Intelligent agent interaction processing method and device

Abstract

The embodiment of the invention provides an agent interaction processing method and device, and the method comprises the steps: obtaining interaction data inputted by a user and interaction environment data of the user in the interaction process of the user and an agent system, performing role label matching according to the interaction data and the interaction environment data to obtain a user role label, and starting from the user role label, scheduling the interaction data to a role agent instance which is deployed by the agent system and corresponds to the user role label, and searching the user role data in the role knowledge base corresponding to the user role label through the role agent instance, and performing interaction response processing according to the user role data and the interaction data to obtain an interaction result, thereby realizing interaction processing matched with the user role in the agent system.

Description

Technical Field

[0001] This document relates to the field of artificial intelligence technology, and in particular to a method and apparatus for intelligent agent interaction processing. Background Technology

[0002] With the rapid development of artificial intelligence technology, the field of intelligent interaction is undergoing a profound upgrade towards personalization and scenario-based applications. The complexity of human-computer interaction is also increasing with the growth of users' diversified needs. In the interaction process of intelligent agent systems, achieving efficient interaction through precise role adaptation has become a key direction for improving user interaction experience. Among them, intelligent agent interaction solutions based on matching interaction roles with user data have been widely used in enterprise office, customer service and other fields. However, as the scenarios of intelligent agent interaction become increasingly rich and the identities and roles of users become more and more diverse, how to better improve the user experience of interacting with intelligent agents has become a focus of attention. Summary of the Invention

[0003] This specification provides one or more embodiments of an intelligent agent interaction processing method, applied to a server-side intelligent agent system. The method includes: acquiring user input interaction data submitted by a client, and acquiring the user's interaction environment data; performing role tag matching based on the interaction data and the interaction environment data to obtain user role tags; scheduling the interaction data to a role proxy instance corresponding to the user role tag in the intelligent agent system's intelligent agent proxy instance library; retrieving user role data from the role knowledge base corresponding to the user role tag through the role proxy instance; and performing interaction response processing based on the user role data and the interaction data to obtain an interaction result.

[0004] This specification provides one or more embodiments of another intelligent agent interaction processing method, applied to a client of an intelligent agent system. The method includes: acquiring interaction data input by a user at the intelligent agent system's interaction interface and submitting it to a server; switching the interaction interface's role tags according to user role tags returned by the server to obtain a role-based interaction interface; the user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data; receiving the interaction result returned by the server and interactively displaying it on the role-based interaction interface; the interaction result is obtained by retrieving user role data and processing interaction responses in the role knowledge base corresponding to the user role tag through a role proxy instance corresponding to the user role tag.

[0005] This specification provides one or more embodiments of an intelligent agent interaction processing device, the device comprising: a data acquisition module configured to acquire user input interaction data submitted by a client and acquire the user's interaction environment data; a tag matching module configured to perform role tag matching based on the interaction data and the interaction environment data to obtain user role tags; a data scheduling module configured to schedule the interaction data to a role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system; and an interaction response processing module configured to retrieve user role data from the role knowledge base corresponding to the user role tag through the role agent instance, and perform interaction response processing based on the user role data and the interaction data to obtain an interaction result.

[0006] This specification provides one or more embodiments of another intelligent agent interaction processing device, the device comprising: a data submission module configured to acquire interaction data input by a user on the interaction interface of an intelligent agent system and submit it to a server; a tag switching module configured to switch role tags on the interaction interface according to user role tags returned by the server to obtain a role interaction interface; the user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data; and an interaction display module configured to receive the interaction results returned by the server and display them interactively on the role interaction interface; the interaction results are obtained by retrieving user role data and processing interaction responses in the role knowledge base corresponding to the user role tags through a role proxy instance corresponding to the user role tags.

[0007] This specification provides one or more embodiments of an intelligent agent interaction processing device, including: a processor; and a memory configured to store computer-executable instructions, which, when executed, cause the processor to: acquire interaction data submitted by a client as user input, and acquire interaction environment data of the user; perform role tag matching based on the interaction data and the interaction environment data to obtain user role tags; schedule the interaction data to a role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system; retrieve user role data in the role knowledge base corresponding to the user role tag through the role agent instance, and perform interaction response processing based on the user role data and the interaction data to obtain an interaction result.

[0008] This specification provides one or more embodiments of another intelligent agent interaction processing device, including: a processor; and a memory configured to store computer-executable instructions, which, when executed, cause the processor to: acquire interaction data input by a user at an interactive interface of an intelligent agent system and submit it to a server; switch role tags on the interactive interface according to user role tags returned by the server to obtain a role-based interactive interface; the user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data; receive the interaction results returned by the server and interactively display them on the role-based interactive interface; the interaction results are obtained by retrieving user role data and processing interaction responses in the role knowledge base corresponding to the user role tags through a role proxy instance corresponding to the user role tags.

[0009] This specification provides one or more embodiments of a computer-readable storage medium for storing computer-executable instructions. When executed, these instructions implement the following process: acquiring user input interaction data submitted by a client, and acquiring the user's interaction environment data; performing role tag matching based on the interaction data and the interaction environment data to obtain a user role tag; scheduling the interaction data to a role agent instance corresponding to the user role tag in the agent agent instance library of the agent system; retrieving user role data from the role knowledge base corresponding to the user role tag through the role agent instance, and performing interaction response processing based on the user role data and the interaction data to obtain an interaction result.

[0010] This specification provides one or more embodiments of another computer-readable storage medium for storing computer-executable instructions, which, when executed, implement the following process: acquiring interaction data input by a user at the interactive interface of an intelligent agent system and submitting it to a server; switching role tags on the interactive interface based on user role tags returned by the server to obtain a role-based interactive interface; the user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data; receiving the interaction results returned by the server and interactively displaying them on the role-based interactive interface; the interaction results are obtained by retrieving user role data and processing interaction responses in the role knowledge base corresponding to the user role tags through a role proxy instance corresponding to the user role tags. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in one or more embodiments of this specification 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 recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1 A schematic diagram illustrating the implementation environment of an intelligent agent interaction processing method provided in one or more embodiments of this specification; Figure 2 A flowchart illustrating an intelligent agent interaction processing method provided in one or more embodiments of this specification; Figure 3 A schematic diagram of the interactive interface of a first client provided for one or more embodiments of this specification; Figure 4 A schematic diagram of the interactive interface of a second client provided in one or more embodiments of this specification; Figure 5 A schematic diagram of the interactive interface of a third client provided in one or more embodiments of this specification; Figure 6 A flowchart illustrating a smart agent interaction processing method applied to a smart agent interaction scenario, provided for one or more embodiments of this specification; Figure 7 A flowchart illustrating another intelligent agent interaction processing method provided in one or more embodiments of this specification; Figure 8 A schematic diagram of an embodiment of an intelligent agent interaction processing device provided in one or more embodiments of this specification; Figure 9 A schematic diagram of another embodiment of an intelligent agent interaction processing device provided in one or more embodiments of this specification; Figure 10 A schematic diagram of the structure of an intelligent agent interaction processing device provided in one or more embodiments of this specification; Figure 11 This is a schematic diagram of the structure of another intelligent agent interaction processing device provided in one or more embodiments of this specification. Detailed Implementation

[0012] To enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of this document.

[0013] The agent interaction processing method provided in one or more embodiments of this specification is applicable to the implementation environment of an agent system. (Refer to...) Figure 1 The implementation environment includes at least: The intelligent agent system server 101, the intelligent agent system client 102, the intelligent agent agent instance library 103, the role agent instances 103-1 to 103-N, and the role knowledge base 104-1 to 104-N; The server 101 is used to obtain user input interaction data and user interaction environment data, perform role tag matching based on the interaction data and interaction environment data to obtain user role tags, schedule the interaction data to the role agent instance 103-1 corresponding to the user role tag in the intelligent agent agent instance library 103 of the intelligent agent system, perform interaction response processing based on user role data and interaction data to obtain interaction results and return them to the client 102; the server 101 can run on one or more servers, a server cluster composed of several servers, or a cloud server of a cloud computing platform; Client 102 is used to obtain the interaction data input by the user in the interaction interface of the intelligent agent system and submit it to server 101. Based on the user role tag returned by server 101, the interaction interface is switched with role tags. The client 102 receives the interaction results returned by the server and displays the interaction in the role interaction interface. Client 102 can run on a user terminal, which can be a mobile phone, personal computer, tablet computer, e-book reader, wearable device, device for information interaction based on AR (Augmented Reality) / VR (Virtual Reality), and laptop computer, etc.

[0014] The agent proxy instance library 103 consists of multiple role proxy instances, specifically 103-1 to 103-N. Each role proxy instance corresponds to a role knowledge base, thus there are multiple role knowledge bases, specifically 104-1 to 104-N. Role proxy instance 103-1 retrieves user role data from the role knowledge base 104-1 corresponding to the user role tag and returns it to the server 101. In this implementation environment, during the interaction between the user and the agent, the client 102 obtains the interaction data input by the user on the interaction interface of the agent system and submits it to the server 101. The server 101 obtains the interaction data and the user's interaction environment data, performs role tag matching based on the interaction data and interaction environment data to obtain the user's role tag, and returns it to the client 102. The client 102 switches the role tag on the interaction interface based on the user's role tag to obtain the role interaction interface. The server 101 schedules the interaction data to the role agent instance 103-1 corresponding to the user's role tag in the agent agent instance library 103 of the agent system. The role agent instance 103-1 retrieves the user's role data in the role knowledge base 104-1 corresponding to the user's role tag and returns it to the server 101. The server 101 performs interaction response processing based on the user's role data and interaction data to obtain the interaction result and returns it to the client 102. The client 102 receives the interaction result returned by the server and displays it on the role interaction interface, thereby realizing the interaction processing adapted to the user's role in the agent system.

[0015] It should be noted that, considering that the user's geographical location, scene data, historical images, and other related data involved in this specification may, to some extent, constitute user privacy, authorization from the user must be obtained before collecting such data to ensure that the data collection operation complies with relevant data management regulations. For example, the user can authorize the data collection when interacting with the intelligent agent for the first time or during user registration. The specific method of data authorization can be to send a data authorization reminder to the user, who can obtain data authorization after confirming the reminder with an instruction. Alternatively, data authorization can also be obtained by signing a data authorization agreement to obtain authorization for data collection or data transmission. This embodiment does not limit this method.

[0016] This specification provides one or more embodiments of an intelligent agent interaction processing method as follows: Reference Figure 2 The intelligent agent interaction processing method provided in this embodiment can be applied to the server side of an intelligent agent system. The method specifically includes steps S202 to S208.

[0017] Step S202: Obtain the user input interaction data submitted by the client, and obtain the user's interaction environment data.

[0018] The intelligent agent system described in this embodiment refers to an intelligent agent or intelligent agent system capable of interacting based on different user roles. Specifically, the intelligent agent system includes a server and a client. The client is used to interact with the user, specifically including acquiring the interaction data input by the user in the client's interaction interface and submitting it to the server, receiving the user role tags returned by the server and switching the role tags, and receiving the interaction results returned by the server and displaying them interactively in the interaction interface. The server is used to process interactions based on different user roles, specifically including matching role tags according to the acquired interaction data and interaction environment data to obtain user role tags, and scheduling the interaction data to the role proxy instance corresponding to the user role tag in the intelligent agent proxy instance library for interaction response processing.

[0019] The agent proxy instance library is used to store role proxy instances for different roles of the corresponding user. Alternatively, the agent proxy instance library can also be composed of role proxy instances for different roles of the corresponding user. In this case, the agent proxy instance library can be a collection of role proxy instances. Specifically, each user may have multiple different roles, that is, a single user may have multiple different role tags. Correspondingly, the agent proxy instance library stores the role proxy instances corresponding to each of the user's roles.

[0020] A role proxy instance refers to an executor deployed or configured by an intelligent agent system to interact with users based on different user roles. Specifically, a role proxy instance can be an intelligent agent or a large language model corresponding to each role label, a sub-intelligent agent of an intelligent agent or a sub-model of a large language model corresponding to each role label, or a proxy instance of an intelligent agent or a large language model corresponding to each role label. A user role label refers to the label of the user's actual identity role, identity type, or interaction scenario when interacting with the intelligent agent system. In this case, the user role label can be an identity type label or an interaction scenario label.

