Artificial intelligence assistant injection method, device, equipment and computer readable medium

Abstract

The present application relates to an injection method, device, equipment and computer-readable medium for an artificial intelligence assistant. The method includes: performing page detection through a browser plug-in included in the system to identify key feature information of the page, wherein the browser plug-in is built based on a standardized API set; obtaining context information of the user in the current business based on the key feature information; injecting the artificial intelligence assistant into the current web page based on the listening state of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information position of the current page based on an isolation container; obtaining a user access request, and responding to the user access request by querying the context information of the current business and the user's historical access data through the artificial intelligence assistant. The present application solves the technical problems of difficulty and high cost in integrating large models in enterprise-level software.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to an injection method, device, equipment and computer-readable medium for an artificial intelligence assistant. Background Art

[0002] With the rapid development of artificial intelligence technology, especially the emergence of large language models (LLM), it has shown great potential in improving work efficiency and assisting decision-making, including the GPT (Generative Pre-trained Transformer) series, AI (Artificial Intelligence) assistants, etc. Large language models are artificial intelligence models designed to understand and generate human language, usually containing tens of billions or more parameters, and trained on massive amounts of text data to obtain a deep understanding of language. However, when applying AI technology in an enterprise environment, there are usually multiple information systems from different suppliers within the enterprise, which use different technology stacks and data formats, and many enterprises are still using old information systems, which often lack open interfaces. Therefore, there are difficulties in integrating enterprise software with AI technology and poor scalability, resulting in low efficiency of enterprise systems. Secondly, fully upgrading or replacing existing systems to support AI functions usually requires a lot of time and financial investment, which is too burdensome and costly for many enterprises.

[0003] In some existing technologies, in order to solve the above problems, the focus is mainly on developing specialized AI-assisted software or developing AI plug-ins for each system. However, these methods are often targeted and specific, so they have long development cycles, high costs, and difficulty in unified management. In addition, general AI assistants on the market are difficult to deeply understand and utilize the specific business environment and data of the enterprise, and cannot achieve deep integration with the enterprise, and have poor flexibility.

[0004] It can be seen that there is currently no more effective solution to the problems of difficulty and high cost in integrating existing enterprise-level software with large model capabilities. Summary of the invention

[0005] The present application provides an injection method, device, equipment and computer-readable medium for an artificial intelligence assistant to solve the technical problems of the difficulty and high cost of integrating large models in the above-mentioned enterprise-level software.

[0006] According to one aspect of an embodiment of the present application, the present application provides an injection method for an artificial intelligence assistant, including: performing page detection through a browser plug-in included in the system to identify key feature information of the page, wherein the browser plug-in is built based on a standardized API set; obtaining context information of the user in the current business based on the key feature information; injecting the artificial intelligence assistant into the current web page based on the listening status of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information location of the current page based on an isolation container; obtaining a user access request, and responding to the user access request by querying the context information of the current business and the user's historical access data through the artificial intelligence assistant.

[0007] Optionally, before performing page detection through the browser plug-in included in the system to identify key feature information of the page, the method also includes: activating the browser plug-in when the user logs into the system and establishing a communication connection with the back-end service; obtaining user permissions and user configuration in the back-end service according to the user login information; performing page conversion based on the user permissions and the user configuration, and monitoring the system's page loading events through the browser plug-in until the page loading event is completed.

[0008] Optionally, the page detection is performed through the browser plug-in included in the system, and the key feature information of the page is identified, including: traversing the elements in the current page through the document object model traversal algorithm integrated in the browser plug-in, and classifying the elements through a selector; extracting element features from each traversed element, the element features including at least one of an element class name, an element attribute, and an element tag name; identifying the extracted element features according to predefined rules through a preset machine learning model, and extracting multiple key feature information of the current page, wherein the preset machine learning model is trained based on various types of element features.

[0009] Optionally, the method of injecting the artificial intelligence assistant into the current web page according to the listening status of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information position of the current page based on the isolation container, includes: creating a first object and a second object associated with the artificial intelligence assistant in the current web page, wherein the first object is a virtual document object model library for rendering the user interface component of the artificial intelligence assistant, and the second object is an isolation container for storing the user interface component; monitoring the change status of the document object model in the browser plug-in, selecting a target injection node according to the change status, and creating an isolation container for the target injection node; creating the user interface component of the artificial intelligence assistant, rendering the user interface component of the artificial intelligence assistant to the second object of the current web page through the first object, and storing the document object model structure of the artificial intelligence assistant in the isolation container of the target injection node; determining the interface area of ​​the main information in the current page, and isolating the interface area of ​​the artificial intelligence assistant from the interface area of ​​the main information based on a preset distance algorithm.

[0010] Optionally, determining the interface area of ​​the main information in the current page and isolating the interface area of ​​the artificial intelligence assistant from the interface area of ​​the main information based on a preset distance algorithm includes: obtaining display interface parameters of the main information in the current page, and determining the interface area of ​​the main information based on the display interface parameters of the main information; and setting the interface parameters of the artificial intelligence assistant outside the interface area of ​​the main information by a preset distance algorithm to determine the interface area of ​​the artificial intelligence assistant.

