Artificial intelligence management system, method, and program

Abstract

To provide AI with versatility and flexibility to reflect user's unique hobbies and preferences, specialize in topics of interest to the user, understand various aspects of national and regional culture, language, law, geography, religion, and education, and adapt to these.SOLUTION: A system for managing artificial intelligence capable of interacting with a user using natural language processing includes: a storage unit configured to store information on the user; an input interface 52 configured to acquire information entered by the user; a context generation unit 541 configured to add user-specific conditions extracted from the user information to input information as context information and send the context information together with the input information to an AI provision server group 4; a response data control unit 542 configured to receive a response from the artificial intelligence; a filtering unit 36b configured to perform filtering based on keywords contained in the response information; and an output I / F 53 configured to output the information filtered by the filtering unit 36b as output information.SELECTED DRAWING: Figure 4

Description

Technical Field

[0001] The present invention relates to a system, method, and program for managing an artificial intelligence capable of interactive communication by a machine learning function using natural language processing (NLP) technologies such as language models and conversational AI (Artificial Intelligence).

Background Art

[0002] In recent years, the progress of artificial intelligence (AI) technology has been remarkable. As one aspect, there is conversational AI that can communicate like a human. This is a technology that utilizes a field of AI called natural language processing (NLP) to understand and respond to human language, such as in chatbots and virtual assistants, and is being applied in various services such as customer support, product recommendations, information provision, and reservation systems. For example, a system that can handle various questions has also been proposed by training a machine reading comprehension (MRC) model using reference texts generated based on natural language conversations as disclosed in Patent Document 1.

[0003] In the dialogue management system disclosed in this Patent Document 1, the system is configured by connecting a client terminal that conducts a dialogue with a user and a dialogue management device via a communication network, and is trained to generate an appropriate answer to a question using an answer label indicating the correct answer to the question and a reference text. In such a dialogue management system, in the field of conversational AI, more accurate dialogue management and question-and-answer capabilities can be realized.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in the dialogue management device disclosed in Patent Document 1 described above, since it does not have a mechanism that takes age restrictions into consideration, there is a possibility that children and minors may come into contact with adult-oriented language usage and content during the dialogue with AI, and the risk of being psychologically affected and the protection in terms of safety are also insufficient. Furthermore, a system design that does not consider age restrictions also has problems in terms of compliance with laws and regulations. In many countries, there are strict laws and regulations regarding content for children and minors, so there is a possibility of violating these standards.

[0006] In addition, when the location (country or region) where the user is located is not considered in the dialogue-type AI service, cultural, linguistic, legal, geographical, religious, and educational problems may occur. For example, since different countries and regions have their own unique cultures, religious beliefs and values differ between regions. In particular, some religions and creeds have specific taboos and doctrines, and the laws and regulations of each country include religious and educational aspects. Without understanding geographical characteristics, local conditions, expressions and nuances specific to each country and region, there is a possibility of providing incorrect information regarding religion and education or causing misunderstandings.

[0007] Therefore, the present invention aims to solve the above problems, reflects the user's unique hobbies and preferences, specializes in topics that the user is interested in, enables AI to understand and adapt to various aspects of culture, language, law, geography, religion, and education in a form that suits the user, takes the user's age into consideration, reduces the risk that the user comes into contact with inappropriate language usage and content through AI, and provides an artificial intelligence management system, management method, and management program that can ensure compliance with laws and regulations.

Means for Solving the Problems

[0008] In order to solve the above problems, the artificial intelligence management system of the present invention A system for managing an artificial intelligence capable of interacting with a user using natural language processing, a storage unit that stores information about the user as user information, an input acquisition unit that acquires information input by the user as input information, a transmission unit that adds user-specific conditions extracted from the user information as context information to the input information and transmits the input information together with the context information to the artificial intelligence, a response reception unit that receives a response from the artificial intelligence as response information, a filtering unit that performs filtering based on keywords included in the response information, and an output unit that outputs information filtered by the filtering unit as output information characterized by comprising.

[0009] Also, the present invention is a method for managing an artificial intelligence capable of interacting with a user using natural language processing, an input acquisition step in which a storage unit stores information about the user as user information and an input acquisition unit acquires information input by the user as input information, a transmission step in which a transmission unit adds user-specific conditions extracted from the user information as context information to the input information and transmits the input information together with the context information to the artificial intelligence, a response reception step in which a response reception unit receives a response from the artificial intelligence as response information, an output step in which a filtering unit performs filtering based on keywords included in the response information and an output unit outputs information filtered by the filtering unit as output information characterized by including.

[0010] In the above invention, the input acquisition unit acquires information about the user's age as part of the input information, The sending unit sends the input information together with the user-specific conditions including age information to the artificial intelligence. The filtering unit performs filtering with reference to the age information. This is preferable.

[0011] The above invention further includes a position information acquisition unit that acquires information regarding the current position. The input acquisition unit acquires the current position information acquired by the position information acquisition unit as part of the input information. The sending unit sends the input information together with the user-specific conditions including the current position information to the artificial intelligence. The output unit outputs the output information including the current position information. This is preferable.

[0012] Note that the system according to the present invention described above can be realized by executing a program of the present invention described in a predetermined language on a computer. That is, by installing the program of the present invention on an IC chip and a memory device of a general-purpose computer such as a mobile terminal device, a smartphone, a wearable terminal, a mobile PC, or other information processing terminals, a personal computer, or a server computer, and executing it on a CPU, a system having each of the above-described functions can be constructed, and the method according to the present invention can be implemented.

[0013] In addition, the program of the present invention can be distributed, for example, through a communication line, and can also be recorded on a computer-readable recording medium and transferred as a package application that operates on a stand-alone computer. Specifically, as this recording medium, in addition to magnetic recording media such as flexible disks and cassette tapes, or optical disks such as CD-ROMs and DVD-ROMs, it can be recorded on various recording media such as RAM cards. And according to the computer-readable recording medium on which this program is recorded, it becomes possible to easily implement the above-described system and method using a general-purpose computer or a dedicated computer, and it is also possible to easily store, transport, and install the program.

