Dialogue content display method based on multi-agent cooperation and related device

Abstract

The application provides a dialogue content display method based on multi-agent cooperation and related devices, the method comprises the following steps: in response to a user's questioning operation, sending the questioning data to the server to obtain a target sub-agent and thinking content, the target sub-agent is the sub-agent with the highest priority when the server generates a reply content; according to the thinking content display rule matched with the function of the target sub-agent, display the thinking content; when receiving the reply content returned by the server, according to the reply content display rule matched with the function of the target sub-agent, display the reply content. In this way, the output characteristics of the sub-agents with different functions are designed to be different, the adaptation degree between the presentation form of the output result and the user demand is enhanced, the user can understand the reply content more easily, and the user experience is optimized.

Description

Technical Field

[0001] This application belongs to the field of artificial intelligence technology, specifically relating to a method and related apparatus for displaying dialogue content based on multi-agent collaboration. Background Technology

[0002] With the rapid development of artificial intelligence technology, agent-based dialogue systems have been widely applied in various fields due to their efficient interactive response capabilities and professional content processing capabilities. To address the diverse user needs in complex scenarios, a multi-agent collaborative technical architecture is gradually becoming mainstream, especially in the financial sector, where it can solve domain-specific challenges posed by massive heterogeneous data and complex terminology.

[0003] However, existing dialogue systems based on multi-agent collaboration use a uniform display strategy for both the thought process and the response content when outputting results. They do not design differentiated display rules for the output characteristics of sub-agents with different functions, resulting in poor adaptability between the presentation of output results and user needs, and the information transmission is not intuitive enough, which affects the user experience. Summary of the Invention

[0004] This application provides a method and related apparatus for displaying dialogue content based on multi-agent collaboration, aiming to design differentiated display rules for the output characteristics of sub-agents with different functions, enhance the adaptability between the presentation of output results and user needs, make it easier for users to understand the response content, and optimize the user experience.

[0005] In a first aspect, embodiments of this application provide a method for displaying dialogue content based on multi-agent collaboration, applied to an electronic device. The electronic device is communicatively connected to a server, the server comprising a parent agent and multiple child agents. The method includes:

[0006] In response to a user's question, the system sends question data to the server to obtain a target sub-agent and thought content. The target sub-agent is the sub-agent with the highest priority when the server generates the response content. The thought content includes the task decision obtained by the server after calling the parent agent to process the question data.

[0007] The thought content is displayed according to the thought content display rules that match the function of the target sub-agent;

[0008] When a response is received from the server, the response is displayed according to the response display rules that match the function of the target sub-agent.

[0009] Secondly, embodiments of this application provide a dialogue content display device based on multi-agent collaboration, applied to an electronic device, wherein the electronic device is communicatively connected to a server, the server including a parent agent and multiple child agents, and the device includes:

[0010] The response unit is used to respond to the user's question operation, send question data to the server, obtain the target sub-intelligent agent and the thinking content. The target sub-intelligent agent is the sub-intelligent agent with the highest priority when the server generates the response content. The thinking content includes the task decision obtained by the server after calling the parent agent to process the question data.

[0011] The first display unit is used to display the thinking content according to the thinking content display rules that match the function of the target sub-intelligent agent;

[0012] The second display unit is used to display the response content according to the response content display rules that match the function of the target sub-agent when the response content returned by the server is received.

[0013] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and one or more programs, the one or more programs being stored in the memory and configured to be executed by the processor, the programs including instructions for performing steps in the method as described in the first aspect of this application.

[0014] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program or instructions stored thereon, wherein the computer program or instructions, when executed by a processor, implement the steps of the method described in the first aspect of this application.

[0015] As can be seen from this embodiment, the electronic device responds to the user's question by sending question data to the server, obtaining a target sub-agent and thought content. The target sub-agent is the highest-priority sub-agent when the server generates response content. The thought content includes the task decision obtained by the server after calling the parent agent to process the question data. Then, the thought content is displayed according to the thought content display rules corresponding to the target sub-agent. When the server returns response content, the response content is displayed according to the response content display rules corresponding to the target sub-agent. In this way, differentiated display rules are designed for the output characteristics of sub-agents with different functions, enhancing the adaptability between the presentation of output results and user needs, making it easier for users to understand the response content and optimizing the user experience. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0017] Figure 1 This is a structural block diagram of a dialogue system based on multi-agent collaboration provided in an embodiment of this application;

[0018] Figure 2 This is a flowchart illustrating a dialogue content display method based on multi-agent collaboration provided in an embodiment of this application;

[0019] Figure 3 This is a simplified example diagram of a dialog interface provided in an embodiment of this application;

[0020] Figure 4 This is a simplified example diagram of a first window provided in an embodiment of this application;

[0021] Figure 5 This is a simplified example diagram of a comprehensive dialog window provided in an embodiment of this application;

[0022] Figure 6 This is a simplified example diagram of the first type of second window provided in the embodiments of this application;

[0023] Figure 7 This is a simplified example diagram of the second type of second window provided in the embodiments of this application;

[0024] Figure 8 This is a simplified example diagram of the third type of second window provided in the embodiments of this application;

[0025] Figure 9 This is a simplified example diagram of the first type of third window provided in the embodiments of this application;

[0026] Figure 10 This is a simplified example diagram of the second type of third window provided in the embodiments of this application;

[0027] Figure 11 This is a simplified example diagram of the third type of third window provided in the embodiments of this application;

[0028] Figure 12 This is a structural block diagram of a dialogue content display device based on multi-agent collaboration provided in an embodiment of this application;

[0029] Figure 13 This is a structural block diagram of another dialogue content display device based on multi-agent collaboration provided in an embodiment of this application;

[0030] Figure 14 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0032] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0033] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0034] Among related technologies, agent-based intelligent dialogue systems are more suitable for solving question-and-answer needs in specific domains compared to general large models. Due to the limitations of single agents in terms of functional boundaries and knowledge reserves, multi-agent collaborative dialogue systems are gradually becoming the mainstream solution for complex scenarios, especially in the financial field, where they can solve domain-specific problems caused by massive heterogeneous data and complex terminology to a certain extent. However, existing dialogue systems still use a uniform display strategy when outputting results, such as displaying them all in plain text. However, the output characteristics of different sub-agents trained based on different needs types vary significantly. Especially in the financial field, dialogue scenarios typically include data retrieval, document analysis, and knowledge question-and-answer needs. A uniform display strategy cannot adapt to these characteristics, making it difficult for users to understand the responses and affecting the user experience.

