Session processing method and system of collaborative multi-agent
Through the collaborative work of the front-end agent with the Q&A database and the back-end AI model, an adaptive UI interface is generated, which solves the problems of robustness and interactive experience in multi-agent conversation processing, and realizes efficient cross-system business processing and users Friendly interaction method.
Patent Information
- Application Number
- CN202411177083.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-12-06
AI Technical Summary
The existing multi-agent conversation processing methods have shortcomings in terms of robustness and user interaction experience, and cannot push fixed content in different scenarios, and the background operation interface is complicated, resulting in low business efficiency and poor user experience.
Receive messages through the front-end agent and judge the matching situation with the Q&A database. If it matches, generate UI elements; if it doesn't match, send them to the back-end AI model for compilation, generate corresponding UI interfaces, and realize one-stop service processing across systems .
It has improved the operational efficiency of the business system several times, lowered the user's usage threshold, transformed into a human-computer dialogue and interaction method, improved the user experience and reduced the AI illusion.
Smart Images

Figure CN119088866A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of artificial intelligence, and in particular to a method and system for collaborative multi-agent conversation processing. Background Art
[0002] With the continuous development of science and technology, people's demand for intelligent services is getting higher and higher. Multi-agent conversation processing can be applied to many fields such as intelligent customer service, smart home, and autonomous driving. The existing multi-agent conversation processing methods have the following problems: the current agent is not robust enough, resulting in the reply content not necessarily being what the user really wants, especially in the actual business operation process, the current agent cannot push fixed content in different scenarios; the current agent's background operation interface is fixed and complicated; the current agent receives messages sent by terminal users and then responds to the messages, and cannot receive messages from external business systems and then actively send the messages to terminal users. Therefore, the existing multi-agent conversation processing methods have the disadvantages of low business efficiency, poor user interaction experience, and serious AI hallucination problems. Summary of the invention
[0003] The present invention provides a method, apparatus, device and storage medium for collaborative multi-agent session processing.
[0004] According to a first aspect of the present disclosure, a collaborative multi-agent conversation processing method is provided.
[0005] The method includes:
[0006] The front-end agent receives the message sent by the terminal user or the external business system and determines the matching situation between the message and the Q&A database in the front-end agent;
[0007] If the message matches the Q&A database, the front-end agent obtains the corresponding feedback information according to the matching data, and sends the primary key ID in the feedback information to the front-end component generation device through the corresponding front-end component SDK called to generate the corresponding UI element; the front-end component SDK receives the corresponding UI element sent by the front-end component generation device and renders the UI element to the front-end agent to generate the corresponding UI interface; wherein the front-end agent is integrated with the front-end component SDK;
[0008] If the message does not match the Q&A database, the front-end agent sends the message to the AI big model in the back-end AI agent, and the AI big model outputs corresponding feedback information and sends it to the front-end agent; the front-end agent recompiles the feedback information, and generates the corresponding UI interface according to the operation of generating the corresponding UI interface using the primary key ID after the recompile;
[0009] The terminal user interacts with the generated UI interface and triggers the control; the front-end component generation device calls the API of the corresponding external business system according to the triggered control; the external business system executes the corresponding business logic according to the called API and sends the execution result to the front-end intelligent agent; the front-end intelligent agent displays the execution result to the terminal user after secondary compilation or transparent transmission.
[0010] In some implementations of the first aspect, sending the primary key ID in the feedback information to the front-end component generating device through the called corresponding front-end component SDK to generate the corresponding UI element includes:
[0011] The front-end agent calls the corresponding front-end component SDK according to the identifier in the feedback information, and sends the primary key ID in the feedback information to the called front-end component SDK;
[0012] The front-end component SDK sends the received primary key ID to the front-end component generation device to obtain the UI component and configuration package corresponding to the primary key ID;
[0013] The front-end component generating device compiles the UI component according to the configuration package to generate corresponding UI elements.
[0014] In some implementations of the first aspect, the front-end component SDK sends the received primary key ID to the front-end component generation device to obtain the UI component and configuration package corresponding to the primary key ID, including:
[0015] The front-end component generating device receives the primary key ID sent by the front-end component SDK, and searches for the corresponding UI component and configuration information in the pre-configured UI component library according to the received primary key ID, and obtains the UI component and configuration package corresponding to the primary key ID.