Advantages of the Invention

[0014] As described above, according to these inventions, regarding conversational AI and other artificial intelligence systems capable of dialogue, by incorporating a mechanism for providing information based on age restrictions and current location while reflecting the user's unique hobbies and preferences, information is provided in a form specialized for topics that the user is interested in, while enhancing the safety and appropriateness of the user, and responses can be made according to the user's current situation (country, region, time zone, legal regulations, religious and educational restrictions) and age. As a result, for example, considering regionality, religion, and education, the risk of children and minors being exposed to adult language and content is reduced, etc., the possibility of being psychologically affected is avoided, and by ensuring compliance with legal regulations, it is possible to respond to strict legal requirements in many countries.

[0015] In addition, according to the present invention, since the information provided to the user can be regulated between the user and the conversational AI, AI developers do not need to be conscious of various ratings, and the development efficiency of AI can be improved.

Brief Description of the Drawings

[0016]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Mode for Carrying Out the Invention

[0017] Hereinafter, with reference to the accompanying drawings, embodiments of the artificial intelligence management system, artificial intelligence management method, and management program according to the present invention will be described in detail. In this embodiment, a case where the method for generating environmental value digital art according to the present invention is used in the power trading service provided by the artificial intelligence management system will be exemplified. Note that the embodiments shown below are examples of devices for embodying the technical idea of the present invention, and the technical idea of the present invention does not specify the materials, shapes, structures, arrangements, etc. of each component as follows. The technical idea of the present invention can be variously modified within the scope of the claims.

[0018] (Overall Configuration of Artificial Intelligence Management System) In this embodiment, an artificial intelligence management service is provided by an artificial intelligence management system. Fig. 1 shows the overall configuration of the artificial intelligence management system 1 according to this embodiment. As shown in the figure, in this embodiment, through the artificial intelligence management system 1 constructed on the Internet 2, an artificial intelligence management service is provided to the user U to manage the information provision by various AIs.

[0019] Specifically, in this service, when providing an information provision service by an AI capable of interacting with the user using natural language processing, the content of the information provided by the AI is managed according to ratings and location information. In this embodiment, various content server groups 4 that provide information provision services by AIs are distributed on the Internet 2, and the provided information management server 3 selects an appropriate AI service server from among these content server groups 4. The user selects according to functions and applications, and for these selected artificial intelligences, while complementing the input information so that the speech conforms to the user-specific characteristics, and filters and processes the responses obtained therefrom according to the user's specific characteristics (age, gender, nationality, current location, religion, hobbies, etc.).

[0020] Specifically, the artificial intelligence management system 1 according to this embodiment is composed of a content server group 4 that provides information provision services by various AIs, a provided information management server 3 connected to these content server groups 4, and information processing terminal devices such as a smartphone 5 and a car navigation system in-vehicle device 6 that are information processing terminals used by each user U, which are interconnected through the Internet 2. In this embodiment, the smartphone 5 and the car navigation system in-vehicle device 6 are described as examples of information processing terminals.

[0021] The Internet 2 is an IP network that uses the communication protocol TCP / IP and is a distributed communication network constructed by interconnecting various communication lines (public lines such as telephone lines, ISDN lines, ADSL lines, optical lines, dedicated lines, fourth-generation (4G) communication methods such as LTE, and communication methods after the fifth generation (5G), etc., as well as wireless communication networks such as Wifi (registered trademark) and Bluetooth (registered trademark)). This IP network also includes LANs such as intranets (corporate internal networks) and home internal networks using 10BASE-T, 100BASE-TX, etc.

[0022] The provided information management server 3 is a server device that selects an appropriate AI service server from among the content server group 4 for the user to select according to functions and applications, complements the input information so that the speech is in accordance with the user's specific characteristics for the selected artificial intelligence, and performs filtering and information processing on the responses obtained therefrom according to the user's specific characteristics. In this embodiment, the provided information management server 3 has a function of once aggregating the input data such as text data and voice data input by the artificial intelligence management system 1 through the smartphone 5 and the in-vehicle device 6 of the car navigation system, and performing filtering and processing. Note that the provided information management server 3 can be realized by a single server device or a group of multiple server devices.

[0023] On the one hand, the smartphone 5 is a portable information processing terminal device that uses wireless communication. A relay point such as a wireless base station 22 and this smartphone 5 can communicate wirelessly, and communication services such as calls and data communication can be received while moving. Examples of this communication method include 3G (3rd Generation) method, LTE (Long Term Evolution) method, 4G method, FDMA method, TDMA method, CDMA method, W-CDMA, and in addition, PHS (Personal Handyphone System) method, etc. Also, this smartphone 5 is equipped with various functions such as a digital camera function, an execution function of application software, a position information acquisition function by GPS (Global Positioning System), etc., and includes mobile computers such as tablet PCs.

[0024] Note that the position information acquisition function regarding the smartphone 5 is a function of acquiring and recording position information indicating the position of its own device. As shown in FIG. 1, examples of this position information acquisition function include a method of detecting the position of its own device by a signal from a satellite 21 like GPS, and a method of detecting the position by the radio wave intensity from a wireless base station 22 of a mobile phone or an access point of Wifi communication.

[0025] And this smartphone 5 is provided with a display unit 53a such as a liquid crystal display as a display unit for displaying information, and is also provided with an operation device such as an operation button for the user to perform input operations. As this operation device, it includes a touch panel 52a as an input unit that is arranged to overlap with the liquid crystal display and acquires an operation signal by a touch operation such as specifying a coordinate position on the liquid crystal display. Specifically, this touch panel 52a is an input device that inputs an operation signal by pressure, electrostatic detection, or the like by a touch operation using the user's fingertip, pen, etc., and is configured by overlapping a liquid crystal display that displays graphics and a touch sensor that receives an operation signal corresponding to the coordinate position of the graphics displayed on this liquid crystal display.

