An intelligent interactive system and method

Abstract

The application relates to an intelligent interaction system and method, and belongs to the technical field of intelligent control. The system comprises a communication connection module, which is arranged on an intranet server and is used for establishing a connection between the intranet server and an external communication server, receiving external user messages and event callbacks, an intention recognition module, which is used for recognizing a user interaction intention corresponding to the external user messages and event callbacks based on a large language model and a preset intention configuration file, a business processing module, which comprises a plurality of business processing units and is used for calling a business processing unit corresponding to the user interaction intention to execute an intranet business operation and generating a response message, and a message pushing module, which is used for pushing the response message to a target user or group through the communication connection module. The application can meet the demand of enterprises for automatic processing of intranet businesses, and improve the intranet business processing efficiency and user interaction convenience.

Description

Technical Field

[0001] This application relates to the field of intelligent control technology, specifically to an intelligent interactive system and method. Background Technology

[0002] In enterprise intranet business management scenarios, related technologies generally rely on manual login to the intranet system to perform operations such as equipment maintenance, software upgrades, and project management. There is a lack of convenient interaction channels between users and intranet business systems. External users cannot directly initiate business commands through daily communication tools and must rely on offline processes or dedicated maintenance platforms to submit requests. This not only results in high interaction barriers and delayed responses but also makes it easy for business execution errors to occur due to information transmission deviations, failing to meet the enterprise's needs for automated processing of intranet business. Summary of the Invention

[0003] This invention provides an intelligent interaction system and method, which aims to solve the problem that existing technologies cannot meet the needs of enterprises for automated processing of intranet business.

[0004] Firstly, an intelligent interaction system is provided, comprising: The communication connection module, deployed on the intranet server, is used to establish a connection between the intranet server and an external communication server, and to receive external user messages and event callbacks. The intent recognition module is used to identify the user interaction intent corresponding to the external user message and the event callback based on the large language model and the preset intent configuration file. The business processing module includes multiple business processing units, which are used to call the business processing unit corresponding to the user interaction intent to perform intranet business operations and generate response messages; The message push module is used to push the response message to the target user or group through the communication connection module.

[0005] Optionally, the communication connection module further includes a message preprocessing unit, used to parse, filter and standardize received external user messages and event callbacks, and to perform cache scheduling through a message queue; The event callbacks include event notifications for user sending messages, user joining or leaving groups, and user triggering interactive commands, which are used to trigger the intent recognition module to start the intent recognition process.

[0006] Optionally, the communication connection module is further configured to complete the connection subscription and identity authentication between the communication connection module and the external communication server through a preset identity identifier and key, and establish a communication link with the external communication server based on a preset communication protocol; The system periodically sends heartbeat data packets to keep the communication link active and monitors the connection status of the communication link.

[0007] Optionally, the intent configuration file is in a structured format, including intent identifier, corresponding processing unit name, storage path, and intent description.

[0008] Optionally, the business processing module includes: A general consultation unit is used to respond to consultation intent and call the large language model to generate a natural language response. The device information query unit is used to respond to the intent of device status query, call the browser automation tool to log in to the intranet device management platform, query and return the device software and hardware version, model and operation record according to the device identifier; The software push unit is used to respond to the software deployment intent, extract key parameters from the external user message and verify them, and call the browser automation tool to log in to the intranet software management platform to perform remote deployment operations; The project management query unit is used to respond to project progress query intents. It calls browser automation tools to log in to the intranet project management platform, obtains project nodes, progress, issue information and personnel information, and generates response content by summarizing it through a large language model. The browser automation tool is used to simulate human browser operation to achieve login, data query, data acquisition and command execution for intranet business systems that do not have open application programming interfaces.

[0009] Optionally, the key parameters extracted by the software push unit include device model, software version, batch information, and control unit identifier; If the key parameters are missing, a supplementary prompt will be sent to the user through the communication connection module.

[0010] Optionally, the project management query unit categorizes project issues by severity level, including serious faults, high-perceived issues, and issues that must be resolved, and filters out long-unresolved issues; personnel information includes work content, progress, and task records.

[0011] Optionally, the intelligent interaction system further includes: The scheduled push module is used to automatically execute preset tasks and generate notification messages according to preset time or business trigger conditions; The message push module is also used to push the notification message to the target user or group through the communication connection module.

[0012] Optionally, the business triggering condition includes a software deployment completion event; Upon receiving the software deployment completion event, the scheduled push module queries the deployment execution result and generates a return visit notification message.

