Intelligent agent creating method, interactive processing method based on intelligent agent and related device

Abstract

The embodiment of the invention provides an agent creating method, an agent-based interactive processing method and a related device.The agent creating method comprises the steps that after a server receives a creating request, sent by a user terminal, of creating an agent of a service category in an application program by a user, the agent is created according to the creating request; according to the agent template corresponding to the service category, the agent of the user under the service category is created, and after it is detected that the agent is created, agent access data is returned to the user terminal, and the user terminal can enter a subprogram of agent access of the agent according to the agent access data. Therefore, the creation of the intelligent agent of the service category in the application program by the user is realized.

Description

Technical Field

[0001] This document relates to the field of artificial intelligence technology, and in particular, to an agent creation method, an interaction processing method based on an agent, and related devices. Background Art

[0002] With the continuous development of Internet technology, agent conversations bring great convenience to users. Users can perform data queries, text generation, and even code generation through agent conversations. With the development of agent conversations, users can quickly obtain the required data in the form of agent conversations without complex and cumbersome queries and processing, improving the user's processing efficiency and enhancing the user's mental ability. How to provide users with more accurate and effective agent conversations is the key point that users and agent providers increasingly focus on. Summary of the Invention

[0003] One or more embodiments of this specification provide an agent creation method, which is applied to a server. The method includes: receiving a creation request sent by a user terminal for creating an agent of a service category in an application program. The creation request is submitted based on user occupation data corresponding to the service category. In response to the creation request, an agent for the user in the service category is created according to an agent template corresponding to the service category. After detecting the agent creation, agent access data is returned to the user terminal. The user terminal enters a subroutine for accessing the agent according to the agent access data.

[0004] One or more embodiments of this specification provide an agent creation method, which is applied to a user terminal. The method includes: obtaining a creation request for creating an agent of a service category in an application program by a user and sending it to a server. The creation request is submitted based on user occupation data corresponding to the service category. An interactive display for the agent creation is performed according to interactive materials corresponding to the user category and identity materials generated based on the user occupation data. According to the agent access data returned by the server after the agent creation, a subroutine for accessing the agent is entered. The agent is created according to an agent template corresponding to the service category.

[0005] One or more embodiments of this specification provide an agent-based interaction processing method, which is applied to an agent under a service category created by a user in an application. The method includes: obtaining interaction data entered by the user through the agent access page. The agent access page is displayed through a subroutine for the agent access of the agent. The agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent. The agent is created according to the agent template corresponding to the service category. The interaction data is input into the large language model in the agent to generate response data, and the response data is obtained. The response data is displayed through the agent access page.

[0006] One or more embodiments of this specification provide an agent creation device, which runs on a server. The device includes: a creation request receiving module, configured to receive a creation request for creating an agent of a service category in an application sent by a user terminal. The creation request is submitted based on user occupation data corresponding to the service category. An agent creation module, configured to create an agent for the user under the service category according to the agent template corresponding to the service category in response to the creation request. After detecting the creation of the agent, a running data return module is run. The running data return module is configured to return agent access data to the user terminal after detecting the creation of the agent. The user terminal enters a subroutine for the agent access of the agent according to the agent access data.

[0007] One or more embodiments of this specification provide an agent creation device, which runs on a user terminal. The device includes: a creation request sending module, configured to obtain a creation request for creating an agent of a service category in an application by the user and send it to the server. The creation request is submitted based on user occupation data corresponding to the service category. An interaction display module, configured to perform an interaction display for the agent creation according to interaction materials corresponding to the user category and identity materials generated based on the user occupation data. A subroutine entry module, configured to enter a subroutine for the agent access of the agent according to the agent access data returned by the server after the agent is created. The agent is created according to the agent template corresponding to the service category.

[0008] One or more embodiments of this specification provide an agent-based interaction processing device that runs on an agent. The device includes: an interaction data acquisition module configured to acquire interaction data entered by a user through the agent accessing a page. The agent accessing the page is presented through a subroutine for the agent access of the agent. The agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent. The agent is created according to an agent template corresponding to the service category. A response data generation module configured to input the interaction data into a large language model in the agent to generate response data and obtain the response data. A response data presentation module configured to present the response data through the agent accessing the page.

[0009] One or more embodiments of this specification provide a server, including: a processor; and a memory configured to store computer-executable instructions, where the computer-executable instructions, when executed, cause the processor to: receive a creation request sent by a user terminal for creating an agent of a service category in an application. The creation request is submitted based on user occupation data corresponding to the service category. In response to the creation request, create an agent for the user in the service category according to an agent template corresponding to the service category. After detecting the creation of the agent, return agent access data to the user terminal. The user terminal enters a subroutine for the agent access of the agent according to the agent access data.

[0010] One or more embodiments of this specification provide a user terminal, including: a processor; and a memory configured to store computer-executable instructions, where the computer-executable instructions, when executed, cause the processor to: acquire a creation request for creating an agent of a service category by a user in an application and send it to a server. The creation request is submitted based on user occupation data corresponding to the service category. Perform an interactive display for the creation of the agent according to interaction materials corresponding to the user category and identity materials generated based on the user occupation data. According to the agent access data returned by the server after the agent is created, enter a subroutine for the agent access of the agent. The agent is created according to an agent template corresponding to the service category.

[0011] One or more embodiments of this specification provide an agent-based interaction processing device, including: a processor; and a memory configured to store computer-executable instructions, which, when executed, cause the processor to: obtain interaction data entered by a user through an agent accessing a page. The agent access page is presented through a subroutine for the agent access of the agent. The agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent. The agent is created according to an agent template corresponding to the service category. Input the interaction data into a large language model in the agent to generate response data and obtain the response data. Present the response data through the agent access page.

[0012] One or more embodiments of this specification provide a first computer-readable storage medium for storing computer-executable instructions, which, when executed, implement the following process: receive a creation request sent by a user terminal for creating an agent of a service category in an application. The creation request is submitted based on user occupation data corresponding to the service category. In response to the creation request, create an agent for the user under the service category according to an agent template corresponding to the service category. After detecting the creation of the agent, return agent access data to the user terminal. The user terminal enters a subroutine for the agent access of the agent according to the agent access data.

[0013] One or more embodiments of this specification provide a second computer-readable storage medium for storing computer-executable instructions, which, when executed, implement the following process: obtain a creation request for creating an agent of a service category by a user in an application and send it to a server. The creation request is submitted based on user occupation data corresponding to the service category. Perform an interactive display for the creation of the agent according to interaction materials corresponding to the user category and identity materials generated based on the user occupation data. According to the agent access data returned by the server after the agent is created, enter a subroutine for the agent access of the agent. The agent is created according to an agent template corresponding to the service category.

[0014] One or more embodiments of this specification provide a third computer-readable storage medium for storing computer-executable instructions, which, when executed, implement the following process: Obtain interaction data entered by a user through an agent accessing a page. The agent access page is presented through a subroutine for the agent to access the agent. The agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent. The agent is created according to an agent template corresponding to the service category. Input the interaction data into a large language model in the agent to generate response data, and obtain the response data. Present the response data through the agent access page. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] To more clearly illustrate the technical solutions in one or more embodiments of this specification or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments described in this specification. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings. Figure 1 Schematic diagram of an implementation environment of an agent creation method provided by one or more embodiments of this specification; Figure 2 Processing flow chart of an agent creation method provided by one or more embodiments of this specification; Figure 3 Schematic diagram of a subroutine list page provided by one or more embodiments of this specification; Figure 4 Schematic diagram of a service access page provided by one or more embodiments of this specification; Figure 5 Schematic diagram of a creation reminder page provided by one or more embodiments of this specification; Figure 6 Schematic diagram of an interaction display page provided by one or more embodiments of this specification; Figure 7 Schematic diagram of an agent access page provided by one or more embodiments of this specification; Figure 8 Schematic diagram of an agent dialogue page provided by one or more embodiments of this specification; Figure 9 Schematic diagram of an agent provided by one or more embodiments of this specification; Figure 10 Processing timing diagram of an agent creation method applied to the creation of a fitness intelligent assistant provided by one or more embodiments of this specification; Figure 11 A sequence diagram for an agent creation method applied to a training agent creation scenario provided for one or more embodiments of this specification; Figure 12 Another flowchart of an agent creation method provided for one or more embodiments of this specification; Figure 13 A flowchart of an interaction processing method based on an agent provided for one or more embodiments of this specification; Figure 14 A schematic diagram of an embodiment of an agent creation device provided for one or more embodiments of this specification; Figure 15 Another schematic diagram of an embodiment of an agent creation device provided for one or more embodiments of this specification; Figure 16 A schematic diagram of an embodiment of an interaction processing device based on an agent provided for one or more embodiments of this specification; Figure 17 A schematic diagram of the structure of a server provided for one or more embodiments of this specification; Figure 18 A schematic diagram of the structure of a user terminal provided for one or more embodiments of this specification; Figure 19 A schematic diagram of the structure of an interaction processing device based on an agent provided for one or more embodiments of this specification. Detailed implementation manners

[0016] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the following will clearly and completely describe the technical solutions in one or more embodiments of this specification in conjunction with the accompanying drawings in one or more embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this document.

[0017] The agent creation method provided for one or more embodiments of this specification is applicable to an implementation environment where a user creates an agent of a service category in an application program. Referring to Figure 1 , this implementation environment at least includes: a server 101 and a user terminal 102; Among them, the server 101 can be a single server, or a server cluster composed of several servers, or one or more cloud servers in a cloud computing platform; it is used for agent creation; The user terminal 102 can specifically be a mobile phone, a personal computer, a tablet computer, an e-reader, a device for information interaction based on VR (Virtual Reality), a vehicle-mounted terminal, an IoT device, a wearable intelligent device, a laptop computer, a desktop computer, etc.; it is used to submit a creation request for an agent to the server 101, and interact with the agent through a subroutine for agent access after the server 101 creates the agent. In this implementation environment, the user submits a creation request for creating an agent of a service category to the server 101 through the user occupation data corresponding to the service category in the application program running on the user terminal 102. After the server 101 receives the creation request for creating an agent of this service category submitted by the user terminal based on the user occupation data corresponding to the service category in the application program, it creates an agent for the user under the service category according to the agent template corresponding to the service category. After the agent is created, it returns agent access data to the user terminal 102, and the user terminal 102 enters the subroutine for agent access according to the agent access data; during the process of the server 101 creating the agent, in order to improve the user's perception, the user terminal 102 can also perform an interactive display of agent creation according to the interactive materials corresponding to the user category and the identity materials generated based on the user occupation data, so as to create an agent for the user under the service category in the application program and perform service processing corresponding to the service category based on this agent.

