Multi-agent cooperative interaction method and device, electronic equipment and readable storage medium

Abstract

This application discloses a multi-agent cooperative interaction method, apparatus, electronic device, and readable storage medium, relating to the field of near-field communication technology. The multi-agent cooperative interaction method includes: when a user terminal device identifies a target NFC device, obtaining the user's query information through the user terminal device, and considering all NFC devices within the area where the target NFC device is located as candidate NFC devices; based on the query information, determining a main responding agent and at least one auxiliary cooperative agent from the agents associated with each candidate NFC device; invoking the main responding agent to generate a first response based on the query information, and invoking the auxiliary cooperative agent to generate a second response based on the query information; fusing the first and second response contents to generate a cooperative response content, and displaying the cooperative response content on the user terminal device.

Description

Technical Field

[0001] This application relates to the field of near-field communication technology, and in particular to a multi-agent cooperative interaction method, device, electronic device, and readable storage medium. Background Technology

[0002] With the development of near field communication technology, NFC (Near Field Communication) technology has been widely used in various interactive products.

[0003] Currently, mainstream NFC interactive products typically adopt a "one-to-one" interaction mode. When a user touches an NFC device, the system only activates the smart agent associated with that NFC device to respond to the user's question; the user cannot interact with multiple smart agents simultaneously. For example, in a smart museum scenario, when a user touches the NFC device corresponding to an exhibit, the system only invokes the explanatory smart agent associated with that exhibit to provide an answer.

[0004] However, users' actual questions often involve multiple fields or require multifaceted knowledge support. A single intelligent agent, limited by its pre-set knowledge scope, finds it difficult to comprehensively and accurately answer users' complex questions. Summary of the Invention

[0005] The main objective of this application is to provide a multi-agent collaborative interaction method, device, electronic device, and readable storage medium, which aims to enable users to interact with multiple agents simultaneously and improve the accuracy and comprehensiveness of answers to complex questions.

[0006] This application provides a multi-agent cooperative interaction method, the method comprising: When the user terminal device identifies the target NFC device, the user's query information is obtained through the user terminal device, and all NFC devices in the area where the target NFC device is located are regarded as candidate NFC devices. Based on the question information, a main responding agent and at least one auxiliary cooperating agent are determined from the agents associated with each candidate NFC device. The main responding agent is the agent with the highest relevance score to the question information, and the auxiliary cooperating agent is an agent other than the main responding agent whose relevance score to the question information is greater than or equal to a preset score threshold. The main response agent is invoked to generate a first response based on the question information, and the auxiliary collaborative agent is invoked to generate a second response based on the question information. The first response content and the second response content are combined to generate collaborative response content, which is then displayed on the user terminal device.

[0007] In one embodiment, the step of determining the primary responding agent and at least one auxiliary cooperating agent from the agents associated with each of the candidate NFC devices based on the query information includes: Obtain the functional description information of the intelligent agent associated with each of the candidate NFC devices; Calculate the semantic similarity between the question information and the functional description information of each of the intelligent agents, and use it as the relevance score between the question information and each of the intelligent agents; The agent with the highest relevance score among all the agents is designated as the main responder agent, and at least one agent with a relevance score greater than or equal to the preset score threshold, other than the main responder agent, is designated as the auxiliary cooperating agent.

[0008] In one embodiment, the step of determining the primary responding agent and at least one auxiliary cooperating agent from the agents associated with each of the candidate NFC devices based on the query information includes: Obtain a keyword library for each candidate NFC device associated with a smart agent, wherein the keyword library includes at least one keyword and the weight value of the keyword; The question information is matched with the keyword database of each of the intelligent agents to obtain the weight value of the keyword matched by the question information in each of the keyword databases; For any of the aforementioned agents, the sum of the weight values ​​of all keywords matched by the question information in the agent's keyword library is calculated to obtain the matching weight sum value, and the sum of the weight values ​​of all keywords in the agent's keyword library is calculated to obtain the total weight sum value; The ratio between the sum of the matching weights and the total weights is used as the relevance score between the question information and the agent. The agent with the highest relevance score among all the agents is designated as the main responder agent, and at least one agent with a relevance score greater than or equal to the preset score threshold, other than the main responder agent, is designated as the auxiliary cooperating agent.

[0009] In one embodiment, the step of determining the primary responding agent and at least one auxiliary cooperating agent from the agents associated with each of the candidate NFC devices based on the query information includes: The intent of the question is identified by performing intent recognition on the question information to obtain the intent category of the question information; Obtain the capability tags of the intelligent agents associated with each of the candidate NFC devices, the capability tags being used to characterize the intent categories that the intelligent agent can process; The intent category of the question information is matched with the capability tags of each agent to determine the relevance score between the question information and each agent. The agent with the highest relevance score among all the agents is designated as the main responder agent, and at least one agent with a relevance score greater than or equal to the preset score threshold, other than the main responder agent, is designated as the auxiliary cooperating agent.

[0010] In one embodiment, the step of selecting all NFC devices within the area where the target NFC device is located as candidate NFC devices includes: Obtain the region identifier of the target NFC device, which is used to characterize the physical or logical region to which the target NFC device belongs; Based on the preset mapping relationship between area identifiers and NFC devices, all NFC devices corresponding to the area identifiers are obtained as candidate NFC devices.

[0011] In one embodiment, before the steps of invoking the main response agent to generate a first response based on the question information and invoking the auxiliary collaborative agent to generate a second response based on the question information, the method further includes: After determining the main response agent and the auxiliary collaborative agent, the user terminal device is controlled to output collaborative interaction prompt information. The collaborative interaction prompt information is used to inform the user of the currently determined main response agent and the auxiliary collaborative agent, as well as the NFC device corresponding to each of the main response agent and the auxiliary collaborative agent. In response to the user's confirmation instruction on the collaborative interaction prompt information, the steps of calling the main response agent to generate a first response based on the question information and calling the auxiliary collaborative agent to generate a second response based on the question information are executed.

[0012] In one embodiment, before the step of fusing the first reply content and the second reply content to generate collaborative reply content, the method further includes: Determine the degree of information overlap between the first response content and the second response content; If the information redundancy is greater than or equal to a preset redundancy threshold, the first reply content is displayed on the user terminal device; If the information redundancy is less than the preset redundancy threshold, then the step of merging the first reply content and the second reply content to generate collaborative reply content is executed.

