System

Abstract

An object of a system according to an embodiment is to provide highly reliable word-of-mouth information.SOLUTION: A system includes a collection unit, a learning unit, a filtering unit, and a presentation unit. The collection unit collects word-of-mouth information. The learning unit analyzes the word-of-mouth information collected by the collection unit and generates a model for identifying malicious word-of-mouth information and posted photos. The filtering unit filters malicious word-of-mouth information and posted photos using the model generated by the learning unit. The presentation unit presents the reliable word-of-mouth information filtered by the filtering unit.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] With conventional technology, there was a problem in that it was difficult to obtain reliable information because word-of-mouth information on the web contained slanderous comments and information that was different from the actual situation.

[0005] The system according to the embodiment aims to provide highly reliable word-of-mouth information. [Means for solving the problem]

[0006] A system according to an embodiment includes a collection unit, a learning unit, a filtering unit, and a presentation unit. The collection unit collects word-of-mouth information. The learning unit analyzes the word-of-mouth information collected by the collection unit and generates a model for identifying malicious word-of-mouth reviews and posted photos. The filtering unit filters out malicious word-of-mouth reviews and posted photos using the model generated by the learning unit. The presentation unit presents highly reliable word-of-mouth reviews filtered by the filtering unit. [Effects of the Invention]

[0007] The system according to the embodiment can provide highly reliable word-of-mouth information. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. DETAILED DESCRIPTION OF THE INVENTION

[0009] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0010] First, the terms used in the following description will be explained.

[0011] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, the processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), or a TPU (Tensor Processing Unit).

[0012] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0013] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0014] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), and Bluetooth (registered trademark).

[0015] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0016] [First embodiment] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0017] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0018] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0019] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0020] The reception device 38 includes a touch panel 38A and a microphone 38B, and receives user input. The touch panel 38A detects contact with a pointer (for example, a pen or a finger) to receive user input by the touch of the pointer. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 (see FIG. 2) acquires the data indicating the user input.

[0021] Output device 40 includes a display 40A and a speaker 40B, and presents data to a user by outputting the data in a form of expression that the user can perceive (e.g., audio and / or text). Display 40A displays visible information such as text and images in accordance with instructions from processor 46. Speaker 40B outputs audio in accordance with instructions from processor 46. Camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0022] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0023] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0024] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0025] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0026] In the smart device 14, the specific processing is performed by the processor 46. The storage 50 stores a specific processing program 60. The specific processing program 60 is used together with the specific processing program 56 by the data processing system 10. The processor 46 reads the specific processing program 60 from the storage 50 and executes the read specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as the control unit 46A in accordance with the specific processing program 60 executed on the RAM 48. Note that the smart device 14 has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can also perform processing similar to that of the specific processing unit 290 using these models.

[0027] Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device (e.g., a generation server) may have the data generation model 58. In this case, the data processing device 12 obtains a processing result (prediction result, etc.) using the data generation model 58 by communicating with the server device having the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device owned by a user (e.g., a mobile phone, a robot, a home appliance, etc.). Next, an example of processing by the data processing system 10 according to the first embodiment will be described.

[0028] (Example 1) A review filtering system according to an embodiment of the present invention analyzes vast amounts of review information and filters out malicious reviews and posted photos. The review filtering system collects review information, uses AI to train the system, filters out malicious reviews and posted photos, and presents only reliable reviews on each review site. For example, the review filtering system collects review information from each review site and stores it in a database. Next, the collected review information is trained using AI. The AI ​​analyzes the review information and generates a model for identifying malicious reviews and posted photos. For example, it can identify slanderous, false reviews, stealth marketing, and the like. The trained model is used to filter malicious reviews and posted photos. For example, it can automatically detect and filter slanderous, false reviews, stealth marketing, and the like. Finally, it presents only reliable reviews on each review site. This allows users to obtain reliable review information and avoid making decisions based on erroneous information. The review filtering system thus prevents the harm caused by malicious reviews and posted photos and provides reliable review information. For example, restaurant review sites display only reliable reviews, allowing users to choose restaurants with confidence. Similarly, product review sites display only reliable reviews from users who have actually used the products, allowing users to choose products based on accurate information.

[0029] A review filtering system according to an embodiment includes a collection unit, a learning unit, a filtering unit, and a presentation unit. The collection unit collects review information. For example, the collection unit collects review information from various review sites and stores the collected information in a database. The collection unit can collect review information from various genres, such as restaurant reviews and product reviews. The learning unit analyzes the review information collected by the collection unit and generates a model for identifying malicious reviews and posted photos. The learning unit, for example, uses AI to generate a model for identifying slanderous comments, reviews that are inaccurate, stealth marketing, and the like. For example, the AI ​​uses a text generation AI (e.g., LLM) to analyze review information and identify malicious reviews. The learning unit can also analyze posted photos and identify malicious photos using a multimodal generation AI. The filtering unit uses the model generated by the learning unit to filter malicious reviews and posted photos. For example, the filtering unit uses the trained model to automatically detect and filter slanderous comments, reviews that are inaccurate, stealth marketing, and the like. For example, the filtering unit can use AI to detect and filter malicious reviews. The presentation unit presents reliable reviews filtered by the filtering unit. The presentation unit presents the reliable reviews on each review site, for example. For example, the presentation unit displays the filtered reliable reviews to users via a website or a mobile application. This allows the review filtering system according to the embodiment to efficiently collect, analyze, filter, and present review information.

[0030] The collection unit can collect review information from each review site and store it in a database. For example, the collection unit collects review information from each review site and stores it in a database. For example, the collection unit can collect information from review sites such as Amazon (registered trademark) and TripAdvisor (registered trademark). The collection unit automatically collects review information from each review site using, for example, web scraping technology. The collection unit can also acquire review information from each review site using an API. For example, the collection unit periodically acquires review information using the API of each review site and stores it in a database. This makes it possible to efficiently collect review information from each review site and store it in a database.

[0031] The learning unit can analyze the collected word-of-mouth information and generate a model for identifying slander, false word-of-mouth, and stealth marketing. The learning unit, for example, analyzes the collected word-of-mouth information and generates a model for identifying slander, false word-of-mouth, and stealth marketing. The learning unit can, for example, use AI to learn and identify specific patterns of slander. The learning unit can also learn and identify characteristics of false word-of-mouth. Furthermore, the learning unit can learn and identify characteristics of stealth marketing. For example, the learning unit uses text generation AI (e.g., LLM) to analyze word-of-mouth information and generate a model for identifying slander, false word-of-mouth, and stealth marketing. This makes it possible to generate a model for identifying slander, false word-of-mouth, stealth marketing, and the like.

[0032] The filtering unit can use the trained model to automatically detect and filter defamatory comments, false reviews, and stealth marketing. The filtering unit can, for example, use the trained model to automatically detect and filter defamatory comments, false reviews, and stealth marketing. The filtering unit can, for example, use AI to automatically detect and filter defamatory comments. The filtering unit can also automatically detect and filter false reviews. The filtering unit can also automatically detect and filter stealth marketing reviews. For example, the filtering unit can use text generation AI (e.g., LLM) to analyze review information and automatically detect and filter defamatory comments, false reviews, and stealth marketing. This makes it possible to automatically detect and filter defamatory comments, false reviews, stealth marketing, etc.

[0033] The presentation unit can present the reliable reviews filtered by the filtering unit on each review site. For example, the presentation unit presents the reliable reviews filtered by the filtering unit on each review site. For example, the presentation unit can display the reliable reviews to users through a website or a mobile application. The presentation unit can also send the reliable reviews to users by email. Furthermore, the presentation unit can share the reliable reviews on social media. For example, the presentation unit can share the filtered reliable reviews on social media such as Facebook (registered trademark) or Twitter (registered trademark). This allows the reliable reviews to be presented on each review site.

[0034] The collection unit can analyze the update frequency of each review site and set an optimal collection schedule. The collection unit, for example, analyzes the update frequency of each review site and sets an optimal collection schedule. The collection unit, for example, can analyze the update frequency of each review site and prioritize information collection from frequently updated sites. The collection unit can also set a regular collection schedule for sites that are updated infrequently and collect information efficiently. Furthermore, for sites that are updated irregularly, the collection unit can dynamically adjust the schedule using AI to collect information at the optimal timing. For example, the collection unit sets an optimal collection schedule based on criteria such as daily updates or weekly updates. As a result, review information can be collected efficiently by setting an optimal collection schedule based on the update frequency of each review site.

[0035] The collection unit can evaluate the reliability of reviews at the time of collection and apply a filter that excludes unreliable information. For example, the collection unit can analyze the past posting history of review posters and exclude reviews from unreliable posters. The collection unit can also analyze the content of reviews and exclude reviews that include slander or overly positive content. Furthermore, the collection unit can filter out unreliable information by taking into account the posting date and time of the review and the number of ratings. For example, the collection unit applies a filter that excludes unreliable information based on criteria such as spam, false information, and exaggerated content. In this way, reliable review information can be collected by evaluating the reliability of reviews and excluding unreliable information.