[0018] It should be noted that considering that relevant data such as user identification data and user occupation data involved in this specification may, to a certain extent, belong to the user's privacy, therefore, if you want to collect relevant data such as the user's user identification data and user occupation data, you can obtain the user's authorization before collecting the data to make the data collection operation comply with relevant data management regulations. For example, data authorization can be performed during the process of the user submitting a creation request for creating an agent of a service category, or during the process of the user first submitting user identification data and user occupation data to the server, or during the process of the user accessing the application program; the specific method of data authorization can be to send a data authorization reminder to the user, and the user can obtain data collection authorization after confirming the reminder through an instruction, or the data authorization method can also be to obtain data collection authorization by signing a data authorization agreement; this embodiment will not elaborate here.

[0019] One or more embodiments of an agent creation method provided in this specification are as follows: Refer to Figure 2 This embodiment provides an agent creation method that can be applied to a server. The method specifically includes steps S202 to S206.

[0020] Step S202: Receive a creation request sent by a user terminal for creating an agent of a service category in an application.

[0021] In this embodiment, the user initiates a creation request for creating an agent of a service category in an application through the user terminal. After the user terminal obtains the creation request for creating an agent of a service category in the application, it sends the request to the server, and the server receives the creation request sent by the user terminal for creating an agent of a service category in the application.

[0022] In practical applications, users have interaction requirements in various fields. In this embodiment, in order to enhance the user's interaction mentality in different fields, for different fields, that is, different service categories, agents are created for the user. Personalized agents are created from two dimensions of "user" and "service category", so that the created agents are more in line with the user's interaction characteristics and are better at the interaction requirements of the service category. The service category includes categories corresponding to different service fields; for example, the training service category corresponding to the training service field, the pet service category corresponding to the pet service field. In addition, it can also be categories corresponding to other service fields, such as the management service category, the catering service category. This embodiment does not make any limitations here.

[0023] The agent includes an entity that can autonomously execute tasks, make decisions, and learn and adjust according to environmental changes. Optionally, the agent includes a large language model and at least one data processing structure. Among them, the data processing structure can be used to execute corresponding data processing tasks. In this embodiment, the agent of the service category in the application includes the agent of the training service category in the third-party payment application; among them, the training service category can be fitness training. The large language model in this embodiment includes the LLM (Large Language Model); in addition, the large language model in this embodiment can also be a pre-trained natural language model. The large language model can adopt the foundation model (Foundation Models), and can also adopt the pre-trained model (Pretrained Models). The specific architecture of the large language model can be a neural network architecture with a large number of parameters, a Transform architecture, or other architectures. In the specific execution process, the large language model can directly adopt the foundation model or the pre-trained model, and can also fine-tune (Supervised Fine-Tuning, SFT) the foundation model or the pre-trained model for intent recognition and other tasks according to the task, so as to obtain a large language model capable of performing tasks such as intent recognition. For example, various open-source large language models such as chatGPT (chat Generative Pre-trained Transformer).

[0024] In order to improve the perception of the identity of the user to whom the agent belongs under the service category, the creation request in this embodiment can be submitted based on the user occupation data corresponding to the service category. In order to ensure the authenticity and effectiveness of the user occupation data, the user occupation data in this embodiment includes the user occupation data certified by the credit authentication service in the application. For example, the user occupation data corresponding to the fitness service category includes the fitness coach certificate and / or user identity data certified by the credit authentication service in the third-party application. In addition, if there is no user qualification data representing the user's skill qualification or professional qualification corresponding to the service category in the credit authentication service, the user identity data certified by the credit authentication service can be used as the user occupation data corresponding to the service category.

[0025] In practical applications, when a user submits data or requests to a server through a user terminal, it is often necessary to rely on a specific page for the user to perceive relevant content and submit data or requests. Therefore, in this embodiment, the user can submit a creation request through the creation reminder page of the agent of the service category displayed on the user terminal; and since it is an agent of the service category in the application program, the creation request is submitted through the application program; the following specifically describes the process of the user terminal submitting a creation request. In the specific execution process, in order to enhance the user's perception of the agent service, the service access entry of the agent service can be displayed on the subprogram list page of the application program, so as to enhance the user's perception of the agent service and also improve the convenience of the user accessing the agent service; the user can view the subprogram access entry and the service entry of the agent service through the subprogram list page in the application program; in order to improve the effectiveness of the agent service in the application program and avoid all users creating agents of each service category provided by the agent service but not reasonably applying the agents, resulting in waste of agent creation resources and management resources, in an optional implementation manner provided in this embodiment, after receiving the list access request for accessing the subprogram list of the application program sent by the user terminal, the server performs the following operations: According to the list access request for accessing the subprogram list sent by the user terminal, detect whether the user meets the admission conditions of the agent service. If so, return the service access entry of the agent service to the user terminal to generate a subprogram list page based on the service access entry and the subprogram access entry. If not, return the subprogram access entry to the user terminal to generate and display a subprogram list page based on the subprogram access entry. Optionally, the admission conditions of the agent service can be the admission conditions configured for the agent service. For example, the user's risk level is the admission risk level, and the user is a user who has been authenticated through the credit authentication service of the application program; in addition, the admission conditions of the agent service can also include the admission conditions corresponding to each service category under the agent service. For example, the admission conditions for the agent of the fitness service category include that the user has a purchase record of fitness tools, and the admission conditions for the agent of the pet service category include that the user owns a pet; it should be noted that when the agent service provides agents of multiple service categories and the user is only admitted to agents of some service categories, after the service access entry of the agent service on the subprogram list page is triggered, the user can only view the relevant data of the admitted service category agents.

[0026] Specifically, after the server receives the list access request sent by the user terminal to access the subprogram list of the application program, if the user meets the access conditions of the agent service, the server returns the service access entry and the subprogram access entry of the agent service to the user terminal, so as to generate and display the subprogram list page based on the service access entry and the subprogram access entry. If the user does not meet the access conditions of the agent service, the server returns the subprogram access entry to the user terminal, so as to generate and display the subprogram list page based on the subprogram access entry; For example, as Figure 3 shown in the subprogram list page, in addition to displaying the recently used subprograms and the subprograms in my favorites, it also displays my widget area, and the access entry of the agent service is displayed in this area.

[0027] In addition, after receiving the list access request, it is also possible not to detect the user, determine the subprogram access entries of the user under at least one associated category, and return the subprogram access entries under at least one associated category and the service access entry of the agent service to the user terminal, so as to generate the subprogram list page based on the service access entry and the subprogram access entry.

[0028] Based on the display of the subprogram list page on the user terminal, the user can enter the agent service through the service access entry in the subprogram list page. However, in the agent service, there are agents of service categories created by the user and agents of service categories not created. In order to improve the user's perception of the creation status of the agent, in an optional implementation provided in this embodiment, if the user terminal detects that the user triggers the service access entry in the subprogram list page, the user terminal sends a service access request for the agent service to the server, and the server performs the following operations: Receive the service access request for the agent service sent by the user terminal; Read the first service category of the agents created by the user and the second service category of the agents not created and return them to the user terminal to display the service access page; Optionally, the service access request is submitted based on the service access entry in the subprogram list page; the service category in the above agent creation of the service category can be any service category in the first service category.

[0029] Specifically, the server reads the first service category of the agents created by the user and the second service category of the agents not created according to the service access request for the agent service submitted based on the service access entry of the agent service in the subprogram list page sent by the user terminal, and returns them to the user terminal. The user terminal generates the service access page according to the first service category of the agents created by the user and the second service category of the agents not created; Optionally, the service category material corresponding to the first service category in the service access page includes the category description of the first service category and the agent access control for performing agent access to the first service category; the service category material corresponding to the second service category includes the category description of the second service category and the access control for performing agent creation access to the second service category.

[0030] For example, Figure 4 The service access page shown includes a fitness service category and a pet service category, wherein if the user has created an intelligent entity of the pet service category, the service category material of the pet service category consists of a service category description corresponding to the pet service category and an “enter intelligent entity” control, and if the user has not created an intelligent entity of the fitness service category, the service category material of the fitness service category consists of a service category description corresponding to the fitness service category and a “create intelligent entity” control; that is, the service access page includes service category material of at least one service category; the service category material of each service category is generated according to the service category description corresponding to the service category and the intelligent entity creation status.

[0031] It should be noted that users can trigger Figure 3 The service access entry of the intelligent agent service displayed on the subroutine list page shown in the figure is as follows Figure 4 The service access page of the intelligent agent service shown in the figure; the service access page can be used to enter the created intelligent agent or create an intelligent agent; in addition, the intelligent agent can also be an intelligent assistant, which can be displayed during the interaction with the user to enhance the user's perception.

[0032] Furthermore, since the created agent is a personalized agent belonging to the user, the user can edit the relevant data of the agent before initiating the agent creation request, such as editing the agent's avatar and logo, so as to mark the one-to-one correspondence between the agent and the user; in order to realize the personalized processing of the user's agent in the service category, in an optional implementation mode provided by this embodiment, the user terminal can submit an access request for agent creation access in the service category based on the service access page of the agent service, and on this basis, the server also performs the following operations: Reading user identification data and user occupation data of the application program according to an access request for creating access to an agent of a service category sent by a user terminal; The user identification data and the user occupation data are returned to the user terminal so as to create a reminder page display according to the user identification data and the user occupation data.

[0033] Correspondingly, the user terminal first obtains the access request submitted by the user for creating an agent of the service category and submits it to the server, then receives the user identification data and user occupation data returned by the server, and generates and displays a creation reminder page based on the user identification data and the access identity material generated based on the user occupation data. Optionally, the creation request is submitted based on the creation reminder page.