[0016] In some implementations of the first aspect, the feedback information further includes:
[0017] Visual feedback information and logical feedback information; among which,
[0018] The visual feedback information includes feedback operation controls and feedback charts; the visual feedback information is displayed in the generated UI interface;
[0019] The logic feedback information includes a feedback algorithm and a code block; the logic feedback information is used to execute various logic judgment processes to generate corresponding UI interface operations.
[0020] In some implementations of the first aspect, the AI big model is an AI model generated by training a general AI big model using a training knowledge base; wherein,
[0021] The training knowledge base includes a function model and a Q&A database.
[0022] In some implementations of the first aspect, the using the training knowledge base to train the general AI big model includes:
[0023] The weights of each knowledge element are trained through question-answer pairs in the Q&A database;
[0024] According to the trained weights of each knowledge element, the general AI model is trained using the corresponding algorithm in the function model library.
[0025] According to a second aspect of the present disclosure, a collaborative multi-agent conversation processing system is provided.
[0026] The system includes:
[0027] A front-end component generating device, an external business system, a back-end AI agent and a terminal user connected to the front-end agent respectively; wherein the front-end component generating device 220 is connected to the front-end component SDK 260 in the front-end agent and the external business system 230 respectively;
[0028] The front-end intelligent agent is used to receive messages sent by terminal users or external business systems and determine the matching status of the messages with the Q&A database in the front-end intelligent agent; if the message matches the Q&A database, the front-end intelligent agent obtains corresponding feedback information according to the matching data, and sends the primary key ID in the feedback information to the front-end component generation device by calling the corresponding front-end component SDK to generate the corresponding UI element; if the message does not match the Q&A database, the front-end intelligent agent sends the message to the AI big model in the back-end AI intelligent agent; it is also used to perform secondary compilation on the feedback information, and generate the corresponding UI interface according to the generation of the corresponding UI interface operation using the secondary compiled primary key ID; it is also used to display the execution result to the terminal user after secondary compilation or transparent transmission;
[0029] The front-end component generating device is used to receive the primary key ID sent by the front-end component SDK; it is also used to generate the corresponding UI element and send it to the corresponding front-end component SDK; it is also used to call the API of the corresponding external business system according to the triggered control;
[0030] The front-end component SDK is used to send the primary key ID to the front-end component generation device; it is also used to receive the corresponding UI elements sent by the front-end component generation device and render the UI elements to the front-end intelligent body to generate the corresponding UI interface;
[0031] The back-end AI agent is used to receive messages sent by the front-end agent; it is also used to output corresponding feedback information through the AI big model and send it to the front-end agent;
[0032] The external business system is used to execute the corresponding business logic according to the called API and send the execution results to the front-end agent;
[0033] It is used by end users to interact with the generated UI interface and trigger controls; it is also used to receive content displayed by the front-end intelligent agent after secondary compilation or transparent transmission.
[0034] According to a third aspect of the present disclosure, an electronic device is provided. The electronic device includes: at least one processor; and a memory connected to the at least one processor in communication; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the method described above.
[0035] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, where the computer instructions are used to cause a computer to execute the method as described above.
[0036] In the present disclosure, the front-end intelligent agent receives a message sent by a terminal user or an external business system and determines whether the message matches the Q&A database in the front-end intelligent agent; if the message matches the Q&A database, the front-end intelligent agent obtains corresponding feedback information based on the matching data, and sends the primary key ID in the feedback information to the front-end component generation device through the corresponding front-end component SDK called to generate the corresponding UI element; the front-end component SDK receives the corresponding UI element sent by the front-end component generation device and renders the UI element to the front-end intelligent agent to generate the corresponding UI interface; if the message does not match the Q&A database, the front-end intelligent agent sends the message to the AI big model in the back-end AI intelligent agent, outputs the corresponding feedback information through the AI big model and sends it to the front-end intelligent agent; the front-end intelligent agent performs secondary compilation on the feedback information, and generates the corresponding UI interface according to the primary key ID after secondary compilation according to the operation of generating the corresponding UI interface; the terminal user interacts with the generated UI interface and triggers the control; the front-end component generation device calls the API of the corresponding external business system according to the triggered control; the external business system executes the corresponding business logic according to the called API and sends the execution result to the front-end intelligent agent; the front-end intelligent agent displays the execution result to the terminal user after secondary compilation or transparent transmission. In this way, the operating efficiency of the business system can be improved several times, and the user's usage threshold can be greatly reduced. Users only need simple training to use the system and do not need to learn complicated background operations. The traditional background operation interaction method is transformed into a human-computer dialogue interaction method, which improves the user's interactive experience, reduces AI illusions, and realizes one-stop business processing across multiple systems.