[0013] Optionally, the business processing module further includes an exception handling unit, which generates an exception prompt message and pushes it to the user when the business operation fails, parameters are missing, or intent recognition is uncertain.

[0014] Secondly, an intelligent interaction method is also provided, including the following steps: It responds to the establishment of a connection between the internal network server and the external communication server, and receives external user messages and event callbacks; Based on the large language model and the preset intent configuration file, the user interaction intent corresponding to the external user message and the event callback is identified; The business processing unit corresponding to the user interaction intent is invoked to perform intranet business operations and generate a response message; The response message is pushed to the target user or group.

[0015] Thirdly, an electronic device is also provided, including a memory and a processor. In a specific example, a computer program is stored on the memory, and when the computer program is executed by the processor, it implements the method of any of the above embodiments.

[0016] Fourthly, a computer-readable storage medium is also provided, on which a computer program is stored, the computer program being loaded by a processor to perform the steps of the method in any of the above embodiments.

[0017] Beneficial effects: This application deploys a communication connection module, an intent recognition module, a business processing module, and a message push module on an intranet server, forming a complete interactive closed loop. The communication connection module ensures secure data exchange between the intranet and extranet, and stable access to various interactive data. Combined with a large language model and preset intent configuration files, it achieves accurate intent recognition of user messages and interactive events. The business processing module, through the collaborative efforts of multiple business units and the use of browser automation tools, automates the operation of intranet business systems without open interfaces. Finally, the message push module uniformly pushes and provides feedback on the business results. This approach meets the enterprise's needs for automated intranet business processing, improving intranet business processing efficiency and user interaction convenience. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the functional modules of an intelligent interactive system provided by an exemplary embodiment of this disclosure; Figure 2 This is a schematic diagram of an intelligent interaction method provided by an exemplary embodiment of this disclosure; Figure 3 This is a schematic diagram of the interaction process of an intelligent interactive system provided by an exemplary embodiment of this disclosure; Figure 4 This is a schematic diagram of the structure of an electronic device provided by an exemplary embodiment of this disclosure.

[0020] Explanation of reference numerals in the attached figures: 1. Intelligent Interaction System; 10. Communication Connection Module; 20. Intent Recognition Module; 30. Business Processing Module; 40. Message Push Module. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] In the description of this application, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0023] "A and / or B" includes the following three combinations: A only, B only, and a combination of A and B.

[0024] The use of "applies to" or "configured to" in this application implies open and inclusive language, which does not exclude the applicability to or configuration to devices performing additional tasks or steps. Additionally, the use of "based on" implies openness and inclusivity, because processes, steps, calculations, or other actions "based on" one or more of the stated conditions or values ​​may in practice be based on additional conditions or values ​​beyond those stated.

[0025] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0026] On the one hand, this embodiment provides an intelligent interactive system, such as Figure 1 As shown, the intelligent interactive system 1 includes a communication connection module 10, an intent recognition module 20, a business processing module 30, and a message push module 40.

[0027] The communication connection module 10, deployed on the intranet server, is used to establish a connection between the intranet server and the external communication server, and to receive external user messages and event callbacks. In other words, the communication connection module 10 serves as the sole entry and exit point for internal and external data interaction, responsible for handling external network message access, authentication of the communication connection module, link maintenance, and message preprocessing. It enables basic communication between the intranet server and the external communication server, and can normally receive various text messages actively sent by users in logged-in communication clients, as well as various interactive event callback data generated by the platform.

[0028] The intent recognition module 20 is used to identify user interaction intents corresponding to external user messages and event callbacks based on a large language model and a pre-set intent configuration file. In other words, the intent recognition module 20 receives pre-processed interaction data and, relying on the large language model and the pre-entered intent configuration file, completes basic semantic parsing and preliminary identification of user interaction intents.

[0029] The business processing module 30 includes multiple business processing units, which are used to call the business processing unit corresponding to the user interaction intent to execute intranet business operations and generate response messages. That is, the business processing module 30, as an intranet business execution module, accurately distributes tasks to different functional business units through the internal business routing unit, completes various interface-less intranet platform automated operations with the help of browser automation tools, and outputs business processing results.

[0030] The message push module 40 is used to push response messages to target users or groups via the communication connection module 10. That is, the message push module 40 combines all business results, timed notifications and abnormal prompts, and sends them back to the communication connection module 10, which then sends them to the user terminal or designated group.