[0021] The interaction data described in this embodiment includes the interaction content input by the user on the client's interactive interface. Specifically, the interaction data can be the interaction content of the user interacting with the intelligent agent in different scenarios, such as the user's interaction content in a work scenario, the user's interaction content in a life scenario, or the user's interaction content in other scenarios. Optionally, the interaction data can be text interaction data, image interaction data, voice interaction data, or operation-type interaction data, such as the user's click command on the interactive interface or the user's configuration trigger command. The interaction environment data refers to the environmental data related to the user's interaction. Specifically, the interaction environment data can include device environment data, such as the client's device type and the client's network communication method, and can also include scene environment data, such as the user's interaction time, the user's interaction location, and / or the user's interaction application.

[0022] In practice, during the interaction between the user and the intelligent agent system, the user inputs interactive data through the interactive interface of the intelligent agent system. The client obtains the interactive data input by the user and submits it to the server. Correspondingly, the server receives the interactive data submitted by the client and obtains the user's interactive environment data, providing a data foundation for subsequent role tag matching based on the interactive data and interactive environment data.

[0023] It should be noted that, to ensure a one-to-one correspondence between role proxy instances and role tags and to accurately respond to the interaction needs of the corresponding roles, the role proxy instances need to be trained and generated. Furthermore, since the training data for role proxy instances comes from user role data, data anonymization and differential privacy computation can be used to improve the security of user data during the training process. Specifically, in one optional implementation method provided in this embodiment, the role proxy instances in the agent proxy instance library are obtained in the following way: Privacy-preserving role data is obtained by performing data anonymization processing and differential privacy calculation on the user role data in the role knowledge base corresponding to any user role tag. The meta-agent agent instance of the intelligent agent system is trained based on privacy role data to obtain the role agent instance corresponding to any user role label.

[0024] The role knowledge base refers to a database deployed in the intelligent agent system that stores the data content corresponding to each user role tag. There is a one-to-one mapping relationship between the role knowledge base and the user role tags, and a one-to-one mapping relationship between the user role tags and the role agent instances. That is, there is a one-to-one mapping relationship between the role knowledge base, the role agent instances, and the user role tags. User role data is stored in the role knowledge base. Specifically, user role data is the data generated when a user interacts with the corresponding role agent instance in the current interactive identity. Optionally, user role data includes the user's historical interaction data.

[0025] Meta-agent proxy instances are the basic models used in an agent system to generate proxy instances for each role. Specifically, a meta-agent proxy instance can be an agent or a large language model, such as a base large language model using the Transformer architecture, or it can be a sub-agent of an agent or a sub-model of a large language model, or it can be a proxy instance of an agent or a large language model.

[0026] Specifically, during the process of data desensitization and differential privacy computation of user role data, a privacy processor can be used for data desensitization and differential privacy computation. The privacy processor can be a lightweight CNN (Convolutional Neural Network) that integrates DD (Data Desensitization) and DP (Differential Privacy) algorithms. By inputting user role data into the CNN for data desensitization and differential privacy computation, privacy role data is obtained.

[0027] During the training of meta-agent agent instances of an intelligent agent system based on privacy role data, the meta-agent agent instances can be trained using a model trainer. The model trainer can be an LSTM (Long Short-Term Memory) network integrating LoRA (Low-Rank Adaptation). By inputting privacy role data into the LSTM, the low-rank parameters of the meta-agent agent instances are adjusted to obtain the role agent instances corresponding to any role label.

[0028] In the process of obtaining the role proxy instance described above, data anonymization and differential privacy calculation can be performed on the user role data in the role knowledge base corresponding to any user role tag to obtain privacy role data. The privacy role data is used as the input for training the meta-agent proxy instance to obtain the role proxy instance. In an optional implementation method provided in this embodiment, the user role data in the role knowledge base is obtained in the following way: Read the user interaction records and corresponding interaction environment records of the intelligent agent system; perform role tag matching on the user interaction records and interaction environment records to obtain user role tags, and store the user interaction records in the role knowledge base corresponding to the obtained user role tags.

[0029] Among them, user interaction records include the user's historical interaction data. Specifically, user interaction records can be records of the user's interaction content with the intelligent agent in different scenarios, such as user interaction content records in work scenarios, user interaction content records in life scenarios, and / or user interaction content records in other scenarios; interaction environment records include the user's historical interaction environment data. Specifically, historical interaction environment data can be device environment data.

[0030] Specifically, by reading the user's interaction records with the intelligent agent system and the corresponding interaction environment records, and performing role tag matching on the read user interaction records and interaction environment records to obtain user role tags, the user interaction records are stored in the role knowledge base corresponding to the user role tags after obtaining the user role tags, thereby obtaining user role data in the role knowledge base.

[0031] Step S204: Perform role tag matching based on the interaction data and the interaction environment data to obtain user role tags.

[0032] As mentioned above, a user role tag refers to the tag indicating the user's actual identity, role type, or interaction scenario when interacting with the intelligent agent system. In this case, the user role tag can be an identity type tag or an interaction scenario tag. In real-world scenarios, the same user may have different identity types when interacting with the intelligent agent system in different scenarios. For example, in a work scenario, user A's role tag might be a programmer tag, while in a life scenario, user A's role tag might be a parent tag. Optionally, the user role tag corresponds one-to-one with the above role proxy instances, and each user role tag corresponds to one of the user's interaction identity types. Since each user has at least one interaction identity type, each user has at least one user role tag.

[0033] In practice, based on the user's input interaction data and interaction environment data, the user's role tags are matched according to the interaction data and interaction environment data to obtain the user's role tags. This provides accurate role basis for subsequently determining the role proxy instance corresponding to the user's role tags and scheduling interaction data to the corresponding role proxy instance.

[0034] In the specific execution process, after acquiring interaction data and interaction environment data, role tag matching is performed based on the interaction data and interaction environment data. In one optional implementation method provided in this embodiment, the user role tag is obtained by performing role tag matching based on interaction data and interaction environment data, including: Feature extraction and feature fusion are performed on interactive data and interactive environment data to obtain fused interactive features; The fused interaction features are input into the role classifier for role classification to obtain user role tags; Specifically, in the process of feature extraction and feature fusion of interactive data and interactive environment data, feature extraction and feature fusion can be performed through a feature extractor. The feature extractor can be a lightweight CNN (Convolutional Neural Network) that integrates NLP (Natural Language Processing). By inputting interactive data and interactive environment data into the CNN for feature extraction and feature fusion, the corresponding fused interactive features are obtained. In the process of inputting the fused interaction features into the role classifier for role classification, the role classifier can be a lightweight MLP (Multi-Layer Perceptron) network that integrates metric learning technology. The MLP network calculates the feature similarity between the input fused interaction features and the preset role label features. Based on the similarity calculation results, the label with the highest similarity among all candidate role labels is selected as the final user role label.

[0035] In this embodiment, the user may be a new role or an existing role during the interaction with the intelligent agent system. Accordingly, during role tag matching, the interaction scenario can be identified, and specific role tags can be matched based on the identification results, thereby improving the efficiency of role tag matching. In one optional implementation of this embodiment, user role tags are obtained by matching role tags based on interaction data and interaction environment data, including: Interaction scenario identification based on interaction data and / or interaction environment data; For new interaction scenarios, the user's role tag is obtained by matching role tags in the full set of role tags of the intelligent agent system based on the interaction data and interaction environment data; for existing interaction scenarios, the user's role tag is obtained by matching role tags in the user's existing role tags based on the interaction data and interaction environment data.

[0036] Specifically, in the process of interactive scene recognition based on interactive data and / or interactive environment data, interactive scene recognition can be performed by a scene recognizer. The scene recognizer can be a lightweight SVM (Support Vector Machine) that integrates SFE (Scene Feature Encoding). By inputting interactive data and / or interactive environment data into the scene recognizer, the corresponding interactive environment features are obtained and compared with the user's existing interactive scene features. If the similarity is lower than the threshold, it is recognized as a new interactive scene. If the similarity is greater than or equal to the threshold, it is determined to be an existing interactive scene.

[0037] When a new interaction scenario is identified, feature extraction and fusion are performed on the interaction data and interaction environment data to obtain fused interaction features. Then, the fused interaction features are compared with the corresponding tag features of all role tags to select the role tag with the highest similarity as the user's current role tag. When an existing interaction scenario is identified, feature extraction and fusion are performed on the interaction data and interaction environment data to obtain fused interaction features. Then, the fused interaction features are compared with the corresponding tag features of the user's existing role tags to determine the final user role tag based on the similarity calculation results.

[0038] It should be noted that the process of obtaining user role tags by matching role tags based on interaction data and interaction environment data can be replaced by: obtaining user role tags by matching role tags in the full set of role tags of the intelligent agent system based on interaction data and interaction environment data; or, if it is a new interaction scenario, obtaining user role tags by matching role tags in the full set of role tags of the intelligent agent system based on interaction data and interaction environment data. Alternatively, it can be replaced with: matching user role tags with existing user role tags based on interaction data and interaction environment data; or, if it is an existing interaction scenario, matching user role tags with existing user role tags based on interaction data and interaction environment data.

[0039] If it is an existing interaction scenario, the interaction data and interaction environment data will be matched with the user's existing role tags.

[0040] In the specific implementation process, in order to improve the effectiveness of interaction data and interaction environment data, these two types of data can be standardized. Then, the user's role tags are matched based on the standardized data to obtain user role tags, thereby improving the accuracy of role tag matching. In an optional implementation provided in this embodiment, before obtaining user role tags by matching role tags based on interaction data and interaction environment data, the following steps are also included: Perform validity checks on interactive data and authenticity checks on interactive environment data.

[0041] The validity verification of interactive data includes: filtering abnormal characters in the interactive data; verifying the integrity and format compliance of the interactive data according to a preset baseline format; deleting interactive data whose content or format does not conform to the baseline format; and sending a re-entry prompt to the client after deletion to remind the user to re-enter valid interactive data. For example, if the interactive data uploaded by the user includes an unencrypted expense report file, and the preset baseline format is an encrypted file, then the expense report file will be deleted, and "Please re-enter" will be displayed on the client. After the user re-uploads the file, the validity verification of the interactive data will be performed again.

[0042] After validating the interaction data, the data that passes the validation can be standardized. Interaction data of the same type can be categorized and core features can be extracted. The parameter format and encoding of the data can be uniformly converted to form standardized interaction data, ensuring that the interaction data structure is uniform and the features are clear.

[0043] The verification of the authenticity of interactive environment data includes: verifying the authenticity of the interactive environment data in conjunction with the client's associated data; deleting interactive environment data that fails verification and sending a verification exception prompt to the client; associated data including client device identifier, interactive address information and / or enterprise intranet credentials; the server verifies the interactive environment data based on the associated data, filters and deletes false interactive environment data, and sends an interactive environment data verification exception prompt to the client after deleting the data, so as to prompt the user to confirm or correct the current interactive environment data. For example, if a user reports interactive environment data as an enterprise office scenario, but the system verifies that the user's interactive address information is a home address and does not have enterprise intranet credentials, the user's interactive environment data is determined to be false data, and the data is deleted. At the same time, the client is displayed "Please confirm whether the current scenario information is accurate, or supplement the enterprise intranet credentials". After the user confirms the scenario information or supplements the relevant credentials, the authenticity of the corrected interactive environment data is verified again.

[0044] In practical applications, users may have already interacted with the intelligent agent system before the current interaction, and the current user role label is different from the previous user role label. In this case, the role label can be switched to improve the user's awareness of the role label. In the specific execution process, after obtaining the user role label by matching the role label based on the interaction data and the interaction environment data, in an optional implementation of this embodiment, it also includes: returning the user role label to the client for role label switching.

[0045] Specifically, after the server completes the user role tag matching, it returns the user role tag to the client. After receiving the user role tag, the client performs role tag switching. Role tag switching includes: switching the interaction interface of the intelligent agent system to the role interaction interface corresponding to the user role tag, or switching the interaction interface of the intelligent agent system to the instance interaction interface of the role proxy instance, or updating the instance interaction interface of the previous role proxy instance of the intelligent agent system to the instance interaction interface of the role proxy instance.

[0046] For example, Figure 3 In the intelligent agent role dialogue interface shown, user A's role label in the previous interaction was "programmer". In the current interaction, based on user A's input of "remind the child to make arrangements", the matched role label is "parent". After the client receives the parent label, the intelligent agent system's interaction interface is switched from the interaction interface corresponding to the programmer label to the interaction interface corresponding to the parent label, and the system responds to the interaction content of "remind the child to make arrangements".

