. This prepares the text data in a form suitable for analysis.

[0795] For example, a comment such as "This service is really great! Ad" is tokenized as "This service is really great" and unnecessary parts are removed.

[0796] Classification and scoring of NPS comments

[0797] Processing performed by the AI ​​module (on the server)

[0798] The pre-processed comments are sent to an AI sentiment analysis module. The server uses AI to analyze the sentiment of each comment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content of each comment. This also generates an overall NPS score for each carrier.

[0799] For example, the server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and sets the NPS score to 8.

[0800] Report Generation

[0801] Processing performed by the server

[0802] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each carrier. It also generates comparative information with other carriers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0803] As a concrete example, the server generates a report containing the information, "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan."

[0804] Report distribution and visualization

[0805] Processing performed by the user terminal

[0806] The generated reports are displayed in real time on a dashboard on the user's device. Users can log in to the dashboard to view the latest reports. Furthermore, they can use the filter and search functions to view detailed information for specific periods or by carrier.

[0807] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as reference for their next marketing initiative.

[0808] The above is a detailed description of the embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, enabling efficient use of human resources and rapid decision-making.

[0809] The processing flow will be explained below.

[0810] Step 1:

[0811] Processing performed by the server

[0812] The server periodically accesses designated websites on the Internet to collect comments about mobile network service providers. Using scraping tools and public APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, username (if available), and applicable carrier name.

[0813] Step 2:

[0814] Processing performed by the server

[0815] The collected comments are filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing is performed, including tokenization (splitting sentences into words), normalization (converting all to lowercase, removing special characters, etc.), and removal of stop words. This prepares the text data in a form suitable for analysis.

[0816] Step 3:

[0817] Processing performed by the server

[0818] The preprocessed comments are sent to the AI ​​sentiment analysis module, which uses an AI model to analyze the sentiment of each comment and classify it as positive, negative, or neutral. It then calculates an NPS score based on the content of each comment and stores it in a database.

[0819] Step 4:

[0820] Processing performed by the server

[0821] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, and generates comparative information with other carriers, including calculating each carrier's overall NPS score and analyzing trends over time.

[0822] Step 5:

[0823] Processing performed by the server

[0824] Based on the aggregated data and analysis results, the server automatically generates a report that includes the overall NPS score, the distribution of positive and negative comments, comparisons with other carriers, and predictions of future changes.

[0825] Step 6:

[0826] Processing performed by the user terminal

[0827] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to the dashboard to check the latest reports. They can also use the filter and search functions to view detailed information for specific periods or by carrier.

[0828] Step 7:

[0829] User actions

[0830] Planning and marketing staff conduct analysis based on the reports generated on the dashboard. This allows them to quickly evaluate the effectiveness of their initiatives and quickly consider and implement any necessary improvements. For example, they can evaluate the effectiveness of a new pricing plan based on information such as "Carrier A's NPS score has improved by 10 points since last month, and the number of positive comments has increased."

[0831] Example 1

[0832] 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."

[0833] It is inefficient to manually collect appropriate feedback in real time from a huge amount of user comments about mobile network service providers, perform sentiment analysis, and automatically generate comprehensive reports including competitive comparisons. There is a need for a system that solves this problem and automates and streamlines the process from data collection to report creation.

[0834] 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.

[0835] In this invention, the server includes a means for collecting comments about mobile network service providers from data sources on the Internet, a means for filtering the collected comments using regular expressions to remove unnecessary data such as spam and advertising links, and a text preprocessing means for tokenizing, normalizing, and removing stop words from the filtered comments. This enables useful information to be extracted from large amounts of data, enabling rapid and accurate sentiment analysis and NPS score evaluation. The server also includes a means for delivering automatically generated reports to user terminals and displaying them in real time, allowing users to make quick decisions based on the latest information.

[0836] "Data Source" refers to any place where information can be collected, such as any website, social media site, forum, review site, or other website accessible on the Internet.

[0837] "Comment" means text data such as feedback, opinions, reviews, etc. posted by a user to a mobile network service provider.

[0838] "Filtering" refers to the process of removing unnecessary information such as spam and advertising links from collected comment data and formatting it into data suitable for analysis.

[0839] "Tokenization" refers to the process of dividing text into the smallest meaningful units, such as words or phrases.

[0840] "Normalization" refers to the process of converting text data into a unified format to improve the accuracy of analysis.

[0841] "Stop words" refer to common words (e.g., wa, ga, no, etc.) that are of little importance and have no meaning in text analysis.

[0842] "Sentiment analysis" refers to the process of analyzing text data and identifying its emotional tendencies (positive, negative, neutral).

[0843] The Net Promoter Score (NPS) is an index that evaluates customer satisfaction and the degree of recommendation, and is scored on a scale of 0 to 10.

[0844] "Positive comments" refer to comments that contain positive or favorable feedback about a service or product.

[0845] "Negative Comments" refers to comments that contain negative or unsatisfactory feedback about a service or product.

[0846] "Report" means a document generated from collected and analyzed data that includes results, trends, comparative information, etc.

[0847] "Dashboard" refers to a web application or software that includes an interface that allows users to view, search, and analyze information in real time.

[0848] This invention relates to a system that collects comments about mobile network service providers from the Internet and uses AI to analyze those comments and generate a report of the results. This system automates the process from data collection to report generation and distribution through the following series of processes.

[0849] Data collection

[0850] Processing performed by the server

[0851] The server periodically accesses designated websites (e.g., social networking sites, forums, review sites, etc.) on the Internet. Specifically, the server uses scraping tools and APIs (e.g., Twitter API) to collect targeted comments. The collected information includes text, timestamp, source URL, username (if available), and corresponding carrier name.

[0852] As a specific example, the server uses the Twitter API to collect tweets containing the hashtag "Career A Review" and store them in a database.

[0853] Data Preprocessing

[0854] Processing performed by the server

[0855] The collected comment data is first filtered using regular expressions to remove unwanted data such as spam and advertising links, then the comment text is tokenized, normalized, and stop words are removed.

[0856] For example, a comment like "This service is really great! Ad" would be tokenized as "This service is really great" and the spam part would be removed.

[0857] Classification and scoring of NPS comments

[0858] Processing performed by the server

[0859] The pre-processed comments are sent to an AI sentiment analysis module, which analyzes the sentiment of each comment and categorizes it as positive, negative, or neutral. It then calculates an NPS score based on the content of each comment and generates an overall NPS score for each carrier.

[0860] As a concrete example, the server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and sets the NPS score to 8.

[0861] Report Generation

[0862] Processing performed by the server

[0863] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, generates comparison information with other carriers, and automatically generates a comprehensive report with forecast results for important fluctuations.

[0864] As a specific example, the server compiles information such as "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan" and includes it in a report.

[0865] Report distribution and visualization

[0866] Processing performed by the user terminal

[0867] The generated reports are displayed in real time on a dashboard on the user's device, and users can log in to view the latest reports and use filters and search functions to view detailed information for specific periods or by carrier.

[0868] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as a reference for their next marketing initiative.

[0869] Examples of prompt statements

[0870] Examples of input prompts for generative AI models include:

[0871] "Please explain the process flow of a system that collects NPS comments about mobile network service providers, performs sentiment analysis and scoring using AI, and compiles the results into a report, including the specific processing steps."

[0872] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0873] Step 1: Data collection

[0874] The server periodically accesses designated data sources (e.g., social media sites, forums, review sites) on the Internet. The server uses scraping tools or social media APIs (e.g., Twitter API) to collect comments about mobile network service providers. As input, the server receives target comments and hashtags (e.g., "Carrier A Review"). As output, the server stores data in a database, including text, timestamp, source URL, username (if available), and the name of the corresponding carrier.

[0875] Specific operation: The server uses the Twitter API to collect tweets containing the hashtag "Career A Review" and stores that data in a database.

[0876] Step 2: Filtering the data

[0877] The server uses regular expressions to filter the collected comment data to remove unwanted data such as spam and advertising links. It receives the collected raw data as input and generates clean filtered data as output.

[0878] Specific operation: The comment "This service is really great! Ad" is converted into clean data using regular expressions to remove unnecessary parts such as "This service is really great."

[0879] Step 3: Tokenize and normalize the data

[0880] The filtered comments are then tokenized and normalized. The server receives the filtered text data as input, tokenizes it (splits the text into words and phrases), normalizes it (converts the text into a uniform format), and removes stop words. As output, it produces tokenized data suitable for analysis.

[0881] Specific behavior: Split the text "This service is really great!" into "This service is really great", removing spelling variations and extra words.

[0882] Step 4: Sentiment analysis and NPS score calculation

[0883] The preprocessed comments are sent to the AI ​​sentiment analysis module. The server receives the tokenized text data as input and performs sentiment analysis using an AI model. It classifies each comment as positive, negative, or neutral and calculates an NPS score based on its content. As output, it generates data containing the sentiment classification result and the NPS score.

[0884] Specific behavior: The server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and calculates an NPS score of 8.

[0885] Step 5: Generate a report

[0886] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, and automatically generates a comprehensive report including comparisons with other carriers and predictions of significant fluctuations. It receives the results of sentiment analysis and NPS scores as input and generates a comprehensive report as output.

[0887] Specific operation: Based on the information that "Carrier A's NPS score this month is 72, an improvement of 10 points from last month," the server compiles a report stating that 50% of users have given positive feedback about the new pricing plan.

[0888] Step 6: Report distribution and visualization

[0889] The user terminal displays the generated reports on a dashboard in real time. Users can log in and view the latest reports. They can also use filters and search functions to check detailed information for specific periods or by carrier. The system receives reports provided by the server as input and displays them on the dashboard as output.

[0890] Specific actions: The planner accesses the dashboard and checks the information that "Career A's NPS score this month is 76, and positive comments have increased by 50%," and uses this information as a reference for the next marketing initiative.

[0891] (Application example 1)

[0892] 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."