[0031] This embodiment deploys a communication connection module 10, an intent recognition module 20, a business processing module 30, and a message push module 40 on an intranet server, forming a complete interactive closed loop. The communication connection module 10 ensures secure data exchange between the internal and external networks and stable access to various interactive data. Combined with a large language model and preset intent configuration files, it achieves accurate intent recognition of user messages and interactive events. The business processing module 30, with the coordinated efforts of multiple business units and the use of browser automation tools, automates the operation of intranet business systems without open interfaces. Finally, the message push module 40 uniformly pushes and provides feedback on the business results. This effectively breaks down the barriers between external network user interaction and intranet business automation, achieving intelligent and automated processing of the entire process from user interaction triggering, intelligent intent recognition, automated intranet business execution, to accurate result push. This significantly reduces intranet operation and maintenance costs and improves intranet business processing efficiency and user interaction convenience.

[0032] In some embodiments, the communication connection module 10 further includes a message preprocessing unit, which is used to parse, filter and standardize the received external user messages and event callbacks, and to perform cache scheduling through a message queue; Event callbacks include event notifications for user-sent messages, user-joined or unjoined groups, and user-triggered interactive commands, which are used to trigger the intent recognition module 20 to start the intent recognition process.

[0033] Specifically, in this embodiment, the communication connection module 10 has a built-in independent message preprocessing unit. After all kinds of raw interactive data pushed by the external communication server are uniformly gathered to the communication connection module 10, it first automatically completes message parsing, filtering of invalid and redundant data, and standardized processing of text format for the received external user messages and various event callback data. It removes invalid and abnormal data such as garbled characters, empty data, and duplicate pushes, ensuring that the interactive data transmitted later is clean, standardized, and uniform. At the same time, the preprocessing unit is equipped with a message queue scheduling mechanism, which can orderly cache, queue, and schedule multiple user messages and event callbacks that suddenly flood in with high concurrency. The data is transmitted to downstream modules in the order of receipt, thereby effectively avoiding problems such as message out-of-order, data loss, process congestion, and identification errors in high-concurrency scenarios, and ensuring the orderliness and stability of data transmission.

[0034] Furthermore, the specific types of event callbacks that this embodiment can respond to cover various user behavior event notifications, such as users actively sending dialogue messages, users joining groups, users leaving groups, and users triggering interactive commands, comprehensively covering diverse interaction scenarios such as community interaction, active command triggering, and daily dialogue. All types of event callbacks can serve as effective trigger signals to directly wake up and start the intent recognition process of the intent recognition module 20, enabling the system not only to passively recognize the intent of user text messages, but also to automatically trigger intelligent recognition based on various user community operations and interactive operations, which can greatly broaden the interaction triggering dimensions of the system.

[0035] In some embodiments, the communication connection module 10 is further configured to complete the connection subscription and identity authentication between the communication connection module 10 and the external communication server through a preset identity identifier and key, and establish a communication link with the external communication server based on a preset communication protocol; It periodically sends heartbeat data packets to keep the communication link active and monitors the connection status of the communication link.

[0036] Specifically, in this embodiment, the communication connection module 10 pre-integrates the official development kit of the external communication service. Staff pre-enter a unique, preset identity identifier and encryption key into the module. After the system powers on, the communication connection module 10 automatically initiates a connection subscription request to the external communication server, completing identity verification and permission confirmation between the communication connection module 10 and the external communication server, and establishing an encrypted communication link. For example, a long-lived connection channel is built using WebSocket or Transmission Control Protocol (TCP). After the link is established, the communication connection module 10 periodically sends heartbeat data packets to the server at fixed time intervals to verify the link connectivity in real time. If any abnormal state such as network interruption, service disconnection, or timeout without response is detected, the module automatically initiates the disconnection reconnection logic, repeating the identity authentication and link establishment process to ensure the long-term stable online communication channel.

[0037] Furthermore, the communication connection module 10 incorporates a message preprocessing unit. Various event callback data, such as ordinary user dialogue messages, user group entry events, user group exit events, and user click interaction commands pushed from the external communication server, are uniformly integrated into the communication connection module 10. The preprocessing unit then sequentially parses these data packets, filters out invalid and redundant information, and standardizes the text format. The processed and compliant interaction data is then stored in a built-in message queue, undergoing ordered caching and scheduling to avoid message out-of-order issues, message loss, and data congestion in high-concurrency scenarios. After scheduling, the valid interaction data is transmitted in an orderly manner to the subsequent intent recognition module 20. Various event callback data can directly trigger the intent recognition module 20 to enter the working recognition state.

