Natural Language Information Processing Method and Natural Language Information Processing System

Abstract

To provide a natural language information processing method and a natural language information processing system.SOLUTION: A method for processing natural language information and a system for processing natural language information are provided. The natural language information processing system includes a cloud server that executes the natural language information processing method, the method including: receiving, after an online interaction program is started, content input by a user through an interaction interface; acquiring semantic features of the content input by the user, using a smart method; acquiring user data and real-time environment information to determine content that suits the semantic features, preference of the user, and the real-time environment information; generate, after performing natural language model processing on the obtained content, interaction content to be introduced to the online interaction program; and output interaction content that suits the semantic features of the user, the preference of the user, and the real-time environment information, to the interaction interface.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a chat robot, and particularly to a natural language information processing method and a natural language information processing system capable of interacting according to a user's semantic meaning, the user's preferences, and real-time environmental information.

Background Art

[0002] Currently, artificial intelligence (AI) in various fields is developing rapidly. Among them, for example, there is a natural language chat robot that can process natural language and automatically generate content, such as the Chat Generative Pre-trained Transformer developed by OpenAI, known as ChatGPT in English. Such a natural language chat robot uses generative artificial intelligence (generative AI) technology to learn a large amount of data and then can generate new data related to the original data. Through deep learning (for example, generative adversarial networks (GAN)), a smart model is constructed.

[0003] Taking ChatGPT as an example, ChatGPT can respond to users in a natural language manner by learning a large amount of network information. However, the general content for answering users is a standard answer due to being learned, and it cannot provide a response related to the user's current state in real time. Although it is called a natural language chat robot, it lacks content related to the user and suitable for the real-time situation.

Summary of the Invention

[0004] This invention provides a natural language information processing method and a natural language information processing system, which, in a dialogue process between a chat robot and a user realized using natural language processing (NLP) and generative artificial intelligence technology, can provide dialogue content that is tailored to the individual's needs and current situation by further taking into account the user's preferences and real-time environmental information.

[0005] In a natural language information processing system implemented by a computer system, a cloud server is provided that executes natural language information processing methods using a processing circuit. The cloud server allows a user to launch an online dialogue program via a user interface, receive content input by the user through the dialogue interface, and then acquire semantic features of the input content, user data, and real-time environmental information. Based on this, it can determine content that matches the semantic features of the input content, the user's preferences obtained from the user data, and the real-time environmental information. Subsequently, natural language model processing is performed in the online dialogue program, dialogue content is generated, and finally, after being introduced into the online dialogue program, the dialogue content can be output. [Brief explanation of the drawing]

[0006] [Figure 1] This is a diagram illustrating an embodiment of a system configuration for executing a natural language information processing method. [Figure 2] This is a diagram illustrating an example of a data structure used to implement a natural language information processing method. [Figure 3] This is a flowchart illustrating an example of a natural language information processing method. [Figure 4] This is a flowchart of another embodiment of a natural language information processing method. [Figure 5] This figure shows an example of a graphical user interface provided by a natural language processing system. [Figure 6] This figure shows an example of a graphical user interface provided by a natural language processing system. [Figure 7] This figure shows an example of a graphical user interface provided by a natural language processing system. [Figure 8] This figure shows an example of a graphical user interface provided by a natural language processing system. [Figure 9] This figure shows an example of a graphical user interface provided by a natural language processing system. [Modes for carrying out the invention]

[0007] This specification discloses a natural language information processing method and a natural language information processing system. The natural language information processing method is executed on a cloud server, and the natural language information processing system is realized. The cloud server provides social media services via a network, allowing users to share text content, image content, and video content after joining. The cloud server further provides chat robots in various fields, allowing users to interact through the services provided by the cloud server. The use of artificial intelligence technology includes chat robots that learn data in various fields using machine learning algorithms and natural language processing technology to provide conversational services. By learning the user's activities on social media and obtaining the user's preferences, the chat robots can provide conversational content that better suits the individual user's needs and current environmental characteristics by taking into account the meaning of the user's conversations, the user's preferences, and real-time environmental information.

[0008] The system implemented by the above-mentioned cloud server can be seen in the example diagram in Figure 1, which shows a system configuration that executes the natural language information processing method.