[0013] Furthermore, to achieve the above objectives, this application also provides a multi-agent collaborative interaction device, the device comprising: The information acquisition module is used to acquire the user's query information through the user terminal device when the user terminal device identifies the target NFC device, and to regard all NFC devices in the area where the target NFC device is located as candidate NFC devices. The agent determination module is used to determine a main responding agent and at least one auxiliary cooperating agent from the agents associated with each candidate NFC device according to the question information. The main responding agent is the agent with the highest relevance score to the question information, and the auxiliary cooperating agent is the agent other than the main responding agent whose relevance score to the question information is greater than or equal to a preset score threshold. The response content generation module is used to call the main response agent to generate a first response content based on the question information, and to call the auxiliary collaborative agent to generate a second response content based on the question information; The reply content display module is used to merge the first reply content and the second reply content to generate collaborative reply content, and to display the collaborative reply content on the user terminal device.

[0014] In addition, to achieve the above objectives, this application also provides an electronic device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the multi-agent cooperative interaction method as described above.

[0015] In addition, to achieve the above objectives, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the multi-agent cooperative interaction method as described above.

[0016] This application provides a multi-agent collaborative interaction method. When a user terminal device identifies a target NFC device, it obtains the user's query information through the user terminal device and considers all NFC devices within the area where the target NFC device is located as candidate NFC devices. This allows all agents within the same area to participate in the interaction, thus providing a foundation for multi-agent collaboration and enabling simultaneous interaction between the user and multiple agents. Then, based on the query information, a primary responding agent and at least one auxiliary cooperating agent are determined from the agents associated with each candidate NFC device. The primary responding agent is defined as the agent with the highest relevance score, and the auxiliary cooperating agent is defined as the agent with a relevance score greater than or equal to a preset threshold. The system can select the most relevant agent to lead the response based on the user's actual question, while retaining other relevant agents to supplement it. This avoids the one-sidedness caused by responses from agents associated with NFC devices identified by the user's terminal device, improving the accuracy and comprehensiveness of answers to complex questions. Next, the main responding agent generates a first response based on the question information, and the auxiliary coordinating agent generates a second response based on the question information. Finally, the first and second responses are merged to generate a collaborative response, which is then displayed on the user's terminal device. This allows the user to obtain a complete answer covering multiple agent knowledge dimensions, further improving the comprehensiveness of answers to complex questions.

[0017] Therefore, the technical solution provided in this application enables users to interact with multiple intelligent agents simultaneously and improves the accuracy and comprehensiveness of answers to complex questions. Attached Figure Description

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

[0019] 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, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating the multi-agent cooperative interaction method provided in the first embodiment of this application; Figure 2 A flowchart illustrating the multi-agent cooperative interaction method provided in the second embodiment of this application; Figure 3 A flowchart illustrating the multi-agent collaborative interaction method provided in the third embodiment of this application; Figure 4A schematic diagram of the module structure of the multi-agent collaborative interaction device provided in the embodiments of this application; Figure 5 This is a schematic diagram of the hardware operating environment involved in the embodiments of this application.

[0021] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0023] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0024] With the development of near field communication technology, NFC (Near Field Communication) technology has been widely used in various interactive products.

[0025] Currently, mainstream NFC interactive products typically adopt a "one-to-one" interaction mode. When a user touches an NFC device, the system only activates the smart agent associated with that NFC device to respond to the user's question; the user cannot interact with multiple smart agents simultaneously. For example, in a smart museum scenario, when a user touches the NFC device corresponding to an exhibit, the system only invokes the explanatory smart agent associated with that exhibit to provide an answer.

[0026] However, users' actual questions often involve multiple fields or require multifaceted knowledge support. A single intelligent agent, limited by its pre-set knowledge scope, finds it difficult to comprehensively and accurately answer users' complex questions.

[0027] Based on this, this application provides a multi-agent collaborative interaction method. When a user terminal device identifies a target NFC device, it obtains the user's query information through the user terminal device and considers all NFC devices within the area where the target NFC device is located as candidate NFC devices, so that all agents within the same area have the opportunity to participate in the interaction, thus providing a foundation for multi-agent collaboration and enabling simultaneous interaction between the user and multiple agents. Then, based on the query information, a primary responding agent and at least one auxiliary collaborative agent are determined from the agents associated with each candidate NFC device. The primary responding agent is defined as the agent with the highest relevance score, and the auxiliary collaborative agent is defined as the agent with a relevance score greater than or equal to a preset threshold. This allows the system to select the most relevant agent to lead the response based on the user's actual question, while retaining other relevant agents to supplement it. This avoids the one-sidedness caused by responses from agents associated with NFC devices identified by the user's terminal device, improving the accuracy and comprehensiveness of answers to complex questions. Next, the main responding agent generates a first response based on the question information, and the auxiliary coordinating agent generates a second response based on the question information. Finally, the first and second responses are merged to generate a collaborative response, which is then displayed on the user's terminal device. This allows the user to obtain a complete answer covering multiple agent knowledge dimensions, further enhancing the comprehensiveness of answers to complex questions.

[0028] Therefore, the technical solution provided in this application enables users to interact with multiple intelligent agents simultaneously and improves the accuracy and comprehensiveness of answers to complex questions.

[0029] The executing entity of the multi-agent cooperative interaction method of this application can be an electronic device with data processing, network communication and program execution functions. For example, it can be a control system, control circuit, etc. that can realize the above functions. This embodiment does not specifically limit it.

[0030] The following description uses an electronic device as the execution subject to illustrate the various embodiments.

[0031] This application proposes a multi-agent cooperative interaction method according to a first embodiment. Please refer to [link / reference]. Figure 1 The multi-agent collaborative interaction method may include steps S10~S40: Step S10: When the user terminal device recognizes the target NFC device, the user's query information is obtained through the user terminal device, and all NFC devices in the area where the target NFC device is located are selected as candidate NFC devices. It should be noted that the user terminal device can be a mobile phone, tablet computer, smart speaker, smartwatch, personal computer, self-service terminal with screen and / or voice interaction functions, etc., and this embodiment does not specifically limit it. The target NFC device is the physical entity with a built-in NFC tag identified by the user terminal device, and each NFC device is associated with a smart agent. The question information is the question content entered by the user in the form of voice or text.

[0032] For example, assuming the user's terminal device is a mobile phone, when the user brings the phone close to the target NFC device, the phone can read the NFC tag information of the target NFC device and notify the user of successful reading through vibration; then, the phone automatically pops up a voice input interface, the user speaks a question, the phone collects the voice through its built-in microphone, and uses a voice recognition engine to convert it into text as the user's question information.