[0037] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, among which:
[0039] Figure 1 A flowchart of a collaborative multi-agent session processing method provided by an embodiment of the present disclosure is shown;
[0040] Figure 2 A structural diagram of a collaborative multi-agent session processing system provided by an embodiment of the present disclosure is shown;
[0041] Figure 3 A block diagram of an exemplary electronic device capable of implementing an embodiment of the present disclosure is shown. DETAILED DESCRIPTION
[0042] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.
[0043] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.
[0044] In response to the problems arising in the background technology, the embodiments of the present disclosure provide a collaborative multi-agent conversation processing method and system. Specifically, the front-end intelligent agent receives a message sent by a terminal user or an external business system and determines whether the message matches the Q&A database in the front-end intelligent agent; if the message matches the Q&A database, the front-end intelligent agent obtains corresponding feedback information based on the matching data, and sends the primary key ID in the feedback information to the front-end component generation device through the called corresponding front-end component SDK to generate the corresponding UI element; the front-end component SDK receives the corresponding UI element sent by the front-end component generation device and renders the UI element to the front-end intelligent agent to generate the corresponding UI interface; if the message does not match the Q&A database, the front-end intelligent agent sends the message to the AI big model in the back-end AI intelligent agent, outputs the corresponding feedback information through the AI big model and sends it to the front-end intelligent agent; the front-end intelligent agent recompiles the feedback information, and generates the corresponding UI interface according to the operation of generating the corresponding UI interface using the secondary compiled primary key ID; the terminal user interacts with the generated UI interface and triggers the control; the front-end component generation device calls the API of the corresponding external business system according to the triggered control; the external business system executes the corresponding business logic according to the called API and sends the execution result to the front-end intelligent agent; the front-end intelligent agent displays the execution result to the terminal user after secondary compilation or transparent transmission. In this way, the efficiency of the business system can be improved several times, and the user's usage threshold can be greatly reduced. Users only need simple training to use the system, and do not need to learn complicated background operations. The traditional background operation interaction method is transformed into a human-computer dialogue interaction method, which improves the user's interactive experience, reduces AI illusions, and realizes one-stop business processing across multiple systems.
[0045] The collaborative multi-agent conversation processing method and system provided by the embodiments of the present disclosure will be described in detail below through specific embodiments in conjunction with the accompanying drawings.
[0046] Figure 1 A flowchart of a collaborative multi-agent session processing method provided by an embodiment of the present disclosure is shown. The method 100 includes the following steps:
[0047] S110, the front-end intelligent agent receives a message sent by a terminal user or an external business system and determines whether the message matches the Q&A database in the front-end intelligent agent.
[0048] S120, if the message matches the Q&A database, the front-end intelligent agent obtains the corresponding feedback information based on the matching data, and sends the primary key ID in the feedback information to the front-end component generation device by calling the corresponding front-end component SDK to generate the corresponding UI element; the front-end component SDK receives the corresponding UI element sent by the front-end component generation device and renders the UI element to the front-end intelligent agent to generate the corresponding UI interface; wherein, the front-end component SDK is integrated in the front-end intelligent agent.
[0049] In some embodiments, the primary key ID in the feedback information is sent to the front-end component generating device through the called corresponding front-end component SDK to generate the corresponding UI element, including:
[0050] The front-end agent calls the corresponding front-end component SDK according to the identifier in the feedback information, and sends the primary key ID in the feedback information to the called front-end component SDK;
[0051] The front-end component SDK sends the received primary key ID to the front-end component generation device to obtain the UI component and configuration package corresponding to the primary key ID;
[0052] The front-end component generating device compiles the UI component according to the configuration package to generate corresponding UI elements.
[0053] In some embodiments, the front-end component SDK sends the received primary key ID to the front-end component generation device to obtain the UI component and configuration package corresponding to the primary key ID, including:
[0054] The front-end component generating device receives the primary key ID sent by the front-end component SDK, and searches for the corresponding UI component and configuration information in the pre-configured UI component library according to the received primary key ID, and obtains the UI component and configuration package corresponding to the primary key ID;
[0055] Furthermore, the primary key ID has a one-to-one correspondence with the component ID in the pre-configured UI component library. The component ID corresponding to the primary key ID can be queried based on the received primary key ID. Each component corresponds to a unique component ID. Therefore, based on the queried component ID, the corresponding UI component and configuration package can be obtained from the pre-configured UI component library.