[0893] Content distribution services are required to collect and analyze viewer feedback in real time, and to efficiently recommend personalized content and propose improvements based on that feedback. Conventional methods have had the problem of being unable to provide appropriate content due to the difficulty of quickly reflecting viewer reactions.

[0894] 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.

[0895] In this invention, the server includes means for collecting comments about a mobile network service provider from websites on the Internet, means for filtering and text preprocessing the collected comments, means for classifying the preprocessed comments into positive, negative, and neutral categories and calculating a Net Promoter Score, means for calculating the distribution of positive and negative comments for each provider and generating comparative information with other providers, means for automatically generating a report including the generated information and predicted fluctuations, means for delivering and displaying the automatically generated report to a user terminal, and means for aggregating viewer responses in real time on the user terminal and recommending content and proposing improvements based on viewing trends, thereby enabling fast and efficient collection and analysis of viewer feedback and the provision of personalized content.

[0896] The "Internet" is a communications infrastructure that connects computer networks around the world and enables the exchange of information.

[0897] A "website" is a collection of information made up of multiple web pages published on the Internet.

[0898] A "mobile network service provider" is a company or organization that provides communications services to mobile phones and other mobile communication devices.

[0899] A "comment" is text information in which a user expresses an opinion or comment on a particular topic.

[0900] "Filtering" is the process of removing unnecessary information from collected data.

[0901] "Text preprocessing" refers to a series of processes used to format text data into a form suitable for analysis.

[0902] "Positive" refers to comments that express positive or favorable sentiment.

[0903] "Negative" refers to comments that express negative or unfavorable sentiment.

[0904] "Neutral" refers to comments that express a neutral sentiment, neither positive nor negative.

[0905] "Net Promoter Score" is a quantified indicator of the likelihood that a customer will recommend a service to others.

[0906] "Provider" refers to a company or organization that provides a particular service or content.

[0907] "Distribution" refers to the frequency or proportion of occurrence in each category of data.

[0908] "Comparative information" is information that shows the differences and similarities between two or more data sets.

[0909] A "report" is a document that summarizes analytical results and statistical data.

[0910] A "user terminal" is a device that a user uses to access the Internet or a system.

[0911] "Real-time" refers to the property that data and events are processed as they occur.

[0912] A "viewer" is a user who views video and audio content.

[0913] "Reactions" refer to the emotions and evaluations that viewers express toward content.

[0914] "Aggregation" is the process of statistically summarizing data.

[0915] "Viewing habits" refers to a viewer's general behavior and preference patterns regarding content.

[0916] "Content" refers to the media that viewers consume, such as television programs, movies, music, articles, etc.

[0917] A "recommendation" is the presentation of particular content to a viewer.

[0918] "Suggestions for improvement" refers to suggestions for improving the quality of content based on viewer reactions and feedback.

[0919] This invention provides a system that collects and analyzes viewer feedback in real time, and recommends personalized content and suggests improvements. Specific implementations of this system are described below.

[0920] Data collection method

[0921] The server collects viewers' comments from specific websites on the Internet (social networking sites, review sites, forums, etc.) using scraping tools and APIs to obtain text information and the following metadata for each comment:

[0922] timestamp

[0923] Source URL

[0924] Username (if available)

[0925] For example, the server uses a social media API to collect posts containing hashtags such as "Series X review" and store them in a database.

[0926] Data Preprocessing

[0927] The server filters the collected comment data and performs the following actions:

[0928] Remove unwanted data such as spam and advertising links (using regular expressions)

[0929] Text preprocessing, including tokenization, normalization, and stop word removal

[0930] This allows the comment to be formatted in a way that is suitable for sentiment analysis. For example, a comment like "This movie was great! Ad" is tokenized as "This movie was great" and unnecessary parts are removed.

[0931] Classification and scoring of NPS comments

[0932] The preprocessed comments are sent to an AI sentiment analysis module. The server uses AI (e.g., NLTK's VADER Sentiment Analyzer) to analyze the sentiment of each comment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content of each comment.

[0933] For example, a comment such as "This movie was amazing!" is considered positive and is assigned an NPS score of 9.

[0934] Report Generation and Delivery

[0935] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each provider, as well as comparisons with other providers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0936] For example, a report might be generated that includes information such as, "Provider A's NPS score this month is 50, up 15 points from last month. 30% of viewers have given positive feedback about new content."

[0937] The generated reports are displayed in real time on a dashboard on the user's device. Users can log in to this dashboard to view the latest reports. Furthermore, they can use the filter and search functions to view detailed information for specific periods or by provider.

[0938] Content recommendations and improvement suggestions

[0939] Viewer responses are collected in real time on the user's device, and content recommendations and suggestions for improvements are made based on viewing trends. For example, it is possible to recommend the next content to watch based on viewer feedback data. In addition, if there is a lot of negative feedback, it is possible to analyze the cause and suggest improvements.

[0940] The specific hardware and software used

[0941] Hardware: Servers, user devices (smartphones, tablets, PCs, etc.)

[0942] Software: Python, BeautifulSoup, requests library, NLTK, VADER Sentiment Analyzer, Matplotlib

[0943] Examples of prompt statements

[0944] Here is an example prompt:

[0945] Analyze the sentiment of this comment: "This movie was amazing! Great story and acting. Advertisement"

[0946] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0947] Step 1:

[0948] The server collects viewer comments from specific websites on the Internet. Using APIs and scraping tools, the server obtains metadata such as comment text information, timestamps, source URLs, and user names. The input is the URL of a specific website, and the output is a list of collected comments and their metadata.

[0949] Step 2:

[0950] The server filters the collected comment data and removes inappropriate data (such as spam or advertising links). It uses regular expressions to detect and remove unnecessary data. The input is the comment data collected in the previous step, and the output is the clean comment data after filtering.

[0951] Step 3:

[0952] The server performs text preprocessing such as tokenization, normalization, and stop word removal. Tokenization splits comments into words, normalization converts uppercase to lowercase, and stop word removal removes common words such as "の" and "は." The input is the filtered comment data, and the output is the preprocessed text data.

[0953] Step 4:

[0954] The server sends the preprocessed comment data to the AI ​​sentiment analysis module for sentiment analysis. It uses NLTK's VADER Sentiment Analyzer to classify each comment's sentiment as positive, negative, or neutral. It then calculates the Net Promoter Score (NPS score) for each comment. The input is the preprocessed text data, and the output is the sentiment classification and NPS score for each comment.

[0955] Step 5:

[0956] The server calculates the distribution of positive and negative comments for each provider and generates comparative information with other providers. It also automatically generates a comprehensive report including predictions of significant fluctuations. The input is the comment data after sentiment analysis and the NPS score, and the output is a report including information for each provider and comparisons.

[0957] Step 6:

[0958] The server delivers the automatically generated reports to the user terminal and displays them. The user terminal displays the reports in real time on a dashboard, allowing the user to search for detailed information for a specific period or by provider. The input is the generated report, and the output is the dashboard display on the user terminal.

[0959] Step 7:

[0960] The user terminal aggregates viewer responses in real time and makes content recommendations and suggestions for improvements based on viewing trends. The input is viewer feedback data, and the output is content recommendations and suggestions for improvements to the viewer.

[0961] 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.

[0962] The present invention relates to a system that collects NPS comments about mobile network service providers from the Internet and generates analysis and result reports using AI and an emotion engine. This system automates processes from data collection to emotion analysis of comments, report generation, and distribution. Furthermore, by combining this system with an emotion engine that can analyze user emotions in real time, highly accurate NPS scoring is achieved. Detailed descriptions of embodiments of the present invention are provided below.

[0963] Data collection from the internet

[0964] Processing performed by the server

[0965] The server periodically accesses designated websites (social media sites, forums, review sites, etc.) on the Internet to collect comments about mobile network service providers. Using scraping tools and APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, username (if available), and applicable carrier name.

[0966] As a specific example, the server uses a social media API to collect tweets containing hashtags such as "Career A Review" and store them in a database.

[0967] Data Preprocessing

[0968] Processing performed by the server

[0969] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization and normalization is performed. This prepares the text data in a form suitable for analysis.

[0970] For example, a comment such as "This service is really great! Ad" is tokenized as "This service is really great" and unnecessary parts are removed.

[0971] Classification and scoring of NPS comments

[0972] Processing performed by the AI ​​module and emotion engine (on the server)

[0973] The preprocessed comments are sent to an AI sentiment analysis module and sentiment engine. The server uses the sentiment engine to analyze user sentiment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content and sentiment data of each comment, which also generates an overall NPS score for each carrier.

[0974] For example, the server might send a comment like "This service is really great" to the sentiment engine, which would then consider it positive and set the NPS score to 8.

[0975] Report Generation

[0976] Processing performed by the server

[0977] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each carrier. It also generates comparative information with other carriers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[0978] As a concrete example, the server generates a report containing the information, "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan."

[0979] Report distribution and visualization

[0980] Processing performed by the user terminal

[0981] The user device displays reports delivered from the server on a dashboard in real time. Users can log in to this dashboard to check the latest reports. They can also use the filter and search functions to check detailed information for specific periods or by carrier.

[0982] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as reference for their next marketing initiative.

[0983] The above is a detailed description of an embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, enabling efficient use of human resources and rapid decision-making. Furthermore, the use of an emotion engine makes it possible to more accurately grasp user emotions and achieve highly accurate NPS scoring.

[0984] The processing flow will be explained below.

[0985] Step 1:

[0986] Processing performed by the server

[0987] The server periodically accesses designated websites on the Internet and uses scraping tools or public APIs to collect comments about mobile network service providers, along with metadata for each comment (timestamp, source URL, user name, and corresponding carrier name), which are then stored in a database.

[0988] Step 2:

[0989] Processing performed by the server

[0990] The collected comment data is filtered, using regular expressions to remove unwanted data such as spam and advertising links, and the comments are tokenized and subjected to text preprocessing, which includes splitting sentences into words, normalization, and stop word removal.