[0034] In the process of specifically submitting a creation request based on the creation reminder page of the agent of the service category, the user terminal obtains the access request submitted by the user for creating an agent of the service category and submits it to the server. In order to enable the user to edit the user identification data and perceive the user occupation data through the creation reminder page, the server reads the user identification data of the user in the application and the user occupation data of the user in the credit authentication service in the application according to the access request sent by the user terminal and returns it to the user terminal. The user terminal generates and displays a creation reminder page based on the user identification data and the access identity material generated based on the user occupation data. Optionally, the server reads the user occupation data based on the service category.

[0035] For example, the user terminal can generate a creation reminder page as Figure 5 shown; as Figure 5 shown, the creation reminder page displays the service category description data "Create your fitness intelligent assistant". The creation reminder page also includes a user category (male or female) selection control 501, an avatar display area 502 in the user identification data, and a name display area 503 in the user identification data. The user can change the avatar by triggering the avatar display area 502 or change the name by triggering the name display area 503; in order to perceive the user's identity, the creation reminder page also displays an access identity material 504 for displaying the user's user occupation certificate and / or user identity data; here, in order to ensure the security of the user identity data, the desensitized user identity data can be displayed; in order to enable the user to submit a creation request based on the creation reminder page, the creation reminder page also displays an "Agree and Experience Immediately" control. If this control is triggered, a creation request will be sent to the server. It should be noted that an authenticated label can be displayed in the access identity material 504 displayed on the creation reminder page to enhance the credibility of the data displayed in the access identity material. Optionally, in order to further improve the effectiveness of identity perception based on the labels in the access identity material, if the user occupation data includes the user occupation certificate, that is, the user qualification data, a qualification certificate can be displayed in the access identity material 504; for example, if the user is authenticated as a fitness coach, a fitness coach label will be displayed in the access identity material 504; if the user occupation data does not include the user occupation certificate, an authenticated label will be displayed in the access identity material 504 according to the authenticated user identity data.

[0036] Among them, the user can trigger, for example Figure 4 In the service category materials of the fitness service category on the service access page shown, submit an access request for creating an agent access for the fitness service category by clicking on "Create Agent". It should be noted that the anonymized user identity data can be returned by the server to the user terminal after anonymizing the user identity data, or can be anonymized by the user terminal for the user identity data returned by the server. This embodiment does not make any limitations in this regard.

[0037] During the process of the user terminal generating the access identity materials based on the user identification data and the user occupation data, the user identification data, the access identity materials, and the service category can be written into the creation reminder page template to obtain the creation reminder page. It should be noted that in addition to the server returning the user identification data and the user occupation data to the user terminal, and the user terminal generating and displaying the creation reminder page based on the user identification data and the user occupation data, the server can also generate the creation reminder page based on the user identification data and the user occupation data and return it to the user terminal, and the user terminal receives and displays the creation reminder page. Optionally, the user occupation data includes user occupation vouchers and / or user identity data.

[0038] It should be noted that the above-provided subprogram list page, service access page, and creation reminder page can be generated and displayed by the user terminal according to the relevant data returned by the server, or can be generated by the server according to the relevant data and returned to the user terminal for display. This embodiment does not make any limitations in this regard; and for other pages mentioned later, they can also be generated by the server and returned to the user terminal, or can be generated by the user terminal according to the relevant data returned by the server. This embodiment will not elaborate further, and the generation of various pages can write the relevant data into the corresponding page templates.

[0039] The above specifically describes the process from the user accessing the application to submitting the creation request. Among them, the user terminal obtains the list access request for the user to access the subprogram list of the application and sends it to the server, and receives the service access entrance of the agent service and the subprogram access entrances under at least one associated category returned by the server after the user logging in to the application meets the access conditions of the agent service; generates and displays the subprogram list page according to the service access entrance, at least one associated category, and the subprogram access entrances under each associated category; correspondingly, the server receives the list access request for the user terminal to access the subprogram list of the application, and returns the service access entrance of the agent service and the subprogram access entrances under at least one associated category after the user logging in to the application meets the access conditions of the agent service; Further, the user terminal obtains a service access request submitted by the user through the service access entry of the intelligent agent service in the subprogram list page, sends it to the server, receives the first service category and the second service category returned by the server, and generates and displays a service access page based on the first service category material corresponding to the first service category and the second service category material corresponding to the second service category; correspondingly, the server receives the service access request for the intelligent agent service sent by the user terminal, reads the first service category and the second service category, and returns them to the user terminal. Still further, the user terminal obtains an access request for creating an intelligent agent for a service category (any first service category, that is, any service category for which no intelligent agent has been created) submitted by the user through the service access page, sends it to the server, receives the user identification and user occupation data returned by the server, and generates and displays a creation reminder page based on the user identification data and the access identity material generated based on the user occupation data; correspondingly, the server reads the user identification data of the application program and the user occupation data authenticated by the user under the credit authentication service according to the access request for creating an intelligent agent for the service category sent by the user terminal, and returns them to the user terminal. Still further, the user terminal obtains a creation request for creating an intelligent agent for a service category in the application program submitted by the user based on the creation reminder page, and sends it to the server, and the server receives the creation request for creating an intelligent agent for a service category in the application program sent by the user terminal. Optionally, the creation request is submitted based on a creation reminder page configured with access identity material corresponding to the user identification data and the user occupation data.

[0040] It should be noted that the creation request includes creation data; optionally, the creation data is collected from the creation reminder page; the creation data includes at least one of the following: intelligent agent name, intelligent agent identification image, user category.

[0041] In specific implementation, since the created intelligent agent is a personalized intelligent agent for the user and the service category, in order to ensure the effectiveness of the created intelligent agent, in an optional implementation manner provided in this embodiment, after receiving the creation request, first perform data detection on the creation data carried in the creation request. If the detection passes, return the user occupation data and the user category included in the creation data to the user terminal, and execute the following step S204. In response to the creation request, create an intelligent agent for the user under the service category according to the intelligent agent template corresponding to the service category; if the detection fails, determine the detection failure keyword, and interact with the user terminal according to the detection failure keyword for creation data editing. Optionally, data detection includes data integrity detection and / or data validity detection; for example, detecting whether the created data contains agent creation parameters, that is, whether it contains the agent name, agent identification image, and user category. If so, it is determined that the data integrity detection passes; if not, it is determined that the data integrity detection fails, and the agent parameters to be supplemented are determined as the detection failure keywords and sent to the user terminal, so that the user terminal generates and displays an agent parameter supplement page corresponding to the detection failure keywords, obtains the agent parameters to be supplemented collected through the agent parameter supplement page and sends them to the server. In addition, the user terminal can also display a creation reminder page with a to-be-supplemented mark for the agent parameters to be supplemented, obtain the collected agent parameters to be supplemented and send them to the server; for another example, detecting whether the agent name included in the creation is a duplicate name or an abnormal name. If not, it is determined that the data validity detection passes; if so, a name abnormality reminder is generated and returned to the user terminal. The user terminal displays a reminder for re-entering the agent name through the creation reminder page, obtains the updated agent name entered and sends it to the server until the data detection of the created data by the server passes.

[0042] Specifically, after receiving the creation request sent by the user terminal for creating an agent of the service category in the application, if the data detection of the created data carried by the creation request passes, the server returns the user's occupation data and the user category included in the created data to the user terminal for interactive display of agent creation, and executes step S204 to create an agent for the user in the service category according to the agent template corresponding to the service category. In addition, in addition to returning the user's occupation data and the user category included in the created data to the user terminal for interactive display of agent creation, the server can also return a creation interaction reminder to the user terminal. In response to the creation interaction reminder, the user terminal performs interactive display of agent creation according to the user's occupation data and the user category.

[0043] It should be noted that step S202 can also be replaced by receiving the creation request sent by the user terminal for creating an agent of the service category in the application and forming a new implementation manner with one or more other processing steps provided in this embodiment. Or, it may not interact with the user terminal, that is, step S202 can also be replaced by obtaining the creation request sent by the user for creating an agent of the service category in the application and forming a new implementation manner with one or more other processing steps provided in this embodiment.

[0044] Step S204, in response to the creation request, create an agent for the user in the service category according to the agent template corresponding to the service category.

[0045] In specific implementation, to improve the efficiency of agent creation, after receiving a creation request sent by a user terminal for creating an agent of a service category in an application, in response to the creation request, an agent for the user under the service category is created according to the agent template corresponding to the service category. The agent template in this embodiment includes an agent template pre-configured with the basic structure of the agent; for example, the agent template includes a reference large language model and at least one data processing structure; the agent templates of different users under the same service category are the same; that is, the agent parameters of the agents just created by different users under the same service category are different, but the structures of the agents are the same.

[0046] In the process of specifically creating an agent for a user under the service category according to the agent template corresponding to the service category, an agent code associated with the agent identifier or agent name in the creation data can be generated first, and the agent code is reviewed. When the agent code passes the review, it is determined that the agent creation is completed.

[0047] It should be noted that agent creation includes front-end agent creation and / or back-end agent creation. To improve the user's perception of the front-end agent, the front-end agent can be in the form of a subroutine, that is, the agent is accessed through the subroutine; to improve the efficiency of agent creation, the subroutine creation template of the application can be used to create a subroutine for agent access; correspondingly, the back-end agent can also be created using the creation process and agent template at the back-end of the application's subroutine.

[0048] Since the creation of an agent takes a certain amount of time, in an optional implementation provided in this embodiment to improve the user experience, after the user terminal sends a creation request to the server, that is, during the process of the server creating the agent, an interactive display of agent creation is performed according to the interaction materials corresponding to the user category and the identity materials generated based on the user's occupation data; In the specific process of the user terminal performing an interactive display of agent creation according to the interaction materials and identity materials, an interactive animation can be generated and played based on the interaction materials and identity materials, so as to improve the user's perception of the agent creation process; specifically, the interaction materials and identity materials can be added to the interactive animation template to obtain and display the interactive animation. Optionally, if the user category is the first category (male), the interaction material is a pre-configured animated image of a male, and if the user category is the second category (female), the interaction material is a pre-configured animated image of a female; the identity materials generated based on the user's occupation data are similar to the above access identity materials, and the user qualification data and the desensitized user identity data can be displayed.