[0056] In some embodiments, a front-end component generating device may generate a control block through multiple control components;
[0057] Specifically, first, a unique primary key ID is defined for the control block and the type of the control block (such as a data display block) is determined. According to the determined type of the control block, the required control components are selected from the UI component library pre-configured in the front-end component generation device. These control components may be buttons, text input boxes, lists, tables, etc. The selected control components are dragged from the pre-configured UI component library to the canvas of the control block using the drag function, and attributes are configured for each selected control component, such as a placeholder for a text input box; trigger events are defined for the selected control components, such as click, input, selection and other events, and response actions corresponding to the trigger events are set. For example, clicking a button will respond to data submission or page jump actions; the workflow corresponding to the control block is configured in the front-end component generation device to ensure the normal operation of the interaction between the control components; finally, the position and size of each control component in the control block are adjusted, and styles are set for the control components and control blocks, including color, font, border, shadow, etc.
[0058] In some embodiments, the front-end component generating device may generate a display block through multiple chart components;
[0059] Specifically, select appropriate chart components (bar charts, pie charts, burndown charts, etc.) according to display requirements, configure properties for each chart component (such as color, font, title, legend, etc.) to ensure the aesthetics and readability of the chart, and set the interactive functions of the chart, such as click to enlarge, hover to display detailed information, etc.; configure the corresponding data source table according to the data requirements of the chart display. The data source table contains all the data fields required for the corresponding chart, and map each data point in the chart component to the corresponding field in the data source table; configure the corresponding data retrieval controls according to user retrieval requirements, such as time period, retrieval by name, filtering by label, etc. These data retrieval controls can help users quickly locate the data sets of interest and improve the flexibility and availability of data display; configure properties for each data retrieval control, such as control name, label, default value, optional value range, etc., to ensure the ease of use of the data retrieval control so that users can easily perform data retrieval operations; integrate the configured chart components and data retrieval controls into the display block to ensure smooth interaction between the chart components and the data retrieval controls, and users can dynamically update the data displayed in the chart through the data retrieval controls.
[0060] In some embodiments, the feedback information further includes:
[0061] Visual feedback information and logical feedback information; among which,
[0062] The visual feedback information includes feedback operation controls and feedback charts; the visual feedback information is displayed in the generated UI interface;
[0063] The logic feedback information includes a feedback algorithm and a code block; the logic feedback information is used to execute various logic judgment processes to generate corresponding UI interface operations.
[0064] S130, if the message does not match the Q&A database, the front-end intelligent agent sends the message to the AI big model in the back-end AI intelligent agent, and the AI big model outputs the corresponding feedback information and sends it to the front-end intelligent agent; the front-end intelligent agent recompiles the feedback information, and generates the corresponding UI interface according to the primary key ID after the recompilation.
[0065] In some embodiments, the AI big model is an AI model generated by training a general AI big model using a training knowledge base; wherein,
[0066] The training knowledge base includes a function model library and a Q&A database;
[0067] The general AI large model can be an LLM model;
[0068] Training methods include fine-tuning, hint engineering, and retrieval-augmented generation methods (RAG).
[0069] Furthermore, the function model library may be a deep learning model library or other algorithm model library.
[0070] In some embodiments, using a training knowledge base to train a general AI big model includes:
[0071] The weights of each knowledge element are trained through question-answer pairs in the Q&A database;
[0072] According to the weights of each trained knowledge element, use the corresponding algorithm in the function model library to train the general AI model;
[0073] For example, if the training method is prompt engineering, the designed prompt words are integrated into the knowledge elements with trained weights, and then the function model library is selected according to each knowledge element and its corresponding weight, and the general AI big model is trained using the corresponding algorithm in the function model library, so that the general AI big model can understand the meaning of the message sent by the terminal user or external business system according to the prompt words and output the corresponding reply. According to the gap between the reply output by the general AI big model and the expected reply, the model parameters of the general AI big model are optimized using the loss function until the loss function converges. At this time, the reply result output by the general AI big model is close to the expected reply. At this time, the general AI big model is a trained AI big model.