[0011] Optionally, obtaining the user access request and responding to the user access request by querying the context information of the current business and the user's historical access data through the artificial intelligence assistant includes: obtaining the user access request and sending it to the back-end engine of the artificial intelligence assistant for request parsing and extracting request information; querying the context information of the current business and the user's historical access data based on the request information to obtain target feedback information, wherein the target feedback information includes multiple types; and responding to the request based on the target feedback information.

[0012] Optionally, after injecting the artificial intelligence assistant into the current web page according to the listening status of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information location of the current page based on the isolation container, the method also includes: monitoring the target injection node and the child nodes of the target injection node; if it is monitored that there are node content changes in the target injection node and the child nodes of the target injection node, triggering the callback function to traverse and respond to all the monitored changes in the node content.

[0013] According to another aspect of an embodiment of the present application, the present application provides an injection device for an artificial intelligence assistant, including: an identification module, used to perform page detection through a browser plug-in included in the system, and identify key feature information of the page, wherein the browser plug-in is built based on a standardized API set; an acquisition module, used to obtain the user's context information in the current business based on the key feature information; an injection module, used to inject the artificial intelligence assistant into the current web page according to the listening status of the browser plug-in, and isolate the user interface component of the artificial intelligence assistant from the main information location of the current page based on an isolation container; a response module, used to obtain a user access request, and respond to the user access request by querying the context information of the current business and the user's historical access data through the artificial intelligence assistant.

[0014] According to another aspect of an embodiment of the present application, the present application provides an electronic device, including a memory, a processor, a communication interface and a communication bus, wherein the memory stores a computer program that can be run on the processor, the memory and the processor communicate through the communication bus and the communication interface, and the processor implements the steps of the injection method of the artificial intelligence assistant when executing the computer program.

[0015] According to another aspect of an embodiment of the present application, the present application provides a computer-readable medium having a non-volatile program code executable by a processor, wherein the program code enables the processor to execute the steps of the injection method of the artificial intelligence assistant.

[0016] The above technical solution provided by the embodiment of the present application has the following advantages compared with the related art:

[0017] This application can integrate the system platform based on cloud service capabilities, such as being applied to an environment where enterprise-level software integrates large model capabilities to improve work efficiency. This application provides an injection method for an artificial intelligence assistant, including: performing page detection through a browser plug-in included in the system to identify key feature information of the page, wherein the browser plug-in is built based on a standardized API set; obtaining the user's context information in the current business based on the key feature information; injecting the artificial intelligence assistant into the current web page based on the monitoring status of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information location of the current page based on an isolation container; obtaining a user access request, and responding to the user access request by querying the context information of the current business and the user's historical access data through the artificial intelligence assistant. This application has high compatibility by installing browser plug-ins in the user's browser based on a standardized API set. It uses web injection technology to seamlessly embed the artificial intelligence assistant into the system in the form of a browser plug-in, and isolates the user interface components of the artificial intelligence assistant from the main information location of the current page based on an isolation container, thereby realizing intelligent integration of heterogeneous systems. Without large-scale transformation of existing systems, artificial intelligence capabilities can be quickly introduced into daily workflows to achieve intelligent upgrades, reduce costs, and adapt to different enterprise systems and work scenarios. It is highly flexible and more conducive to improving work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0020] Figure 1 A schematic diagram of a hardware environment for an optional artificial intelligence assistant injection method provided according to an embodiment of the present application;

[0021] Figure 2 A schematic flow chart of an optional artificial intelligence assistant injection method provided according to an embodiment of the present application;

[0022] Figure 3 A schematic diagram of relationship interaction corresponding to another optional injection method of an artificial intelligence assistant provided according to an embodiment of the present application;

[0023] Figure 4A block diagram of an optional artificial intelligence assistant injection device provided according to an embodiment of the present application;

[0024] Figure 5 A schematic diagram of an optional electronic device structure provided in an embodiment of the present application. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0026] In the subsequent description, the suffixes such as "module", "component" or "unit" used to represent elements are only used to facilitate the description of the present application and have no specific meaning. Therefore, "module" and "component" can be used interchangeably.

[0027] In order to solve the problems mentioned in the background technology, according to one aspect of the embodiments of the present application, an embodiment of an injection method for an artificial intelligence assistant is provided.

[0028] Optionally, in the embodiment of the present application, the above-mentioned artificial intelligence assistant injection method can be applied to Figure 1 In the hardware environment composed of the terminal 101 and the server 103 shown in FIG. Figure 1 As shown, the server 103 is connected to the terminal 101 via a network, and can be used to provide services for the terminal or a client installed on the terminal. A database 105 can be set on the server or independently of the server to provide data storage services for the server 103. The above-mentioned network includes but is not limited to: a wide area network, a metropolitan area network or a local area network, and the terminal 101 includes but is not limited to a PC, a mobile phone, a tablet computer, etc.

[0029] The injection method of the artificial intelligence assistant in the embodiment of the present application can be executed by the server 103, or it can be executed jointly by the server 103 and the terminal 101.

[0030] like Figure 2 As shown, the method comprises the following steps:

[0031] Step S202, performing page detection through a browser plug-in included in the system to identify key feature information of the page, wherein the browser plug-in is constructed based on a standardized API set.