[0009] The cloud server 100 shown in the figure is realized by a computer system, database, and network, and has various functional modules realized through the cooperation of software and hardware wafers, as shown in the figure, including a natural language processing module 101, a machine learning module 103, an external system interface module 105, and a user interface module 107. The natural language processing module 101 is a module for executing natural language information processing methods. The natural language processing module 101 realizes a chat robot with natural language information processing capabilities. The machine learning module 103 executes machine learning algorithms and, in addition to learning natural language models, can learn the user's activities on the network using deep learning methods to obtain information preferred by the user and provide the chat robot with dialogue content that suits the user's preferences. The external system interface module 105 provided by the cloud server 100 executes connections with external systems (e.g., a first external system 111, a second external system 112) (e.g., via network 10), circuits for obtaining data via an application program interface (API), and related usage software. The user interface module 107 provided by the cloud server 100 allows the user device 150 to connect to the cloud server 100 via its network connection function and run a server that provides network services. The user device 150 then runs an application program that retrieves the corresponding services and obtains the services provided by the cloud server 100.

[0010] According to the configuration shown in the diagram, the cloud server 100 has an internal or externally connected database and provides data services through the cloud server 100. The video database 110 shown in the diagram provides video content that each end user uploads and shares, which is stored via the network 10 by the user device 150. The content uploaded and shared by users may also include text content and image content. The user database 120 stores user data, including personal user data and uploaded text content, image content, and video content. It also acquires user activity data from the network services provided by the cloud server 100, such as content viewed, following, likes, shares, and subscriptions, thereby forming a user profile. Furthermore, as conversational content is continuously generated over time, the user database 120 can store and update user data, including a record of the user's history of conversations, in chronological order so that it becomes a conversation record provided to a natural language model for training a machine learning algorithm. The vector database 130 is a database for storing structured information obtained by performing calculations using vector algorithms on various text content, image content, and video content. This structured information is used to compare various data that are tailored to the user's personalization.

[0011] According to the schematic system configuration diagram shown in the figure, the cloud server 100 acquires data from external systems, specifically the first external system 111 and the second external system 112, via network 10 or a connection using a specific protocol. The first external system 111 and the second external system 112 are, for example, servers set up by a government or company to provide open data. The cloud server 100 can acquire real-time information tailored to its needs, such as real-time weather, real-time road conditions, real-time news, and real-time location and related network information, through the application program interfaces provided by each external system via the external system interface module 105.

[0012] The user device 150 executes an application program that can acquire services provided by the cloud server 100, such as a social media service provided by the cloud server 100. The user device 150 then executes a corresponding social media application program and acquires the social media service via the user interface module 107. In particular, the cloud server 100 provides a natural language chat robot via the natural language processing module 101, and the user can interact with the chat robot via the dialogue interface 115. Meanwhile, the cloud server 100 learns activity data from users using various services of the cloud server 100 via the machine learning module 103, and the machine learning module 103 learns the user's interest characteristics and builds user data. This activity data includes activity data acquired by the user using the social media application program and the dialogue interface 115.

[0013] It should be noted that the various text content, image content, and video content acquired by the cloud server 100 are unstructured data, and may be converted into vectorized data using an encoder method to facilitate the acquisition of meaning from the various text content, image content, and video content, making data retrieval easier. Furthermore, the vectorized data is used for comparison in user keyword searches, and a distance function is used to calculate the distance between the search keyword and the vectorized data in the database. A closer distance indicates greater data similarity. This allows the user to search for data using the vector database 130.

[0014] According to the embodiment, the vector database 130 of the cloud server 100 can support a multi-mode search service for text content, image content, etc. The vector database 130 provides structured data; for example, after converting various text content, image content, and video content into text, vectorized data is obtained by performing calculations using a vector algorithm. This vectorized data can be applied to the search service and can also be used in a natural language processing program. The natural language processing program uses a natural language model to map the vectorized data to a vector space. Taking a word entered by the user as an example, a word vector is obtained by performing calculations using a vector algorithm.

[0015] In the embodiment, the natural language information processing method provided herein is actually a chat robot executed on a cloud server 100. The chat robot can interact with the user in natural language, including text and voice, and in addition to responding to information input by the user, it may also acquire user data in advance from the cloud server 100 before the interaction and obtain the user's personality and habits from that user data, or it may obtain real-time status from external systems (first external system 111, second external system 112), for example, it may obtain the weather and news for the user's location. As a result, the content that the chat robot responds to can reflect not only the user's preferences but also the real-world situation.