[0074] In some embodiments, AI enhancement means are used to train the weights of each knowledge element;
[0075] For example, retrieval enhancement technology or prompt engineering technology is used to analyze factors such as the context, question type, and answer quality of question and answer pairs in the Q&A database, and a reasonable weight is assigned to each knowledge element (such as concept, rule, etc.).
[0076] In some embodiments, the main interface for configuring the front-end agent, back-end AI agent, and Q&A database has the following functions:
[0077] Create front-end agents and back-end AI agents;
[0078] Set the global configuration entry. Click the entry to expand the global management related function menu;
[0079] Switch the agent group, that is, switch the front-end agent and the back-end AI agent;
[0080] Perform global settings, including setting agent groups, global IDs, and other information;
[0081] Set up a single agent configuration entry to configure related functions for a single agent, such as configuring a Q&A database;
[0082] Perform personalized configuration, including personalizing the name of the intelligent agent, etc.;
[0083] Furthermore, the Q&A database is configured, including:
[0084] Create Q&A question-answer pairs, and use the created Q&A question-answer pairs as Q&A database;
[0085] The answer in a Q&A question and answer pair usually consists of text, an identifier and a primary key ID. The identifier is used to determine the front-end component SDK that needs to be called, and the primary key ID is used to determine the block that needs to be called; the text is the content of the answer.
[0086] S140, the terminal user interacts with the generated UI interface and triggers the control; the front-end component generation device calls the API of the corresponding external business system according to the triggered control; the external business system executes the corresponding business logic according to the called API and sends the execution result to the front-end intelligent agent; the front-end intelligent agent displays the execution result to the terminal user after secondary compilation or transparent transmission.
[0087] In some embodiments, the business logic includes data query, event notification, business operation, other API calls, etc.
[0088] The above is an introduction to the method embodiment. The following further illustrates the scheme of the present disclosure by adopting a specific embodiment of the method. Specific embodiment 1:
[0090] The interaction process between the end user or external business system and the front-end agent is as follows:
[0091] The terminal user or external business system inputs the question text, the front-end intelligent agent receives and recognizes the question text, and then searches it in the Q&A database. If a text consistent with the question text is retrieved, a drop-down list is automatically displayed for the user or external business system to select the corresponding text. After the terminal user or external business system selects the text consistent with the question text, the front-end intelligent agent searches the Q&A database for the corresponding answer, compiles the corresponding answer for the second time, and then displays it to the terminal user or external business system; wherein the answer content after the second compilation consists of text, identifier, and primary key ID, the identifier is used to call the corresponding front-end component SDK to obtain the corresponding block, the front-end component SDK receives the primary key ID and text and performs a secondary compilation on the primary key ID. The key ID and text are compiled to generate corresponding controls. The front-end component SDK sends the received primary key ID to the front-end component generation device to generate corresponding UI elements and renders the UI elements to the front-end intelligent body to generate a corresponding UI interface. The terminal user or external business system interacts with the front-end intelligent body through the corresponding controls in the generated UI interface; for example, the user or the external business system enters the question text "operate into the warehouse", and the generated UI interface displays the corresponding reply "OK, please operate" and the corresponding controls. Click the corresponding controls to perform the operation into the warehouse; for another example, the user or the external business system enters the question text as "take a look at this year's sales", and the generated UI interface will display this year's sales in the form of a chart. Specific embodiment 2:
[0093] The interaction process between the end user, the front-end agent and the back-end AI agent is as follows:
[0094] The terminal user inputs the question text, and the front-end intelligent agent searches the question text and finds that it is not in the Q&A database, so the question text is sent to the back-end AI intelligent agent, and the back-end AI intelligent agent uses the AI big model to think about the question text and get a corresponding reply; the primary key ID in the reply information is sent to the front-end component generation device to generate the corresponding UI element and send it to the front-end component SDK, and the front-end component SDK renders the received UI element to the front-end intelligent agent to generate the corresponding UI interface; in the generated UI interface, the corresponding front-end component SDK is called through the identifier in the reply information to obtain the corresponding block, and the front-end component SDK generates the corresponding block according to the primary key ID and text in the received reply information. The terminal user interacts with the front-end agent through the corresponding controls in the generated UI interface; for example, the question text entered by the terminal user is "What is the weather like in Wuhan on August 10, 2024?", the generated UI interface will evoke the control block and trigger the corresponding controls, and then give the corresponding reply "The weather in Wuhan on August 10, 2024 is as follows: cloudy to sunny during the day, the temperature is between 28℃ and 35℃, and the wind force is southwesterly level 3"; for another example, the question text entered by the terminal user is "Help me summarize today's inventory situation", the generated UI interface will evoke the display block and trigger the corresponding controls, and then display the inventory situation to the terminal user in the form of a chart.