[0035] To address the aforementioned issues, embodiments of this application provide a method and related apparatus for displaying dialogue content based on multi-agent collaboration.

[0036] The following describes a dialogue system based on multi-agent collaboration provided by an embodiment of this application.

[0037] Please see Figure 1 , Figure 1 This is a structural block diagram of a dialogue system based on multi-agent collaboration provided in an embodiment of this application. Figure 1 As shown, the dialogue system 10 based on multi-agent collaboration includes an electronic device 11 and a server 12. The server 12 includes a parent agent and multiple child agents, and the electronic device 11 is communicatively connected to the server 12. The child agents can be implemented as data retrieval agents, document analysis agents, and knowledge question-answering agents. The data retrieval agent retrieves target fields from target data sources, the document analysis agent reads and analyzes the content of target documents to extract viewpoints or conclusions, and the knowledge question-answering agent searches for explanations or associations of target terms from a knowledge graph. When a user asks a question on the electronic device 11, the electronic device 11, based on data interaction with the server 12, executes the dialogue content display method based on multi-agent collaboration as described in this application embodiment. The electronic device 11 refers to a smart device used by the user, such as a mobile phone, tablet computer, or personal computer. The server 12 can be a single server, a server cluster consisting of several servers, or a cloud computing service center; no specific limitation is made here.

[0038] The following describes a method for displaying dialogue content based on multi-agent collaboration, provided by an embodiment of this application.

[0039] Please see Figure 2 , Figure 2 This is a flowchart illustrating a dialogue content display method based on multi-agent collaboration provided in an embodiment of this application, applicable to, for example... Figure 1 In the electronic device 11 shown, such as Figure 2 As shown, the method includes:

[0040] S201, in response to the user's question, send question data to the server to obtain the target sub-agent and the thought content.

[0041] The target sub-agent is the highest priority sub-agent when the server generates response content. The priority includes resource allocation priority and rule restriction priority. That is, the target sub-agent that matches the user's needs can be allocated more resources and stricter rules to perform data processing tasks, thereby improving the exploration depth of the target sub-agent and ensuring the reliability of its output content.

[0042] The thought process includes the task decision obtained after the server calls the parent agent to process the query data. The task decision is used to characterize the data processing task assigned by the parent agent to the child agent after completing the initial processing.

[0043] The target sub-agent can be selected by the user. If the user selects a target sub-agent, the electronic device determines the target sub-agent based on the user's selection. If the user does not select a target sub-agent but performs a question operation, the server calls the parent agent to process the question data, quantify the user's needs, and determine the target sub-agent that best matches the user's needs. Specifically, for example... Figure 3 As shown, the dialogue interface 30 includes an agent selection area 31, a question input area 32, and a submit function control 33. The agent selection area 31 includes multiple agent labels. In this example, the multiple agent labels include a first agent label 311, a second agent label 312, and a third agent label 313. The first agent label 311 indicates a data retrieval agent, the second agent label 312 indicates a document analysis agent, and the third agent label 313 indicates a knowledge question answering agent. In this example, the user's questioning operation includes question entry in the question input area 32 and clicking the submit function control 33. If the user has not clicked any agent label before clicking the submit function control 33, the question data is sent to the server, which determines the target sub-agent and returns it to the electronic device. If the user has clicked an agent label before clicking the submit function control 33, the electronic device responds to the click operation and determines that the agent indicated by that agent label is the target sub-agent. Optionally, after responding to the selection operation, the corresponding agent label can be displayed differently to mark it as selected.

[0044] S202, Display the thought content according to the thought content display rules that match the function of the target sub-agent.

[0045] Displaying the thought content on the client side allows users to clearly understand the thought logic and task decisions of the dialogue system in the current response process. Displaying the thought content through rules that match the function of the target sub-agent makes it easier for users to grasp the key information in this question-and-answer interaction, further understand the thought logic and task decisions, and help reduce the difficulty for users to understand the response content.

[0046] Understandably, when the server returns the thought content, it will simultaneously call the sub-agent to continue executing the data processing task according to the displayed task decision until the response content is obtained.

[0047] S203, when the response content returned by the server is received, the response content is displayed according to the response content display rules that match the function of the target sub-agent.

[0048] The response content is the result of the server calling the parent agent to fuse the output content of all sub-agents after obtaining the output content of all sub-agents. The client side displays the response content according to the rules that match the function of the target sub-agent, presenting the response content in a form that is more in line with the user's needs, thus reducing the difficulty for the user to understand the response content.

[0049] As can be seen in this example, the electronic device responds to the user's question by sending question data to the server, obtaining the target sub-agent and its thought content. The target sub-agent is the highest-priority sub-agent when the server generates the response content. The thought content includes the task decision obtained by the server after calling the parent agent to process the question data. Then, the thought content is displayed according to the display rules corresponding to the target sub-agent. When the server returns the response content, the response content is displayed according to the display rules corresponding to the target sub-agent. In this way, differentiated display rules are designed for the output characteristics of sub-agents with different functions, enhancing the adaptability between the presentation of the output results and user needs, making it easier for users to understand the response content and optimizing the user experience.