[0049] For example, as Figure 6An interactive display page showing a certain interactive display time shows agent reminder data "Hi, I'm your exclusive fitness assistant", the first type of animated image, and an identity card or credit card.

[0050] It should be noted that, as Figure 6 The content shown on the interactive display page at different times is different. For example, the animated image can wave from far to near, and the identity card can slide out from below; the specific animation form can be configured according to actual needs, and this embodiment does not limit it here. Optionally, the user can enter the credit authentication service through the identity card to view the user qualification data and user identity data.

[0051] Step S206, after detecting the creation of the agent, return agent access data to the user terminal.

[0052] In this embodiment, after creating an agent for the user under the service category according to the agent template corresponding to the service category, agent access data is returned to the user terminal; optionally, the user terminal enters a subroutine for agent access according to the agent access data.

[0053] The agent access data includes the subroutine identifier of the subroutine for agent access and / or the data shown on the agent access page displayed through the subroutine.

[0054] In specific implementation, after the server detects the creation of the agent, it returns agent access data to the user terminal. Since the subroutine for agent access corresponds to the agent, during the process of returning the agent access data to the user terminal, the agent access data can be sent to the user terminal through the agent; in addition, the server can first return the subroutine identifier of the subroutine for agent access to the user terminal, and the user terminal enters the subroutine for agent access according to the subroutine identifier. The user terminal sends an agent access request to the agent through the subroutine, and the agent, that is, the backend agent, reads and / or generates agent recommendation data according to the agent access request and returns it to the subroutine. Optionally, the agent recommendation data includes user occupation data, user category, and / or interaction reminder data.

[0055] In the specific execution process, in order to enhance the user's perception of the created intelligent agent, in an optional implementation provided in this embodiment, after the user terminal enters the subroutine for intelligent agent access according to the intelligent agent access data returned by the server after the intelligent agent is created, it first generates access identity materials based on the user occupation data carried in the intelligent agent access data, and then generates and displays an intelligent agent access page based on the access identity materials, interaction materials, and interaction reminder data carried in the intelligent agent access data; or after the user terminal enters the subroutine for intelligent agent access, it first reads the stored identity materials or access identity materials, interaction materials, and interaction reminder data corresponding to the service category, and generates and displays an intelligent agent access page based on the identity materials or access identity materials, interaction materials, and interaction reminder data corresponding to the service category.

[0056] In addition, after the user terminal enters the subroutine for intelligent agent access according to the intelligent agent access data returned by the server, it can also send an intelligent agent access request to the intelligent agent through the subroutine. The intelligent agent reads and / or generates intelligent agent recommendation data or intelligent agent recommendation instructions according to the intelligent agent access request and returns them to the subroutine. The user terminal receives the intelligent agent recommendation data through the subroutine and generates and displays an intelligent agent access page based on the intelligent agent recommendation data, or the user terminal receives the intelligent agent recommendation instructions through the subroutine and reads the stored intelligent agent recommendation data according to the intelligent agent recommendation instructions, and generates and displays an intelligent agent access page based on the intelligent agent recommendation data.

[0057] For example, as Figure 7 shown in the intelligent agent access page, that is, the intelligent agent access home page displays an identity business card 701, interaction materials 702, and interaction reminder data 703; in addition, it may also include user identification data; optionally, the interaction reminder data includes a quick interaction reminder text and a quick service entry. The user can efficiently submit interaction data through the quick service entry, saving the time consumed by entering interaction data through voice or text; among them, the quick service entry can be configured according to the intelligent agent template. For example, if the intelligent agent template is configured with "explanation of the auxiliary uses of the fitness intelligent assistant", then "introduce what you can assist me to do" can be used as the quick service entry; for another example, if the large language model in the intelligent agent template has a course management function, then "course management" can be used as the quick service entry; in addition, the quick service entry can also be configured according to the user qualification data. If it is detected that the user occupation data contains the user qualification data, then "certificate management" is used as the quick service entry; in addition, the quick service entry can also be pre-configured, which is not limited in this embodiment. In addition to configuring a quick service entry in the interactive reminder data, a separate quick service entry can also be configured. For example, there are three quick service entries of knowledge management, course management, and certificate management shown in the quick service entry area 704. It should be noted that the determination method of the quick service entry included in the interactive reminder data can be the same as or different from the determination method of the quick service entry configured in the quick service entry area. The specific quick service entries can be the same or different, and can be specifically configured according to actual needs. This embodiment does not make a limitation here.

[0058] It should be noted that in order to improve the convenience of entering interactive data by accessing the page through the agent, the interactive data can be submitted by triggering the quick service entry in the agent access page, or the interactive data can be submitted by means of text entry and / or voice entry in the agent access page. Optionally, the agent access page is configured with a text box for text entry and / or a voice control for voice entry; after detecting that the voice control is triggered, the user terminal collects the user voice as interactive data according to the authorization data of the voice collection control.

[0059] In specific implementation, after the user terminal enters the interactive data through the agent access page, it sends it to the server or to the agent. In the specific interaction process, the server refers to the agent; after receiving the interactive data sent by the user terminal, the agent inputs the interactive data into the large language model in the agent to generate response data, obtains the response data, and returns it to the user terminal; specifically, after receiving the interactive data, the agent first queries whether there is response data configured for the interactive data. If so, it returns the response data to the user terminal; if not, it inputs the interactive data into the large language model to generate response data, obtains the response data, and returns it to the user terminal. In order to improve the effectiveness of the generated response data, the large language model first performs intent recognition on the interactive data to obtain the interactive intent, and then generates the response data according to the data generation strategy corresponding to the interactive intent to obtain the response data; in an optional implementation manner provided by this embodiment, the large language model can generate the response data in the following manner: Perform intent recognition on the interactive data through the intent recognition module to obtain the interactive intent; Determine the data generation module corresponding to the interactive intent, and generate initial response data for the interactive data through the data generation module to obtain the initial response data; Summarize the interactive data and the initial response data through the data summary module to obtain the response data.

[0060] In the specific execution process, first, the interaction data is input into the intent recognition sub-model in the large language model for intent recognition to obtain the interaction intent, determine the data generation module corresponding to the interaction intent, and input the interaction data into the data generation module to generate the initial response data. After obtaining the initial response data, finally, in order to improve the user's perception of the response data and the interaction data, the interaction data and the initial response data are summarized through the data summary module to obtain the response data.

[0061] Specifically, data generation modules corresponding to each interaction intent are configured in the large language model. After determining the interaction intent of the interaction data, the interaction data is used to generate response data through the data generation module corresponding to the interaction intent to obtain the response data.

[0062] The following takes the process of generating two types of initial response data as an example to illustrate the process of generating initial response data by different data generation modules.

[0063] (1) In an optional implementation provided in this embodiment, the data generation module corresponding to the knowledge management intent generates the initial response data in the following manner: Query knowledge data in the preset knowledge base according to the intent data included in the interaction data to obtain the first knowledge data, and / or query knowledge data in the dataset according to the intent data to obtain the second knowledge data; Input the first knowledge data, the second knowledge data, and the third knowledge data included in the interaction data into the summary sub-module for data summary to obtain the initial response data; Specifically, after the user terminal obtains the intent data entered by the user through the agent access page, if the collection prompt data corresponding to the intent data is detected, the collection prompt data is displayed through the agent access page, and the intent data and the data to be processed entered by the user based on the collection prompt data are determined as the interaction data; if the collection prompt data is not detected, the intent data is directly sent as the interaction data to the agent; Under the knowledge management intent, after the user terminal obtains the intent data of knowledge management entered by the user through the agent access page and detects the intent data, then through, for example Figure 8The intelligent agent conversation page shown displays collection prompt data. The user can upload files and / or images based on the "file upload" control and / or "image upload" control in the collection prompt data, and / or input voice data or text data in the form of voice input or text input as the third knowledge data, which is sent to the intelligent agent together with the intent data as interaction data. After receiving the interaction data, the intelligent agent inputs the interaction data into the large language model to generate response data. During the process of generating response data, the large language model first performs intent recognition on the interaction data to obtain the knowledge management intent, and then according to the data generation strategy corresponding to the knowledge management intent, first performs retrieval enhancement based on the intent data to obtain the first knowledge data, and queries for knowledge data in the dataset of the intelligent agent according to the intent data to obtain the second knowledge data. Then, the first knowledge data, the second knowledge data, and the third knowledge data included in the interaction data are input into the summary sub-module for data summary to obtain the initial response data. After that, the interaction data and the initial response data are input into the data summary module for data summary to obtain the response data that at least includes the intent data and the initial response data.

[0064] (2) In an optional implementation manner provided in this embodiment, the data generation module corresponding to the question-and-answer intent generates the initial response data in the following manner: Query for the response data in the database according to the question data included in the interaction data; If the query is successful, determine the data obtained from the query as the initial response data; If the query fails, input the question data into the question-and-answer sub-module to generate an answer to obtain the initial response data.

[0065] Specifically, during the process of generating the initial response data, the data generation module corresponding to the question-and-answer intent first queries the response data of the question data included in the interaction data in the database of the intelligent agent. If the query fails, the question data is input into the question-and-answer sub-module to generate an answer to obtain the initial response data.

[0066] In practical applications, due to the strong randomness of the interaction data submitted by users, the data generation module corresponding to the interaction intention in the agent template may not be comprehensive enough, which may lead to an inability to accurately and effectively respond to the interaction data of users. Therefore, in an alternative implementation provided in this embodiment, if the data generation module corresponding to the determined interaction intention is empty, the agent configures the corresponding data generation module according to the interaction intention, that is, the configuration of the agent tool. In this way, the accuracy and effectiveness of the agent's response to the interaction data of various interaction intentions are improved, and further the user's perception of the agent is enhanced. Optionally, the agent tool includes a module in the agent responsible for calling external tools and services; for example, a RAG (Retrieval-Augmented Generation) module, a weather module, etc.

