system

Abstract

We provide the system. [Solution] A means of collecting word-of-mouth data, Means for performing pre-processing, Methods for conducting sentiment analysis, A method for calculating satisfaction scores for each company, Methods for creating rankings against competitors, The means for displaying the aforementioned ranking, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a persona chatbot control method performed by at least one processor, the method including: receiving a user utterance; adding the user utterance to a prompt including an instruction sentence related to an explanation of a 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

Summary of the Invention

Problems to be Solved by the Invention

[0006] "User review data" refers to information that includes opinions and evaluations of products and services posted online by users.

[0007] "Collection methods" refer to the technologies and processes used to obtain word-of-mouth data from multiple websites online.

[0008] "Preprocessing" is the process of removing noise and unnecessary information from raw data and preparing it into a format that can be analyzed.

[0009] "Sentiment analysis" is a technique that analyzes the content of text data and classifies it into emotions such as positive, negative, and neutral.

[0010] "Generative AI" is a type of artificial intelligence technology, specifically referring to models that analyze and generate text data through natural language processing.

[0011] A "satisfaction score" is a numerical evaluation value based on customer opinions and sentiment analysis results.

[0012] "Ranking creation methods" refer to the technologies and processes used to rank companies based on their customer satisfaction scores.

[0013] "Display means" refers to interfaces and technologies used to visually present analysis results and rankings to users. [Brief explanation of the drawing]

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

Embodiments for Carrying Out the Invention

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

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

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

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

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

[0020] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0022] [First Embodiment]

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

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

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

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

[0027] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

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

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

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

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

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

[0035] This system collects review data from multiple websites, preprocesses and performs sentiment analysis on it, calculates satisfaction scores for each company, and creates and displays rankings with competitors. The specific processing of this invention is as follows:

[0036] 1. Data Collection

[0037] The device collects customer review data from various websites (e.g., social media, review sites). For example, the device uses the Twitter API to collect tweets related to a specific hashtag and saves them in JSON format.

[0038] 2. Data preprocessing

[0039] The server receives JSON data sent from the terminal and stores it in the database. Next, the server removes noise. For example, it filters out spam tweets and advertisements using regular expressions. This results in clean data.

[0040] 3. Opinion analysis

[0041] The server uses generated AI to perform sentiment analysis on clean data. Specifically, it uses the BERT model to classify text as positive, negative, or neutral. For example, a tweet like "This product is great" would be classified as positive.

[0042] 4. Calculation of satisfaction score

[0043] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. Next, the total score for each company is tallied. For example, if company A has 15 positive opinions and 5 negative opinions, company A's satisfaction score will be +10.

[0044] 5. Competitive analysis

[0045] The server creates a ranking based on each company's satisfaction score. This ranking is ordered from highest score to lowest.

[0046] 6. Ranking Display

[0047] The server displays the rankings it generates in the user interface (UI). Users can access the web dashboard to check the latest rankings.

[0048] For example, the rankings for companies A, B, and C will be displayed as follows:

[0049] 1. Company C (Satisfaction Score: +12)

[0050] 2. Company A (Satisfaction Score: +10)

[0051] 3. Company B (Satisfaction Score: +8)

[0052] In this way, users can check customer satisfaction rankings for each company in real time and understand the competitive landscape. This system aggregates real customer feedback and visually displays company evaluations, helping companies quickly identify areas for improvement and their strengths.

[0053] The following describes the processing flow.

[0054] Step 1:

[0055] The device collects customer review data from various websites (e.g., social media, review sites). The device uses the Twitter API to send GET requests to retrieve tweets related to specific keywords or hashtags. The retrieved data is stored on the device in JSON format.

[0056] Step 2:

[0057] The device sends the collected review data to the server. The device uses a REST API to send the collected data to the server via a POST request. The sent data is temporarily stored in a database on the server.

[0058] Step 3:

[0059] The server preprocesses the received JSON data. Specifically, the server uses regular expressions to remove noise such as spam and advertisements from the data. It also removes unnecessary spaces and line breaks to normalize the text.

[0060] Step 4:

[0061] The server then performs sentiment analysis on the pre-processed data. Using the BERT model, a generative AI, it determines whether each review text is positive, negative, or neutral. For example, a review that says "This product is great" would be classified as positive.

[0062] Step 5:

[0063] The server assigns a satisfaction score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. This score is stored in the database.

[0064] Step 6:

[0065] The server calculates a total score for each company based on the satisfaction scores. For example, if company A receives 15 positive comments and 5 negative comments, company A's satisfaction score will be +10. The compiled scores are stored in a database.

[0066] Step 7:

[0067] The server creates a ranking based on each company's satisfaction score. The scores for each company are sorted in descending order, and a ranking list is generated. This ranking list is stored in a database.

[0068] Step 8:

[0069] The server updates the UI to display the latest rankings on the web dashboard accessed by the user. Using HTML and JavaScript, a web page is generated to visually display the ranking list to the user.

[0070] Step 9:

[0071] Users access the web dashboard from their browser and view updated rankings in real time. By accessing the dashboard URL in their browser, users can quickly grasp comparisons and statuses of companies.

[0072] This series of steps results in an automated system that handles everything from data collection to analysis and display.

[0073] (Example 1)

[0074] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0075] There is a need for a system that can accurately evaluate a company's customer satisfaction and instantly compare it with competitors. In particular, there is a lack of efficient methods for collecting and analyzing review data from multiple websites. In addition, accuracy and efficiency are challenges in the process of analyzing the sentiment of this data using AI models and calculating satisfaction scores for each company.

[0076] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0077] In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing the data, means for performing sentiment analysis using a generative AI model, means for calculating satisfaction scores for each company, means for creating rankings with competitors, and means for displaying the rankings. This makes it possible to efficiently collect and pre-process word-of-mouth data from multiple websites and perform highly accurate sentiment analysis using a generative AI model. Furthermore, it is possible to quickly calculate satisfaction scores for each company and visually display real-time rankings with competitors.

[0078] "Word-of-mouth data" refers to text data of opinions and impressions about products and services posted by users on online platforms.

[0079] "Preprocessing" is the process of removing noise from collected data and preparing it in a format that is easy to analyze.

[0080] A "generative AI model" is an artificial intelligence model trained using machine learning, and is particularly used for sentiment analysis and text classification.

[0081] "Sentiment analysis" is the process of analyzing the text contained in word-of-mouth data and classifying its content as either positive, negative, or neutral.

[0082] A "satisfaction score" is an indicator of customer satisfaction for each company, calculated by assigning a score to each review data based on the results of sentiment analysis and then aggregating these scores.

[0083] "Ranking creation" is the process of arranging companies in descending order based on their customer satisfaction scores, thereby enabling visual comparisons with competitors.

[0084] "Ranking display" refers to displaying the created ranking results on the user interface so that users can view them.

[0085] This invention is a system that collects, preprocesses, and performs sentiment analysis on word-of-mouth data, calculates a satisfaction score for each company based on the results, and creates and displays a ranking with competitors. The following shows how this system is specifically implemented.

[0086] In implementing this system, the terminal first plays the role of collecting word-of-mouth data. The terminal uses a Python script to access the Twitter API. Specifically, the terminal collects tweets related to a specific hashtag via the Twitter API and saves them in JSON format. In this case, the terminal hardware is a client PC, and the software used is Python and the Twitter API.

[0087] The server receives JSON data sent from the terminal and stores it in a database. Since the stored data requires preprocessing, the server uses regular expressions to remove noise. By using regular expressions, spam tweets and advertisements are filtered out, generating clean data. The server hardware consists of a server machine, and the software used is Python and its regular expression package.

[0088] Next, the server performs sentiment analysis. For this purpose, it uses a generative AI model, specifically the BERT model. The server uses the Hugging Face Transformers library to classify tweets as positive, negative, or neutral. The server hardware continues to be a server machine.

[0089] Based on the sentiment analysis results, the server assigns a score to each review. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. These scores are then aggregated to calculate a satisfaction score for each company. Python and its aggregation library, NumPy, are used to aggregate the scores.

[0090] The server creates a ranking based on the aggregated scores. This ranking is sorted in descending order based on each company's satisfaction score. The server hardware and software will continue to use Python.

[0091] Finally, the user accesses a web dashboard to view the rankings. The server displays the aforementioned rankings in a user interface (UI). This UI is built using HTML, CSS, JavaScript, and a backend framework (e.g., Django or Flask).

[0092] As a concrete example, here is an example of a prompt message that uses the Twitter API to collect review data for the hashtag "good product":

[0093] Collect the latest tweets related to the hashtag "good product" and perform a sentiment analysis on them. Classify them as positive, negative, or neutral, and provide the number of each.

[0094] As described above, the present invention is a system that efficiently collects word-of-mouth data from many websites and performs sentiment analysis using a generative AI model, thereby enabling accurate evaluation of customer satisfaction for each company and real-time comparison with competitors.

[0095] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0096] Step 1: Data Collection

[0097] The device collects word-of-mouth data from various websites (e.g., social media, review sites). Specifically, the device uses a Python script to access the Twitter API. The device collects tweets related to a specific hashtag, "good product," and saves them in JSON format. The input is a specific hashtag, and the output is a JSON file of tweet data related to that hashtag. In terms of operation, it uses the Twitter API to collect tweets and saves them to local storage in JSON format.

[0098] Step 2: Data Preprocessing

[0099] The server receives JSON data sent from the terminal and stores it in a database. Next, the server uses regular expressions to remove noise from the data. For example, it filters out spam tweets and advertisements to obtain clean data. The input is collected tweet data in JSON format, and the output is clean tweet data with noise removed. Specifically, it uses a regular expression package to filter out advertising links and spam messages.

[0100] Step 3: Sentiment Analysis

[0101] The server generates clean, pre-processed data and performs sentiment analysis using an AI model. Specifically, it uses the BERT model with the Hugging Face Transformers library to classify each tweet as positive, negative, or neutral. The input is clean data, and the output is the result of the sentiment analysis for each tweet. In terms of operation, it calls the sentiment analysis model and classifies the sentiment of each tweet.

[0102] Step 4: Calculating the satisfaction score

[0103] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. Then, it compiles the total score for each company. The input is the result of the sentiment analysis, and the output is the satisfaction score for each company. Specifically, it compiles the sentiment score for each tweet and calculates the total score for each company.

[0104] Step 5: Competitive Analysis

[0105] The server creates a ranking of each company based on their satisfaction scores. The input is the satisfaction score for each company, and the output is the ranking. Specifically, it sorts the scores for each company in descending order and generates a ranking list.

[0106] Step 6: Display the rankings

[0107] Users can access a web dashboard to view their rankings. The server displays the generated rankings in the user interface (UI). The input is ranking data, and the output is a visual display of the rankings for the user. Specifically, HTML, CSS, and JavaScript are used to display the rankings on the web page, allowing users to visually confirm them.

[0108] (Application Example 1)

[0109] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0110] In recent years, consumer reviews have come to have a significant impact on a company's reputation. However, manually analyzing vast amounts of review data is difficult, making it challenging to grasp a company's relative ranking against its competitors in real time. To solve this problem, a system is needed that efficiently collects, preprocesses, and analyzes review data, calculates satisfaction scores for each company, and visualizes competitor rankings on smartphones.

[0111] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0112] In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing the data, and means for performing sentiment analysis. This makes it possible to efficiently collect consumer word-of-mouth data, process it into clean data, and perform sentiment classification using generative AI. Furthermore, by creating a ranking of competitors based on the calculated satisfaction score and displaying it on a smartphone, companies can grasp the competitive situation in real time.

[0113] "Word-of-mouth data" refers to reviews and comments posted by customers and users on online platforms regarding products and services.

[0114] "Preprocessing" refers to the process of removing noise from collected data and converting it into a format suitable for analysis.

[0115] "Sentiment analysis" is a method of analyzing text data and classifying its emotional tendencies (positive, negative, neutral).

[0116] The "satisfaction score" is a numerical value that represents an overall evaluation, calculated by assigning positive scores to positive reviews and negative scores to negative reviews based on the results of sentiment analysis.

[0117] A "ranking" refers to a list that ranks companies based on satisfaction scores and allows for comparison with competitors.

[0118] "A means of displaying rankings on a smartphone" refers to a function that visually displays customer satisfaction rankings for each company on the smartphone screen.

[0119] "Generative AI" is a general term for artificial intelligence that generates and analyzes text, images, and other data, and is a technology that enables highly accurate analysis, particularly in natural language processing.

[0120] This invention is a system that consistently performs the collection, preprocessing, sentiment analysis, satisfaction score calculation, competitor analysis, and ranking display of word-of-mouth data. It is implemented using a server, a smartphone, and a generative AI model.

[0121] The system program follows the following main processing steps:

[0122] 1. Data Collection

[0123] The server collects user review data from multiple websites, including social media and review sites. For example, it uses the Twitter API to retrieve tweets related to a specific hashtag. The collected data is stored in JSON format.

[0124] 2. Data preprocessing

[0125] The server receives the collected JSON data and stores it in a database. Next, it cleans up the data. For example, it filters out spam and advertisements using regular expressions and removes noise. This results in clean data suitable for analysis.

[0126] 3. Sentiment analysis

[0127] The server uses a generative AI model (e.g., BERT model) to perform sentiment analysis on clean data. It classifies text data as positive, negative, or neutral. This process is accelerated using CUDA-enabled GPUs.

[0128] 4. Calculation of satisfaction score

[0129] The server assigns a score to each review based on the sentiment analysis results. Positive reviews receive a score of +1, negative reviews receive -1, and neutral reviews receive 0. The total scores for each company are then tallied to calculate the satisfaction score.

[0130] 5. Competitive analysis

[0131] The server creates a ranking based on each company's satisfaction score. This ranking is ordered from highest to lowest score, visually representing the competitive landscape.

[0132] 6. Ranking Display

[0133] The smartphone displays ranking data sent from the server to the user. Users can check real-time rankings through the smartphone application.

[0134] Specific examples and prompt statements

[0135] As a concrete example, the following shows the prompt text that a user would input to the generated AI model.

[0136] Example of a prompt:

[0137] "Use this app to perform sentiment analysis on tweets about Product A and display a competitor satisfaction ranking."

[0138] This prompt allows users to easily grasp real-time customer review results and competitive landscape for specific products or companies.

[0139] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0140] Step 1:

[0141] The device collects user review data from multiple websites, including social media and review sites. It uses the Twitter API to collect tweets related to specific hashtags. The input is a specific hashtag or keyword, and the output is the corresponding tweet data (in JSON format). This collected data is sent to a server and stored in a database.

[0142] Step 2:

[0143] The server receives JSON data sent from the terminal and stores it in the database. It then preprocesses the data. Specifically, it uses regular expressions to filter out spam tweets and advertisements and removes noise. The input is raw JSON data, and the output is clean JSON data.

[0144] Step 3:

[0145] The server uses a generative AI model (e.g., the BERT model) to perform sentiment analysis on clean data. Specifically, it classifies text data into positive, negative, and neutral. The input is clean JSON data, and the output is data with the sentiment classification result for each tweet added. A CUDA-enabled GPU is used for the analysis.

[0146] Step 4:

[0147] The server assigns a score to each review based on the sentiment analysis results. Positive reviews receive a score of +1, negative reviews receive -1, and neutral reviews receive 0. The total scores for each company are aggregated to calculate the satisfaction score. The input is data including the sentiment classification results, and the output is the satisfaction score for each company.

[0148] Step 5:

[0149] The server creates a ranking of each company against its competitors based on their satisfaction scores. This ranking is ordered from highest score to lowest. The input is the satisfaction score for each company, and the output is a list of company rankings.

[0150] Step 6:

[0151] The smartphone displays ranking data transmitted from the server to the user. Users can check real-time rankings through the smartphone application. The input is company ranking data transmitted from the server, and the output is the ranking display in the user interface.

[0152] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0153] This invention combines a system that collects word-of-mouth data, performs preprocessing, sentiment analysis, calculates satisfaction scores, creates rankings, and displays rankings, with a sentiment engine that recognizes user emotions to perform even more advanced analysis and feedback.

[0154] overview

[0155] This system primarily consists of a server, terminals, and users. The emotion engine recognizes emotions from the user's voice or text input and sends this information to the server, which can then be reflected in the ranking display.

[0156] Data collection

[0157] The device collects customer review data from various websites (e.g., social media, review sites). The device retrieves tweets and reviews using specific keywords and hashtags and saves them in JSON format. For example, the device collects reviews from Twitter using the hashtag "product A review".

[0158] Data preprocessing

[0159] The server receives JSON data sent from the terminal and stores it in the database. The server uses regular expressions to remove noise and unnecessary information, creating a clean dataset. For example, it can filter out advertisements and spam messages.

[0160] sentiment analysis

[0161] The server uses the BERT model, a generated AI, to perform sentiment analysis on text data. The analysis results are classified as positive, negative, or neutral. For example, the text "This product is great!" is classified as positive.

[0162] Calculation of satisfaction score

[0163] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. The satisfaction scores for each company are compiled and stored in a database. For example, if company A has 15 positive reviews and 5 negative reviews, company A's satisfaction score will be +10.

[0164] Ranking creation

[0165] The server creates a ranking based on each company's satisfaction score. The total score for each company is sorted in descending order, and the ranking is generated.

[0166] Ranking display

[0167] The server displays the ranking results on a web dashboard that users access. The rankings are displayed visually using HTML and JavaScript. Users can access the dashboard in their browser and check the latest rankings. For example, it might show Company A in 1st place, Company B in 2nd place, and Company C in 3rd place.