[0038] This embodiment utilizes a communication connection module 10 to complete two-way authentication and connection subscription using a preset identity identifier and key. It constructs a dedicated encrypted long-term communication link by combining a preset communication protocol. By periodically sending heartbeat data packets, it monitors the link status in real time and automatically reconnects in case of anomalies. This effectively ensures the legality, security, and long-term stability of data interaction between the intranet and external communication servers. At the same time, the built-in message preprocessing unit parses, filters, standardizes, and schedules the received user messages and various event callbacks in an orderly manner, achieving standardized and orderly transmission of interactive data. This enables stable triggering of the intent recognition process, thereby significantly improving the overall data transmission quality and the continuity and reliability of intelligent interaction of the system.

[0039] In some embodiments, the intent configuration file is in a structured format, including an intent identifier, a corresponding processing unit name, a storage path, and an intent description.

[0040] Specifically, in this embodiment, the intent recognition module 20 locally stores a preset structured intent configuration file. The configuration file contains all preset interactive intents one by one. Each intent information is filled with a unique intent identifier, the name of the corresponding business processing unit, the storage path of the unit program, and a textual explanation of the intent, so as to realize a one-to-one binding relationship between intents and execution units.

[0041] After the preprocessed user messages and event callback data are transmitted to the intent recognition module 20, the intent recognition module 20 automatically calls the locally deployed large language model to first perform semantic understanding and semantic parsing on the user's input colloquial text, short instructions, and event trigger information. Then, it accurately compares and matches the parsing results with the intent keywords and semantic features stored in the intent configuration file to quickly determine the specific user interaction intent corresponding to the current interaction behavior, generate the intent instruction corresponding to the user interaction intent, and transmit it to the business processing module 30.

[0042] This embodiment establishes a one-to-one binding relationship between intents and business processing units by using a structured intent configuration file that includes intent identifiers, corresponding processing unit names, storage paths, and intent descriptions. Combined with a large language model, semantic parsing and precise comparison and matching are performed on preprocessed user messages and event callbacks. This enables the rapid and accurate determination of user interaction intents and the output of corresponding intent commands, thereby standardizing intent recognition logic and business distribution criteria. This helps improve the accuracy, stability, and regularity of system intent recognition and ensures that subsequent intranet business processing units can be accurately invoked.

[0043] In some embodiments, the service processing module 30 includes: A general consultation unit is used to respond to consultation intent and call a large language model to generate a natural language response. The device information query unit is used to respond to the intent of device status query, call the browser automation tool to log in to the intranet device management platform, query and return the device software and hardware version, model and operation record according to the device identifier; The software push unit is used to respond to software deployment intentions, extract and verify key parameters from external user messages, and call browser automation tools to log in to the intranet software management platform to perform remote deployment operations. The project management query unit is used to respond to project progress query intents. It calls browser automation tools to log in to the intranet project management platform, obtains project nodes, progress, issue information and personnel information, and generates response content by summarizing it through a large language model. Among them, the browser automation tool is used to simulate human browser operation to achieve login, data query, data acquisition and command execution for intranet business systems that do not have open application programming interfaces.

[0044] Specifically, this embodiment configures corresponding business processing units for different user interaction intentions to achieve accurate matching between intentions and business processing.

[0045] The general consultation unit corresponds to routine consultation interactions. When the business routing unit determines that the current user's interaction intent is a routine consultation, common sense Q&A, or process consultation, the task is distributed to the general consultation unit without needing to log in to the intranet business system. The general consultation unit directly calls the system's built-in large language model to semantically understand and answer the user's questions and routine inquiries. Based on the natural language generation capabilities of the large language model, it automatically generates fluent, standardized, and usage-specific natural language text responses without logging into any intranet business platform or retrieving intranet business data. After generation, the responses are organized into a response message and transmitted to the message push module 40 for unified distribution, enabling rapid response to lightweight interaction needs such as routine Q&A, process consultation, and common sense inquiries.

[0046] The device information query unit targets device status queries, focusing on intranet device operation and maintenance query scenarios. After a user initiates a query command related to device status, hardware, or operation, the business routing unit distributes the task to the device information query unit. The device information query unit, using a browser automation tool, fully simulates the human browser operation process, automatically completing operations such as logging into the intranet device management platform, accessing pages, and performing conditional searches. Based on the device identifier carried in the user message, it accurately queries the corresponding device's hardware model, software and hardware version parameters, historical operation records, and other data. After obtaining complete device information, it generates a corresponding business response message, achieving unmanned automatic querying of device information.