[0047] Furthermore, during the role tag switching process, it is also possible to detect whether the user tag is the same as the previous user role tag that matches the previous interaction data of the intelligent agent system to determine whether to switch the role tag. Specifically, in an optional implementation provided in this embodiment, if yes, the step of scheduling the interaction data to the role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system is executed; if no, the user role tag is returned to the client to perform user role switching.

[0048] In addition to the above-mentioned implementation methods for switching role tags, the interactive interface can also be updated. Specifically, after obtaining the user role tag by matching role tags based on interaction data and interaction environment data, an optional implementation method provided in this embodiment further includes: returning the user role tag to the client to update the interactive interface.

[0049] Specifically, after the server completes the user role tag matching, it returns the user role tag to the client. After receiving the user role tag, the client updates the interactive interface. The interactive interface update includes: updating the interactive interface of the intelligent agent system to the role interactive interface corresponding to the user role tag or the instance interactive interface of the role proxy instance, or updating the instance interactive interface of the previous role proxy instance to the instance interactive interface of the role proxy instance.

[0050] Similarly, during the process of updating the interactive interface, it is also possible to detect whether the user tag is the same as the previous user role tag that matched the previous interactive data of the intelligent agent system to determine whether to update the interactive interface. For example, it can detect whether the user role tag is the same as the previous user role tag that matched the previous interactive data of the intelligent agent system. If yes, execute the following steps to schedule the interactive data to the role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system. If no, return the user role tag to the client to update the interactive interface.

[0051] Step S206: Dispatch the interaction data to the role agent instance corresponding to the user role tag in the agent agent instance library of the agent system.

[0052] The agent proxy instance library stores multiple different role proxy instances, or the agent proxy instance library is composed of multiple different role proxy instances. In this case, the agent proxy instance library can be a collection of role proxy instances. Based on this, after obtaining the user role tag, the server determines the role proxy instance corresponding to the user role tag and schedules the user's input interaction data to the role proxy instance corresponding to the user role tag in the agent proxy instance library of the agent system. This makes the role proxy instance match the user's actual needs and improves the accuracy of agent interaction.

[0053] In the specific execution process, after determining the role proxy instance corresponding to the user, the interaction data needs to be scheduled to the role proxy instance for user role data retrieval and interaction response processing. Before or after scheduling the interaction data, the user's user configuration for the role proxy instance can be read and loaded into the role proxy instance. Here, the user configuration refers to the user's settings for the role proxy instance. For example, the user sets a simple role style for the programmer proxy instance to output results in a concise way.

[0054] In addition, a scheduling component can be set up or deployed in the intelligent agent system. Optionally, the interaction data is scheduled to the role agent instance by the scheduling component of the intelligent agent system. The scheduling component reads the user configuration for the role agent instance and loads the user configuration into the role agent. Here, the scheduling component refers to the executor deployed in the intelligent agent system to perform data scheduling and configuration loading. Specifically, the scheduling component can be used to schedule the interaction data input by the user on the client to the corresponding role agent instance, and can also be used to load the user configuration for the role agent instance into the corresponding role agent instance. Optionally, in actual applications, the same user may have one or more role tags. In this case, the scheduling component can also call the role agent instance corresponding to one or more user role tags.

[0055] In the specific execution process, after obtaining the user role tag, the scheduling component reads the user configuration for the role proxy instance and loads the user configuration into the corresponding role proxy instance so that the role proxy instance conforms to the user's settings, enhancing the user's personalized experience. It also schedules the user input interaction data submitted by the client to the role proxy instance corresponding to the user role tag, so that the role proxy instance can obtain the interaction data and provide a basis for subsequent interaction response processing.

[0056] It should be noted that the process of scheduling interactive data to the role agent instance corresponding to the user role tag in the agent agent instance library of the agent system can be replaced by: scheduling interactive data to the role agent instance corresponding to the user role tag in the agent agent instance library of the agent system through the scheduling component of the agent system.

[0057] In practical applications, when a user has multiple role tags, the interaction between the user and the intelligent agent system requires multiple role proxy instances to perform collaborative processing. In this case, the scheduling component of the intelligent agent system can call the role proxy instances corresponding to multiple user role tags to perform collaborative processing, thereby improving the efficiency of collaborative interaction and the flexibility of scenario adaptation.

[0058] In the specific execution process, after obtaining user role tags by matching role tags based on interaction data and interaction environment data, it is also possible to detect whether there are multiple user role tags. If so, the scheduling component of the intelligent agent system calls multiple role agent instances corresponding to user role tags for interactive and collaborative processing; If not, execute the step of scheduling the interaction data to the role agent instance corresponding to the user role tag in the agent agent instance library of the agent system.

[0059] Specifically, after obtaining the user role tag, the server counts the number of user role tags. If there is only one user role tag, the interaction data is scheduled to the role proxy instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system. If there are multiple user role tags, the scheduling component of the intelligent agent system calls multiple role proxy instances corresponding to the user role tags for interactive collaborative processing.

[0060] Furthermore, in an optional implementation of this embodiment, multiple user role tag corresponding role proxy instances are invoked for interactive collaborative processing, including: Perform intent recognition on the interaction data, and generate role interaction data corresponding to each user role tag based on the intent recognition results; The interaction data of each role is routed to the role proxy instance corresponding to each user role tag to retrieve user role data and process interaction responses to obtain role response data. The role response data is then fused to obtain interaction response data.

[0061] Among them, role interaction data refers to the data generated based on the intent recognition results for processing by each role agent instance. Role interaction data includes the data in the interaction data that is adapted to the corresponding role agent instance; role response data refers to the response data generated by the role agent instance after retrieving user role data in the corresponding role knowledge base and performing corresponding response processing based on the received corresponding role interaction data; and interaction response data refers to the final response result generated after integrating the role response data output by multiple role agent instances.

[0062] In the specific execution process, the interaction data is first subjected to intent recognition. Based on the results of intent recognition, role interaction data corresponding to each user role tag is generated. Each role interaction data is routed to the corresponding role proxy instance to retrieve user role data and process interaction response to obtain role response data. Then, the obtained role response data is fused to obtain interaction response data.

[0063] For example, a programmer might enter in the client, "Help me write a short message to post on social media, summarizing my work today and sharing my life." This programmer's interaction data contains both programmer-tagged and parent-tagged data. In this case, intent recognition of the interaction data can be achieved through an intent recognition module. Specifically, this module could be an intent recognition network using an attention mechanism to directly classify the interaction data and generate role-based interaction data. Alternatively, it could use a Transformer encoder to extract features from the interaction data and obtain interaction feature vectors. These feature vectors could then be input into an intent recognition network using a convolutional neural network architecture to generate role-based interaction data based on the intent recognition results. Through intent recognition, programmer interaction data corresponding to the programmer tag and parent interaction data corresponding to the parent tag are generated, allowing role proxy instances to obtain their corresponding role-based interaction data. Then, the programmer and parent proxy instances perform user role data retrieval and interaction response processing on their respective role-based interaction data to obtain their respective role-based response data. Finally, the server merges the role-based response data from both instances to obtain the final interaction response data.

[0064] It should be noted that the process of scheduling interactive data to the role agent instance corresponding to the user role tag in the agent agent instance library of the agent system can be replaced by: calling multiple role agent instances corresponding to user role tags for interactive collaborative processing through the scheduling component of the agent system; or, it can also be replaced by: detecting whether there are multiple user role tags, and if so, calling multiple role agent instances corresponding to user role tags for interactive collaborative processing through the scheduling component of the agent system.

[0065] Step S208: Retrieve user role data from the role knowledge base corresponding to the user role tag through the role proxy instance, and perform interactive response processing based on the user role data and interaction data to obtain the interaction result.

[0066] In practice, after the interaction data is dispatched to the role proxy instance, the role proxy instance retrieves the user role data in the role knowledge base corresponding to the user role tag. After retrieval, the role proxy instance performs interaction response processing based on the user role data and interaction data and obtains the interaction result. Here, the interaction result refers to the output result obtained by the role proxy instance based on the user role data and interaction data to respond to the user interaction. The interaction result is adapted to the user interaction scenario and usage habits. Specifically, the interaction result can be a text interaction result, a voice interaction result, or a document interaction result.

[0067] In practical applications, to improve the accuracy of interactive responses, increase data support dimensions, and enhance collaborative efficiency in group processing, it is also possible to retrieve associated user data and / or group role data located in the same interactive group as the user. In one optional implementation of this embodiment, user role data is retrieved from the role knowledge base corresponding to the user role tag, including: Retrieve the user's user role data from the role knowledge base, and retrieve the associated role data of the user's authorized credentials and / or the group role data of the interactive groups the user participates in from the role knowledge base.

[0068] Among them, authorization credentials refer to the user's permission to access associated user data. Authorization credentials can record the authorizing party, the scope of authorized data, and / or the validity period of the permission; associated users refer to other users who have a collaborative or management relationship with the current user and whose authorization credentials the current user has obtained; associated role data refers to the user role data of associated users, which is a necessary supplementary data for the current user to process interactive responses; and interaction groups refer to collaborative groups formed by the current user based on association relationships. Specifically, interaction groups can be family groups or project groups, and members in the group can participate in intelligent agent interactions and share group role data.

[0069] Specifically, the role proxy instance can not only retrieve the user's role data in the role knowledge base, but also retrieve the associated role data of the user's authorized credentials and the group role data of the interactive groups in which the user participates, thereby improving the collaborative efficiency of group processing.

[0070] In real-world scenarios, user-submitted interaction data often contains mixed or ambiguous intents. Directly processing this data may lead to misjudgment of the interaction intent, resulting in a disconnect between the interaction response and the user's actual needs. Therefore, intent recognition can be performed on the interaction data to improve the accuracy of intent recognition, thereby enhancing the accuracy of the interaction response. In one optional implementation of this embodiment, interaction response processing based on user role data and interaction data is used to obtain the interaction result, including: The interaction intent is obtained by performing intent recognition on the interaction data, and the interaction intent features obtained by feature extraction of the interaction data and interaction intent are fused with user configuration features and role data features to obtain fused features; The inference engine that matches the intent type of the interaction intent is determined and the standardized constraint parameters corresponding to the user configuration are loaded. The inference engine after the fused feature input is loaded is then used for interaction processing to obtain the interaction result.

[0071] The inference engine refers to the inference module that adapts to different interaction intents. Each inference engine has built-in processing logic adapted to the intent type. The standardized constraint parameters are parameters configured by the user to impose constraints on the inference engine. Users can control the output of the inference engine by configuring the standardized constraint parameters. For example, when the user configures the response to be simplified, the standardized constraint parameters are to output only the core result without any additional explanatory content. When the user configures the output to be in the form of an image, the standardized constraint parameters are to output only the image that reflects the result without any other types of output content.

[0072] Specifically, in the process of intent recognition, in order to improve the accuracy of intent recognition, the intent recognition module can be used to obtain the interactive intent from the interactive data. For example, the intent recognition module can be an intent recognition network using the Attention mechanism, which can directly classify the interactive data and obtain the interactive intent; or, the Transformer encoder can be used to extract features from the interactive data and obtain interactive feature vectors, and then input the interactive feature vectors into an intent recognition network using a CNN architecture to obtain the interactive intent.

[0073] During the feature extraction process, feature extraction and feature fusion can be performed using a feature extractor. The feature extractor can be a lightweight CNN (Convolutional Neural Network) that integrates NLP (Natural Language Processing). By inputting the interaction intent and interaction data into the CNN for feature extraction, interaction intent features are obtained. Then, the interaction intent features, user configuration features, and role data features are input into the CNN for feature fusion to obtain fused features.

[0074] Subsequently, the inference engine that matches the intent type of the interaction intent is determined and the standardized constraint parameters corresponding to the user configuration are loaded. The inference engine after the fused feature input is loaded is used for interaction processing to obtain the interaction result. The inference engine can be an inference module using the Transformer architecture or an inference module that integrates the Attention mechanism.

[0075] In practical applications, users may need to customize role proxy instances, adjust role data access permissions, or edit role-specific data during interaction with the intelligent agent system. To improve the user's personalized experience, this embodiment provides an optional implementation that allows for configuration of role proxy instances in the following manner: If a role configuration instruction is received from the client, query at least one user role tag for the user; Generate a configuration interface containing a configuration control for at least one role agent instance corresponding to a user role tag and return it to the client.