[0033] For example, assuming the user's terminal device is a tablet computer, when the user brings the tablet computer close to the target NFC device, the tablet computer can read the NFC tag information of the target NFC device and directly display a text input box on the screen; the user can get the user's question information by entering a question in the text input box.

[0034] In one feasible implementation, step S10 may include steps S11-S12: Step S11: Obtain the region identifier of the target NFC device. The region identifier is used to characterize the physical or logical region to which the target NFC device belongs. It should be noted that a region identifier is information that uniquely identifies a region, and can be a numeric ID, a string name, a QR code value, etc. This embodiment does not impose specific limitations on it. A physical region refers to a physical area defined based on spatial location, geographical boundaries, or the arrangement of objects. NFC devices have a fixed spatial affiliation within this area. For example, multiple NFC devices placed on the same display stand, regardless of their theme or function, belong to the same physical region.

[0035] A logical area refers to a non-physical region defined based on abstract dimensions such as content theme, functional category, user preferences, or business rules. NFC devices within this region have the same or similar attribute tags, regardless of their physical location. For example, all NFC devices with storytelling capabilities (such as teddy bears, moon dolls, and star dolls) can be grouped into the story function group, forming the same logical area.

[0036] Step S12: Based on the preset mapping relationship between area identifiers and NFC devices, obtain all NFC devices corresponding to the area identifiers as candidate NFC devices.

[0037] It should be noted that the mapping relationship between area identifiers and NFC devices can be recorded using methods such as databases, configuration files, and mapping lists. This embodiment does not impose any specific limitations on this.

[0038] In this embodiment, the region identifier of the target NFC device is first obtained to clarify the physical or logical region to which the target NFC device belongs. Then, based on the preset mapping relationship between region identifiers and NFC devices, all other NFC devices within the region where the target NFC device is located are quickly and accurately identified as candidate NFC devices. This provides a reliable candidate NFC device determination mechanism for achieving collaborative interaction among multiple agents, laying a data foundation for dynamically selecting primary and secondary agents based on query information, and ensuring the accuracy and flexibility of regional linkage.

[0039] Step S20: Based on the question information, determine the main responding agent and at least one auxiliary cooperating agent from the agents associated with each candidate NFC device. The main responding agent is the agent with the highest relevance score to the question information, and the auxiliary cooperating agent is the agent other than the main responding agent whose relevance score to the question information is greater than or equal to a preset score threshold. It should be noted that an intelligent agent is an intelligent application or entity capable of autonomous action, environmental perception, decision-making, and interaction with the environment. The relevance score is a numerical value used to quantify the degree of matching between the question information and the intelligent agent; a higher relevance score indicates that the intelligent agent is more suitable to answer the question. A preset score threshold serves as the basis for determining whether the relevance score between the intelligent agent and the question information is high. This threshold can be a default value or can be flexibly set by the user according to actual circumstances; this embodiment does not impose specific limitations on it.

[0040] In a first feasible implementation, step S20 may include steps S211 to S213: Step S211: Obtain the functional description information of the smart agent associated with each candidate NFC device; It should be noted that the job description information is a piece of natural language text used to summarize and describe the knowledge domain, the types of questions the agent is good at answering, and its role positioning or functional scope. For example, the job description of the agent "Storytelling King" could be "good at telling fairy tales, role-playing, and bedtime stories".

[0041] Step S212: Calculate the semantic similarity between the question information and the functional description information of each agent, and use it as the relevance score between the question information and each agent; It should be noted that semantic similarity is a numerical value used to measure the semantic closeness between two texts: a question and a job description. Unlike keyword matching, semantic similarity can capture synonyms, near-synonyms, and contextual semantic relationships.

[0042] When calculating the semantic similarity between the question information and the functional description information of each agent, a pre-trained semantic matching model can be used to encode the question information and functional description information into semantic vectors respectively, and then the cosine similarity between the two semantic vectors can be calculated as the semantic similarity. Alternatively, the question information and functional description information can be segmented into words, and each word can be converted into a word vector. Then, the weighted average of all word vectors corresponding to the question information can be calculated as the overall vector of the question information, and the weighted average of all word vectors corresponding to the functional description information can be calculated as the overall vector of the functional description information. Finally, the cosine similarity between the two overall vectors can be calculated as the semantic similarity. This embodiment does not specifically limit the implementation of step S212.

[0043] Step S213: Select the agent with the highest relevance score among all agents as the main responder agent, and select at least one agent with a relevance score greater than or equal to a preset score threshold, other than the main responder agent, as an auxiliary cooperating agent.

[0044] In this implementation, firstly, by acquiring the functional description information of the intelligent agents associated with each candidate NFC device, a quantifiable capability profile is established for each intelligent agent. Then, by calculating the semantic similarity between the question information and each functional description information, a precise semantic match between the user's question and the intelligent agent's capabilities is achieved. Subsequently, by ranking by relevance score and filtering by threshold, the main responding intelligent agent (the one with the highest score) and the auxiliary collaborative intelligent agents (those who meet the score) are dynamically determined. This allows the system to intelligently select the most relevant intelligent agent to lead the answer based on the question content, while retaining other relevant intelligent agents to participate in supplementation. This effectively avoids the one-sided problem caused by the answering only by the intelligent agents associated with the NFC device identified by the user terminal device, ensuring the accuracy and comprehensiveness of the answer from the source.

[0045] In a second feasible implementation, step S20 may include steps S221 to S225: Step S221: Obtain the keyword library of the smart agent associated with each candidate NFC device. The keyword library includes at least one keyword and the weight value of the keyword. It's important to note that an agent's keyword library is a set of keywords associated with the agent. Each keyword in the library represents a domain the agent excels in or a core concept it can recognize. For example, a weather assistant's keyword library could include weather, rain, sunny, temperature, and forecast. The weight value of a keyword is used to characterize its importance to the agent's functions.

[0046] Step S222: Match the question information with the keyword database of each agent to obtain the weight value of the keywords matched by the question information in each keyword database; It should be noted that the keywords matched by the question information in the keyword database are those keywords in the keyword database that are the same as or semantically similar to the words contained in the question information.

[0047] When matching the query information with the keyword databases of each agent, the query information can be segmented into words, and the segmentation results can be precisely compared with the keywords in the keyword database. If the query information contains a certain keyword, the weight value of that keyword is used as the matching result. For example, assuming the query information is "How is the weather today?", after segmentation, it contains "weather", so it matches "weather" in the keyword database of "weather assistant", with a weight value of 0.9. Alternatively, the query information and keywords can be morphologically restored before matching (e.g., restoring "it's raining" to "rain"), and then precise matching can be performed. This method can handle morphological changes of words and improve the matching coverage. In addition to precise matching, synonyms or semantically similar words can be introduced for extended matching. For example, if the keyword database contains "weather", when "weather" appears in the query information, since the two are semantically similar, the system still considers it a successful match and can appropriately reduce the weight value according to the degree of similarity. This embodiment does not specifically limit the implementation of step S222.