[0016] As shown in Figure 2, which illustrates an example of a data structure for implementing a natural language information processing method, the data is divided into social media platform data 21, user data (e.g., user description files) 23, and user activity data (user profile data) 25.

[0017] Social media platform data 21 is data that is not made public within the system. The system, which executes a natural language information processing method, acquires viewer data 211 of users accessing various content provided by the cloud server 100, and creator data 212 of users who provide various content to the system. Social media platform data 21 further includes business data 213 for the system to provide companies with built company data for advertising, and various location data 214 related to acquiring geographical locations so that the system can provide appropriate services.

[0018] User data 23 is data that can be made public in the system and includes data edited by the user himself / herself. The data edited by the user himself / herself includes viewer data 231 obtained from various user activity data. The viewer data 231 may include the user's preference data as a viewer obtained through machine learning, such as recent preference data, past preference data, and preference data regarding location.

[0019] The creator data 232 in the user data 23 is the relevant data of the user as a creator and includes the type preferred by the creator and the location-related data of the creator obtained by the system through machine learning. The creator data 232 may include, for example, the data of the user as a creator, and the type preferred by the learned creator and the location including the geographical location or a specific location of one place.

[0020] The business data 233 in the user data 23 includes the business type of the enterprise and product features obtained by the system through machine learning when the user is an enterprise.

[0021] The user activity data 25 is data that is not made public in the system and includes statistical data on the activities of the user in various services provided by the cloud server 100 and data obtained through machine learning. The user activity data 25 mainly includes viewer data 251, creator data 252, and business data 253.

[0022] The viewer data 251 is the viewing rate, viewing time, and activity data such as following, preference, comment, and reserved subscription of the services provided by the user on the cloud server 100. The creator data 252 is the statistical data of the user as a creator, for example, the followers of the channel or account, the number of views of the created content, and the account viewing rate, etc. The business data 253 is the followers, the number of content views, and the overall impression data, etc. obtained when the user is an enterprise.

[0023] The social media platform data 21, user data 23, and user activity data 25 collected and learned by the cloud server 100 form the basis for the conversational service realized using natural language processing and generative artificial intelligence as described herein. By performing calculations on the above various data using processing circuits in the cloud server 100, it becomes possible to provide a chat robot that is tailored to the user's personal and real-time needs.

[0024] As an example, in the cloud server 100, the natural language model executed may perform a vector algorithm on the content entered by the user through the dialogue interface, the user's preferences, and real-time environmental information to mark the acquired text, calculate the vector of each word, and then, after searching the database based on the vector distance between words, acquire relevant content to generate dialogue content that matches the user's preferences and real-time environmental information. In the online dialogue process, a program may be executed to perform machine translation, document summarization, document generation, etc., on the data that has been converted into text using a transformation model. Referring to the flow description of the embodiment shown in Figure 4, the chat robot may generate dialogue content by obtaining the meaning of words during the user's dialogue.

[0025] In the flow shown in Figure 3, the user first connects to the cloud server 100 using an application program running on the user's device and uses the provided services to view its content. The graphical user interface for manipulating and viewing the text content, image content, and video content provided by the cloud server 100 to the user (for example, by the user interface module 107 shown in Figure 1) is a map interface 50 with an electronic map as the background, which is launched after the application program is executed by the user's device, as shown in Figure 5. Figure 5 schematically shows how the link icons associated with each video content are displayed on the map interface 50 according to the geographical information associated with it. The link points associated with multiple geographical locations as shown in the figure may include one or more video link points, and in the figure, video link points 501, 502, and 503 are used as examples.

[0026] According to one embodiment, as shown in Figure 5, a user may, while viewing the content of the map interface 50, launch an online dialogue program by touching or using a specific hand gesture to click the dialogue link icon for dialogue 512, or by clicking a link point in the map interface 50 that provides dialogue, using the link icons for play 511, dialogue 512, help 513, search 514, and user main page 515 shown below (step S301).