[0050] In one possible example, displaying the thought content according to the thought content display rules matching the function of the target sub-agent includes: displaying a first window, the first window including a first display area, a second display area, and a third display area; in the first display area, displaying the question data and highlighting keywords in the question data that match the function of the target sub-agent; in the second display area, displaying a first priority label, the first priority label being used to indicate that the server prioritizes calling the target sub-agent to generate response content; and in the third display area, displaying a first task decision, the target task decision being used to indicate the data processing task assigned by the server to the first sub-agent.

[0051] Among them, such as Figure 4As shown, after determining the target sub-agent and the thought content, a first window 40 is displayed. The first window 40 includes a first display area 41, a second display area 42, and a third display area 43. The question data is displayed in the first display area 41, and keywords in the question data that match the function of the target sub-agent are highlighted. For example, the question data displayed in the first display area 41 is: "I want to check the profitability of Company A in the third quarter of 2024." The keywords matching the function of the data retrieval agent include: Company A, third quarter of 2024, profitability. These keywords are highlighted, for example, by bolding. A first priority label 421 is displayed in the second display area 42, indicating that the server prioritizes calling the target sub-agent to generate the response content. Optionally, a second priority label 422 is also included, indicating that the server synchronously calls other sub-agents to generate the response content. For example, the first priority label 421 is: Prioritize calling the [Data Retrieval Agent], and the second priority label 422 is: Synchronously call the [Knowledge Question Answering Agent]. The third display area 43 displays the first task decision 431, which characterizes the data processing task assigned by the server to the target sub-agent. Optionally, it also includes a second task decision 432, which characterizes the data processing task assigned by the server to other sub-agents. Here, "profitability" is an indicator name that does not exist in the structured database. When the server calls the parent agent for pre-processing, it quantifies "profitability" into indicator fields in the database, such as operating revenue, net profit, and gross profit margin, and then makes task decisions based on the quantified indicator fields. For example, the first task decision 431 is: the search conditions are "Subject = Company A; Time = Third Quarter of 2024; Indicator Name 1 = Operating Revenue; Indicator Name 2 = Net Profit; Indicator Name 3 = Gross Profit Margin", retrieving data from the CSMAR database; the second task decision 432 is: explaining "profitability" and its corresponding core indicators.

[0052] The first window 40 also includes a window state identifier 44, which indicates the window state of the first window 40, which is an expanded state. In this example, the window state is an expanded state. The user can switch the first window 40 to a collapsed state by clicking the window state identifier 44.

[0053] As can be seen in this example, the question data is displayed in the first display area and keywords that match the function of the target sub-agent are highlighted. The first priority label associated with the target sub-agent is displayed in the second display area, and the first task decision associated with the target sub-agent is displayed in the third display area. The thinking logic and decision content in this dialogue are displayed in the first window to help users understand the subsequent output response content and optimize the user experience.

[0054] In one possible example, displaying the response content according to a display rule matching the function of the target sub-agent includes: displaying a second window, the second window including a fourth display area and a fifth display area; displaying a first response content in the fourth display area in a panel format matching the function of the target sub-agent, the first response content indicating the data processing result corresponding to the target sub-agent; and displaying a second response content in the fifth display area, the second response content indicating the data processing result corresponding to other sub-agents, the second response content being supplementary content to the first response content.

[0055] Among them, such as Figure 5 As shown, while displaying the second window 50, the window state of the first window 40 is automatically switched to the collapsed state to free up display space for the reply content. At the same time, the window state indicator 44 of the first window 40 is retained. Users can switch the window state of the first window 40 to the expanded state by clicking the window state indicator 44 to meet the needs of on-demand tracing.

[0056] The second window 50 includes a fourth display area 51 and a fifth display area 52. The fourth display area 51 displays core results, and the fifth display area 52 displays supplementary content. In the fourth display area 51, the first response content, i.e., the data processing results corresponding to the target sub-agent, is displayed in the form of a preset panel 511. The preset panel 511 is functionally matched with the target sub-agent. This functional matching is specifically reflected in the overall layout and information organization of the preset panel being highly compatible with the processing characteristics of the target sub-agent, ensuring that key conclusions, core data, or core viewpoints in the first response content are clearly highlighted, avoiding inefficient information transmission due to a disconnect between the display format and content characteristics. The fifth display area 52 displays the second response content, i.e., the data processing results corresponding to other sub-agents. The second response content is supplementary content to the first response content. The fifth display area 52 is located below the fourth display area 51 to represent the priority relationship between the second response content displayed in the fifth display area 52 and the first response content displayed in the fourth display area 51.

[0057] As can be seen, in this example, by differentiating the functional division and content presentation of the fourth and fifth display areas in the second window, the hierarchical relationship between the core results corresponding to the target sub-agent and the auxiliary content corresponding to other sub-agents is clarified. This solves the problems of single display strategy and mixed main and auxiliary information in the existing technology, enabling users to quickly locate key content, reducing the difficulty for users to understand the response content, and optimizing the user experience.

[0058] In one possible example, the target sub-agent is a data retrieval agent. The step of displaying the first response content in the fourth display area in a form that matches the function of the target sub-agent includes: displaying the first response content in the fourth display area in the form of a structured data table. The structured data table includes an indicator name column, a value column, and a data source column. The indicator name column is used to display the indicator name corresponding to the target data. The value column is used to display the specific value of the indicator. The data source column is used to display the target data source.

[0059] In the scenario where the target sub-agent is a data retrieval agent, the user's needs focus on obtaining the values ​​of indicator fields in the underlying database. In this example, the first response content retrieved is displayed in the fourth display area in the form of a structured data table, matching the core functional characteristics of the data retrieval agent.