[0048] Step S223: For any agent, calculate the sum of the weight values ​​of all keywords matched by the question information in the agent's keyword library to obtain the matching weight sum value, and calculate the sum of the weight values ​​of all keywords in the agent's keyword library to obtain the total weight sum value. Step S224: The ratio between the sum of matching weights and the total sum of weights is used as the relevance score between the query information and the agent. Step S225: Select the agent with the highest relevance score among all agents as the main responder agent, and select at least one agent with a relevance score greater than or equal to a preset score threshold, other than the main responder agent, as an auxiliary cooperating agent.

[0049] In this implementation, firstly, by acquiring the keyword library and weight values ​​of the intelligent agents associated with each candidate NFC device, a quantifiable capability profile is established for each intelligent agent. The weight values ​​reflect the importance of different keywords to the intelligent agent's functions. Then, by matching the question information with each keyword library and calculating the ratio of the matching weights to the total weights as a relevance score, a quantitative assessment of the relevance between the question information and the intelligent agent is achieved. This ratio effectively reflects the proportion of semantic coverage of the intelligent agent's core capabilities by the question information. Finally, through relevance score ranking and threshold filtering, the primary responding intelligent agent (the one with the highest score) and auxiliary collaborative intelligent agents (those meeting the score threshold) are dynamically determined. This allows the system to intelligently select the most relevant intelligent agent to lead the answer based on the question content, while retaining other relevant intelligent agents to participate in supplementation. This effectively avoids the one-sidedness caused by answers from intelligent agents associated with NFC devices identified only by the user terminal device, ensuring the accuracy and comprehensiveness of the answer from the source.

[0050] It is understandable that this implementation method has the advantages of high computational efficiency, strong interpretability, and ease of debugging and maintenance, and is especially suitable for scenarios with limited resources or high real-time requirements. By reasonably configuring keywords and weight values, developers can precisely control the triggering conditions of each agent, enabling the system to quickly and accurately select the most relevant agent to lead the answer based on the question information, while retaining other relevant agents to participate in supplementing the answer, providing an efficient and controllable agent selection basis for subsequent multi-agent collaborative generation of response content.

[0051] In a third feasible implementation, step S20 may include steps S231 to S234: Step S231: Perform intent recognition on the question information to obtain the intent category of the question information; It should be noted that intent recognition is a natural language processing technique used to identify the intent or purpose expressed behind a user's input text. For example, the intent of the question "What's the weather like today?" is "to check the weather." Intent categories are predefined intent classification labels, such as checking the weather, telling a story, or setting reminders. These can be flexibly defined according to the application scenario, and this embodiment does not impose any specific limitations on them.

[0052] When performing intent recognition on the query information to obtain the intent category of the query information, in one feasible implementation, an intent classification model can be pre-trained. The query information can then be input into the intent classification model, which outputs the probability that the query information belongs to each intent category. The category with the highest probability is taken as the intent category of the query information. In another feasible implementation, intent recognition can be performed by combining keyword matching and rule logic. For example, keywords such as "weather," "rain," and "temperature" can be set as trigger words for the intent of "query the weather." When the query information contains these keywords, it is initially determined to be "query the weather." Further subdivision into subcategories such as "reason for querying the weather" can be achieved through rules (such as whether it contains interrogative words like "why"). This embodiment does not specifically limit the implementation of step S231.

[0053] Step S232: Obtain the capability tags of the intelligent agents associated with each candidate NFC device. The capability tags are used to characterize the intent categories that the intelligent agents can process. It should be noted that capability tags are identification information used to describe the types of intents that an agent can process, and each agent can correspond to one or more capability tags. For example, the capability tags for the agent "Weather Assistant" could be "Check the weather" or "Check the temperature"; the capability tags for the agent "Storyteller" could be "Tell a story" or "Role-playing".

[0054] Step S233: Match the intent category of the question information with the capability labels of each agent to determine the relevance score between the question information and each agent; It should be noted that matching the intent category of the question with the agent's capability label is essentially comparing the intent category of the question with the agent's capability label to determine whether the agent has the ability to process that intent category.

[0055] When matching the intent category of a question with the capability tags of each agent to determine the relevance score, one feasible implementation can be to perform string matching between the intent category of the question and the capability tags of each agent. If the intent category completely matches a capability tag, the relevance score is 1 (or a preset high score); if they do not match, the score is 0 (or a preset low score). For example, assuming the intent category of the question is "check the weather," and the capability tag of the agent "Weather Assistant" includes "check the weather," the relevance score is 1; if the capability tag of the agent "Scientist" does not include "check the weather," the relevance score is 0. In another feasible implementation, the intent category of the question and the capability tags of each agent can be organized hierarchically (e.g., "Entertainment - Storytelling," "Entertainment - Role-playing"). When there is a hierarchical inclusion relationship between the intent category of the question and the capability tag of the agent, a partial score is given. For example, assuming the intent category of the question is "Entertainment," and the capability tag of the agent "Storyteller" is "Storytelling," since "Storytelling" is a subcategory of "Entertainment," a relevance score of 0.8 can be given. This embodiment does not specifically limit the implementation method of step S233.

[0056] Step S234: Select the agent with the highest relevance score among all agents as the main responder agent, and select at least one agent with a relevance score greater than or equal to a preset score threshold, other than the main responder agent, as an auxiliary cooperating agent.

[0057] In this implementation, the user's question is first converted into a structured intent category by performing intent recognition on the question information, so as to accurately grasp the user's real needs from a semantic level. Then, by obtaining the capability tags of the intelligent agents associated with each candidate NFC device, a clear intent processing capability profile is established for each intelligent agent. Next, by matching the intent category with the capability tag, it is quickly and accurately determined whether each intelligent agent has the ability to process the current user's intent, and a relevance score is determined based on the degree of matching. Finally, by sorting the relevance scores and filtering by threshold, the main responding intelligent agent (the one with the highest score) and the auxiliary collaborative intelligent agents (those with the target score) are dynamically determined. This allows the system to intelligently select the most relevant intelligent agent to lead the answer based on the question content, while retaining other relevant intelligent agents to participate in the supplement. This effectively avoids the one-sided problem caused by the answering only by the intelligent agent associated with the NFC device identified by the user terminal device, and ensures the accuracy and comprehensiveness of the answer from the source.