[0095] According to an embodiment of the present disclosure, the front-end intelligent agent receives a message sent by a terminal user or an external business system and determines whether the message matches the Q&A database in the front-end intelligent agent; if the message matches the Q&A database, the front-end intelligent agent obtains corresponding feedback information based on the matching data, and sends the primary key ID in the feedback information to the front-end component generation device through the called corresponding front-end component SDK to generate the corresponding UI element; the front-end component SDK receives the corresponding UI element sent by the front-end component generation device and renders the UI element to the front-end intelligent agent to generate the corresponding UI interface; if the message does not match the Q&A database, the front-end intelligent agent sends the message to the AI big model in the back-end AI intelligent agent, outputs the corresponding feedback information through the AI big model and sends it to the front-end intelligent agent; the front-end intelligent agent performs secondary compilation on the feedback information, and generates the corresponding UI interface according to the generation of the corresponding UI interface operation using the secondary compiled primary key ID; the terminal user interacts with the generated UI interface and triggers the control; the front-end component generation device calls the API of the corresponding external business system according to the triggered control; the external business system executes the corresponding business logic according to the called API and sends the execution result to the front-end intelligent agent; the front-end intelligent agent displays the execution result to the terminal user after secondary compilation or transparent transmission. In this way, the efficiency of the business system can be improved several times, and the user's usage threshold can be greatly reduced. Users only need simple training to use the system, and do not need to learn complicated background operations. The traditional background operation interaction method is transformed into a human-computer dialogue interaction method, which improves the user's interactive experience, reduces AI illusions, and realizes one-stop business processing across multiple systems.
[0096] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present disclosure is not limited by the order of the actions described, because according to the present disclosure, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present disclosure.
[0097] The above is an introduction to the method embodiment. The following is a further explanation of the scheme disclosed in the present invention through an apparatus embodiment.
[0098] Figure 2 The structure diagram of a collaborative multi-agent session processing system provided by an embodiment of the present disclosure is shown. The system 200 includes:
[0099] A front-end component generation device 220, an external business system 230, a back-end AI agent 240 and a terminal user 250 respectively connected to the front-end agent 210; wherein the front-end component generation device 220 is respectively connected to the front-end component SDK 260 in the front-end agent and the external business system 230;
[0100] The front-end intelligent agent 210 is used to receive messages sent by terminal users or external business systems and determine the matching status of the messages with the Q&A database in the front-end intelligent agent; if the message matches the Q&A database, the front-end intelligent agent obtains corresponding feedback information based on the matching data, and sends the primary key ID in the feedback information to the front-end component generation device by calling the corresponding front-end component SDK to generate the corresponding UI element; if the message does not match the Q&A database, the front-end intelligent agent sends the message to the AI big model in the back-end AI intelligent agent; it is also used to perform secondary compilation on the feedback information, and generate the corresponding UI interface according to the primary key ID after secondary compilation according to the operation of generating the corresponding UI interface; it is also used to display the execution result to the terminal user after secondary compilation or transparent transmission.
[0101] In some embodiments, the front-end agent 210 is specifically used to:
[0102] The primary key ID in the feedback information is sent to the front-end component generation device through the corresponding front-end component SDK called to generate the corresponding UI element, including:
[0103] The front-end agent calls the corresponding front-end component SDK according to the identifier in the feedback information, and sends the primary key ID in the feedback information to the called front-end component SDK;
[0104] The front-end component SDK sends the received primary key ID to the front-end component generation device to obtain the UI component and configuration package corresponding to the primary key ID;
[0105] The front-end component generating device compiles the UI component according to the configuration package to generate corresponding UI elements.
[0106] In some embodiments, the front-end agent 210 is further used to:
[0107] The front-end component SDK sends the received primary key ID to the front-end component generation device to obtain the UI component and configuration package corresponding to the primary key ID, including:
[0108] The front-end component generating device receives the primary key ID sent by the front-end component SDK, and searches for the corresponding UI component and configuration information in the pre-configured UI component library according to the received primary key ID, and obtains the UI component and configuration package corresponding to the primary key ID.