[0047] For example, the device information query unit automatically extracts the device identifier filled in by the user from the user interaction message. This device identifier includes unique identification codes such as the vehicle unique identifier, device hardware serial number, and vehicle control unit number. After extraction, the unit launches the built-in browser automation tool, simulating the entire manual operation process of opening a browser, entering the intranet address, filling in the intranet login account and password, and completing the intranet permission verification, successfully logging into the intranet device management platform. Then, the browser automation tool automatically enters the extracted device identifier into the platform's search box, completes a precise search query, and automatically obtains all relevant data such as the device's corresponding hardware model parameters, device built-in software version information, device historical operation records, and device online / offline status. Finally, the obtained raw data is organized and formatted into concise and easy-to-understand text content to generate a device query response message.

[0048] The software push unit corresponds to the software deployment intent and is used to realize remote software upgrades and deployment operations for intranet devices. After recognizing the user's interactive commands related to vehicle software upgrades and remote program deployment, the business routing unit distributes the task to the software push unit. The software push unit can intelligently parse the external user message content corresponding to the interactive commands, extract the key information required by the business, log in to the intranet software management platform using browser automation tools, and automatically complete remote operations such as device selection, software version matching, and deployment command submission. It can complete batch deployment and upgrade tasks of intranet software without manual background operation, realizing the automated execution of software deployment business.

[0049] For example, the software push unit first performs intelligent content extraction on the interactive text sent by the user, automatically obtaining several key business parameters contained in the text, such as the device model, the version number of the software to be pushed, the software push batch number, and the unique identifier of the vehicle control unit. After extraction, the unit automatically performs parameter integrity verification. If it detects missing key parameters, incorrect parameter input, or incorrect parameter format, it generates a parameter completion prompt and pushes it to the user via the message push module 40 and the communication connection module 10, reminding the user to supplement and improve the corresponding information. If the parameters are complete and correct, it calls a browser automation tool to simulate manual login to the intranet software management push platform, selects the target push devices in batches according to the extracted parameters, and automatically issues software remote deployment and version upgrade execution instructions, completing the fully automated operation of the intranet backend software push process. After the operation is completed, the execution process information is recorded, and a software push execution response message is generated.

[0050] The project management query unit corresponds to the intent of project progress query and is adapted to the business scenario of intranet project control. When a user initiates a query request related to project progress, project issues, or personnel work, the task is distributed to the project management query unit. The project management query unit logs into the intranet project management platform through a browser automation tool, automatically obtains full-dimensional raw business data such as project execution nodes, real-time progress, existing issues, and personnel work logs, and calls a large language model to integrate, summarize, and streamline the scattered and fragmented raw data to generate a clear and concise project summary response, providing users with visualized project operation query results.

[0051] Furthermore, the project management query unit incorporates built-in automatic classification and organization logic, categorizing acquired project issues into three levels based on their impact: critical issues, high-impact usage issues, and routine, mandatory issues. It also automatically filters and statistically analyzes long-standing, unresolved, and overdue project issues. In addition, it uniformly summarizes and integrates personnel information such as daily work content, phased work progress, and completed and incomplete task records for the responsible personnel. Finally, based on a large language model, it integrates, simplifies, and summarizes a large amount of scattered raw project data, organizing it into a clear and focused project summary, ultimately generating project query response messages.

[0052] The browser automation tool in this embodiment serves as a carrier for adapting to closed intranet systems. Designed for various intranet business systems that lack open application programming interfaces and cannot achieve direct program integration, it can completely simulate the entire process of manual browser login, page operation, data reading, command submission, and information retrieval. It enables efficient execution of external network interaction commands to intranet business systems without requiring program modifications to existing intranet business systems or the addition of new data integration interfaces. This significantly lowers the threshold for intelligent intranet interaction transformation and is compatible with relevant legacy intranet office platforms and closed intranet management systems within enterprises.

[0053] In some embodiments, the key parameters extracted by the software push unit include device model, software version, batch information, and control unit identifier; In the event of missing key parameters, supplementary prompts are sent to the user via the communication connection module 10.

[0054] Specifically, this embodiment determines the types of key business parameters that the software push unit needs to extract for the software deployment business. These parameters include the device model of the device to be upgraded, the software version to be deployed, the batch information corresponding to the software push, and the identifier of the target vehicle control unit. The execution target and deployment standard of this software push can be uniquely identified through these key business parameters.

[0055] After receiving the user's software deployment intent command, the software push unit automatically performs intelligent parsing and keyword extraction on the user-input external message text, batch captures the four types of key parameters mentioned above, and immediately performs parameter integrity verification. If the verification result shows any missing parameters such as missing device model, unknown software version, no batch information, or missing control unit identifier, the system will not perform the intranet software deployment operation to avoid business anomalies such as deployment errors, device push errors, and batch upgrade failures due to incomplete parameters.