[0060] For example, such as Figure 6 As shown, a structured data table 53 is displayed in the fourth display area 51. The structured data table 53 includes an indicator name column 531, a value column 532, and a data source column 533. The indicator name column 531 vertically arranges all indicator names extracted by the data retrieval agent from the target data source that directly match the user's needs. For example, if the query is "I want to check the profitability of Company A in the third quarter of 2024," based on the parent agent's preprocessing, "profitability" is quantified into indicator fields in the structured database: operating revenue, net profit, and gross profit margin. These indicator fields are then displayed sequentially in the indicator name column 531. The value column 532 corresponds one-to-one with the indicator name column 531, displaying the specific values ​​of each indicator field. For example, if the data retrieval agent retrieves Company A's financial statements for the third quarter of 2024 and obtains operating revenue of 860 million yuan, net profit of 120 million yuan, and a gross profit margin of 32.5%, these specific values ​​are displayed in the value column 532. In this section, column 533 displays the target data source in text format, visualizing the traceability of the data. For example, the revenue figure comes from page 12 of Company A's Q3 2024 financial statement, the net profit figure comes from page 15 of Company A's Q3 2024 financial statement, and the gross profit margin figure comes from page 18 of Company A's Q3 2024 financial statement. Key indicator fields can be highlighted using bolding, color blocking, or other differentiated styles to help users quickly locate core data.

[0061] In this example, based on the first response displayed in the fourth display area 51, the second response, i.e., the data processing result corresponding to the knowledge-answering agent, is displayed in the fifth display area 52. For example, in this example, the fifth display area 52 is used to display the interpretation of the term "profitability" in the question data, such as: Profitability is a company's ability to generate profits and realize capital appreciation through operating activities. Core indicators for measuring a company's profitability generally include: operating revenue, net profit, and gross profit margin.

[0062] As can be seen, in this example, the content of the first response is presented in the form of a structured data table, which accurately matches the core functions of the intelligent retrieval agent, forms a clear correspondence, makes the presentation of multi-indicator information more intuitive, meets the precise and efficient needs of data retrieval scenarios, and optimizes the user experience.

[0063] In one possible example, the target sub-agent is a document parsing agent. The step of displaying the first response content in the fourth display area in a form that matches the function of the target sub-agent includes: displaying the first response content in the fourth display area in the form of a two-column comparison panel, wherein the two-column comparison panel includes a first display panel and a second display panel, wherein the first display panel is used to display the analysis conclusion and the second display panel is used to display the original text fragments associated with the analysis conclusion.

[0064] In the scenario where the target sub-agent is a document parsing agent, the user's needs focus on obtaining the viewpoints or conclusions of the original document excerpts. In this example, the first response content is displayed in the fourth display area in the form of a two-column comparison panel, matching the core functional characteristics of the document parsing agent.

[0065] For example, a user uploads the file "Technical White Paper for Product X" and simultaneously enters the question "Please read the uploaded file and analyze the core functional modules of Product X." The server determines that the target sub-agent is a document analysis agent, and then presents the first response in a two-column comparison panel in the fourth display area. For example... Figure 7 As shown, a double-column comparison panel 54 is displayed in the fourth display area 51. The double-column comparison panel 54 includes a first display panel 541 located on the left and a second display panel 542 located on the right.

[0066] The first display panel 541 displays the analysis conclusions output by the document analysis agent, arranged in a structured manner according to logical hierarchy, with core viewpoints and sub-arguments, and the content corresponding to the core viewpoints is displayed in bold. The content corresponding to the core viewpoints is displayed in bold at the top, for example, "Core functional modules: The core functions of product X focus on two major modules: portable deployment management and intelligent operation and maintenance optimization, covering the entire lifecycle needs of equipment deployment and problem troubleshooting." The content corresponding to the sub-arguments is displayed in segments with preset interval indentation, for example, the first segment is "Module 1: Portable deployment management supports the same configuration and batch operation in the cloud, without the need for on-site debugging of each device," and the second segment is "Module 2: Intelligent operation and maintenance optimization has the capabilities of scenario-based RF configuration, automatic fault identification, and progressive optimization."

[0067] The second display panel 542 is used to display original text fragments related to the above analysis conclusions, and each original text fragment ends with a positional citation mark, which is used to indicate the specific location of the original text fragment in the original document. For example, the corresponding content of the core functional module is: Product X is built around the core needs of enterprise wireless network construction, and two core functional modules are constructed: a portable deployment management module to ensure rapid network setup and an intelligent operation and maintenance optimization module to improve network stability, jointly covering the entire life cycle management scenario of the device - Chapter 3, P4, Paragraph 1 of the "Product X Technical White Paper"; the corresponding content of sub-module 1 is: After Product X is added to the site, it can automatically synchronize configuration from the cloud. Administrators do not need to log in to operate on a single device. They can complete the Internet access configuration in batches through the APP tool, which greatly simplifies the deployment process - Chapter 3, P5, Paragraph 2 of the "Product X Technical White Paper"; the corresponding content of sub-module 2 is: Intelligent operation and maintenance optimization provides four radio frequency adjustment methods: scenario-based configuration, fine configuration, intelligent configuration, and progressive optimization. At the same time, it supports automatic identification of various network problems and realizes rapid fault location and repair - Chapter 5, P12, Paragraph 3 of the "Product X Technical White Paper".

[0068] In this example, based on the first response content displayed in the fourth display area 51, the second response content, i.e., the data processing result corresponding to the knowledge question-answering agent, is displayed in the fifth display area 52. For example, in this example, the fifth display area 52 is used to display the overall framework of the "X Product Technical White Paper" in the form of a mind map. For instance, the "X Product Technical White Paper" is first divided into chapters, including Chapter 1 Overview, Chapter 2 Technical Architecture, and Chapter 3 Core Functional Modules. Then, the specific content of each chapter is divided. For example, Chapter 3 Core Functional Modules can be further divided into "3.1 Portable Deployment Management Module" and "3.2 Intelligent Operation and Maintenance Optimization Module" based on sub-section themes, thereby helping users quickly understand the content of the target document.