[0032] In this embodiment, the above-mentioned browser plug-in can allow users to customize the functions and appearance of their browsers according to their personal needs and preferences, and also has the characteristics of enhanced browser functions, high security, high ease of use, etc., such as being able to automatically fill in forms, automatically log in, etc. These functions can improve browsing efficiency, security and comfort. Among them, the user's browser plug-in can be downloaded and installed from the browser's application store or third-party websites, etc., and can realize the automatic update mechanism of the plug-in to ensure that users always use the latest version, and adopt the default automatic update mechanism of the browser plug-in to ensure that enterprises can obtain the latest AI capabilities and functional improvements in a timely manner. The browser plug-in is built based on a standardized API (Application Programming Interface) set, so that a cross-browser compatible plug-in can be developed to support mainstream browsers such as Chrome, Firefox, Edge, etc., and the browser provides a native Web Extensions API to ensure compatibility with mainstream browsers. Among them, the Web Extensions API is a cross-browser extension development standard, which is supported by mainstream browsers such as Chrome, Firefox and Edge. In this embodiment, by using a standardized API set to build a browser plug-in, the need to write different codes for different browsers can be reduced to a greater extent.

[0033] In this embodiment, the above-mentioned system includes but is not limited to various types of management systems at the enterprise level, and the management systems include customer relationship management (CRM) system, human resource management (HRM) system, financial management system, project management system, enterprise knowledge management system, etc.

[0034] Among them, the system architecture corresponding to the system may specifically include browser plug-ins, backend services and management background. The browser plug-in may include a core engine, a UI rendering module, a data processing module and an AI interaction module. The core engine is responsible for the main logic processing and control flow of the plug-in; the UI rendering module is responsible for generating and managing the user interface of the AI ​​assistant; the data processing module is responsible for processing the extraction, conversion and storage of page data; and the AI ​​interaction module is responsible for managing the communication with the backend AI service. Compared with the traditional system integration method, plug-in deployment can be promoted on a large scale in a short time, and the open plug-in architecture allows enterprises and third-party developers to continuously expand the functions of the AI ​​assistant. The above-mentioned backend services may include an API gateway, an authentication and authorization service, an AI engine and a database, wherein the API gateway is responsible for processing the communication between the plug-in and the backend service; the authentication and authorization service is responsible for managing user identity authentication and access control; the AI ​​engine is responsible for providing AI capabilities such as natural language processing and machine learning; and the database is responsible for storing user configuration, interaction history and other data. The above-mentioned management background may include plug-in configuration management, user management and data analysis, wherein the plug-in configuration management is responsible for allowing administrators to customize plug-in behavior and functions; user management is responsible for managing enterprise user accounts and permissions; and data analysis is responsible for providing statistics and analysis of plug-in usage.

[0035] In this embodiment, when a user logs in to the system to access, the browser plug-in can monitor the page and scan the page based on the monitored situation. Since the current page includes various types of information, in order to more accurately and quickly identify the current page, key feature information can be extracted from the scanned data. Among them, the key feature information includes but is not limited to key elements and structures, such as tag names, class names, attributes, etc.

[0036] Step S204: Acquire the context information of the user in the current service according to the key feature information.

[0037] In this embodiment, after extracting the key feature information, a context search can be performed on the current web page based on the key feature information to obtain context information related to the key feature information. In some examples, the context information corresponding to the current business can be extracted by the Context Extractor tool based on the page analysis results (key feature information), wherein the extracted context information includes but is not limited to the customer information that the user is viewing, the form data filled in, etc.

[0038] Step S206, injecting the artificial intelligence assistant into the current web page according to the monitoring status of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information location of the current page based on the isolation container.

[0039] In this embodiment, the browser plug-in is in a monitoring state, including monitoring the status of each node. When the IA assistant is injected into the current web page, it is injected into a certain node. When the browser plug-in monitors that a certain node state can be injected into the IA assistant, it will control the execution of the injection operation. In order to ensure that the added AI assistant does not block the main information in the current page and affect the visual effect, the user interface component (UI component) of the AI ​​assistant can be isolated from the position corresponding to the main information of the current page by constructing an isolation container, where the main information can include the page content, form, etc. being browsed in the current web page.

[0040] In some examples, the Assistant Injector tool can be used to inject the UI components of the AI ​​assistant into the appropriate location of the current web page, and the injection location takes the original page layout into consideration during the injection process to ensure that important content of the current page is not obscured.

[0041] Step S208, obtaining a user access request, and responding to the user access request by querying the context information of the current business and the user's historical access data through the artificial intelligence assistant.

[0042] In this embodiment, combined with Figure 3 As shown, Figure 3 The relationship between browser plug-ins, enterprise internal systems and AI assistants is provided. After the user accesses the web page based on the browser plug-in and injects the AI ​​assistant into the current web page to serve the user, the user can send a user access request to the system's AI assistant. When assistant A receives the user access request, it can parse the user access request and query the enterprise system based on the parsing result to extract the context information corresponding to the current business. In addition, the user's historical access data related to the current business can also be queried, and the request response can be made in combination with the context information corresponding to the current business and the user's historical access data, which can more accurately locate the user's needs.