[0109] In some embodiments, the front-end agent 210 is further used to:
[0110] Feedback also includes:
[0111] Visual feedback information and logical feedback information; among which,
[0112] The visual feedback information includes feedback operation controls and feedback charts; the visual feedback information is displayed in the generated UI interface;
[0113] The logic feedback information includes a feedback algorithm and a code block; the logic feedback information is used to execute various logic judgment processes to generate corresponding UI interface operations.
[0114] The front-end component generating device 220 is used to receive the primary key ID sent by the front-end component SDK; it is also used to generate corresponding UI elements and send them to the corresponding front-end component SDK; it is also used to call the API of the corresponding external business system according to the triggered control.
[0115] The external business system 230 is used to execute the corresponding business logic according to the called API and send the execution results to the front-end intelligent agent.
[0116] The back-end AI agent 240 is used to receive messages sent by the front-end agent; it is also used to output corresponding feedback information through the AI big model and send it to the front-end agent.
[0117] In some embodiments, the backend AI agent 240 is specifically used to:
[0118] The AI big model is an AI model generated by training the general AI big model using the training knowledge base;
[0119] The training knowledge base includes a function model library and a Q&A database.
[0120] In some embodiments, the backend AI agent 240 is further used to:
[0121] Use the training knowledge base to train the general AI model, including:
[0122] The weights of each knowledge element are trained through question-answer pairs in the Q&A database;
[0123] According to the trained weights of each knowledge element, the general AI model is trained using the corresponding algorithm in the function model library.
[0124] The terminal user 250 is used to interact with the generated UI interface and trigger controls; it is also used to receive the content displayed by the front-end intelligent body after secondary compilation or transparent transmission.
[0125] The front-end component SDK 260 is used to send a primary key ID to the front-end component generation device; it is also used to receive corresponding UI elements sent by the front-end component generation device and render the UI elements to the front-end intelligent body to generate a corresponding UI interface.
[0126] Understandably, Figure 2 Each module / unit in the illustrated system 200 has the function of implementing each step in the method 100 provided in the embodiment of the present disclosure and can achieve its corresponding technical effect, which will not be described in detail here for the sake of brevity.
[0127] Figure 3 A block diagram of an exemplary electronic device capable of implementing an embodiment of the present disclosure is shown. The electronic device 300 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device 300 may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.
[0128] like Figure 3 As shown, the electronic device 300 includes a computing unit 301, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 302 or a computer program loaded from a storage unit 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 can also be stored. The computing unit 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An I / O interface 305 is also connected to the bus 304.
[0129] A number of components in the electronic device 300 are connected to the I / O interface 305, including: an input unit 306, such as a keyboard, a mouse, etc.; an output unit 307, such as various types of displays, speakers, etc.; a storage unit 308, such as a disk, an optical disk, etc.; and a communication unit 309, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 309 allows the electronic device 300 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0130] The computing unit 301 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 301 performs the various methods and processes described above, such as method 100. For example, in some embodiments, the method 100 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 308. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 300 via the ROM 302 and / or the communication unit 309. When the computer program is loaded into the RAM 303 and executed by the computing unit 301, one or more steps of the method 100 described above may be performed. Alternatively, in other embodiments, the computing unit 301 may be configured to perform the method 100 in any other appropriate manner (e.g., by means of firmware).
[0131] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0132] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0133] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0134] It should be noted that the present disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute method 100 and achieve the corresponding technical effect achieved by executing the method in the embodiment of the present disclosure. For the sake of concise description, they are not repeated here.
[0135] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).
[0136] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.
[0137] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.
[0138] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.
[0139] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.
Claims
1. A collaborative multi-agent conversation processing method, characterized in that: include: The front-end agent receives the message sent by the terminal user or the external business system and determines the matching situation between the message and the Q&A database in the front-end agent; If the message matches the Q&A database, the front-end agent obtains the corresponding feedback information according to the matching data, and sends the primary key ID in the feedback information to the front-end component generation device to generate the corresponding UI element by calling the corresponding front-end component SDK; The front-end component SDK receives the corresponding UI elements sent by the front-end component generation device and renders the UI elements to the front-end intelligent body to generate a corresponding UI interface; wherein the front-end component SDK is integrated in the front-end intelligent body; If the message does not match the Q&A database, the front-end agent sends the message to the AI big model in the back-end AI agent, and the AI big model outputs corresponding feedback information and sends it to the front-end agent; the front-end agent recompiles the feedback information, and generates the corresponding UI interface according to the operation of generating the corresponding UI interface using the primary key ID after the recompile; The terminal user interacts with the generated UI interface and triggers the control; the front-end component generation device calls the API of the corresponding external business system according to the triggered control; the external business system executes the corresponding business logic according to the called API and sends the execution result to the front-end intelligent agent; the front-end intelligent agent displays the execution result to the terminal user after secondary compilation or transparent transmission.