[0056] Meanwhile, this embodiment sets up a comprehensive missing parameter prompt mechanism. When a key parameter is detected to be missing, the software push unit generates a parameter supplement prompt message and calls the message transmission capability of the communication connection module 10 to accurately push the prompt content to the target user who initiated the command, clearly informing the user of the type of missing parameter, guiding the user to complete the business parameters and then reissue the deployment command, effectively avoiding invalid business execution, thereby improving the accuracy, security and interactive experience of remote deployment of intranet software.

[0057] In some embodiments, the intelligent interaction system further includes: The scheduled push module is used to automatically execute preset tasks and generate notification messages according to preset time or business trigger conditions; The message push module 40 is also used to push notification messages to target users or groups through the communication connection module 10.

[0058] Specifically, the function of the timed push module in this embodiment to automatically execute preset tasks and generate notification messages is parallel to the real-time interaction process, thus enabling the system to have a dual start-up mode of autonomous timed wake-up and event-triggered wake-up.

[0059] For the scheduled push mode, a fixed execution time point is preset in the scheduled push module. When the system reaches the preset time, the scheduled push module automatically starts the preset summary and statistics task to uniformly summarize and statistically analyze data such as the overall progress of the intranet project, the operating status of all devices, and the completion status of batch software pushes, and automatically generate routine notification messages such as daily work summaries and weekly progress reports.

[0060] For business event triggering modes, the timed push module monitors the execution status of the intranet software push unit in real time. Once it detects that all intranet software remote deployment and version push operations have been completed, it determines that the business conditions for software deployment completion have been triggered. It automatically delays and initiates a secondary inspection query of the device upgrade results, automatically verifies the number of devices that have been successfully upgraded, the number of devices that failed to upgrade, the cause of abnormal failures, and other follow-up data, and automatically generates a software push completion follow-up notification message based on the verification results.

[0061] Then, all timed notification messages are aggregated into the message push module 40. These notification messages fall into three main categories: business response messages generated upon completion of real-time user transactions, timed notification messages generated autonomously by the timed push module, and exception notification messages generated by the exception handling unit. The message push module 40 formats and matches the notification messages to their recipients, precisely categorizing and pushing them to specific individual users or designated internal communication groups according to requirements. After formatting, all messages to be pushed are uniformly transmitted back to the intranet communication connection module 10. The communication connection module 10, based on the established stable long-term connection channel, encapsulates the messages according to the message format of the external communication server, obtaining the encapsulation results. Finally, all encapsulation results are pushed to the external communication client pages used by users daily, completing a fully closed-loop intelligent interaction process from external network-initiated interaction, automatic intranet processing, and result feedback.

[0062] In some embodiments, the business triggering condition includes a software deployment completion event; Upon receiving a software deployment completion event, the scheduled push module queries the deployment execution results and generates a follow-up notification message.

[0063] Specifically, the timed push module in this embodiment monitors the execution status of the intranet software push service in real time, and can continuously capture feedback signals when the software push unit completes remote software deployment, batch device version upgrades, and other complete executions, and accurately identify software deployment completion events.

[0064] When the scheduled push module detects a software deployment completion event, it automatically triggers a preset business follow-up task. There is no need for manual triggering of instructions or manual querying of deployment results. It proactively performs a secondary query and verification of the deployment execution data on the intranet software management platform, accurately obtaining full-dimensional execution result data such as the number of devices deployed, information on successfully upgraded devices, details of devices that failed to upgrade, reasons for abnormal failures, and overall deployment completion progress.

[0065] The scheduled push module automatically integrates and summarizes data information based on the deployment execution results obtained from the query, generating a clear and complete follow-up notification message to fully record the overall execution status of this software deployment. Finally, through the message push module 40 in conjunction with the communication connection module 10, the follow-up notification message is accurately pushed to the target users or groups, allowing users to monitor the software deployment effect in real time. This achieves a closed-loop business process for software distribution, deployment execution, and result follow-up, thereby effectively improving the integrity, controllability, and ease of operation and maintenance of remote software deployment within the intranet.

[0066] In some embodiments, the business processing module 30 further includes an exception handling unit, which generates an exception prompt message and pushes it to the user when the business operation fails, parameters are missing, or intent recognition is uncertain.

[0067] Specifically, in this embodiment, the exception handling unit is deployed and integrated inside the business processing module 30, which monitors the entire business execution process of all functional business units in real time and comprehensively monitors abnormal system operation scenarios.