[0991] Step 3:

[0992] Processing performed by the server

[0993] The preprocessed comments are sent to the sentiment engine and AI sentiment analysis module. The sentiment engine analyzes the user sentiment of each comment and classifies it as positive, negative, or neutral. The AI ​​sentiment analysis module calculates an NPS score for each comment based on the sentiment data, which then derives an overall NPS score for each carrier.

[0994] Step 4:

[0995] Processing performed by the server

[0996] The server periodically compiles NPS data, calculates the distribution of positive and negative comments for each carrier, and generates comparative information with other carriers, including calculating each carrier's overall NPS score and analyzing trends over time.

[0997] Step 5:

[0998] Processing performed by the server

[0999] Based on the aggregated data and analysis results, the server automatically generates a report that includes the overall NPS score, the distribution of positive and negative comments, comparison information with other carriers, and a forecast of future fluctuations, allowing users to grasp the necessary information at a glance.

[1000] Step 6:

[1001] Processing performed by the user terminal

[1002] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to the dashboard to check the latest reports. They can also use the filter and search functions to view detailed information for specific periods or by carrier.

[1003] Step 7:

[1004] User actions

[1005] Planning and marketing staff conduct analysis based on the reports generated on the dashboard. This allows them to quickly evaluate the effectiveness of their initiatives and quickly consider and implement any necessary improvements. For example, they can evaluate the effectiveness of a new pricing plan based on information such as "Carrier A's NPS score has improved by 10 points since last month, and the number of positive comments has increased."

[1006] Example 2

[1007] 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."

[1008] The vast amount of comment data available online varies in content and format, making it difficult to manually collect and analyze, and therefore difficult to efficiently calculate NPS scores. Furthermore, accurate understanding of user sentiment and reflecting it in marketing and service improvements requires advanced analysis. There is a need for a system that can solve these problems, achieve highly accurate NPS scoring, and generate reports quickly.

[1009] 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.

[1010] In this invention, the server includes means for collecting evaluation comments about a communication service provider from Internet sources, including designated websites and social networks, means for filtering and text preprocessing the collected comments, means for classifying the preprocessed comments into positive, negative, and neutral categories and calculating a Net Promoter Score, means for calculating the distribution of positive and negative comments for each communication service provider and generating comparative information with other communication service providers, means for automatically generating a report including the generated information and predicted fluctuations, and means for delivering and displaying the automatically generated report to a user device. This allows for efficient calculation of NPS scores while accurately grasping user sentiment, and enables rapid report generation and delivery.

[1011] "Designated websites and social networks" are pre-selected internet websites, social media platforms, and other online platforms used to collect information.

[1012] A "communications service provider" is a company or organization that provides communications services such as mobile communications and internet access.

[1013] "Evaluation comments" are written evaluations such as feedback, reviews, opinions, etc., given by users to communication service providers.

[1014] "Filtering" is the process of removing unnecessary data to extract only useful data.

[1015] "Text preprocessing" refers to the process of preparing data for later analysis, such as tokenizing and normalizing comment data.

[1016] "Classifying into positive, negative, and neutral" is the process of classifying emotions and evaluations into three categories based on the content of comments.

[1017] "Net Promoter Score" is an index that scores the likelihood that users will recommend a service to others.

[1018] "Distribution of positive and negative comments" is data that shows the percentage of positive and negative comments for a particular period or career.

[1019] "Comparative information" is information for comparing the ratings and performance of different communication service providers.

[1020] A "report including predicted change results" is a report that predicts NPS scores and evaluation data that change over time and summarizes the results.

[1021] A "user device" is a device such as a computer, smartphone, or tablet that a user uses to receive and display reports.

[1022] The present invention relates to a system that collects evaluation comments about telecommunications service providers from the Internet and generates analysis and result reports using AI and an emotion engine. This system automates processes from data collection to emotion analysis of comments, report generation, and distribution. Furthermore, by combining it with an emotion engine that can analyze user emotions in real time, highly accurate Net Promoter Scoring is achieved. The following describes in detail an embodiment of the present invention.

[1023] Data collection from the internet

[1024] The server accesses designated websites and social networks (e.g., social networking sites, forums, review sites) to periodically collect comments about telecommunications service providers. Using scraping tools (e.g., the BeautifulSoup library) or APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, user name, and the name of the applicable carrier.

[1025] Data Preprocessing

[1026] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization and normalization is performed. This prepares the text data in a form suitable for analysis.

[1027] Classification and scoring of NPS comments

[1028] The preprocessed comments are sent to an AI sentiment analysis module and sentiment engine. The server uses the sentiment engine to analyze user sentiment and categorize it as positive, negative, or neutral. It then calculates an NPS score based on the content and sentiment data of each comment, which then generates an overall NPS score for each telecommunications service provider.

[1029] Report Generation

[1030] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each telecommunications service provider. It also generates comparative information with other telecommunications service providers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[1031] Report distribution and visualization

[1032] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to this dashboard and check the latest reports. They can also use the filter and search functions to check detailed information for specific periods or by communication service provider.

[1033] As a concrete example of a prompt sentence, the following input is given to the generative AI model:

[1034] "Analyze the sentiment of NPS comments using the data below and generate a report.

[1035] Example data:

[1036] Comment: "This service is absolutely amazing!"

[1037] Timestamp: "2023-10-13 12:00:00"

[1038] Source URL: "https: / / example.com / review / 12345"

[1039] Username: "user123"

[1040] Applicable carrier name: "Carrier A"

[1041] This concludes the details of the embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, making it possible to accurately grasp user sentiment and efficiently calculate NPS scores.

[1042] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1043] Processing Steps:

[1044] Step 1: Collecting data from the internet

[1045] Processing performed by the server

[1046] Input: The URL or API endpoint of a specified website or social network (social network, forum, review site, etc.).

[1047] Specific operation: Uses Python's BeautifulSoup library to scrape and retrieve comments containing specified keywords (e.g., "Telecommunications Service Provider A Review"). If necessary, uses the social media API to collect data in real time.

[1048] Output: Collected comment data (comment text, timestamp, source URL, username, carrier name).

[1049] Step 2: Preprocessing the data

[1050] Processing performed by the server

[1051] Input: Collected comment data.

[1052] What it does: It uses regular expressions to filter out unwanted data like spam and advertising links, then uses the NLTK library to tokenize and normalize the comment text.

[1053] Output: Filtered and preprocessed comment data.

[1054] Step 3: Categorize and score NPS comments

[1055] Processing performed by the server

[1056] Input: Preprocessed comment data.

[1057] What it does: It uses HuggingFace's Transformers library to run a sentiment analysis model to classify comments as positive, negative, or neutral, and then calculates an NPS score based on the classification (8 for positive, 2 for negative, 5 for neutral).

[1058] Output: Comment data with sentiment classification results and NPS scores.

[1059] Step 4: Generate a report

[1060] Processing performed by the server

[1061] Input: Comment data with sentiment classification results and NPS scores.

[1062] What it does: Aggregates the distribution of positive, negative, and neutral comments for each telecommunications service provider, calculates an overall NPS score, performs comparative analysis with other telecommunications service providers, and automatically generates a comprehensive report with predicted results of important fluctuations.

[1063] Output: An automatically generated report containing the aggregated and analyzed results.

[1064] Step 5: Report distribution and visualization

[1065] Processing performed by the user terminal

[1066] Input: Report delivered from the server.

[1067] Specific operation: The report is reflected in a dashboard that displays the report in real time on the user's device. On the dashboard, users can search and filter for detailed information for specific periods or by communication service provider.

[1068] Output: The latest report and details displayed to the user.

[1069] This flow of processing steps automates the entire system, allowing users to efficiently check the analysis results of NPS comments.

[1070] (Application example 2)

[1071] 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."

[1072] Currently, systems for collecting and analyzing feedback on the efficiency and performance of factory robots are inadequate, preventing effective use of this information. This makes it difficult to quickly and accurately develop specific measures to improve factory operational efficiency and robot performance. Furthermore, when the process from collecting feedback to analyzing, generating, and distributing reports is carried out manually, it results in wasted human resources and makes it difficult to reflect information in a timely manner.

[1073] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for collecting comments about objects from websites on the Internet; means for filtering and text preprocessing the collected comments; means for classifying the comments into positive, negative, and neutral based on the preprocessed comments and calculating a performance score; means for calculating the distribution of positive and negative comments for each object and generating comparative information with other objects; means for automatically generating a report including the generated information and predicted fluctuations; means for delivering and displaying the automatically generated report to a user terminal; means for a user to view the report in real time via a dashboard and check detailed information; and means for using AI and an emotion engine to collect and analyze feedback comments about the efficiency and performance of objects. This makes it possible to quickly and automatically collect and analyze feedback about the efficiency and performance of factory robots and generate and distribute highly accurate performance reports.

[1074] 1. "Internet website" refers to an information site on the Internet that is accessible to users.

[1075] 2. "Object" refers to the factory robot or other equipment that is the subject of evaluation in this invention.

[1076] 3. "Comments" refers to the written opinions and feedback of users regarding the subject matter.

[1077] 4. "Collection methods" refers to the technologies and methods used to extract comments from websites on the Internet.

[1078] 5. "Filtering" refers to the process of removing unnecessary information and noise from acquired comment data.

[1079] 6. "Text preprocessing" refers to procedures that transform comment text into a form suitable for analysis, such as tokenization and normalization.

[1080] 7. "Positive" refers to comments that have a favorable opinion or reputation about the subject.

[1081] 8. "Negative" refers to comments that have a negative opinion or reputation about the subject.

[1082] 9. "Neutral" refers to comments that have a neutral opinion or do not evoke any particular strong emotion toward the subject.

[1083] 10. "Performance Score" refers to a numerical rating that indicates the efficiency or performance of an object.

[1084] 11. "Means for calculating distribution" refers to the techniques and methods for calculating the ratio of positive and negative comments from the collected comment data.

[1085] 12. "Comparative information" refers to data and statistics that allow the evaluation of an object to be compared with other objects.

[1086] 13. "Means for automatically generating reports" refers to technologies and methods for periodically generating reports based on analysis results.