[0058] Understandably, this implementation method offers stronger advantages in structure and interpretability compared to the previous two. Specifically, intent recognition can strip away surface noise from the text, directly addressing the user's core needs; the matching results of capability tags are intuitive and clear, facilitating system debugging and user understanding. Through this mechanism, the system can accurately select the most suitable agent to lead the response based on the user's intent, while simultaneously selecting agents with other relevant capabilities to participate in collaboration. This multi-agent collaborative approach to answering user questions more closely reflects the user's actual needs, effectively improving the relevance and accuracy of the responses.

[0059] The above are only three feasible implementation methods of step S20 provided in this embodiment. This embodiment does not limit the implementation method of step S30.

[0060] Step S30: Invoke the main response agent to generate the first response content based on the question information, and invoke the auxiliary collaborative agent to generate the second response content based on the question information; It should be noted that, in this embodiment, the response content generated by the main responding agent based on the question information is referred to as the first response content; and the response content generated by the auxiliary coordinating agent based on the question information is referred to as the second response content.

[0061] When the main response agent generates the first response content based on the question information, and the auxiliary collaborative agent generates the second response content based on the question information, the generation can be synchronous and parallel, or asynchronous and serial. For example, the main response agent can be called first to generate the first response content, and then the first response content, along with the question information, can be passed to the auxiliary collaborative agent as context, allowing the auxiliary collaborative agent to generate more targeted supplementary content based on its understanding of the first response content. For example, if the main response agent returns "Tomorrow will be cloudy turning sunny," the auxiliary collaborative agent, based on the main response agent's response content and the question information, can generate "This weather is suitable for outdoor activities; it is recommended to take precautions against sunburn." This embodiment does not specifically limit the implementation method of step S30.

[0062] Step S40: Merge the first reply content and the second reply content to generate collaborative reply content, and display the collaborative reply content on the user's terminal device.

[0063] It should be noted that merging the first and second response contents involves combining, deduplicating, sorting, or structuring the first and second response contents to form a unified output as collaborative response content. When displaying the collaborative response content on a user terminal device, it can be displayed directly on the user terminal device screen, or it can be played by the user terminal device via voice broadcast. This embodiment does not specifically limit this.

[0064] In one feasible implementation, when merging the first and second response contents to generate collaborative response content and displaying it on the user's terminal device, text similarity calculation can be performed on the first and second response contents to identify and remove duplicate information. Then, the deduplicated first and second response contents are concatenated in the order of the first response content generated by the main response agent first, followed by the second response content generated by the auxiliary collaborative agent, to obtain the collaborative response content, which is then displayed on the user's terminal device.

[0065] In another feasible implementation, source tags can be added to each part of the information in the first response content and the second response content. The source tag includes at least the identity of the intelligent agent that generated the information and the corresponding NFC device identity. The information after adding the source tags is spliced ​​together in the order of the first response content generated by the main responding intelligent agent first and the second response content generated by the auxiliary cooperating intelligent agent last. When displaying the spliced ​​content on the user terminal device, the source tags are displayed in association with the visual elements corresponding to the NFC device.

[0066] The source tag is the metadata information attached to the response content, used to identify which smart agent generated the content and what the corresponding NFC device is. The source tag can be presented in the form of text, icons, colors, or identification codes. Visual elements are graphical elements used to represent NFC devices on the user interface, such as device thumbnails, colored dots, custom icons, and device name labels. Each visual element corresponds one-to-one with an NFC device to help users quickly identify the source of the information.

[0067] This implementation adds source tags to each part of the information in the first and second response content, splices them together in the order of main response first and auxiliary response last, and then displays them in association with the visual elements corresponding to the NFC device. This achieves traceability and visual differentiation of collaborative response content, which greatly optimizes the transparency of information presentation and user experience while ensuring the integrity of information.

[0068] The above are only two feasible implementations of step S40 provided in this embodiment. This embodiment does not specifically limit the implementation of step S40.

[0069] Based on the above, this embodiment provides a multi-agent collaborative interaction method. When a user terminal device identifies a target NFC device, it obtains the user's query information through the user terminal device and considers all NFC devices within the area where the target NFC device is located as candidate NFC devices. This allows all agents within the same area to participate in the interaction, thus providing a foundation for multi-agent collaboration and enabling simultaneous interaction between the user and multiple agents. Then, based on the query information, a primary responding agent and at least one auxiliary collaborative agent are determined from the agents associated with each candidate NFC device. The primary responding agent is defined as the agent with the highest relevance score, and the auxiliary collaborative agent is defined as the agent with a relevance score greater than or equal to a preset threshold. The system employs an agent-based approach, enabling it to select the most relevant agent to lead the response based on the user's actual question, while retaining other relevant agents to supplement it. This avoids the one-sidedness caused by responses from agents associated with NFC devices identified by the user's terminal device, improving the accuracy and comprehensiveness of answers to complex questions. Next, the main responding agent generates a first response based on the question information, and the auxiliary coordinating agent generates a second response based on the question information. Finally, the first and second responses are merged to generate a collaborative response, which is then displayed on the user's terminal device. This allows the user to obtain a complete answer covering multiple agent knowledge dimensions, further enhancing the comprehensiveness of answers to complex questions.

[0070] Therefore, the technical solution provided in this embodiment can enable users to interact with multiple intelligent agents simultaneously and improve the accuracy and comprehensiveness of answers to complex questions.

[0071] Based on the first embodiment described above, a second embodiment of the multi-agent cooperative interaction method of this application is proposed. For the second embodiment, please refer to... Figure 2 Before step S30, the multi-agent cooperative interaction method may further include steps S301-S302: Step S301: After determining the main responding agent and the auxiliary cooperating agent, control the user terminal device to output cooperative interaction prompt information. The cooperative interaction prompt information is used to inform the user of the currently determined main responding agent and auxiliary cooperating agent, as well as the NFC device corresponding to each of the main responding agent and the auxiliary cooperating agent. Step S302: In response to the user's confirmation instruction on the collaborative interaction prompt information, the steps of calling the main response agent to generate the first response content based on the question information and calling the auxiliary collaborative agent to generate the second response content based on the question information are executed.