[0168] Combination of emotional engines

[0169] Users access a web dashboard and enter feedback via text or voice. The emotion engine analyzes this feedback in real time and categorizes the user's emotions as positive, negative, or neutral. The results are sent to the server and reflected in the ranking display. For example, if a user enters feedback saying "This product is very good," the emotion engine will determine this to be positive and it will be reflected in the ranking display.

[0170] Specific example

[0171] As a concrete example, suppose a device collects 50 tweets from Twitter using the hashtag "product A review," of which 30 are positive, 15 are negative, and 5 are neutral. These tweets are pre-processed on a server, and sentiment analysis is performed using a generative AI. As a result, a satisfaction score of +15 is calculated for company A, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking.

[0172] In this way, a sophisticated system is realized that not only collects, analyzes, and displays user reviews, but also reflects user sentiment in real time.

[0173] The following describes the processing flow.

[0174] Step 1:

[0175] The device collects customer review data from various websites (e.g., social media, review sites). Specifically, the device uses specific keywords or hashtags (e.g., "product A review") to retrieve tweets from Twitter via GET requests. This tweet data is stored on the device in JSON format.

[0176] Step 2:

[0177] The device sends the collected review data to the server. The device uses a REST API to send the collected data to the server via a POST request. The sent data is temporarily stored in a database on the server.

[0178] Step 3:

[0179] The server preprocesses the received JSON data. The server uses regular expressions to remove noise (e.g., spam, advertisements) from the data, creating a clean dataset. For example, it removes advertising messages such as "Free sample giveaway!".

[0180] Step 4:

[0181] The server uses the BERT model, a generative AI, to perform sentiment analysis on pre-processed review data. It determines whether the text is positive, negative, or neutral and stores the results in a database. For example, a tweet like "This product is great" would be classified as positive.

[0182] Step 5:

[0183] The server assigns a satisfaction score to each review based on the sentiment analysis results. Positive opinions are given a score of +1, negative opinions a score of -1, and neutral opinions a score of 0, and these scores are stored in the database. For example, if there are 30 positive reviews and 15 negative reviews, Company A's score will be +15.

[0184] Step 6:

[0185] The server collects satisfaction scores from each company and creates a ranking. The scores for each company are sorted in descending order to generate a ranking list, which is then saved to the database. For example, Company A might be ranked 1st, Company B 2nd, and Company C 3rd.

[0186] Step 7:

[0187] The UI for displaying the server-generated rankings on the web dashboard is updated. The server uses HTML and JavaScript to generate a web page that visually displays the ranking list to the user. For example, company A is displayed as number 1 in the latest rankings.

[0188] Step 8:

[0189] Users access the web dashboard and then access the dashboard URL in their browser to view the rankings. Users can see the company satisfaction rankings in real time.

[0190] Step 9:

[0191] Users provide feedback. On the web dashboard, users offer their opinions on products and services through text or voice input. This input is sent to the sentiment engine.

[0192] Step 10:

[0193] The emotion engine analyzes user feedback in real time. It uses an emotion analysis algorithm to analyze voice and text input, classifying it as positive, negative, or neutral. The results are then sent to the server.

[0194] Step 11:

[0195] The server receives the results from the emotion engine and reflects them in the ranking display. For example, if new positive user feedback is reflected, Company A's satisfaction score will be increased by +1, and the ranking will be updated again.

[0196] This series of steps completes a system where data collection, analysis, display, and feedback integration are all automated.

[0197] (Example 2)

[0198] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0199] Traditional systems often require manual collection and analysis of customer review data, which is time-consuming and labor-intensive. Furthermore, it's difficult to reflect user feedback in real time, leading to issues with the accuracy of company satisfaction scores and rankings. Additionally, the inability to collect data from multiple sources simultaneously and perform comprehensive analysis makes accurate evaluation difficult.

[0200] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing, means for performing sentiment analysis using a generative AI model, means for calculating satisfaction scores for each company, means for creating rankings with competitors, means for displaying the rankings, means for analyzing user feedback and recognizing emotions, and means for reflecting the analysis results in the rankings. This automates all processes from word-of-mouth data collection and analysis to ranking creation and display, and further enables the reflection of real-time user feedback.

[0201] "Word-of-mouth data" refers to ratings and opinions posted by customers or users about products and services on online platforms (e.g., social media, review sites).

[0202] "Collection methods" refer to methods and devices for obtaining word-of-mouth data from various information sources.

[0203] "Preprocessing" refers to the process of removing noise and unnecessary information from collected data and converting it into a format suitable for analysis.

[0204] A "generative AI model" refers to a machine learning model used to automatically generate or analyze text and data.

[0205] "Sentiment analysis" refers to an analytical method for identifying emotions (positive, negative, neutral) from text data.

[0206] A "satisfaction score" refers to a score that quantifies the results of sentiment analysis and indicates an overall evaluation of a company or product.

[0207] A "ranking" refers to a list of companies or products based on customer satisfaction scores.

[0208] "Display means" refers to methods or devices that allow users to visually confirm results.

[0209] "Feedback" refers to the opinions and evaluations that users provide about a product or service.

[0210] An "emotion engine" refers to technology that identifies emotions in real time from user feedback.

[0211] This invention combines a system that collects, preprocesses, analyzes sentiment, calculates satisfaction scores, creates rankings, and displays rankings with a sentiment engine that recognizes user emotions, enabling more advanced analysis and feedback. This automates the entire process from collecting and analyzing review data to creating and displaying rankings, and also allows for the incorporation of real-time user feedback.

[0212] overview

[0213] This system primarily consists of a server, terminals, and users. The emotion engine recognizes emotions from the user's voice or text input and sends this information to the server, which can then be reflected in the ranking display.

[0214] Data collection

[0215] The device uses web scraping tools (e.g., Beautiful Soup, Scrapy) to collect user reviews from social media and review sites. For example, it might search for tweets using the hashtag "product A review" and save the data in JSON format. This allows for the unified collection of data from various sources.

[0216] Data preprocessing

[0217] The server receives JSON data sent from the terminal and stores it in a database (e.g., MySQL®, PostgreSQL). Next, the server uses regular expressions and natural language processing (NLP) libraries (e.g., NLTK, SpaCy) to remove noise and unnecessary information, creating a clean dataset. This improves the quality of the data and allows for more accurate subsequent processing.

[0218] sentiment analysis

[0219] The server uses a generated AI model (e.g., BERT, GPT-3®) to perform sentiment analysis on text data. Specifically, it classifies customer review text into positive, negative, and neutral. For example, the text "This product is the best!" would be classified as positive. This allows for an understanding of general customer sentiment trends.

[0220] Calculation of satisfaction score

[0221] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. The satisfaction scores for each company are then compiled and stored in a database. For example, if company A receives 15 positive reviews and 5 negative reviews, company A's satisfaction score will be +10.

[0222] Ranking creation

[0223] The server creates a ranking based on customer satisfaction scores for each company. The total scores for each company are then sorted in descending order to generate the ranking. This makes it easy to compare companies with competitors and allows for quick information provision to users.

[0224] Ranking display

[0225] The server displays the ranking results on a web dashboard that users access. The rankings are displayed visually using HTML and JavaScript. Users can access the dashboard in their browser and check the latest rankings. For example, it might show Company A in 1st place, Company B in 2nd place, and Company C in 3rd place.

[0226] Combination of emotional engines

[0227] Users access a web dashboard and enter feedback via text or voice. The emotion engine analyzes this feedback in real time and categorizes the user's emotions as positive, negative, or neutral. The results are sent to the server and reflected in the ranking display. For example, if a user enters feedback saying "This product is very good," the emotion engine will determine this to be positive and it will be reflected in the ranking display.

[0228] Specific example

[0229] As a concrete example, suppose a device collects 50 tweets from Twitter using the hashtag "product A review," with 30 being positive, 15 negative, and 5 neutral. These tweets are preprocessed on a server, and sentiment analysis is performed using a generative AI model. As a result, a satisfaction score of +15 is calculated for company A, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking.

[0230] In this way, a sophisticated system is realized that not only collects, analyzes, and displays user reviews, but also reflects user sentiment in real time.

[0231] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0232] Step 1: Data Collection

[0233] The device uses a web scraping tool (e.g., Beautiful Soup, Scrapy) to collect user review data from social media and review sites. Input requires the URL of the target site and search keywords (e.g., "product A review"). Based on this input, the device extracts the necessary user review data from the website and saves it in JSON format. The output is the collected user review data (in JSON format).

[0234] Step 2: Data transmission

[0235] The device sends the collected review data (in JSON format) to the server. The input requires the collected review data and the destination server address. The device then forwards the data to the server based on this input. The output is the review data sent to the server.

[0236] Step 3: Data Preprocessing

[0237] The server receives the JSON data and stores it in a database (e.g., MySQL, PostgreSQL). Next, the server uses regular expressions and natural language processing (NLP) libraries (e.g., NLTK, SpaCy) to remove noise and unwanted information, creating a clean dataset. The input is the received raw data (in JSON format). The output is a pre-processed, clean dataset.

[0238] Step 4: Sentiment Analysis

[0239] The server uses a generative AI model (e.g., BERT, GPT-3) to perform sentiment analysis on text data. Specifically, it classifies word-of-mouth text into positive, negative, and neutral. The input requires a pre-processed, clean dataset. The server uses this to perform sentiment analysis and assigns sentiment labels to each text. The output is the data with sentiment labels assigned.

[0240] Step 5: Calculating the Satisfaction Score

[0241] The server assigns a score to each review based on the sentiment analysis results. The input requires data with sentiment labels. Positive opinions are assigned a score of +1, negative opinions a score of -1, and neutral opinions a score of 0. The satisfaction scores for each company are then compiled and stored in a database. The output is the satisfaction score for each company.

[0242] Step 6: Create Rankings

[0243] The server creates a ranking based on customer satisfaction scores for each company. The input requires customer satisfaction scores for each company. The server sorts these scores in descending order to generate the ranking and saves this data to a database. The output is the ranking for each company.

[0244] Step 7: Display the rankings

[0245] The server displays the ranking results on a web dashboard accessed by the user. The input requires generated ranking data. The server uses HTML and JavaScript to visually display the rankings. The output is the latest ranking displayed on the dashboard.

[0246] Step 8: Gathering Feedback

[0247] The user accesses the web dashboard and enters feedback via text or voice. The input requires the user's feedback as text or voice data. The feedback is sent to the server. The output is the feedback data sent to the server.

[0248] Step 9: Analysis using the Emotion Engine

[0249] The emotion engine analyzes user feedback in real time and classifies the user's emotions as positive, negative, or neutral. The input is user feedback data. The output is feedback data with emotion labels attached.

[0250] Step 10: Reflection in rankings

[0251] The server receives the analysis results from the emotion engine and reflects them in the ranking display. The input requires feedback data with emotion labels. The server uses this data to update the rankings and reflects them in the dashboard in real time. The output is the updated rankings.

[0252] (Application Example 2)

[0253] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0254] Traditional word-of-mouth data analysis systems have struggled to reflect customers' real-time emotions, making immediate improvements in customer satisfaction impossible. Furthermore, systems that generate rankings based on a comprehensive evaluation of multiple ratings suffer from low variability due to their reliance on static data. As a result, the latest user emotions and feedback are not reflected in business decisions or marketing strategies, preventing a more accurate response to customer needs.

[0255] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing, means for performing sentiment analysis, means for calculating satisfaction scores for each company, means for creating rankings with competitors, means for displaying the rankings, means for receiving voice or text feedback from users and analyzing it using a sentiment engine, and means for reflecting the analysis results in the rankings in real time. This makes it possible to immediately analyze real-time user feedback and reflect it in the rankings.

[0256] "Word-of-mouth data" refers to a collection of digital opinions such as reviews, comments, and ratings posted by customers on the internet about products and services.

[0257] "Preprocessing" is the process of removing unnecessary information and noise from raw data and organizing it so that it can be analyzed accurately.

[0258] "Sentiment analysis" is a technique that analyzes the emotions contained in opinions and feedback within text data and classifies them into categories such as positive, negative, and neutral.

[0259] A "satisfaction score" is an index that quantifies customer satisfaction with a product or service based on customer feedback.

[0260] A "ranking" is a list that evaluates and ranks multiple companies or products based on criteria such as satisfaction scores.

[0261] An "emotion engine" is software or hardware that analyzes voice or text feedback from users in real time and classifies their emotions.

[0262] A "user" is someone who accesses the system, views user reviews, and provides feedback.

[0263] "Real-time" refers to a situation where data or events are processed and reflected immediately the moment they occur.

[0264] The system implementing this invention mainly consists of a server, a terminal, and a user. The specific roles of each component are described below.

[0265] server

[0266] The server plays a central role in collecting, pre-processing, and performing sentiment analysis on review data. Specifically, this includes the following methods:

[0267] 1. Data Collection Method: Receive word-of-mouth data sent from the device and store it in a database. The data is stored in JSON format.

[0268] 2. Preprocessing method: Regular expressions are used to remove noise and unwanted information, creating a clean dataset. For example, advertisements and spam messages are filtered out.

[0269] 3. Sentiment Analysis Method: The BERT model, a generative AI model, is used to analyze the sentiment of the collected text data. The analysis results are classified into positive, negative, and neutral.

[0270] 4. Satisfaction Score Calculation Method: Based on the results of sentiment analysis, a score is assigned to each review, and a total score is calculated for each company. For example, positive opinions are given a score of +1, negative opinions a score of -1, and neutral opinions a score of 0.

[0271] 5. Ranking creation method: The total score for each company is sorted in descending order, and a ranking is generated.

[0272] 6. Ranking Display Method: Ranking results will be displayed on a web dashboard. The display will be visually generated using HTML and JavaScript.

[0273] terminal

[0274] The device is responsible for collecting user review data and sending it to the server. Specifically, this includes the following methods:

[0275] 1. Data Collection Methods: Collect word-of-mouth data from multiple websites (e.g., social media, review sites). For example, obtain word-of-mouth data from Twitter using specific keywords or hashtags.

[0276] User

[0277] Users are responsible for accessing the system, providing feedback, and checking rankings. Specifically, this includes the following:

[0278] 1. Feedback input method: Access the web dashboard and enter feedback in text or voice.

[0279] 2. Emotion Analysis Method: The emotion engine analyzes user feedback in real time, classifies emotions as positive, negative, or neutral, and sends the results to the server.

[0280] 3. How to check rankings: Check the latest rankings on the web dashboard. Rankings are updated in real time.

[0281] Hardware and software to be used

[0282] Hardware:

[0283] Server (for data storage and processing)

[0284] Smartphone (for user feedback input)

[0285] Software:

[0286] transformers library (using BERT model)

[0287] requests library (for data collection)

[0288] Regular expression library (for data preprocessing)

[0289] softmax function (Scipy library; for sentiment score calculation)

[0290] Specific example

[0291] A specific example is shown. Suppose a terminal collects 50 tweets from Twitter with the hashtag "Product A Review", where 30 are positive, 15 are negative, and 5 are neutral. These tweets are preprocessed on the server and sentiment analysis is performed using a generative AI model. As a result, the satisfaction score for Company A is calculated as +15 and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good", the sentiment engine determines this as positive and it is reflected in the ranking in real time.

[0292] Examples of prompt sentences

[0293] "Analyze the sentiment of the following text: 'This product is great.' Classify the sentiment score as positive, negative, or neutral."

[0294] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0295] Step 1:

[0296] The device collects user review data from multiple websites (e.g., social media and review sites). For example, it retrieves reviews from Twitter using specific keywords or hashtags. The input is a specific keyword such as "product A review," and the output is user review data in JSON format. This data is initially saved for later analysis.

[0297] Step 2:

[0298] The server receives review data sent from the terminal and stores it in a database. The input is JSON data sent from the terminal, and the output is storage data stored in the database. A regular expression library is used to remove noise and unwanted information, such as advertisements and spam messages.

[0299] Step 3:

[0300] The server generates a clean dataset. The input is pre-processed review data using regular expressions, and the output is clean JSON formatted data. At this stage, the data is ready for analysis, so we move on to the next step. Specifically, we remove unnecessary strings and HTML tags using regular expressions.

[0301] Step 4:

[0302] The server performs sentiment analysis using the BERT model, which is a generative AI model. The input is each review text in the clean dataset, and the output is the sentiment score (positive, negative, neutral) for each text. Specifically, each text is input into the BERT model to obtain the sentiment score.

[0303] Step 5:

[0304] Based on the results of the sentiment analysis, the server assigns scores to each review. The input is the sentiment analysis result, and the output is the satisfaction score for each company. For example, a positive opinion is assigned a score of +1, a negative opinion is assigned a score of -1, and a neutral opinion is assigned a score of 0. Thus, the total score for each company is calculated.

[0305] Step 6:

[0306] The server sorts the total scores for each company in descending order and generates a ranking. The input is the satisfaction score, and the output is the ranking for each company. For example, the company with the highest total score ranks first.

[0307] Step 7:

[0308] The server displays the ranking results on the web dashboard. The input is the generated ranking, and the output is the visually displayed ranking information. Using HTML and JavaScript, users can view the latest ranking in a browser.

[0309] Step 8:

[0310] The user accesses the web dashboard and inputs feedback in text or voice. The input is the user's feedback text or voice, and the output is the feedback information sent to the sentiment engine.

[0311] Step 9:

[0312] The emotion engine analyzes user feedback in real time. The input is user feedback, and the output is a real-time emotion score. Specifically, user feedback is input into a BERT model to obtain an emotion score.

[0313] Step 10:

[0314] The server reflects the analysis results in the rankings in real time. The input is the real-time sentiment score, and the output is the updated ranking. This ensures that the rankings display the latest user feedback immediately.