[1087] 14. "User terminal" refers to a device used to receive and display reports, such as a personal computer or smartphone.

[1088] 15. "Means of delivery" refers to the technology or method for sending automatically generated reports to user terminals.

[1089] 16. "Means of display" refers to the technology or method for displaying the delivered report in an easy-to-read format on the user's device.

[1090] 17. "Dashboard" refers to an interface for aggregating and visually displaying various data and reports.

[1091] 18. "Real-time viewing means" refers to technologies and methods that allow users to instantly view the latest information and reports.

[1092] 19. "Means for checking detailed information" refers to technologies and methods that allow users to search and view detailed data for specific periods or subjects.

[1093] 20. "Feedback comments" refers to user evaluations and opinions regarding an object.

[1094] 21. “AI” refers to artificial intelligence technologies, including machine learning models and data analysis algorithms.

[1095] 22. “Sentiment Engine” refers to technology or software used to analyze comment data to identify user sentiment.

[1096] This invention relates to a system that collects feedback on the efficiency and performance of factory robots from websites on the internet, analyzes the feedback, and generates and distributes reports. The system uses AI and an emotion engine to analyze the sentiment of comments and calculate a performance score.

[1097] Program generation and processing steps

[1098] Server-side processing

[1099] 1. Data Collection:

[1100] The server uses web scraping tools such as Scrapy to collect comments about factory robots from social media, review sites, and forums on the Internet, such as comments containing the hashtag "factory robot performance."

[1101] 2. Data preprocessing:

[1102] Collected comments are filtered, tokenized, and normalized using Python, which includes removing spam and ads, and eliminating unnecessary data.

[1103] 3. Sentiment Analysis and Scoring:

[1104] The preprocessed comments are fed into a sentiment analysis model powered by TensorFlow / Keras, where they are classified as positive, negative, or neutral and a performance score is calculated.

[1105] 4. Report Generation:

[1106] Based on the aggregated data, a report is automatically generated using pandas and matplotlib, including the distribution of positive and negative comments, as well as a comparison with other robots.

[1107] 5. Report Delivery:

[1108] The reports are delivered to the user's device and displayed on a dashboard in real time.

[1109] Processing on the user terminal side

[1110] 1. Dashboard view:

[1111] The delivered reports are displayed in real time on the user's device using JavaScript (especially React.js), and users can view detailed information for specific periods or for each factory robot through this dashboard.

[1112] Examples of concrete examples and prompts

[1113] Examples:

[1114] When a user accesses the dashboard, the "Performance Report for Factory Robot A this Month" will show "Performance score increased by 45% and positive comments increased by 60%." Users can also use the filter function to specify a specific period and view performance data for that period in more detail.

[1115] Example prompt sentence:

[1116] The prompt sentence is an example of input to the generative AI model, and looks like this:

[1117] Generate a monthly report for a factory robot performance feedback system. Separate positive, negative, and neutral comments, calculate a score for each, and output the report.

[1118] This completes the description of an embodiment of the present invention, which enables the rapid and automatic collection and analysis of feedback on the efficiency and performance of factory robots, and the generation and distribution of highly accurate performance reports.

[1119] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1120] Step 1: Data collection

[1121] The server uses a web scraping tool such as Scrapy to collect comments about factory robots from social media, review sites, and forums on the Internet. For example, it collects comments containing the hashtag "factory robot performance." The input is the comment data obtained from social media and review sites, and the output is a dataset containing these comments.

[1122] Step 2: Filtering the data

[1123] The server filters the collected comment data. It uses regular expressions in Python to remove unnecessary data, such as spam and advertisements. Specifically, it detects comments containing "advertisements" or "spam links" and does not send them to the preprocessing step. The input is the initial comment data, and the output is the cleaned comment data.

[1124] Step 3: Preprocessing the text

[1125] The server preprocesses the filtered comments, performing tokenization and normalization to make the text suitable for analysis. Specifically, it splits sentences into tokens and removes unnecessary symbols and special characters. For example, it converts a comment such as "This factory robot is great" into "This factory robot is great." The input is filtered comment data, and the output is preprocessed text data.

[1126] Step 4: Sentiment analysis

[1127] The preprocessed comments are sent to a sentiment analysis model (generative AI model) using TensorFlow / Keras. The server performs data calculations to classify comments into positive, negative, or neutral. Specifically, the model predicts the sentiment of each comment and classifies it as a category. For example, a comment saying "great" is classified as positive. The input is the preprocessed text data, and the output is the sentiment category (positive, negative, neutral) and its confidence score.

[1128] Step 5: Calculating the Performance Score

[1129] The server calculates a performance score based on the results of the sentiment analysis. It aggregates the percentage of positive and negative comments and converts them into a score. Specifically, it calculates the performance score from the ratio of positive comments and generates an overall score from, for example, "60% positive, 40% negative." The input is sentiment category data, and the output is a performance score.

[1130] Step 6: Generate reports

[1131] The server automatically generates a report using pandas and matplotlib based on the aggregated data. This report includes the distribution of positive and negative comments and comparison information with other robots. Specifically, it creates a graph of the progress of scores and distribution, and compiles it into a report document. The input is the aggregated data, and the output is a report document.

[1132] Step 7: Report Delivery

[1133] The server delivers the generated report to the user's device. The report is automatically sent via email or a cloud platform. Specifically, the report file is delivered to the user using an email sending API or cloud storage service. The input is the report document, and the output is a notification that the report has been sent.

[1134] Step 8: View the dashboard

[1135] On the user's device, JavaScript (particularly React.js) is used to display the delivered reports in real time on a dashboard. Through this dashboard, users can check detailed information for a specific period or for each factory robot. Specifically, specific data can be displayed using date filters and search functions. The input is the report data, and the output is a visualized report display.

[1136] These are the specific processing steps of the system program that realizes this application example, which allows feedback on the efficiency and performance of factory robots to be collected and analyzed sequentially, with the results immediately available.

[1137] 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.

[1138] 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.

[1139] 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.

[1140] [Fourth embodiment]

[1141] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1142] 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.

[1143] 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).

[1144] 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.

[1145] 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.

[1146] 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).

[1147] 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.

[1148] 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.

[1149] 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.

[1150] 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.

[1151] 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.

[1152] 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.

[1153] 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."

[1154] The present invention relates to a system that collects NPS comments about mobile network service providers from the Internet and generates analysis and result reports using AI. This system automates processes from data collection to sentiment analysis of comments, report generation, and distribution. Detailed descriptions of embodiments of the present invention are provided below.

[1155] Data collection from the internet

[1156] Processing performed by the server

[1157] The server periodically accesses designated websites (e.g., social networking sites, forums, review sites) on the Internet to collect comments about mobile network service providers. The server uses scraping tools and APIs to obtain text information and the following metadata for each comment:

[1158] timestamp

[1159] Source URL

[1160] Username (if available)

[1161] Applicable carrier name

[1162] As a specific example, the server uses a social media API to collect tweets containing hashtags such as "Career A Review" and store them in a database.

[1163] Data Preprocessing

[1164] Processing performed by the server

[1165] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization, normalization, and stop word removal is performed. This prepares the text data in a form suitable for analysis.

[1166] For example, a comment such as "This service is really great! Ad" is tokenized as "This service is really great" and unnecessary parts are removed.

[1167] Classification and scoring of NPS comments

[1168] Processing performed by the AI ​​module (on the server)

[1169] The pre-processed comments are sent to an AI sentiment analysis module. The server uses AI to analyze the sentiment of each comment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content of each comment. This also generates an overall NPS score for each carrier.

[1170] For example, the server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and sets the NPS score to 8.

[1171] Report Generation

[1172] Processing performed by the server

[1173] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each carrier. It also generates comparative information with other carriers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[1174] As a concrete example, the server generates a report containing the information, "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan."

[1175] Report distribution and visualization

[1176] Processing performed by the user terminal

[1177] The generated reports are displayed in real time on a dashboard on the user's device. Users can log in to the dashboard to view the latest reports. Furthermore, they can use the filter and search functions to view detailed information for specific periods or by carrier.

[1178] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as reference for their next marketing initiative.

[1179] The above is a detailed description of the embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, enabling efficient use of human resources and rapid decision-making.

[1180] The processing flow will be explained below.

[1181] Step 1:

[1182] Processing performed by the server

[1183] The server periodically accesses designated websites on the Internet to collect comments about mobile network service providers. Using scraping tools and public APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, username (if available), and applicable carrier name.

[1184] Step 2:

[1185] Processing performed by the server

[1186] The collected comments are filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing is performed, including tokenization (splitting sentences into words), normalization (converting all to lowercase, removing special characters, etc.), and removal of stop words. This prepares the text data in a form suitable for analysis.

[1187] Step 3:

[1188] Processing performed by the server

[1189] The preprocessed comments are sent to the AI ​​sentiment analysis module, which uses an AI model to analyze the sentiment of each comment and classify it as positive, negative, or neutral. It then calculates an NPS score based on the content of each comment and stores it in a database.

[1190] Step 4:

[1191] Processing performed by the server

[1192] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, and generates comparative information with other carriers, including calculating each carrier's overall NPS score and analyzing trends over time.

[1193] Step 5:

[1194] Processing performed by the server

[1195] Based on the aggregated data and analysis results, the server automatically generates a report that includes the overall NPS score, the distribution of positive and negative comments, comparisons with other carriers, and predictions of future changes.

[1196] Step 6:

[1197] Processing performed by the user terminal

[1198] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to the dashboard to check the latest reports. They can also use the filter and search functions to view detailed information for specific periods or by carrier.

[1199] Step 7:

[1200] User actions

[1201] Planning and marketing staff conduct analysis based on the reports generated on the dashboard. This allows them to quickly evaluate the effectiveness of their initiatives and quickly consider and implement any necessary improvements. For example, they can evaluate the effectiveness of a new pricing plan based on information such as "Carrier A's NPS score has improved by 10 points since last month, and the number of positive comments has increased."

[1202] Example 1

[1203] 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."