[0027] On the other hand, Figure 6 provides a schematic diagram of another link method for launching an online interactive program. For example, after a user selects one of the video link points (video link points 501, 502, or 503) on the map interface 50, the screen launches a video playback page 60 that plays a video shared or created by a user, as shown in Figure 6. In Figure 6, the geographical location 601 of the video is further displayed, and several link icons such as favorites 603, dialogue 604, collect 605, and share 606 are displayed in the sidebar. The user may also launch the online interactive program by clicking the dialogue link icon for dialogue 604 (step S301).

[0028] Next, the conversational interface is activated, and the user inputs text content, image content, or specific video content through the conversational interface (for example, inputting a link to share video content), and the cloud server 100 may receive the content input by the user via the user interface module 107 (step S303). According to the embodiment, the online conversational program is actually a chat robot that uses a natural language model, and may interact with the user through the conversational interface and execute a natural language information processing method for the content that the user inputs each time. For example, one can refer to the conversational interface 70 shown in Figure 7, the conversational interface 80 shown in Figure 8, and the conversational interface 90 shown in Figure 9, etc., and the conversational interface shown in each example provides an input field for the user to input content and a conversational display area that displays the conversational content output by the chat robot and the content input by the user.

[0029] At this time, the cloud server 100 acquires the content entered by the user via the user interface module 107. The content received through the dialogue interface may be text content, audio content, or video content. If the received content is audio content or video content, the audio content or video content may be converted into text content using a text conversion program, and then semantic analysis may be performed to acquire semantic features (step S305). While the above programs are running, the cloud server 100 acquires user data from the user database 120 and acquires real-time environment information from an external system (via the external system interface module 105 shown in Figure 1) (step S307).

[0030] Subsequently, the system may determine (or screen after investigating the database) content that matches the semantic features of the content entered by the user, the user's preferences obtained from user data, and real-time environmental information (step S309). After executing natural language model processing in the online dialogue program, dialogue content is generated (step S311). The dialogue content is then introduced into the online dialogue program and output to the dialogue interface (step S313).

[0031] Furthermore, when the natural language model of the cloud server 100 is running, multidimensional information is recorded using a database or system memory, and this multidimensional information may include the history of conversations in the same online dialogue program. As a result, before the chat robot generates a dialogue, in addition to considering the semantic features of the user's conversation, the user's preferences, and real-time environmental information as in step S309, the chat robot further considers the history of conversations in the user's current online dialogue program (step S315), so that the dialogue content generated by the natural language model (step S311) has dialogue content that is appropriate to the current situation.

[0032] For example, since the user's current emotions and needs are often reflected in the history of their conversations in the same online dialogue program, as shown in the embodiment in Figure 1, the natural language processing module 101 in the cloud server 100 can use a natural language model to simultaneously consider the user's word meaning, user preferences, real-time environmental information, and history of conversations, and can continue the same conversation situation when generating dialogue content. For example, when it is possible to generate dialogue using natural language on the same dialogue theme, the terminology may have the same tone as the previous dialogue content (reflecting the user's emotions such as joy, anger, sadness, etc.), and the chat robot may learn the user's way of expressing emotions from the history of conversations.

[0033] For a related diagram, see Figure 7. As shown in Figure 7, the dialogue interface 70 has dialogue content 701, 702, and 703 between the user and the chat robot. The chat robot further searches a database based on user semantic features obtained from dialogue content 702 to provide recommended video content 704, and below the dialogue interface 70 is an input field 705 where the user can further input dialogue content.

[0034] Another mode is the dialogue interface 80 shown in Figure 8. As shown in this example, when an online dialogue program is launched, the system may directly provide natural language dialogue content 801, 802, and 804 based on the user's preferences and real-time information, and directly provide recommended video content 803, and the user may then respond to the dialogue content using the input field 805 in the dialogue interface 80.

[0035] In another embodiment, the dialogue content generated by the natural language model executed in the chat robot, based on the semantic features of the user-input content, user preferences, and real-time environmental information, may further include providing multiple recommendation options, multiple recommended video content, and / or multiple recommended friend links, as can be seen in the example in Figure 9.

[0036] In the online dialogue program, the dialogue interface 90 shown in Figure 9 includes dialogue content 901 generated by the chat robot based on the user's semantic features obtained. In this example, the semantics allow the chat robot to determine that the user has selected a specific item, and therefore it provides multiple recommendation options 902. In particular, the chat robot provides recommendation options to the user based on real-time environmental information obtained by the system from an external system. For example, the chat robot may provide recommendation options 902 based on real-time weather, road conditions, time, and the user's location. If it is mealtime, the system may provide meal options from restaurants that are open near the user's location, taking into account the user's eating habits.