[0068] In situations where business operations fail, such as browser automation tools failing to log in to the intranet platform, the intranet platform server becoming unresponsive, software push command submission failing, or device data retrieval yielding no results, the exception handling unit promptly captures abnormal operation signals. Simultaneously, when the intent recognition module 20 encounters semantic ambiguity, polysemous statements failing to accurately determine the interaction intent, or when user-submitted key business parameters are severely missing or have conflicting parameter logic, the exception handling unit synchronously initiates its working logic, automatically matching the corresponding exception scenario to generate an exception prompt message. This message replaces invalid blank responses and system error codes, and is then uniformly sent to the user terminal via the push link, effectively improving the overall intelligent interaction user experience.

[0069] On the other hand, this embodiment provides an intelligent interaction method, please refer to... Figure 2 The method includes the following steps: Step 100: In response to the establishment of a connection between the intranet server and the external communication server, and the receipt of external user messages and event callbacks. That is, a communication link is established with the external communication server through the communication connection module deployed on the intranet server, and external user messages and event callbacks are received through the communication link.

[0070] Step 200: Based on the large language model and the preset intent configuration file, identify the user interaction intent corresponding to external user messages and event callbacks.

[0071] Step 300: Invoke the business processing unit corresponding to the user interaction intent to execute the intranet business operation and generate a response message.

[0072] Step 400: Push the response message to the target user or group.

[0073] Please see Figure 3 , Figure 3 The diagram illustrates the complete interaction flow of an intelligent interaction system 1. Specifically, it includes: a user initiating a conversational interaction on a communication client already logged into an external communication server, generating external user messages and event callbacks such as dialogue messages, group operations, or interaction commands. A communication connection module 10, deployed on an intranet server, establishes a long-term communication link with the external communication server based on a preset identity identifier and key, and continuously receives the external user messages and event callback data while the link remains active. Upon receiving the data, the communication connection module 10 enters the external user message receiving and preprocessing stage, parsing, filtering, formatting, and queuing the raw data to achieve the regularization and orderly distribution of the interactive data. The preprocessed data is then uniformly forwarded by the communication connection module 10 to the intent recognition module 20, where natural language preprocessing is used to clean, segment, and semantically regularize the user's text content, providing input data for intent recognition.

[0074] Assuming the device in this embodiment is a vehicle, the normalized text data is sent to the intent recognition module 20. The system accurately parses and identifies the user's corresponding interactive intents such as consultation, vehicle query, software push, and project query based on the application programming interface (API) of the intranet large language model and the structured intent configuration file.

[0075] After completing intent recognition, the system enters the business processing module 30. Based on the identified user intents, the system automatically routes business branches, leading to the corresponding business processing flows. When the intent is identified as a general consultation intent (such as casual conversation), the general consultation unit in the system performs a natural language response generation operation, using the large language model to call the Model Context Protocol (MCP) tool to generate the corresponding natural language response content. When the intent is identified as a vehicle information query intent, the device information query unit in the system sequentially performs vehicle identifier (such as Vehicle Identification Number (VIN)) extraction, vehicle information query, and result formatting steps. It logs into the intranet device management platform using a browser automation tool, obtains vehicle hardware and software parameters and operating records based on unique identifiers such as vehicle version information, and summarizes and outputs the results through the large language model's API. When the intent is identified as a software push deployment intent, the system's software push unit sequentially completes key parameter extraction and verification, and the Over-the-Air Technology (OTA) push execution process, automatically extracting key parameters such as vehicle model, software version, and batch information. With key parameters verified to be correct, the system logs into the OTA management platform via a browser automation tool to perform remote deployment operations for intranet software, such as downloading and pushing software packages. If key parameters are missing, a user-provided supplementary prompt is generated, and push information is returned via the API of the large language model. When the system identifies a project management query intent or a personnel information query intent, the system's project management query unit sequentially performs project information acquisition, issue classification and filtering, and personnel information aggregation operations. It automatically collects intranet project nodes, progress, fault issues, and personnel information, classifies project issues by severity level, filters long-unresolved issues, completes multi-dimensional data integration, and summarizes and outputs the results via the API of the large language model.

[0076] Meanwhile, the system supports a scheduled task triggering mechanism, which can automatically start tasks based on preset time or business trigger conditions such as software deployment completion. For example, when the set time is reached, it can automatically summarize and output project information and personnel information; or, it can automatically start a timer thread to automatically trigger a vehicle upgrade status query task two hours after the OTA push operation is completed, generating and returning push information.