[0069] As can be seen, in this example, the first response is presented in the form of a two-column comparison panel, which accurately matches the core functions of the document analysis agent. By binding the analysis conclusions with related original text fragments, the conclusions and evidence are visualized and traced, which enhances the credibility of the analysis results, deeply meets the usage needs of professional document analysis scenarios, and optimizes the user experience.

[0070] In one possible example, the target sub-agent is a knowledge-based question-answering agent. The step of displaying the first response content in the fourth display area in a form matching the function of the target sub-agent includes: displaying the first response content in the form of a knowledge card in the fourth display area. The knowledge card includes a core explanation area and a related knowledge area. The core explanation area displays explanatory text for the target term involved in the question data, and the related knowledge area displays a related knowledge graph, which includes other terms associated with the target term.

[0071] In the scenario where the target sub-agent is a knowledge-based question-answering agent, the user's needs focus on terminology explanation or knowledge-based question-answering. In this example, the first answer is displayed in the form of a knowledge card in the fourth display area, matching the core functional characteristics of the knowledge-based question-answering agent.

[0072] For example, if a user inputs the question "What does non-GAAP net profit mean?", and the server determines that the target sub-agent is a knowledge-based question-answering agent, then the first answer will be displayed in the fourth display area in the form of a knowledge card. Figure 8 As shown, a knowledge card 55 is displayed in the fourth display area 51. The knowledge card 55 includes a core explanation area 551 and an associated knowledge area 552.

[0073] The core explanation area 551 displays explanatory text for the target terms involved in the query data. For example, "net profit excluding non-recurring gains and losses" refers to the profit amount obtained by a company when calculating net profit after deducting gains and losses that are not directly related to its daily operations and are not sustainable. It is a core indicator reflecting a company's sustainable profitability. The name of the target term "net profit excluding non-recurring gains and losses" is displayed in bold, large font. Key expressions in the explanation text such as "not directly related to its daily operations and not sustainable" and "sustainable profitability" are also highlighted in bold to help users quickly grasp the core points.

[0074] The associated knowledge area 552 is used to display an associated knowledge graph, which includes other terms associated with the target term. In this example, the central node of the associated knowledge graph is net profit excluding non-recurring items, surrounded by several associated nodes, such as net profit and non-recurring gains and losses, each connected to the central node. Each associated term is labeled with a brief association logic, such as "Net Profit: The basis for calculating net profit excluding non-recurring items, including non-recurring gains and losses," or "Non-recurring gains and losses: The core part of net profit excluding non-recurring items that needs to be excluded." Furthermore, the associated nodes are configured as interactive controls. When a user clicks on any associated term node, a pop-up window displays a dedicated knowledge card for that associated term, helping the user efficiently acquire more knowledge information.

[0075] In this example, based on the first response content displayed in the fourth display area 51, the second response content, i.e., the data processing result corresponding to the data retrieval agent, is displayed in the fifth display area 52. For example, in this example, the fifth display area 52 is used to display two core indicators directly related to non-recurring net profit extracted by the data retrieval agent from Company A's publicly disclosed financial report for the third quarter of 2024: net profit, specifically RMB 120 million, and non-recurring gains and losses, specifically RMB 30 million. It can be seen that Company A's non-recurring net profit for the third quarter of 2024 was RMB 90 million.

[0076] As can be seen, in this example, the first answer is presented in the form of a knowledge card, which accurately matches the core function of the knowledge question-answering agent. Through the presentation structure of the core explanation area and the related knowledge area, it not only meets the user's core need for terminology explanation, but also extends related knowledge through the related knowledge graph, which deeply fits the usage needs of the knowledge question-answering scenario and optimizes the user experience.

[0077] In one possible example, after displaying the response content according to the response content display rules matching the function of the target sub-agent, the method further includes: displaying a third window, the third window including modification function controls matching the function of the target sub-agent; wherein, when the target sub-agent is a data retrieval agent, the modification function controls include a retrieval condition modification control, which is used to adjust the data retrieval scope of the data retrieval agent; when the target sub-agent is a document parsing agent, the modification function controls include a parsing condition modification control, which is used to adjust the range of document paragraphs analyzed by the document parsing agent; when the target sub-agent is a knowledge question answering agent, the modification function controls include a knowledge dimension modification control, which is used to adjust the terminology explanation granularity or the scope of related knowledge expansion of the knowledge question answering agent.

[0078] In this example, after displaying the response content according to the rules that match the function of the target sub-agent, a third window is used as the response adjustment interaction area to meet the user's secondary need to further optimize the results based on the initial response. The window has built-in modification function controls, allowing users to quickly adjust the search, parsing or question-and-answer conditions without having to re-enter the complete question, thereby improving the interaction efficiency and response accuracy.

[0079] When the target sub-agent is a data retrieval agent, the modification function controls include search condition modification controls to allow users to adjust the data retrieval scope of the data retrieval agent. For example, if the user's initial query is "I want to check the profitability of Company A in the third quarter of 2024," and the data retrieval agent outputs a structured table containing operating revenue, net profit, and gross profit margin in the second window, and the user wants to expand the search time range to the second half of 2024 and add a non-GAAP net profit indicator, this requirement can be met through the search condition modification controls in the third window.

[0080] For example, such as Figure 9 As shown, the third window 90 includes a time modification area 91, an indicator modification area 92, and a re-search function control 93. The time modification area 91 includes a time range display box 911 and a first modification function control 912. The time range display box 911 defaults to displaying "July 1, 2024 to September 30, 2024 (Third Quarter of 2024)". Users can click the first modification function control 912 to switch the time range display box 911 to an interactive state and then modify it to "July 1, 2024 to December 31, 2024". The indicator modification area 92 displays multiple selected indicators, including operating revenue, net profit, and gross profit margin. Each indicator has a corresponding indicator deletion function control 921, which users can click to delete indicators they don't need to keep. The indicator modification area 92 also includes an indicator addition function control 922, which users can click to select non-GAAP net profit to add an indicator. After the user completes the above adjustments, click the re-retrieval function control 93, and the server will execute a new data processing task, and the front-end display interface will be updated in real time.