[0036] The collection unit can apply different collection algorithms depending on the genre and category of the review during collection. For example, the collection unit applies different collection algorithms depending on the genre and category of the review during collection. For example, when collecting restaurant reviews, the collection unit can adjust the collection algorithm based on the type of food and the service rating. Furthermore, when collecting product reviews, the collection unit can change the collection algorithm depending on the product category and purpose of use. Furthermore, when collecting travel reviews, the collection unit can optimize the collection algorithm based on the ratings of tourist destinations and accommodations. For example, the collection unit classifies genres and categories based on criteria such as product category and service category, and applies the corresponding collection algorithm. In this way, more appropriate review information can be collected by adjusting the collection algorithm depending on the genre and category of the review.

[0037] The collection unit can prioritize collecting highly relevant word-of-mouth information by taking into account the user's geographical location information during collection. For example, the collection unit prioritizes collecting highly relevant word-of-mouth information by taking into account the user's geographical location information during collection. For example, the collection unit can identify the user's current location using GPS data and prioritize collecting word-of-mouth information in the vicinity. The collection unit can also estimate the user's geographical location using an IP address and prioritize collecting word-of-mouth information related to that area. For example, the collection unit prioritizes collecting word-of-mouth information about restaurants close to the user's current location. The collection unit can also prioritize collecting word-of-mouth information about tourist destinations where the user is traveling. Furthermore, the collection unit can prioritize collecting word-of-mouth information about stores and services in the area where the user lives. In this way, highly relevant word-of-mouth information can be prioritized by taking into account the user's geographical location information.

[0038] The collection unit can analyze the user's social media activities at the time of collection and collect related word-of-mouth information. For example, the collection unit can analyze the user's social media activities at the time of collection and collect related word-of-mouth information. For example, the collection unit can collect word-of-mouth information about places where the user has checked in on social media. The collection unit can also analyze the content of the user's social media posts and collect related word-of-mouth information. Furthermore, the collection unit can collect related word-of-mouth information by referring to the activities of the user's friends on social media. For example, the collection unit preferentially collects word-of-mouth information about restaurants where the user has checked in on social media. The collection unit can also analyze the content of the user's social media posts and collect reviews of related products. Furthermore, the collection unit can analyze the activities of the user's friends on social media and collect word-of-mouth information about related tourist destinations. In this way, related word-of-mouth information can be efficiently collected by analyzing the user's social media activities.

[0039] The collection unit can customize the collection method by reflecting the user's past feedback when collecting data. For example, the collection unit customizes the collection method by reflecting the user's past feedback when collecting data. For example, the collection unit can prioritize collecting information from review sites that the user has previously given high ratings to. The collection unit can also adjust the collection algorithm based on feedback provided by the user in the past. Furthermore, the collection unit can analyze the user's past feedback and suggest an optimal collection method. For example, the collection unit prioritizes collecting information from review sites for restaurants that the user has previously given high ratings to. The collection unit can also adjust the product review collection algorithm based on feedback provided by the user in the past. Furthermore, the collection unit can analyze the user's past feedback and suggest an optimal collection method for review information on tourist destinations. In this way, the collection method can be customized by reflecting the user's past feedback to collect more appropriate information.

[0040] The learning unit can optimize the learning algorithm by referring to past learning data during learning. For example, the learning unit can optimize the learning algorithm by referring to past learning data during learning. For example, the learning unit can identify slanderous patterns based on past learning data and optimize the learning algorithm. The learning unit can also reflect stealth marketing characteristics in the learning algorithm by referring to past learning data. Furthermore, the learning unit can analyze past learning data and optimize an algorithm for identifying word-of-mouth that differs from the actual situation. For example, the learning unit can optimize the learning algorithm by referring to previous training data or past models. In this way, by referring to past learning data, the learning algorithm can be optimized and accuracy can be improved.

[0041] The learning unit can generate different learning models during learning depending on the genre and category of the review. For example, the learning unit generates different learning models during learning depending on the genre and category of the review. For example, the learning unit can generate a learning model specialized for restaurant reviews to analyze food ratings and service quality. The learning unit can also generate a learning model specialized for product reviews to analyze product performance and usability. Furthermore, the learning unit can generate a learning model specialized for travel reviews to analyze ratings of tourist spots and accommodations. For example, the learning unit classifies genres and categories based on criteria such as product category and service category, and generates a learning model accordingly. In this way, more appropriate analysis is possible by generating a learning model according to the genre and category of the review.

[0042] The learning unit can update the learning data by reflecting user feedback during learning. For example, the learning unit updates the learning data by reflecting user feedback during learning. For example, the learning unit can update the learning data based on feedback provided by the user to improve the accuracy of the algorithm. The learning unit can also reflect user feedback to improve the accuracy of detecting slander and stealth marketing. Furthermore, the learning unit can analyze user feedback and review the criteria for selecting learning data. For example, the learning unit updates the learning data based on survey results and review content. In this way, the learning data can be updated by reflecting user feedback, thereby improving the accuracy of the algorithm.

[0043] The learning unit can weight the learning data based on the time of submission of the review during learning. For example, the learning unit weights the learning data based on the time of submission of the review during learning. For example, the learning unit can weight the most recent review information highly and reflect this in the learning algorithm. The learning unit can also weight older review information less highly and reflect this in the learning algorithm. Furthermore, the learning unit can assign appropriate weights based on the time of submission of the review to optimize the learning data. For example, the learning unit evaluates and weights the time of submission based on criteria such as the date and time of posting or whether the review is before or after an event. In this way, more appropriate learning is possible by weighting the learning data based on the time of submission of the review.

[0044] The learning unit can integrate information from different data sources to enrich the learning data during learning. For example, the learning unit can integrate information from different data sources to enrich the learning data during learning. For example, the learning unit can integrate information collected from various review sites to enrich the learning data. The learning unit can also integrate information from social media and blogs to enrich the learning data. Furthermore, the learning unit can integrate information from public databases and news sites to enrich the learning data. For example, the learning unit collects information from different data sources such as social media and news sites, and integrates it to enrich the learning data. In this way, by integrating information from different data sources, the learning data can be enriched and its accuracy can be improved.

[0045] The learning unit can evaluate the reliability of reviews and exclude low-reliability data during learning. For example, the learning unit evaluates the reliability of reviews and excludes low-reliability data during learning. For example, the learning unit can analyze the past posting history of review posters and exclude low-reliability data. The learning unit can also analyze the content of reviews and exclude data that contains slander or overly positive content. Furthermore, the learning unit can filter low-reliability data by taking into account the posting date and time of the review and the number of ratings. For example, the learning unit excludes low-reliability data based on criteria such as spam, false information, and exaggerated content. In this way, by evaluating the reliability of reviews and excluding low-reliability data, more reliable training data can be used.

[0046] The filtering unit can improve the accuracy of filtering by taking into account the interrelationships between reviews when filtering. The filtering unit can improve the accuracy of filtering by taking into account the interrelationships between reviews when filtering, for example. The filtering unit can, for example, analyze the content of the reviews and group and filter reviews that are related to each other. The filtering unit can also take into account the interrelationships between the posters of the reviews and preferentially filter highly reliable information. Furthermore, the filtering unit can filter information that is related to each other by taking into account the number of ratings and comments on the reviews. For example, the filtering unit evaluates the interrelationships between reviews based on criteria such as co-occurrence networks and highly related keywords, and improves the accuracy of filtering. In this way, the accuracy of filtering can be improved by taking into account the interrelationships between reviews.

[0047] The filtering unit can perform filtering while taking into consideration the attribute information of the person who submitted the review. For example, the filtering unit performs filtering while taking into consideration the attribute information of the person who submitted the review. The filtering unit can filter highly reliable information while taking into consideration, for example, the age and gender of the person who submitted the review. The filtering unit can also filter highly reliable information by analyzing the past posting history of the person who submitted the review. Furthermore, the filtering unit can filter highly reliable information while taking into consideration the region and occupation of the person who submitted the review. For example, the filtering unit evaluates the attribute information of the person who submitted the review based on criteria such as age, gender, and region, and performs filtering. In this way, more reliable information can be filtered by taking into consideration the attribute information of the person who submitted the review.

[0048] The filtering unit can weight the filtering based on the frequency of reviews submitted during filtering. For example, the filtering unit weights the filtering based on the frequency of reviews submitted during filtering. For example, the filtering unit can weight reviews that are posted frequently, thereby filtering out highly reliable information. The filtering unit can also weight reviews that are submitted infrequently, thereby filtering out less reliable information. Furthermore, the filtering unit can weight reviews appropriately based on the frequency of reviews submitted, thereby improving the accuracy of filtering. For example, the filtering unit evaluates and weights the submission frequency based on criteria such as the number of posts and the number of posts within a period. As a result, more appropriate filtering is possible by weighting the filtering based on the frequency of reviews submitted.