[0072] It should be noted that the collaborative interaction prompt information is used to display the prompt content of the current collaborative interaction configuration to the user. This may include the identity information of the main responding intelligent agent (such as name and role), the identity information of the auxiliary collaborative intelligent agents, and the identifiers of the NFC devices corresponding to each intelligent agent (such as device name and icon). The collaborative interaction prompt information can be presented through text, voice, images, or a combination of these formats; this embodiment does not impose specific limitations on this. The confirmation command is the user's response to the collaborative interaction prompt information, indicating that the user acknowledges the currently selected intelligent agent combination and agrees to continue generating subsequent responses. The confirmation command may be clicking the confirmation button on the screen, speaking a voice command such as "confirm" or "okay," or refraining from a negative operation within a certain timeframe; this embodiment does not impose specific limitations on this.

[0073] This embodiment introduces an interactive confirmation step after determining the primary and secondary intelligent agents and before actually calling the agents to generate response content. First, by outputting collaborative interaction prompts, the system transparently informs the user of the agent selection results in a visual or audio format, allowing the user to clearly understand the composition of the agents to be involved in the collaboration. Then, by responding to a confirmation command, the user decides whether the subsequent response generation is appropriate. The user can confirm the suitability of the current combination according to their needs, or make adjustments if necessary (such as with editable prompts), avoiding the system automatically executing collaborative combinations that may not meet the user's expectations. Therefore, this embodiment significantly improves the naturalness, transparency, and user satisfaction of the interaction while ensuring the accuracy of multi-agent collaboration.

[0074] Based on the first and / or second embodiments described above, a third embodiment of the multi-agent cooperative interaction method of this application is proposed. In the third embodiment, please refer to... Figure 3 Before step S40, the multi-agent cooperative interaction method may further include steps S401 to S403: Step S401: Determine the degree of information overlap between the first response content and the second response content; It should be noted that information redundancy is a numerical value used to quantify the degree of information overlap between the first and second responses. A higher information redundancy indicates more semantic, factual, or expressive overlap between the two responses.

[0075] When determining the information redundancy between the first and second responses, in one feasible implementation, a text similarity algorithm (such as cosine similarity, edit distance, etc.) can be used to calculate the similarity between the first and second responses as the information redundancy. In another feasible implementation, a pre-trained semantic model can be used to encode the first and second responses into semantic vectors, and then the cosine similarity between the two semantic vectors can be calculated as the information redundancy. This embodiment does not specifically limit the implementation of step S401.

[0076] Step S402: If the information redundancy is greater than or equal to the preset redundancy threshold, then the first reply content is displayed on the user terminal device. Step S403: If the information redundancy is less than the preset redundancy threshold, then the step of merging the first reply content and the second reply content to generate collaborative reply content is executed.

[0077] It should be noted that the preset repetition threshold is used as the basis for determining whether the content of the first reply and the content of the second reply are highly overlapping. It can be a default value or it can be flexibly set by the user according to the actual situation. This embodiment does not make specific limitations on it.

[0078] This embodiment introduces a mechanism for information redundancy judgment and selective fusion before fusing the first and second response contents. First, by determining the information redundancy between the first and second response contents, the degree of information overlap between the responses generated by the main responding agent and the auxiliary coordinating agent is quantitatively evaluated. Then, the degree of information overlap is compared with a preset redundancy threshold. When the redundancy is greater than or equal to the preset threshold, it is determined that the second response content highly overlaps with the first response content and lacks incremental information. In this case, only the first response content can be displayed to avoid redundant information interfering with the user, ensuring the simplicity and readability of the displayed content. When the redundancy is less than the preset threshold, it is determined that the two responses have complementary value. A fusion step is then performed, organically integrating the information from both before displaying it, allowing the user to obtain a more comprehensive and multi-dimensional answer. Therefore, this embodiment effectively solves the information redundancy problem that may occur in multi-agent collaboration, improving display efficiency while ensuring information integrity. This allows users to quickly obtain core information without being disturbed by duplicate content, further optimizing the interactive experience.

[0079] Based on the first, second, and / or third embodiments described above, a fourth embodiment of the multi-agent cooperative interaction method of this application is proposed. In the fourth embodiment, after step S40, the multi-agent cooperative interaction method may further include: After displaying the collaborative response content on the user's terminal device, the system receives user feedback on the collaborative response content. If the feedback is positive, the system records the combination of the main responding agent and the auxiliary collaborating agent participating in this collaboration as the preferred combination. If the feedback is negative, the system records the combination of the main responding agent and the auxiliary collaborating agent participating in this collaboration as the combination to be optimized. In subsequent interactions, when it is detected that the question information matches the triggering conditions of the preferred combination, the preferred combination is used to generate response content based on the question information.

[0080] It should be noted that the evaluation feedback refers to the user's expression of satisfaction with the content of this collaborative response. This can be explicit feedback (such as clicking "helpful / useless," rating, or selecting an emoji) or implicit feedback (such as dwell time, whether to ask follow-up questions, etc.). Positive feedback indicates that the user approves of the collaborative result, while negative feedback indicates that the user does not approve. The optimal combination refers to the agent combination that has been validated by positive user feedback and is considered to be effective. Combinations to be optimized refer to agent combinations that have been validated by negative user feedback and are considered ineffective; the system can mark these as low priority or use them as negative samples for subsequent model adjustments. The triggering condition is the condition used to determine whether the current question information is suitable for a certain optimal combination. The triggering condition can be specific keywords contained in the question information, a match between the intent category and the capabilities of the agents in the combination, etc., but this embodiment does not specifically limit these.

[0081] In this embodiment, user feedback on collaborative responses is first received to quantify subjective satisfaction into recordable tags. Then, based on the evaluation results, the participating agent combinations are categorized and recorded as preferred or unoptimized combinations to form reusable experiential knowledge. In subsequent interactions, when a question matches the triggering conditions of a preferred combination, that preferred combination is prioritized, thereby reducing redundant calculations, improving response speed, and gradually converging the system's agent selection strategy towards user preferences. Thus, this embodiment enables the system to learn from each interaction, continuously optimize collaborative effects, avoid recurring combinations that dissatisfy users, and significantly improve the user experience.

[0082] This application also provides a multi-agent collaborative interaction device. Please refer to... Figure 4 Multi-agent collaborative interaction devices may include: Information acquisition module 10 is used to acquire user query information through user terminal device when user terminal device recognizes target NFC device, and to regard all NFC devices in the area where target NFC device is located as candidate NFC devices. The agent determination module 20 is used to determine the main responding agent and at least one auxiliary cooperating agent from the agents associated with each candidate NFC device according to the question information. The main responding agent is the agent with the highest relevance score to the question information, and the auxiliary cooperating agent is the agent other than the main responding agent whose relevance score to the question information is greater than or equal to a preset score threshold. The response content generation module 30 is used to call the main response agent to generate the first response content based on the question information, and to call the auxiliary collaborative agent to generate the second response content based on the question information; The reply content display module 40 is used to merge the first reply content and the second reply content to generate collaborative reply content, and display the collaborative reply content on the user's terminal device.