[0315] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0316] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0317] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0318] [Second Embodiment]

[0319] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0320] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0321] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0322] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0323] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0324] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0325] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0326] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0327] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0328] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0329] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0330] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0331] This system collects review data from multiple websites, preprocesses and performs sentiment analysis on it, calculates satisfaction scores for each company, and creates and displays rankings with competitors. The specific processing of this invention is as follows:

[0332] 1. Data Collection

[0333] The device collects customer review data from various websites (e.g., social media, review sites). For example, the device uses the Twitter API to collect tweets related to a specific hashtag and saves them in JSON format.

[0334] 2. Data preprocessing

[0335] The server receives JSON data sent from the terminal and stores it in the database. Next, the server removes noise. For example, it filters out spam tweets and advertisements using regular expressions. This results in clean data.

[0336] 3. Opinion analysis

[0337] The server uses generated AI to perform sentiment analysis on clean data. Specifically, it uses the BERT model to classify text as positive, negative, or neutral. For example, a tweet like "This product is great" would be classified as positive.

[0338] 4. Calculation of satisfaction score

[0339] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. Next, the total score for each company is tallied. For example, if company A has 15 positive opinions and 5 negative opinions, company A's satisfaction score will be +10.

[0340] 5. Competitive analysis

[0341] The server creates a ranking based on each company's satisfaction score. This ranking is ordered from highest score to lowest.

[0342] 6. Ranking Display

[0343] The server displays the rankings it generates in the user interface (UI). Users can access the web dashboard to check the latest rankings.

[0344] For example, the rankings for companies A, B, and C will be displayed as follows:

[0345] 1. Company C (Satisfaction Score: +12)

[0346] 2. Company A (Satisfaction Score: +10)

[0347] 3. Company B (Satisfaction Score: +8)

[0348] In this way, users can check customer satisfaction rankings for each company in real time and understand the competitive landscape. This system aggregates real customer feedback and visually displays company evaluations, helping companies quickly identify areas for improvement and their strengths.

[0349] The following describes the processing flow.

[0350] Step 1:

[0351] The device collects customer review data from various websites (e.g., social media, review sites). The device uses the Twitter API to send GET requests to retrieve tweets related to specific keywords or hashtags. The retrieved data is stored on the device in JSON format.

[0352] Step 2:

[0353] The device sends the collected review data to the server. The device uses a REST API to send the collected data to the server via a POST request. The sent data is temporarily stored in a database on the server.

[0354] Step 3:

[0355] The server preprocesses the received JSON data. Specifically, the server uses regular expressions to remove noise such as spam and advertisements from the data. It also removes unnecessary spaces and line breaks to normalize the text.

[0356] Step 4:

[0357] The server then performs sentiment analysis on the pre-processed data. Using the generative AI BERT model, it determines whether each review text is positive, negative, or neutral. For example, a review saying "This product is great" would be classified as positive.

[0358] Step 5:

[0359] The server assigns a satisfaction score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. This score is stored in the database.

[0360] Step 6:

[0361] The server calculates a total score for each company based on the satisfaction scores. For example, if company A receives 15 positive comments and 5 negative comments, company A's satisfaction score will be +10. The compiled scores are stored in a database.

[0362] Step 7:

[0363] The server creates a ranking based on each company's satisfaction score. The scores for each company are sorted in descending order, and a ranking list is generated. This ranking list is stored in a database.

[0364] Step 8:

[0365] The server updates the UI to display the latest rankings on the web dashboard accessed by the user. HTML and JavaScript are used to generate a web page that visually displays the ranking list to the user.

[0366] Step 9:

[0367] Users access the web dashboard from their browser and view updated rankings in real time. By accessing the dashboard URL in their browser, users can quickly grasp comparisons and statuses of companies.

[0368] This series of steps results in an automated system that handles everything from data collection to analysis and display.

[0369] (Example 1)

[0370] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0371] There is a need for a system that can accurately evaluate a company's customer satisfaction and instantly compare it with competitors. In particular, there is a lack of efficient methods for collecting and analyzing review data from multiple websites. In addition, accuracy and efficiency are challenges in the process of analyzing the sentiment of this data using AI models and calculating satisfaction scores for each company.

[0372] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0373] In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing the data, means for performing sentiment analysis using a generative AI model, means for calculating satisfaction scores for each company, means for creating rankings with competitors, and means for displaying the rankings. This makes it possible to efficiently collect and pre-process word-of-mouth data from multiple websites and perform highly accurate sentiment analysis using a generative AI model. Furthermore, it is possible to quickly calculate satisfaction scores for each company and visually display real-time rankings with competitors.

[0374] "Word-of-mouth data" refers to text data of opinions and impressions about products and services posted by users on online platforms.

[0375] "Preprocessing" is the process of removing noise from collected data and preparing it in a format that is easy to analyze.

[0376] A "generative AI model" is an artificial intelligence model trained using machine learning, and is particularly used for sentiment analysis and text classification.

[0377] "Sentiment analysis" is the process of analyzing the text contained in word-of-mouth data and classifying its content as either positive, negative, or neutral.

[0378] A "satisfaction score" is an indicator of customer satisfaction for each company, calculated by assigning a score to each review data based on the results of sentiment analysis and then aggregating these scores.

[0379] "Ranking creation" is the process of arranging companies in descending order based on their customer satisfaction scores, thereby enabling visual comparisons with competitors.

[0380] "Ranking display" refers to displaying the created ranking results on the user interface so that users can view them.

[0381] This invention is a system that collects, preprocesses, and performs sentiment analysis on word-of-mouth data, calculates a satisfaction score for each company based on the results, and creates and displays a ranking with competitors. The following shows how this system is specifically implemented.

[0382] In implementing this system, the terminal first plays the role of collecting word-of-mouth data. The terminal uses a Python script to access the Twitter API. Specifically, the terminal collects tweets related to a specific hashtag via the Twitter API and saves them in JSON format. In this case, the terminal hardware is a client PC, and the software used is Python and the Twitter API.

[0383] The server receives JSON data sent from the terminal and stores it in a database. Since the stored data requires preprocessing, the server uses regular expressions to remove noise. By using regular expressions, spam tweets and advertisements are filtered out, generating clean data. The server hardware consists of a server machine, and the software used is Python and its regular expression package.

[0384] Next, the server performs sentiment analysis. For this purpose, it uses a generative AI model, specifically the BERT model. The server uses the Hugging Face Transformers library to classify tweets as positive, negative, or neutral. The server hardware continues to be a server machine.

[0385] Based on the sentiment analysis results, the server assigns a score to each review. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. These scores are then aggregated to calculate a satisfaction score for each company. Python and its aggregation library, NumPy, are used to aggregate the scores.

[0386] The server creates a ranking based on the aggregated scores. This ranking is sorted in descending order based on each company's satisfaction score. The server hardware and software will continue to use Python.

[0387] Finally, the user accesses a web dashboard to view the rankings. The server displays the aforementioned rankings in a user interface (UI). This UI is built using HTML, CSS, JavaScript, and a backend framework (e.g., Django or Flask).

[0388] As a concrete example, here is an example of a prompt message that uses the Twitter API to collect review data for the hashtag "good product":

[0389] Collect the latest tweets related to the hashtag "good product" and perform a sentiment analysis on them. Classify them as positive, negative, or neutral, and provide the number of each.

[0390] As described above, the present invention is a system that efficiently collects word-of-mouth data from many websites and performs sentiment analysis using a generative AI model, thereby enabling accurate evaluation of customer satisfaction for each company and real-time comparison with competitors.

[0391] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0392] Step 1: Data Collection

[0393] The device collects word-of-mouth data from various websites (e.g., social media, review sites). Specifically, the device uses a Python script to access the Twitter API. The device collects tweets related to a specific hashtag, "good product," and saves them in JSON format. The input is a specific hashtag, and the output is a JSON file of tweet data related to that hashtag. In terms of operation, it uses the Twitter API to collect tweets and saves them to local storage in JSON format.

[0394] Step 2: Data Preprocessing

[0395] The server receives JSON data sent from the terminal and stores it in a database. Next, the server uses regular expressions to remove noise from the data. For example, it filters out spam tweets and advertisements to obtain clean data. The input is collected tweet data in JSON format, and the output is clean tweet data with noise removed. Specifically, it uses a regular expression package to filter out advertising links and spam messages.

[0396] Step 3: Sentiment Analysis

[0397] The server generates clean, pre-processed data and performs sentiment analysis using an AI model. Specifically, it uses the BERT model with the Hugging Face Transformers library to classify each tweet as positive, negative, or neutral. The input is clean data, and the output is the result of the sentiment analysis for each tweet. In terms of operation, it calls the sentiment analysis model and classifies the sentiment of each tweet.

[0398] Step 4: Calculating the satisfaction score

[0399] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. Then, it compiles the total score for each company. The input is the result of the sentiment analysis, and the output is the satisfaction score for each company. Specifically, it compiles the sentiment score for each tweet and calculates the total score for each company.

[0400] Step 5: Competitive Analysis

[0401] The server creates a ranking of each company based on their satisfaction scores. The input is the satisfaction score for each company, and the output is the ranking. Specifically, it sorts the scores for each company in descending order and generates a ranking list.

[0402] Step 6: Display the rankings

[0403] Users can access a web dashboard to view their rankings. The server displays the generated rankings in the user interface (UI). The input is ranking data, and the output is a visual display of the rankings for the user. Specifically, HTML, CSS, and JavaScript are used to display the rankings on the web page, allowing users to visually confirm them.

[0404] (Application Example 1)

[0405] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0406] In recent years, consumer reviews have come to have a significant impact on a company's reputation. However, manually analyzing vast amounts of review data is difficult, making it challenging to grasp a company's relative ranking against its competitors in real time. To solve this problem, a system is needed that efficiently collects, preprocesses, and analyzes review data, calculates satisfaction scores for each company, and visualizes competitor rankings on smartphones.

[0407] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0408] In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing the data, and means for performing sentiment analysis. This makes it possible to efficiently collect consumer word-of-mouth data, process it into clean data, and perform sentiment classification using generative AI. Furthermore, by creating a ranking of competitors based on the calculated satisfaction score and displaying it on a smartphone, companies can grasp the competitive situation in real time.

[0409] "Word-of-mouth data" refers to reviews and comments posted by customers and users on online platforms regarding products and services.

[0410] "Preprocessing" refers to the process of removing noise from collected data and converting it into a format suitable for analysis.

[0411] "Sentiment analysis" is a method of analyzing text data and classifying its emotional tendencies (positive, negative, neutral).

[0412] The "satisfaction score" is a numerical value that represents an overall evaluation, calculated by assigning positive scores to positive reviews and negative scores to negative reviews based on the results of sentiment analysis.

[0413] A "ranking" refers to a list that ranks companies based on satisfaction scores and allows for comparison with competitors.

[0414] "A means of displaying rankings on a smartphone" refers to a function that visually displays customer satisfaction rankings for each company on the smartphone screen.

[0415] "Generative AI" is a general term for artificial intelligence that generates and analyzes text, images, and other data, and is a technology that enables highly accurate analysis, particularly in natural language processing.

[0416] This invention is a system that consistently performs the collection, preprocessing, sentiment analysis, satisfaction score calculation, competitor analysis, and ranking display of word-of-mouth data. It is implemented using a server, a smartphone, and a generative AI model.

[0417] The system program follows the following main processing steps:

[0418] 1. Data Collection

[0419] The server collects user review data from multiple websites, including social media and review sites. For example, it uses the Twitter API to retrieve tweets related to a specific hashtag. The collected data is stored in JSON format.

[0420] 2. Data preprocessing

[0421] The server receives the collected JSON data and stores it in a database. Next, it cleans up the data. For example, it filters out spam and advertisements using regular expressions and removes noise. This results in clean data suitable for analysis.

[0422] 3. Sentiment analysis

[0423] The server uses a generative AI model (e.g., BERT model) to perform sentiment analysis on clean data. It classifies text data as positive, negative, or neutral. This process is accelerated using CUDA-enabled GPUs.

[0424] 4. Calculation of satisfaction score

[0425] The server assigns a score to each review based on the sentiment analysis results. Positive reviews receive a score of +1, negative reviews receive -1, and neutral reviews receive 0. The total scores for each company are then tallied to calculate the satisfaction score.

[0426] 5. Competitive analysis

[0427] The server creates a ranking based on each company's satisfaction score. This ranking is ordered from highest to lowest score, visually representing the competitive landscape.

[0428] 6. Ranking Display

[0429] The smartphone displays ranking data sent from the server to the user. Users can check the real-time rankings through the smartphone application.

[0430] Specific examples and prompt statements

[0431] As a concrete example, the following shows the prompt text that a user would input to the generated AI model.

[0432] Example of a prompt:

[0433] "Use this app to perform sentiment analysis on tweets about Product A and display a competitor satisfaction ranking."

[0434] This prompt allows users to easily grasp real-time customer review results and competitive landscape for specific products or companies.

[0435] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0436] Step 1:

[0437] The device collects user review data from multiple websites, including social media and review sites. It uses the Twitter API to collect tweets related to specific hashtags. The input is a specific hashtag or keyword, and the output is the corresponding tweet data (in JSON format). This collected data is sent to a server and stored in a database.

[0438] Step 2:

[0439] The server receives JSON data sent from the terminal and stores it in the database. It then preprocesses the data. Specifically, it uses regular expressions to filter out spam tweets and advertisements, and removes noise. The input is raw JSON data, and the output is clean JSON data.

[0440] Step 3:

[0441] The server uses a generative AI model (e.g., the BERT model) to perform sentiment analysis on clean data. Specifically, it classifies text data into positive, negative, and neutral. The input is clean JSON data, and the output is data with the sentiment classification result for each tweet added. A CUDA-enabled GPU is used for the analysis.

[0442] Step 4:

[0443] The server assigns a score to each review based on the sentiment analysis results. Positive reviews receive a score of +1, negative reviews receive -1, and neutral reviews receive 0. The total scores for each company are aggregated to calculate the satisfaction score. The input is data including the sentiment classification results, and the output is the satisfaction score for each company.

[0444] Step 5:

[0445] The server creates a ranking of each company against its competitors based on their satisfaction scores. This ranking is ordered from highest score to lowest. The input is the satisfaction score for each company, and the output is a list of company rankings.

[0446] Step 6:

[0447] The smartphone displays ranking data transmitted from the server to the user. Users can check real-time rankings through the smartphone application. The input is company ranking data transmitted from the server, and the output is the ranking display in the user interface.

[0448] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0449] This invention combines a system that collects word-of-mouth data, performs preprocessing, sentiment analysis, calculates satisfaction scores, creates rankings, and displays rankings, with a sentiment engine that recognizes user emotions to perform even more advanced analysis and feedback.

[0450] overview

[0451] This system primarily consists of a server, terminals, and users. The emotion engine recognizes emotions from the user's voice or text input and sends this information to the server, which can then be reflected in the ranking display.

[0452] Data collection

[0453] The device collects customer review data from various websites (e.g., social media, review sites). The device retrieves tweets and reviews using specific keywords and hashtags and saves them in JSON format. For example, the device collects reviews from Twitter using the hashtag "product A review".

[0454] Data preprocessing

[0455] The server receives JSON data sent from the terminal and stores it in the database. The server uses regular expressions to remove noise and unnecessary information, creating a clean dataset. For example, it can filter out advertisements and spam messages.

[0456] sentiment analysis

[0457] The server uses the BERT model, a generated AI, to perform sentiment analysis on text data. The analysis results are classified as positive, negative, or neutral. For example, the text "This product is great!" is classified as positive.

[0458] Calculation of satisfaction score

[0459] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. The satisfaction scores for each company are compiled and stored in a database. For example, if company A has 15 positive reviews and 5 negative reviews, company A's satisfaction score will be +10.

[0460] Ranking creation

[0461] The server creates a ranking based on each company's satisfaction score. The total score for each company is sorted in descending order, and the ranking is generated.

[0462] Ranking display

[0463] The server displays the ranking results on a web dashboard that users access. The rankings are displayed visually using HTML and JavaScript. Users can access the dashboard in their browser and check the latest rankings. For example, it might show Company A in 1st place, Company B in 2nd place, and Company C in 3rd place.

[0464] Combination of emotional engines

[0465] Users access a web dashboard and enter feedback via text or voice. The emotion engine analyzes this feedback in real time and categorizes the user's emotions as positive, negative, or neutral. The results are sent to the server and reflected in the ranking display. For example, if a user enters feedback saying "This product is very good," the emotion engine will determine this to be positive and it will be reflected in the ranking display.

[0466] Specific example

[0467] As a concrete example, suppose a device collects 50 tweets from Twitter using the hashtag "product A review," of which 30 are positive, 15 are negative, and 5 are neutral. These tweets are pre-processed on a server, and sentiment analysis is performed using a generative AI. As a result, a satisfaction score of +15 is calculated for company A, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking.

[0468] In this way, a sophisticated system is realized that not only collects, analyzes, and displays user reviews, but also reflects user sentiment in real time.

[0469] The following describes the processing flow.

[0470] Step 1:

[0471] The device collects customer review data from various websites (e.g., social media, review sites). Specifically, the device uses specific keywords or hashtags (e.g., "product A review") to retrieve tweets from Twitter via GET requests. This tweet data is stored on the device in JSON format.

[0472] Step 2:

[0473] The device sends the collected review data to the server. The device uses a REST API to send the collected data to the server via a POST request. The sent data is temporarily stored in a database on the server.

[0474] Step 3:

[0475] The server preprocesses the received JSON data. The server uses regular expressions to remove noise (e.g., spam, advertisements) from the data, creating a clean dataset. For example, it removes advertising messages such as "Free sample giveaway!".