[0049] The filtering unit can perform filtering taking into account the geographical distribution of reviews. For example, the filtering unit can perform filtering taking into account the geographical distribution of reviews. For example, the filtering unit can prioritize filtering reviews close to the user's current location. The filtering unit can also analyze the geographical distribution of reviews and filter reviews related to a specific region. Furthermore, the filtering unit can filter highly reliable information taking into account the geographical distribution of reviews. For example, the filtering unit evaluates the geographical distribution and performs filtering based on criteria such as GPS data or the number of posts by region. This enables more appropriate filtering by taking into account the geographical distribution of reviews.

[0050] The filtering unit can improve the accuracy of filtering by referring to literature related to the word-of-mouth reviews during filtering. For example, the filtering unit can improve the accuracy of filtering by referring to literature related to the word-of-mouth reviews during filtering. For example, the filtering unit can analyze the content of the word-of-mouth reviews and filter by referring to related academic papers and articles. The filtering unit can also analyze the content of the word-of-mouth reviews and filter by referring to related news articles. Furthermore, the filtering unit can analyze the content of the word-of-mouth reviews and filter by referring to related blogs and reviews. For example, the filtering unit evaluates related literature based on criteria such as academic papers and industry reports, and improves the accuracy of filtering. In this way, the accuracy of filtering can be improved by referring to literature related to the word-of-mouth reviews.

[0051] The filtering unit can perform filtering taking into account the market value of the word-of-mouth when filtering. For example, the filtering unit performs filtering taking into account the market value of the word-of-mouth when filtering. For example, the filtering unit can analyze the content of the word-of-mouth and preferentially filter information with high market value. The filtering unit can also analyze the content of the word-of-mouth and exclude information with low market value. Furthermore, the filtering unit can evaluate the market value of the word-of-mouth and apply appropriate filtering criteria. For example, the filtering unit evaluates the market value based on criteria such as sales data and customer ratings and performs filtering. In this way, more appropriate filtering is possible by taking into account the market value of the word-of-mouth.

[0052] The presentation unit can select the optimal display method by referring to the user's past operation history when presenting the information. For example, the presentation unit can select the optimal display method by referring to the user's past operation history when presenting the information. For example, the presentation unit can preferentially apply the display method of a review site that the user has previously given a high rating to. The presentation unit can also suggest the optimal display method based on the user's past operation history. Furthermore, the presentation unit can analyze the user's past operation history and select the optimal display method. For example, the presentation unit evaluates the past operation history based on criteria such as click history and browsing history, and selects the optimal display method. In this way, the optimal display method can be selected by referring to the user's past operation history, and more appropriate information can be provided.

[0053] The presentation unit can customize the display content according to the user's current areas of interest at the time of presentation. For example, the presentation unit customizes the display content according to the user's current areas of interest at the time of presentation. For example, the presentation unit can preferentially display reviews in genres in which the user is currently interested. The presentation unit can also customize and display related reviews based on the user's current areas of interest. Furthermore, the presentation unit can analyze the user's current areas of interest and suggest optimal display content. For example, the presentation unit evaluates the user's current areas of interest based on criteria such as search history and topics being followed, and customizes the display content. In this way, more appropriate information can be provided by customizing the display content according to the user's current areas of interest.

[0054] The presentation unit can improve the display method by reflecting user feedback at the time of presentation. For example, the presentation unit can improve the display method by reflecting user feedback at the time of presentation. For example, the presentation unit can improve the display method based on feedback provided by the user and provide an optimal display. The presentation unit can also reflect user feedback and highlight and display highly reliable word-of-mouth reviews. Furthermore, the presentation unit can analyze user feedback and identify areas for improvement in the display method. For example, the presentation unit improves the display method based on survey results and review content. In this way, the display method can be improved by reflecting user feedback, and more appropriate information can be provided.

[0055] The presentation unit can select the optimal display method by taking into consideration the user's device information when presenting information. For example, the presentation unit selects the optimal display method by taking into consideration the user's device information when presenting information. For example, if the user is using a smartphone, the presentation unit can provide a display method that matches the screen size. Furthermore, if the user is using a tablet, the presentation unit can also provide a display method that is optimized for a large screen. Furthermore, if the user is using a smartwatch, the presentation unit can also provide a simple and highly visible display method. For example, the presentation unit evaluates device information based on criteria such as the device type, screen size, and OS, and selects the optimal display method. In this way, the optimal display method can be selected by taking into consideration the user's device information, and more appropriate information can be provided.

[0056] The presentation unit can make the display content multilingual when presented in accordance with the user's language setting. For example, the presentation unit can make the display content multilingual when presented in accordance with the user's language setting. The presentation unit can automatically translate the display content based on the language setting of the user's device, for example. The presentation unit can also provide a language switching function when the user uses multiple languages. Furthermore, when the user selects a specific language, the presentation unit can provide the display content in that language. For example, the presentation unit evaluates the language setting based on criteria such as the browser's language setting and the user's profile information, and makes the display content multilingual. This makes it possible to provide more appropriate information by making the display content multilingual according to the user's language setting.

[0057] The presentation unit can analyze the user's social media activity at the time of presentation and prioritize displaying related reviews. For example, the presentation unit can analyze the user's social media activity at the time of presentation and prioritize displaying related reviews. For example, the presentation unit can prioritize displaying reviews related to places where the user has checked in on social media. The presentation unit can also analyze the content of the user's posts on social media and prioritize displaying related reviews. Furthermore, the presentation unit can also prioritize displaying related reviews with reference to the activities of the user's friends on social media. For example, the presentation unit evaluates the social media activity based on criteria such as check-in information, post content, and friend activities on social media, and prioritizes displaying related reviews. In this way, by analyzing the user's social media activity, related reviews can be prioritized and more appropriate information can be provided.

[0058] The system according to the embodiment is not limited to the above-described example, and various modifications are possible, for example, as follows.

[0059] The collection unit can analyze the update frequency of each review site and set an optimal collection schedule. For example, it can analyze the update frequency of each review site and prioritize collecting information from frequently updated sites. It can also set a regular collection schedule for sites that are updated infrequently, allowing for efficient information collection. Furthermore, for sites that are updated irregularly, AI can dynamically adjust the schedule and collect information at the optimal time. This allows for efficient collection of review information by setting an optimal collection schedule based on the update frequency of each review site.

[0060] The collection unit can evaluate the reliability of reviews during collection and apply a filter to exclude unreliable information. For example, it can analyze the past posting history of review posters and exclude reviews from unreliable posters. It can also analyze the content of reviews and exclude reviews that contain slander or overly positive content. It can also filter out unreliable information by taking into account the posting date and time of the review and the number of ratings. In this way, it is possible to collect highly reliable review information by evaluating the reliability of reviews and excluding unreliable information.

[0061] During learning, the learning unit can optimize the learning algorithm by referring to past learning data. For example, it can identify defamatory patterns based on past learning data and optimize the learning algorithm. It can also refer to past learning data to reflect the characteristics of stealth marketing in the learning algorithm. Furthermore, it can analyze past learning data and optimize the algorithm for identifying reviews that are inconsistent with the reality. In this way, by referring to past learning data, it is possible to optimize the learning algorithm and improve accuracy.

[0062] The filtering unit can improve the accuracy of filtering by taking into account the interrelationships between reviews during filtering. For example, it can analyze the content of reviews and group and filter reviews that are mutually related. It can also take into account the interrelationships between review posters and preferentially filter highly reliable information. It can also take into account the number of ratings and comments on reviews to filter information that is mutually related. In this way, by taking into account the interrelationships between reviews, it is possible to improve the accuracy of filtering.

[0063] When presenting information, the presentation unit can select the optimal display method by referring to the user's past operation history. For example, it can preferentially apply the display method of a review site that the user has previously given a high rating. It can also suggest the optimal display method based on the user's past operation history. It can also analyze the user's past operation history and select the optimal display method. In this way, by referring to the user's past operation history, it is possible to select the optimal display method and provide more appropriate information.

[0064] The processing flow of the first embodiment will be briefly explained below.

[0065] Step 1: The collection unit collects word-of-mouth information. For example, the collection unit collects word-of-mouth information from various review sites and stores it in a database. The collection unit can collect word-of-mouth information from various genres, such as restaurant reviews and product reviews. Step 2: The learning unit analyzes the review information collected by the collection unit and generates a model for identifying malicious reviews and posted photos. For example, the learning unit uses AI to generate a model for identifying slanderous or false reviews, stealth marketing, etc. For example, the AI ​​uses text generation AI (e.g., LLM) to analyze review information and identify malicious reviews. The learning unit can also use multimodal generation AI to analyze posted photos and identify malicious posted photos. Step 3: The filtering unit uses the model generated by the learning unit to filter out malicious reviews and posted photos. For example, the filtering unit uses the learned model to automatically detect and filter out slanderous or false reviews, stealth marketing, etc. For example, the filtering unit can use AI to detect and filter malicious reviews. Step 4: The presentation unit presents the reliable reviews filtered by the filtering unit. For example, the presentation unit presents the reliable reviews on each review site. For example, the presentation unit displays the filtered reliable reviews to the user through a website or a mobile application.