[0026] (Internal Structure of Each Device) Next, the internal structure of each device constituting the artificial intelligence management system 1 described above will be explained. FIG. 2 is a block diagram showing the internal configuration of the provided information management server 3 according to the present embodiment, and FIG. 3 is a block diagram showing the internal configuration of the smartphone 5 according to the present embodiment. Note that the "module" used in the description refers to a functional unit configured by hardware such as a device or equipment, software having its function, or a combination thereof, and achieving a predetermined operation.

[0027] (1) Provided Information Management Server 3 First, the internal configuration of the provided information management server 3 will be explained. The provided information management server 3 is a server device arranged on the Internet 2 and can transmit and receive data to and from the content server group 4 and each smartphone 5 through the Internet 2 for an information providing service by AI.

[0028] Specifically, the provided information management server 3 includes a communication interface 31 for performing data communication through the Internet 2, an authentication unit 33 for authenticating the authority of users and user terminals, a location information management unit 32 for collecting and managing the location information of each user terminal, a provided information control unit 36 for executing control of processing for providing information in a form of dialogue (chat) with the user U, an information distribution unit 34 for distributing information to each user, and various database groups 35a to 35c.

[0029] As the database group, a filtering database 35a for storing information required for filtering, information processing, and context generation, a user database 35b for accumulating information about users, and an external information database 35c for accumulating external information such as current situation information, weather information, and economic information collected from the Internet are included. Each of these databases may be a single database, or may be divided into a plurality of databases and set as a relational database in which relations are set between the respective databases to associate the data with each other.

[0030] The filtering database 35a stores materials and templates necessary for filtering, processing, adding context, and replacing data for input data from users and response data from various AIs. Specifically, it includes lists of inappropriate words, phrases, topics, etc., content filtering information such as criteria for content considered inappropriate according to age, and alternative information that is referred to in the case of inappropriate responses by the AI and is used to generate alternative responses.

[0031] The information stored in the user database 35b includes authentication information associating identifiers (user ID, terminal ID) for identifying the user or the information processing terminal used by the user with a password, etc., and also includes personal information of the user associated with the user ID, the model of the terminal device, etc. In addition, the user database 35b stores an authentication history (access history) for each user or each user terminal.

[0032] Also, the user database 35b can store image information of official identity certificates such as driver's licenses, passports, and My Number cards for the purpose of detailed user authentication and identity verification. And the user database 35b stores image data and text data extracted from these identity certificates, including information such as the face photo, name, address, issue date, expiration date of a driver's license, information such as the face photo, name, nationality, issue date, expiration date of a passport, and information such as the face photo, name, address, issue date, expiration date, My Number of a My Number card, which are stored as user information.

[0033] Furthermore, the user database 35b stores a profile related to the user's unique characteristics for each user. This profile includes personal information such as the age, gender, language, interests, past conversation history of individual users, information used for context analysis and generation of personalized responses, as well as standard response templates such as examples of appropriate responses and examples of alternative responses for various questions and topics.

[0034] As information stored in the external information database 35c, it includes external information such as current news information, weather information, traffic information, stock price and exchange rate information, real estate price information, sports and entertainment information, etc. Specifically, the external information database 35c aggregates information for capturing the social background and environment when the right of publicity data is generated, and includes geographical information such as religions, customs, languages, etc. of each country, and information on laws, regulations, and guidelines. In particular, as information necessary for the above-described processing such as filtering and processing, it includes location information-linked information, age declaration information, and hobby-related information.

[0035] As current location-linked information, in addition to geographical information such as the country, city, and place (e.g., restaurant, museum, park) where the user is currently located, it stores weather information such as the temperature and weather conditions (sunny, rainy, snowy, etc.) at the user's current location, local event information regarding events and festivals held around the current location, and cultural and legal information such as cultural habits and legal regulations (e.g., smoking prohibited areas, holidays, etc.) in that area.

[0036] As information related to age restrictions, it includes the age group targeted by specific content or services (e.g., 12 years old and above, 18 years old and above), content whose access is restricted according to age (e.g., R-rated movies, adult games), and legal age restrictions: restrictions based on the age defined by law (e.g., legal drinking age, driving license acquisition age).

[0037] Also, as information related to hobbies, it includes individual preferences such as movies, music, books, sports (e.g., rock music, science fiction movies, soccer), favorite genres, activities and events related to hobbies (e.g., attending concerts, reading clubs, hiking). Also, as information related to an individual's interests, it includes personal concerns (e.g., cooking, traveling, technology), and consumption preferences such as preferences for specific brands and products (e.g., Apple products, organic food, luxury cars).

[0038] The authentication unit 33 is a module that establishes a communication session with each smartphone 5 and the AI information providing service through the communication interface 31 and performs authentication processing for each established communication session. As this authentication processing, authentication information is acquired from the smartphone 5 of the user who is the accessor, the user database 35b is referred to, the user and the like are identified, and their authorities are authenticated. The authentication results (user ID, authentication time, session ID, etc.) by this authentication unit 33 are sent to the provided information control unit 36 and accumulated as an authentication history in the user database 35b.

[0039] The location information management unit 32 is a module that acquires the location information acquired on the smartphone 5 side of the user and transmitted to the provided information management server 3. The location information management unit 32 associates the identifiers (user ID, terminal ID, etc.) of the user and the user terminal device identified by the authentication processing by the authentication unit 33 with their location information and accumulates it as a usage history in the user database 35b.

[0040] The provided information control unit 36 is a module that manages the input data sent from the user to the AI service and the response data from the AI to the user, and includes a user characteristic analysis unit 37, a location-linked control unit 36a, and a filtering unit 36b. The user characteristic analysis unit 37 has a rating management unit 37a and a hobby preference management unit 37b as subsystems for analyzing the attributes and preferences of the user and realizing an individualized dialogue based on these.