[0083] In one embodiment, the agent determination module 20 is further configured to: Obtain the functional description information of the intelligent agents associated with each candidate NFC device; Calculate the semantic similarity between the question information and the functional description information of each agent, and use it as the relevance score between the question information and each agent. The agent with the highest relevance score among all agents is designated as the primary responder agent, and at least one agent with a relevance score greater than or equal to a preset score threshold, excluding the primary responder agent, is designated as an auxiliary cooperating agent.

[0084] In one embodiment, the agent determination module 20 is further configured to: Obtain the keyword library of the smart agents associated with each candidate NFC device. The keyword library shall include at least one keyword and the weight value of the keyword. The question information is matched with the keyword database of each agent to obtain the weight value of the keywords matched by the question information in each keyword database; For any agent, calculate the sum of the weights of all keywords matched by the query information in the agent's keyword library to obtain the matching weight sum value, and calculate the sum of the weights of all keywords in the agent's keyword library to obtain the total weight sum value; The ratio between the sum of matching weights and the total sum of weights is used as the relevance score between the question information and the agent. The agent with the highest relevance score among all agents is designated as the primary responder agent, and at least one agent with a relevance score greater than or equal to a preset score threshold, excluding the primary responder agent, is designated as an auxiliary cooperating agent.

[0085] In one embodiment, the agent determination module 20 is further configured to: Perform intent recognition on the question information to obtain the intent category of the question information; Obtain the capability tags of the intelligent agents associated with each candidate NFC device. The capability tags are used to characterize the types of intents that the intelligent agents can process. The intent category of the question information is matched with the capability labels of each agent to determine the relevance score between the question information and each agent. The agent with the highest relevance score among all agents is designated as the primary responder agent, and at least one agent with a relevance score greater than or equal to a preset score threshold, excluding the primary responder agent, is designated as an auxiliary cooperating agent.

[0086] In one embodiment, the information acquisition module 10 is further configured to: Obtain the region identifier of the target NFC device. The region identifier is used to characterize the physical or logical region to which the target NFC device belongs. Based on the preset mapping relationship between area identifiers and NFC devices, all NFC devices corresponding to the area identifiers are obtained as candidate NFC devices.

[0087] In one embodiment, the response content generation module 30 is further configured to: After determining the primary responding agent and the auxiliary cooperating agent, the user terminal device is controlled to output cooperative interaction prompt information. The cooperative interaction prompt information is used to inform the user of the currently determined primary responding agent and auxiliary cooperating agent, as well as the NFC device corresponding to each of the primary responding agent and the auxiliary cooperating agent. In response to the user's confirmation instruction for the collaborative interaction prompt, the steps of calling the main response agent to generate the first response content based on the question information and calling the auxiliary collaborative agent to generate the second response content based on the question information are executed.

[0088] In one embodiment, the response content display module 40 is further configured to: Determine the degree of information overlap between the first and second responses; If the information redundancy is greater than or equal to the preset redundancy threshold, the first reply content will be displayed on the user's terminal device. If the information redundancy is less than the preset redundancy threshold, then the step of merging the first and second response content to generate collaborative response content is executed.

[0089] The multi-agent collaborative interaction device provided in this application embodiment enables users to interact simultaneously with multiple agents and improves the accuracy and comprehensiveness of answers to complex questions. Compared with the prior art, the beneficial effects of the multi-agent collaborative interaction device provided in this application embodiment are the same as those of the multi-agent collaborative interaction method provided in the above embodiments, and other technical features in this multi-agent collaborative interaction device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0090] This application also provides an electronic device, which may include: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the multi-agent cooperative interaction method described above.

[0091] The following is for reference. Figure 5 It shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiments of this application. Figure 5 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0092] like Figure 5 As shown, the electronic device may include a processing unit 101 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in read-only memory 102 or a program loaded from storage device 103 into random access memory 104. Random access memory 104 also stores various programs and data required for the operation of the electronic device. The processing unit 101, read-only memory 102, and random access memory 104 are interconnected via bus 105. Input / output interface 106 is also connected to bus 105. Typically, the following systems can be connected to input / output interface 106: input devices 107 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 108 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 103 including, for example, magnetic tape, hard disks, etc.; and communication devices 109. Communication device 109 allows the electronic device to communicate wirelessly or wiredly with other devices to exchange data. Although the diagrams show electronic devices with various systems, it should be understood that it is not required to implement or have all of the systems shown. More or fewer systems may be implemented alternatively.

[0093] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 103, or installed from read-only memory 102. When the computer program is executed by processing device 101, it performs the functions defined in the methods of the embodiments of this application.

[0094] The electronic device provided in this application embodiment employs the multi-agent collaborative interaction method described in the above embodiments, enabling users to interact simultaneously with multiple agents and improving the accuracy and comprehensiveness of answers to complex questions. Compared with the prior art, the beneficial effects of the electronic device provided in this application embodiment are the same as those of the multi-agent collaborative interaction method described in the above embodiments, and other technical features of the electronic device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0095] It should be understood that various parts of the embodiments of this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0096] 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 variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included 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 above claims.

[0097] This application also provides a computer-readable storage medium storing a computer program that can run on a processor. The computer program is used to execute the multi-agent cooperative interaction method described in the above embodiments.

[0098] The computer-readable storage medium provided in this application embodiment may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0099] The aforementioned computer-readable storage medium may be included in an electronic device or may exist independently without being assembled into an electronic device.

[0100] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by an electronic device, the electronic device causes the following actions: when a user terminal device identifies a target NFC device, the electronic device obtains the user's query information through the user terminal device and identifies all NFC devices within the area where the target NFC device is located as candidate NFC devices; based on the query information, the electronic device determines a primary responding agent and at least one auxiliary cooperating agent from the agents associated with each candidate NFC device, wherein the primary responding agent is the agent with the highest relevance score to the query information, and the auxiliary cooperating agent is an agent other than the primary responding agent whose relevance score to the query information is greater than or equal to a preset score threshold; the electronic device invokes the primary responding agent to generate a first response based on the query information and invokes the auxiliary cooperating agent to generate a second response based on the query information; the electronic device merges the first and second response contents to generate a collaborative response, and displays the collaborative response on the user terminal device.