[0043] In an embodiment of the present invention, by installing a browser plug-in in the user's browser based on a standardized API set, the compatibility is high. The artificial intelligence assistant is seamlessly embedded into the system using web page injection technology in the form of a browser plug-in, and the UI component of the AI ​​assistant is isolated from the main information position of the current page based on an isolation container, thereby realizing the intelligent integration of heterogeneous systems, and quickly introducing artificial intelligence capabilities into daily workflows without large-scale transformation of existing systems, realizing intelligent upgrades, reducing costs, and being able to adapt to different enterprise systems and work scenarios. It is more widely applicable to the daily work of employees in the enterprise, with high flexibility and more conducive to improving work efficiency.

[0044] In an optional embodiment, before step S202, the method further includes:

[0045] When the user logs into the system, the browser plug-in is activated and a communication connection is established with the backend service;

[0046] Obtaining user permissions and user configuration in the backend service according to user login information;

[0047] The page conversion is performed based on the user authority and the user configuration, and the page loading event of the system is monitored through the browser plug-in until the page loading event is completed.

[0048] In this embodiment, before providing services to users through the AI ​​assistant, page initialization is required. When the user logs in to the enterprise system, the browser plug-in will be automatically activated, and the browser plug-in will establish a secure communication connection with the backend service. Based on the communication connection, user permissions and user configurations can be obtained from the backend service. User configuration is personalized configuration. Identity authentication and page conversion are performed based on user permissions and user configurations, which not only realizes identity authentication to obtain corresponding permissions, but also provides personalized services for different users.

[0049] Furthermore, after the system is started, the page loading event can be monitored based on the browser plug-in, and page analysis can be started after the page loading is completed. The current page can be scanned through the Page Analyzer tool to identify key elements and structures.

[0050] In this embodiment, by automatically activating the browser plug-in when the user logs into the system, the browser plug-in activation time can be reduced and the response speed can be improved. The system security and personalized service functions can be improved through identity authentication and personalized configuration.

[0051] In an optional embodiment, the above step S102 specifically includes:

[0052] Traversing the elements in the current page through the document object model traversal algorithm integrated in the browser plug-in, and classifying the elements through the selector;

[0053] Extracting element features from each traversed element, wherein the element features include at least one of an element class name, an element attribute, and an element tag name;

[0054] The extracted element features are identified according to predefined rules through a preset machine learning model to extract multiple key feature information of the current page, wherein the preset machine learning model is trained based on various types of element features.

[0055] In this embodiment, the elements in the current page can be scanned and classified by using DOM (Document Object Model) traversal technology combined with XPath and CSS selectors to implement a rule engine, which contains predefined pattern matching rules for identifying common UI components such as forms, buttons, data tables, etc. Then, element features can be extracted from each traversed element, including at least one of element class name, element attribute, element tag name, etc. In order to provide recognition accuracy, a preset machine learning model can also be used to identify the extracted element features according to predefined rules, and further extract multiple key feature information of the current page. Among them, the preset machine learning model includes but is not limited to decision trees, random forests, etc., and the preset machine learning model is trained based on the HTML structure, CSS class name, attributes and other features of each type of element to ensure high accuracy when used.

[0056] In some examples, a class named PageAnalyzer can be defined to analyze and identify elements in the current web page or document, extract element features such as tag names, class names, attributes, etc., and further combine predefined rules (rules) and machine learning models (mlModel) to achieve in-depth analysis of element features, thereby obtaining the final recognition result, that is, obtaining key feature information.

[0057] In this embodiment, by utilizing DOM traversal technology and combining selectors, the elements in the current page can be identified and classified, and adding a machine learning model for element recognition can improve recognition accuracy.

[0058] In an optional embodiment, the above step S206 specifically includes:

[0059] Creating a first object and a second object associated with the artificial intelligence assistant in the current webpage, wherein the first object is a virtual document object model library for rendering a user interface component of the artificial intelligence assistant, and the second object is an isolation container for storing the user interface component;

[0060] Monitoring the change state of the document object model in the browser plug-in, selecting a target injection node according to the change state, and creating an isolation container for the target injection node;

[0061] Creating a user interface component of the artificial intelligence assistant, rendering the user interface component of the artificial intelligence assistant into the second object of the current webpage through the first object, and storing the document object model structure of the artificial intelligence assistant in the isolation container of the target injection node;

[0062] The interface area of ​​the main information in the current page is determined, and based on a preset distance algorithm, the interface area of ​​the artificial intelligence assistant is isolated from the interface area of ​​the main information.

[0063] In this embodiment, JavaScript's Mutation Observer API is used to monitor DOM changes. During the injection process of the AI ​​assistant, two objects can be instantiated in the constructor of the Assistant Injector, including the first object virtual DOM and the second object styleIsolator. The first object is a virtual DOM library for rendering UI components, and a lightweight virtual DOM library is implemented that can be used to efficiently manage and update the UI components of the AI ​​assistant. At the same time, Web Components technology is used to create an encapsulated DOM tree using Shadow DOM to form an isolation container, in which the UI components of the AI ​​assistant are isolated to avoid style conflicts and ensure that the AI ​​assistant UI components are isolated from the original style of the current page to avoid style conflicts. A CSS reset is applied inside the Shadow DOM to ensure that the AI ​​assistant UI is not affected by the global style of the page; and flexbox or grid layout is used to ensure that the AI ​​assistant UI can adapt to different page layouts.