[0066] (Example 2) A review filtering system according to an embodiment of the present invention analyzes vast amounts of review information and filters out malicious reviews and posted photos. The review filtering system collects review information, uses AI to train the system, filters out malicious reviews and posted photos, and presents only reliable reviews on each review site. For example, the review filtering system collects review information from each review site and stores it in a database. Next, the collected review information is trained using AI. The AI ​​analyzes the review information and generates a model for identifying malicious reviews and posted photos. For example, it can identify slanderous, false reviews, stealth marketing, and the like. The trained model is used to filter malicious reviews and posted photos. For example, it can automatically detect and filter slanderous, false reviews, stealth marketing, and the like. Finally, it presents only reliable reviews on each review site. This allows users to obtain reliable review information and avoid making decisions based on erroneous information. The review filtering system thus prevents the harm caused by malicious reviews and posted photos and provides reliable review information. For example, restaurant review sites display only reliable reviews, allowing users to choose restaurants with confidence. Similarly, product review sites display only reliable reviews from users who have actually used the products, allowing users to choose products based on accurate information.

[0067] A review filtering system according to an embodiment includes a collection unit, a learning unit, a filtering unit, and a presentation unit. The collection unit collects review information. For example, the collection unit collects review information from various review sites and stores the collected information in a database. The collection unit can collect review information from various genres, such as restaurant reviews and product reviews. The learning unit analyzes the review information collected by the collection unit and generates a model for identifying malicious reviews and posted photos. The learning unit, for example, uses AI to generate a model for identifying slanderous comments, reviews that are inaccurate, stealth marketing, and the like. For example, the AI ​​uses a text generation AI (e.g., LLM) to analyze review information and identify malicious reviews. The learning unit can also analyze posted photos and identify malicious photos using a multimodal generation AI. The filtering unit uses the model generated by the learning unit to filter malicious reviews and posted photos. For example, the filtering unit uses the trained model to automatically detect and filter slanderous comments, reviews that are inaccurate, stealth marketing, and the like. For example, the filtering unit can use AI to detect and filter malicious reviews. The presentation unit presents reliable reviews filtered by the filtering unit. The presentation unit presents the reliable reviews on each review site, for example. For example, the presentation unit displays the filtered reliable reviews to users via a website or a mobile application. This allows the review filtering system according to the embodiment to efficiently collect, analyze, filter, and present review information.

[0068] The collection unit can collect review information from each review site and store it in a database. For example, the collection unit collects review information from each review site and stores it in a database. For example, the collection unit can collect information from review sites such as Amazon, Yelp, and TripAdvisor. For example, the collection unit automatically collects review information from each review site using web scraping technology. The collection unit can also acquire review information from each review site using an API. For example, the collection unit periodically acquires review information using the API of each review site and stores it in a database. This makes it possible to efficiently collect review information from each review site and store it in a database.

[0069] The learning unit can analyze the collected word-of-mouth information and generate a model for identifying slander, false word-of-mouth, and stealth marketing. The learning unit, for example, analyzes the collected word-of-mouth information and generates a model for identifying slander, false word-of-mouth, and stealth marketing. The learning unit can, for example, use AI to learn and identify specific patterns of slander. The learning unit can also learn and identify characteristics of false word-of-mouth. Furthermore, the learning unit can learn and identify characteristics of stealth marketing. For example, the learning unit uses text generation AI (e.g., LLM) to analyze word-of-mouth information and generate a model for identifying slander, false word-of-mouth, and stealth marketing. This makes it possible to generate a model for identifying slander, false word-of-mouth, stealth marketing, and the like.

[0070] The filtering unit can use the trained model to automatically detect and filter defamatory comments, false reviews, and stealth marketing. The filtering unit can, for example, use the trained model to automatically detect and filter defamatory comments, false reviews, and stealth marketing. The filtering unit can, for example, use AI to automatically detect and filter defamatory comments. The filtering unit can also automatically detect and filter false reviews. The filtering unit can also automatically detect and filter stealth marketing reviews. For example, the filtering unit can use text generation AI (e.g., LLM) to analyze review information and automatically detect and filter defamatory comments, false reviews, and stealth marketing. This makes it possible to automatically detect and filter defamatory comments, false reviews, stealth marketing, etc.

[0071] The presentation unit can present the reliable reviews filtered by the filtering unit on each review site. For example, the presentation unit presents the reliable reviews filtered by the filtering unit on each review site. For example, the presentation unit can display the reliable reviews to users through a website or a mobile application. The presentation unit can also send the reliable reviews to users by email. Furthermore, the presentation unit can share the reliable reviews on social media. For example, the presentation unit can share the filtered reliable reviews on social media such as Facebook or Twitter. This allows the reliable reviews to be presented on each review site.

[0072] The collection unit can estimate the user's emotions and adjust the timing of collecting word-of-mouth information based on the estimated user emotions. The collection unit, for example, estimates the user's emotions and adjusts the timing of collecting word-of-mouth information based on the estimated user emotions. The collection unit can estimate the user's emotions using text analysis, for example. The collection unit can also estimate the user's emotions using facial expression recognition technology. Furthermore, the collection unit can estimate the user's emotions using behavioral data analysis. For example, if the user is feeling stressed, the collection unit can delay the collection timing and collect information when the user is relaxed. Furthermore, if the user is excited, the collection unit can advance the collection timing and collect the latest word-of-mouth information in real time. Furthermore, if the user is tired, the collection unit can adjust the collection timing and collect information while the user is resting. In this way, by adjusting the collection timing according to the user's emotions, word-of-mouth information can be collected at a more appropriate time.

[0073] The collection unit can analyze the update frequency of each review site and set an optimal collection schedule. The collection unit, for example, analyzes the update frequency of each review site and sets an optimal collection schedule. The collection unit, for example, can analyze the update frequency of each review site and prioritize information collection from frequently updated sites. The collection unit can also set a regular collection schedule for sites that are updated infrequently and collect information efficiently. Furthermore, for sites that are updated irregularly, the collection unit can dynamically adjust the schedule using AI to collect information at the optimal timing. For example, the collection unit sets an optimal collection schedule based on criteria such as daily updates or weekly updates. As a result, review information can be collected efficiently by setting an optimal collection schedule based on the update frequency of each review site.

[0074] The collection unit can evaluate the reliability of reviews at the time of collection and apply a filter that excludes unreliable information. For example, the collection unit can analyze the past posting history of review posters and exclude reviews from unreliable posters. The collection unit can also analyze the content of reviews and exclude reviews that include slander or overly positive content. Furthermore, the collection unit can filter out unreliable information by taking into account the posting date and time of the review and the number of ratings. For example, the collection unit applies a filter that excludes unreliable information based on criteria such as spam, false information, and exaggerated content. In this way, reliable review information can be collected by evaluating the reliability of reviews and excluding unreliable information.

[0075] The collection unit can apply different collection algorithms depending on the genre and category of the review during collection. For example, the collection unit applies different collection algorithms depending on the genre and category of the review during collection. For example, when collecting restaurant reviews, the collection unit can adjust the collection algorithm based on the type of food and the service rating. Furthermore, when collecting product reviews, the collection unit can change the collection algorithm depending on the product category and purpose of use. Furthermore, when collecting travel reviews, the collection unit can optimize the collection algorithm based on the ratings of tourist destinations and accommodations. For example, the collection unit classifies genres and categories based on criteria such as product category and service category, and applies the corresponding collection algorithm. In this way, more appropriate review information can be collected by adjusting the collection algorithm depending on the genre and category of the review.

[0076] The collection unit can estimate the user's emotions and determine the priority of the word-of-mouth information to be collected based on the estimated user's emotions. The collection unit can, for example, estimate the user's emotions and determine the priority of the word-of-mouth information to be collected based on the estimated user's emotions. The collection unit can, for example, estimate the user's emotions using text analysis. The collection unit can also estimate the user's emotions using facial expression recognition technology. Furthermore, the collection unit can also estimate the user's emotions using behavioral data analysis. For example, the collection unit can prioritize collecting reliable word-of-mouth information when the user is feeling anxious. Furthermore, the collection unit can prioritize collecting the latest word-of-mouth information when the user is excited. Furthermore, the collection unit can prioritize collecting detailed word-of-mouth information when the user is relaxed. In this way, by determining the priority of the word-of-mouth information to be collected according to the user's emotions, more appropriate information can be preferentially collected.