[0081] When the target sub-agent is a document parsing agent, the modification function controls include parsing condition modification controls to allow users to adjust the range of document paragraphs analyzed by the document parsing agent. For example, if the user's initial question is to analyze the core functional modules of product X based on the "Technical White Paper of Product X," and after the document parsing agent outputs the first response in the second window, the user wants to focus on the content of Chapter 5 of the white paper, this requirement can be met through the parsing condition modification controls in the third window.

[0082] For example, such as Figure 10 As shown, the third window 90 includes a parsing condition display box 94, a second modification function control 95, and a re-parsing function control 96. The parsing condition display box 94 defaults to displaying the original parsing range as "Full Text." Users can click the second modification function control 95 to switch the parsing condition display box 94 to an interactive state and then change it to "Chapter 5." After the user completes the above adjustments, clicking the re-parsing function control 96 will cause the server to execute a new data processing task, and the front-end display interface will update in real time.

[0083] When the target sub-agent is a knowledge-based question-answering agent, the modification controls include knowledge dimension modification controls to allow users to adjust the terminology explanation level and the scope of related knowledge expansion. For example, if the user's initial question is "What does non-GAAP net profit mean?", and the knowledge-based question-answering agent outputs a knowledge card containing the core explanation and related graphs in the second window, and the user requests "a more accessible explanation and to clarify the difference between this and net profit attributable to the parent company," this request can be met through the knowledge dimension modification controls in the third window.

[0084] For example, such as Figure 11 As shown, the third window 90 includes an explanation granularity modification area 97, a related knowledge expansion scope modification area 98, and a re-answer function control 99. The explanation granularity modification area 97 includes an explanation granularity display box 971 and a third modification function control 972. The explanation granularity display box 971 defaults to displaying the original explanation granularity as "professional." Users can click the third modification function control 972 to switch the explanation granularity display box 971 to an interactive state and then change it to "common." The related knowledge expansion scope modification area 98 includes a related knowledge expansion scope display box 981 and a fourth modification function control 982. The related knowledge expansion scope display box 981 defaults to displaying the original related indicator expansion scope as "basic related." Users can click the fourth modification function control 982 to switch the related knowledge expansion scope display box 981 to an interactive state and change it to "comparative," while simultaneously entering the name of the indicator to be compared, "net profit attributable to parent company." After the user completes the above adjustments, clicking the re-answer function control 99 will cause the server to execute a new data processing task, and the front-end display interface will update in real time.

[0085] As can be seen, in this example, the interactive entry point for quickly adjusting conditions is provided through a third window, which solves the rigid problem of traditional intelligent agents terminating after a single response. This not only improves the adaptability of the intelligent agent but also strengthens the user's control over the response content, perfectly matching the user-centric interaction design logic and optimizing the user experience.

[0086] For examples consistent with the above embodiments, please refer to... Figure 12 , Figure 12 This is a structural block diagram of a dialogue content display device based on multi-agent collaboration provided in an embodiment of this application. The dialogue content display device 1200 based on multi-agent collaboration includes: a response unit 1201, used to respond to a user's question operation, send question data to the server, obtain a target sub-agent and thought content, wherein the target sub-agent is the sub-agent with the highest priority when the server generates response content, and the thought content includes the task decision obtained by the server calling the parent agent to process the question data; a first display unit 1202, used to display the thought content according to the thought content display rules that match the function of the target sub-agent; and a second display unit 1203, used to display the response content according to the response content display rules that match the function of the target sub-agent when the server returns response content.

[0087] In one possible example, regarding the display of the thought content according to the display rules matching the function of the target sub-agent, the first display unit 1202 is specifically configured to: display a first window, the first window including a first display area, a second display area, and a third display area; in the first display area, display the question data and highlight keywords in the question data that match the function of the target sub-agent; in the second display area, display a first priority label, the first priority label being used to indicate that the server prioritizes calling the target sub-agent to generate response content; and in the third display area, display a first task decision, the first task decision being used to indicate the data processing task assigned by the server to the target sub-agent.

[0088] In one possible example, regarding the display of the response content according to the response content display rules matching the function of the target sub-agent, the second display unit 1203 is specifically configured to: display a second window, the second window including a fourth display area and a fifth display area; in the fourth display area, display a first response content in a panel format matching the function of the target sub-agent, the first response content indicating the data processing result corresponding to the target sub-agent; and in the fifth display area, display a second response content, the second response content indicating the data processing result corresponding to other sub-agents, the second response content being supplementary content to the first response content.

[0089] In one possible example, the target sub-agent is a data retrieval agent. Regarding the display of the first response content in the fourth display area in a form matching the function of the target sub-agent, the second display unit 1203 is specifically used to: display the first response content in the fourth display area in the form of a structured data table. The structured data table includes an indicator name column, a value column, and a data source column. The indicator name column displays the indicator name corresponding to the target data, the value column displays the specific value of the indicator, and the data source column displays the target data source.

[0090] In one possible example, the target sub-agent is a document parsing agent. In the fourth display area, the first response content is displayed in a form that matches the function of the target sub-agent. The second display unit 1203 is specifically used to display the first response content in the fourth display area in the form of a two-column comparison panel. The two-column comparison panel includes a first display panel and a second display panel. The first display panel is used to display the analysis conclusion, and the second display panel is used to display the original text fragment associated with the analysis conclusion.