[1204] It is inefficient to manually collect appropriate feedback in real time from a huge amount of user comments about mobile network service providers, perform sentiment analysis, and automatically generate comprehensive reports including competitive comparisons. There is a need for a system that solves this problem and automates and streamlines the process from data collection to report creation.

[1205] 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.

[1206] In this invention, the server includes a means for collecting comments about mobile network service providers from data sources on the Internet, a means for filtering the collected comments using regular expressions to remove unnecessary data such as spam and advertising links, and a text preprocessing means for tokenizing, normalizing, and removing stop words from the filtered comments. This enables useful information to be extracted from large amounts of data, enabling rapid and accurate sentiment analysis and NPS score evaluation. The server also includes a means for delivering automatically generated reports to user terminals and displaying them in real time, allowing users to make quick decisions based on the latest information.

[1207] "Data Source" refers to any place where information can be collected, such as any website, social media site, forum, review site, or other website accessible on the Internet.

[1208] "Comment" means text data such as feedback, opinions, reviews, etc. posted by a user to a mobile network service provider.

[1209] "Filtering" refers to the process of removing unnecessary information such as spam and advertising links from collected comment data and formatting it into data suitable for analysis.

[1210] "Tokenization" refers to the process of dividing text into the smallest meaningful units, such as words or phrases.

[1211] "Normalization" refers to the process of converting text data into a unified format to improve the accuracy of analysis.

[1212] "Stop words" refer to common words (e.g., wa, ga, no, etc.) that are of little importance and have no meaning in text analysis.

[1213] "Sentiment analysis" refers to the process of analyzing text data and identifying its emotional tendencies (positive, negative, neutral).

[1214] The Net Promoter Score (NPS) is an index that evaluates customer satisfaction and the degree of recommendation, and is scored on a scale of 0 to 10.

[1215] "Positive comments" refer to comments that contain positive or favorable feedback about a service or product.

[1216] "Negative Comments" refers to comments that contain negative or unsatisfactory feedback about a service or product.

[1217] "Report" means a document generated from collected and analyzed data that includes results, trends, comparative information, etc.

[1218] "Dashboard" refers to a web application or software that includes an interface that allows users to view, search, and analyze information in real time.

[1219] This invention relates to a system that collects comments about mobile network service providers from the Internet and uses AI to analyze those comments and generate a report of the results. This system automates the process from data collection to report generation and distribution through the following series of processes.

[1220] Data collection

[1221] Processing performed by the server

[1222] The server periodically accesses designated websites (e.g., social networking sites, forums, review sites, etc.) on the Internet. Specifically, the server uses scraping tools and APIs (e.g., Twitter API) to collect targeted comments. The collected information includes text, timestamp, source URL, username (if available), and corresponding carrier name.

[1223] As a specific example, the server uses the Twitter API to collect tweets containing the hashtag "Career A Review" and store them in a database.

[1224] Data Preprocessing

[1225] Processing performed by the server

[1226] The collected comment data is first filtered using regular expressions to remove unwanted data such as spam and advertising links, then the comment text is tokenized, normalized, and stop words are removed.

[1227] For example, a comment like "This service is really great! Ad" would be tokenized as "This service is really great" and the spam part would be removed.

[1228] Classification and scoring of NPS comments

[1229] Processing performed by the server

[1230] The pre-processed comments are sent to an AI sentiment analysis module, which analyzes the sentiment of each comment and categorizes it as positive, negative, or neutral. It then calculates an NPS score based on the content of each comment and generates an overall NPS score for each carrier.

[1231] As a concrete example, the server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and sets the NPS score to 8.

[1232] Report Generation

[1233] Processing performed by the server

[1234] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, generates comparison information with other carriers, and automatically generates a comprehensive report with forecast results for important fluctuations.

[1235] As a specific example, the server compiles information such as "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan" and includes it in a report.

[1236] Report distribution and visualization

[1237] Processing performed by the user terminal

[1238] The generated reports are displayed in real time on a dashboard on the user's device, and users can log in to view the latest reports and use filters and search functions to view detailed information for specific periods or by carrier.

[1239] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as a reference for their next marketing initiative.

[1240] Examples of prompt statements

[1241] Examples of input prompts for generative AI models include:

[1242] "Please explain the process flow of a system that collects NPS comments about mobile network service providers, performs sentiment analysis and scoring using AI, and compiles the results into a report, including the specific processing steps."

[1243] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1244] Step 1: Data collection

[1245] The server periodically accesses designated data sources (e.g., social media sites, forums, review sites) on the Internet. The server uses scraping tools or social media APIs (e.g., Twitter API) to collect comments about mobile network service providers. As input, the server receives target comments and hashtags (e.g., "Carrier A Review"). As output, the server stores data in a database, including text, timestamp, source URL, username (if available), and the name of the corresponding carrier.

[1246] Specific operation: The server uses the Twitter API to collect tweets containing the hashtag "Career A Review" and stores that data in a database.

[1247] Step 2: Filtering the data

[1248] The server uses regular expressions to filter the collected comment data to remove unwanted data such as spam and advertising links. It receives the collected raw data as input and generates clean filtered data as output.

[1249] Specific operation: The comment "This service is really great! Ad" is converted into clean data using regular expressions to remove unnecessary parts such as "This service is really great."

[1250] Step 3: Tokenize and normalize the data

[1251] The filtered comments are then tokenized and normalized. The server receives the filtered text data as input, tokenizes it (splits the text into words and phrases), normalizes it (converts the text into a uniform format), and removes stop words. As output, it produces tokenized data suitable for analysis.

[1252] Specific behavior: Split the text "This service is really great!" into "This service is really great", removing spelling variations and extra words.

[1253] Step 4: Sentiment analysis and NPS score calculation

[1254] The preprocessed comments are sent to the AI ​​sentiment analysis module. The server receives the tokenized text data as input and performs sentiment analysis using an AI model. It classifies each comment as positive, negative, or neutral and calculates an NPS score based on its content. As output, it generates data containing the sentiment classification result and the NPS score.

[1255] Specific behavior: The server sends the comment "This service is really great" to the AI ​​sentiment analysis module, which determines it as positive and calculates an NPS score of 8.

[1256] Step 5: Generate a report

[1257] The server periodically aggregates NPS data, calculates the distribution of positive and negative comments for each carrier, and automatically generates a comprehensive report including comparisons with other carriers and predictions of significant fluctuations. It receives the results of sentiment analysis and NPS scores as input and generates a comprehensive report as output.

[1258] Specific operation: Based on the information that "Carrier A's NPS score this month is 72, an improvement of 10 points from last month," the server compiles a report stating that 50% of users have given positive feedback about the new pricing plan.

[1259] Step 6: Report distribution and visualization

[1260] The user terminal displays the generated reports on a dashboard in real time. Users can log in and view the latest reports. They can also use filters and search functions to check detailed information for specific periods or by carrier. The system receives reports provided by the server as input and displays them on the dashboard as output.

[1261] Specific actions: The planner accesses the dashboard and checks the information that "Career A's NPS score this month is 76, and positive comments have increased by 50%," and uses this information as a reference for the next marketing initiative.

[1262] (Application example 1)

[1263] 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."

[1264] Content distribution services are required to collect and analyze viewer feedback in real time, and to efficiently recommend personalized content and propose improvements based on that feedback. Conventional methods have had the problem of being unable to provide appropriate content due to the difficulty of quickly reflecting viewer reactions.

[1265] 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.

[1266] In this invention, the server includes means for collecting comments about a mobile network service provider from websites on the Internet, means for filtering and text preprocessing the collected comments, means for classifying the preprocessed comments into positive, negative, and neutral categories and calculating a Net Promoter Score, means for calculating the distribution of positive and negative comments for each provider and generating comparative information with other providers, means for automatically generating a report including the generated information and predicted fluctuations, means for delivering and displaying the automatically generated report to a user terminal, and means for aggregating viewer responses in real time on the user terminal and recommending content and proposing improvements based on viewing trends, thereby enabling fast and efficient collection and analysis of viewer feedback and the provision of personalized content.

[1267] The "Internet" is a communications infrastructure that connects computer networks around the world and enables the exchange of information.

[1268] A "website" is a collection of information made up of multiple web pages published on the Internet.

[1269] A "mobile network service provider" is a company or organization that provides communications services to mobile phones and other mobile communication devices.

[1270] A "comment" is text information in which a user expresses an opinion or comment on a particular topic.

[1271] "Filtering" is the process of removing unnecessary information from collected data.

[1272] "Text preprocessing" refers to a series of processes used to format text data into a form suitable for analysis.

[1273] "Positive" refers to comments that express positive or favorable sentiment.

[1274] "Negative" refers to comments that express negative or unfavorable sentiment.

[1275] "Neutral" refers to comments that express a neutral sentiment, neither positive nor negative.

[1276] "Net Promoter Score" is a quantified indicator of the likelihood that a customer will recommend a service to others.

[1277] "Provider" refers to a company or organization that provides a particular service or content.

[1278] "Distribution" refers to the frequency or proportion of occurrence in each category of data.

[1279] "Comparative information" is information that shows the differences and similarities between two or more data sets.

[1280] A "report" is a document that summarizes analytical results and statistical data.

[1281] A "user terminal" is a device that a user uses to access the Internet or a system.

[1282] "Real-time" refers to the property that data and events are processed as they occur.

[1283] A "viewer" is a user who views video and audio content.

[1284] "Reactions" refer to the emotions and evaluations that viewers express toward content.

[1285] "Aggregation" is the process of statistically summarizing data.

[1286] "Viewing habits" refers to a viewer's general behavior and preference patterns regarding content.

[1287] "Content" refers to the media that viewers consume, such as television programs, movies, music, articles, etc.

[1288] A "recommendation" is the presentation of particular content to a viewer.

[1289] "Suggestions for improvement" refers to suggestions for improving the quality of content based on viewer reactions and feedback.

[1290] This invention provides a system that collects and analyzes viewer feedback in real time, and recommends personalized content and suggests improvements. Specific implementations of this system are described below.