[0077] The collection unit can prioritize collecting highly relevant word-of-mouth information by taking into account the user's geographical location information during collection. For example, the collection unit prioritizes collecting highly relevant word-of-mouth information by taking into account the user's geographical location information during collection. For example, the collection unit can identify the user's current location using GPS data and prioritize collecting word-of-mouth information in the vicinity. The collection unit can also estimate the user's geographical location using an IP address and prioritize collecting word-of-mouth information related to that area. For example, the collection unit prioritizes collecting word-of-mouth information about restaurants close to the user's current location. The collection unit can also prioritize collecting word-of-mouth information about tourist destinations where the user is traveling. Furthermore, the collection unit can prioritize collecting word-of-mouth information about stores and services in the area where the user lives. In this way, highly relevant word-of-mouth information can be prioritized by taking into account the user's geographical location information.

[0078] The collection unit can analyze the user's social media activities at the time of collection and collect related word-of-mouth information. For example, the collection unit can analyze the user's social media activities at the time of collection and collect related word-of-mouth information. For example, the collection unit can collect word-of-mouth information about places where the user has checked in on social media. The collection unit can also analyze the content of the user's social media posts and collect related word-of-mouth information. Furthermore, the collection unit can collect related word-of-mouth information by referring to the activities of the user's friends on social media. For example, the collection unit preferentially collects word-of-mouth information about restaurants where the user has checked in on social media. The collection unit can also analyze the content of the user's social media posts and collect reviews of related products. Furthermore, the collection unit can analyze the activities of the user's friends on social media and collect word-of-mouth information about related tourist destinations. In this way, related word-of-mouth information can be efficiently collected by analyzing the user's social media activities.

[0079] The collection unit can customize the collection method by reflecting the user's past feedback when collecting data. For example, the collection unit customizes the collection method by reflecting the user's past feedback when collecting data. For example, the collection unit can prioritize collecting information from review sites that the user has previously given high ratings to. The collection unit can also adjust the collection algorithm based on feedback provided by the user in the past. Furthermore, the collection unit can analyze the user's past feedback and suggest an optimal collection method. For example, the collection unit prioritizes collecting information from review sites for restaurants that the user has previously given high ratings to. The collection unit can also adjust the product review collection algorithm based on feedback provided by the user in the past. Furthermore, the collection unit can analyze the user's past feedback and suggest an optimal collection method for review information on tourist destinations. In this way, the collection method can be customized by reflecting the user's past feedback to collect more appropriate information.

[0080] The learning unit can estimate the user's emotions and select training data based on the estimated user's emotions. For example, the learning unit can estimate the user's emotions and select training data based on the estimated user's emotions. For example, the learning unit can estimate the user's emotions using text analysis. The learning unit can also estimate the user's emotions using facial expression recognition technology. Furthermore, the learning unit can estimate the user's emotions using behavioral data analysis. For example, if the user is feeling anxious, the learning unit can select highly reliable word-of-mouth information as training data. Furthermore, if the user is excited, the learning unit can select the most recent word-of-mouth information as training data. Furthermore, if the user is relaxed, the learning unit can select detailed word-of-mouth information as training data. In this way, by selecting training data based on the user's emotions, more appropriate training data can be used.

[0081] The learning unit can optimize the learning algorithm by referring to past learning data during learning. For example, the learning unit can optimize the learning algorithm by referring to past learning data during learning. For example, the learning unit can identify slanderous patterns based on past learning data and optimize the learning algorithm. The learning unit can also reflect stealth marketing characteristics in the learning algorithm by referring to past learning data. Furthermore, the learning unit can analyze past learning data and optimize an algorithm for identifying word-of-mouth that differs from the actual situation. For example, the learning unit can optimize the learning algorithm by referring to previous training data or past models. In this way, by referring to past learning data, the learning algorithm can be optimized and accuracy can be improved.

[0082] The learning unit can generate different learning models during learning depending on the genre and category of the review. For example, the learning unit generates different learning models during learning depending on the genre and category of the review. For example, the learning unit can generate a learning model specialized for restaurant reviews to analyze food ratings and service quality. The learning unit can also generate a learning model specialized for product reviews to analyze product performance and usability. Furthermore, the learning unit can generate a learning model specialized for travel reviews to analyze ratings of tourist spots and accommodations. For example, the learning unit classifies genres and categories based on criteria such as product category and service category, and generates a learning model accordingly. In this way, more appropriate analysis is possible by generating a learning model according to the genre and category of the review.

[0083] The learning unit can update the learning data by reflecting user feedback during learning. For example, the learning unit updates the learning data by reflecting user feedback during learning. For example, the learning unit can update the learning data based on feedback provided by the user to improve the accuracy of the algorithm. The learning unit can also reflect user feedback to improve the accuracy of detecting slander and stealth marketing. Furthermore, the learning unit can analyze user feedback and review the criteria for selecting learning data. For example, the learning unit updates the learning data based on survey results and review content. In this way, the learning data can be updated by reflecting user feedback, thereby improving the accuracy of the algorithm.

[0084] The learning unit can estimate the user's emotions and adjust the frequency of learning based on the estimated user emotions. For example, the learning unit can estimate the user's emotions and adjust the frequency of learning based on the estimated user emotions. For example, the learning unit can estimate the user's emotions using text analysis. The learning unit can also estimate the user's emotions using facial expression recognition technology. Furthermore, the learning unit can estimate the user's emotions using behavioral data analysis. For example, if the user is feeling anxious, the learning unit can increase the frequency of learning to provide more reliable information. If the user is excited, the learning unit can adjust the frequency of learning to provide the latest information. Furthermore, if the user is relaxed, the learning unit can adjust the frequency of learning to provide more detailed information. In this way, by adjusting the frequency of learning based on the user's emotions, learning can be performed at a more appropriate time.

[0085] The learning unit can weight the learning data based on the time of submission of the review during learning. For example, the learning unit weights the learning data based on the time of submission of the review during learning. For example, the learning unit can weight the most recent review information highly and reflect this in the learning algorithm. The learning unit can also weight older review information less highly and reflect this in the learning algorithm. Furthermore, the learning unit can assign appropriate weights based on the time of submission of the review to optimize the learning data. For example, the learning unit evaluates and weights the time of submission based on criteria such as the date and time of posting or whether the review is before or after an event. In this way, more appropriate learning is possible by weighting the learning data based on the time of submission of the review.

[0086] The learning unit can integrate information from different data sources to enrich the learning data during learning. For example, the learning unit can integrate information from different data sources to enrich the learning data during learning. For example, the learning unit can integrate information collected from various review sites to enrich the learning data. The learning unit can also integrate information from social media and blogs to enrich the learning data. Furthermore, the learning unit can integrate information from public databases and news sites to enrich the learning data. For example, the learning unit collects information from different data sources such as social media and news sites, and integrates it to enrich the learning data. In this way, by integrating information from different data sources, the learning data can be enriched and its accuracy can be improved.

[0087] The learning unit can evaluate the reliability of reviews and exclude low-reliability data during learning. For example, the learning unit evaluates the reliability of reviews and excludes low-reliability data during learning. For example, the learning unit can analyze the past posting history of review posters and exclude low-reliability data. The learning unit can also analyze the content of reviews and exclude data that contains slander or overly positive content. Furthermore, the learning unit can filter low-reliability data by taking into account the posting date and time of the review and the number of ratings. For example, the learning unit excludes low-reliability data based on criteria such as spam, false information, and exaggerated content. In this way, by evaluating the reliability of reviews and excluding low-reliability data, more reliable training data can be used.

[0088] The filtering unit can estimate a user's emotion and adjust filtering criteria based on the estimated user's emotion. For example, the filtering unit can estimate a user's emotion and adjust filtering criteria based on the estimated user's emotion. For example, the filtering unit can estimate a user's emotion using text analysis. The filtering unit can also estimate a user's emotion using facial expression recognition technology. Furthermore, the filtering unit can estimate a user's emotion using behavioral data analysis. For example, if the user is feeling anxious, the filtering unit can apply strict filtering criteria to provide only reliable information. If the user is excited, the filtering unit can apply flexible filtering criteria to provide the latest information. Furthermore, if the user is relaxed, the filtering unit can apply filtering criteria that include detailed information. As a result, more appropriate filtering can be achieved by adjusting the filtering criteria based on the user's emotion.

[0089] The filtering unit can improve the accuracy of filtering by taking into account the interrelationships between reviews when filtering. The filtering unit can improve the accuracy of filtering by taking into account the interrelationships between reviews when filtering, for example. The filtering unit can, for example, analyze the content of the reviews and group and filter reviews that are related to each other. The filtering unit can also take into account the interrelationships between the posters of the reviews and preferentially filter highly reliable information. Furthermore, the filtering unit can filter information that is related to each other by taking into account the number of ratings and comments on the reviews. For example, the filtering unit evaluates the interrelationships between reviews based on criteria such as co-occurrence networks and highly related keywords, and improves the accuracy of filtering. In this way, the accuracy of filtering can be improved by taking into account the interrelationships between reviews.