[0067] In addition to configuring the agent tool to update the agent, the large language model can also be trained to improve the quality of the response data generated by the large language model. Specifically, in an alternative implementation provided in this embodiment, after the interaction data is input into the large language model in the agent to generate response data and the response data is obtained, the agent also performs the following operations: Input the response data and the detection prompt data into the large language model for response quality detection to obtain a response quality detection result; Train the large language model according to the response quality detection result, input the interaction data into the trained large language model for secondary response data generation, and return the response data obtained from the secondary response data generation to the subroutine.

[0068] It should be noted that as Figure 7 shown, the agent access home page displays an identity card, interaction materials, and interaction reminder data. During the interaction with the user, in order to enhance the user's perception of the interaction content, that is, the conversation content, it can be jumped to the agent conversation page as Figure 8 shown for the display of the conversation content. Optionally, the agent access page includes the agent access home page and / or the agent conversation page.

[0069] In addition, the large language module may not be configured in the agent, and an intention recognition model, at least one data processing module, and a data summary model are configured. For example, Figure 9 shown in the agent schematic diagram, including an intention recognition model, a RAG module, agent tools 1 to agent tools n, and a summary model; the agent can be updated and response data can be generated according to the above relevant content, which will not be elaborated in this embodiment.

[0070] It should be noted that step S206 can also be replaced by, after detecting the creation of the agent, returning agent access data to the user terminal and forming a new implementation manner with one or more other processing steps provided in this embodiment; or being replaced by returning agent access data to the user terminal and forming a new implementation manner with one or more other processing steps provided in this embodiment.

[0071] After creating an agent of a service category, subsequent agent access can be entered through the service access page; in order to enhance the user's perception of agent access, in an optional implementation manner provided in this embodiment, after the agent receives an agent access request sent by the user terminal, the following operations are performed: According to the agent access request sent by the user terminal for the agent, read the user's historical access time to the agent; Perform access time interval detection based on the historical access time and the agent access time carried in the agent access request; Generate corresponding agent access data according to the detection result and return it to the user terminal for display through a subroutine on the user terminal.

[0072] Specifically, after receiving the agent access request sent by the user terminal, if the time interval between the agent access time and the historical access time is greater than the interval threshold, the agent recommendation data or agent recommendation instruction is used as the agent access data and returned to the user terminal; optionally, the user terminal displays the agent access home page; if the time interval is less than or equal to the interval threshold, the historical interaction data or historical interaction instruction is returned to the user terminal; optionally, the user terminal displays the agent conversation page including the historical interaction data.

[0073] In addition to creating and accessing agents through the service access page, agents can also be created and accessed through the user occupation data display page corresponding to the service category in the credit authentication service. Specifically, except for the different initiation locations, other processes are similar to those of creating and accessing agents through the service access page. Refer to the relevant content and this embodiment will not be elaborated here.

[0074] In the specific implementation process, after a user creates an agent, the user can share the agent with other users. For example, a fitness coach can share the agent with fitness trainees so that the fitness trainees can view the knowledge and courses in the agent. In addition, the fitness trainees can also ask questions through the agent. Specifically, in an optional implementation manner provided in this embodiment, the agent receives an agent access request sent by a user terminal, determines an access category based on the access user identifier carried in the agent access request and the agent user identifier of the user to whom the agent belongs, generates agent access data corresponding to the access category, and returns it to the user terminal. Optionally, the access category includes a first access category and a second access category. Among them, the first access category includes the access category for a user to access their own agent, and the second access category includes the access category for other users to access the agent of the user to whom the agent belongs.

[0075] Specifically, during the process of a fitness coach accessing their own agent and a fitness trainee accessing the fitness coach's agent, the service quick access entry may be different, and the others can be the same or configured differently according to actual needs. This embodiment does not make any limitations here.

[0076] During specific implementation, the agent accesses through a subroutine of the user terminal. To improve the user's perception of the subroutine for agent access, in an optional implementation manner provided in this embodiment, during the process of the user accessing the subroutine list, the server detects whether the user is in the agent based on the list access request sent by the user terminal for accessing the subroutine list. If not, it returns the subroutine access entry and / or service access entry of other subroutines to the user terminal. If so, it returns the agent access entry and the subroutine access entry of other subroutines to the user terminal to generate a subroutine list page based on the agent access entry and the subroutine access entry. Optionally, after the agent access entry is triggered, it enters the subroutine for agent access to perform agent access. It should be noted that if the subroutine for agent access is associated, the subroutine list page is generated according to the association category. That is, the agent access entry is configured under the corresponding association category.

[0077] In summary, in the process of creating an agent in cooperation with the user terminal, after the server receives a creation request sent by the user terminal to create an agent of a service category in the application, the server creates an agent for the user under the service category according to the agent template corresponding to the service category. In this way, the agent is created through the agent template, improving the creation efficiency of the agent. Further, the server returns agent access data to the user terminal. The user terminal enters a subroutine for agent access according to the agent access data and interacts with the agent through the agent access page displayed by the subroutine. In this way, an agent for the user under the service category is created to provide service data under the service category through the agent, improving the effectiveness and convenience of the user's access to and management of service data under the service category. In the process of creating the agent, the user's professional data corresponding to the service category authenticated by the credit authentication service is also used as one of the creation parameters, improving the effectiveness of the created agent and enhancing the trust of the user or the accessing user in the agent; In the process of using the user's professional data as one of the creation parameters to create an agent for the user under the service category according to the agent template corresponding to the service category, the user's professional data and the creation data carried in the creation request can be written into the agent template to obtain the agent. In addition, a subroutine for agent access can be created according to the subroutine template for creating subroutines in the application. The identifier of the created subroutine is the agent identifier carried in the creation request, that is, the name of the created subroutine is the agent name carried in the creation request, and the subroutine image is the agent image carried in the creation request.

[0078] The above steps S202 to S206 provided in this embodiment can be executed by the server. It should be noted that the above steps S202 to S206 executed by the server can cooperate with the steps S1202 to S1206 executed by the user terminal in the following embodiment during the execution process. Therefore, when reading this embodiment, please refer to the corresponding content of steps S1202 to S1206 provided in the following method embodiment, and when reading the following method embodiment, please refer to the corresponding content of steps S202 to S206 provided in this embodiment.

[0079] The following takes the application of an agent creation method provided in this embodiment in the fitness intelligent assistant creation scenario as an example to further illustrate the agent creation method provided in this embodiment. As Figure 10 described, the agent creation method applied to the fitness intelligent assistant creation scenario specifically includes the following steps.

[0080] Step S1004, receive a list access request sent by the user terminal to access the subroutine list of the application.

[0081] Step S1006: After the user logging in to the application meets the access conditions for the intelligent assistant service, return the service access entrance of the intelligent assistant service and the subprogram access entrances under at least one associated category to the user terminal.

[0082] Step S1014: Receive the service access request for the intelligent assistant service sent by the user terminal, read the first service category and the second service category, and return them to the user terminal.

[0083] Step S1018: According to the access request for creating a fitness intelligent assistant sent by the user terminal, read the user identification data of the application and the fitness qualification data certified by the user under the credit authentication service, and return them to the user terminal.

[0084] Step S1024: Receive the creation request for creating a fitness intelligent assistant in the application sent by the user terminal.

[0085] Step S1026: Perform data detection on the creation data included in the creation request.

[0086] Step S1028: In the case of passing the detection, return an interactive display reminder to the user terminal, and create the user's fitness intelligent assistant according to the fitness intelligent assistant template.

[0087] Step S1032: After the fitness intelligent assistant is created, return the fitness intelligent assistant access data for entering the subprogram for accessing the fitness intelligent assistant to the user terminal.

[0088] The above steps S1004 to S1006, step S1014, step S1018, steps S1026 to S1028, and step S1032 provided in this embodiment are executed by the server. It should be noted that the above steps S1004 to S1006, step S1014, step S1018, steps S1026 to S1028, and step S1032 executed by the server, and the steps S1002, steps S1008 to S1012, step S1016, steps S1020 to S1024, step S1030, and step S1034 executed by the user terminal in the following embodiment can cooperate with each other during the execution process. Therefore, when reading this embodiment, please refer to the corresponding content of steps S1002, steps S1008 to S1012, step S1016, steps S1020 to S1024, step S1030, and step S1034 provided in the following method embodiment. When reading the following method embodiment, please refer to the corresponding content of steps S1004 to S1006, step S1014, step S1018, steps S1026 to S1028, and step S1032 provided in this embodiment.

[0089] It should be noted that any one or any combination of steps from step S1002 to step S1034 can be combined with any one or any combination of steps from step S202 to step S206 according to the needs of implementation and deployment to form a new implementation method; in addition, according to the actual deployment needs, any one or any combination of technical features from step S1002 to step S1034 can be selected and combined with any one or more technical features provided in the above steps S202 to S206 to form a new implementation method; or, any one or any combination of technical features in steps S1002 to S1034 can also be replaced by any one or more technical feature combinations provided in the above steps S202 to S206 according to the actual deployment needs to form a new implementation method, which will not be elaborated here one by one.

[0090] The following takes the application of an agent creation method provided in this embodiment in the training agent creation scenario as an example to further illustrate the agent creation method provided in this embodiment. As Figure 11 shown, the agent creation method applied to the training agent creation scenario specifically includes the following steps.

[0091] Step S1104, receive a training agent access request sent by a user terminal.

[0092] Step S1106, detect whether the user creates a training agent; If so, return training agent access data to the user terminal to enter the subroutine for accessing the training agent; If not, execute step S1108.

[0093] Step S1108, read the user identification data of the application program and the training qualification data authenticated by the user under the credit authentication service and return them to the user terminal.

[0094] Step S1114, receive a creation request sent by the user terminal to create a training agent in the application program.

[0095] Step S1116, return an interactive display reminder to the user terminal and create a training agent for the user according to the training agent template.

[0096] Step S1120, after the training agent is created, return the training agent access data for entering the subroutine for accessing the training agent to the user terminal.