[0064] Furthermore, during the DOM monitoring process, when the target injection node is monitored, the UI component is rendered into an isolated container through the virtual DOM library that renders the UI component, and the DOM structure is stored in the isolated container of the target injection node. The interface area of ​​the main information in the current page is further determined, and based on a preset distance algorithm, the interface area of ​​the artificial intelligence assistant is isolated from the interface area of ​​the main information, so as to isolate the AI ​​assistant UI component from the original style of the current page.

[0065] Among them, the preset distance algorithm includes the use of fixed or absolute positioning, combined with JavaScript to dynamically calculate the best position to avoid blocking important content. The injected AI assistant can achieve a foldable interface when displayed, allowing users to minimize the AI ​​assistant when needed. In addition, event delegation and prevention of event bubbling are used to ensure that the interaction of the AI ​​assistant does not affect the original page: virtual DOM and lazy loading technology are used to reduce the impact on page performance; customizable AI interaction components are developed, including dialog boxes, smart prompts, automatic filling and other functions.

[0066] In some examples, a class named AssistantInjector can be defined to inject an AI assistant (or similar UI component) into a target injection node (targetNode), while forming an isolation container to ensure that the style of the injected assistant is isolated to avoid affecting the style of the original page, and to respond to DOM changes of the target injection node or its child nodes to update the position or status of the AI ​​assistant.

[0067] In an optional embodiment, after injecting the AI ​​assistant, the method further includes:

[0068] Monitoring the target injection node and the child nodes of the target injection node;

[0069] If it is monitored that the target injection node and the child nodes of the target injection node have node content changes, the callback function is triggered to traverse and respond to all the monitored changed node contents.

[0070] In this embodiment, Mutation Observer is a powerful API that can monitor changes in the DOM tree, such as monitoring the increase or decrease of the target injection node and its child nodes or changes in attributes. When a node change is monitored, a callback function can be triggered to traverse and respond to all the monitored node contents, thereby realizing the automatic update mechanism of the plug-in. The automatic update mechanism of the plug-in ensures that enterprises can obtain the latest AI capabilities and functional improvements in a timely manner, ensuring that users always use the latest version.

[0071] In some examples, when the user needs to invoke the AI ​​assistant, multiple ways are provided to invoke the AI ​​assistant, including but not limited to registering a global shortcut key using the browser extension API, creating a custom right-click menu item using the browser extension API, or adding a button to the browser toolbar.

[0072] In this embodiment, the rendering and style isolation of the UI components of the AI ​​assistant are achieved by utilizing a lightweight virtual DOM library and Shadow DOM technology, which efficiently manages and updates the UI components of the AI ​​assistant while avoiding style conflicts with the original page.

[0073] In an optional embodiment, the above-mentioned determining the interface area of ​​the main information in the current page and isolating the interface area of ​​the artificial intelligence assistant from the interface area of ​​the main information based on a preset distance algorithm specifically includes:

[0074] Acquire display interface parameters of the main information in the current page, and determine the interface area of ​​the main information based on the display interface parameters of the main information;

[0075] The interface parameters of the artificial intelligence assistant are set outside the interface area of ​​the main information through a preset distance algorithm to determine the interface area of ​​the artificial intelligence assistant.

[0076] In this embodiment, in order to avoid blocking the main information of the current page, the display interface parameters of the main information in the current page can be obtained, and the interface area of ​​the main information can be determined based on the display interface parameters of the main information, wherein the display interface parameters include parameter information such as the length from the left / right side, width, outer border size, border radius size, etc. Then, by using fixed or absolute positioning, combined with JavaScript to dynamically calculate the optimal position, the interface parameters of the AI ​​assistant are determined to ensure that the display interface of the AI ​​assistant does not block important content in the original page. Furthermore, modern front-end frameworks such as React or Vue.js can be used to implement componentized UI development and implement a lightweight state management system for managing the internal state of the AI ​​assistant. Use CSS-in-JS technology to ensure the modularity and maintainability of styles, such as styled-components.

[0077] In this embodiment, by combining the display interface parameters of the main information for regional positioning, the interface area of ​​the main information can be accurately estimated, and the distance algorithm is used to calculate the optimal display position of the AI ​​assistant, which can avoid the AI ​​assistant's interface from obstructing the main information of the original page as much as possible.

[0078] In an optional embodiment, the above step S208 specifically includes:

[0079] Obtain user access requests and send them to the engine of the artificial intelligence assistant at the back end to parse the requests and extract request information;

[0080] Based on the request information, query the context information of the current service and the user's historical access data to obtain target feedback information, where the target feedback information includes multiple types;

[0081] A request response is made based on the target feedback information.