[0091] In one possible example, the target sub-agent is a knowledge-based question-answering agent. The first response content is displayed in the fourth display area in a form matching the function of the target sub-agent. Specifically, the second display unit 1203 is used to: display the first response content in the fourth display area in the form of a knowledge card. The knowledge card includes a core explanation area and a related knowledge area. The core explanation area displays explanatory text for the target term involved in the question data, and the related knowledge area displays a related knowledge graph, which includes other terms associated with the target term.

[0092] In one possible example, after displaying the response content according to the response content display rules matching the function of the target sub-agent, the dialogue content display device 1200 based on multi-agent collaboration is further configured to: display a third window, the third window including modification function controls matching the function of the target sub-agent; wherein, when the target sub-agent is a data retrieval agent, the modification function controls include retrieval condition modification controls, the retrieval condition modification controls being used to adjust the data retrieval range of the data retrieval agent; when the target sub-agent is a document parsing agent, the modification function controls include parsing condition modification controls, the parsing condition modification controls being used to adjust the range of document paragraphs analyzed by the document parsing agent; when the target sub-agent is a knowledge question answering agent, the modification function controls include knowledge dimension modification controls, the knowledge dimension modification controls being used to adjust the terminology explanation granularity or the scope of related knowledge expansion of the knowledge question answering agent.

[0093] It is understood that since the method embodiments and the device embodiments are different presentations of the same technical concept, the content of the method embodiment section in this application should be adapted to the device embodiment section in a synchronous manner, and will not be repeated here.

[0094] When using integrated units, such as Figure 13 As shown, Figure 13 This is a structural block diagram of another dialogue content display device based on multi-agent collaboration provided in this application embodiment. Figure 13 The multi-agent collaborative dialogue content display device 1200 includes a processing module 122 and a communication module 121. The processing module 122 controls and manages the actions of the multi-agent collaborative dialogue content display device, for example, executing the steps of the response unit 1201, the first display unit 1202, and the second display unit 1203, and / or performing other processes of the technology described herein. The communication module 121 supports interaction between the multi-agent collaborative dialogue content display device and other devices. Figure 13 As shown, the dialogue content display device based on multi-agent collaboration may further include a storage module 123, which is used to store the program code and data of the dialogue content display device based on multi-agent collaboration.

[0095] All relevant content in each scenario involved in the above method embodiments can be referenced from the functional descriptions of the corresponding functional modules, and will not be repeated here. The above-mentioned dialogue content display device 1200 based on multi-agent collaboration can execute the above-mentioned... Figure 2 The method shown is a dialogue content display method based on multi-agent collaboration.

[0096] Based on the description of the above method and device embodiments, please refer to... Figure 14 , Figure 14 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 14 The illustrated electronic device includes a memory 1401, a processor 1402, a communication interface 1403, and a bus 1404. The memory 1401, processor 1402, and communication interface 1403 are interconnected via the bus 1404. Specifically, the electronic device may refer to the electronic device 11 in the above embodiment.

[0097] The memory 1401 may be a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM).

[0098] The memory 1401 can store programs. When the program stored in the memory 1401 is executed by the processor 1402, the processor 1402 and the communication interface 1403 are used to execute the various steps of the dialogue content display method based on multi-agent collaboration of this application embodiment.

[0099] The processor 1402 may be a general-purpose central processing unit (CPU), microprocessor, application specific integrated circuit (ASIC), graphics processing unit (GPU), or one or more integrated circuits, used to execute relevant programs to achieve the functions required by the units in the electronic device of this application embodiment, or to execute the dialogue content display method based on multi-agent collaboration of this application method embodiment.

[0100] The processor 1402 can also be an integrated circuit chip with signal processing capabilities. In implementation, each step of the multi-agent collaborative dialogue content display method of this application can be completed by the integrated logic circuitry in the hardware of the processor 1402 or by software instructions. The processor 1402 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in the memory 1401. The processor 1402 reads the information in the memory 1401 and, in conjunction with its hardware, performs the functions required by the units included in the electronic device of this application embodiment, or executes the dialogue content display method based on multi-agent collaboration of this application method embodiment.

[0101] Communication interface 1403 uses transceiver devices, such as, but not limited to, transceivers, to enable communication between electronic devices and other devices or communication networks. For example, data can be acquired through communication interface 1403.

[0102] Bus 1404 may include a pathway for transmitting information between various components of an electronic device (e.g., memory 1401, processor 1402, communication interface 1403).

[0103] It should be noted that, although Figure 14 The illustrated electronic device only shows the memory 1401, processor 1402, and communication interface 1403. However, those skilled in the art should understand that in specific implementations, the electronic device may also include other components necessary for normal operation. Furthermore, depending on specific needs, those skilled in the art should understand that the electronic device may also include hardware components for implementing other additional functions. Moreover, those skilled in the art should understand that the electronic device may only include the components necessary for implementing the embodiments of this application, and may not necessarily include... Figure 14 All the devices shown.

[0104] This application also provides a computer storage medium storing a computer program / instructions thereon, which, when executed by a processor, implements some or all of the steps of any of the methods described in the above method embodiments.

[0105] In the several embodiments provided in this application, it should be understood that the disclosed methods and apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for example, the division of units is merely a logical functional division, and there may be other division methods in actual implementation; for example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between devices or units, and may be electrical, mechanical, or other forms.

[0106] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0107] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented, in whole or in part, as a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a read-only memory, or random access memory, or a magnetic medium, such as a floppy disk, hard disk, magnetic tape, magnetic disk, or an optical medium, such as a digital universal optical disc, or a semiconductor medium, such as a solid-state drive.

[0108] The above description is merely a specific implementation of the embodiments of this application, but the protection scope of the embodiments of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the embodiments of this application should be covered within the protection scope of the embodiments of this application. Therefore, the protection scope of the embodiments of this application should be determined by the protection scope of the claims.