[0097] The above steps S1104 to S1108, steps S1114 to S1116, and step S1120 provided in this embodiment are executed by the server. It should be noted that the above steps S1104 to S1108, steps S1114 to S1116, and step S1120 executed by the server can cooperate with the steps S1102, steps S1110 to S1112, steps S1118, and step S1122 executed by the user terminal in the following embodiments during the execution process. Therefore, when reading this embodiment, please refer to the corresponding content of steps S1102, steps S1110 to S1112, steps S1118, and step S1122 provided in the following method embodiment, and when reading the following method embodiment, please refer to the corresponding content of steps S1104 to S1108, steps S1114 to S1116, and step S1120 provided in this embodiment.

[0098] It should be noted that any one step or any combination of multiple steps among steps S1102 to S1122 can be combined with any one step or any combination of multiple steps among the above steps S202 to S206 to form a new implementation manner according to the needs of implementation deployment; in addition, according to the actual deployment needs, any one or any combination of technical features among steps S1102 to S1122 can be combined with any one or more technical features provided by the above steps S202 to S206 to form a new implementation manner; or, any one or any combination of technical features among steps S1102 to S1122 can also be replaced by any one or more technical feature combinations provided by the above steps S202 to S206 according to the actual deployment needs to form a new implementation manner, which will not be elaborated here one by one.

[0099] One or more embodiments of another intelligent agent creation method provided in this specification are as follows: Refer to Figure 12 , the intelligent agent creation method provided in this embodiment, the method specifically includes steps S1202 to S1206.

[0100] Step S1202, obtain a creation request for creating an intelligent agent of a service category in the application program and send it to the server.

[0101] In this embodiment, the user initiates a creation request for creating an intelligent agent of a service category in the application program through the user terminal running on the user terminal. After the user terminal obtains the creation request for creating an intelligent agent of a service category in the application program, it sends it to the server, and the server receives the creation request for creating an intelligent agent of a service category in the application program sent by the user terminal.

[0102] In practical applications, users have interaction requirements in various fields. In this embodiment, in order to enhance the user's interaction mentality in different fields, for different fields, that is, different service categories, an agent is created for the user. An individualized agent is created from two dimensions of "user" and "service category", so that the created agent is more in line with the user's interaction characteristics and is better at meeting the interaction requirements of the service category. The service category includes categories corresponding to different service fields; for example, the training service category corresponding to the training service field, the pet service category corresponding to the pet service field. In addition, it can also be categories corresponding to other service fields, such as the management service category, the catering service category. This embodiment does not make any limitations here.

[0103] The agent includes an entity that can autonomously execute tasks, make decisions, and learn and adjust according to environmental changes. Optionally, the agent includes a large language model and at least one data processing structure. Among them, the data processing structure can be used to execute corresponding data processing tasks. In this embodiment, the agent of the service category in the application program includes the agent of the training service category in the third-party payment application program; among them, the training service category can be fitness training.

[0104] In order to enhance the perception of the identity of the user to whom the agent belongs under the service category, the creation request in this embodiment can be submitted based on the user occupation data corresponding to the service category. In order to ensure the authenticity and validity of the user occupation data, the user occupation data in this embodiment includes the user occupation data authenticated by the credit authentication service in the application program. For example, the user occupation data corresponding to the fitness service category includes the fitness coach certificate and / or user identity data authenticated by the credit authentication service in the third-party application program. In addition, if there is no user qualification data indicating the user's skill qualification or occupation qualification corresponding to the service category in the credit authentication service, the authenticated user identity data in the credit authentication service can be used as the user occupation data corresponding to the service category.

[0105] In practical applications, when a user submits data or requests to the server through a user terminal, it is often necessary to rely on a specific page for the user to perceive relevant content and submit data or requests. Therefore, in this embodiment, the user can submit a creation request through the creation reminder page of the agent of the service category displayed on the user terminal; and since it is the agent of the service category in the application program, the creation request is submitted through the application program; the following specifically describes the process of the user terminal submitting a creation request; In the specific execution process, in order to enhance the user's perception of the agent service, the service access entry of the agent service can be displayed on the subprogram list page of the application program. In this way, the user's perception of the agent service is enhanced, and the convenience of the user accessing the agent service is also improved; the user can view the subprogram access entry and the service entry of the agent service through the subprogram list page in the application program; in order to improve the effectiveness of the agent service in the application program and avoid all users creating agents of each service category provided by the agent service, but not reasonably applying the agents, resulting in waste of agent creation resources and management resources, in an optional implementation manner provided in this embodiment, after the server receives the list access request for accessing the subprogram list of the application program sent by the user terminal, the following operations are performed: According to the list access request for accessing the subprogram list sent by the user terminal, detect whether the user meets the admission conditions of the agent service; If so, return the service access entry of the agent service to the user terminal to generate a subprogram list page based on the service access entry and the subprogram access entry; If not, return the subprogram access entry to the user terminal to generate and display a subprogram list page based on the subprogram access entry; Optionally, the admission conditions of the agent service can be the admission conditions configured for the agent service. For example, the user's risk level is the admission risk level, and the user is a user who has passed the identity authentication service of the application program; in addition, the admission conditions of the agent service can also include the admission conditions corresponding to each service category under the agent service. For example, the admission conditions of the agent for the fitness service category include that the user has a purchase record of fitness tools, and the admission conditions of the agent for the pet service category include that the user owns a pet; it should be noted that when the agent service provides agents of multiple service categories and the user is only admitted to some service categories of agents, after the service access entry of the agent service on the subprogram list page is triggered, the user can only view the relevant data of the admitted service category of agents.

[0106] Specifically, after the server receives the list access request for accessing the subprogram list of the application program sent by the user terminal, if the user meets the admission conditions of the agent service, the service access entry and the subprogram access entry of the agent service are returned to the user terminal to generate and display a subprogram list page based on the service access entry and the subprogram access entry; if the user does not meet the admission conditions of the agent service, the subprogram access entry is returned to the user terminal to generate and display a subprogram list page based on the subprogram access entry; For example, such as Figure 3The displayed subroutine list page shows not only the recently used subroutines and the subroutines in my favorites, but also my widget area, where the access entrances to the agent services are shown.

[0107] In addition, after receiving a list access request, it is also possible not to detect the user, determine the subroutine access entrances of the user under at least one associated category, and return to the user terminal the subroutine access entrances under at least one associated category and the service access entrance of the agent service, so as to generate a subroutine list page based on the service access entrance and the subroutine access entrance.

[0108] Based on the display of the subroutine list page on the user terminal, the user can enter the agent service through the service access entrance in the subroutine list page. However, in the agent service, there are agents of service categories created by the user and agents of service categories not created. In order to improve the user's perception of the creation status of the agents, in an optional implementation provided in this embodiment, if the user terminal detects that the user triggers the service access entrance in the subroutine list page, it sends a service access request for the agent service to the server, and the server performs the following operations: Receive the service access request for the agent service sent by the user terminal; Read the first service category of the agents created by the user and the second service category of the agents not created and return them to the user terminal to display the service access page; Optionally, the service access request is submitted based on the service access entrance in the subroutine list page; the service category in the above agent creation of the service category can be any service category in the first service category.

[0109] Specifically, the server reads the first service category of the agents created by the user and the second service category of the agents not created according to the service access request for the agent service submitted based on the service access entrance in the subroutine list page sent by the user terminal and returns them to the user terminal. The user terminal generates a service access page according to the first service category of the agents created and the second service category of the agents not created. Optionally, the service category materials corresponding to the first service category in the service access page include the category description of the first service category and the agent access control for accessing the agents of the first service category; the service category materials corresponding to the second service category include the category description of the second service category and the access control for accessing the creation of agents of the second service category.

[0110] For example, as Figure 4The service access page shown includes a fitness service category and a pet service category, wherein if the user has created an intelligent entity of the pet service category, the service category material of the pet service category consists of a service category description corresponding to the pet service category and an “enter intelligent entity” control, and if the user has not created an intelligent entity of the fitness service category, the service category material of the fitness service category consists of a service category description corresponding to the fitness service category and a “create intelligent entity” control; that is, the service access page includes service category material of at least one service category; the service category material of each service category is generated according to the service category description corresponding to the service category and the intelligent entity creation status.

[0111] It should be noted that users can trigger Figure 3 The service access entry of the intelligent agent service displayed on the subroutine list page shown in the figure is as follows Figure 4 The service access page of the intelligent agent service shown in the figure; the service access page can be used to enter the created intelligent agent or create an intelligent agent; in addition, the intelligent agent can also be an intelligent assistant, which can be displayed during the interaction with the user to enhance the user's perception.

[0112] Furthermore, since the created agent is a personalized agent belonging to the user, the user can edit the relevant data of the agent before initiating the agent creation request, such as editing the agent's avatar and logo, so as to mark the one-to-one correspondence between the agent and the user; in order to realize the personalized processing of the user's agent in the service category, in an optional implementation mode provided by this embodiment, the user terminal can submit an access request for agent creation access in the service category based on the service access page of the agent service, and on this basis, the server also performs the following operations: Reading user identification data and user occupation data of the application program according to an access request for creating access to an agent of a service category sent by a user terminal; The user identification data and the user occupation data are returned to the user terminal so as to create a reminder page display according to the user identification data and the user occupation data.

[0113] Correspondingly, the user terminal first obtains the access request submitted by the user for creating access to the intelligent agent of the service category and submits it to the server, then receives the user identification data and user occupation data returned by the server, generates and displays the creation reminder page according to the user identification data and the access identity material generated based on the user occupation data. Optionally, the creation request is submitted based on the creation reminder page.

[0114] During the process of submitting a creation request on the creation reminder page for a service category-based agent, the user terminal obtains the access request submitted by the user for creating access to the service category-based agent and submits it to the server. To enable the user to edit user identification data and perceive user occupation data through the creation reminder page, the server reads the user identification data of the user in the application and the user occupation data of the user in the credit authentication service in the application according to the access request sent by the user terminal and returns it to the user terminal. The user terminal generates and displays the creation reminder page based on the user identification data and the access identity material generated based on the user occupation data. Optionally, the server reads the user occupation data based on the service category.