[0082] In this embodiment, the user's input can be sent to the backend AI assistant engine through the MultiModalInteractionHandler tool for data analysis and processing. The AI ​​assistant engine combines the data analysis results to query the current business context and the user's historical access data, and searches for data associated with the current business to generate target feedback information for feedback. Among them, the response of the AI ​​assistant can be presented to the user through CopilotUIRender, and the target feedback information includes but is not limited to text, voice or visual data. Among them, the current business context and the user's historical access data can be queried in no particular order. If the result of the former query is sufficient to meet the response requirements, the latter can be queried without continuing to reduce query time and improve response efficiency.

[0083] Furthermore, combined with Figure 3 As shown, users can have multiple rounds of conversations with the AI ​​assistant to explore or solve problems in depth. Furthermore, based on user confirmation, the AI ​​assistant can perform certain operations directly on the page, such as automatically filling in forms. Complex operations may require users to confirm twice to ensure security. The interaction records between users and AI assistants are securely synchronized to the backend for optimizing AI models and personalized services. In addition, based on user feedback and usage data, the system can continuously optimize AI models and plug-in functions, and regularly push plug-in updates to provide smarter and more personalized services.

[0084] In this embodiment, by instantly parsing user access requests and quickly generating target feedback information based on context information and user historical access data, users can be provided with more personalized and accurate responses; using the AI ​​assistant's engine to parse requests not only improves the accuracy and efficiency of request processing.

[0085] In order to better illustrate the technical solution of the present invention, several specific embodiments are provided below for illustration.

[0086] When used in a customer relationship management system, when a salesperson visits the customer details page of the CRM system, the AI ​​assistant can: automatically analyze the customer's historical interaction records and generate a customer portrait summary. Recommend the most suitable product or service based on customer characteristics and historical transaction status. Provide an intelligent email draft function to automatically generate personalized follow-up emails based on the context. Analyze the sales funnel, predict the probability of closing a deal, and provide suggestions to improve the success rate.

[0087] When used in human resource management systems, in the recruitment process, AI assistants can: automatically analyze job applicants’ resumes, extract key information and score the match with job requirements. Generate a personalized list of interview questions based on the job description. Provide real-time candidate background information and question suggestions during the interview process. Assist in generating standardized interview evaluation reports and provide data visualization analysis.

[0088] When used in financial management systems, when financial personnel use financial systems, AI assistants can: monitor key financial indicators in real time, automatically detect abnormal values ​​and provide reminders. Analyze historical financial data, generate trend reports and future forecasts. Provide intelligent data filling and formula recommendations when making financial statements. Provide cash flow optimization and investment advice based on current financial conditions.

[0089] When used in a project management system, when project managers use project management tools, AI assistants can: Provide resource allocation and time estimation suggestions for new projects based on historical project data. Intelligently identify project risk points and provide prevention and mitigation strategies. Automatically generate project progress reports and highlight delayed tasks that need attention. Analyze the workload of team members and provide suggestions for task reallocation.

[0090] When used in enterprise knowledge management systems and using the internal knowledge base of an enterprise, AI assistants can: Understand the user's query intent and provide accurate document recommendations. Automatically summarize long documents and generate key point summaries. Personalize knowledge push based on the user's role and historical queries. Assist users to quickly create and classify new knowledge documents.

[0091] Through the above-mentioned multiple embodiments, we can see the application potential of the present invention in different enterprise systems. The AI ​​assistant can not only improve daily work efficiency, but also provide intelligent support for decision-making, thereby comprehensively improving the operational efficiency of the enterprise.

[0092] According to another aspect of the embodiment of the present application, Figure 4 As shown, an injection device for an artificial intelligence assistant is provided, including:

[0093] The identification module 401 is used to detect the page through the browser plug-in included in the system, and identify the key feature information of the page, wherein the browser plug-in is constructed based on a standardized API set;

[0094] The acquisition module 403 is used to acquire the context information of the user in the current service according to the key feature information;

[0095] An injection module 405, for injecting the artificial intelligence assistant into the current web page according to the monitoring state of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information position of the current page based on the isolation container;

[0096] The response module 407 is used to obtain the user access request and respond to the user access request by querying the context information of the current business and the user's historical access data through the artificial intelligence assistant.

[0097] It should be noted that the identification module 401 in this embodiment can be used to execute step S202 in the embodiment of the present application, the acquisition module 403 in this embodiment can be used to execute step S204 in the embodiment of the present application, the injection module 405 in this embodiment can be used to execute step S206 in the embodiment of the present application, and the response module 407 in this embodiment can be used to execute step S208 in the embodiment of the present application.

[0098] It should be noted that the examples and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the contents disclosed in the above embodiments. It should be noted that the above modules as part of the device can be run in Figure 1 In the hardware environment shown, it can be implemented by software or by hardware.

[0099] Optionally, the device further comprises:

[0100] An activation module is used to activate the browser plug-in and establish a communication connection with the back-end service when the user logs in to the system; a user information acquisition module is used to obtain user permissions and user configuration in the back-end service according to the user login information; a page monitoring module is used to perform page conversion based on the user permissions and the user configuration, and monitor the system's page loading events through the browser plug-in until the page loading event is completed.