[0476] Step 4:

[0477] The server uses the BERT model, a generative AI, to perform sentiment analysis on pre-processed review data. It determines whether the text is positive, negative, or neutral and stores the results in a database. For example, a tweet like "This product is great" would be classified as positive.

[0478] Step 5:

[0479] The server assigns a satisfaction score to each review based on the sentiment analysis results. Positive opinions are given a score of +1, negative opinions a score of -1, and neutral opinions a score of 0, and these scores are stored in the database. For example, if there are 30 positive reviews and 15 negative reviews, Company A's score will be +15.

[0480] Step 6:

[0481] The server collects satisfaction scores from each company and creates a ranking. The scores for each company are sorted in descending order to generate a ranking list, which is then saved to the database. For example, Company A might be ranked 1st, Company B 2nd, and Company C 3rd.

[0482] Step 7:

[0483] The UI for displaying the server-generated rankings on the web dashboard is updated. The server uses HTML and JavaScript to generate a web page that visually displays the ranking list to the user. For example, company A is displayed as number 1 in the latest rankings.

[0484] Step 8:

[0485] Users access the web dashboard and then access the dashboard URL in their browser to view the rankings. Users can see the company satisfaction rankings in real time.

[0486] Step 9:

[0487] Users provide feedback. On the web dashboard, users offer their opinions on products and services through text or voice input. This input is sent to the sentiment engine.

[0488] Step 10:

[0489] The emotion engine analyzes user feedback in real time. It uses an emotion analysis algorithm to analyze voice and text input, classifying it as positive, negative, or neutral. The results are then sent to the server.

[0490] Step 11:

[0491] The server receives the results from the emotion engine and reflects them in the ranking display. For example, if new positive user feedback is reflected, Company A's satisfaction score will be increased by +1, and the ranking will be updated again.

[0492] This series of steps completes a system where data collection, analysis, display, and feedback integration are all automated.

[0493] (Example 2)

[0494] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0495] Traditional systems often require manual collection and analysis of customer review data, which is time-consuming and labor-intensive. Furthermore, it's difficult to reflect user feedback in real time, leading to issues with the accuracy of company satisfaction scores and rankings. Additionally, the inability to collect data from multiple sources simultaneously and perform comprehensive analysis makes accurate evaluation difficult.

[0496] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing, means for performing sentiment analysis using a generative AI model, means for calculating satisfaction scores for each company, means for creating rankings with competitors, means for displaying the rankings, means for analyzing user feedback and recognizing emotions, and means for reflecting the analysis results in the rankings. This automates all processes from word-of-mouth data collection and analysis to ranking creation and display, and further enables the reflection of real-time user feedback.

[0497] "Word-of-mouth data" refers to ratings and opinions posted by customers or users about products and services on online platforms (e.g., social media, review sites).

[0498] "Collection methods" refer to methods and devices for obtaining word-of-mouth data from various information sources.

[0499] "Preprocessing" refers to the process of removing noise and unnecessary information from collected data and converting it into a format suitable for analysis.

[0500] A "generative AI model" refers to a machine learning model used to automatically generate or analyze text and data.

[0501] "Sentiment analysis" refers to an analytical method for identifying emotions (positive, negative, neutral) from text data.

[0502] A "satisfaction score" refers to a score that quantifies the results of sentiment analysis and indicates an overall evaluation of a company or product.

[0503] A "ranking" refers to a list of companies or products based on customer satisfaction scores.

[0504] "Display means" refers to methods or devices that allow users to visually confirm results.

[0505] "Feedback" refers to the opinions and evaluations that users provide about a product or service.

[0506] An "emotion engine" refers to technology that identifies emotions in real time from user feedback.

[0507] This invention combines a system that collects, preprocesses, analyzes sentiment, calculates satisfaction scores, creates rankings, and displays rankings with a sentiment engine that recognizes user emotions, enabling more advanced analysis and feedback. This automates the entire process from collecting and analyzing review data to creating and displaying rankings, and also allows for the incorporation of real-time user feedback.

[0508] overview

[0509] This system primarily consists of a server, terminals, and users. The emotion engine recognizes emotions from the user's voice or text input and sends this information to the server, which can then be reflected in the ranking display.

[0510] Data collection

[0511] The device uses web scraping tools (e.g., Beautiful Soup, Scrapy) to collect user reviews from social media and review sites. For example, it might search for tweets using the hashtag "product A review" and save the data in JSON format. This allows for the unified collection of data from various sources.

[0512] Data preprocessing

[0513] The server receives JSON data sent from the terminal and stores it in a database (e.g., MySQL, PostgreSQL). Next, the server uses regular expressions and natural language processing (NLP) libraries (e.g., NLTK, SpaCy) to remove noise and unnecessary information, creating a clean dataset. This improves the quality of the data and allows for more accurate subsequent processing.

[0514] sentiment analysis

[0515] The server uses a generative AI model (e.g., BERT, GPT-3) to perform sentiment analysis on text data. Specifically, it classifies customer reviews into positive, negative, and neutral. For example, the text "This product is the best!" would be classified as positive. This allows for an understanding of general customer sentiment trends.

[0516] Calculation of satisfaction score

[0517] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. The satisfaction scores for each company are then compiled and stored in a database. For example, if company A receives 15 positive reviews and 5 negative reviews, company A's satisfaction score will be +10.

[0518] Ranking creation

[0519] The server creates a ranking based on customer satisfaction scores for each company. The total scores for each company are then sorted in descending order to generate the ranking. This makes it easy to compare companies with competitors and allows for quick information provision to users.

[0520] Ranking display

[0521] The server displays the ranking results on a web dashboard that users access. The rankings are displayed visually using HTML and JavaScript. Users can access the dashboard in their browser and check the latest rankings. For example, it might show Company A in 1st place, Company B in 2nd place, and Company C in 3rd place.

[0522] Combination of emotional engines

[0523] Users access a web dashboard and enter feedback via text or voice. The emotion engine analyzes this feedback in real time and categorizes the user's emotions as positive, negative, or neutral. The results are sent to the server and reflected in the ranking display. For example, if a user enters feedback saying "This product is very good," the emotion engine will determine this to be positive and it will be reflected in the ranking display.

[0524] Specific example

[0525] As a concrete example, suppose a device collects 50 tweets from Twitter using the hashtag "product A review," with 30 being positive, 15 negative, and 5 neutral. These tweets are preprocessed on a server, and sentiment analysis is performed using a generative AI model. As a result, a satisfaction score of +15 is calculated for company A, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking.

[0526] In this way, a sophisticated system is realized that not only collects, analyzes, and displays user reviews, but also reflects user sentiment in real time.

[0527] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0528] Step 1: Data Collection

[0529] The device uses a web scraping tool (e.g., Beautiful Soup, Scrapy) to collect user review data from social media and review sites. Input requires the URL of the target site and search keywords (e.g., "product A review"). Based on this input, the device extracts the necessary user review data from the website and saves it in JSON format. The output is the collected user review data (in JSON format).

[0530] Step 2: Data transmission

[0531] The device collects review data (in JSON format) and sends it to the server. The input requires the collected review data and the destination server address. The device then forwards the data to the server based on this input. The output is the review data sent to the server.

[0532] Step 3: Data Preprocessing

[0533] The server receives the JSON data and stores it in a database (e.g., MySQL, PostgreSQL). Next, the server uses regular expressions and natural language processing (NLP) libraries (e.g., NLTK, SpaCy) to remove noise and unwanted information, creating a clean dataset. The input is the received raw data (in JSON format). The output is a pre-processed, clean dataset.

[0534] Step 4: Sentiment Analysis

[0535] The server uses a generative AI model (e.g., BERT, GPT-3) to perform sentiment analysis on text data. Specifically, it classifies word-of-mouth text into positive, negative, and neutral. The input requires a pre-processed, clean dataset. The server uses this to perform sentiment analysis and assigns sentiment labels to each text. The output is the data with sentiment labels assigned.

[0536] Step 5: Calculating the Satisfaction Score

[0537] The server assigns a score to each review based on the sentiment analysis results. The input requires data with sentiment labels. Positive opinions are assigned a score of +1, negative opinions a score of -1, and neutral opinions a score of 0. The satisfaction scores for each company are then compiled and stored in a database. The output is the satisfaction score for each company.

[0538] Step 6: Create Rankings

[0539] The server creates a ranking based on customer satisfaction scores for each company. The input requires customer satisfaction scores for each company. The server sorts these scores in descending order to generate the ranking and saves this data to a database. The output is the ranking for each company.

[0540] Step 7: Display the rankings

[0541] The server displays the ranking results on a web dashboard accessed by the user. The input requires generated ranking data. The server uses HTML and JavaScript to visually display the rankings. The output is the latest ranking displayed on the dashboard.

[0542] Step 8: Gathering Feedback

[0543] The user accesses the web dashboard and enters feedback via text or voice. The input requires the user's feedback as text or voice data. The feedback is sent to the server. The output is the feedback data sent to the server.

[0544] Step 9: Analysis using the Emotion Engine

[0545] The emotion engine analyzes user feedback in real time and classifies the user's emotions as positive, negative, or neutral. The input is user feedback data. The output is feedback data with emotion labels attached.

[0546] Step 10: Reflection in rankings

[0547] The server receives the analysis results from the emotion engine and reflects them in the ranking display. The input requires feedback data with emotion labels. The server uses this data to update the rankings and reflects them in the dashboard in real time. The output is the updated rankings.

[0548] (Application Example 2)

[0549] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0550] Traditional word-of-mouth data analysis systems have struggled to reflect customers' real-time emotions, making immediate improvements in customer satisfaction impossible. Furthermore, systems that generate rankings based on a comprehensive evaluation of multiple ratings suffer from low variability due to their reliance on static data. As a result, the latest user emotions and feedback are not reflected in business decisions or marketing strategies, preventing a more accurate response to customer needs.

[0551] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing, means for performing sentiment analysis, means for calculating satisfaction scores for each company, means for creating rankings with competitors, means for displaying the rankings, means for receiving voice or text feedback from users and analyzing it using a sentiment engine, and means for reflecting the analysis results in the rankings in real time. This makes it possible to immediately analyze real-time user feedback and reflect it in the rankings.

[0552] "Word-of-mouth data" refers to a collection of digital opinions such as reviews, comments, and ratings posted by customers on the internet about products and services.

[0553] "Preprocessing" is the process of removing unnecessary information and noise from raw data and organizing it so that it can be analyzed accurately.

[0554] "Sentiment analysis" is a technique that analyzes the emotions contained in opinions and feedback within text data and classifies them into categories such as positive, negative, and neutral.

[0555] A "satisfaction score" is an index that quantifies customer satisfaction with a product or service based on customer feedback.

[0556] A "ranking" is a list that evaluates and ranks multiple companies or products based on criteria such as satisfaction scores.

[0557] An "emotion engine" is software or hardware that analyzes voice or text feedback from users in real time and classifies their emotions.

[0558] A "user" is someone who accesses the system, views user reviews, and provides feedback.

[0559] "Real-time" refers to a situation where data or events are processed and reflected immediately the moment they occur.

[0560] The system implementing this invention mainly consists of a server, a terminal, and a user. The specific roles of each component are described below.

[0561] server

[0562] The server plays a central role in collecting, pre-processing, and performing sentiment analysis on review data. Specifically, this includes the following methods:

[0563] 1. Data Collection Method: Receive word-of-mouth data sent from the device and store it in a database. The data is stored in JSON format.

[0564] 2. Preprocessing method: Regular expressions are used to remove noise and unwanted information, creating a clean dataset. For example, advertisements and spam messages are filtered out.

[0565] 3. Sentiment Analysis Method: The BERT model, a generative AI model, is used to analyze the sentiment of the collected text data. The analysis results are classified into positive, negative, and neutral.

[0566] 4. Satisfaction Score Calculation Method: Based on the results of sentiment analysis, a score is assigned to each review, and a total score is calculated for each company. For example, positive opinions are given a score of +1, negative opinions are given a score of -1, and neutral opinions are given a score of 0.

[0567] 5. Ranking creation method: The total score for each company is sorted in descending order, and a ranking is generated.

[0568] 6. Ranking Display Method: Ranking results will be displayed on a web dashboard. The display will be visually generated using HTML and JavaScript.

[0569] terminal

[0570] The device is responsible for collecting user review data and sending it to the server. Specifically, this includes the following methods:

[0571] 1. Data Collection Methods: Collect word-of-mouth data from multiple websites (e.g., social media, review sites). For example, obtain word-of-mouth data from Twitter using specific keywords or hashtags.

[0572] User

[0573] Users are responsible for accessing the system, providing feedback, and checking rankings. Specifically, this includes the following:

[0574] 1. Feedback input method: Access the web dashboard and enter feedback in text or voice.

[0575] 2. Emotion Analysis Method: The emotion engine analyzes user feedback in real time, classifies emotions as positive, negative, or neutral, and sends the results to the server.

[0576] 3. How to check rankings: Check the latest rankings on the web dashboard. Rankings are updated in real time.

[0577] Hardware and software to be used

[0578] Hardware:

[0579] Server (for data storage and processing)

[0580] Smartphone (for user feedback input)

[0581] software:

[0582] Transformers library (using the BERT model)

[0583] Requests Library (for data collection)

[0584] Regular expression library (for data preprocessing)

[0585] softmax function (Scipy library; used for calculating sentiment scores)

[0586] Specific example

[0587] Let's look at a concrete example. Suppose a device collects 50 tweets from Twitter using the hashtag "product A review," with 30 positive, 15 negative, and 5 neutral. These tweets are pre-processed on a server, and sentiment analysis is performed using a generative AI model. As a result, company A's satisfaction score is calculated to be +15, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking in real time.

[0588] Example of a prompt

[0589] "Analyze the sentiment of the following text: 'This product is great.' Classify the sentiment score as positive, negative, or neutral."

[0590] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0591] Step 1:

[0592] The device collects user review data from multiple websites (e.g., social media and review sites). For example, it retrieves reviews from Twitter using specific keywords or hashtags. The input is a specific keyword such as "product A review," and the output is user review data in JSON format. This data is initially saved for later analysis.

[0593] Step 2:

[0594] The server receives review data sent from the terminal and stores it in a database. The input is JSON data sent from the terminal, and the output is storage data stored in the database. A regular expression library is used to remove noise and unwanted information, such as advertisements and spam messages.

[0595] Step 3:

[0596] The server generates a clean dataset. The input is pre-processed review data using regular expressions, and the output is clean JSON formatted data. At this stage, the data is ready for analysis, so we move on to the next step. Specifically, we remove unnecessary strings and HTML tags using regular expressions.

[0597] Step 4:

[0598] The server performs sentiment analysis using the BERT model, a generative AI model. The input is each review text from the clean dataset, and the output is the sentiment score (positive, negative, neutral) for each text. Specifically, each text is input into the BERT model to obtain the sentiment score.

[0599] Step 5:

[0600] The server assigns a score to each review based on the sentiment analysis results. The input is the sentiment analysis results, and the output is the satisfaction score for each company. For example, positive opinions are given a score of +1, negative opinions a score of -1, and neutral opinions a score of 0. This allows the total score for each company to be calculated.

[0601] Step 6:

[0602] The server sorts the total scores for each company in descending order and generates a ranking. The input is the satisfaction score, and the output is the ranking for each company. For example, the company with the highest total score will be ranked first.

[0603] Step 7:

[0604] The server displays the ranking results on a web dashboard. The input is the generated ranking, and the output is the visually displayed ranking information. Using HTML and JavaScript, users can view the latest rankings in their browser.

[0605] Step 8:

[0606] Users access a web dashboard and enter feedback in text or voice. The input is the user's feedback in text or voice, and the output is the feedback information sent to the sentiment engine.

[0607] Step 9:

[0608] The emotion engine analyzes user feedback in real time. The input is user feedback, and the output is a real-time emotion score. Specifically, user feedback is input into a BERT model to obtain an emotion score.

[0609] Step 10:

[0610] The server reflects the analysis results in the rankings in real time. The input is the real-time sentiment score, and the output is the updated ranking. This ensures that the rankings display the latest user feedback immediately.

[0611] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0612] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0613] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0614] [Third Embodiment]

[0615] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0616] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0617] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0618] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0619] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0620] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0621] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0622] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0623] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0624] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0625] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0626] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0627] This system collects review data from multiple websites, preprocesses and performs sentiment analysis on it, calculates satisfaction scores for each company, and creates and displays rankings with competitors. The specific processing of this invention is as follows:

[0628] 1. Data Collection

[0629] The device collects customer review data from various websites (e.g., social media, review sites). For example, the device uses the Twitter API to collect tweets related to a specific hashtag and saves them in JSON format.

[0630] 2. Data preprocessing

[0631] The server receives JSON data sent from the terminal and stores it in the database. Next, the server removes noise. For example, it filters out spam tweets and advertisements using regular expressions. This results in clean data.

[0632] 3. Opinion analysis

[0633] The server uses generated AI to perform sentiment analysis on clean data. Specifically, it uses the BERT model to classify text as positive, negative, or neutral. For example, a tweet like "This product is great" would be classified as positive.

[0634] 4. Calculation of satisfaction score

[0635] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. Next, the total score for each company is tallied. For example, if company A has 15 positive opinions and 5 negative opinions, company A's satisfaction score will be +10.

[0636] 5. Competitive analysis

[0637] The server creates a ranking based on each company's satisfaction score. This ranking is ordered from highest score to lowest.

[0638] 6. Ranking Display

[0639] The server displays the rankings it generates in the user interface (UI). Users can access the web dashboard to check the latest rankings.