[0090] The filtering unit can perform filtering while taking into consideration the attribute information of the person who submitted the review. For example, the filtering unit performs filtering while taking into consideration the attribute information of the person who submitted the review. The filtering unit can filter highly reliable information while taking into consideration, for example, the age and gender of the person who submitted the review. The filtering unit can also filter highly reliable information by analyzing the past posting history of the person who submitted the review. Furthermore, the filtering unit can filter highly reliable information while taking into consideration the region and occupation of the person who submitted the review. For example, the filtering unit evaluates the attribute information of the person who submitted the review based on criteria such as age, gender, and region, and performs filtering. In this way, more reliable information can be filtered by taking into consideration the attribute information of the person who submitted the review.

[0091] The filtering unit can weight the filtering based on the frequency of reviews submitted during filtering. For example, the filtering unit weights the filtering based on the frequency of reviews submitted during filtering. For example, the filtering unit can weight reviews that are posted frequently, thereby filtering out highly reliable information. The filtering unit can also weight reviews that are submitted infrequently, thereby filtering out less reliable information. Furthermore, the filtering unit can weight reviews appropriately based on the frequency of reviews submitted, thereby improving the accuracy of filtering. For example, the filtering unit evaluates and weights the submission frequency based on criteria such as the number of posts and the number of posts within a period. As a result, more appropriate filtering is possible by weighting the filtering based on the frequency of reviews submitted.

[0092] The filtering unit can estimate a user's emotion and adjust the order in which the filtering results are displayed based on the estimated user's emotion. The filtering unit can, for example, estimate a user's emotion and adjust the order in which the filtering results are displayed based on the estimated user's emotion. The filtering unit can, for example, estimate a user's emotion using text analysis. The filtering unit can also estimate a user's emotion using facial expression recognition technology. The filtering unit can also estimate a user's emotion using behavioral data analysis. For example, the filtering unit can prioritize displaying highly reliable reviews when the user is feeling anxious. The filtering unit can also prioritize displaying the most recent reviews when the user is excited. The filtering unit can also prioritize displaying detailed reviews when the user is relaxed. In this way, by adjusting the order in which the filtering results are displayed based on the user's emotion, more appropriate information can be provided.

[0093] The filtering unit can perform filtering taking into account the geographical distribution of reviews. For example, the filtering unit can perform filtering taking into account the geographical distribution of reviews. For example, the filtering unit can prioritize filtering reviews close to the user's current location. The filtering unit can also analyze the geographical distribution of reviews and filter reviews related to a specific region. Furthermore, the filtering unit can filter highly reliable information taking into account the geographical distribution of reviews. For example, the filtering unit evaluates the geographical distribution and performs filtering based on criteria such as GPS data or the number of posts by region. This enables more appropriate filtering by taking into account the geographical distribution of reviews.

[0094] The filtering unit can improve the accuracy of filtering by referring to literature related to the word-of-mouth reviews during filtering. For example, the filtering unit can improve the accuracy of filtering by referring to literature related to the word-of-mouth reviews during filtering. For example, the filtering unit can analyze the content of the word-of-mouth reviews and filter by referring to related academic papers and articles. The filtering unit can also analyze the content of the word-of-mouth reviews and filter by referring to related news articles. Furthermore, the filtering unit can analyze the content of the word-of-mouth reviews and filter by referring to related blogs and reviews. For example, the filtering unit evaluates related literature based on criteria such as academic papers and industry reports, and improves the accuracy of filtering. In this way, the accuracy of filtering can be improved by referring to literature related to the word-of-mouth reviews.

[0095] The filtering unit can perform filtering taking into account the market value of the word-of-mouth when filtering. For example, the filtering unit performs filtering taking into account the market value of the word-of-mouth when filtering. For example, the filtering unit can analyze the content of the word-of-mouth and preferentially filter information with high market value. The filtering unit can also analyze the content of the word-of-mouth and exclude information with low market value. Furthermore, the filtering unit can evaluate the market value of the word-of-mouth and apply appropriate filtering criteria. For example, the filtering unit evaluates the market value based on criteria such as sales data and customer ratings and performs filtering. In this way, more appropriate filtering is possible by taking into account the market value of the word-of-mouth.

[0096] The presentation unit can estimate the user's emotions and adjust the display method of the reviews based on the estimated user emotions. The presentation unit can, for example, estimate the user's emotions and adjust the display method of the reviews based on the estimated user emotions. The presentation unit can, for example, estimate the user's emotions using text analysis. The presentation unit can also estimate the user's emotions using facial expression recognition technology. Furthermore, the presentation unit can also estimate the user's emotions using behavioral data analysis. For example, if the user is feeling anxious, the presentation unit can highlight and display highly reliable reviews. If the user is excited, the presentation unit can highlight and display the most recent reviews. Furthermore, if the user is relaxed, the presentation unit can highlight and display detailed reviews. In this way, more appropriate information can be provided by adjusting the display method of the reviews based on the user's emotions.

[0097] The presentation unit can select the optimal display method by referring to the user's past operation history when presenting the information. For example, the presentation unit can select the optimal display method by referring to the user's past operation history when presenting the information. For example, the presentation unit can preferentially apply the display method of a review site that the user has previously given a high rating to. The presentation unit can also suggest the optimal display method based on the user's past operation history. Furthermore, the presentation unit can analyze the user's past operation history and select the optimal display method. For example, the presentation unit evaluates the past operation history based on criteria such as click history and browsing history, and selects the optimal display method. In this way, the optimal display method can be selected by referring to the user's past operation history, and more appropriate information can be provided.

[0098] The presentation unit can customize the display content according to the user's current areas of interest at the time of presentation. For example, the presentation unit customizes the display content according to the user's current areas of interest at the time of presentation. For example, the presentation unit can preferentially display reviews in genres in which the user is currently interested. The presentation unit can also customize and display related reviews based on the user's current areas of interest. Furthermore, the presentation unit can analyze the user's current areas of interest and suggest optimal display content. For example, the presentation unit evaluates the user's current areas of interest based on criteria such as search history and topics being followed, and customizes the display content. In this way, more appropriate information can be provided by customizing the display content according to the user's current areas of interest.

[0099] The presentation unit can improve the display method by reflecting user feedback at the time of presentation. For example, the presentation unit can improve the display method by reflecting user feedback at the time of presentation. For example, the presentation unit can improve the display method based on feedback provided by the user and provide an optimal display. The presentation unit can also reflect user feedback and highlight and display highly reliable word-of-mouth reviews. Furthermore, the presentation unit can analyze user feedback and identify areas for improvement in the display method. For example, the presentation unit improves the display method based on survey results and review content. In this way, the display method can be improved by reflecting user feedback, and more appropriate information can be provided.

[0100] The presentation unit can estimate the user's emotions and adjust the display order of the reviews based on the estimated user emotions. The presentation unit can, for example, estimate the user's emotions and adjust the display order of the reviews based on the estimated user emotions. The presentation unit can, for example, estimate the user's emotions using text analysis. The presentation unit can also estimate the user's emotions using facial expression recognition technology. Furthermore, the presentation unit can also estimate the user's emotions using behavioral data analysis. For example, if the user is feeling anxious, the presentation unit can prioritize displaying highly reliable reviews. Furthermore, if the user is excited, the presentation unit can prioritize displaying the most recent reviews. Furthermore, if the user is relaxed, the presentation unit can prioritize displaying detailed reviews. In this way, by adjusting the display order of the reviews based on the user's emotions, more appropriate information can be provided.

[0101] The presentation unit can select the optimal display method by taking into consideration the user's device information when presenting information. For example, the presentation unit selects the optimal display method by taking into consideration the user's device information when presenting information. For example, if the user is using a smartphone, the presentation unit can provide a display method that matches the screen size. Furthermore, if the user is using a tablet, the presentation unit can also provide a display method that is optimized for a large screen. Furthermore, if the user is using a smartwatch, the presentation unit can also provide a simple and highly visible display method. For example, the presentation unit evaluates device information based on criteria such as the device type, screen size, and OS, and selects the optimal display method. In this way, the optimal display method can be selected by taking into consideration the user's device information, and more appropriate information can be provided.

[0102] The presentation unit can make the display content multilingual when presented in accordance with the user's language setting. For example, the presentation unit can make the display content multilingual when presented in accordance with the user's language setting. The presentation unit can automatically translate the display content based on the language setting of the user's device, for example. The presentation unit can also provide a language switching function when the user uses multiple languages. Furthermore, when the user selects a specific language, the presentation unit can provide the display content in that language. For example, the presentation unit evaluates the language setting based on criteria such as the browser's language setting and the user's profile information, and makes the display content multilingual. This makes it possible to provide more appropriate information by making the display content multilingual according to the user's language setting.