[0076] The role configuration command refers to the operation command initiated by the user in the client interaction interface to customize the role agent instance or user role label. For example, the role configuration command can be to configure the financial agent instance tool or to adjust the access scope of teaching data.

[0077] Specifically, after receiving the role configuration command submitted by the client, the server queries the user's role tags, matches the corresponding role proxy instance for the retrieved user role tags, generates a configuration interface for the role proxy instance's configuration controls, and returns it to the client so that the user can customize the configuration. For example, if a user has matched the finance role tag and clicks the role configuration button on the client to submit a role configuration command, after receiving the role configuration command, the server queries the user's user role tags, generates a configuration interface for the user's finance role tag that includes configuration controls for the finance role proxy instance, such as reimbursement rule template selection and file archive path settings, and returns the generated configuration interface to the client so that the user can customize the configuration.

[0078] Furthermore, in practical applications, since users may have already completed custom configurations for some role agent instances during their past use of the intelligent agent system, but as the scenario changes, users need to further adjust and optimize based on the existing configurations. Therefore, when the user triggers the configuration control, it is necessary to return the existing user configuration and role data identifier of the role agent instance to provide the user with a basis for adjustment. In an optional implementation method provided in this embodiment, it also includes: Based on the trigger command of any configuration control submitted by the client, query the user configuration of the role agent instance corresponding to any configuration control, and perform a role data query in the role knowledge base corresponding to any configuration control to obtain the role data identifier; return the user configuration and role data identifier to the client; wherein, the user configuration includes: the tool configuration deployed by the role agent instance and / or the role style configuration of the role agent instance; the role data identifier includes: the project identifier of the data project, the user identifier of the interactive user, and / or, the template identifier of the data template.

[0079] Based on this, if a user needs to edit the user role data corresponding to the role data identifier, an optional implementation of this embodiment further includes: Based on the access request for the target role data identifier submitted by the client, query the user role data corresponding to the target role data identifier in the role knowledge base corresponding to any configuration control and return it to the client so as to edit the user role data corresponding to the target role data identifier.

[0080] Specifically, when a user submits an access request for a target role data identifier, the server queries the target role data corresponding to the target role data identifier in the role knowledge base corresponding to any configuration control and returns it to the client, so that the user can edit the user role data corresponding to the target role data identifier.

[0081] For example, Figure 4 The role agent management interface shown contains a list of multiple role agent instances, including programmer, parent, and hybrid role agents. Users can configure role agent instances based on this list, specifically by triggering the corresponding agent instance in the list to select one. After the user selects a role agent instance, the client's interactive interface is displayed. Figure 5 The editing page shown allows users to edit the core capabilities of a role agent instance and personalize settings such as core capabilities, tone of voice, language style, and priority. Users can also edit user role data in the role knowledge base. Core capabilities include: project progress tracking, automatic email generation, data analysis reports, meeting minutes organization, and customer relationship organization.

[0082] In specific implementation, after obtaining the interaction result, in order to ensure that the returned interaction result conforms to the user's current user role tag, it is necessary to detect the user role tag to improve the accuracy of the interaction and the user's experience. In an optional implementation provided in this embodiment, after obtaining the interaction result by performing interaction response processing based on user role data and interaction data, it further includes: Detect whether the user role label matches the user role label in the previous interaction data of the intelligent agent system's interaction interface; if yes, return the interaction result to the client; if no, write the role switching field of the user role label into the interaction result and return it to the client.

[0083] The role switching field is a switching instruction written by the server when it determines that the current user role label is inconsistent with the previous user role label. After the client receives the role switching field, it will switch to the corresponding interactive interface to ensure that the client interface is adapted to the user's current user role label. The role switching field includes a switching status indicator and / or current role label information. The switching status indicator includes yes and no, and the current role label information includes the current user role label. For example, when the client receives a switching status indicator of yes and the current role label information is the finance label, the client interface will be switched to the interactive interface corresponding to the finance personnel.

[0084] Specifically, after obtaining the interaction result, the user role tag is checked to determine whether it is the same as the user role tag that matched the previous interaction data. If they are the same, the server returns the obtained interaction result to the client; otherwise, the role switching field is written into the interaction result and returned to the client.

[0085] In summary, the intelligent agent interaction processing method for the server side of an intelligent agent system provided in this embodiment first obtains the user's input interaction data submitted by the client and the user's interaction environment data during the interaction between the user and the intelligent agent, providing data support for subsequent role tag matching. Then, role tag matching is performed based on the interaction data and interaction environment data to obtain the user's role tag. Next, the corresponding role proxy instance is determined based on the user's role tag, and the interaction data is scheduled to the role proxy instance in the intelligent agent system's intelligent agent proxy instance library. Finally, the user's role data is retrieved from the role knowledge base corresponding to the user's role tag through the role proxy instance, and the interaction response is processed based on the user's role data and interaction data to obtain the interaction result, thereby improving the personalized experience of the user's interaction with the intelligent agent.

[0086] Furthermore, in the process of retrieving user role data, in order to improve the accuracy of interactive responses, increase the dimensions of data support, and enhance the collaborative efficiency of group processing, it is also possible to retrieve associated user data related to the user and / or group role data located in the same interactive group as the user; thereby, based on the retrieved role data and interactive data, interactive response processing can be performed to improve the collaborative efficiency and accuracy of the group.

[0087] Steps S202 to S208 provided in this embodiment can be executed by the server of the intelligent agent system. It should be noted that the steps S202 to S208 executed by the server of the intelligent agent system and steps S702 to S706 executed by the client of the intelligent agent in the following embodiment can cooperate with each other during execution. Therefore, when reading this embodiment, please refer to the corresponding content of steps S702 to S706 provided in the following method embodiment, and when reading the following method embodiment, please refer to the corresponding content of steps S202 to S208 provided in this embodiment.

[0088] The following example uses an intelligent agent interaction processing method provided in this embodiment applied to the server side of an intelligent agent system, taking an application based on user roles for interaction as an example. Figure 6 The intelligent agent interaction processing method provided in this embodiment will be further explained below. Figure 6 The intelligent agent interaction processing method applied to intelligent agent systems specifically includes the following steps.

[0089] Step S606: Obtain the interaction data submitted by the client of the intelligent agent system and the user's interaction environment data.

[0090] Step S608: Extract features from the interaction data and interaction environment data and fuse the features to obtain fused interaction features.

[0091] Step S610: Perform interactive scene recognition based on interactive environment data and interactive data.

[0092] Step S612: Based on the interaction data and interaction environment data, perform role tag matching on the user's existing role tags to obtain the user's role tags.

[0093] Step S614: Based on the interaction data and interaction environment data, perform role tag matching in the full set of role tags of the intelligent agent system to obtain the user role tag.

[0094] Step S616: Return the user role label to the client.

[0095] Step S620: The interaction data is scheduled to the role agent instance corresponding to the user role tag in the agent agent instance library of the agent system through the scheduling component.

[0096] Step S622: The role proxy instance retrieves the group role data of the interaction group to which the user belongs from the role knowledge base.

[0097] Step S624: Perform interactive response processing based on user role data, group role data, and interaction data, and obtain the interaction results.

[0098] Step S626: Return the interaction result to the client.

[0099] Steps S606 to S616 and S620 to S626 provided in this embodiment are executed by the server of the intelligent agent system. It should be noted that the steps S606 to S616 and S620 to S626 executed by the server of the intelligent agent system can cooperate with the steps S602, S604, S618, and S628 executed by the client in the following embodiments during execution. Therefore, when reading this embodiment, please refer to the corresponding content of steps S602, S604, S618, and S628 provided in the following method embodiments. When reading the following method embodiments, please refer to the corresponding content of steps S606 to S616 and S620 to S626 provided in this embodiment.

[0100] It should be noted that any one or more of steps S606 to S616 and steps S620 to S626 can be combined with any one or more of steps S202 to S208 to form a new implementation method according to the needs of implementation and deployment. In addition, any one or more technical features in steps S606 to S616 and steps S620 to S626 can be selected and combined with any one or more technical features provided in steps S202 to S208 to form a new implementation method according to the actual deployment needs. Alternatively, any one or more technical features in steps S606 to S616 and steps S620 to S626 can also be replaced with any one or more technical features provided in steps S202 to S208 to form a new implementation method according to the actual deployment needs. These will not be elaborated on here.

[0101] One or more embodiments of another intelligent agent interaction processing method provided in this specification are as follows: Reference Figure 7 The intelligent agent interaction processing method provided in this embodiment can be applied to the client of the intelligent agent system. The method specifically includes steps S702 to S706.

[0102] Step S702: Obtain the interaction data input by the user in the interaction interface of the intelligent agent system and submit it to the server.

[0103] The intelligent agent system described in this embodiment refers to an intelligent agent or intelligent agent system capable of interacting based on different user roles. Specifically, the intelligent agent system includes a server and a client. The client is used to interact with the user, specifically including acquiring the interaction data input by the user in the client's interaction interface and submitting it to the server, receiving the user role tags returned by the server and switching the role tags, and receiving the interaction results returned by the server and displaying them interactively in the interaction interface. The server is used to process interactions based on different user roles, specifically including matching role tags according to the acquired interaction data and interaction environment data to obtain user role tags, and scheduling the interaction data to the role proxy instance corresponding to the user role tag in the intelligent agent proxy instance library for interaction response processing.

[0104] The agent proxy instance library is used to store role proxy instances for different roles of the corresponding user. Alternatively, the agent proxy instance library can also be composed of role proxy instances for different roles of the corresponding user. In this case, the agent proxy instance library can be a collection of role proxy instances. Specifically, each user may have multiple different roles, that is, a single user may have multiple different role tags. Correspondingly, the agent proxy instance library stores the role proxy instances corresponding to each of the user's roles.

[0105] A role proxy instance refers to an executor deployed or configured by an intelligent agent system to interact with users based on different user roles. Specifically, a role proxy instance can be an intelligent agent or a large language model corresponding to each role label, a sub-intelligent agent of an intelligent agent or a sub-model of a large language model corresponding to each role label, or a proxy instance of an intelligent agent or a large language model corresponding to each role label. A user role label refers to the label of the user's actual identity role, identity type, or interaction scenario when interacting with the intelligent agent system. In this case, the user role label can be an identity type label or an interaction scenario label.

[0106] The interaction data described in this embodiment includes the interaction content input by the user on the client's interactive interface. Specifically, the interaction data can be the interaction content of the user interacting with the intelligent agent in different scenarios, such as the user's interaction content in a work scenario, the user's interaction content in a life scenario, or the user's interaction content in other scenarios. Optionally, the interaction data can be text interaction data, image interaction data, voice interaction data, or operation-type interaction data, such as the user's click command on the interactive interface or the user's configuration trigger command. The interaction environment data refers to the environmental data related to the user's interaction. Specifically, the interaction environment data can include device environment data, such as the client's device type and the client's network communication method, and can also include scene environment data, such as the user's interaction time, the user's interaction location, and / or the user's interaction application.

[0107] In practice, during the interaction between the user and the intelligent agent system, the user inputs interactive data through the interactive interface of the intelligent agent system. The client obtains the interactive data input by the user and submits it to the server. Correspondingly, the server receives the interactive data submitted by the client and obtains the user's interactive environment data, providing a data foundation for subsequent role tag matching based on the interactive data and interactive environment data.

[0108] It should be noted that, to ensure a one-to-one correspondence between role proxy instances and role tags and to accurately respond to the interaction needs of the corresponding roles, the role proxy instances need to be trained and generated. Furthermore, since the training data for role proxy instances comes from user role data, data anonymization and differential privacy computation can be used to improve the security of user data during the training process. Specifically, in one optional implementation method provided in this embodiment, the role proxy instances in the agent proxy instance library are obtained in the following way: Privacy-preserving role data is obtained by performing data anonymization processing and differential privacy calculation on the user role data in the role knowledge base corresponding to any user role tag. The meta-agent agent instance of the intelligent agent system is trained based on privacy role data to obtain the role agent instance corresponding to any user role label.