[0640] For example, the rankings for companies A, B, and C will be displayed as follows:

[0641] 1. Company C (Satisfaction Score: +12)

[0642] 2. Company A (Satisfaction Score: +10)

[0643] 3. Company B (Satisfaction Score: +8)

[0644] In this way, users can check customer satisfaction rankings for each company in real time and understand the competitive landscape. This system aggregates real customer feedback and visually displays company evaluations, helping companies quickly identify areas for improvement and their strengths.

[0645] The following describes the processing flow.

[0646] Step 1:

[0647] The device collects customer review data from various websites (e.g., social media, review sites). The device uses the Twitter API to send GET requests to retrieve tweets related to specific keywords or hashtags. The retrieved data is stored on the device in JSON format.

[0648] Step 2:

[0649] The device sends the collected review data to the server. The device uses a REST API to send the collected data to the server via a POST request. The sent data is temporarily stored in a database on the server.

[0650] Step 3:

[0651] The server preprocesses the received JSON data. Specifically, the server uses regular expressions to remove noise such as spam and advertisements from the data. It also removes unnecessary spaces and line breaks to normalize the text.

[0652] Step 4:

[0653] The server then performs sentiment analysis on the pre-processed data. Using the generative AI BERT model, it determines whether each review text is positive, negative, or neutral. For example, a review saying "This product is great" would be classified as positive.

[0654] Step 5:

[0655] The server assigns a satisfaction score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. This score is stored in the database.

[0656] Step 6:

[0657] The server calculates a total score for each company based on the satisfaction scores. For example, if company A receives 15 positive comments and 5 negative comments, company A's satisfaction score will be +10. The compiled scores are stored in a database.

[0658] Step 7:

[0659] The server creates a ranking based on each company's satisfaction score. The scores for each company are sorted in descending order, and a ranking list is generated. This ranking list is stored in a database.

[0660] Step 8:

[0661] The server updates the UI to display the latest rankings on the web dashboard accessed by the user. HTML and JavaScript are used to generate a web page that visually displays the ranking list to the user.

[0662] Step 9:

[0663] Users access the web dashboard from their browser and view updated rankings in real time. By accessing the dashboard URL in their browser, users can quickly grasp comparisons and statuses of companies.

[0664] This series of steps results in an automated system that handles everything from data collection to analysis and display.

[0665] (Example 1)

[0666] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0667] There is a need for a system that can accurately evaluate a company's customer satisfaction and instantly compare it with competitors. In particular, there is a lack of efficient methods for collecting and analyzing review data from multiple websites. In addition, accuracy and efficiency are challenges in the process of analyzing the sentiment of this data using AI models and calculating satisfaction scores for each company.

[0668] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0669] In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing the data, means for performing sentiment analysis using a generative AI model, means for calculating satisfaction scores for each company, means for creating rankings with competitors, and means for displaying the rankings. This makes it possible to efficiently collect and pre-process word-of-mouth data from multiple websites and perform highly accurate sentiment analysis using a generative AI model. Furthermore, it is possible to quickly calculate satisfaction scores for each company and visually display real-time rankings with competitors.

[0670] "Word-of-mouth data" refers to text data of opinions and impressions about products and services posted by users on online platforms.

[0671] "Preprocessing" is the process of removing noise from collected data and preparing it in a format that is easy to analyze.

[0672] A "generative AI model" is an artificial intelligence model trained using machine learning, and is particularly used for sentiment analysis and text classification.

[0673] "Sentiment analysis" is the process of analyzing the text contained in word-of-mouth data and classifying its content as either positive, negative, or neutral.

[0674] A "satisfaction score" is an indicator of customer satisfaction for each company, calculated by assigning a score to each review data based on the results of sentiment analysis and then aggregating these scores.

[0675] "Ranking creation" is the process of arranging companies in descending order based on their customer satisfaction scores, thereby enabling visual comparisons with competitors.

[0676] "Ranking display" refers to displaying the created ranking results on the user interface so that users can view them.

[0677] This invention is a system that collects, preprocesses, and performs sentiment analysis on word-of-mouth data, calculates a satisfaction score for each company based on the results, and creates and displays a ranking with competitors. The following shows how this system is specifically implemented.

[0678] In implementing this system, the terminal first plays the role of collecting word-of-mouth data. The terminal uses a Python script to access the Twitter API. Specifically, the terminal collects tweets related to a specific hashtag via the Twitter API and saves them in JSON format. In this case, the terminal hardware is a client PC, and the software used is Python and the Twitter API.

[0679] The server receives JSON data sent from the terminal and stores it in a database. Since the stored data requires preprocessing, the server uses regular expressions to remove noise. By using regular expressions, spam tweets and advertisements are filtered out, generating clean data. The server hardware consists of a server machine, and the software used is Python and its regular expression package.

[0680] Next, the server performs sentiment analysis. For this purpose, it uses a generative AI model, specifically the BERT model. The server uses the Hugging Face Transformers library to classify tweets as positive, negative, or neutral. The server hardware continues to be a server machine.

[0681] Based on the sentiment analysis results, the server assigns a score to each review. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. These scores are then aggregated to calculate a satisfaction score for each company. Python and its aggregation library, NumPy, are used to aggregate the scores.

[0682] The server creates a ranking based on the aggregated scores. This ranking is sorted in descending order based on each company's satisfaction score. The server hardware and software will continue to use Python.

[0683] Finally, the user accesses a web dashboard to view the rankings. The server displays the aforementioned rankings in a user interface (UI). This UI is built using HTML, CSS, JavaScript, and a backend framework (e.g., Django or Flask).

[0684] As a concrete example, here is an example of a prompt message that uses the Twitter API to collect review data for the hashtag "good product":

[0685] Collect the latest tweets related to the hashtag "good product" and perform a sentiment analysis on them. Classify them as positive, negative, or neutral, and provide the number of each.

[0686] As described above, the present invention is a system that efficiently collects word-of-mouth data from many websites and performs sentiment analysis using a generative AI model, thereby enabling accurate evaluation of customer satisfaction for each company and real-time comparison with competitors.

[0687] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0688] Step 1: Data Collection

[0689] The device collects word-of-mouth data from various websites (e.g., social media, review sites). Specifically, the device uses a Python script to access the Twitter API. The device collects tweets related to a specific hashtag, "good product," and saves them in JSON format. The input is a specific hashtag, and the output is a JSON file of tweet data related to that hashtag. In terms of operation, it uses the Twitter API to collect tweets and saves them to local storage in JSON format.

[0690] Step 2: Data Preprocessing

[0691] The server receives JSON data sent from the terminal and stores it in a database. Next, the server uses regular expressions to remove noise from the data. For example, it filters out spam tweets and advertisements to obtain clean data. The input is collected tweet data in JSON format, and the output is clean tweet data with noise removed. Specifically, it uses a regular expression package to filter out advertising links and spam messages.

[0692] Step 3: Sentiment Analysis

[0693] The server generates clean, pre-processed data and performs sentiment analysis using an AI model. Specifically, it uses the BERT model with the Hugging Face Transformers library to classify each tweet as positive, negative, or neutral. The input is clean data, and the output is the result of the sentiment analysis for each tweet. In terms of operation, it calls the sentiment analysis model and classifies the sentiment of each tweet.

[0694] Step 4: Calculating the satisfaction score

[0695] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. Then, it compiles the total score for each company. The input is the result of the sentiment analysis, and the output is the satisfaction score for each company. Specifically, it compiles the sentiment score for each tweet and calculates the total score for each company.

[0696] Step 5: Competitive Analysis

[0697] The server creates a ranking of each company based on their satisfaction scores. The input is the satisfaction score for each company, and the output is the ranking. Specifically, it sorts the scores for each company in descending order and generates a ranking list.

[0698] Step 6: Display the rankings

[0699] Users can access a web dashboard to view their rankings. The server displays the generated rankings in the user interface (UI). The input is ranking data, and the output is a visual display of the rankings for the user. Specifically, HTML, CSS, and JavaScript are used to display the rankings on the web page, allowing users to visually confirm them.

[0700] (Application Example 1)

[0701] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0702] In recent years, consumer reviews have come to have a significant impact on a company's reputation. However, manually analyzing vast amounts of review data is difficult, making it challenging to grasp a company's relative ranking against its competitors in real time. To solve this problem, a system is needed that efficiently collects, preprocesses, and analyzes review data, calculates satisfaction scores for each company, and visualizes competitor rankings on smartphones.

[0703] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0704] In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing the data, and means for performing sentiment analysis. This makes it possible to efficiently collect consumer word-of-mouth data, process it into clean data, and perform sentiment classification using generative AI. Furthermore, by creating a ranking of competitors based on the calculated satisfaction score and displaying it on a smartphone, companies can grasp the competitive situation in real time.

[0705] "Word-of-mouth data" refers to reviews and comments posted by customers and users on online platforms regarding products and services.

[0706] "Preprocessing" refers to the process of removing noise from collected data and converting it into a format suitable for analysis.

[0707] "Sentiment analysis" is a method of analyzing text data and classifying its emotional tendencies (positive, negative, neutral).

[0708] The "satisfaction score" is a numerical value that represents an overall evaluation, calculated by assigning positive scores to positive reviews and negative scores to negative reviews based on the results of sentiment analysis.

[0709] A "ranking" refers to a list that ranks companies based on satisfaction scores and allows for comparison with competitors.

[0710] "A means of displaying rankings on a smartphone" refers to a function that visually displays customer satisfaction rankings for each company on the smartphone screen.

[0711] "Generative AI" is a general term for artificial intelligence that generates and analyzes text, images, and other data, and is a technology that enables highly accurate analysis, particularly in natural language processing.

[0712] This invention is a system that consistently performs the collection, preprocessing, sentiment analysis, satisfaction score calculation, competitor analysis, and ranking display of word-of-mouth data. It is implemented using a server, a smartphone, and a generative AI model.

[0713] The system program follows the following main processing steps:

[0714] 1. Data Collection

[0715] The server collects user review data from multiple websites, including social media and review sites. For example, it uses the Twitter API to retrieve tweets related to a specific hashtag. The collected data is stored in JSON format.

[0716] 2. Data preprocessing

[0717] The server receives the collected JSON data and stores it in a database. Next, it cleans up the data. For example, it filters out spam and advertisements using regular expressions and removes noise. This results in clean data suitable for analysis.

[0718] 3. Sentiment analysis

[0719] The server uses a generative AI model (e.g., BERT model) to perform sentiment analysis on clean data. It classifies text data as positive, negative, or neutral. This process is accelerated using CUDA-enabled GPUs.

[0720] 4. Calculation of satisfaction score

[0721] The server assigns a score to each review based on the sentiment analysis results. Positive reviews receive a score of +1, negative reviews receive -1, and neutral reviews receive 0. The total scores for each company are then tallied to calculate the satisfaction score.

[0722] 5. Competitive analysis

[0723] The server creates a ranking based on each company's satisfaction score. This ranking is ordered from highest to lowest score, visually representing the competitive landscape.

[0724] 6. Ranking Display

[0725] The smartphone displays ranking data sent from the server to the user. Users can check the real-time rankings through the smartphone application.

[0726] Specific examples and prompt statements

[0727] As a concrete example, the following shows the prompt text that a user would input to the generated AI model.

[0728] Example of a prompt:

[0729] "Use this app to perform sentiment analysis on tweets about Product A and display a competitor satisfaction ranking."

[0730] This prompt allows users to easily grasp real-time customer review results and competitive landscape for specific products or companies.

[0731] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0732] Step 1:

[0733] The device collects user review data from multiple websites, including social media and review sites. It uses the Twitter API to collect tweets related to specific hashtags. The input is a specific hashtag or keyword, and the output is the corresponding tweet data (in JSON format). This collected data is sent to a server and stored in a database.

[0734] Step 2:

[0735] The server receives JSON data sent from the terminal and stores it in the database. It then preprocesses the data. Specifically, it uses regular expressions to filter out spam tweets and advertisements, and removes noise. The input is raw JSON data, and the output is clean JSON data.

[0736] Step 3:

[0737] The server uses a generative AI model (e.g., the BERT model) to perform sentiment analysis on clean data. Specifically, it classifies text data into positive, negative, and neutral. The input is clean JSON data, and the output is data with the sentiment classification result for each tweet added. A CUDA-enabled GPU is used for the analysis.

[0738] Step 4:

[0739] The server assigns a score to each review based on the sentiment analysis results. Positive reviews receive a score of +1, negative reviews receive -1, and neutral reviews receive 0. The total scores for each company are aggregated to calculate the satisfaction score. The input is data including the sentiment classification results, and the output is the satisfaction score for each company.

[0740] Step 5:

[0741] The server creates a ranking of each company against its competitors based on their satisfaction scores. This ranking is ordered from highest score to lowest. The input is the satisfaction score for each company, and the output is a list of company rankings.

[0742] Step 6:

[0743] The smartphone displays ranking data transmitted from the server to the user. Users can check real-time rankings through the smartphone application. The input is company ranking data transmitted from the server, and the output is the ranking display in the user interface.

[0744] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0745] This invention combines a system that collects word-of-mouth data, performs preprocessing, sentiment analysis, calculates satisfaction scores, creates rankings, and displays rankings, with a sentiment engine that recognizes user emotions to perform even more advanced analysis and feedback.

[0746] overview

[0747] This system primarily consists of a server, terminals, and users. The emotion engine recognizes emotions from the user's voice or text input and sends this information to the server, which can then be reflected in the ranking display.

[0748] Data collection

[0749] The device collects customer review data from various websites (e.g., social media, review sites). The device retrieves tweets and reviews using specific keywords and hashtags and saves them in JSON format. For example, the device collects reviews from Twitter using the hashtag "product A review".

[0750] Data preprocessing

[0751] The server receives JSON data sent from the terminal and stores it in the database. The server uses regular expressions to remove noise and unnecessary information, creating a clean dataset. For example, it can filter out advertisements and spam messages.

[0752] sentiment analysis

[0753] The server uses the BERT model, a generated AI, to perform sentiment analysis on text data. The analysis results are classified as positive, negative, or neutral. For example, the text "This product is great!" is classified as positive.

[0754] Calculation of satisfaction score

[0755] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. The satisfaction scores for each company are compiled and stored in a database. For example, if company A has 15 positive reviews and 5 negative reviews, company A's satisfaction score will be +10.

[0756] Ranking creation

[0757] The server creates a ranking based on each company's satisfaction score. The total score for each company is sorted in descending order, and the ranking is generated.

[0758] Ranking display

[0759] The server displays the ranking results on a web dashboard that users access. The rankings are displayed visually using HTML and JavaScript. Users can access the dashboard in their browser and check the latest rankings. For example, it might show Company A in 1st place, Company B in 2nd place, and Company C in 3rd place.

[0760] Combination of emotional engines

[0761] Users access a web dashboard and enter feedback via text or voice. The emotion engine analyzes this feedback in real time and categorizes the user's emotions as positive, negative, or neutral. The results are sent to the server and reflected in the ranking display. For example, if a user enters feedback saying "This product is very good," the emotion engine will determine this to be positive and it will be reflected in the ranking display.

[0762] Specific example

[0763] As a concrete example, suppose a device collects 50 tweets from Twitter using the hashtag "product A review," of which 30 are positive, 15 are negative, and 5 are neutral. These tweets are pre-processed on a server, and sentiment analysis is performed using a generative AI. As a result, a satisfaction score of +15 is calculated for company A, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking.

[0764] In this way, a sophisticated system is realized that not only collects, analyzes, and displays user reviews, but also reflects user sentiment in real time.

[0765] The following describes the processing flow.

[0766] Step 1:

[0767] The device collects customer review data from various websites (e.g., social media, review sites). Specifically, the device uses specific keywords or hashtags (e.g., "product A review") to retrieve tweets from Twitter via GET requests. This tweet data is stored on the device in JSON format.

[0768] Step 2:

[0769] The device sends the collected review data to the server. The device uses a REST API to send the collected data to the server via a POST request. The sent data is temporarily stored in a database on the server.

[0770] Step 3:

[0771] The server preprocesses the received JSON data. The server uses regular expressions to remove noise (e.g., spam, advertisements) from the data, creating a clean dataset. For example, it removes advertising messages such as "Free sample giveaway!".

[0772] Step 4:

[0773] The server uses the BERT model, a generative AI, to perform sentiment analysis on pre-processed review data. It determines whether the text is positive, negative, or neutral and stores the results in a database. For example, a tweet like "This product is great" would be classified as positive.

[0774] Step 5:

[0775] The server assigns a satisfaction score to each review based on the sentiment analysis results. Positive opinions are given a score of +1, negative opinions a score of -1, and neutral opinions a score of 0, and these scores are stored in the database. For example, if there are 30 positive reviews and 15 negative reviews, Company A's score will be +15.

[0776] Step 6:

[0777] The server collects satisfaction scores from each company and creates a ranking. The scores for each company are sorted in descending order to generate a ranking list, which is then saved to the database. For example, Company A might be ranked 1st, Company B 2nd, and Company C 3rd.

[0778] Step 7:

[0779] The UI for displaying the server-generated rankings on the web dashboard is updated. The server uses HTML and JavaScript to generate a web page that visually displays the ranking list to the user. For example, company A is displayed as number 1 in the latest rankings.

[0780] Step 8:

[0781] Users access the web dashboard and then access the dashboard URL in their browser to view the rankings. Users can see the company satisfaction rankings in real time.

[0782] Step 9:

[0783] Users provide feedback. On the web dashboard, users offer their opinions on products and services through text or voice input. This input is sent to the sentiment engine.

[0784] Step 10:

[0785] The emotion engine analyzes user feedback in real time. It uses an emotion analysis algorithm to analyze voice and text input, classifying it as positive, negative, or neutral. The results are then sent to the server.