[0041] The user characteristic analysis unit 37 includes a rating management unit 37a and a hobby management unit 37b. The rating management unit 37a is a module that regulates the content of the provided information in real time based on various rating criteria, and uses user-specific characteristics, such as information regarding the user's age, to control access to appropriate content. In this embodiment, filtering and information processing according to the rating type (category) are performed according to the user-specific characteristics. As this rating type, in addition to general ones such as age, those based on violent, enemy, political, religious, economic, and cultural criteria are provided.

[0042] The rating management unit 37a has a function of automatically discriminating the above-described rating types. For example, according to the current time (calendar, time), it determines whether it is appropriate for the time zone, day of the week, calendar holiday, holiday, time zone, and other dates and times important for an individual, country, or region, the location, the language used, the moving method or means such as stopped, walking, running, automobile, airplane, etc. due to speed and acceleration, temperature, humidity, and altitude, extracts an appropriate rating type, and executes content regulation. In addition, as a method for discriminating these other rating types, those based on manual input by the user himself / herself such as name, gender, age, hobby, religion, and values can be mentioned.

[0043] In addition, the rating items restricted or controlled by this rating management unit 37a include · Conversation: Text, voice, image, video presented by the AI to humans · Thinking: The decision-making process by which the AI arrives at a conversation or action · Action: Operation of the system or application and so on.

[0044] The Hobby Management Department 37b collects and analyzes information on the user's interests and preferences, and integrates this into the dialogue context. The Location Linkage Control Unit 36a acquires information on the current location and sends this location information, combined with the user's input information, to the artificial intelligence. This enables dialogue and information provision based on the user's current location.

[0045] The Filtering Unit 36b is a module that performs information filtering based on keywords included in the input information by the user or the response information from the AI. The filtering here includes not only deletion of information, but also processing such as replacing with alternative information, changing the expression or nuance, etc. The Filtering Unit 36b is equipped with the following functions. · Keyword-based filtering Extract keywords from the input information and response information, and evaluate the appropriateness of the content based on these keywords. Depending on this evaluation, processing such as deletion, processing, replacement, etc. of the information is executed as filtering processing as appropriate. · Deletion and replacement of content: Delete information determined to be inappropriate or unnecessary, and in some cases, replace it with appropriate alternative information according to the user's age, preferences, and cultural background. · Modification of expression: Not only simply delete inappropriate content, but also adjust the expression and nuance, and process and output it in a more appropriate form. · Consideration of context The information is appropriately adjusted based on the context of the dialogue, such as the user's specific conditions and current location information. · Ensuring safety and compliance with regulations Filtering ensures the safety of the user and complies with local regulations. · Adaptability: Update the filtering criteria through continuous learning to adapt to the changing needs of users and social standards.

[0046] With these functions, the filtering unit 36b plays an important role in providing a useful, safe, and highly satisfactory interaction experience for the user, and also serves as an element for ensuring privacy protection and compliance with legal requirements while maintaining the quality of the user experience.

[0047] The information distribution unit 34 is a module that presents a website related to the provided information management service in response to access from the user, provides a my page unique to each user, etc., and enables the user to individually view the interaction with the AI on the user-specific page.

[0048] (2) Smartphone 5 Next, the internal configuration of the smartphone 5 will be described. As shown in FIG. 3, the smartphone 5 is schematically composed of a communication interface 51, an input interface 52, an output interface 53, an application execution unit 54, and a memory 55.

[0049] The communication interface 51 is a communication device for making calls or data communication, and has functions for non-contact communication such as wireless and contact (wired) communication by means of a cable, an adapter, etc. The input interface 52 is a device for inputting user operations such as a mouse, a keyboard, operation buttons, and a touch panel 52a. Further, the output interface 53 is a device for outputting video and sound such as a display and a speaker. In particular, this output interface 53 includes a display unit 53a such as a liquid crystal display, and this display unit is superimposed on the touch panel 52a which is an input interface.

[0050] Memory 55 is a storage device that stores an operating system (OS), firmware, programs for various applications, and other data. In this memory 55, in addition to the user ID that identifies the user, application data downloaded from the provision information management server 3 is accumulated, and various data processed by the application execution unit 54 is also accumulated. In particular, in this embodiment, the memory 55 stores the right information and the like acquired from the provision information management server 3.

[0051] The application execution unit 54 is a module that executes applications such as general OSs, game applications, and browser software, and is usually realized by a CPU or the like. In this application execution unit 54, by executing the program of the provision information management application according to the present invention, the dialogue management unit 543, the display data generation unit 546, the display control unit 545, and the position information acquisition unit 544 are virtually constructed. [[ID=]6]

[0052] The dialogue management unit 543 cooperates with the management server 3 and the AI provision server group 4 to · An input acquisition unit that acquires the information input by the user as input information · A transmission unit that adds the user-specific conditions extracted from the user information as context information to the input information and transmits the input information together with the input information to the artificial intelligence · A response reception unit that receives the response from the artificial intelligence as response information · A filtering unit that performs filtering based on the keywords included in the response information · An output unit that outputs the information filtered by the filtering unit as output information is a module that realizes the functions of

[0053] The dialogue management unit 543 refers to the data input through the input I / F 52 and the data stored in the memory 55, obtains the information input by the user as input information as an input acquisition unit, and adds the user-specific conditions extracted from the user information obtained by this input acquisition unit as context information to the input information in the context generation unit 541. At the same time, as a function of the sending unit, it sends the input information together with the AI-providing server group 4 (artificial intelligence). Then, the dialogue management unit 543 functions as a response receiving unit, and the response data control unit 542 receives the response from the AI-providing server group 4 (artificial intelligence) as response information, and while cooperating with the filtering unit 36b of the management server 3, it functions as a filtering unit that performs filtering based on the keywords included in the response information. This filtered information is output as output information through the output I / F 53 which is the output unit. Specifically, the dialogue management unit 543 has a context generation unit 541, a response data control unit 542, and a synchronization processing unit 540.