[0109] The role knowledge base refers to a database deployed in the intelligent agent system that stores the data content corresponding to each user role tag. There is a one-to-one mapping relationship between the role knowledge base and the user role tags, and a one-to-one mapping relationship between the user role tags and the role agent instances. That is, there is a one-to-one mapping relationship between the role knowledge base, the role agent instances, and the user role tags. User role data is stored in the role knowledge base. Specifically, user role data is the data generated when a user interacts with the corresponding role agent instance in the current interactive identity. Optionally, user role data includes the user's historical interaction data.

[0110] Meta-agent proxy instances are the basic models used in an agent system to generate proxy instances for each role. Specifically, a meta-agent proxy instance can be an agent or a large language model, such as a base large language model using the Transformer architecture, or it can be a sub-agent of an agent or a sub-model of a large language model, or it can be a proxy instance of an agent or a large language model.

[0111] Specifically, during the process of data desensitization and differential privacy computation of user role data, a privacy processor can be used for data desensitization and differential privacy computation. The privacy processor can be a lightweight CNN (Convolutional Neural Network) that integrates DD (Data Desensitization) and DP (Differential Privacy) algorithms. By inputting user role data into the CNN for data desensitization and differential privacy computation, privacy role data is obtained.

[0112] During the training of meta-agent agent instances of an intelligent agent system based on privacy role data, the meta-agent agent instances can be trained using a model trainer. The model trainer can be an LSTM (Long Short-Term Memory) network integrating LoRA (Low-Rank Adaptation). By inputting privacy role data into the LSTM, the low-rank parameters of the meta-agent agent instances are adjusted to obtain the role agent instances corresponding to any role label.

[0113] In the process of obtaining the role proxy instance described above, data anonymization and differential privacy calculation can be performed on the user role data in the role knowledge base corresponding to any user role tag to obtain privacy role data. The privacy role data is used as the input for training the meta-agent proxy instance to obtain the role proxy instance. In an optional implementation method provided in this embodiment, the user role data in the role knowledge base is obtained in the following way: Read the user interaction records and corresponding interaction environment records of the intelligent agent system; perform role tag matching on the user interaction records and interaction environment records to obtain user role tags, and store the user interaction records in the role knowledge base corresponding to the obtained user role tags.

[0114] Among them, user interaction records include the user's historical interaction data. Specifically, user interaction records can be records of the user's interaction content with the intelligent agent in different scenarios, such as user interaction content records in work scenarios, user interaction content records in life scenarios, and / or user interaction content records in other scenarios; interaction environment records include the user's historical interaction environment data. Specifically, historical interaction environment data can be device environment data.

[0115] Specifically, by reading the user's interaction records with the intelligent agent system and the corresponding interaction environment records, and performing role tag matching on the read user interaction records and interaction environment records to obtain user role tags, the user interaction records are stored in the role knowledge base corresponding to the user role tags after obtaining the user role tags, thereby obtaining user role data in the role knowledge base.

[0116] Step S704: Switch the role tags of the interactive interface according to the user role tags returned by the server to obtain the role interactive interface; the user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data.

[0117] As mentioned above, a user role tag refers to the tag indicating the user's actual identity, role type, or interaction scenario when interacting with the intelligent agent system. In this case, the user role tag can be an identity type tag or an interaction scenario tag. In real-world scenarios, the same user may have different identity types when interacting with the intelligent agent system in different scenarios. For example, in a work scenario, user A's role tag might be a programmer tag, while in a life scenario, user A's role tag might be a parent tag. Optionally, the user role tag corresponds one-to-one with the above role proxy instances, and each user role tag corresponds to one of the user's interaction identity types. Since each user has at least one interaction identity type, each user has at least one user role tag.

[0118] In practice, based on the user's input interaction data and interaction environment data, the user's role tags are matched according to the interaction data and interaction environment data to obtain the user's role tags. This provides accurate role basis for subsequently determining the role proxy instance corresponding to the user's role tags and scheduling interaction data to the corresponding role proxy instance.

[0119] In the specific execution process, after acquiring interaction data and interaction environment data, role tag matching is performed based on the interaction data and interaction environment data. In one optional implementation method provided in this embodiment, the user role tag is obtained by performing role tag matching based on interaction data and interaction environment data, including: Feature extraction and feature fusion are performed on interactive data and interactive environment data to obtain fused interactive features; The fused interaction features are input into the role classifier for role classification to obtain user role tags; Specifically, in the process of feature extraction and feature fusion of interactive data and interactive environment data, feature extraction and feature fusion can be performed through a feature extractor. The feature extractor can be a lightweight CNN (Convolutional Neural Network) that integrates NLP (Natural Language Processing). By inputting interactive data and interactive environment data into the CNN for feature extraction and feature fusion, the corresponding fused interactive features are obtained. In the process of inputting the fused interaction features into the role classifier for role classification, the role classifier can be a lightweight MLP (Multi-Layer Perceptron) network that integrates metric learning technology. The MLP network calculates the feature similarity between the input fused interaction features and the preset role label features. Based on the similarity calculation results, the label with the highest similarity among all candidate role labels is selected as the final user role label.

[0120] In this embodiment, the user may be a new role or an existing role during the interaction with the intelligent agent system. Accordingly, during role tag matching, the interaction scenario can be identified, and specific role tags can be matched based on the identification results, thereby improving the efficiency of role tag matching. In one optional implementation of this embodiment, user role tags are obtained by matching role tags based on interaction data and interaction environment data, including: Interaction scenario identification based on interaction data and / or interaction environment data; For new interaction scenarios, the user's role tag is obtained by matching role tags in the full set of role tags of the intelligent agent system based on the interaction data and interaction environment data; for existing interaction scenarios, the user's role tag is obtained by matching role tags in the user's existing role tags based on the interaction data and interaction environment data.

[0121] Specifically, in the process of interactive scene recognition based on interactive data and / or interactive environment data, interactive scene recognition can be performed by a scene recognizer. The scene recognizer can be a lightweight SVM (Support Vector Machine) that integrates SFE (Scene Feature Encoding). By inputting interactive data and / or interactive environment data into the scene recognizer, the corresponding interactive environment features are obtained and compared with the user's existing interactive scene features. If the similarity is lower than the threshold, it is recognized as a new interactive scene. If the similarity is greater than or equal to the threshold, it is determined to be an existing interactive scene.

[0122] When a new interaction scenario is identified, feature extraction and fusion are performed on the interaction data and interaction environment data to obtain fused interaction features. Then, the fused interaction features are compared with the corresponding tag features of all role tags to select the role tag with the highest similarity as the user's current role tag. When an existing interaction scenario is identified, feature extraction and fusion are performed on the interaction data and interaction environment data to obtain fused interaction features. Then, the fused interaction features are compared with the corresponding tag features of the user's existing role tags to determine the final user role tag based on the similarity calculation results.

[0123] It should be noted that the process of obtaining user role tags by matching role tags based on interaction data and interaction environment data can be replaced by: obtaining user role tags by matching role tags in the full set of role tags of the intelligent agent system based on interaction data and interaction environment data; or, if it is a new interaction scenario, obtaining user role tags by matching role tags in the full set of role tags of the intelligent agent system based on interaction data and interaction environment data. Alternatively, it can be replaced with: matching user role tags with existing user role tags based on interaction data and interaction environment data; or, if it is an existing interaction scenario, matching user role tags with existing user role tags based on interaction data and interaction environment data.

[0124] If it is an existing interaction scenario, the interaction data and interaction environment data will be matched with the user's existing role tags.

[0125] In the specific implementation process, in order to improve the effectiveness of interaction data and interaction environment data, these two types of data can be standardized. Then, the user's role tags are matched based on the standardized data to obtain user role tags, thereby improving the accuracy of role tag matching. In an optional implementation provided in this embodiment, before obtaining user role tags by matching role tags based on interaction data and interaction environment data, the following steps are also included: Perform validity checks on interactive data and authenticity checks on interactive environment data.

[0126] The validity verification of interactive data includes: filtering abnormal characters in the interactive data; verifying the integrity and format compliance of the interactive data according to a preset baseline format; deleting interactive data whose content or format does not conform to the baseline format; and sending a re-entry prompt to the client after deletion to remind the user to re-enter valid interactive data. For example, if the interactive data uploaded by the user includes an unencrypted expense report file, and the preset baseline format is an encrypted file, then the expense report file will be deleted, and "Please re-enter" will be displayed on the client. After the user re-uploads the file, the validity verification of the interactive data will be performed again.

[0127] After validating the interaction data, the data that passes the validation can be standardized. Interaction data of the same type can be categorized and core features can be extracted. The parameter format and encoding of the data can be uniformly converted to form standardized interaction data, ensuring that the interaction data structure is uniform and the features are clear.

[0128] The verification of the authenticity of interactive environment data includes: verifying the authenticity of the interactive environment data in conjunction with the client's associated data; deleting interactive environment data that fails verification and sending a verification exception prompt to the client; associated data including client device identifier, interactive address information and / or enterprise intranet credentials; the server verifies the interactive environment data based on the associated data, filters and deletes false interactive environment data, and sends an interactive environment data verification exception prompt to the client after deleting the data, so as to prompt the user to confirm or correct the current interactive environment data. For example, if a user reports interactive environment data as an enterprise office scenario, but the system verifies that the user's interactive address information is a home address and does not have enterprise intranet credentials, the user's interactive environment data is determined to be false data, and the data is deleted. At the same time, the client is displayed "Please confirm whether the current scenario information is accurate, or supplement the enterprise intranet credentials". After the user confirms the scenario information or supplements the relevant credentials, the authenticity of the corrected interactive environment data is verified again.

[0129] In practical applications, users may have already interacted with the intelligent agent system before the current interaction, and the current user role label is different from the previous user role label. In this case, the role label can be switched to improve the user's awareness of the role label. In the specific execution process, after obtaining the user role label by matching the role label based on the interaction data and the interaction environment data, in an optional implementation of this embodiment, it also includes: returning the user role label to the client for role label switching.

[0130] Specifically, after the server completes the user role tag matching, it returns the user role tag to the client. After receiving the user role tag, the client performs role tag switching. Role tag switching includes: switching the interaction interface of the intelligent agent system to the role interaction interface corresponding to the user role tag, or switching the interaction interface of the intelligent agent system to the instance interaction interface of the role proxy instance, or updating the instance interaction interface of the previous role proxy instance of the intelligent agent system to the instance interaction interface of the role proxy instance.

[0131] For example, Figure 3 In the intelligent agent role dialogue interface shown, user A's role label in the previous interaction was programmer. In the current interaction, based on user A's input of "remind the child to make arrangements", the matched role label is parent. After the client receives the parent label, the intelligent agent system's interaction interface is switched from the interaction interface corresponding to the programmer label to the interaction interface corresponding to the parent label.

[0132] Furthermore, during the role tag switching process, it is also possible to detect whether the user tag is the same as the previous user role tag that matches the previous interaction data of the intelligent agent system to determine whether to switch the role tag. Specifically, in an optional implementation provided in this embodiment, if yes, the step of scheduling the interaction data to the role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system is executed; if no, the user role tag is returned to the client to perform user role switching.

[0133] In addition to the above-mentioned implementation methods for switching role tags, the interactive interface can also be updated. Specifically, after obtaining the user role tag by matching role tags based on interaction data and interaction environment data, an optional implementation method provided in this embodiment further includes: returning the user role tag to the client to update the interactive interface.

[0134] Specifically, after the server completes the user role tag matching, it returns the user role tag to the client. After receiving the user role tag, the client updates the interactive interface. The interactive interface update includes: updating the interactive interface of the intelligent agent system to the role interactive interface corresponding to the user role tag or the instance interactive interface of the role proxy instance, or updating the instance interactive interface of the previous role proxy instance to the instance interactive interface of the role proxy instance.

[0135] Similarly, during the process of updating the interactive interface, it is also possible to detect whether the user tag is the same as the previous user role tag that matched the previous interactive data of the intelligent agent system to determine whether to update the interactive interface. For example, it can detect whether the user role tag is the same as the previous user role tag that matched the previous interactive data of the intelligent agent system. If yes, execute the following steps to schedule the interactive data to the role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system. If no, return the user role tag to the client to update the interactive interface.

[0136] Step S706: Receive the interaction result returned by the server and display it interactively on the role interaction interface; the interaction result is obtained by retrieving user role data and processing interaction responses in the role knowledge base corresponding to the user role tag through the role proxy instance corresponding to the user role tag.