[0077] Ultimately, all business response messages, scheduled notification messages, and exception alert messages are uniformly converged into the message push module 40. Through the external user message sending link and a stable WebSocket long connection communication channel, they are accurately pushed to the target users or enterprise groups. This achieves intelligent and automated interactive operation of the entire process, from user interaction triggering, identity security verification, intelligent message preprocessing, accurate intent recognition, multi-branch automated intranet business processing, data regularization and generation, scheduled business follow-up, to the final closed-loop message push.

[0078] This embodiment also provides an electronic device; please refer to [link / reference]. Figure 4 It includes a memory and a processor. In a specific example, the memory stores a computer program, which, when executed by the processor, implements the method of any of the above embodiments.

[0079] This embodiment also provides a computer-readable storage medium having a computer program stored thereon, the computer program being loaded by a processor to perform the steps of any of the methods in the above embodiments.

[0080] In the embodiments of this application, the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc.

[0081] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0082] The above provides a detailed description of an intelligent interactive system 1 and method provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope according to the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. An intelligent interactive system, characterized in that, The intelligent interaction system includes: The communication connection module, deployed on the intranet server, is used to establish a connection between the intranet server and an external communication server, and to receive external user messages and event callbacks. The intent recognition module is used to identify the user interaction intent corresponding to the external user message and the event callback based on the large language model and the preset intent configuration file. The business processing module includes multiple business processing units, which are used to call the business processing unit corresponding to the user interaction intent to perform intranet business operations and generate response messages; The message push module is used to push the response message to the target user or group through the communication connection module.

2. The intelligent interactive system according to claim 1, characterized in that, The communication connection module also includes a message preprocessing unit, which is used to parse, filter and standardize the received external user messages and event callbacks, and to cache and schedule them through a message queue. The event callbacks include event notifications for user sending messages, user joining or leaving groups, and user triggering interactive commands, which are used to trigger the intent recognition module to start the intent recognition process.

3. The intelligent interactive system according to claim 1, characterized in that, The communication connection module is also used to complete the connection subscription and identity authentication between the communication connection module and the external communication server through a preset identity identifier and key, and to establish a communication link with the external communication server based on a preset communication protocol; The system periodically sends heartbeat data packets to keep the communication link active and monitors the connection status of the communication link.

4. The intelligent interactive system according to claim 1, characterized in that, The intent configuration file is in a structured format, including intent identifier, corresponding processing unit name, storage path, and intent description.

5. The intelligent interactive system according to claim 1, characterized in that, The business processing module includes: A general consultation unit is used to respond to consultation intent and call the large language model to generate a natural language response. The device information query unit is used to respond to the intent of device status query, call the browser automation tool to log in to the intranet device management platform, query and return the device software and hardware version, model and operation record according to the device identifier; The software push unit is used to respond to the software deployment intent, extract key parameters from the external user message and verify them, and call the browser automation tool to log in to the intranet software management platform to perform remote deployment operations; The project management query unit is used to respond to project progress query intents. It calls browser automation tools to log in to the intranet project management platform, obtains project nodes, progress, issue information and personnel information, and generates response content by summarizing it through a large language model. The browser automation tool is used to simulate human browser operation to achieve login, data query, data acquisition and command execution for intranet business systems that do not have open application programming interfaces.

6. The intelligent interactive system according to claim 5, characterized in that, The key parameters extracted by the software push unit include device model, software version, batch information, and control unit identifier; If the key parameters are missing, a supplementary prompt will be sent to the user through the communication connection module.

7. The intelligent interactive system according to claim 5, characterized in that, The project management query unit categorizes project issues by severity level, including serious faults, high-perceived issues, and issues that must be resolved, and filters out long-unresolved issues; personnel information includes work content, progress, and task records.

8. The intelligent interactive system according to claim 1, characterized in that, The intelligent interaction system also includes: The scheduled push module is used to automatically execute preset tasks and generate notification messages according to preset time or business trigger conditions; The message push module is also used to push the notification message to the target user or group through the communication connection module.

9. The intelligent interactive system according to claim 8, characterized in that, The business triggering conditions include software deployment completion events; Upon receiving the software deployment completion event, the scheduled push module queries the deployment execution result and generates a return visit notification message.

10. An intelligent interaction method, characterized in that, The method includes the following steps: It responds to the establishment of a connection between the internal network server and the external communication server, and receives external user messages and event callbacks; Based on the large language model and the preset intent configuration file, the user interaction intent corresponding to the external user message and the event callback is identified; The business processing unit corresponding to the user interaction intent is invoked to perform intranet business operations and generate a response message; The response message is pushed to the target user or group.

Images