[0101] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0102] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0103] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0104] The computer-readable storage medium provided in this application embodiment stores computer-readable program instructions for executing the above-described multi-agent cooperative interaction method, enabling users to interact simultaneously with multiple agents and improving the accuracy and comprehensiveness of answers to complex questions. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application embodiment are the same as the beneficial effects of the multi-agent cooperative interaction method provided in the above embodiments, and will not be repeated here.

[0105] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the multi-agent cooperative interaction method described above.

[0106] The computer program product provided in this application embodiment enables users to interact simultaneously with multiple intelligent agents and improves the accuracy and comprehensiveness of answers to complex questions. Compared with the prior art, the beneficial effects of the computer program product provided in this application embodiment are the same as those of the multi-agent collaborative interaction method provided in the above embodiments, and will not be repeated here.

[0107] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A multi-agent cooperative interaction method, characterized in that, The method includes: When the user terminal device identifies the target NFC device, the user's query information is obtained through the user terminal device, and all NFC devices in the area where the target NFC device is located are regarded as candidate NFC devices. Based on the question information, a main responding agent and at least one auxiliary cooperating agent are determined from the agents associated with each candidate NFC device. The main responding agent is the agent with the highest relevance score to the question information, and the auxiliary cooperating agent is an agent other than the main responding agent whose relevance score to the question information is greater than or equal to a preset score threshold. The main response agent is invoked to generate a first response based on the question information, and the auxiliary collaborative agent is invoked to generate a second response based on the question information. The first response content and the second response content are combined to generate collaborative response content, which is then displayed on the user terminal device.

2. The method as described in claim 1, characterized in that, The step of determining the main responding agent and at least one auxiliary cooperating agent from the agents associated with each of the candidate NFC devices based on the query information includes: Obtain the functional description information of the intelligent agent associated with each of the candidate NFC devices; Calculate the semantic similarity between the question information and the functional description information of each of the intelligent agents, and use it as the relevance score between the question information and each of the intelligent agents; The agent with the highest relevance score among all the agents is designated as the main responder agent, and at least one agent with a relevance score greater than or equal to the preset score threshold, other than the main responder agent, is designated as the auxiliary cooperating agent.

3. The method as described in claim 1, characterized in that, The step of determining the main responding agent and at least one auxiliary cooperating agent from the agents associated with each of the candidate NFC devices based on the query information includes: Obtain a keyword library for each candidate NFC device associated with a smart agent, wherein the keyword library includes at least one keyword and the weight value of the keyword; The question information is matched with the keyword database of each of the intelligent agents to obtain the weight value of the keyword matched by the question information in each of the keyword databases; For any of the aforementioned agents, the sum of the weight values ​​of all keywords matched by the question information in the agent's keyword library is calculated to obtain the matching weight sum value, and the sum of the weight values ​​of all keywords in the agent's keyword library is calculated to obtain the total weight sum value; The ratio between the sum of the matching weights and the total weights is used as the relevance score between the question information and the agent. The agent with the highest relevance score among all the agents is designated as the main responder agent, and at least one agent with a relevance score greater than or equal to the preset score threshold, other than the main responder agent, is designated as the auxiliary cooperating agent.

4. The method as described in claim 1, characterized in that, The step of determining the main responding agent and at least one auxiliary cooperating agent from the agents associated with each of the candidate NFC devices based on the query information includes: The intent of the question is identified by performing intent recognition on the question information to obtain the intent category of the question information; Obtain the capability tags of the intelligent agents associated with each of the candidate NFC devices, the capability tags being used to characterize the intent categories that the intelligent agent can process; The intent category of the question information is matched with the capability tags of each agent to determine the relevance score between the question information and each agent. The agent with the highest relevance score among all the agents is designated as the main responder agent, and at least one agent with a relevance score greater than or equal to the preset score threshold, other than the main responder agent, is designated as the auxiliary cooperating agent.

5. The method as described in claim 1, characterized in that, The step of selecting all NFC devices within the area where the target NFC device is located as candidate NFC devices includes: Obtain the region identifier of the target NFC device, which is used to characterize the physical or logical region to which the target NFC device belongs; Based on the preset mapping relationship between area identifiers and NFC devices, all NFC devices corresponding to the area identifiers are obtained as candidate NFC devices.

6. The method according to any one of claims 1 to 5, characterized in that, Before the steps of invoking the main response agent to generate the first response content based on the question information and invoking the auxiliary collaborative agent to generate the second response content based on the question information, the method further includes: After determining the main response agent and the auxiliary collaborative agent, the user terminal device is controlled to output collaborative interaction prompt information. The collaborative interaction prompt information is used to inform the user of the currently determined main response agent and the auxiliary collaborative agent, as well as the NFC device corresponding to each of the main response agent and the auxiliary collaborative agent. In response to the user's confirmation instruction on the collaborative interaction prompt information, the steps of calling the main response agent to generate a first response based on the question information and calling the auxiliary collaborative agent to generate a second response based on the question information are executed.

7. The method according to any one of claims 1 to 5, characterized in that, Before the step of merging the first response content and the second response content to generate collaborative response content, the method further includes: Determine the degree of information overlap between the first response content and the second response content; If the information redundancy is greater than or equal to a preset redundancy threshold, the first reply content is displayed on the user terminal device; If the information redundancy is less than the preset redundancy threshold, then the step of merging the first reply content and the second reply content to generate collaborative reply content is executed.

8. A multi-agent collaborative interaction device, characterized in that, The device includes: The information acquisition module is used to acquire the user's query information through the user terminal device when the user terminal device recognizes the target NFC device, and to regard all NFC devices in the area where the target NFC device is located as candidate NFC devices. The agent determination module is used to determine a main responding agent and at least one auxiliary cooperating agent from the agents associated with each candidate NFC device according to the question information. The main responding agent is the agent with the highest relevance score to the question information, and the auxiliary cooperating agent is the agent other than the main responding agent whose relevance score to the question information is greater than or equal to a preset score threshold. The response content generation module is used to call the main response agent to generate a first response content based on the question information, and to call the auxiliary collaborative agent to generate a second response content based on the question information; The reply content display module is used to merge the first reply content and the second reply content to generate collaborative reply content, and to display the collaborative reply content on the user terminal device.

9. An electronic device, characterized in that, The electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the multi-agent cooperative interaction method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the multi-agent cooperative interaction method as described in any one of claims 1 to 7.

Images