[0054] The context generation unit 541 is a module that generates context information based on the input information from the user. It includes the age information and hobby information provided by the user characteristic analysis unit 37, and the current location information provided by the location linkage control unit. The context generated here is added to the input information and sent to the server side, enabling a more personalized dialogue with the user.

[0055] The response data acquisition unit 542 receives the response information from the AI, verifies its content, and the response information verified and corrected by the filtering unit 36b is acquired before being output to the user, and necessary processing is performed so that the response generated by the AI conforms to the user's needs and context.

[0056] The synchronization processing unit 540 is a module that synchronizes the processing on the smartphone 5 side with the processing on the provided information management server 3 side. Specifically, on the provided information management server 3 side, graphic processing is performed to a certain extent, and the material data generated in this way is transmitted to the smartphone 5 side. The actual occurrence of event processing and the graphic processing therefor are executed by the GUI control unit 545a on the smartphone 5 side based on the occurrence conditions received from the provided information management server 3.

[0057] In particular, in the AI rating, regarding the input information acquired on the smartphone 5 side, the synchronization processing unit 540 refers to the user information stored on the provided information management server 3 side through this synchronization processing unit 540 and passes it to the context generation unit 541. The context generation unit 541 extracts user-specific conditions from the user information acquired through this synchronization processing unit 540, adds them as context information to the input information, and sends them together with the input information to the AI-providing server group 4 (artificial intelligence). Also, when the response data control unit 542 receives the response from the AI-providing server group 4 (artificial intelligence) as response information, based on the keywords included in the response information, it performs filtering in cooperation with the filtering unit 36b of the provided information management server 3 through the synchronization processing unit 540, and outputs the filtered information as output information.

[0058] The display data generation unit 546 is a module that generates display data for display on the display unit 53a. The display data is data generated by combining, in addition to graphic data, image data, character data, video data, audio, and other data. For example, it generates screens such as a list display of portrait rights and a detailed display of specific provided information. The display processing of the display data generated by this display data generation unit 546 is controlled by the display control unit 545.

[0059] The display control unit 545 is a module that controls specific screen displays based on the data generated by the display data generation unit 546. In this embodiment, the display control unit 545 has a GUI control unit 545a, realizes interaction with the user via the touch panel 52a and the display unit 53a, and performs appropriate screen transitions and information updates according to the user's operations.

[0060] The GUI control unit 545a is a module responsible for the display and control of the graphical user interface (GUI) on the display of the smartphone 5. This GUI control unit 545a displays various graphical elements such as buttons, menus, icons, and windows on the display so that the user can intuitively operate the application, and performs appropriate reactions and actions in response to inputs such as touch operations from the user. The GUI control unit 545a operates in cooperation with the display data generation unit 546.

[0061] Based on the data and information generated by the display data generation unit 546, the GUI control unit 545a presents it to the user in a visual and easy-to-operate form. For example, when displaying a list of portrait rights, the GUI control unit 545a controls design elements such as the design, layout, and color usage of the list. Furthermore, the GUI control unit 545a also has the role of calling appropriate application functions in response to operations from the user. For example, when the user taps the "Record new voice data" button, it receives the operation, sends a command to the dialogue management unit 543, and starts the voice data recording process.

[0062] The location information acquisition unit 544 is a module that has the function of acquiring current location information by using the built-in GPS sensor of the smartphone 5 and communication information with a wireless base station. Specifically, the location information acquisition unit 544 acquires the information by means of the Global Positioning System (GPS) that uses artificial satellites, base station positioning by triangulation based on the radio wave intensity from a base station and base station information, Wifi positioning using a database that combines the SSID (Service Set ID) of Wifi and radio wave conditions with longitude and latitude, and so on.

[0063] Also, in this embodiment, the location information acquisition unit 544 has a time information acquisition unit 544a. The location information and time information acquired by this location information acquisition unit 544 are associated as metadata, which is related data, at the time of generating the provided information, and are stored as information on when and where the voice data was acquired. With this function, it becomes possible to add a temporal and geographical context to the voice data, and to enrich the story and background behind the voice data.

[0064] (3) In-vehicle device for car navigation system Next, the internal configuration of the in-vehicle device 6 for a car navigation system will be described. As shown in FIG. 4, the in-vehicle device 6 for a car navigation system is schematically configured from a communication interface 61, an input interface 62, an output interface 63, an application execution unit 64, and a memory 65.

[0065] The communication interface 61 is a communication device for data communication as a car navigation system, and has a function of performing non-contact communication by wireless or the like, and contact (wired) communication by a cable, adapter means, or the like. The input interface 62 is a device for inputting operations of a driver such as a touch panel and buttons. The output interface 63 is a device for outputting video and sound, such as a display for displaying navigation information and a speaker for outputting a guidance voice. In particular, the output interface 63 includes a display unit 63a such as a navigation map, and this display unit is superimposed on the touch panel 62a which is an input interface.

[0066] The memory 65 is a storage device for storing the OS and applications of the navigation system, map data 65a, and other data. In this memory 65, information related to the running of the vehicle is accumulated, and various data processed by the application execution unit 64 are accumulated. In particular, in this embodiment, the memory 65 stores information for reflecting in the dialogue with the AI in conjunction with the running data of the vehicle such as the running speed, acceleration, and direction input from the running data acquisition unit 66.

[0067] The application execution unit 64 is a module for executing applications of the navigation system, and is usually realized by a CPU or the like. In this application execution unit 64, when the program of the navigation management application is executed, the dialogue management unit 643 and the display data generation unit 646 that constitute the artificial intelligence management system are virtually constructed.

[0068] The dialogue management unit 643 manages the dialogue with the AI based on the context-added input information sent from the context generation unit 641. Also, keywords included in the input information from the user and the response information from the AI are detected via the filtering unit, and appropriate filtering processing is performed. Thereby, control is performed so that an appropriate and safe response is made to the driver. The dialogue management unit 643 includes the context generation unit 641 and the response data acquisition unit 642.