[0786] Step 11:

[0787] The server receives the results from the emotion engine and reflects them in the ranking display. For example, if new positive user feedback is reflected, Company A's satisfaction score will be increased by +1, and the ranking will be updated again.

[0788] This series of steps completes a system where data collection, analysis, display, and feedback integration are all automated.

[0789] (Example 2)

[0790] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0791] Traditional systems often require manual collection and analysis of customer review data, which is time-consuming and labor-intensive. Furthermore, it's difficult to reflect user feedback in real time, leading to issues with the accuracy of company satisfaction scores and rankings. Additionally, the inability to collect data from multiple sources simultaneously and perform comprehensive analysis makes accurate evaluation difficult.

[0792] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing, means for performing sentiment analysis using a generative AI model, means for calculating satisfaction scores for each company, means for creating rankings with competitors, means for displaying the rankings, means for analyzing user feedback and recognizing emotions, and means for reflecting the analysis results in the rankings. This automates all processes from word-of-mouth data collection and analysis to ranking creation and display, and further enables the reflection of real-time user feedback.

[0793] "Word-of-mouth data" refers to ratings and opinions posted by customers or users about products and services on online platforms (e.g., social media, review sites).

[0794] "Collection methods" refer to methods and devices for obtaining word-of-mouth data from various information sources.

[0795] "Preprocessing" refers to the process of removing noise and unnecessary information from collected data and converting it into a format suitable for analysis.

[0796] A "generative AI model" refers to a machine learning model used to automatically generate or analyze text and data.

[0797] "Sentiment analysis" refers to an analytical method for identifying emotions (positive, negative, neutral) from text data.

[0798] A "satisfaction score" refers to a score that quantifies the results of sentiment analysis and indicates an overall evaluation of a company or product.

[0799] A "ranking" refers to a list of companies or products based on customer satisfaction scores.

[0800] "Display means" refers to methods or devices that allow users to visually confirm results.

[0801] "Feedback" refers to the opinions and evaluations that users provide about a product or service.

[0802] An "emotion engine" refers to technology that identifies emotions in real time from user feedback.

[0803] This invention combines a system that collects, preprocesses, analyzes sentiment, calculates satisfaction scores, creates rankings, and displays rankings with a sentiment engine that recognizes user emotions, enabling more advanced analysis and feedback. This automates the entire process from collecting and analyzing review data to creating and displaying rankings, and also allows for the incorporation of real-time user feedback.

[0804] overview

[0805] This system primarily consists of a server, terminals, and users. The emotion engine recognizes emotions from the user's voice or text input and sends this information to the server, which can then be reflected in the ranking display.

[0806] Data collection

[0807] The device uses web scraping tools (e.g., Beautiful Soup, Scrapy) to collect user reviews from social media and review sites. For example, it might search for tweets using the hashtag "product A review" and save the data in JSON format. This allows for the unified collection of data from various sources.

[0808] Data preprocessing

[0809] The server receives JSON data sent from the terminal and stores it in a database (e.g., MySQL, PostgreSQL). Next, the server uses regular expressions and natural language processing (NLP) libraries (e.g., NLTK, SpaCy) to remove noise and unnecessary information, creating a clean dataset. This improves the quality of the data and allows for more accurate subsequent processing.

[0810] sentiment analysis

[0811] The server uses a generative AI model (e.g., BERT, GPT-3) to perform sentiment analysis on text data. Specifically, it classifies customer reviews into positive, negative, and neutral. For example, the text "This product is the best!" would be classified as positive. This allows for an understanding of general customer sentiment trends.

[0812] Calculation of satisfaction score

[0813] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. The satisfaction scores for each company are then compiled and stored in a database. For example, if company A receives 15 positive reviews and 5 negative reviews, company A's satisfaction score will be +10.

[0814] Ranking creation

[0815] The server creates a ranking based on customer satisfaction scores for each company. The total scores for each company are then sorted in descending order to generate the ranking. This makes it easy to compare companies with competitors and allows for quick information provision to users.

[0816] Ranking display

[0817] The server displays the ranking results on a web dashboard that users access. The rankings are displayed visually using HTML and JavaScript. Users can access the dashboard in their browser and check the latest rankings. For example, it might show Company A in 1st place, Company B in 2nd place, and Company C in 3rd place.

[0818] Combination of emotional engines

[0819] Users access a web dashboard and enter feedback via text or voice. The emotion engine analyzes this feedback in real time and categorizes the user's emotions as positive, negative, or neutral. The results are sent to the server and reflected in the ranking display. For example, if a user enters feedback saying "This product is very good," the emotion engine will determine this to be positive and it will be reflected in the ranking display.

[0820] Specific example

[0821] As a concrete example, suppose a device collects 50 tweets from Twitter using the hashtag "product A review," with 30 being positive, 15 negative, and 5 neutral. These tweets are preprocessed on a server, and sentiment analysis is performed using a generative AI model. As a result, a satisfaction score of +15 is calculated for company A, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking.

[0822] In this way, a sophisticated system is realized that not only collects, analyzes, and displays user reviews, but also reflects user sentiment in real time.

[0823] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0824] Step 1: Data Collection

[0825] The device uses a web scraping tool (e.g., Beautiful Soup, Scrapy) to collect user review data from social media and review sites. Input requires the URL of the target site and search keywords (e.g., "product A review"). Based on this input, the device extracts the necessary user review data from the website and saves it in JSON format. The output is the collected user review data (in JSON format).

[0826] Step 2: Data transmission

[0827] The device collects review data (in JSON format) and sends it to the server. The input requires the collected review data and the destination server address. The device then forwards the data to the server based on this input. The output is the review data sent to the server.

[0828] Step 3: Data Preprocessing

[0829] The server receives the JSON data and stores it in a database (e.g., MySQL, PostgreSQL). Next, the server uses regular expressions and natural language processing (NLP) libraries (e.g., NLTK, SpaCy) to remove noise and unwanted information, creating a clean dataset. The input is the received raw data (in JSON format). The output is a pre-processed, clean dataset.

[0830] Step 4: Sentiment Analysis

[0831] The server uses a generative AI model (e.g., BERT, GPT-3) to perform sentiment analysis on text data. Specifically, it classifies word-of-mouth text into positive, negative, and neutral. The input requires a pre-processed, clean dataset. The server uses this to perform sentiment analysis and assigns sentiment labels to each text. The output is the data with sentiment labels assigned.

[0832] Step 5: Calculating the Satisfaction Score

[0833] The server assigns a score to each review based on the sentiment analysis results. The input requires data with sentiment labels. Positive opinions are assigned a score of +1, negative opinions a score of -1, and neutral opinions a score of 0. The satisfaction scores for each company are then compiled and stored in a database. The output is the satisfaction score for each company.

[0834] Step 6: Create Rankings

[0835] The server creates a ranking based on customer satisfaction scores for each company. The input requires customer satisfaction scores for each company. The server sorts these scores in descending order to generate the ranking and saves this data to a database. The output is the ranking for each company.

[0836] Step 7: Display the rankings

[0837] The server displays the ranking results on a web dashboard accessed by the user. The input requires generated ranking data. The server uses HTML and JavaScript to visually display the rankings. The output is the latest ranking displayed on the dashboard.

[0838] Step 8: Gathering Feedback

[0839] The user accesses the web dashboard and enters feedback via text or voice. The input requires the user's feedback as text or voice data. The feedback is sent to the server. The output is the feedback data sent to the server.

[0840] Step 9: Analysis using the Emotion Engine

[0841] The emotion engine analyzes user feedback in real time and classifies the user's emotions as positive, negative, or neutral. The input is user feedback data. The output is feedback data with emotion labels attached.

[0842] Step 10: Reflection in rankings

[0843] The server receives the analysis results from the emotion engine and reflects them in the ranking display. The input requires feedback data with emotion labels. The server uses this data to update the rankings and reflects them in the dashboard in real time. The output is the updated rankings.

[0844] (Application Example 2)

[0845] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0846] Traditional word-of-mouth data analysis systems have struggled to reflect customers' real-time emotions, making immediate improvements in customer satisfaction impossible. Furthermore, systems that generate rankings based on a comprehensive evaluation of multiple ratings suffer from low variability due to their reliance on static data. As a result, the latest user emotions and feedback are not reflected in business decisions or marketing strategies, preventing a more accurate response to customer needs.

[0847] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing, means for performing sentiment analysis, means for calculating satisfaction scores for each company, means for creating rankings with competitors, means for displaying the rankings, means for receiving voice or text feedback from users and analyzing it using a sentiment engine, and means for reflecting the analysis results in the rankings in real time. This makes it possible to immediately analyze real-time user feedback and reflect it in the rankings.

[0848] "Word-of-mouth data" refers to a collection of digital opinions such as reviews, comments, and ratings posted by customers on the internet about products and services.

[0849] "Preprocessing" is the process of removing unnecessary information and noise from raw data and organizing it so that it can be analyzed accurately.

[0850] "Sentiment analysis" is a technique that analyzes the emotions contained in opinions and feedback within text data and classifies them into categories such as positive, negative, and neutral.

[0851] A "satisfaction score" is an index that quantifies customer satisfaction with a product or service based on customer feedback.

[0852] A "ranking" is a list that evaluates and ranks multiple companies or products based on criteria such as satisfaction scores.

[0853] An "emotion engine" is software or hardware that analyzes voice or text feedback from users in real time and classifies their emotions.

[0854] A "user" is someone who accesses the system, views user reviews, and provides feedback.

[0855] "Real-time" refers to a situation where data or events are processed and reflected immediately the moment they occur.

[0856] The system implementing this invention mainly consists of a server, a terminal, and a user. The specific roles of each component are described below.

[0857] server

[0858] The server plays a central role in collecting, pre-processing, and performing sentiment analysis on review data. Specifically, this includes the following methods:

[0859] 1. Data Collection Method: Receive word-of-mouth data sent from the device and store it in a database. The data is stored in JSON format.

[0860] 2. Preprocessing method: Regular expressions are used to remove noise and unwanted information, creating a clean dataset. For example, advertisements and spam messages are filtered out.

[0861] 3. Sentiment Analysis Method: The BERT model, a generative AI model, is used to analyze the sentiment of the collected text data. The analysis results are classified into positive, negative, and neutral.

[0862] 4. Satisfaction Score Calculation Method: Based on the results of sentiment analysis, a score is assigned to each review, and a total score is calculated for each company. For example, positive opinions are given a score of +1, negative opinions are given a score of -1, and neutral opinions are given a score of 0.

[0863] 5. Ranking creation method: The total score for each company is sorted in descending order, and a ranking is generated.

[0864] 6. Ranking Display Method: Ranking results will be displayed on a web dashboard. The display will be visually generated using HTML and JavaScript.

[0865] terminal

[0866] The device is responsible for collecting user review data and sending it to the server. Specifically, this includes the following methods:

[0867] 1. Data Collection Methods: Collect word-of-mouth data from multiple websites (e.g., social media, review sites). For example, obtain word-of-mouth data from Twitter using specific keywords or hashtags.

[0868] User

[0869] Users are responsible for accessing the system, providing feedback, and checking rankings. Specifically, this includes the following:

[0870] 1. Feedback input method: Access the web dashboard and enter feedback in text or voice.

[0871] 2. Emotion Analysis Method: The emotion engine analyzes user feedback in real time, classifies emotions as positive, negative, or neutral, and sends the results to the server.

[0872] 3. How to check rankings: Check the latest rankings on the web dashboard. Rankings are updated in real time.

[0873] Hardware and software to be used

[0874] Hardware:

[0875] Server (for data storage and processing)

[0876] Smartphone (for user feedback input)

[0877] software:

[0878] Transformers library (using the BERT model)

[0879] Requests Library (for data collection)

[0880] Regular expression library (for data preprocessing)

[0881] softmax function (Scipy library; used for calculating sentiment scores)

[0882] Specific example

[0883] Let's look at a concrete example. Suppose a device collects 50 tweets from Twitter using the hashtag "product A review," with 30 positive, 15 negative, and 5 neutral. These tweets are pre-processed on a server, and sentiment analysis is performed using a generative AI model. As a result, company A's satisfaction score is calculated to be +15, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking in real time.

[0884] Example of a prompt

[0885] "Analyze the sentiment of the following text: 'This product is great.' Classify the sentiment score as positive, negative, or neutral."

[0886] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0887] Step 1:

[0888] The device collects user review data from multiple websites (e.g., social media and review sites). For example, it retrieves reviews from Twitter using specific keywords or hashtags. The input is a specific keyword such as "product A review," and the output is user review data in JSON format. This data is initially saved for later analysis.

[0889] Step 2:

[0890] The server receives review data sent from the terminal and stores it in a database. The input is JSON data sent from the terminal, and the output is storage data stored in the database. A regular expression library is used to remove noise and unwanted information, such as advertisements and spam messages.

[0891] Step 3:

[0892] The server generates a clean dataset. The input is pre-processed review data using regular expressions, and the output is clean JSON formatted data. At this stage, the data is ready for analysis, so we move on to the next step. Specifically, we remove unnecessary strings and HTML tags using regular expressions.

[0893] Step 4:

[0894] The server performs sentiment analysis using the BERT model, a generative AI model. The input is each review text from the clean dataset, and the output is the sentiment score (positive, negative, neutral) for each text. Specifically, each text is input into the BERT model to obtain the sentiment score.

[0895] Step 5:

[0896] The server assigns a score to each review based on the sentiment analysis results. The input is the sentiment analysis results, and the output is the satisfaction score for each company. For example, positive opinions are given a score of +1, negative opinions a score of -1, and neutral opinions a score of 0. This allows the total score for each company to be calculated.

[0897] Step 6:

[0898] The server sorts the total scores for each company in descending order and generates a ranking. The input is the satisfaction score, and the output is the ranking for each company. For example, the company with the highest total score will be ranked first.

[0899] Step 7:

[0900] The server displays the ranking results on a web dashboard. The input is the generated ranking, and the output is the visually displayed ranking information. Using HTML and JavaScript, users can view the latest rankings in their browser.

[0901] Step 8:

[0902] Users access a web dashboard and enter feedback in text or voice. The input is the user's feedback in text or voice, and the output is the feedback information sent to the sentiment engine.

[0903] Step 9:

[0904] The emotion engine analyzes user feedback in real time. The input is user feedback, and the output is a real-time emotion score. Specifically, user feedback is input into a BERT model to obtain an emotion score.

[0905] Step 10:

[0906] The server reflects the analysis results in the rankings in real time. The input is the real-time sentiment score, and the output is the updated ranking. This ensures that the rankings display the latest user feedback immediately.

[0907] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0908] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0909] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0910] [Fourth Embodiment]

[0911] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0912] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0913] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0914] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0915] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0916] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0917] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0918] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0919] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0920] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0921] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0922] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0923] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0924] This system collects review data from multiple websites, preprocesses and performs sentiment analysis on it, calculates satisfaction scores for each company, and creates and displays rankings with competitors. The specific processing of this invention is as follows:

[0925] 1. Data Collection

[0926] The device collects customer review data from various websites (e.g., social media, review sites). For example, the device uses the Twitter API to collect tweets related to a specific hashtag and saves them in JSON format.

[0927] 2. Data preprocessing

[0928] The server receives JSON data sent from the terminal and stores it in the database. Next, the server removes noise. For example, it filters out spam tweets and advertisements using regular expressions. This results in clean data.

[0929] 3. Opinion analysis

[0930] The server uses generated AI to perform sentiment analysis on clean data. Specifically, it uses the BERT model to classify text as positive, negative, or neutral. For example, a tweet like "This product is great" would be classified as positive.

[0931] 4. Calculation of satisfaction score

[0932] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. Next, the total score for each company is tallied. For example, if company A has 15 positive opinions and 5 negative opinions, company A's satisfaction score will be +10.

[0933] 5. Competitive analysis

[0934] The server creates a ranking based on each company's satisfaction score. This ranking is ordered from highest score to lowest.

[0935] 6. Ranking Display

[0936] The server displays the rankings it generates in the user interface (UI). Users can access the web dashboard to check the latest rankings.

[0937] For example, the rankings for companies A, B, and C will be displayed as follows:

[0938] 1. Company C (Satisfaction Score: +12)

[0939] 2. Company A (Satisfaction Score: +10)

[0940] 3. Company B (Satisfaction Score: +8)

[0941] In this way, users can check customer satisfaction rankings for each company in real time and understand the competitive landscape. This system aggregates real customer feedback and visually displays company evaluations, helping companies quickly identify areas for improvement and their strengths.

[0942] The following describes the processing flow.

[0943] Step 1:

[0944] The device collects customer review data from various websites (e.g., social media, review sites). The device uses the Twitter API to send GET requests to retrieve tweets related to specific keywords or hashtags. The retrieved data is stored on the device in JSON format.

[0945] Step 2:

[0946] The device sends the collected review data to the server. The device uses a REST API to send the collected data to the server via a POST request. The sent data is temporarily stored in a database on the server.

[0947] Step 3:

[0948] The server preprocesses the received JSON data. Specifically, the server uses regular expressions to remove noise such as spam and advertisements from the data. It also removes unnecessary spaces and line breaks to normalize the text.

[0949] Step 4:

[0950] The server then performs sentiment analysis on the pre-processed data. Using the generative AI BERT model, it determines whether each review text is positive, negative, or neutral. For example, a review saying "This product is great" would be classified as positive.

[0951] Step 5:

[0952] The server assigns a satisfaction score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. This score is stored in the database.

[0953] Step 6:

[0954] The server calculates a total score for each company based on the satisfaction scores. For example, if company A receives 15 positive comments and 5 negative comments, company A's satisfaction score will be +10. The compiled scores are stored in a database.