[0109] The device embodiments described above are merely illustrative. The units and modules described as separate components may or may not be physically separate. Furthermore, some or all of the units and modules can be selected to achieve the purpose of this embodiment, depending on actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0110] While this application discloses the above information, it is not limited thereto. Any person skilled in the art can easily conceive of variations or substitutions without departing from the spirit and scope of this application, and can make various alterations and modifications, including combinations of the different functions and implementation steps described above, as well as software and hardware implementation methods, all of which are within the protection scope of this application.

Claims

1. A method for displaying dialogue content based on multi-agent collaboration, characterized in that, Applied to an electronic device that is communicatively connected to a server, the server comprising a parent agent and multiple child agents, the method includes: In response to a user's question, the system sends question data to the server to obtain a target sub-agent and thought content. The target sub-agent is the sub-agent with the highest priority when the server generates the response content. The thought content includes the task decision obtained by the server after calling the parent agent to process the question data. The thought content is displayed according to the thought content display rules that match the function of the target sub-agent; When the server returns a response, the response is displayed according to the response display rules that match the function of the target sub-agent. The step of displaying the response content according to the response content display rules matching the function of the target sub-agent includes: displaying a second window, which includes a fourth display area and a fifth display area; when the target sub-agent is a data retrieval agent, displaying the first response content in the form of a structured data table in the fourth display area, the structured data table including an indicator name column, a value column, and a data source column, the indicator name column being used to display the indicator name corresponding to the target data, the value column being used to display the specific value of the indicator, and the data source column being used to display the target data source; and displaying the second response content in the fifth display area, the second response content including the data processing results corresponding to the knowledge question answering agent; when the target sub-agent is a document parsing agent, displaying the first response content in the form of a two-column comparison panel in the fourth display area. The dual-column comparison panel includes a first display panel and a second display panel. The first display panel displays the analysis conclusion, and the second display panel displays the original text fragment associated with the analysis conclusion. In the fifth display area, a second response is displayed, which includes the data processing results corresponding to the knowledge-based question-answering agent. When the target sub-agent is a knowledge-based question-answering agent, the first response is displayed in the fourth display area in the form of a knowledge card. The knowledge card includes a core explanation area and a related knowledge area. The core explanation area displays the explanation text for the target term involved in the question data, and the related knowledge area displays a related knowledge graph, which includes other terms associated with the target term. In the fifth display area, the second response is displayed, which includes the data processing results corresponding to the data retrieval agent.

2. The method according to claim 1, characterized in that, The step of displaying the thought content according to the thought content display rules that match the function of the target sub-agent includes: Display a first window, which includes a first display area, a second display area, and a third display area; In the first display area, the question data is displayed, and keywords in the question data that match the function of the target sub-agent are highlighted; In the second display area, a first priority label is displayed, which is used to indicate that the server prioritizes calling the target sub-agent to generate response content; In the third display area, a first task decision is displayed, which is used to characterize the data processing task assigned by the server to the target sub-agent.

3. The method according to claim 1, characterized in that, After displaying the response content according to the response content display rules that match the function of the target sub-agent, the method further includes: A third window is displayed, which includes modification function controls that match the functionality of the target sub-agent; wherein... When the target sub-agent is a data retrieval agent, the modification function control includes a retrieval condition modification control, which is used to adjust the data retrieval range of the data retrieval agent; When the target sub-agent is a document parsing agent, the modification function control includes a parsing condition modification control, which is used to adjust the range of document paragraphs analyzed by the document parsing agent; When the target sub-agent is a knowledge-answering agent, the modification function control includes a knowledge dimension modification control, which is used to adjust the terminology explanation granularity or the scope of related knowledge expansion of the knowledge-answering agent.

4. A dialogue content display device based on multi-agent collaboration, characterized in that, Applied to an electronic device, the electronic device is communicatively connected to a server, the server including a parent agent and multiple child agents, the device includes: The response unit is used to respond to the user's question operation, send question data to the server, obtain the target sub-intelligent agent and the thinking content. The target sub-intelligent agent is the sub-intelligent agent with the highest priority when the server generates the response content. The thinking content includes the task decision obtained by the server after calling the parent agent to process the question data. The first display unit is used to display the thinking content according to the thinking content display rules that match the function of the target sub-intelligent agent; The second display unit is used to display the response content according to the response content display rules that match the function of the target sub-agent when the response content returned by the server is received; The step of displaying the response content according to the response content display rules matching the function of the target sub-agent includes: displaying a second window, which includes a fourth display area and a fifth display area; when the target sub-agent is a data retrieval agent, displaying the first response content in the form of a structured data table in the fourth display area, the structured data table including an indicator name column, a value column, and a data source column, the indicator name column being used to display the indicator name corresponding to the target data, the value column being used to display the specific value of the indicator, and the data source column being used to display the target data source; and displaying the second response content in the fifth display area, the second response content including the data processing results corresponding to the knowledge question answering agent; when the target sub-agent is a document parsing agent, displaying the first response content in the form of a two-column comparison panel in the fourth display area. The dual-column comparison panel includes a first display panel and a second display panel. The first display panel displays the analysis conclusion, and the second display panel displays the original text fragment associated with the analysis conclusion. In the fifth display area, a second response is displayed, which includes the data processing results corresponding to the knowledge-based question-answering agent. When the target sub-agent is a knowledge-based question-answering agent, the first response is displayed in the fourth display area in the form of a knowledge card. The knowledge card includes a core explanation area and a related knowledge area. The core explanation area displays the explanation text for the target term involved in the question data, and the related knowledge area displays a related knowledge graph, which includes other terms associated with the target term. In the fifth display area, the second response is displayed, which includes the data processing results corresponding to the data retrieval agent.

5. An electronic device, characterized in that, It includes a processor, a memory, and one or more programs, said one or more programs being stored in the memory and configured to be executed by the processor, said programs including instructions for performing the steps in the method as claimed in any one of claims 1-3.

6. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method according to any one of claims 1-3.

Images