[1291] Data collection method

[1292] The server collects viewers' comments from specific websites on the Internet (social networking sites, review sites, forums, etc.) using scraping tools and APIs to obtain text information and the following metadata for each comment:

[1293] timestamp

[1294] Source URL

[1295] Username (if available)

[1296] For example, the server uses a social media API to collect posts containing hashtags such as "Series X review" and store them in a database.

[1297] Data Preprocessing

[1298] The server filters the collected comment data and performs the following actions:

[1299] Remove unwanted data such as spam and advertising links (using regular expressions)

[1300] Text preprocessing, including tokenization, normalization, and stop word removal

[1301] This allows the comment to be formatted in a way that is suitable for sentiment analysis. For example, a comment like "This movie was great! Ad" is tokenized as "This movie was great" and unnecessary parts are removed.

[1302] Classification and scoring of NPS comments

[1303] The preprocessed comments are sent to an AI sentiment analysis module. The server uses AI (e.g., NLTK's VADER Sentiment Analyzer) to analyze the sentiment of each comment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content of each comment.

[1304] For example, a comment such as "This movie was amazing!" is considered positive and is assigned an NPS score of 9.

[1305] Report Generation and Delivery

[1306] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each provider, as well as comparisons with other providers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[1307] For example, a report might be generated that includes information such as, "Provider A's NPS score this month is 50, up 15 points from last month. 30% of viewers have given positive feedback about new content."

[1308] The generated reports are displayed in real time on a dashboard on the user's device. Users can log in to this dashboard to view the latest reports. Furthermore, they can use the filter and search functions to view detailed information for specific periods or by provider.

[1309] Content recommendations and improvement suggestions

[1310] Viewer responses are collected in real time on the user's device, and content recommendations and suggestions for improvements are made based on viewing trends. For example, it is possible to recommend the next content to watch based on viewer feedback data. In addition, if there is a lot of negative feedback, it is possible to analyze the cause and suggest improvements.

[1311] The specific hardware and software used

[1312] Hardware: Servers, user devices (smartphones, tablets, PCs, etc.)

[1313] Software: Python, BeautifulSoup, requests library, NLTK, VADER Sentiment Analyzer, Matplotlib

[1314] Examples of prompt statements

[1315] Here is an example prompt:

[1316] Analyze the sentiment of this comment: "This movie was amazing! Great story and acting. Advertisement"

[1317] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1318] Step 1:

[1319] The server collects viewer comments from specific websites on the Internet. Using APIs and scraping tools, the server obtains metadata such as comment text information, timestamps, source URLs, and user names. The input is the URL of a specific website, and the output is a list of collected comments and their metadata.

[1320] Step 2:

[1321] The server filters the collected comment data and removes inappropriate data (such as spam or advertising links). It uses regular expressions to detect and remove unnecessary data. The input is the comment data collected in the previous step, and the output is the clean comment data after filtering.

[1322] Step 3:

[1323] The server performs text preprocessing such as tokenization, normalization, and stop word removal. Tokenization splits comments into words, normalization converts uppercase to lowercase, and stop word removal removes common words such as "の" and "は." The input is the filtered comment data, and the output is the preprocessed text data.

[1324] Step 4:

[1325] The server sends the preprocessed comment data to the AI ​​sentiment analysis module for sentiment analysis. It uses NLTK's VADER Sentiment Analyzer to classify each comment's sentiment as positive, negative, or neutral. It then calculates the Net Promoter Score (NPS score) for each comment. The input is the preprocessed text data, and the output is the sentiment classification and NPS score for each comment.

[1326] Step 5:

[1327] The server calculates the distribution of positive and negative comments for each provider and generates comparative information with other providers. It also automatically generates a comprehensive report including predictions of significant fluctuations. The input is the comment data after sentiment analysis and the NPS score, and the output is a report including information for each provider and comparisons.

[1328] Step 6:

[1329] The server delivers the automatically generated reports to the user terminal and displays them. The user terminal displays the reports in real time on a dashboard, allowing the user to search for detailed information for a specific period or by provider. The input is the generated report, and the output is the dashboard display on the user terminal.

[1330] Step 7:

[1331] The user terminal aggregates viewer responses in real time and makes content recommendations and suggestions for improvements based on viewing trends. The input is viewer feedback data, and the output is content recommendations and suggestions for improvements to the viewer.

[1332] 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.

[1333] The present invention relates to a system that collects NPS comments about mobile network service providers from the Internet and generates analysis and result reports using AI and an emotion engine. This system automates processes from data collection to emotion analysis of comments, report generation, and distribution. Furthermore, by combining this system with an emotion engine that can analyze user emotions in real time, highly accurate NPS scoring is achieved. Detailed descriptions of embodiments of the present invention are provided below.

[1334] Data collection from the internet

[1335] Processing performed by the server

[1336] The server periodically accesses designated websites (social media sites, forums, review sites, etc.) on the Internet to collect comments about mobile network service providers. Using scraping tools and APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, username (if available), and applicable carrier name.

[1337] As a specific example, the server uses a social media API to collect tweets containing hashtags such as "Career A Review" and store them in a database.

[1338] Data Preprocessing

[1339] Processing performed by the server

[1340] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization and normalization is performed. This prepares the text data in a form suitable for analysis.

[1341] For example, a comment such as "This service is really great! Ad" is tokenized as "This service is really great" and unnecessary parts are removed.

[1342] Classification and scoring of NPS comments

[1343] Processing performed by the AI ​​module and emotion engine (on the server)

[1344] The preprocessed comments are sent to an AI sentiment analysis module and sentiment engine. The server uses the sentiment engine to analyze user sentiment and classify it as positive, negative, or neutral. An NPS score is then calculated based on the content and sentiment data of each comment, which also generates an overall NPS score for each carrier.

[1345] For example, the server might send a comment like "This service is really great" to the sentiment engine, which would then consider it positive and set the NPS score to 8.

[1346] Report Generation

[1347] Processing performed by the server

[1348] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each carrier. It also generates comparative information with other carriers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[1349] As a concrete example, the server generates a report containing the information, "Carrier A's NPS score this month is 72, a 10-point improvement from last month. 50% of users have given positive feedback about the new pricing plan."

[1350] Report distribution and visualization

[1351] Processing performed by the user terminal

[1352] The user device displays reports delivered from the server on a dashboard in real time. Users can log in to this dashboard to check the latest reports. They can also use the filter and search functions to check detailed information for specific periods or by carrier.

[1353] As a specific example, a planning manager can access the dashboard and check information such as "Career A's NPS score this month is 76, and positive comments have increased by 50%," and use this information as reference for their next marketing initiative.

[1354] The above is a detailed description of an embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, enabling efficient use of human resources and rapid decision-making. Furthermore, the use of an emotion engine makes it possible to more accurately grasp user emotions and achieve highly accurate NPS scoring.

[1355] The processing flow will be explained below.

[1356] Step 1:

[1357] Processing performed by the server

[1358] The server periodically accesses designated websites on the Internet and uses scraping tools or public APIs to collect comments about mobile network service providers, along with metadata for each comment (timestamp, source URL, user name, and corresponding carrier name), which are then stored in a database.

[1359] Step 2:

[1360] Processing performed by the server

[1361] The collected comment data is filtered, using regular expressions to remove unwanted data such as spam and advertising links, and the comments are tokenized and subjected to text preprocessing, which includes splitting sentences into words, normalization, and stop word removal.

[1362] Step 3:

[1363] Processing performed by the server

[1364] The preprocessed comments are sent to the sentiment engine and AI sentiment analysis module. The sentiment engine analyzes the user sentiment of each comment and classifies it as positive, negative, or neutral. The AI ​​sentiment analysis module calculates an NPS score for each comment based on the sentiment data, which then derives an overall NPS score for each carrier.

[1365] Step 4:

[1366] Processing performed by the server

[1367] The server periodically compiles NPS data, calculates the distribution of positive and negative comments for each carrier, and generates comparative information with other carriers, including calculating each carrier's overall NPS score and analyzing trends over time.

[1368] Step 5:

[1369] Processing performed by the server

[1370] Based on the aggregated data and analysis results, the server automatically generates a report that includes the overall NPS score, the distribution of positive and negative comments, comparison information with other carriers, and a forecast of future fluctuations, allowing users to grasp the necessary information at a glance.

[1371] Step 6:

[1372] Processing performed by the user terminal

[1373] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to the dashboard to check the latest reports. They can also use the filter and search functions to view detailed information for specific periods or by carrier.

[1374] Step 7:

[1375] User actions

[1376] Planning and marketing staff conduct analysis based on the reports generated on the dashboard. This allows them to quickly evaluate the effectiveness of their initiatives and quickly consider and implement any necessary improvements. For example, they can evaluate the effectiveness of a new pricing plan based on information such as "Carrier A's NPS score has improved by 10 points since last month, and the number of positive comments has increased."

[1377] Example 2

[1378] 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."

[1379] The vast amount of comment data available online varies in content and format, making it difficult to manually collect and analyze, and therefore difficult to efficiently calculate NPS scores. Furthermore, accurate understanding of user sentiment and reflecting it in marketing and service improvements requires advanced analysis. There is a need for a system that can solve these problems, achieve highly accurate NPS scoring, and generate reports quickly.

[1380] 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.

[1381] In this invention, the server includes means for collecting evaluation comments about a communication service provider from Internet sources, including designated websites and social networks, means for filtering and text preprocessing the collected comments, means for classifying the preprocessed comments into positive, negative, and neutral categories and calculating a Net Promoter Score, means for calculating the distribution of positive and negative comments for each communication service provider and generating comparative information with other communication service providers, means for automatically generating a report including the generated information and predicted fluctuations, and means for delivering and displaying the automatically generated report to a user device. This allows for efficient calculation of NPS scores while accurately grasping user sentiment, and enables rapid report generation and delivery.

[1382] "Designated websites and social networks" are pre-selected internet websites, social media platforms, and other online platforms used to collect information.

[1383] A "communications service provider" is a company or organization that provides communications services such as mobile communications and internet access.