[0037] Furthermore, if the user expresses a desire to watch video content, the recommendation option 902 above may include multiple recommended video content. If the user expresses a desire to find friends with similar interests, the recommendation option 902 above may include multiple recommended friend links.

[0038] Furthermore, when the user then uses the input field 906 to input dialogue content 903 in response to several recommendation options 902, the chat robot will respond with dialogue content 904 based on the meaning of the dialogue content 903, and then provide several more recommendation contents 905 based on the meaning of the above dialogue contents. Following the above example, when the user responds that they would like one of the meals, the chat robot will provide restaurant options that correspond to the meal the user wants to eat, based on real-time weather, road conditions, and the user's location obtained by the system from an external system, and if the weather is bad or the roads are congested, it will recommend restaurant options that are easier for the user to get to.

[0039] As yet another embodiment, you may refer to the flowchart of another embodiment of the natural language information processing method shown in Figure 4.

[0040] In the flow shown in Figure 4, the user launches an online dialogue program via an application program (step S401), interacts with a chat robot, and the system receives the dialogue content entered by the user (step S403) to further acquire the user's semantic features. As an example, semantic features may be obtained by performing conversion operations and vector algorithm operations using a natural language processing module on the cloud server 100 (step S405).

[0041] As an example of the natural language information processing method provided by the present invention, natural language is learned using artificial intelligence technology, and after understanding the natural language, text classification and grammatical analysis are performed. When processing dialogue content input by the user, a transformation model (Google in 2017) is used. TM The deep learning method (submitted by the brain team) may be used to process user input with chronologically ordered natural language content, and if the input content is not text content, it needs to be converted to text before obtaining the text. Thus, in online dialogue programs, this conversion model may be used to perform machine translation, document summarization, document generation, etc.

[0042] After acquiring the semantic features of the user dialogue content, the system analyzes the user's preferences and current location, or the locations the user is interested in based on the dialogue content, and acquires real-time environmental information from an external system in real time based on that location (step S407). The real-time environmental information may include information acquired in real time from one or more external systems, such as real-time weather, real-time road conditions, real-time news, and real-time location-related network information (POIs (points of interface) on a map, POI evaluations, etc.), or any combination thereof.

[0043] Subsequently, the system uses a vector database to calculate the closest answer based on user semantic features, user preferences, and real-time environmental information, or by adding historical dialogue records (step S409). It should be noted that the data in the vector database is structured information obtained using a vector algorithm, and the system may also obtain words with relatively similar semantics from content obtained by vector distance. For example, the vector distance between the word "computer" in the dialogue content and the word "calculate" in the database is relatively close, while the vector distance between the word "computer" and the word "run" is relatively far.

[0044] In this embodiment, dialogue content that matches the user's preferences and real-time environment information may be generated by adding a history dialogue record to the content input by the user, the content the user is interested in, and real-time environment information according to the needs, executing a vector algorithm, marking the acquired text, calculating the vector of each word, and acquiring related content based on the vector distance between words. Furthermore, in this embodiment, executing a vector algorithm on the history dialogue record described in the cloud server 100 may generate dialogue content that matches the user's current mood, for example, by continuing the same topic in the history dialogue record or using terminology that corresponds to the mood obtained through analysis.

[0045] Furthermore, the system further searches the video database 110 based on the above information to obtain suitable video content (step S411), the chat robot generates dialogue content using natural language processing and generative artificial intelligence technology (step S413), and outputs the dialogue content to the dialogue interface (step S415). In addition, in the embodiment, the system may continue the above steps in the chat process, and the chat robot may interact with the user using natural language (text or voice) and provide real-time content (video content, text content) that the user is interested in. [Explanation of Symbols]