[0115] For example, the user terminal can generate a creation reminder page as Figure 5 shown; as Figure 5 shown, the creation reminder page displays the service category description data "Create your fitness intelligent assistant". The creation reminder page also includes a user category (male or female) selection control 501, an avatar display area 502 in the user identification data, and a name display area 503 in the user identification data. The user can change the avatar by triggering the avatar display area 502 or change the name by triggering the name display area 503. To perceive the user's identity, the creation reminder page also displays access identity material 504, which is used to display the user's user occupation certificate and / or user identity data. Here, to ensure the security of the user identity data, the desensitized user identity data can be displayed. To enable the user to submit a creation request based on the creation reminder page, the creation reminder page also displays an "Agree and Experience Immediately" control. If this control is triggered, a creation request will be sent to the server. It should be noted that an authenticated label can be displayed in the access identity material 504 shown on the creation reminder page to enhance the credibility of the data shown in the access identity material. Among them, the user can submit an access request for creating access to the fitness service category-based agent by triggering the "Create Agent" in the service category material of the fitness service category in the service access page as Figure 4 shown. It should be noted that the desensitized user identity data can be returned to the user terminal by the server after desensitizing the user identity data, or the user terminal can desensitize the user identity data returned by the server. This embodiment does not make a limitation here.

[0116] In the process of generating access identity materials based on user identification data and user occupation data, the user terminal can write the user identification data, access identity materials, and service category into the creation reminder page template to obtain the creation reminder page; it should be noted that, in addition to the server returning the user identification data and user occupation data to the user terminal, and the user terminal generating and displaying the creation reminder page based on the user identification data and user occupation data, the server can also generate the creation reminder page based on the user identification data and user occupation data and return it to the user terminal, and the user terminal receives and displays the creation reminder page. Optionally, the user occupation data includes user occupation vouchers and / or user identity data.

[0117] It should be noted that the above-provided subroutine list page, service access page, and creation reminder page can be generated and displayed by the user terminal according to the relevant data returned by the server, or can be generated by the server according to the relevant data and returned to the user terminal for display. This embodiment does not make any limitations here; moreover, for other pages mentioned later, they can also be generated by the server and returned to the user terminal, or can be generated by the user terminal according to the relevant data returned by the server. This embodiment will not elaborate further here, and the generation of various pages can write the relevant data into the corresponding page template.

[0118] The above specifically describes the process from the user accessing the application to submitting a creation request. Among them, the user terminal obtains the list access request for the user to access the subroutine list of the application and sends it to the server, and receives the service access entry of the agent service and the subroutine access entries under at least one associated category returned by the server after the user logging in to the application meets the access conditions of the agent service; generates and displays the subroutine list page according to the service access entry, at least one associated category, and the subroutine access entries under each associated category; correspondingly, the server receives the list access request for the user to access the subroutine list of the application sent by the user terminal, and returns the service access entry of the agent service and the subroutine access entries under at least one associated category after the user logging in to the application meets the access conditions of the agent service; Furthermore, the user terminal obtains the service access request submitted by the user through the service access entry of the agent service in the subroutine list page and sends it to the server, receives the first service category and the second service category returned by the server, and generates and displays the service access page according to the first service category material corresponding to the first service category and the second service category material corresponding to the second service category; correspondingly, the server receives the service access request for the agent service sent by the user terminal, reads the first service category and the second service category and returns them to the user terminal; Furthermore, the user terminal obtains an access request for creating an agent for a service category (any first service category, that is, any service category for which no agent has been created) submitted by the user through the service access page and sends it to the server, receives the user identification and user occupation data returned by the server, and generates and displays a creation reminder page based on the user identification data and the access identity material generated based on the user occupation data; correspondingly, the server reads the user identification data of the application and the user occupation data authenticated by the user under the credit authentication service according to the access request for creating an agent for the service category sent by the user terminal and returns it to the user terminal; Furthermore, the user terminal obtains a creation request for creating an agent for the service category in the application submitted by the user based on the creation reminder page and sends it to the server, and the server receives the creation request for creating an agent for the service category in the application sent by the user terminal. Optionally, the creation request is submitted based on the creation reminder page configured with the access identity material corresponding to the user identification data and the user occupation data.

[0119] It should be noted that the creation request contains creation data; optionally, the creation data is collected from the creation reminder page; the creation data includes at least one of the following: agent name, agent identification image, user category.

[0120] In specific implementation, since the created agent is a personalized agent for the user and the service category, in order to ensure the effectiveness of the created agent, in an optional implementation manner provided in this embodiment, after receiving the creation request, the server first performs data detection on the creation data carried in the creation request. If the detection passes, it returns the user occupation data and the user category included in the creation data to the user terminal, and executes in response to the creation request to create an agent for the user under the service category according to the agent template corresponding to the service category; if the detection fails, it determines the detection failure keyword and interacts with the user terminal according to the detection failure keyword for creating data editing; Optionally, the data detection includes data integrity detection and / or data validity detection; for example, detecting whether the created data contains agent creation parameters, that is, whether it contains the agent name, the agent identification image, and the user category. If so, it is determined that the data integrity detection passes; if not, it is determined that the data integrity detection fails, and the agent parameters to be supplemented are determined as the detection failure keywords and sent to the user terminal, so that the user terminal generates and displays an agent parameter supplement page corresponding to the detection failure keywords, obtains the agent parameters to be supplemented collected through the agent parameter supplement page and sends them to the server. In addition, the user terminal can also display a creation reminder page with a to-be-supplemented mark for the agent parameters to be supplemented, obtain the collected agent parameters to be supplemented and send them to the server; for another example, detecting whether the agent name included in the creation is a duplicate name or an abnormal name. If not, it is determined that the data validity detection passes; if so, a name abnormality reminder is generated and returned to the user terminal. The user terminal displays a reminder for re-entering the agent name through the creation reminder page, obtains the updated agent name entered and sends it to the server until the server passes the data detection of the created data.

[0121] Specifically, after the server receives a creation request sent by the user terminal for creating an agent of a service category in the application, if the data detection of the creation data carried in the creation request passes, the server returns the user's professional data and the user category included in the creation data to the user terminal for interactive display of agent creation, and creates an agent for the user under the service category according to the agent template corresponding to the service category. In addition to returning the user's professional data and the user category included in the creation data to the user terminal for interactive display of agent creation, the server can also return a creation interaction reminder to the user terminal. In response to the creation interaction reminder, the user terminal performs interactive display of agent creation according to the user's professional data and the user category.

[0122] Step S1204, perform interactive display of the agent creation according to the interaction materials corresponding to the user category and the identity materials generated based on the user's professional data.

[0123] In specific implementation, in order to improve the efficiency of agent creation, after the server receives a creation request sent by the user terminal for creating an agent of a service category in the application, in response to the creation request, the server creates an agent for the user under the service category according to the agent template corresponding to the service category. The agent template in this embodiment includes an agent template that pre-configures the basic structure of the agent; for example, the agent template includes a benchmark large language model and at least one data processing structure; the agent templates of different users under the same service category are the same; that is, the agent parameters of the agents just created by different users under the same service category are different, but the structures of the agents are the same.

[0124] In the process of specifically creating an agent for a user under a service category according to the agent template corresponding to the service category, the agent code associated with the agent identifier or agent name in the creation data can be first generated, and the agent code can be audited. When the agent code passes the audit, it is determined that the agent creation is completed.

[0125] It should be noted that agent creation includes front-end agent creation and / or back-end agent creation. To enhance the user's perception of the front-end agent, the front-end agent can be in the form of a subroutine, that is, the agent is accessed through the subroutine. To improve the efficiency of agent creation, the subroutine creation template of the application can be used to create a subroutine for agent access. Correspondingly, the back-end agent can also be created using the creation process and agent template at the back-end of the application's subroutine.

[0126] Since the creation of an agent takes a certain amount of time, in an optional implementation provided in this embodiment, in order to enhance the user's mind, after the user terminal sends a creation request to the server, that is, during the process of the server creating the agent, the interactive display of agent creation is performed according to the interactive materials corresponding to the user category and the identity materials generated based on the user's occupation data; In the specific process of the user terminal performing the interactive display of agent creation according to the interactive materials and identity materials, an interactive animation can be generated and played based on the interactive materials and identity materials, thereby enhancing the user's perception of the agent creation process. Specifically, the interactive materials and identity materials can be added to the interactive animation template to obtain the interactive animation and display it. Optionally, if the user category is the first category (male), the interactive material is the pre-configured animated image of a male. If the user category is the second category (female), the interactive material is the pre-configured animated image of a female. The identity materials generated based on the user's occupation data are similar to the above access identity materials, and the user qualification data and the desensitized user identity data can be displayed.

[0127] For example, as Figure 6 shown in an interactive display page at a certain interactive display time, there is agent reminder data "Hi, I'm your exclusive fitness assistant", the animated image of the first category, and an identity business card.

[0128] It should be noted that, as Figure 6 shown, the display content of the interactive display page is different at different times. For example, the animated image can wave from far to near, and the identity business card can slide out from below. The specific animation form can be configured according to actual needs, and this embodiment does not limit it here. Optionally, the user can enter the credit authentication service through the identity business card to view the user qualification data and user identity data.

[0129] Step S1206: According to the agent access data returned by the server after the agent is created, enter the subroutine for agent access of the agent.

[0130] In this embodiment, after the server creates an agent for the user under the service category according to the agent template corresponding to the service category, it returns the agent access data to the user terminal; optionally, the user terminal enters the subroutine for agent access of the agent according to the agent access data.

[0131] The agent access data includes the subroutine identifier of the subroutine for agent access and / or the data displayed on the agent access page displayed by the subroutine.

[0132] In specific implementation, after the server detects the creation of the agent, it returns the agent access data to the user terminal. Since the subroutine for agent access corresponds to the agent, during the process of returning the agent access data to the user terminal, the agent can send the agent access data to the user terminal; in addition, the server can first return the subroutine identifier for agent access to the user terminal, and the user terminal enters the subroutine for agent access according to the subroutine identifier. The user terminal sends an agent access request to the agent through the subroutine, and the agent, that is, the backend agent, reads and / or generates agent recommendation data according to the agent access request and returns it to the subroutine. Optionally, the agent recommendation data includes user occupation data, user category, and / or interaction reminder data. Optionally, the agent is created according to the agent template corresponding to the service category.