[0137] In practice, after the interaction data is dispatched to the role proxy instance, the role proxy instance retrieves the user role data in the role knowledge base corresponding to the user role tag. After retrieval, the role proxy instance performs interaction response processing based on the user role data and interaction data and obtains the interaction result. Here, the interaction result refers to the output result obtained by the role proxy instance based on the user role data and interaction data to respond to the user interaction. The interaction result is adapted to the user interaction scenario and usage habits. Specifically, the interaction result can be a text interaction result, a voice interaction result, or a document interaction result.

[0138] In practical applications, to improve the accuracy of interactive responses, increase data support dimensions, and enhance collaborative efficiency in group processing, it is also possible to retrieve associated user data and / or group role data located in the same interactive group as the user. In one optional implementation of this embodiment, user role data is retrieved from the role knowledge base corresponding to the user role tag, including: Retrieve the user's user role data from the role knowledge base, and retrieve the associated role data of the user's authorized credentials and / or the group role data of the interactive groups the user participates in from the role knowledge base.

[0139] Among them, authorization credentials refer to the user's permission to access associated user data. Authorization credentials can record the authorizing party, the scope of authorized data, and / or the validity period of the permission; associated users refer to other users who have a collaborative or management relationship with the current user and whose authorization credentials the current user has obtained; associated role data refers to the user role data of associated users, which is a necessary supplementary data for the current user to process interactive responses; and interaction groups refer to collaborative groups formed by the current user based on association relationships. Specifically, interaction groups can be family groups or project groups, and members in the group can participate in intelligent agent interactions and share group role data.

[0140] Specifically, the role proxy instance can not only retrieve the user's role data in the role knowledge base, but also retrieve the associated role data of the user's authorized credentials and the group role data of the interactive groups in which the user participates, thereby improving the collaborative efficiency of group processing.

[0141] In real-world scenarios, user-submitted interaction data often contains mixed or ambiguous intents. Directly processing this data may lead to misjudgment of the interaction intent, resulting in a disconnect between the interaction response and the user's actual needs. Therefore, intent recognition can be performed on the interaction data to improve the accuracy of intent recognition, thereby enhancing the accuracy of the interaction response. In one optional implementation of this embodiment, interaction response processing based on user role data and interaction data is used to obtain the interaction result, including: The interaction intent is obtained by performing intent recognition on the interaction data, and the interaction intent features obtained by feature extraction of the interaction data and interaction intent are fused with user configuration features and role data features to obtain fused features; The inference engine that matches the intent type of the interaction intent is determined and the standardized constraint parameters corresponding to the user configuration are loaded. The inference engine after the fused feature input is loaded is then used for interaction processing to obtain the interaction result.

[0142] The inference engine refers to the inference module that adapts to different interaction intents. Each inference engine has built-in processing logic adapted to the intent type. The standardized constraint parameters are parameters configured by the user to impose constraints on the inference engine. Users can control the output of the inference engine by configuring the standardized constraint parameters. For example, when the user configures the response to be simplified, the standardized constraint parameters are to output only the core result without any additional explanatory content. When the user configures the output to be in the form of an image, the standardized constraint parameters are to output only the image that reflects the result without any other types of output content.

[0143] Specifically, in the process of intent recognition, in order to improve the accuracy of intent recognition, the intent recognition module can be used to obtain the interactive intent from the interactive data. For example, the intent recognition module can be an intent recognition network using the Attention mechanism, which can directly classify the interactive data and obtain the interactive intent; or, the Transformer encoder can be used to extract features from the interactive data and obtain interactive feature vectors, and then input the interactive feature vectors into an intent recognition network using a CNN architecture to obtain the interactive intent.

[0144] During the feature extraction process, feature extraction and feature fusion can be performed using a feature extractor. The feature extractor can be a lightweight CNN (Convolutional Neural Network) that integrates NLP (Natural Language Processing). By inputting the interaction intent and interaction data into the CNN for feature extraction, interaction intent features are obtained. Then, the interaction intent features, user configuration features, and role data features are input into the CNN for feature fusion to obtain fused features.

[0145] Subsequently, the inference engine that matches the intent type of the interaction intent is determined and the standardized constraint parameters corresponding to the user configuration are loaded. The inference engine after the fused feature input is loaded is used for interaction processing to obtain the interaction result. The inference engine can be an inference module using the Transformer architecture or an inference module that integrates the Attention mechanism.

[0146] In practical applications, users may need to customize role proxy instances, adjust role data access permissions, or edit role-specific data during interaction with the intelligent agent system. To improve the user's personalized experience, this embodiment provides an optional implementation that allows for configuration of role proxy instances in the following manner: If a role configuration instruction is received from the client, query at least one user role tag for the user; Generate a configuration interface containing a configuration control for at least one role agent instance corresponding to a user role tag and return it to the client.

[0147] The role configuration command refers to the operation command initiated by the user in the client interaction interface to customize the role agent instance or user role label. For example, the role configuration command can be to configure the financial agent instance tool or to adjust the access scope of teaching data.

[0148] Specifically, after receiving the role configuration command submitted by the client, the server queries the user's role tags, matches the corresponding role proxy instance for the retrieved user role tags, generates a configuration interface for the role proxy instance's configuration controls, and returns it to the client so that the user can customize the configuration. For example, if a user has matched the finance role tag and clicks the role configuration button on the client to submit a role configuration command, after receiving the role configuration command, the server queries the user's user role tags, generates a configuration interface for the user's finance role tag that includes configuration controls for the finance role proxy instance, such as reimbursement rule template selection and file archive path settings, and returns the generated configuration interface to the client so that the user can customize the configuration.

[0149] Furthermore, in practical applications, since users may have already completed custom configurations for some role agent instances during their past use of the intelligent agent system, but as the scenario changes, users need to further adjust and optimize based on the existing configurations. Therefore, when the user triggers the configuration control, it is necessary to return the existing user configuration and role data identifier of the role agent instance to provide the user with a basis for adjustment. In an optional implementation method provided in this embodiment, it also includes: Based on the trigger command of any configuration control submitted by the client, query the user configuration of the role agent instance corresponding to any configuration control, and perform a role data query in the role knowledge base corresponding to any configuration control to obtain the role data identifier; return the user configuration and role data identifier to the client; wherein, the user configuration includes: the tool configuration deployed by the role agent instance and / or the role style configuration of the role agent instance; the role data identifier includes: the project identifier of the data project, the user identifier of the interactive user, and / or, the template identifier of the data template.

[0150] Based on this, if a user needs to edit the user role data corresponding to the role data identifier, an optional implementation of this embodiment further includes: Based on the access request for the target role data identifier submitted by the client, query the user role data corresponding to the target role data identifier in the role knowledge base corresponding to any configuration control and return it to the client so as to edit the user role data corresponding to the target role data identifier.

[0151] Specifically, when a user submits an access request for a target role data identifier, the server queries the target role data corresponding to the target role data identifier in the role knowledge base corresponding to any configuration control and returns it to the client, so that the user can edit the user role data corresponding to the target role data identifier.

[0152] For example, Figure 4The role agent management interface shown contains multiple role agent instances in its list. Users can configure role agent instances based on this list, specifically by triggering the corresponding agent instance in the list to select one. After the user selects a role agent instance, the client's interactive interface will be displayed. Figure 5 The editing page shown allows users to edit the core capabilities of a role's proxy instance and personalize settings such as tone of voice, language style, and priority. Users can also edit user role data in the role knowledge base.

[0153] In specific implementation, after obtaining the interaction result, in order to ensure that the returned interaction result conforms to the user's current user role tag, it is necessary to detect the user role tag to improve the accuracy of the interaction and the user's experience. In an optional implementation provided in this embodiment, after obtaining the interaction result by performing interaction response processing based on user role data and interaction data, it further includes: Detect whether the user role label matches the user role label in the previous interaction data of the intelligent agent system's interaction interface; if yes, return the interaction result to the client; if no, write the role switching field of the user role label into the interaction result and return it to the client.

[0154] The role switching field is a switching instruction written by the server when it determines that the current user role label is inconsistent with the previous user role label. After the client receives the role switching field, it will switch to the corresponding interactive interface to ensure that the client interface is adapted to the user's current user role label. The role switching field includes a switching status indicator and / or current role label information. The switching status indicator includes yes and no, and the current role label information includes the current user role label. For example, when the client receives a switching status indicator of yes and the current role label information is the finance label, the client interface will be switched to the interactive interface corresponding to the finance personnel.

[0155] Specifically, after obtaining the interaction result, the user role tag is checked to determine whether it is the same as the user role tag that matched the previous interaction data. If they are the same, the server returns the obtained interaction result to the client; otherwise, the role switching field is written into the interaction result and returned to the client.

[0156] The following example uses the application of an intelligent agent interaction processing method provided in this embodiment to a client-side intelligent agent system as an example, combined with... Figure 6 The intelligent agent interaction processing method provided in this embodiment will be further explained below. Figure 6 This method for intelligent agent interaction based on user roles includes the following steps: Step S602: Collect the interactive data input by the user on the client's interactive interface.

[0157] Step S604: Submit interaction data to the server of the intelligent agent system.

[0158] Step S618: Receive the user role tags sent by the server and switch the role tags on the interactive interface.

[0159] Step S628: Receive the interaction results sent by the server and display them interactively on the interactive interface.

[0160] It should be noted that any one or more of steps S602, S604, S618, and S628 can be combined with any one or more of steps S702 to S706 to form a new implementation method according to the needs of implementation and deployment. In addition, any one or more technical features in steps S602, S604, S618, and S628 can be selected and combined with any one or more technical features provided in steps S702 to S706 to form a new implementation method according to the actual deployment needs. Alternatively, any one or more technical features in steps S602, S604, S618, and S628 can be replaced with any one or more technical features provided in steps S702 to S706 to form a new implementation method according to the actual deployment needs. These will not be elaborated on here.

[0161] This specification provides an embodiment of an intelligent agent interaction processing device as follows: In the above embodiments, an intelligent agent interaction processing method is provided, and correspondingly, an intelligent agent interaction processing device is also provided, which will be described below with reference to the accompanying drawings.

[0162] Reference Figure 8 This illustration shows a schematic diagram of an embodiment of an intelligent agent interaction processing device provided in this embodiment.

[0163] Since the apparatus embodiments correspond to the method embodiments, the descriptions are relatively simple. For relevant parts, please refer to the corresponding descriptions of the method embodiments provided above. The apparatus embodiments described below are merely illustrative.

[0164] This embodiment provides an intelligent agent interaction processing device, the device comprising: The data acquisition module 802 is configured to acquire interactive data of user input submitted by the client, and to acquire the user's interactive environment data; The tag matching module 804 is configured to perform role tag matching based on the interaction data and the interaction environment data to obtain user role tags; The data scheduling module 806 is configured to schedule the interactive data to the role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system; The interaction response processing module 808 is configured to retrieve user role data from the role knowledge base corresponding to the user role tag through the role proxy instance, and perform interaction response processing based on the user role data and the interaction data to obtain the interaction result.

[0165] For ease of description, the above devices are described by dividing them into various modules or units based on their functions. Of course, when implementing one or more of these specifications, the functions of each module or unit can be implemented in the same or different software and / or hardware, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0166] Another embodiment of the intelligent agent interaction processing device provided in this specification is as follows: In the above embodiments, another intelligent agent interaction processing method is provided, and correspondingly, another intelligent agent interaction processing device is also provided, which will be described below with reference to the accompanying drawings.

[0167] Reference Figure 9 This illustration shows a schematic diagram of another embodiment of the intelligent agent interaction processing device provided in this embodiment.

[0168] Since the apparatus embodiments correspond to the method embodiments, the descriptions are relatively simple. For relevant parts, please refer to the corresponding descriptions of the method embodiments provided above. The apparatus embodiments described below are merely illustrative.

[0169] This embodiment provides an intelligent agent interaction processing device, the device comprising: The data submission module 902 is configured to acquire interactive data input by the user at the interactive interface of the intelligent agent system and submit it to the server. The tag switching module 904 is configured to switch the role tags of the interactive interface according to the user role tags returned by the server, and obtain a role interactive interface; the user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data. The interactive display module 906 is configured to receive the interactive results returned by the server and display them interactively on the role interaction interface; the interactive results are obtained by the role proxy instance corresponding to the user role tag in the role knowledge base corresponding to the user role tag through user role data retrieval and interactive response processing.