2. The method according to claim 1, characterized in that The sending of the primary key ID in the feedback information to the front-end component generating device by calling the corresponding front-end component SDK to generate the corresponding UI element includes: The front-end agent calls the corresponding front-end component SDK according to the identifier in the feedback information, and sends the primary key ID in the feedback information to the called front-end component SDK; The front-end component SDK sends the received primary key ID to the front-end component generation device to obtain the UI component and configuration package corresponding to the primary key ID; The front-end component generating device compiles the UI component according to the configuration package to generate corresponding UI elements.
3. The method according to claim 2, characterized in that The front-end component SDK sends the received primary key ID to the front-end component generation device to obtain the UI component and configuration package corresponding to the primary key ID, including: The front-end component generating device receives the primary key ID sent by the front-end component SDK, and searches for the corresponding UI component and configuration information in the pre-configured UI component library according to the received primary key ID, and obtains the UI component and configuration package corresponding to the primary key ID.
4. The method according to claim 2, characterized in that: The feedback information also includes: Visual feedback information and logical feedback information; among which, The visual feedback information includes feedback operation controls and feedback charts; the visual feedback information is displayed in the generated UI interface; The logic feedback information includes a feedback algorithm and a code block; the logic feedback information is used to execute various logic judgment processes to generate corresponding UI interface operations.
5. The method according to claim 1, characterized in that The AI big model is an AI model generated by training the general AI big model using the training knowledge base; wherein, The training knowledge base includes a function model library and a Q&A database.
6. The method according to claim 5, characterized in that The use of the training knowledge base to train the general AI large model includes: The weights of each knowledge element are trained through question-answer pairs in the Q&A database; According to the trained weights of each knowledge element, the general AI model is trained using the corresponding algorithm in the function model library.
7. A collaborative multi-agent conversation processing system, characterized in that: include: A front-end component generating device, an external business system, a back-end AI agent and a terminal user connected to the front-end agent respectively; wherein the front-end component generating device 220 is connected to the front-end component SDK 260 in the front-end agent and the external business system 230 respectively; The front-end intelligent agent is used to receive messages sent by terminal users or external business systems and determine the matching status of the messages with the Q&A database in the front-end intelligent agent; if the message matches the Q&A database, the front-end intelligent agent obtains corresponding feedback information according to the matching data, and sends the primary key ID in the feedback information to the front-end component generation device by calling the corresponding front-end component SDK to generate the corresponding UI element; if the message does not match the Q&A database, the front-end intelligent agent sends the message to the AI big model in the back-end AI intelligent agent; it is also used to perform secondary compilation on the feedback information, and generate the corresponding UI interface according to the generation of the corresponding UI interface operation using the secondary compiled primary key ID; it is also used to display the execution result to the terminal user after secondary compilation or transparent transmission; The front-end component generating device is used to receive the primary key ID sent by the front-end component SDK; it is also used to generate the corresponding UI element and send it to the corresponding front-end component SDK; it is also used to call the API of the corresponding external business system according to the triggered control; The front-end component SDK is used to send the primary key ID to the front-end component generation device; it is also used to receive the corresponding UI elements sent by the front-end component generation device and render the UI elements to the front-end intelligent body to generate the corresponding UI interface; The back-end AI agent is used to receive messages sent by the front-end agent; it is also used to output corresponding feedback information through the AI big model and send it to the front-end agent; The external business system is used to execute the corresponding business logic according to the called API and send the execution results to the front-end agent; It is used by end users to interact with the generated UI interface and trigger controls; it is also used to receive content displayed by the front-end intelligent agent after secondary compilation or transparent transmission.
8. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 6.
9. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-6.
Citation Information
Cited By
Construction method of message processing system and message processing method and device
CN120567814A
Method for constructing message processing system, message processing method and device
CN120567814B
Tourism AI assistant system and method based on multi-agent cooperation
CN120994790A
Multi-modal data visual display method and system based on large model
CN121030007A
Multi-agent cooperation and dynamic UI interaction method based on streaming protocol
CN122346347A