[0069] The context generation unit 641 is a module that generates context information based on the input information from the driver. It includes the age information and hobby information provided by the user characteristic analysis unit 37, and the current position information provided by the position linkage control unit. The context generated here is added to the input information and sent to the in-vehicle system side, enabling more personalized navigation for the driver.

[0070] The response data acquisition unit 642 receives the response information from the AI, verifies its content, and performs necessary processing on the response information verified and corrected by the filtering unit 36b before it is output to the driver, so that the response generated by the AI conforms to the driver's needs and context.

[0071] The response data acquisition unit 642 has the function of receiving the response information from the navigation AI and verifying its content. The response including the information on the driving state of the vehicle provided by the driving data acquisition unit 66 is appropriately corrected or adjusted by the filtering unit 36b and then output to the driver.

[0072] The synchronization processing unit 640 is a module that synchronizes the processing of the in-vehicle device 6 of the car navigation system and the processing on the side of the provided information management server 3. Specifically, on the side of the provided information management server 3, some processing related to navigation is performed to a certain extent, and the generated navigation data is transmitted to the side of the in-vehicle device 6 of the car navigation system. The actual driving management and navigation processing are executed on the side of the in-vehicle device 6 of the car navigation system based on the data received from the provided information management server 3.

[0073] The display control unit 645 is a module that generates navigation data for display on the display unit 63a. The display data is generated by combining map data, driving data, destination information, traffic information, and other data. The display processing of the display data generated by this display data generation unit 646 is controlled by the display control unit 645.

[0074] The display control unit 645 is a module that controls specific screen displays based on the navigation data generated by the display data generation unit 646. In this embodiment, the display control unit 645 realizes interaction with the driver via the touch panel 62a and the display unit 63a, and performs appropriate screen transitions and information updates according to the driver's operations.

[0075] The position information acquisition unit 644 has a function of acquiring current position information by using the built-in GPS sensor of the car navigation system vehicle-mounted device 6 and communication information with a wireless base station. Specifically, the position information acquisition unit 644 acquires information through the Global Positioning System (GPS), base station positioning, Wi-Fi positioning, etc. In this embodiment, the position information acquisition unit 644 has a time information acquisition unit 644a, and the position information and time information acquired by this position information acquisition unit 644 are associated as metadata, which is related data at the time of generating navigation data, and is stored as information on when and where the navigation data was acquired. This function enables the addition of temporal and geographical contexts to the navigation data, and can provide more detailed route guidance and situation explanations to the driver.

[0076] The position information and time information acquired by the position information acquisition unit 644 are utilized as important metadata in the generation of navigation data, accurately calculating and providing the current position, predicted route, distance to the destination, and estimated time of arrival of the vehicle during driving. This information also functions as real-time driving assistance information such as changes in traffic conditions, road construction, and detour guidance. Thereby, the car navigation system vehicle-mounted device 6 provides intuitive and timely navigation information to the driver, promoting a safe and efficient driving experience.

[0077] The GUI control unit 545a is responsible for the display and control of the graphical user interface (GUI) on the display of the in-vehicle device 6 of the car navigation system. This GUI control unit 545a displays graphical elements such as various buttons, menus, icons, and windows on the display so that the driver can intuitively operate the application, and performs reactions and actions according to touch operations. For example, when selecting a specific destination, the GUI control unit 545a controls design elements such as the design, layout, and color usage of the destination list, and has the role of calling an appropriate navigation function according to the operation from the user.

[0078] As described above, the in-vehicle device 6 of the car navigation system is characterized in that each internal part operates in cooperation to provide an intuitive and detailed navigation service for the driver. Thereby, the safety of the vehicle during driving is improved, the driving efficiency is optimized, and a comfortable driving experience for the driver is realized.

[0079] (Rating Management Method for Artificial Intelligence) By operating the artificial intelligence management system 1 described above, the artificial intelligence management method according to the present invention can be implemented. In this embodiment, one of the artificial intelligence managements includes an AI rating management method. FIG. 5 shows the operation related to rating control in the artificial intelligence management system, and FIG. 6 shows the operation related to position linkage control in the artificial intelligence management system. Note that the processing procedures described below are only examples, and each process may be changed as much as possible. Also, for the processing procedures described below, steps can be omitted, replaced, and added as appropriate according to the embodiment.

[0080] (1) Rating Control Rating control will be described. As shown in FIG. 5, first, profile matching & viewing restriction category identification is performed. Here, user-specific profile information (age, gender, interests, past interaction history, etc.) is collated with the user database 35b by the user characteristic analysis unit 37. An appropriate rating type for each user, for example, a viewing restriction category (e.g., age restriction or restriction related to sensitive topics, etc.) is identified and reflected in subsequent processes. Next, filtering & sentiment analysis is performed. This process filters the data input by the user and the requests from the user, and performs sentiment analysis. Thereby, the feelings and intentions of the user are understood, and information provision according to them is made possible.

[0081] Furthermore, context information addition is performed. In this process, additional information related to the context of the request or interaction from the user is collected and analyzed by the context generation unit 541. For example, information such as the current geographical location, time zone, and recent news events is considered. Then, the input information together with the context is sent to the server side. This input information is used as basic data for the AI service side to handle. The AI generates a response based on the input information from the user and sends this to the dialogue management unit 543.

[0082] When a response returns from each AI service to the provision information management server 3, a filtering process is executed. The generated response is confirmed again through the filtering process as to whether it is appropriate based on the user's profile and viewing restriction category. At this time, if there is inappropriate content or expression, they are removed or corrected. And finally, the filtered response is generated and sent to the user. This response is customized according to the user's needs, composed of appropriate information and expression, and provides a sensitive and appropriate dialogue and information personalized for each user.