[1384] "Evaluation comments" are written evaluations such as feedback, reviews, opinions, etc., given by users to communication service providers.

[1385] "Filtering" is the process of removing unnecessary data to extract only useful data.

[1386] "Text preprocessing" refers to the process of preparing data for later analysis, such as tokenizing and normalizing comment data.

[1387] "Classifying into positive, negative, and neutral" is the process of classifying emotions and evaluations into three categories based on the content of comments.

[1388] "Net Promoter Score" is an index that scores the likelihood that users will recommend a service to others.

[1389] "Distribution of positive and negative comments" is data that shows the percentage of positive and negative comments for a particular period or career.

[1390] "Comparative information" is information for comparing the ratings and performance of different communication service providers.

[1391] A "report including predicted change results" is a report that predicts NPS scores and evaluation data that change over time and summarizes the results.

[1392] A "user device" is a device such as a computer, smartphone, or tablet that a user uses to receive and display reports.

[1393] The present invention relates to a system that collects evaluation comments about telecommunications service providers from the Internet and generates analysis and result reports using AI and an emotion engine. This system automates processes from data collection to emotion analysis of comments, report generation, and distribution. Furthermore, by combining it with an emotion engine that can analyze user emotions in real time, highly accurate Net Promoter Scoring is achieved. The following describes in detail an embodiment of the present invention.

[1394] Data collection from the internet

[1395] The server accesses designated websites and social networks (e.g., social networking sites, forums, review sites) to periodically collect comments about telecommunications service providers. Using scraping tools (e.g., the BeautifulSoup library) or APIs, the server extracts the comments section of each page and stores it along with the following metadata: timestamp, source URL, user name, and the name of the applicable carrier.

[1396] Data Preprocessing

[1397] The collected comment data is first filtered. The server uses regular expressions to remove unnecessary data such as spam and advertising links. Next, text preprocessing such as tokenization and normalization is performed. This prepares the text data in a form suitable for analysis.

[1398] Classification and scoring of NPS comments

[1399] The preprocessed comments are sent to an AI sentiment analysis module and sentiment engine. The server uses the sentiment engine to analyze user sentiment and categorize it as positive, negative, or neutral. It then calculates an NPS score based on the content and sentiment data of each comment, which then generates an overall NPS score for each telecommunications service provider.

[1400] Report Generation

[1401] The server periodically compiles NPS data and calculates the distribution of positive and negative comments for each telecommunications service provider. It also generates comparative information with other telecommunications service providers. Based on this information, a comprehensive report is automatically generated, including predictions of significant fluctuations.

[1402] Report distribution and visualization

[1403] The user terminal displays the reports delivered from the server on a dashboard in real time. Users can log in to this dashboard and check the latest reports. They can also use the filter and search functions to check detailed information for specific periods or by communication service provider.

[1404] As a concrete example of a prompt sentence, the following input is given to the generative AI model:

[1405] "Analyze the sentiment of NPS comments using the data below and generate a report.

[1406] Example data:

[1407] Comment: "This service is absolutely amazing!"

[1408] Timestamp: "2023-10-13 12:00:00"

[1409] Source URL: "https: / / example.com / review / 12345"

[1410] Username: "user123"

[1411] Applicable carrier name: "Carrier A"

[1412] This concludes the details of the embodiment of the present invention. This system automates the process from collecting NPS comments to analyzing them, generating reports, and distributing them, making it possible to accurately grasp user sentiment and efficiently calculate NPS scores.

[1413] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1414] Processing Steps:

[1415] Step 1: Collecting data from the internet

[1416] Processing performed by the server

[1417] Input: The URL or API endpoint of a specified website or social network (social network, forum, review site, etc.).

[1418] Specific operation: Uses Python's BeautifulSoup library to scrape and retrieve comments containing specified keywords (e.g., "Telecommunications Service Provider A Review"). If necessary, uses the social media API to collect data in real time.

[1419] Output: Collected comment data (comment text, timestamp, source URL, username, carrier name).

[1420] Step 2: Preprocessing the data

[1421] Processing performed by the server

[1422] Input: Collected comment data.

[1423] What it does: It uses regular expressions to filter out unwanted data like spam and advertising links, then uses the NLTK library to tokenize and normalize the comment text.

[1424] Output: Filtered and preprocessed comment data.

[1425] Step 3: Categorize and score NPS comments

[1426] Processing performed by the server

[1427] Input: Preprocessed comment data.

[1428] What it does: It uses HuggingFace's Transformers library to run a sentiment analysis model to classify comments as positive, negative, or neutral, and then calculates an NPS score based on the classification (8 for positive, 2 for negative, 5 for neutral).

[1429] Output: Comment data with sentiment classification results and NPS scores.

[1430] Step 4: Generate a report

[1431] Processing performed by the server

[1432] Input: Comment data with sentiment classification results and NPS scores.

[1433] What it does: Aggregates the distribution of positive, negative, and neutral comments for each telecommunications service provider, calculates an overall NPS score, performs comparative analysis with other telecommunications service providers, and automatically generates a comprehensive report with predicted results of important fluctuations.

[1434] Output: An automatically generated report containing the aggregated and analyzed results.

[1435] Step 5: Report distribution and visualization

[1436] Processing performed by the user terminal

[1437] Input: Report delivered from the server.

[1438] Specific operation: The report is reflected in a dashboard that displays the report in real time on the user's device. On the dashboard, users can search and filter for detailed information for specific periods or by communication service provider.

[1439] Output: The latest report and details displayed to the user.

[1440] This flow of processing steps automates the entire system, allowing users to efficiently check the analysis results of NPS comments.

[1441] (Application example 2)

[1442] 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."

[1443] Currently, systems for collecting and analyzing feedback on the efficiency and performance of factory robots are inadequate, preventing effective use of this information. This makes it difficult to quickly and accurately develop specific measures to improve factory operational efficiency and robot performance. Furthermore, when the process from collecting feedback to analyzing, generating, and distributing reports is carried out manually, it results in wasted human resources and makes it difficult to reflect information in a timely manner.

[1444] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for collecting comments about objects from websites on the Internet; means for filtering and text preprocessing the collected comments; means for classifying the comments into positive, negative, and neutral based on the preprocessed comments and calculating a performance score; means for calculating the distribution of positive and negative comments for each object and generating comparative information with other objects; means for automatically generating a report including the generated information and predicted fluctuations; means for delivering and displaying the automatically generated report to a user terminal; means for a user to view the report in real time via a dashboard and check detailed information; and means for using AI and an emotion engine to collect and analyze feedback comments about the efficiency and performance of objects. This makes it possible to quickly and automatically collect and analyze feedback about the efficiency and performance of factory robots and generate and distribute highly accurate performance reports.

[1445] 1. "Internet website" refers to an information site on the Internet that is accessible to users.

[1446] 2. "Object" refers to the factory robot or other equipment that is the subject of evaluation in this invention.

[1447] 3. "Comments" refers to the written opinions and feedback of users regarding the subject matter.

[1448] 4. "Collection methods" refers to the technologies and methods used to extract comments from websites on the Internet.

[1449] 5. "Filtering" refers to the process of removing unnecessary information and noise from acquired comment data.

[1450] 6. "Text preprocessing" refers to procedures that transform comment text into a form suitable for analysis, such as tokenization and normalization.

[1451] 7. "Positive" refers to comments that have a favorable opinion or reputation about the subject.

[1452] 8. "Negative" refers to comments that have a negative opinion or reputation about the subject.

[1453] 9. "Neutral" refers to comments that have a neutral opinion or do not evoke any particular strong emotion toward the subject.

[1454] 10. "Performance Score" refers to a numerical rating that indicates the efficiency or performance of an object.

[1455] 11. "Means for calculating distribution" refers to the techniques and methods for calculating the ratio of positive and negative comments from the collected comment data.

[1456] 12. "Comparative information" refers to data and statistics that allow the evaluation of an object to be compared with other objects.

[1457] 13. "Means for automatically generating reports" refers to technologies and methods for periodically generating reports based on analysis results.

[1458] 14. "User terminal" refers to a device used to receive and display reports, such as a personal computer or smartphone.

[1459] 15. "Means of delivery" refers to the technology or method for sending automatically generated reports to user terminals.

[1460] 16. "Means of display" refers to the technology or method for displaying the delivered report in an easy-to-read format on the user's device.

[1461] 17. "Dashboard" refers to an interface for aggregating and visually displaying various data and reports.

[1462] 18. "Real-time viewing means" refers to technologies and methods that allow users to instantly view the latest information and reports.

[1463] 19. "Means for checking detailed information" refers to technologies and methods that allow users to search and view detailed data for specific periods or subjects.

[1464] 20. "Feedback comments" refers to user evaluations and opinions regarding an object.

[1465] 21. “AI” refers to artificial intelligence technologies, including machine learning models and data analysis algorithms.

[1466] 22. “Sentiment Engine” refers to technology or software used to analyze comment data to identify user sentiment.

[1467] This invention relates to a system that collects feedback on the efficiency and performance of factory robots from websites on the internet, analyzes the feedback, and generates and distributes reports. The system uses AI and an emotion engine to analyze the sentiment of comments and calculate a performance score.

[1468] Program generation and processing steps

[1469] Server-side processing

[1470] 1. Data Collection:

[1471] The server uses web scraping tools such as Scrapy to collect comments about factory robots from social media, review sites, and forums on the Internet, such as comments containing the hashtag "factory robot performance."

[1472] 2. Data preprocessing:

[1473] Collected comments are filtered, tokenized, and normalized using Python, which includes removing spam and ads, and eliminating unnecessary data.

[1474] 3. Sentiment Analysis and Scoring:

[1475] The preprocessed comments are fed into a sentiment analysis model powered by TensorFlow / Keras, where they are classified as positive, negative, or neutral and a performance score is calculated.

[1476] 4. Report Generation:

[1477] Based on the aggregated data, a report is automatically generated using pandas and matplotlib, including the distribution of positive and negative comments, as well as a comparison with other robots.