[0955] Step 7:

[0956] The server creates a ranking based on each company's satisfaction score. The scores for each company are sorted in descending order, and a ranking list is generated. This ranking list is stored in a database.

[0957] Step 8:

[0958] The server updates the UI to display the latest rankings on the web dashboard accessed by the user. HTML and JavaScript are used to generate a web page that visually displays the ranking list to the user.

[0959] Step 9:

[0960] Users access the web dashboard from their browser and view updated rankings in real time. By accessing the dashboard URL in their browser, users can quickly grasp comparisons and statuses of companies.

[0961] This series of steps results in an automated system that handles everything from data collection to analysis and display.

[0962] (Example 1)

[0963] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0964] There is a need for a system that can accurately evaluate a company's customer satisfaction and instantly compare it with competitors. In particular, there is a lack of efficient methods for collecting and analyzing review data from multiple websites. In addition, accuracy and efficiency are challenges in the process of analyzing the sentiment of this data using AI models and calculating satisfaction scores for each company.

[0965] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0966] In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing the data, means for performing sentiment analysis using a generative AI model, means for calculating satisfaction scores for each company, means for creating rankings with competitors, and means for displaying the rankings. This makes it possible to efficiently collect and pre-process word-of-mouth data from multiple websites and perform highly accurate sentiment analysis using a generative AI model. Furthermore, it is possible to quickly calculate satisfaction scores for each company and visually display real-time rankings with competitors.

[0967] "Word-of-mouth data" refers to text data of opinions and impressions about products and services posted by users on online platforms.

[0968] "Preprocessing" is the process of removing noise from collected data and preparing it in a format that is easy to analyze.

[0969] A "generative AI model" is an artificial intelligence model trained using machine learning, and is particularly used for sentiment analysis and text classification.

[0970] "Sentiment analysis" is the process of analyzing the text contained in word-of-mouth data and classifying its content as either positive, negative, or neutral.

[0971] A "satisfaction score" is an indicator of customer satisfaction for each company, calculated by assigning a score to each review data based on the results of sentiment analysis and then aggregating these scores.

[0972] "Ranking creation" is the process of arranging companies in descending order based on their customer satisfaction scores, thereby enabling visual comparisons with competitors.

[0973] "Ranking display" refers to displaying the created ranking results on the user interface so that users can view them.

[0974] This invention is a system that collects, preprocesses, and performs sentiment analysis on word-of-mouth data, calculates a satisfaction score for each company based on the results, and creates and displays a ranking with competitors. The following shows how this system is specifically implemented.

[0975] In implementing this system, the terminal first plays the role of collecting word-of-mouth data. The terminal uses a Python script to access the Twitter API. Specifically, the terminal collects tweets related to a specific hashtag via the Twitter API and saves them in JSON format. In this case, the terminal hardware is a client PC, and the software used is Python and the Twitter API.

[0976] The server receives JSON data sent from the terminal and stores it in a database. Since the stored data requires preprocessing, the server uses regular expressions to remove noise. By using regular expressions, spam tweets and advertisements are filtered out, generating clean data. The server hardware consists of a server machine, and the software used is Python and its regular expression package.

[0977] Next, the server performs sentiment analysis. For this purpose, it uses a generative AI model, specifically the BERT model. The server uses the Hugging Face Transformers library to classify tweets as positive, negative, or neutral. The server hardware continues to be a server machine.

[0978] Based on the sentiment analysis results, the server assigns a score to each review. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. These scores are then aggregated to calculate a satisfaction score for each company. Python and its aggregation library, NumPy, are used to aggregate the scores.

[0979] The server creates a ranking based on the aggregated scores. This ranking is sorted in descending order based on each company's satisfaction score. The server hardware and software will continue to use Python.

[0980] Finally, the user accesses a web dashboard to view the rankings. The server displays the aforementioned rankings in a user interface (UI). This UI is built using HTML, CSS, JavaScript, and a backend framework (e.g., Django or Flask).

[0981] As a concrete example, here is an example of a prompt message that uses the Twitter API to collect review data for the hashtag "good product":

[0982] Collect the latest tweets related to the hashtag "good product" and perform a sentiment analysis on them. Classify them as positive, negative, or neutral, and provide the number of each.

[0983] As described above, the present invention is a system that efficiently collects word-of-mouth data from many websites and performs sentiment analysis using a generative AI model, thereby enabling accurate evaluation of customer satisfaction for each company and real-time comparison with competitors.

[0984] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0985] Step 1: Data Collection

[0986] The device collects word-of-mouth data from various websites (e.g., social media, review sites). Specifically, the device uses a Python script to access the Twitter API. The device collects tweets related to a specific hashtag, "good product," and saves them in JSON format. The input is a specific hashtag, and the output is a JSON file of tweet data related to that hashtag. In terms of operation, it uses the Twitter API to collect tweets and saves them to local storage in JSON format.

[0987] Step 2: Data Preprocessing

[0988] The server receives JSON data sent from the terminal and stores it in a database. Next, the server uses regular expressions to remove noise from the data. For example, it filters out spam tweets and advertisements to obtain clean data. The input is collected tweet data in JSON format, and the output is clean tweet data with noise removed. Specifically, it uses a regular expression package to filter out advertising links and spam messages.

[0989] Step 3: Sentiment Analysis

[0990] The server generates clean, pre-processed data and performs sentiment analysis using an AI model. Specifically, it uses the BERT model with the Hugging Face Transformers library to classify each tweet as positive, negative, or neutral. The input is clean data, and the output is the result of the sentiment analysis for each tweet. In terms of operation, it calls the sentiment analysis model and classifies the sentiment of each tweet.

[0991] Step 4: Calculating the satisfaction score

[0992] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. Then, it compiles the total score for each company. The input is the result of the sentiment analysis, and the output is the satisfaction score for each company. Specifically, it compiles the sentiment score for each tweet and calculates the total score for each company.

[0993] Step 5: Competitive Analysis

[0994] The server creates a ranking of each company based on their satisfaction scores. The input is the satisfaction score for each company, and the output is the ranking. Specifically, it sorts the scores for each company in descending order and generates a ranking list.

[0995] Step 6: Display the rankings

[0996] Users can access a web dashboard to view their rankings. The server displays the generated rankings in the user interface (UI). The input is ranking data, and the output is a visual display of the rankings for the user. Specifically, HTML, CSS, and JavaScript are used to display the rankings on the web page, allowing users to visually confirm them.

[0997] (Application Example 1)

[0998] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0999] In recent years, consumer reviews have come to have a significant impact on a company's reputation. However, manually analyzing vast amounts of review data is difficult, making it challenging to grasp a company's relative ranking against its competitors in real time. To solve this problem, a system is needed that efficiently collects, preprocesses, and analyzes review data, calculates satisfaction scores for each company, and visualizes competitor rankings on smartphones.

[1000] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1001] In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing the data, and means for performing sentiment analysis. This makes it possible to efficiently collect consumer word-of-mouth data, process it into clean data, and perform sentiment classification using generative AI. Furthermore, by creating a ranking of competitors based on the calculated satisfaction score and displaying it on a smartphone, companies can grasp the competitive situation in real time.

[1002] "Word-of-mouth data" refers to reviews and comments posted by customers and users on online platforms regarding products and services.

[1003] "Preprocessing" refers to the process of removing noise from collected data and converting it into a format suitable for analysis.

[1004] "Sentiment analysis" is a method of analyzing text data and classifying its emotional tendencies (positive, negative, neutral).

[1005] The "satisfaction score" is a numerical value that represents an overall evaluation, calculated by assigning positive scores to positive reviews and negative scores to negative reviews based on the results of sentiment analysis.

[1006] A "ranking" refers to a list that ranks companies based on satisfaction scores and allows for comparison with competitors.

[1007] "A means of displaying rankings on a smartphone" refers to a function that visually displays customer satisfaction rankings for each company on the smartphone screen.

[1008] "Generative AI" is a general term for artificial intelligence that generates and analyzes text, images, and other data, and is a technology that enables highly accurate analysis, particularly in natural language processing.

[1009] This invention is a system that consistently performs the collection, preprocessing, sentiment analysis, satisfaction score calculation, competitor analysis, and ranking display of word-of-mouth data. It is implemented using a server, a smartphone, and a generative AI model.

[1010] The system program follows the following main processing steps:

[1011] 1. Data Collection

[1012] The server collects user review data from multiple websites, including social media and review sites. For example, it uses the Twitter API to retrieve tweets related to a specific hashtag. The collected data is stored in JSON format.

[1013] 2. Data preprocessing

[1014] The server receives the collected JSON data and stores it in a database. Next, it cleans up the data. For example, it filters out spam and advertisements using regular expressions and removes noise. This results in clean data suitable for analysis.

[1015] 3. Sentiment analysis

[1016] The server uses a generative AI model (e.g., BERT model) to perform sentiment analysis on clean data. It classifies text data as positive, negative, or neutral. This process is accelerated using CUDA-enabled GPUs.

[1017] 4. Calculation of satisfaction score

[1018] The server assigns a score to each review based on the sentiment analysis results. Positive reviews receive a score of +1, negative reviews receive -1, and neutral reviews receive 0. The total scores for each company are then tallied to calculate the satisfaction score.

[1019] 5. Competitive analysis

[1020] The server creates a ranking based on each company's satisfaction score. This ranking is ordered from highest to lowest score, visually representing the competitive landscape.

[1021] 6. Ranking Display

[1022] The smartphone displays ranking data sent from the server to the user. Users can check the real-time rankings through the smartphone application.

[1023] Specific examples and prompt statements

[1024] As a concrete example, the following shows the prompt text that a user would input to the generated AI model.

[1025] Example of a prompt:

[1026] "Use this app to perform sentiment analysis on tweets about Product A and display a competitor satisfaction ranking."

[1027] This prompt allows users to easily grasp real-time customer review results and competitive landscape for specific products or companies.

[1028] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1029] Step 1:

[1030] The device collects user review data from multiple websites, including social media and review sites. It uses the Twitter API to collect tweets related to specific hashtags. The input is a specific hashtag or keyword, and the output is the corresponding tweet data (in JSON format). This collected data is sent to a server and stored in a database.

[1031] Step 2:

[1032] The server receives JSON data sent from the terminal and stores it in the database. It then preprocesses the data. Specifically, it uses regular expressions to filter out spam tweets and advertisements, and removes noise. The input is raw JSON data, and the output is clean JSON data.

[1033] Step 3:

[1034] The server uses a generative AI model (e.g., the BERT model) to perform sentiment analysis on clean data. Specifically, it classifies text data into positive, negative, and neutral. The input is clean JSON data, and the output is data with the sentiment classification result for each tweet added. A CUDA-enabled GPU is used for the analysis.

[1035] Step 4:

[1036] The server assigns a score to each review based on the sentiment analysis results. Positive reviews receive a score of +1, negative reviews receive -1, and neutral reviews receive 0. The total scores for each company are aggregated to calculate the satisfaction score. The input is data including the sentiment classification results, and the output is the satisfaction score for each company.

[1037] Step 5:

[1038] The server creates a ranking of each company against its competitors based on their satisfaction scores. This ranking is ordered from highest score to lowest. The input is the satisfaction score for each company, and the output is a list of company rankings.

[1039] Step 6:

[1040] The smartphone displays ranking data transmitted from the server to the user. Users can check real-time rankings through the smartphone application. The input is company ranking data transmitted from the server, and the output is the ranking display in the user interface.

[1041] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1042] This invention combines a system that collects word-of-mouth data, performs preprocessing, sentiment analysis, calculates satisfaction scores, creates rankings, and displays rankings, with a sentiment engine that recognizes user emotions to perform even more advanced analysis and feedback.

[1043] overview

[1044] This system primarily consists of a server, terminals, and users. The emotion engine recognizes emotions from the user's voice or text input and sends this information to the server, which can then be reflected in the ranking display.

[1045] Data collection

[1046] The device collects customer review data from various websites (e.g., social media, review sites). The device retrieves tweets and reviews using specific keywords and hashtags and saves them in JSON format. For example, the device collects reviews from Twitter using the hashtag "product A review".

[1047] Data preprocessing

[1048] The server receives JSON data sent from the terminal and stores it in the database. The server uses regular expressions to remove noise and unnecessary information, creating a clean dataset. For example, it can filter out advertisements and spam messages.

[1049] sentiment analysis

[1050] The server uses the BERT model, a generated AI, to perform sentiment analysis on text data. The analysis results are classified as positive, negative, or neutral. For example, the text "This product is great!" is classified as positive.

[1051] Calculation of satisfaction score

[1052] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. The satisfaction scores for each company are compiled and stored in a database. For example, if company A has 15 positive reviews and 5 negative reviews, company A's satisfaction score will be +10.

[1053] Ranking creation

[1054] The server creates a ranking based on each company's satisfaction score. The total score for each company is sorted in descending order, and the ranking is generated.

[1055] Ranking display

[1056] The server displays the ranking results on a web dashboard that users access. The rankings are displayed visually using HTML and JavaScript. Users can access the dashboard in their browser and check the latest rankings. For example, it might show Company A in 1st place, Company B in 2nd place, and Company C in 3rd place.

[1057] Combination of emotional engines

[1058] Users access a web dashboard and enter feedback via text or voice. The emotion engine analyzes this feedback in real time and categorizes the user's emotions as positive, negative, or neutral. The results are sent to the server and reflected in the ranking display. For example, if a user enters feedback saying "This product is very good," the emotion engine will determine this to be positive and it will be reflected in the ranking display.

[1059] Specific example

[1060] As a concrete example, suppose a device collects 50 tweets from Twitter using the hashtag "product A review," of which 30 are positive, 15 are negative, and 5 are neutral. These tweets are pre-processed on a server, and sentiment analysis is performed using a generative AI. As a result, a satisfaction score of +15 is calculated for company A, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking.

[1061] In this way, a sophisticated system is realized that not only collects, analyzes, and displays user reviews, but also reflects user sentiment in real time.

[1062] The following describes the processing flow.

[1063] Step 1:

[1064] The device collects customer review data from various websites (e.g., social media, review sites). Specifically, the device uses specific keywords or hashtags (e.g., "product A review") to retrieve tweets from Twitter via GET requests. This tweet data is stored on the device in JSON format.

[1065] Step 2:

[1066] The device sends the collected review data to the server. The device uses a REST API to send the collected data to the server via a POST request. The sent data is temporarily stored in a database on the server.

[1067] Step 3:

[1068] The server preprocesses the received JSON data. The server uses regular expressions to remove noise (e.g., spam, advertisements) from the data, creating a clean dataset. For example, it removes advertising messages such as "Free sample giveaway!".

[1069] Step 4:

[1070] The server uses the BERT model, a generative AI, to perform sentiment analysis on pre-processed review data. It determines whether the text is positive, negative, or neutral and stores the results in a database. For example, a tweet like "This product is great" would be classified as positive.

[1071] Step 5:

[1072] The server assigns a satisfaction score to each review based on the sentiment analysis results. Positive opinions are given a score of +1, negative opinions a score of -1, and neutral opinions a score of 0, and these scores are stored in the database. For example, if there are 30 positive reviews and 15 negative reviews, Company A's score will be +15.

[1073] Step 6:

[1074] The server collects satisfaction scores from each company and creates a ranking. The scores for each company are sorted in descending order to generate a ranking list, which is then saved to the database. For example, Company A might be ranked 1st, Company B 2nd, and Company C 3rd.

[1075] Step 7:

[1076] The UI for displaying the server-generated rankings on the web dashboard is updated. The server uses HTML and JavaScript to generate a web page that visually displays the ranking list to the user. For example, company A is displayed as number 1 in the latest rankings.

[1077] Step 8:

[1078] Users access the web dashboard and then access the dashboard URL in their browser to view the rankings. Users can see the company satisfaction rankings in real time.

[1079] Step 9:

[1080] Users provide feedback. On the web dashboard, users offer their opinions on products and services through text or voice input. This input is sent to the sentiment engine.

[1081] Step 10:

[1082] The emotion engine analyzes user feedback in real time. It uses an emotion analysis algorithm to analyze voice and text input, classifying it as positive, negative, or neutral. The results are then sent to the server.

[1083] Step 11:

[1084] The server receives the results from the emotion engine and reflects them in the ranking display. For example, if new positive user feedback is reflected, Company A's satisfaction score will be increased by +1, and the ranking will be updated again.

[1085] This series of steps completes a system where data collection, analysis, display, and feedback integration are all automated.

[1086] (Example 2)

[1087] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1088] Traditional systems often require manual collection and analysis of customer review data, which is time-consuming and labor-intensive. Furthermore, it's difficult to reflect user feedback in real time, leading to issues with the accuracy of company satisfaction scores and rankings. Additionally, the inability to collect data from multiple sources simultaneously and perform comprehensive analysis makes accurate evaluation difficult.

[1089] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing, means for performing sentiment analysis using a generative AI model, means for calculating satisfaction scores for each company, means for creating rankings with competitors, means for displaying the rankings, means for analyzing user feedback and recognizing emotions, and means for reflecting the analysis results in the rankings. This automates all processes from word-of-mouth data collection and analysis to ranking creation and display, and further enables the reflection of real-time user feedback.

[1090] "Word-of-mouth data" refers to ratings and opinions posted by customers or users about products and services on online platforms (e.g., social media, review sites).

[1091] "Collection methods" refer to methods and devices for obtaining word-of-mouth data from various information sources.

[1092] "Preprocessing" refers to the process of removing noise and unnecessary information from collected data and converting it into a format suitable for analysis.

[1093] A "generative AI model" refers to a machine learning model used to automatically generate or analyze text and data.