[0170] For ease of description, the above devices are described by dividing them into various modules or units based on their functions. Of course, when implementing one or more of these specifications, the functions of each module or unit can be implemented in the same or different software and / or hardware, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0171] This specification provides an embodiment of an intelligent agent interaction processing device as follows: Corresponding to the agent interaction processing method described above, based on the same technical concept, one or more embodiments of this specification also provide an agent interaction processing device, which is used to execute the agent interaction processing method provided above. Figure 10 This is a schematic diagram of the structure of an intelligent agent interaction processing device provided for one or more embodiments of this specification.

[0172] This embodiment provides an intelligent agent interaction processing device, including: like Figure 10As shown, device 1000 mainly consists of a communication interface 1002, a user interface 1004, a processor 1006, and a data storage 1008. These components are interconnected and communicate with each other via a system bus, network, or other connection mechanism 1010. The communication interface 1002 enables device 1000 to communicate with other devices, access networks, and transmission networks via analog or digital modulation. For example, the communication interface 1002 may include a chipset and antenna for wireless communication with a radio access network or access point. Furthermore, the communication interface 1002 can be a wired interface such as Ethernet, Token Ring, or a USB port, or a wireless interface such as Wi-Fi, Bluetooth, Global Positioning System (GPS), or a wide-area wireless interface (e.g., WiMAX or LTE). Of course, the communication interface 1002 can also support other forms of physical layer interfaces and standard or proprietary communication protocols. The communication interface 1002 may also include multiple physical communication interfaces, such as Wi-Fi, Bluetooth, and wide-area wireless interfaces. The user interface 1004 includes receiving user input and providing output to the user. Therefore, the user interface 1004 may include input components such as a keypad, keyboard, touch-sensitive or presence-sensitive panel, computer mouse, trackball, joystick, microphone, still camera, and video camera, and output components such as a display screen (which may be combined with a touch-sensitive panel), CRT, LCD, LED, display using DLP technology, printer, and other similar devices known or developed in the future. The user interface 1004 may also generate auditory output via speakers, speaker jacks, audio output ports, audio output devices, headphones, and other similar devices known or developed in the future. In some embodiments, the user interface 1004 may include software, circuitry, or other forms of logic capable of transmitting data to and receiving data from external user input / output devices. Additionally or alternatively, the device 1000 may support remote access from other devices via communication interface 1002 or another physical interface (not shown). The user interface 1004 may be configured to receive user input, the position and movement of which may be indicated by indicators or cursors described herein. The user interface 1004 may also be configured as a display device for rendering or displaying text fragments.

[0173] Processor 1006 may include one or more general-purpose processors and / or dedicated processors. Data storage 1008 may include one or more volatile and / or non-volatile storage components, and may be integrated wholly or partially with processor 1006. Data storage 1008 may include removable and non-removable components.

[0174] Processor 1006 is capable of executing program instructions 1018 (e.g., compiled or uncompiled program logic and / or machine code) stored in data storage 1008 to perform the various functions described herein. Data storage 1008 may contain a non-transitory computer-readable medium on which program instructions are stored, which, when executed by device 1000, enable device 1000 to perform any methods, processes, or functions disclosed in this specification and / or the accompanying drawings. Execution of program instructions 1018 by processor 1006 may result in processor 1006 using data 1012. For example, program instructions 1018 may include an operating system 1022 (e.g., an operating system kernel, device drivers, and / or other modules) installed on device 1000 and one or more application programs 1020 (e.g., a browser, social application, or game application). Similarly, data 1012 may include operating system data 1016 and application data 1014. Operating system data 1016 is primarily accessible to operating system 1022, while application data 1014 is primarily accessible to one or more application programs 1020. Application data 1014 may reside in a file system visible or hidden to the user of device 1000. Application 1020 may communicate with operating system 1012 via one or more application programming interfaces (APIs). These APIs facilitate application 1020 reading and / or writing application data 1014, transmitting or receiving information via communication interface 1002, receiving or displaying information on user interface 1004, etc. In some terms, application 1020 may be simply referred to as an "app". Furthermore, application 1020 may be downloaded to device 1000 through one or more online app stores or app markets. However, applications may also be installed on device 1000 in other ways, such as through a web browser or a physical interface on device 1000 (e.g., a USB port).

[0175] In one specific embodiment, the intelligent agent interaction processing device includes a memory and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer-executable instructions for the intelligent agent interaction processing device, and is configured to be executed by one or more processors. The one or more programs include computer-executable instructions for performing the following: Obtain the user input interaction data submitted by the client, and obtain the user's interaction environment data; User role tags are obtained by matching role tags based on the interaction data and the interaction environment data. The interactive data is dispatched to the role agent instance corresponding to the user role tag in the intelligent agent system's intelligent agent agent instance library; The role proxy instance retrieves user role data from the role knowledge base corresponding to the user role tag, and performs interactive response processing based on the user role data and the interaction data to obtain the interaction result.

[0176] Another embodiment of the intelligent agent interaction processing device provided in this specification is as follows: Corresponding to the other intelligent agent interaction processing method described above, based on the same technical concept, one or more embodiments of this specification also provide another intelligent agent interaction processing device, which is used to execute the other intelligent agent interaction processing method provided above. Figure 11 This is a schematic diagram of the structure of another intelligent agent interaction processing device provided in one or more embodiments of this specification.

[0177] This embodiment provides an intelligent agent interaction processing device, including: like Figure 11As shown, device 1100 mainly consists of a communication interface 1102, a user interface 1104, a processor 1106, and a data storage 1108. These components are interconnected and communicate with each other via a system bus, network, or other connection mechanism 1110. The communication interface 1102 enables device 1100 to communicate with other devices, access networks, and transmission networks via analog or digital modulation. For example, the communication interface 1102 may include a chipset and antenna for wireless communication with a radio access network or access point. Furthermore, the communication interface 1102 can be a wired interface such as Ethernet, Token Ring, or a USB port, or a wireless interface such as Wi-Fi, Bluetooth, Global Positioning System (GPS), or a wide-area wireless interface (e.g., WiMAX or LTE). Of course, the communication interface 1102 can also support other forms of physical layer interfaces and standard or proprietary communication protocols. The communication interface 1102 may also include multiple physical communication interfaces, such as Wi-Fi, Bluetooth, and wide-area wireless interfaces. The user interface 1104 includes receiving user input and providing output to the user. Therefore, user interface 1104 may include input components such as a keypad, keyboard, touch-sensitive or presence-sensitive panel, computer mouse, trackball, joystick, microphone, still camera, and video camera, and output components such as a display screen (which may be combined with a touch-sensitive panel), CRT, LCD, LED, display using DLP technology, printer, and other similar devices known or developed in the future. User interface 1104 may also generate auditory output via speakers, speaker jacks, audio output ports, audio output devices, headphones, and other similar devices known or developed in the future. In some embodiments, user interface 1104 may include software, circuitry, or other forms of logic capable of transmitting data to and receiving data from external user input / output devices. Additionally or alternatively, device 1100 may support remote access from other devices via communication interface 1102 or another physical interface (not shown). User interface 1104 may be configured to receive user input, the position and movement of which may be indicated by indicators or cursors described herein. User interface 1104 may also be configured as a display device for rendering or displaying text fragments.

[0178] Processor 1106 may include one or more general-purpose processors and / or dedicated processors. Data storage 1108 may include one or more volatile and / or non-volatile storage components and may be integrated wholly or partially with processor 1106. Data storage 1108 may include removable and non-removable components.

[0179] Processor 1106 is capable of executing program instructions 1118 (e.g., compiled or uncompiled program logic and / or machine code) stored in data store 1108 to perform the various functions described herein. Data store 1108 may contain a non-transitory computer-readable medium on which program instructions are stored, which, when executed by device 1100, enable device 1100 to perform any methods, processes, or functions disclosed in this specification and / or the accompanying drawings. Execution of program instructions 1118 by processor 1106 may result in processor 1106 using data 1112. For example, program instructions 1118 may include an operating system 1122 (e.g., an operating system kernel, device drivers, and / or other modules) installed on device 1100 and one or more application programs 1120 (e.g., a browser, social application, or game application). Similarly, data 1112 may include operating system data 1116 and application data 1114. Operating system data 1116 is primarily accessible to operating system 1122, while application data 1114 is primarily accessible to one or more application programs 1120. Application data 1114 may reside in a file system visible or hidden from the user of device 1100. Application 1120 may communicate with operating system 1112 via one or more application programming interfaces (APIs). These APIs facilitate application 1120 reading and / or writing application data 1114, transmitting or receiving information via communication interface 1102, receiving or displaying information on user interface 1104, etc. In some terms, application 1120 may be simply referred to as an "app". Furthermore, application 1120 may be downloaded to device 1100 through one or more online app stores or app markets. However, applications may also be installed on device 1100 in other ways, such as through a web browser or a physical interface on device 1100 (e.g., a USB port).

[0180] In one specific embodiment, the intelligent agent interaction processing device includes a memory and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer-executable instructions for the intelligent agent interaction processing device, and is configured to be executed by one or more processors. The one or more programs include computer-executable instructions for performing the following: Obtain the interaction data input by the user in the interactive interface of the intelligent agent system and submit it to the server; The user role tags are switched on the interactive interface according to the user role tags returned by the server to obtain a role-based interactive interface; the user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data. The system receives the interaction results returned by the server and displays them interactively on the role interaction interface. The interaction results are obtained by the role proxy instance corresponding to the user role tag in the role knowledge base corresponding to the user role tag through user role data retrieval and interaction response processing.

[0181] This specification provides an embodiment of a computer-readable storage medium as follows: Corresponding to the intelligent agent interaction processing method described above, based on the same technical concept, one or more embodiments of this specification also provide a computer-readable storage medium.

[0182] The computer-readable storage medium provided in this embodiment is used to store computer-executable instructions, which, when executed, implement the following process: Obtain the user input interaction data submitted by the client, and obtain the user's interaction environment data; User role tags are obtained by matching role tags based on the interaction data and the interaction environment data. The interactive data is dispatched to the role agent instance corresponding to the user role tag in the intelligent agent system's intelligent agent agent instance library; The role proxy instance retrieves user role data from the role knowledge base corresponding to the user role tag, and performs interactive response processing based on the user role data and the interaction data to obtain the interaction result.

[0183] It should be noted that the embodiments of a computer-readable storage medium described in this specification and the embodiments of an intelligent agent interaction processing method described in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can be referred to the implementation of the corresponding method described above, and the repeated parts will not be described again.

[0184] Another embodiment of a computer-readable storage medium provided in this specification is as follows: Corresponding to the other intelligent agent interaction processing method described above, based on the same technical concept, one or more embodiments of this specification also provide another computer-readable storage medium.

[0185] The computer-readable storage medium provided in this embodiment is used to store computer-executable instructions, which, when executed, implement the following process: Obtain the interaction data input by the user in the interactive interface of the intelligent agent system and submit it to the server; The user role tags are switched on the interactive interface according to the user role tags returned by the server to obtain a role-based interactive interface; the user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data. The system receives the interaction results returned by the server and displays them interactively on the role interaction interface. The interaction results are obtained by the role proxy instance corresponding to the user role tag in the role knowledge base corresponding to the user role tag through user role data retrieval and interaction response processing.

[0186] It should be noted that the embodiments of another computer-readable storage medium described in this specification and the embodiments of another intelligent agent interaction processing method described in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can be referred to the implementation of the corresponding method described above, and the repeated parts will not be described again.

[0187] This specification provides an example of a computer program product as follows: Corresponding to the intelligent agent interaction processing method described above, based on the same technical concept, one or more embodiments of this specification also provide a computer program product.

[0188] A computer program product includes a computer program / instructions that, when executed by a processor, perform the following steps: Obtain the user input interaction data submitted by the client, and obtain the user's interaction environment data; User role tags are obtained by matching role tags based on the interaction data and the interaction environment data. The interactive data is dispatched to the role agent instance corresponding to the user role tag in the intelligent agent system's intelligent agent agent instance library; The role proxy instance retrieves user role data from the role knowledge base corresponding to the user role tag, and performs interactive response processing based on the user role data and the interaction data to obtain the interaction result.

[0189] It should be noted that the embodiments of a computer program product described in this specification and the embodiments of an intelligent agent interaction processing method described in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can be referred to the implementation of the corresponding method described above, and the repeated parts will not be described again.