[0083] (2) Location-linked control Next, position-linked control will be described. As shown in Fig. 6, first, profile matching & viewing restriction category identification is performed. Here, profile information unique to the driver (user) (such as age, gender, interests, past interaction history, driving history, etc.) is matched against the user database 35b by the user characteristic analysis unit 37. An appropriate viewing restriction category (for example, age restriction or restriction regarding sensitive topics, etc.) for each user is identified and also matched against the current position information obtained from the position information acquisition unit 644 such as in-vehicle GPS. Through this double matching, the system grasps information linked to the driver's personal characteristics and the physical position at that time.

[0084] Next, filtering & sentiment analysis is performed. This process filters the data input by the user and the requests from the user and performs sentiment analysis. Thereby, the user's feelings and intentions are understood, enabling information provision according to that.

[0085] Furthermore, context information addition is performed. In this process, additional information related to the context of the user's requests and interactions is collected and analyzed by the context generation unit 541. For example, information such as the current geographical location, time zone, recent news events, etc. is considered. Then, the input information together with the context is sent to the server side. This input information is used as the basic data for the AI service side to handle. The AI generates a response based on the input information from the user and transmits this to the dialogue management unit 543.

[0086] When a response comes back from each AI service to the provided information management server 3, filtering processing is executed. The generated response is again confirmed for appropriateness based on the user's profile and viewing restriction classification through the filtering processing. At this time, if there is inappropriate content or expression, they are removed or corrected. And finally, the filtered response is generated and sent to the driver. This response provides real-time customized information linked to the current location according to the driver's hobbies and preferences. For example, when the driver is interested in art, a personalized navigation experience is realized, such as guiding exhibition information nearby.

[0087] Here, it has been described that the above-mentioned rating control and position-linked control operate independently of each other. However, in this embodiment, both of these controls cooperate with each other to comprehensively consider age, hobbies, and the current position in real time to identify the rating type and execute the selection, regulation, and processing of information. For example, when the user is 7 years old and uses a tourist guide AI, · User rating: G (General audiences: all ages) · AI rating: PG-13 (Parents strongly cautioned: under 13 years old) It is classified into the above rating type and restricted from having conversations about brothels, violent shops, or cultures, histories, and politics that violate public order and good customs.

[0088] As another example, when the user is 30 years old and uses a tourist guide AI, conversations about culture, history, and politics are possible regardless of whether it is related to brothels, violent shops, or not. Whether to obtain the user's consent is entrusted to the AI service provider.

[0089] The user currently engaged in the conversation is identified by the human rating recognition function of the rating management unit 37a. This human rating recognition function serves as an input interface for each AI, and the AI assesses the human rating based on personal information registered on the user's AI device (such as a smartphone), the user's own voice, image, and video captured by a camera and speaker linked to this system, and the voice, image, and video recognized by various recognition systems. The user's personal information referenced at this time includes the user's ID, physical identification such as a card, and information from manual settings (such as adult, family, or child).

[0090] (Example of change) The above-described embodiment is merely an example of the present invention, and therefore the present invention is not limited to the above-described embodiment, and various modifications can be made depending on the design, etc., as long as they do not deviate from the technical concept of the present invention.

[0091] For example, by linking with disaster information as the external information mentioned above, in the event of an emergency such as a disaster, a "Call Red signal" can be sent to the AI to switch to pause, stop, pause, standby, emergency mode, alert mode, public interest mode, alert mode, etc. Note that this signal can be: Call Yellow: Emergency standby / Temporarily deactivated ·Call Blue: Complete cancellation of normal state and state of emergency · Role of AI in Call Yellow / Red Various commands can be set to have AI speak, think, and act (behave) to ensure the social safety of citizens, such as emergency evacuation guidance, ensuring safety, and providing disaster information.

[0092] (Actions and Effects) As described above, according to the present embodiment, as shown in FIGS. 7 and 8, for a conversational AI or other artificial intelligence system capable of dialogue, by incorporating a mechanism for providing information based on age restrictions and current location while reflecting the user's unique hobbies and preferences, information is provided in a form specialized for topics that the user is interested in, while enhancing the safety and appropriateness of the user, and responses can be made according to the user's current situation (country or region, time zone, legal regulations, religious and educational restrictions) and age. Further, in the present embodiment, according to age, gender, region (country, religious circle, etc.), user-specific hobbies and preferences, etc., inappropriate or off-topic responses for the user are filtered, and furthermore, alternative responses according to the characteristics of the user are made, so that inappropriate responses can be prevented in real time while continuing natural dialogue.

[0093] In the present embodiment, as shown in FIGS. 7 and 8, the artificial intelligence rating system can change the content of dialogue and information provided by the AI according to the characteristics and situation of the user. For example, as shown in FIG. 7, for questions of the same content, in addition to age and gender, location information such as country and cultural circle is referred to, and responses considering age, gender, religion, and habits can be output, or responses can be made as if not knowing. Cases of responding as if not knowing include filtering information outside the scope of the hobbies and interests of the user being talked to, or diverting the conversation to other topics to avoid the topic.

[0094] Also, in the explanation of works in art museums and the guidance of historical cultures such as Kyoto and world heritage sites, the AI can provide information suitable for the user's age and knowledge level so that the content of the explanation differs between elementary school students and adults. Furthermore, as shown in FIG. 8, by referring to the user's age and usual action history (including browsing and viewing history of websites and videos, driving history of car navigation, biometric information history such as a pedometer (registered trademark), etc.), responses considering the user's hobbies, preferences, and biometric information can be output according to the characteristics of the family members in the vehicle, and information can be filtered or information processing such as responding as if not knowing can be performed to protect privacy according to the member composition of the passengers.

[0095] In addition, this AI rating system can automatically avoid inappropriate language and content according to the user's attributes. For example, it automatically adjusts to avoid using violent or sexual language. In particular, religious, ethnic, and political topics are sensitive issues depending on the user's attributes, so these topics are also handled carefully.

[0096] Furthermore, the response to current events and emergencies is also one of the important functions of the AI rating system. For example, during the period of national mourning when His Majesty the Emperor passes away, appropriate consideration can be given to the content spoken by the AI. Also, during a disaster, the AI can be switched to an emergency mode to provide evacuation guidance instead of normal tourist guidance. In this case, it is also possible to cooperate with the earthquake early warning, and the AI can notify the approach of the earthquake in multiple languages.