[0101] Optionally, the identification module is specifically used to: traverse the elements in the current page through the document object model traversal algorithm integrated in the browser plug-in, and classify the elements through a selector; extract element features from each traversed element, the element features including at least one of element class name, element attributes, and element tag name; identify the extracted element features according to predefined rules through a preset machine learning model, and extract multiple key feature information of the current page, wherein the preset machine learning model is trained based on various types of element features.

[0102] Optionally, the injection module is specifically used to: create a first object and a second object associated with the artificial intelligence assistant in the current web page, wherein the first object is a virtual document object model library for rendering the user interface component of the artificial intelligence assistant, and the second object is an isolation container for storing the user interface component; monitor the change state of the document object model in the browser plug-in, select a target injection node according to the change state, and create an isolation container for the target injection node; create the user interface component of the artificial intelligence assistant, render the user interface component of the artificial intelligence assistant to the second object of the current web page through the first object, and store the document object model structure of the artificial intelligence assistant in the isolation container of the target injection node; determine the interface area of ​​the main information in the current page, and isolate the interface area of ​​the artificial intelligence assistant from the interface area of ​​the main information based on a preset distance algorithm.

[0103] Optionally, the injection module is also specifically used to: obtain display interface parameters of the main information in the current page, and determine the interface area of ​​the main information based on the display interface parameters of the main information; set the interface parameters of the artificial intelligence assistant outside the interface area of ​​the main information through a preset distance algorithm to determine the interface area of ​​the artificial intelligence assistant.

[0104] Optionally, the response module is specifically used to: obtain user access requests and send them to the engine of the artificial intelligence assistant at the back end for request parsing and extracting request information; based on the request information, query the context information of the current business and the user's historical access data to obtain target feedback information, and the target feedback information includes multiple types; and respond to the request based on the target feedback information.

[0105] Optionally, the device also includes: a node status monitoring module, which is used to monitor the target injection node and the child nodes of the target injection node; a node change response module, which is used to trigger a callback function to traverse and respond to all the changed node contents monitored if it is monitored that there are node content changes in the target injection node and the child nodes of the target injection node.

[0106] According to another aspect of the embodiments of the present application, the present application provides an electronic device, such as Figure 5 As shown, it includes a memory 501, a processor 503, a communication interface 505 and a communication bus 507. The memory 501 stores a computer program that can be run on the processor 503. The memory 501 and the processor 503 communicate through the communication interface 505 and the communication bus 507. When the processor 503 executes the computer program, the steps of the above method are implemented.

[0107] The memory and processor in the above electronic device communicate via a communication bus and a communication interface. The communication bus may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus may be divided into an address bus, a data bus, a control bus, etc.

[0108] The memory may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0109] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0110] According to another aspect of the embodiments of the present application, a computer program product or a computer program is provided, the computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the steps of any of the above embodiments.

[0111] Optionally, in an embodiment of the present application, the computer-readable medium is configured to store program codes for the processor to perform the following steps:

[0112] Step S202, performing page detection through a browser plug-in included in the system to identify key feature information of the page, wherein the browser plug-in is constructed based on a standardized API set.

[0113] Step S204: Acquire the context information of the user in the current service according to the key feature information.

[0114] Step S206, injecting the artificial intelligence assistant into the current web page according to the monitoring status of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information location of the current page based on the isolation container.

[0115] Step S208, obtaining a user access request, and responding to the user access request by querying the context information of the current business and the user's historical access data through the artificial intelligence assistant.

[0116] Optionally, the specific examples in this embodiment can refer to the examples described in the above embodiments, and this embodiment will not be repeated here. And when the embodiments of this application are specifically implemented, they can refer to the above embodiments and have corresponding technical effects.

[0117] It is understood that the embodiments described herein may be implemented in hardware, software, firmware, middleware, microcode, or a combination thereof. For hardware implementation, the processing unit may be implemented in one or more application specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field programmable gate arrays (FPGA), general purpose processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions described in the present application, or a combination thereof.

[0118] For software implementation, the technology described herein can be implemented by a unit that performs the functions described herein. The software code can be stored in a memory and executed by a processor. The memory can be implemented in the processor or outside the processor.

[0119] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0120] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0121] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0122] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0123] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0124] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium, including several instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a ROM, a RAM, a magnetic disk or an optical disk. It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the term "include", "include" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements that are not explicitly listed, or also includes elements inherent to such a process, method, article or device. Without more constraints, an element defined by the phrase "comprising a..." does not exclude the existence of other identical elements in the process, method, article or apparatus comprising the element.