[0190] Another example of a computer program product provided in this specification is as follows: Corresponding to the other intelligent agent interaction processing method described above, based on the same technical concept, one or more embodiments of this specification also provide another computer program product.

[0191] A computer program product includes a computer program / instructions that, when executed by a processor, perform the following steps: Obtain the interaction data input by the user in the interactive interface of the intelligent agent system and submit it to the server; The user role tags are switched on the interactive interface according to the user role tags returned by the server to obtain a role-based interactive interface; the user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data. The system receives the interaction results returned by the server and displays them interactively on the role interaction interface. The interaction results are obtained by the role proxy instance corresponding to the user role tag in the role knowledge base corresponding to the user role tag through user role data retrieval and interaction response processing.

[0192] It should be noted that the embodiments of another computer program product described in this specification and the embodiments of another intelligent agent interaction processing method described in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can be referred to the implementation of the corresponding method described above, and the repeated parts will not be described again.

[0193] The various embodiments in this specification are described in a progressive manner. For the same or similar parts between the various embodiments, please refer to each other. Each embodiment focuses on describing the differences from other embodiments. For example, the device embodiments, equipment embodiments, computer-readable storage medium embodiments, and computer program product embodiments are all similar to the method embodiments, so the descriptions are relatively simple. For reading the relevant content of the device embodiments, equipment embodiments, computer-readable storage medium embodiments, and computer program product embodiments, please refer to the description of the method embodiments.

[0194] While one or more embodiments of this specification provide method steps as described in the embodiments or flowcharts, it is understood that the order of steps listed in the embodiments or flowcharts is merely one possible execution order among many steps, and does not represent the only execution order. Therefore, when the claims involve method steps, any changes or adjustments to the order of such steps, or the parallelism between steps, are also within the scope of protection of the claims. This specification uses specific terms to describe embodiments of this specification. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic related to at least one embodiment of this specification. Therefore, it should be emphasized and noted that "an embodiment," "one embodiment," or "an alternative embodiment" mentioned twice or more in different locations in this specification do not necessarily refer to the same embodiment. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0195] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0196] In the 1930s, improvements to a technology could be clearly distinguished as either hardware improvements (e.g., improvements to the circuit structure of diodes, transistors, switches, etc.) or software improvements (improvements to the methodology). However, with technological advancements, many improvements to the methodology today can be considered direct improvements to the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved methodology into the hardware circuit. Therefore, it cannot be said that an improvement to the methodology cannot be implemented using hardware physical modules. For example, a Programmable Logic Device (PLD) (such as a Field Programmable Gate Array (FPGA)) is such an integrated circuit whose logic function is determined by the user programming the device. Designers can program and "integrate" a digital system onto a PLD themselves, without needing chip manufacturers to design and manufacture dedicated integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing integrated circuit chips, this programming is mostly implemented using "logic compiler" software. Similar to the software compiler used in program development, the original code before compilation must also be written in a specific programming language, called a Hardware Description Language (HDL). There are many HDLs, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, and RHDL (Ruby Hardware Description Language). Currently, the most commonly used are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also understand that by simply performing some logic programming on the method flow using one of these hardware description languages ​​and programming it into an integrated circuit, the hardware circuit implementing the logical method flow can be easily obtained.

[0197] The controller can be implemented in any suitable manner. For example, it can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicon Labs C8051F320. A memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also recognize that, in addition to implementing the controller in purely computer-readable program code form, the same functionality can be achieved by logically programming the method steps to make the controller take the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the means included therein for implementing various functions can also be considered as structures within the hardware component. Alternatively, the means for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0198] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0199] For ease of description, the above apparatus is described by dividing it into various functional units. Of course, when implementing the embodiments of this specification, the functions of each unit can be implemented in one or more software and / or hardware.

[0200] Those skilled in the art will understand that one or more embodiments of this specification can be provided as a method, system, or computer program product. Therefore, one or more embodiments of this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0201] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0202] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0203] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0204] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0205] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0206] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer-readable storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0207] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising at least one…" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0208] One or more embodiments of this specification can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a particular task or implement a particular abstract data type. One or more embodiments of this specification can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0209] The above description is merely an embodiment of this document and is not intended to limit the scope of this document. Various modifications and variations can be made to this document by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this document should be included within the scope of the claims of this document.

Claims

1. A method for intelligent agent interaction processing, applied to the server side of an intelligent agent system, the method comprising: Obtain the user input interaction data submitted by the client, and obtain the user's interaction environment data; User role tags are obtained by matching role tags based on the interaction data and the interaction environment data. The interactive data is dispatched to the role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system; The role proxy instance retrieves user role data from the role knowledge base corresponding to the user role tag, and performs interactive response processing based on the user role data and the interaction data to obtain the interaction result.

2. The intelligent agent interaction processing method according to claim 1, after the step of obtaining user role tags by matching role tags based on the interaction data and the interaction environment data is executed, and before the step of scheduling the interaction data to the role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system is executed, it further includes: Return the user role label to the client for role label switching; The role tag switching includes: switching the interactive interface of the intelligent agent system to the role interactive interface corresponding to the user role tag or the instance interactive interface of the role proxy instance, or updating the instance interactive interface of the previous role proxy instance to the instance interactive interface of the role proxy instance.

3. The intelligent agent interaction processing method according to claim 1, wherein the interaction data is scheduled to the role agent instance by the scheduling component of the intelligent agent system; The scheduling component reads the user configuration for the role proxy instance and loads the user configuration into the role proxy instance.

4. The intelligent agent interaction processing method according to claim 3, wherein the step of obtaining the interaction result by performing interaction response processing based on the user role data and the interaction data includes: The interaction data is subjected to intent recognition to obtain the interaction intent, and the interaction intent features obtained by feature extraction of the interaction data and the interaction intent are fused with user configuration features and role data features to obtain fused features; The inference engine that matches the intent type of the interaction intent is determined and the standardized constraint parameters corresponding to the user configuration are loaded. The fused features are then input into the loaded inference engine for interaction processing to obtain the interaction result.

5. The intelligent agent interaction processing method according to claim 1, wherein retrieving user role data from the role knowledge base corresponding to the user role tag includes: Retrieve the user's user role data in the role knowledge base, and retrieve the associated role data of the associated user corresponding to the user's authorization credentials in the role knowledge base; And / or, retrieve group role data of the interactive groups in which the user participates from the role knowledge base.

6. The intelligent agent interaction processing method according to claim 1, after the step of obtaining user role tags by matching role tags based on the interaction data and the interaction environment data is executed, and before the step of scheduling the interaction data to the role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system is executed, it further includes: Detect whether the number of user role tags is multiple; If so, the scheduling component of the intelligent agent system calls multiple role agent instances corresponding to user role tags for interactive collaborative processing; If not, proceed with the step of scheduling the interaction data to the role agent instance corresponding to the user role tag in the agent agent instance library of the agent system.

7. The intelligent agent interaction processing method according to claim 6, wherein calling multiple user role tag corresponding role proxy instances for interactive collaborative processing includes: The interaction data is subjected to intent recognition, and role interaction data corresponding to each user role tag is generated based on the intent recognition results; The interaction data of each role is routed to the role proxy instance corresponding to each user role tag to retrieve user role data and process interaction response to obtain role response data, and the role response data is then fused to obtain interaction response data.

8. The agent interaction processing method according to claim 1, wherein the role agent instances in the agent agent instance library are obtained in the following manner: Privacy-preserving role data is obtained by performing data anonymization processing and differential privacy calculation on the user role data in the role knowledge base corresponding to any user role tag. The meta-agent agent instance of the intelligent agent system is trained based on the privacy role data to obtain the role agent instance corresponding to any user role label; the meta-agent agent instance includes the base large language model.

9. The intelligent agent interaction processing method according to claim 1, wherein the user role data in the role knowledge base is obtained in the following manner: Read the user interaction records and corresponding interaction environment records of the intelligent agent system; The user interaction records and the interaction environment records are matched with role tags to obtain user role tags, and the user interaction records are stored in the role knowledge base corresponding to the obtained user role tags.

10. The intelligent agent interaction processing method according to claim 1, further comprising: If a role configuration instruction submitted by the client is received, query at least one user role tag of the user; A configuration interface containing configuration controls for the role proxy instances corresponding to the at least one user role tag is generated and returned to the client.

11. The intelligent agent interaction processing method according to claim 10, further comprising: Based on the trigger command of any configuration control submitted by the client, query the user configuration of the role agent instance corresponding to the configuration control, and perform a role data query in the role knowledge base corresponding to the configuration control to obtain the role data identifier; The user configuration and the role data identifier are returned to the client.

12. The intelligent agent interaction processing method according to claim 11, wherein the user configuration includes: Tool configuration for deploying role proxy instances and / or role style configuration for role proxy instances; The role data identifiers include: the project identifier of the data project, the user identifier of the interacting user, and / or the template identifier of the data template.

13. The intelligent agent interaction processing method according to claim 11, further comprising: Based on the access request for the target role data identifier submitted by the client, the user role data corresponding to the target role data identifier is queried in the role knowledge base corresponding to any configuration control and returned to the client, so as to edit the user role data corresponding to the target role data identifier.

14. The intelligent agent interaction processing method according to claim 1, after the step of obtaining the interaction result by performing interaction response processing based on the user role data and the interaction data is executed, it further includes: Detect whether the user role tag is the same as the user role tag matched in the previous interaction data of the interaction interface of the intelligent agent system; If so, return the interaction result to the client; If not, write the role switching field of the user role tag into the interaction result and return it to the client.

15. A method for intelligent agent interaction processing, the method comprising: Obtain the interaction data input by the user in the interactive interface of the intelligent agent system and submit it to the server; Based on the user role tags returned by the server, the interactive interface is switched to obtain a role-based interactive interface; The user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data; The system receives the interaction results returned by the server and displays them interactively on the role interaction interface. The interaction results are obtained by the role proxy instance corresponding to the user role tag in the role knowledge base corresponding to the user role tag through user role data retrieval and interaction response processing.

16. The intelligent agent interaction processing method according to claim 15, wherein switching the role tags of the interaction interface based on the user role tags returned by the server to obtain a role interaction interface includes: Switch the interactive interface of the intelligent agent system to the role interactive interface corresponding to the user role tag or the instance interactive interface of the role proxy instance, or update the instance interactive interface of the previous role proxy instance to the instance interactive interface of the role proxy instance.

17. An intelligent agent interaction processing device, the device comprising: The data acquisition module is configured to acquire interactive data of user input submitted by the client, and to acquire the user's interactive environment data; The tag matching module is configured to perform role tag matching based on the interaction data and the interaction environment data to obtain user role tags; The data scheduling module is configured to schedule the interactive data to the role agent instance corresponding to the user role tag in the intelligent agent agent instance library of the intelligent agent system; The interaction response processing module is configured to retrieve user role data from the role knowledge base corresponding to the user role tag through the role proxy instance, and perform interaction response processing based on the user role data and the interaction data to obtain the interaction result.

18. An intelligent agent interaction processing device, the device comprising: The data submission module is configured to acquire interactive data input by the user in the interactive interface of the intelligent agent system and submit it to the server. The tag switching module is configured to switch the role tags of the interactive interface according to the user role tags returned by the server, and obtain a role interactive interface; The user role tags are obtained by matching role tags based on the interaction data and the user's interaction environment data; The interactive display module is configured to receive the interactive results returned by the server and display them interactively on the role interaction interface; the interactive results are obtained by the role proxy instance corresponding to the user role tag in the role knowledge base corresponding to the user role tag to retrieve user role data and process interactive responses.

19. An intelligent agent interaction processing device, comprising: processor; And, a memory configured to store computer-executable instructions, which, when executed, cause the processor to: Obtain the user input interaction data submitted by the client, and obtain the user's interaction environment data; User role tags are obtained by matching role tags based on the interaction data and the interaction environment data. The interactive data is dispatched to the role agent instance corresponding to the user role tag in the intelligent agent system's intelligent agent agent instance library; The role proxy instance retrieves user role data from the role knowledge base corresponding to the user role tag, and performs interactive response processing based on the user role data and the interaction data to obtain the interaction result.

20. A computer-readable storage medium for storing computer-executable instructions that, when executed, implement the steps of the method of claim 1 or 15.

Images