[0103] The presentation unit can analyze the user's social media activity at the time of presentation and prioritize displaying related reviews. For example, the presentation unit can analyze the user's social media activity at the time of presentation and prioritize displaying related reviews. For example, the presentation unit can prioritize displaying reviews related to places where the user has checked in on social media. The presentation unit can also analyze the content of the user's posts on social media and prioritize displaying related reviews. Furthermore, the presentation unit can also prioritize displaying related reviews with reference to the activities of the user's friends on social media. For example, the presentation unit evaluates the social media activity based on criteria such as check-in information, post content, and friend activities on social media, and prioritizes displaying related reviews. In this way, by analyzing the user's social media activity, related reviews can be prioritized and more appropriate information can be provided. === Hard Collateral 1-1 === Each of the multiple elements including the collection unit, learning unit, filtering unit, and presentation unit described above is realized, for example, by at least one of the smart device 14 and the data processing device 12. For example, the collection unit is realized by the control unit 46A of the smart device 14 and collects review information from each review site and stores it in the database 24. The learning unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and analyzes the collected review information using AI to generate a model for identifying malicious reviews and posted photos. The filtering unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and filters malicious reviews and posted photos using the learned model. The presentation unit is realized, for example, by the control unit 46A of the smart device 14 and presents the filtered, reliable reviews on each review site. === Hard Collateral 1-2 === Each of the multiple elements including the collection unit, learning unit, filtering unit, and presentation unit described above is realized, for example, by at least one of the smart glasses 214 and the data processing device 12. For example, the collection unit is realized by the control unit 46A of the smart glasses 214 and collects review information from each review site and stores it in the database 24. The learning unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and analyzes the collected review information using AI to generate a model for identifying malicious reviews and posted photos. The filtering unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and filters malicious reviews and posted photos using the learned model. The presentation unit is realized, for example, by the control unit 46A of the smart glasses 214 and presents the filtered, reliable reviews on each review site. === Hard Collateral 1-3 === Each of the multiple elements including the collection unit, learning unit, filtering unit, and presentation unit described above is realized, for example, by at least one of the headset-type terminal 314 and the data processing device 12. For example, the collection unit is realized by the control unit 46A of the headset-type terminal 314, and collects review information from each review site and stores it in the database 24. The learning unit is realized, for example, by the specific processing unit 290 of the data processing device 12, and analyzes the collected review information using AI and generates a model for identifying malicious reviews and posted photos. The filtering unit is realized, for example, by the specific processing unit 290 of the data processing device 12, and filters malicious reviews and posted photos using the learned model. The presentation unit is realized, for example, by the control unit 46A of the headset-type terminal 314, and presents the filtered, reliable reviews on each review site. === Hard Collateral 1-4 === Each of the multiple elements including the collection unit, learning unit, filtering unit, and presentation unit described above is realized, for example, by at least one of the robot 414 and the data processing device 12. For example, the collection unit is realized by the control unit 46A of the robot 414 and collects review information from each review site and stores it in the database 24. The learning unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and analyzes the collected review information using AI to generate a model for identifying malicious reviews and posted photos. The filtering unit is realized, for example, by the specific processing unit 290 of the data processing device 12 and filters malicious reviews and posted photos using the learned model. The presentation unit is realized, for example, by the control unit 46A of the robot 414 and presents the filtered, reliable reviews on each review site.

[0104] The system according to the embodiment is not limited to the above-described example, and various modifications are possible, for example, as follows.

[0105] The collection unit can estimate the user's emotions and determine the priority of word-of-mouth information to be collected based on the estimated user's emotions. For example, if the user is feeling anxious, highly reliable word-of-mouth information can be collected with priority. Also, if the user is excited, the latest word-of-mouth information can be collected with priority. Furthermore, if the user is relaxed, detailed word-of-mouth information can be collected with priority. In this way, by determining the priority of word-of-mouth information to be collected according to the user's emotions, more appropriate information can be collected with priority.

[0106] The learning unit can estimate the user's emotions and select training data based on the estimated user emotions. For example, if the user is feeling anxious, highly reliable word-of-mouth information can be selected as training data. Also, if the user is excited, the latest word-of-mouth information can be selected as training data. Furthermore, if the user is relaxed, detailed word-of-mouth information can be selected as training data. In this way, by selecting training data based on the user's emotions, more appropriate training data can be used.

[0107] The filtering unit can estimate the user's emotions and adjust the filtering criteria based on the estimated user emotions. For example, if the user is feeling anxious, strict filtering criteria can be applied to provide only reliable information. Alternatively, if the user is excited, flexible filtering criteria can be applied to provide the latest information. Furthermore, if the user is relaxed, filtering criteria including detailed information can be applied. This allows for more appropriate filtering by adjusting the filtering criteria based on the user's emotions.

[0108] The presentation unit can estimate the user's emotions and adjust the way reviews are displayed based on the estimated user emotions. For example, if the user is feeling anxious, the most reliable reviews can be highlighted and displayed. If the user is excited, the most recent reviews can be highlighted and displayed. Furthermore, if the user is relaxed, detailed reviews can be highlighted and displayed. In this way, by adjusting the way reviews are displayed based on the user's emotions, more appropriate information can be provided.

[0109] The presentation unit can estimate the user's emotions and adjust the display order of reviews based on the estimated user emotions. For example, if the user is feeling anxious, it can prioritize displaying highly reliable reviews. Also, if the user is excited, it can prioritize displaying the most recent reviews. Furthermore, if the user is relaxed, it can prioritize displaying detailed reviews. In this way, by adjusting the display order of reviews based on the user's emotions, it is possible to provide more appropriate information.

[0110] The collection unit can analyze the update frequency of each review site and set an optimal collection schedule. For example, it can analyze the update frequency of each review site and prioritize collecting information from frequently updated sites. It can also set a regular collection schedule for sites that are updated infrequently, allowing for efficient information collection. Furthermore, for sites that are updated irregularly, AI can dynamically adjust the schedule and collect information at the optimal time. This allows for efficient collection of review information by setting an optimal collection schedule based on the update frequency of each review site.

[0111] The collection unit can evaluate the reliability of reviews during collection and apply a filter to exclude unreliable information. For example, it can analyze the past posting history of review posters and exclude reviews from unreliable posters. It can also analyze the content of reviews and exclude reviews that contain slander or overly positive content. It can also filter out unreliable information by taking into account the posting date and time of the review and the number of ratings. In this way, it is possible to collect highly reliable review information by evaluating the reliability of reviews and excluding unreliable information.

[0112] During learning, the learning unit can optimize the learning algorithm by referring to past learning data. For example, it can identify defamatory patterns based on past learning data and optimize the learning algorithm. It can also refer to past learning data to reflect the characteristics of stealth marketing in the learning algorithm. Furthermore, it can analyze past learning data and optimize the algorithm for identifying reviews that are inconsistent with the reality. In this way, by referring to past learning data, it is possible to optimize the learning algorithm and improve accuracy.

[0113] The filtering unit can improve the accuracy of filtering by taking into account the interrelationships between reviews during filtering. For example, it can analyze the content of reviews and group and filter reviews that are mutually related. It can also take into account the interrelationships between review posters and preferentially filter highly reliable information. It can also take into account the number of ratings and comments on reviews to filter information that is mutually related. In this way, by taking into account the interrelationships between reviews, it is possible to improve the accuracy of filtering.

[0114] When presenting information, the presentation unit can select the optimal display method by referring to the user's past operation history. For example, it can preferentially apply the display method of a review site that the user has previously given a high rating. It can also suggest the optimal display method based on the user's past operation history. It can also analyze the user's past operation history and select the optimal display method. In this way, by referring to the user's past operation history, it is possible to select the optimal display method and provide more appropriate information.

[0115] The processing flow of the second embodiment will be briefly explained below.

[0116] Step 1: The collection unit collects word-of-mouth information. For example, the collection unit collects word-of-mouth information from various review sites and stores it in a database. The collection unit can collect word-of-mouth information from various genres, such as restaurant reviews and product reviews. Step 2: The learning unit analyzes the review information collected by the collection unit and generates a model for identifying malicious reviews and posted photos. For example, the learning unit uses AI to generate a model for identifying slanderous or false reviews, stealth marketing, etc. For example, the AI ​​uses text generation AI (e.g., LLM) to analyze review information and identify malicious reviews. The learning unit can also use multimodal generation AI to analyze posted photos and identify malicious posted photos. Step 3: The filtering unit uses the model generated by the learning unit to filter out malicious reviews and posted photos. For example, the filtering unit uses the learned model to automatically detect and filter out slanderous or false reviews, stealth marketing, etc. For example, the filtering unit can use AI to detect and filter malicious reviews. Step 4: The presentation unit presents the reliable reviews filtered by the filtering unit. For example, the presentation unit presents the reliable reviews on each review site. For example, the presentation unit displays the filtered reliable reviews to the user through a website or a mobile application.