[1478] 5. Report Delivery:

[1479] The reports are delivered to the user's device and displayed on a dashboard in real time.

[1480] Processing on the user terminal side

[1481] 1. Dashboard view:

[1482] The delivered reports are displayed in real time on the user's device using JavaScript (especially React.js), and users can view detailed information for specific periods or for each factory robot through this dashboard.

[1483] Examples of concrete examples and prompts

[1484] Examples:

[1485] When a user accesses the dashboard, the "Performance Report for Factory Robot A this Month" will show "Performance score increased by 45% and positive comments increased by 60%." Users can also use the filter function to specify a specific period and view performance data for that period in more detail.

[1486] Example prompt sentence:

[1487] The prompt sentence is an example of input to the generative AI model, and looks like this:

[1488] Generate a monthly report for a factory robot performance feedback system. Separate positive, negative, and neutral comments, calculate a score for each, and output the report.

[1489] This completes the description of an embodiment of the present invention, which enables the rapid and automatic collection and analysis of feedback on the efficiency and performance of factory robots, and the generation and distribution of highly accurate performance reports.

[1490] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1491] Step 1: Data collection

[1492] The server uses a web scraping tool such as Scrapy to collect comments about factory robots from social media, review sites, and forums on the Internet. For example, it collects comments containing the hashtag "factory robot performance." The input is the comment data obtained from social media and review sites, and the output is a dataset containing these comments.

[1493] Step 2: Filtering the data

[1494] The server filters the collected comment data. It uses regular expressions in Python to remove unnecessary data, such as spam and advertisements. Specifically, it detects comments containing "advertisements" or "spam links" and does not send them to the preprocessing step. The input is the initial comment data, and the output is the cleaned comment data.

[1495] Step 3: Preprocessing the text

[1496] The server preprocesses the filtered comments, performing tokenization and normalization to make the text suitable for analysis. Specifically, it splits sentences into tokens and removes unnecessary symbols and special characters. For example, it converts a comment such as "This factory robot is great" into "This factory robot is great." The input is filtered comment data, and the output is preprocessed text data.

[1497] Step 4: Sentiment analysis

[1498] The preprocessed comments are sent to a sentiment analysis model (generative AI model) using TensorFlow / Keras. The server performs data calculations to classify comments into positive, negative, or neutral. Specifically, the model predicts the sentiment of each comment and classifies it as a category. For example, a comment saying "great" is classified as positive. The input is the preprocessed text data, and the output is the sentiment category (positive, negative, neutral) and its confidence score.

[1499] Step 5: Calculating the Performance Score

[1500] The server calculates a performance score based on the results of the sentiment analysis. It aggregates the percentage of positive and negative comments and converts them into a score. Specifically, it calculates the performance score from the ratio of positive comments and generates an overall score from, for example, "60% positive, 40% negative." The input is sentiment category data, and the output is a performance score.

[1501] Step 6: Generate reports

[1502] The server automatically generates a report using pandas and matplotlib based on the aggregated data. This report includes the distribution of positive and negative comments and comparison information with other robots. Specifically, it creates a graph of the progress of scores and distribution, and compiles it into a report document. The input is the aggregated data, and the output is a report document.

[1503] Step 7: Report Delivery

[1504] The server delivers the generated report to the user's device. The report is automatically sent via email or a cloud platform. Specifically, the report file is delivered to the user using an email sending API or cloud storage service. The input is the report document, and the output is a notification that the report has been sent.

[1505] Step 8: View the dashboard

[1506] On the user's device, JavaScript (particularly React.js) is used to display the delivered reports in real time on a dashboard. Through this dashboard, users can check detailed information for a specific period or for each factory robot. Specifically, specific data can be displayed using date filters and search functions. The input is the report data, and the output is a visualized report display.

[1507] These are the specific processing steps of the system program that realizes this application example, which allows feedback on the efficiency and performance of factory robots to be collected and analyzed sequentially, with the results immediately available.

[1508] 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.

[1509] 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.

[1510] 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.

[1511] 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.

[1512] FIG. 9 is a diagram illustrating 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 actions 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.

[1513] 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.

[1514] 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).

[1515] 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.

[1516] 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."

[1517] 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.

[1518] 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).

[1519] 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.

[1520] 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.

[1521] 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.

[1522] 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.

[1523] 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.

[1524] 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.

[1525] 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.

[1526] 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.

[1527] 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.

[1528] 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.

[1529] The following is further disclosed regarding the above embodiment.

[1530] (Claim 1)

[1531] a means for collecting comments about a mobile network service provider from websites on the Internet;

[1532] a means for filtering and text preprocessing the collected comments;

[1533] A means to classify pre-processed comments as positive, negative, or neutral and calculate a Net Promoter Score;

[1534] A means for calculating the distribution of positive and negative comments for each carrier and generating comparative information with other carriers;

[1535] means for automatically generating a report including the generated information and predicted results of the fluctuations;

[1536] A system including means for delivering and displaying automatically generated reports to a user terminal.

[1537] (Claim 2)

[1538] The system of claim 1, wherein the collected comments are filtered based on specific rules and unnecessary data is removed using regular expressions.

[1539] (Claim 3)

[1540] 10. The system of claim 1, further comprising means for enabling a user to view the generated reports in real time via a dashboard and search for detailed information for a specific period or carrier.

[1541] "Example 1"

[1542] (Claim 1)

[1543] a means for collecting comments about mobile network service providers from data sources on the Internet;

[1544] A means of filtering collected comments using regular expressions to remove unwanted data such as spam and advertising links;

[1545] a text preprocessing means for tokenizing, normalizing and removing stop words from the filtered comments;

[1546] A means to classify sentiment as positive, negative, or neutral based on pre-processed comments and calculate a Net Promoter Score;

[1547] A means for calculating the distribution of positive and negative comments for each carrier and generating comparative information with other carriers;

[1548] means for automatically generating a report including the generated information and predicted results of the fluctuations;

[1549] A system including means for delivering automatically generated reports to user terminals and displaying them in real time.

[1550] (Claim 2)

[1551] The system of claim 1, wherein the collected comments are filtered based on specific rules and unnecessary data is removed using regular expressions.

[1552] (Claim 3)

[1553] 10. The system of claim 1, further comprising means for enabling a user to view the generated reports in real time via a dashboard and search for detailed information for a specific period or carrier.

[1554] "Application Example 1"

[1555] (Claim 1)

[1556] a means for collecting comments about a mobile network service provider from websites on the Internet;

[1557] a means for filtering and text preprocessing the collected comments;

[1558] A means to classify pre-processed comments as positive, negative, or neutral and calculate a Net Promoter Score;

[1559] A means for calculating the distribution of positive and negative comments for each provider and generating comparative information with other providers;

[1560] means for automatically generating a report including the generated information and predicted results of the fluctuations;

[1561] means for delivering and displaying the automatically generated report to a user terminal;

[1562] A system that includes a means of aggregating viewer responses in real time on user devices and recommending content and suggesting areas for improvement based on viewing trends.

[1563] (Claim 2)

[1564] The system of claim 1, wherein the collected comments are filtered based on specific rules and unnecessary data is removed using regular expressions.

[1565] (Claim 3)

[1566] 10. The system of claim 1, further comprising means for enabling a user to view generated reports in real time via a dashboard and search for detailed information for a specific period or by provider.

[1567] "Example 2: Combining Emotion Engines"

[1568] (Claim 1)

[1569] means for collecting evaluation comments about the communications service provider from internet sources, including designated websites and social networks;

[1570] a means for filtering and text preprocessing the collected comments;

[1571] A means to classify pre-processed comments as positive, negative, or neutral and calculate a Net Promoter Score;

[1572] A means for calculating a distribution of positive and negative comments for each communication service provider and generating comparative information with other communication service providers;

[1573] means for automatically generating a report including the generated information and predicted results of the fluctuations;

[1574] A system including means for delivering and displaying automatically generated reports to a user device.

[1575] (Claim 2)

[1576] The system of claim 1, wherein the collected comments are filtered based on specific rules and unnecessary data is removed using regular expressions.

[1577] (Claim 3)

[1578] 2. The system according to claim 1, further comprising means for enabling a user to view the generated report in real time on a display device and search for detailed information for a specific period or for each communication service provider.

[1579] "Application example 2 when combining emotion engines"

[1580] (Claim 1)

[1581] A means for collecting comments about an object from a website on the Internet;

[1582] a means for filtering and text preprocessing the collected comments;

[1583] A means to classify preprocessed comments into positive, negative, and neutral categories and calculate a performance score;

[1584] A means for calculating the distribution of positive and negative comments for each subject and generating comparative information with other subjects;

[1585] means for automatically generating a report including the generated information and predicted results of the fluctuations;

[1586] means for delivering and displaying the automatically generated report to a user terminal;

[1587] A way for users to view reports in real time through a dashboard and see detailed information,

[1588] A means for using AI and emotion engines to collect and analyze feedback comments regarding the efficiency and performance of the object;

[1589] A system including:

[1590] (Claim 2)

[1591] The system of claim 1, wherein the collected comments are filtered based on specific rules and unnecessary data is removed using regular expressions.

[1592] (Claim 3)

[1593] 10. The system of claim 1, further comprising means for enabling a user to view the generated reports in real time via a dashboard and search for detailed information for a specific period or subject. [Explanation of symbols]

[1594] 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 comments about a mobile network service provider from websites on the Internet; a means for filtering and text preprocessing the collected comments; A means to classify pre-processed comments as positive, negative, or neutral and calculate a Net Promoter Score; A means for calculating the distribution of positive and negative comments for each carrier and generating comparative information with other carriers; means for automatically generating a report including the generated information and predicted results of the fluctuations; A system including means for delivering and displaying automatically generated reports to a user terminal.

2. 2. The system according to claim 1, wherein the collected comments are filtered based on specific rules and unnecessary data is removed using regular expressions.

3. 10. The system of claim 1, further comprising means for enabling a user to view the generated reports in real time via a dashboard and search for detailed information for a specific period or carrier.

Images