[0097] As a result, for example, by considering regionality, religion, and education, it is possible to avoid the possibility of being psychologically affected, such as reducing the risk of children and minors being exposed to adult language and content, and to ensure compliance with legal regulations, thus meeting the strict legal requirements in many countries.

[0098] It should be noted that the present invention is not limited to the above-described embodiments as they are, and at the implementation stage, the components can be modified and embodied without departing from the gist thereof. Also, various inventions can be formed by appropriately combining a plurality of components disclosed in the above embodiments. For example, some components can be deleted from all the components shown in the embodiments.

Explanation of Reference Numerals

[0099] U... User 1... Artificial Intelligence Management System 2... Internet 3... Provided Information Management Server 4... Content Server Group 5... Smartphone 6… In-vehicle device of car navigation system 21… Satellite 22… Wireless base station 31… Communication interface 32… Location information management unit 33… Authentication unit 34… Information distribution unit 35a… Filtering database 35b… User database 35c… External information database 36… Provided information control unit 36a… Location-linked control unit 36b… Filtering unit 37… User characteristic analysis unit 37a… Rating management unit 37b… Hobby management unit 51… Communication interface 52… Input interface 52a… Touch panel 53… Output interface 53a… Display unit 54… Application execution unit 55… Memory 61… Communication interface 62… Input interface 62a… Touch panel 63… Output interface 63a… Display unit 64… Application execution unit 65… Memory 65a… Map data 66… Travel data acquisition unit 540… Synchronization processing unit 541… Context generation unit 542… Response data acquisition unit 543… Dialogue management unit 544… Location information acquisition unit 544a… Time information acquisition unit 545… Display control unit 545a… GUI control unit 546… Display data generation unit 640… Synchronization processing unit 641…Context generation unit 642…Response data acquisition unit 643…Dialogue management unit 644…Location information acquisition unit 644a…Time information acquisition unit 645…Display control unit 645a…GUI control unit 646…Display data generation unit

Claims

1. A system for managing an artificial intelligence capable of interacting with a user using natural language processing, comprising: a storage unit that stores information about the user as user information; an input acquisition unit that acquires information input by the user as input information; a transmission unit that adds user-specific conditions extracted from the user information as context information to the input information and transmits the input information together with the user information to the artificial intelligence; a response reception unit that receives a response from the artificial intelligence as response information; a filtering unit that performs filtering based on keywords included in the response information; and an output unit that outputs information filtered by the filtering unit as output information. An artificial intelligence management system, characterized by comprising the above components.

2. The input acquisition unit acquires information about the user's age as part of the input information, the transmission unit includes age information in the user-specific conditions and transmits the input information together with the user information to the artificial intelligence, and the filtering unit performs filtering with reference to the age information. The artificial intelligence management system according to Claim 1, characterized by the above features.

3. The artificial intelligence management system further comprises a position information acquisition unit that acquires information about the current position, the input acquisition unit acquires the current position information acquired by the position information acquisition unit as part of the input information, the transmission unit includes the current position information in the user-specific conditions and transmits the input information together with the user information to the artificial intelligence, and the output unit includes the current position information in the output information for output. The artificial intelligence management system according to Claim 1, characterized by the above features.

4. A method for managing an artificial intelligence capable of interacting with a user using natural language processing, comprising: an input acquisition step in which a storage unit stores information about the user as user information and an input acquisition unit acquires information input by the user as input information; a transmission step in which a transmission unit adds user-specific conditions extracted from the user information as context information to the input information and transmits the input information together with the user information to the artificial intelligence; a response reception step in which a response reception unit receives a response from the artificial intelligence as response information; Based on the keywords included in the reply information, the filtering unit performs filtering, and an output step in which an output unit outputs the information filtered by the filtering unit as output information An artificial intelligence management method characterized by including the above.

5. In the input acquisition step, the input acquisition unit acquires information regarding the age of the user as part of the input information, In the sending step, the sending unit includes age information in the user-specific conditions and sends it together with the input information to the artificial intelligence, In the output step, the filtering unit performs filtering with reference to the age information The artificial intelligence management method according to claim 4, characterized in that.

6. The method further includes a position information acquisition step in which a position information acquisition unit acquires information regarding the current position, In the input acquisition step, the input acquisition unit acquires the current position information acquired by the position information acquisition unit as part of the input information, In the sending step, the sending unit includes the current position information in the user-specific conditions and sends it together with the input information to the artificial intelligence, In the output step, the output unit includes the current position information in the output information and outputs it The artificial intelligence management method according to claim 4, characterized in that.

7. A program for managing an artificial intelligence capable of interacting with a user using natural language processing, the information processing terminal device being A storage unit that stores information regarding the user as user information, An input acquisition unit that acquires information input by the user as input information, A sending unit that adds user-specific conditions extracted from the user information as context information to the input information and sends it together with the input information to the artificial intelligence, A reply receiving unit that receives a reply from the artificial intelligence as reply information, A filtering unit that performs filtering based on keywords included in the reply information An output unit that outputs the information filtered by the filtering unit as output information An artificial intelligence management program characterized by functioning as.

8. The input acquisition unit acquires information regarding the age of the user as part of the input information, The sending unit includes age information in the user-specific conditions and sends it together with the input information to the artificial intelligence, The filtering unit performs filtering with reference to the age information The artificial intelligence management program according to claim 7, characterized in that.

9. The program further causes the information processing terminal device to function as a position information acquisition unit that acquires information regarding the current position, The input acquisition unit acquires the current position information acquired by the position information acquisition unit as part of the input information, The sending unit sends the current position information together with the input information to the artificial intelligence, including the current position information in the user-specific conditions, The output unit outputs the output information including the current position information The artificial intelligence management program according to claim 7, characterized in that.

Images