[0046] 10: Network 100: Cloud Server 101: Natural Language Processing Module 103: Machine Learning Module 105: External System Interface Module 107: User Interface Module 110: Video Database 111: First External System 112: Second External System 115: Interactive Interface 120: User Database 130: Vector Database 150: User device 21: Social media platform data 211: Viewer Data 212: Creator Data 213: Business Data 214: Location data 23: User Data 25: User activity data 231, 251: Viewer data 232, 252: Creator Data 233, 253: Business Data 50: Map Interface 501, 502, 503: Video link points 511: Playback 512, 604: Dialogue 513: Help 514: Search 515: User Main Page 60: Video Playback Page 601: Geographic location 603: preference 605: Collection 606: Share 70, 80, 90: Interactive Interfaces 701, 702, 703, 801, 802, 804, 901, 903, 904: Dialogue content 704, 803: Recommended video content 705, 805, 906: Input fields 902: Recommendation Options 905: Recommended Content S301-S315, S401-S415: Natural Language Information Processing Flow

Claims

1. A natural language information processing method executed on a cloud server, This involves launching an online dialogue program and receiving content entered by the user through the dialogue interface, The process involves obtaining semantic features of the content entered by the user, To acquire user data and real-time environmental information, The process involves determining content that matches the semantic features of the content input by the user, the user's preferences obtained from the user data, and the real-time environmental information, then performing natural language model processing in the online dialogue program, and subsequently generating dialogue content. This includes introducing the aforementioned dialogue content into the online dialogue program and outputting it to the dialogue interface, The content received through the aforementioned dialogue interface is text content, audio content, or video content. If the received content is audio content or video content, a text conversion program converts the audio content or video content into text, and then performs semantic analysis to obtain semantic features of the text. The natural language model executed on the aforementioned cloud server uses a transformation model to execute programs for machine translation, document summarization, and document generation to generate the dialogue content. A natural language processing method that generates conversational content that matches the user's preferences and real-time environment information by executing a vector algorithm on the content entered by the user, the user's preferences and the real-time environment information on the cloud server, marking the acquired text, calculating the vector of each word, and acquiring related content based on the vector distance between words.

2. Furthermore, the natural language information processing method according to claim 1, wherein the vector algorithm is executed on the historical dialogue record described in the cloud server to generate the dialogue content.

3. The natural language information processing method according to claim 1, wherein the real-time environmental information includes information obtained in real time from one or more external systems, one or any combination of real-time weather, real-time road conditions, real-time news, and real-time location-related network information.

4. The natural language information processing method according to claim 1, wherein the dialogue content includes text content generated after performing the natural language model processing and video content obtained by examining a video database.

5. The online dialogue program operates a chat robot using the natural language model, interacts with the user through the dialogue interface, and executes the natural language information processing method for the content that the user inputs each time. A natural language information processing method according to any one of claims 1 to 4, wherein the dialogue content generated by the natural language model executed in the chat robot based on the semantic features of the content input by the user, the user's preferences, and the real-time environmental information includes providing a plurality of recommendation options, a plurality of recommendation video contents, and / or a plurality of recommendation friend links.

6. A natural language information processing system implemented by a computer system, Includes a cloud server that executes a natural language information processing method using a processing circuit, The aforementioned natural language information processing method is This involves launching an online dialogue program and receiving content entered by the user through the dialogue interface, The process involves obtaining semantic features of the content entered by the user, To acquire user data and real-time environmental information, The process involves determining content that matches the semantic features of the content input by the user, the user's preferences obtained from the user data, and the real-time environmental information, then performing natural language model processing in the online dialogue program, and subsequently generating dialogue content. This includes introducing the dialogue content into the online dialogue program and outputting the dialogue content to the dialogue interface, The content received through the aforementioned dialogue interface is text content, audio content, or video content. If the received content is audio content or video content, a text conversion program converts the audio content or video content into text, and then performs semantic analysis to obtain semantic features of the text. The natural language model executed on the aforementioned cloud server uses a transformation model to execute programs for machine translation, document summarization, and document generation to generate the dialogue content. A natural language processing system that generates conversational content that matches the user's preferences and real-time environment information by executing a vector algorithm on the content entered by the user, the user's preferences and the real-time environment information on the cloud server, marking the acquired text, calculating the vector of each word, and acquiring related content based on the vector distance between words.

7. The natural language information processing system according to claim 6, wherein the cloud server provides an external system interface for connecting to one or more external systems to acquire real-time environmental information including one or any combination of real-time weather, real-time road conditions, real-time news, and real-time location-related network information.

Images