[0125] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

Claims

1. The method for injecting an artificial intelligence assistant is characterized in that: include: Performing page detection through a browser plug-in included in the system to identify key feature information of the page, wherein the browser plug-in is built based on a standardized API set; Acquire context information of the user in the current business according to the key feature information; Injecting the artificial intelligence assistant into the current web page according to the monitoring state of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information position of the current page based on the isolation container; Obtaining a user access request, and responding to the user access request by querying the context information of the current service and the user's historical access data through the artificial intelligence assistant; The method of injecting an artificial intelligence assistant into a current web page according to the monitoring state of the browser plug-in, and isolating the user interface component of the artificial intelligence assistant from the main information position of the current page based on an isolation container, includes: creating a first object and a second object associated with the artificial intelligence assistant in the current web page, wherein the first object is a virtual document object model library for rendering the user interface component of the artificial intelligence assistant, and the second object is an isolation container for storing the user interface component; monitoring the change state of the document object model in the browser plug-in, selecting a target injection node according to the change state, and creating an isolation container for the target injection node; creating the user interface component of the artificial intelligence assistant, rendering the user interface component of the artificial intelligence assistant into the second object of the current web page through the first object, and storing the document object model structure of the artificial intelligence assistant in the isolation container of the target injection node; determining the interface area of ​​the main information in the current page, and isolating the interface area of ​​the artificial intelligence assistant from the interface area of ​​the main information based on a preset distance algorithm; The method of determining the interface area of ​​the main information in the current page and isolating the interface area of ​​the artificial intelligence assistant from the interface area of ​​the main information based on a preset distance algorithm includes: obtaining display interface parameters of the main information in the current page, and determining the interface area of ​​the main information based on the display interface parameters of the main information; and setting the interface parameters of the artificial intelligence assistant outside the interface area of ​​the main information by a preset distance algorithm to determine the interface area of ​​the artificial intelligence assistant.

2. The method according to claim 1, characterized in that Before performing page detection through the browser plug-in included in the system to identify key feature information of the page, the method further includes: When the user logs into the system, the browser plug-in is activated and a communication connection is established with the backend service; Obtaining user permissions and user configuration in the backend service according to user login information; The page conversion is performed based on the user authority and the user configuration, and the page loading event of the system is monitored through the browser plug-in until the page loading event is completed.

3. The method according to claim 2, characterized in that The browser plug-in included in the system is used to detect the page and identify the key feature information of the page, including: Traversing the elements in the current page through the document object model traversal algorithm integrated in the browser plug-in, and classifying the elements through the selector; Extracting element features from each traversed element, wherein the element features include at least one of an element class name, an element attribute, and an element tag name; The extracted element features are identified according to predefined rules through a preset machine learning model to extract multiple key feature information of the current page, wherein the preset machine learning model is trained based on various types of element features.

4. The method according to any one of claims 1 to 3, characterized in that The obtaining of the user access request and the querying of the context information of the current service and the user's historical access data by the artificial intelligence assistant to respond to the user access request include: Obtain user access requests and send them to the engine of the artificial intelligence assistant at the back end to parse the requests and extract request information; Based on the request information, query the context information of the current service and the user's historical access data to obtain target feedback information, where the target feedback information includes multiple types; A request response is made based on the target feedback information.

5. The method according to claim 1, characterized in that After injecting the artificial intelligence assistant into the current webpage according to the monitoring state of the browser plug-in and isolating the user interface component of the artificial intelligence assistant from the main information position of the current page based on the isolation container, the method further includes: Monitoring the target injection node and the child nodes of the target injection node; If it is monitored that the target injection node and the child nodes of the target injection node have node content changes, the callback function is triggered to traverse and respond to all the monitored changed node contents.

6. An injection device for an artificial intelligence assistant, characterized in that: include: An identification module, used to detect pages through a browser plug-in included in the system, and identify key feature information of the pages, wherein the browser plug-in is constructed based on a standardized API set; An acquisition module, used to acquire context information of the user in the current business according to the key feature information; An injection module, used to inject the artificial intelligence assistant into the current web page according to the monitoring state of the browser plug-in, and isolate the user interface component of the artificial intelligence assistant from the main information position of the current page based on the isolation container; A response module, used to obtain a user access request and respond to the user access request by querying the context information of the current business and the user's historical access data through the artificial intelligence assistant; The injection module is specifically used to: create a first object and a second object associated with the artificial intelligence assistant in the current web page, wherein the first object is a virtual document object model library for rendering the user interface component of the artificial intelligence assistant, and the second object is an isolation container for storing the user interface component; monitor the change state of the document object model in the browser plug-in, select a target injection node according to the change state, and create an isolation container for the target injection node; create the user interface component of the artificial intelligence assistant, render the user interface component of the artificial intelligence assistant to the second object of the current web page through the first object, and store the document object model structure of the artificial intelligence assistant in the isolation container of the target injection node; determine the interface area of ​​the main information in the current page, and isolate the interface area of ​​the artificial intelligence assistant from the interface area of ​​the main information based on a preset distance algorithm; The injection module is also used to: obtain the display interface parameters of the main information in the current page, and determine the interface area of ​​the main information based on the display interface parameters of the main information; set the interface parameters of the artificial intelligence assistant outside the interface area of ​​the main information through a preset distance algorithm to determine the interface area of ​​the artificial intelligence assistant.

7. An electronic device, comprising a memory, a processor, a communication interface and a communication bus, wherein the memory stores a computer program that can be run on the processor, and the memory and the processor communicate through the communication bus and the communication interface, characterized in that: When the processor executes the computer program, the artificial intelligence assistant injection method described in any one of claims 1 to 5 is implemented.

8. A computer-readable medium having a non-volatile program code executable by a processor, characterized in that: The program code enables the processor to execute the artificial intelligence assistant injection method described in any one of claims 1 to 5.

Images