[0117] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0118] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> Examples of generative AIs include the data generation model 58, such as a neural network model (e.g., a neural network model), and a neural network model (e.g., a neural network model). The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating speech, text data indicating text, and image data indicating an image is also input to the data generation model 58. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specification processing unit 290 performs the above-mentioned specification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation model 58 includes AIs other than the generative AI. The AI ​​other than the generative AI may be, for example, linear regression, logistic regression, decision tree, random forest, support vector machine (SVM), k-means clustering, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processes of each of the above-mentioned parts are performed by AI, the processes may be performed in part or entirely by AI, but are not limited to these examples. The processes performed by AI, including the generative AI, may be replaced with rule-based processes.

[0119] Furthermore, the processing by the data processing system 10 described above is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart device 14, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart device 14. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the smart device 14 or an external device, and the smart device 14 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0120] The correspondence between each part and the device or control part is not limited to the above example, and various modifications are possible.

[0121] [Second embodiment] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0122] 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.

[0123] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN and / or a LAN.

[0124] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0125] The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to instructions from the processor 46.

[0126] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the user's surroundings (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0127] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0128] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0129] 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.

[0130] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0131] In the smart glasses 214, the specific processing is performed by the processor 46. A specific processing program 60 is stored in the storage 50. The processor 46 reads the specific processing program 60 from the storage 50 and executes the read specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as the control unit 46A in accordance with the specific processing program 60 executed on the RAM 48. The smart glasses 214 also have a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can perform processing similar to that of the specific processing unit 290 using these models.

[0132] Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 communicates with the server device having the data generation model 58 to obtain a processing result (such as a prediction result) using the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device (for example, a mobile phone, a robot, a home appliance, etc.) owned by a user.

[0133] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0134] The data generation model 58 is a so-called generative AI. An example of the data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 receives a prompt containing an instruction, as well as inference data such as voice data representing speech, text data representing text, and image data representing an image. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The identification processing unit 290 performs the above-mentioned identification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation model 58 includes AI other than the generative AI. The AI ​​other than the generative AI may be, for example, linear regression, logistic regression, decision tree, random forest, support vector machine (SVM), k-means clustering, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processes of each of the above-mentioned parts are performed by AI, the processes may be performed in part or entirely by AI, but are not limited to these examples. The processes performed by AI, including the generative AI, may be replaced with rule-based processes.

[0135] The data processing system 210 according to the second embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 210 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart glasses 214, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart glasses 214. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the smart glasses 214 or an external device, etc., and the smart glasses 214 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

[0136] The correspondence between each part and the device or control part is not limited to the above example, and various modifications are possible.

[0137] [Third embodiment] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0138] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0139] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN and / or a LAN.

[0140] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0141] The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to instructions from the processor 46.

[0142] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the user's surroundings (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0143] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0144] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0145] 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.

[0146] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0147] In the headset type terminal 314, the identification process is performed by the processor 46. A identification program 60 is stored in the storage 50. The processor 46 reads the identification program 60 from the storage 50 and executes the read identification program 60 on the RAM 48. The identification process is realized by the processor 46 operating as a control unit 46A in accordance with the identification program 60 executed on the RAM 48. Note that the headset type terminal 314 has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can also perform processing similar to that of the identification processing unit 290 using these models.

[0148] Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 communicates with the server device having the data generation model 58 to obtain a processing result (such as a prediction result) using the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device (for example, a mobile phone, a robot, a home appliance, etc.) owned by a user.

[0149] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0150] The data generation model 58 is a so-called generative AI. An example of the data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 receives a prompt containing an instruction, as well as inference data such as voice data representing speech, text data representing text, and image data representing an image. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The identification processing unit 290 performs the above-mentioned identification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation model 58 includes AI other than the generative AI. The AI ​​other than the generative AI may be, for example, linear regression, logistic regression, decision tree, random forest, support vector machine (SVM), k-means clustering, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processes of each of the above-mentioned parts are performed by AI, the processes may be performed in part or entirely by AI, but are not limited to these examples. The processes performed by AI, including the generative AI, may be replaced with rule-based processes.

[0151] The data processing system 310 according to the third embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 310 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the headset type terminal 314, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the headset type terminal 314. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the headset type terminal 314 or an external device, etc., and the headset type terminal 314 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

[0152] The correspondence between each part and the device or control part is not limited to the above example, and various modifications are possible.

[0153] [Fourth embodiment] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0154] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0155] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN and / or a LAN.

[0156] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[0157] The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to instructions from the processor 46.

[0158] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS image sensor or a CCD image sensor, and captures images of the user's surroundings (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0159] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0160] The control 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 emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[0161] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0162] 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.

[0163] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0164] In the robot 414, the processor 46 performs the identification process. The storage 50 stores the identification program 60. The processor 46 reads the identification program 60 from the storage 50 and executes the read identification program 60 on the RAM 48. The identification process is realized by the processor 46 operating as the control unit 46A in accordance with the identification program 60 executed on the RAM 48. The robot 414 also has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can perform the same process as the identification processing unit 290 using these models.

[0165] Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 communicates with the server device having the data generation model 58 to obtain a processing result (such as a prediction result) using the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device (for example, a mobile phone, a robot, a home appliance, etc.) owned by a user.

[0166] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[0167] The data generation model 58 is a so-called generative AI. An example of the data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 receives a prompt containing an instruction, as well as inference data such as voice data representing speech, text data representing text, and image data representing an image. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The identification processing unit 290 performs the above-mentioned identification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation model 58 includes AI other than the generative AI. The AI ​​other than the generative AI may be, for example, linear regression, logistic regression, decision tree, random forest, support vector machine (SVM), k-means clustering, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processes of each of the above-mentioned parts are performed by AI, the processes may be performed in part or entirely by AI, but are not limited to these examples. The processes performed by AI, including the generative AI, may be replaced with rule-based processes.

[0168] The data processing system 410 according to the fourth embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 410 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the robot 414, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the robot 414. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the robot 414 or an external device, etc., and the robot 414 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

[0169] The correspondence between each part and the device or control part is not limited to the above example, and various modifications are possible.

[0170] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0171] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion encompasses both emotions and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[0172] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[0173] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[0174] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is expressed, and when they approach the ideal, a state of pleasure is expressed. Emotions can also be created for robots, cars, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is expressed, and when they approach the ideal, a state of pleasure is expressed. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on speech emotion recognition and brain physiological signal analysis systems for emotions, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the area called "reaction," where sensation is dominant. The right half of the emotion map lists emotions belonging to the area called "situation," where situational awareness is dominant.

[0175] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[0176] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[0177] In the above embodiment, an example was given in which a specific process is performed by one computer 22, but the technology disclosed herein is not limited to this, and distributed processing of the specific process may be performed by multiple computers including computer 22.

[0178] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[0179] 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.

[0180] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[0181] The hardware resource for executing a specific process can be any of the following types of processors: A processor, for example, is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. A processor also includes a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[0182] The hardware resource that executes the specific process may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific process may be a single processor.

[0183] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[0184] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[0185] In the above example, the first to fourth embodiments have been described separately, but some or all of these embodiments may be combined. The smart device 14, smart glasses 214, headset terminal 314, and robot 414 are merely examples, and they may be combined, or other devices may be used. In the above example, the first and second embodiments have been described separately, but they may be combined.

[0186] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, in order to avoid confusion and to facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[0187] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[0188] [Explanation of symbols]

[0189] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot

Claims

1. A collection department that collects word-of-mouth information; a learning unit that analyzes the word-of-mouth information collected by the collection unit and generates a model for identifying malicious word-of-mouth reviews and posted photos; a filtering unit that filters malicious reviews and posted photos using the model generated by the learning unit; a presentation unit that presents highly reliable word-of-mouth reviews filtered by the filtering unit. A system characterized by:

2. The collecting unit Collect review information from each review site and store it in a database 2. The system of claim 1.

3. The learning unit Analyze collected review information and generate a model to identify slanderous and false reviews, as well as stealth marketing.

2. The system of claim 1.

4. The filtering unit Using trained models, we automatically detect and filter defamatory content, false reviews, and stealth marketing.

2. The system of claim 1.

5. The presentation unit Reliable reviews filtered by the filtering unit are presented on each review site.

2. The system of claim 1.

6. The collecting unit Estimate user emotions and adjust the timing of collecting word-of-mouth information based on the estimated user emotions 2. The system of claim 1.

7. The collecting unit Analyze the update frequency of each review site and set the optimal collection schedule 2. The system of claim 1.

8. The collecting unit As they are collected, we assess the reliability of the reviews and apply filters to filter out unreliable information.

2. The system of claim 1.

Images