[1094] "Sentiment analysis" refers to an analytical method for identifying emotions (positive, negative, neutral) from text data.

[1095] A "satisfaction score" refers to a score that quantifies the results of sentiment analysis and indicates an overall evaluation of a company or product.

[1096] A "ranking" refers to a list of companies or products based on customer satisfaction scores.

[1097] "Display means" refers to methods or devices that allow users to visually confirm results.

[1098] "Feedback" refers to the opinions and evaluations that users provide about a product or service.

[1099] An "emotion engine" refers to technology that identifies emotions in real time from user feedback.

[1100] This invention combines a system that collects, preprocesses, analyzes sentiment, calculates satisfaction scores, creates rankings, and displays rankings with a sentiment engine that recognizes user emotions, enabling more advanced analysis and feedback. This automates the entire process from collecting and analyzing review data to creating and displaying rankings, and also allows for the incorporation of real-time user feedback.

[1101] overview

[1102] This system primarily consists of a server, terminals, and users. The emotion engine recognizes emotions from the user's voice or text input and sends this information to the server, which can then be reflected in the ranking display.

[1103] Data collection

[1104] The device uses web scraping tools (e.g., Beautiful Soup, Scrapy) to collect user reviews from social media and review sites. For example, it might search for tweets using the hashtag "product A review" and save the data in JSON format. This allows for the unified collection of data from various sources.

[1105] Data preprocessing

[1106] The server receives JSON data sent from the terminal and stores it in a database (e.g., MySQL, PostgreSQL). Next, the server uses regular expressions and natural language processing (NLP) libraries (e.g., NLTK, SpaCy) to remove noise and unnecessary information, creating a clean dataset. This improves the quality of the data and allows for more accurate subsequent processing.

[1107] sentiment analysis

[1108] The server uses a generative AI model (e.g., BERT, GPT-3) to perform sentiment analysis on text data. Specifically, it classifies customer reviews into positive, negative, and neutral. For example, the text "This product is the best!" would be classified as positive. This allows for an understanding of general customer sentiment trends.

[1109] Calculation of satisfaction score

[1110] The server assigns a score to each review based on the sentiment analysis results. Positive opinions receive a score of +1, negative opinions receive -1, and neutral opinions receive 0. The satisfaction scores for each company are then compiled and stored in a database. For example, if company A receives 15 positive reviews and 5 negative reviews, company A's satisfaction score will be +10.

[1111] Ranking creation

[1112] The server creates a ranking based on customer satisfaction scores for each company. The total scores for each company are then sorted in descending order to generate the ranking. This makes it easy to compare companies with competitors and allows for quick information provision to users.

[1113] Ranking display

[1114] The server displays the ranking results on a web dashboard that users access. The rankings are displayed visually using HTML and JavaScript. Users can access the dashboard in their browser and check the latest rankings. For example, it might show Company A in 1st place, Company B in 2nd place, and Company C in 3rd place.

[1115] Combination of emotional engines

[1116] Users access a web dashboard and enter feedback via text or voice. The emotion engine analyzes this feedback in real time and categorizes the user's emotions as positive, negative, or neutral. The results are sent to the server and reflected in the ranking display. For example, if a user enters feedback saying "This product is very good," the emotion engine will determine this to be positive and it will be reflected in the ranking display.

[1117] Specific example

[1118] As a concrete example, suppose a device collects 50 tweets from Twitter using the hashtag "product A review," with 30 being positive, 15 negative, and 5 neutral. These tweets are preprocessed on a server, and sentiment analysis is performed using a generative AI model. As a result, a satisfaction score of +15 is calculated for company A, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking.

[1119] In this way, a sophisticated system is realized that not only collects, analyzes, and displays user reviews, but also reflects user sentiment in real time.

[1120] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1121] Step 1: Data Collection

[1122] The device uses a web scraping tool (e.g., Beautiful Soup, Scrapy) to collect user review data from social media and review sites. Input requires the URL of the target site and search keywords (e.g., "product A review"). Based on this input, the device extracts the necessary user review data from the website and saves it in JSON format. The output is the collected user review data (in JSON format).

[1123] Step 2: Data transmission

[1124] The device collects review data (in JSON format) and sends it to the server. The input requires the collected review data and the destination server address. The device then forwards the data to the server based on this input. The output is the review data sent to the server.

[1125] Step 3: Data Preprocessing

[1126] The server receives the JSON data and stores it in a database (e.g., MySQL, PostgreSQL). Next, the server uses regular expressions and natural language processing (NLP) libraries (e.g., NLTK, SpaCy) to remove noise and unwanted information, creating a clean dataset. The input is the received raw data (in JSON format). The output is a pre-processed, clean dataset.

[1127] Step 4: Sentiment Analysis

[1128] The server uses a generative AI model (e.g., BERT, GPT-3) to perform sentiment analysis on text data. Specifically, it classifies word-of-mouth text into positive, negative, and neutral. The input requires a pre-processed, clean dataset. The server uses this to perform sentiment analysis and assigns sentiment labels to each text. The output is the data with sentiment labels assigned.

[1129] Step 5: Calculating the Satisfaction Score

[1130] The server assigns a score to each review based on the sentiment analysis results. The input requires data with sentiment labels. Positive opinions are assigned a score of +1, negative opinions a score of -1, and neutral opinions a score of 0. The satisfaction scores for each company are then compiled and stored in a database. The output is the satisfaction score for each company.

[1131] Step 6: Create Rankings

[1132] The server creates a ranking based on customer satisfaction scores for each company. The input requires customer satisfaction scores for each company. The server sorts these scores in descending order to generate the ranking and saves this data to a database. The output is the ranking for each company.

[1133] Step 7: Display the rankings

[1134] The server displays the ranking results on a web dashboard accessed by the user. The input requires generated ranking data. The server uses HTML and JavaScript to visually display the rankings. The output is the latest ranking displayed on the dashboard.

[1135] Step 8: Gathering Feedback

[1136] The user accesses the web dashboard and enters feedback via text or voice. The input requires the user's feedback as text or voice data. The feedback is sent to the server. The output is the feedback data sent to the server.

[1137] Step 9: Analysis using the Emotion Engine

[1138] The emotion engine analyzes user feedback in real time and classifies the user's emotions as positive, negative, or neutral. The input is user feedback data. The output is feedback data with emotion labels attached.

[1139] Step 10: Reflection in rankings

[1140] The server receives the analysis results from the emotion engine and reflects them in the ranking display. The input requires feedback data with emotion labels. The server uses this data to update the rankings and reflects them in the dashboard in real time. The output is the updated rankings.

[1141] (Application Example 2)

[1142] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1143] Traditional word-of-mouth data analysis systems have struggled to reflect customers' real-time emotions, making immediate improvements in customer satisfaction impossible. Furthermore, systems that generate rankings based on a comprehensive evaluation of multiple ratings suffer from low variability due to their reliance on static data. As a result, the latest user emotions and feedback are not reflected in business decisions or marketing strategies, preventing a more accurate response to customer needs.

[1144] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for collecting word-of-mouth data, means for pre-processing, means for performing sentiment analysis, means for calculating satisfaction scores for each company, means for creating rankings with competitors, means for displaying the rankings, means for receiving voice or text feedback from users and analyzing it using a sentiment engine, and means for reflecting the analysis results in the rankings in real time. This makes it possible to immediately analyze real-time user feedback and reflect it in the rankings.

[1145] "Word-of-mouth data" refers to a collection of digital opinions such as reviews, comments, and ratings posted by customers on the internet about products and services.

[1146] "Preprocessing" is the process of removing unnecessary information and noise from raw data and organizing it so that it can be analyzed accurately.

[1147] "Sentiment analysis" is a technique that analyzes the emotions contained in opinions and feedback within text data and classifies them into categories such as positive, negative, and neutral.

[1148] A "satisfaction score" is an index that quantifies customer satisfaction with a product or service based on customer feedback.

[1149] A "ranking" is a list that evaluates and ranks multiple companies or products based on criteria such as satisfaction scores.

[1150] An "emotion engine" is software or hardware that analyzes voice or text feedback from users in real time and classifies their emotions.

[1151] A "user" is someone who accesses the system, views user reviews, and provides feedback.

[1152] "Real-time" refers to a situation where data or events are processed and reflected immediately the moment they occur.

[1153] The system implementing this invention mainly consists of a server, a terminal, and a user. The specific roles of each component are described below.

[1154] server

[1155] The server plays a central role in collecting, pre-processing, and performing sentiment analysis on review data. Specifically, this includes the following methods:

[1156] 1. Data Collection Method: Receive word-of-mouth data sent from the device and store it in a database. The data is stored in JSON format.

[1157] 2. Preprocessing method: Regular expressions are used to remove noise and unwanted information, creating a clean dataset. For example, advertisements and spam messages are filtered out.

[1158] 3. Sentiment Analysis Method: The BERT model, a generative AI model, is used to analyze the sentiment of the collected text data. The analysis results are classified into positive, negative, and neutral.

[1159] 4. Satisfaction Score Calculation Method: Based on the results of sentiment analysis, a score is assigned to each review, and a total score is calculated for each company. For example, positive opinions are given a score of +1, negative opinions are given a score of -1, and neutral opinions are given a score of 0.

[1160] 5. Ranking creation method: The total score for each company is sorted in descending order, and a ranking is generated.

[1161] 6. Ranking Display Method: Ranking results will be displayed on a web dashboard. The display will be visually generated using HTML and JavaScript.

[1162] terminal

[1163] The device is responsible for collecting user review data and sending it to the server. Specifically, this includes the following methods:

[1164] 1. Data Collection Methods: Collect word-of-mouth data from multiple websites (e.g., social media, review sites). For example, obtain word-of-mouth data from Twitter using specific keywords or hashtags.

[1165] User

[1166] Users are responsible for accessing the system, providing feedback, and checking rankings. Specifically, this includes the following:

[1167] 1. Feedback input method: Access the web dashboard and enter feedback in text or voice.

[1168] 2. Emotion Analysis Method: The emotion engine analyzes user feedback in real time, classifies emotions as positive, negative, or neutral, and sends the results to the server.

[1169] 3. How to check rankings: Check the latest rankings on the web dashboard. Rankings are updated in real time.

[1170] Hardware and software to be used

[1171] Hardware:

[1172] Server (for data storage and processing)

[1173] Smartphone (for user feedback input)

[1174] software:

[1175] Transformers library (using the BERT model)

[1176] Requests Library (for data collection)

[1177] Regular expression library (for data preprocessing)

[1178] softmax function (Scipy library; used for calculating sentiment scores)

[1179] Specific example

[1180] Let's look at a concrete example. Suppose a device collects 50 tweets from Twitter using the hashtag "product A review," with 30 positive, 15 negative, and 5 neutral. These tweets are pre-processed on a server, and sentiment analysis is performed using a generative AI model. As a result, company A's satisfaction score is calculated to be +15, and a ranking is generated. When a user accesses the dashboard and enters feedback such as "This product is very good," the sentiment engine determines this to be positive, and it is reflected in the ranking in real time.

[1181] Example of a prompt

[1182] "Analyze the sentiment of the following text: 'This product is great.' Classify the sentiment score as positive, negative, or neutral."

[1183] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1184] Step 1:

[1185] The device collects user review data from multiple websites (e.g., social media and review sites). For example, it retrieves reviews from Twitter using specific keywords or hashtags. The input is a specific keyword such as "product A review," and the output is user review data in JSON format. This data is initially saved for later analysis.

[1186] Step 2:

[1187] The server receives review data sent from the terminal and stores it in a database. The input is JSON data sent from the terminal, and the output is storage data stored in the database. A regular expression library is used to remove noise and unwanted information, such as advertisements and spam messages.

[1188] Step 3:

[1189] The server generates a clean dataset. The input is pre-processed review data using regular expressions, and the output is clean JSON formatted data. At this stage, the data is ready for analysis, so we move on to the next step. Specifically, we remove unnecessary strings and HTML tags using regular expressions.

[1190] Step 4:

[1191] The server performs sentiment analysis using the BERT model, a generative AI model. The input is each review text from the clean dataset, and the output is the sentiment score (positive, negative, neutral) for each text. Specifically, each text is input into the BERT model to obtain the sentiment score.

[1192] Step 5:

[1193] The server assigns a score to each review based on the sentiment analysis results. The input is the sentiment analysis results, and the output is the satisfaction score for each company. For example, positive opinions are given a score of +1, negative opinions a score of -1, and neutral opinions a score of 0. This allows the total score for each company to be calculated.

[1194] Step 6:

[1195] The server sorts the total scores for each company in descending order and generates a ranking. The input is the satisfaction score, and the output is the ranking for each company. For example, the company with the highest total score will be ranked first.

[1196] Step 7:

[1197] The server displays the ranking results on a web dashboard. The input is the generated ranking, and the output is the visually displayed ranking information. Using HTML and JavaScript, users can view the latest rankings in their browser.

[1198] Step 8:

[1199] Users access a web dashboard and enter feedback in text or voice. The input is the user's feedback in text or voice, and the output is the feedback information sent to the sentiment engine.

[1200] Step 9:

[1201] The emotion engine analyzes user feedback in real time. The input is user feedback, and the output is a real-time emotion score. Specifically, user feedback is input into a BERT model to obtain an emotion score.

[1202] Step 10:

[1203] The server reflects the analysis results in the rankings in real time. The input is the real-time sentiment score, and the output is the updated ranking. This ensures that the rankings display the latest user feedback immediately.

[1204] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1205] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1206] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1207] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1208] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1209] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1210] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1211] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1212] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1213] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1214] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1215] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1216] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1217] 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.

[1218] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1219] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1220] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1221] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1222] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1223] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1224] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[1225] The following is further disclosed regarding the embodiments described above.

[1226] (Claim 1)

[1227] Methods for collecting word-of-mouth data,

[1228] Means for performing pre-processing,

[1229] Methods for conducting sentiment analysis,

[1230] A method for calculating satisfaction scores for each company,

[1231] Methods for creating rankings against competitors,

[1232] The means for displaying the aforementioned ranking,

[1233] A system that includes this.

[1234] (Claim 2)

[1235] The system according to claim 1, which performs sentiment analysis using generative AI.

[1236] (Claim 3)

[1237] The system according to claim 1, wherein the means for collecting word-of-mouth data includes means for obtaining data from multiple websites.

[1238] "Example 1"

[1239] (Claim 1)

[1240] Methods for collecting word-of-mouth data,

[1241] Means for performing pre-processing,

[1242] A method for performing sentiment analysis using a generative AI model,

[1243] A method for calculating satisfaction scores for each company,

[1244] Methods for creating rankings against competitors,

[1245] The means for displaying the aforementioned ranking,

[1246] A system that includes this.

[1247] (Claim 2)

[1248] The system according to claim 1, comprising means for obtaining data from multiple websites.

[1249] (Claim 3)

[1250] The system according to claim 1, comprising means for assigning scores to word-of-mouth data based on the results of sentiment analysis, and for aggregating those scores to calculate a satisfaction score.

[1251] "Application Example 1"

[1252] (Claim 1)

[1253] Methods for collecting word-of-mouth data,

[1254] Means for performing pre-processing,

[1255] Methods for conducting sentiment analysis,

[1256] A method for calculating satisfaction scores for each company,

[1257] Methods for creating rankings against competitors,

[1258] A method for displaying rankings on a smartphone,

[1259] A system that includes this.

[1260] (Claim 2)

[1261] The system according to claim 1, which performs sentiment analysis using generative AI.

[1262] (Claim 3)

[1263] The system according to claim 1, wherein the means for collecting word-of-mouth data includes means for obtaining data from multiple websites.

[1264] "Example 2 of combining an emotion engine"

[1265] (Claim 1)

[1266] Methods for collecting word-of-mouth data,

[1267] Means for performing pre-processing,

[1268] A method for performing sentiment analysis using a generative AI model,

[1269] A method for calculating satisfaction scores for each company,

[1270] Methods for creating rankings against competitors,

[1271] The means for displaying the aforementioned ranking,

[1272] A means of analyzing user feedback and recognizing emotions,

[1273] A means of reflecting the aforementioned analysis results in the ranking,

[1274] A system that includes this.

[1275] (Claim 2)

[1276] The system according to claim 1, which performs sentiment analysis using a generative AI model.

[1277] (Claim 3)

[1278] The system according to claim 1, wherein the means for collecting word-of-mouth data includes means for obtaining data from multiple sources.

[1279] "Application example 2 when combining with an emotional engine"

[1280] (Claim 1)

[1281] Methods for collecting word-of-mouth data,

[1282] Means for performing pre-processing,

[1283] Methods for conducting sentiment analysis,

[1284] A method for calculating satisfaction scores for each company,

[1285] Methods for creating rankings against competitors,

[1286] The means for displaying the aforementioned ranking,

[1287] A means of receiving voice or text feedback from users and analyzing it using an emotion engine,

[1288] A means of reflecting the analysis results in the ranking in real time,

[1289] A system that includes this.

[1290] (Claim 2)

[1291] The system according to claim 1, which performs sentiment analysis using generative AI.

[1292] (Claim 3)

[1293] The system according to claim 1, wherein the means for collecting word-of-mouth data includes means for obtaining data from multiple websites. [Explanation of Symbols]

[1294] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. Methods for collecting word-of-mouth data, Means for performing pre-processing, Methods for conducting sentiment analysis, A method for calculating satisfaction scores for each company, Methods for creating rankings against competitors, The means for displaying the aforementioned ranking, A system that includes this.

2. The system according to claim 1, which performs sentiment analysis using generative AI.

3. The system according to claim 1, wherein the means for collecting word-of-mouth data includes means for obtaining data from multiple websites.

Images