[0133] In the specific execution process, in order to improve the user's perception of the created agent, in an optional implementation manner provided in this embodiment, after the user terminal enters the subroutine for agent access according to the agent access data returned by the server after the agent is created, it first generates access identity materials according to the user occupation data carried in the agent access data, and then generates and displays an agent access page according to the access identity materials, interaction materials, and interaction reminder data carried in the agent access data; or after the user terminal enters the subroutine for agent access, it first reads the stored identity materials or access identity materials, interaction materials, and interaction reminder data corresponding to the service category, and generates and displays an agent access page according to the identity materials or access identity materials, interaction materials, and interaction reminder data corresponding to the service category.

[0134] In addition, after the user terminal enters the subroutine for agent access according to the agent access data returned by the server, it can also send an agent access request to the agent through the subroutine. The agent reads and / or generates agent recommendation data or agent recommendation instructions according to the agent access request and returns them to the subroutine. The user terminal receives the agent recommendation data through the subroutine, generates an agent access page according to the agent recommendation data, and displays it. Alternatively, the user terminal receives the agent recommendation instructions through the subroutine, reads the stored agent recommendation data according to the agent recommendation instructions, generates an agent access page according to the agent recommendation data, and displays it.

[0135] For example, as Figure 7 shown in the agent access page, that is, the agent access home page displays an identity card 701, interactive materials 702, and interactive reminder data 703; optionally, the interactive reminder data includes a quick interaction reminder text and a quick service entry. The user can efficiently submit interaction data through the quick service entry, saving the time consumed by entering interaction data through voice or text; among them, the quick service entry can be configured according to the agent template. For example, if the agent template is configured with "auxiliary use description of the fitness intelligent assistant", then "introduce what you can assist me to do" can be used as the quick service entry; for another example, if the large language model in the agent template has a course management function, then "course management" can be used as the quick service entry; in addition, the quick service entry can also be configured according to the user qualification data. If it is detected that the user occupation data contains the user qualification data, then "certificate management" is used as the quick service entry; in addition, the quick service entry can also be pre-configured, which is not limited in this embodiment; it should be noted that the identity cards or credit cards displayed on different pages can be the same or different, and can be specifically generated according to the actual configuration, which is not limited in this embodiment; In addition to configuring the quick service entry in the interactive reminder data, a separate quick service entry can also be configured. For example, three quick service entries of knowledge management, course management, and certificate management are displayed in the quick service entry area 704; it should be noted that the determination method of the quick service entry included in the interactive reminder data and the determination method of the quick service entry configured in the quick service entry area can be the same or different, and the specific quick service entries can be the same or different, and can be specifically configured according to the actual needs, which is not limited in this embodiment.

[0136] It should be noted that, in order to improve the convenience of entering interaction data by accessing the page through the agent, interaction data can be submitted by triggering the quick service entry in the page accessed by the agent, or by entering text and / or voice in the page accessed by the agent. Optionally, the page accessed by the agent is configured with a text box for text entry and / or a voice control for voice entry; after detecting that the voice control is triggered, the user terminal collects the user's voice as interaction data according to the authorization data of the voice collection control.

[0137] In specific implementation, after the user terminal enters the interaction data through the page accessed by the agent, it sends the data to the server or to the agent. In the specific interaction process, the server refers to the agent; after receiving the interaction data sent by the user terminal, the agent inputs the interaction data into the large language model in the agent to generate response data, obtains the response data and returns it to the user terminal; specifically, after receiving the interaction data, the agent first queries whether there is response data configured for the interaction data. If so, it returns the response data to the user terminal; if not, it inputs the interaction data into the large language model to generate response data, obtains the response data and returns it to the user terminal. In order to improve the effectiveness of the generated response data, the large language model first performs intent recognition on the interaction data to obtain the interaction intent, and then generates response data according to the data generation strategy corresponding to the interaction intent; in an optional implementation manner provided in this embodiment, the large language model can generate response data in the following manner: Perform intent recognition on the interaction data through the intent recognition module to obtain the interaction intent; Determine the data generation module corresponding to the interaction intent, and generate initial response data for the interaction data through the data generation module to obtain the initial response data; Summarize the interaction data and the initial response data through the data summary module to obtain the response data.

[0138] In the specific execution process, first input the interaction data into the intent recognition sub-model in the large language model for intent recognition to obtain the interaction intent, determine the data generation module corresponding to the interaction intent, and input the interaction data into the data generation module to generate initial response data. Finally, in order to improve the user's perception of the response data and the interaction data, summarize the interaction data and the initial response data through the data summary module to obtain the response data.

[0139] Specifically, configure the data generation module corresponding to each interaction intent in the large language model. After determining the interaction intent of the interaction data, generate response data for the interaction data through the data generation module corresponding to the interaction intent to obtain the response data.

[0140] The following takes the process of generating two initial response data as an example to illustrate the process of generating initial response data by different data generation modules.

[0141] (1) In an optional implementation manner provided in this embodiment, the data generation module corresponding to the knowledge management intention generates initial response data in the following manner: Query knowledge data in a preset knowledge base according to the intention data included in the interaction data to obtain first knowledge data, and / or query knowledge data in a dataset according to the intention data to obtain second knowledge data; Input the first knowledge data, the second knowledge data, and the third knowledge data included in the interaction data into a summary sub-module for data summary to obtain initial response data; Specifically, after the user terminal obtains the intention data entered by the user through the intelligent agent access page, if the acquisition prompt data corresponding to the intention data is detected, the acquisition prompt data is displayed through the intelligent agent access page, and the intention data and the data to be processed entered by the user based on the acquisition prompt data are determined as the interaction data; if the acquisition prompt data is not detected, the intention data is directly used as the interaction data and sent to the intelligent agent; Under the knowledge management intention, after the user terminal obtains the intention data of knowledge management entered by the user through the intelligent agent access page and detects the intention data, it displays the acquisition prompt data through the intelligent agent dialogue page shown in, for example Figure 8 The user can upload files and / or images based on the "file upload" control and / or "image upload" control in the acquisition prompt data, and / or enter voice data or text data as the third knowledge data through voice input or text input and send it to the intelligent agent together with the intention data as interaction data; after receiving the interaction data, the intelligent agent inputs the interaction data into a large language model for generating response data. During the process of generating response data by the large language model, it first performs intention recognition on the interaction data to obtain the knowledge management intention, and then according to the data generation strategy corresponding to the knowledge management intention, first performs retrieval enhancement according to the intention data to obtain first knowledge data, and queries knowledge data in the dataset of the intelligent agent according to the intention data to obtain second knowledge data, and then inputs the first knowledge data, the second knowledge data, and the third knowledge data included in the interaction data into the summary sub-module for data summary to obtain initial response data; thereafter, the interaction data and the initial response data are input into a data summary module for data summary to obtain response data including at least the intention data and the initial response data.

[0142] (2) In an optional implementation manner provided in this embodiment, the data generation module corresponding to the question-and-answer intention generates initial response data in the following manner: Query for response data in the database according to the question data included in the interaction data; If the query is successful, determine the retrieved data as the initial response data; If the query fails, input the question data into the Q&A sub-module to generate an answer, and obtain the initial response data.

[0143] Specifically, during the process of generating the initial response data, the data generation module corresponding to the Q&A intent first queries the response data of the question data included in the interaction data in the agent's database. If the query fails, it inputs the question data into the Q&A sub-module to generate an answer and obtains the initial response data.

[0144] In practical applications, due to the strong randomness of the interaction data submitted by users, the data generation module corresponding to the interaction intent in the agent template may not be comprehensive enough, which may lead to the inability to accurately and effectively respond to the user's interaction data. Therefore, in an alternative implementation provided in this embodiment, if the determined data generation module corresponding to the interaction intent is empty, the agent configures the corresponding data generation module according to the interaction intent, that is, the configuration of the agent tool. In this way, the accuracy and effectiveness of the agent's response to the interaction data of various interaction intents are improved, and thus the user's perception of the agent is enhanced. Optionally, the agent tool includes a module in the agent responsible for calling external tools and services; such as the RAG module, weather module, etc.

[0145] In addition to configuring the agent tool to update the agent, the large language model can also be trained to improve the quality of the response data generated by the large language model. Specifically, in an alternative implementation provided in this embodiment, after inputting the interaction data into the large language model in the agent to generate response data and obtaining the response data, the agent also performs the following operations: Input the response data and the detection prompt data into the large language model for response quality detection to obtain the response quality detection result; Perform model training on the large language model according to the response quality detection result, input the interaction data into the trained large language model for secondary response data generation, and return the response data obtained from the secondary response data generation to the subroutine.

[0146] It should be noted that as Figure 7 shown, the agent access home page displays an identity card, interaction materials, and interaction reminder data. During the interaction with the user, in order to enhance the user's perception of the interaction content, that is, the conversation content, it can be jumped to the agent conversation page as Figure 8 shown for displaying the conversation content. Optionally, the agent access page includes the agent access home page and / or the agent conversation page.

[0147] In addition, the agent may not be configured with a large language module, but with an intention recognition model, at least one data processing module and a data summarization model, for example Figure 9 The agent schematic diagram shown includes a decision model, wherein the decision model includes a large language model and a function fine-tuning module, and the large language model here can be used for intent recognition; in addition, the agent also includes a RAG module, agent tools 1 to agent tools n, and a summary model; the agent update and response data generation can be performed according to the above-mentioned related contents, and this embodiment will not be repeated here. Among them, the RAG module may include a preset knowledge base, a vector recall submodule, and a summary submodule; the vector recall module user vectorizes the intent data in the interaction data and then performs vector recall from the preset knowledge base.

[0148] After creating an agent of the service category, the agent can be accessed through the service access page. In order to enhance the user's perception of agent access, in an optional implementation provided by this embodiment, after receiving an agent access request sent by a user terminal, the agent performs the following operations: Read the user's historical access time to the agent according to the agent access request to the agent sent by the user terminal; Perform access time interval detection based on historical access time and agent access time carried in agent access request; The corresponding agent access data is generated according to the detection results and returned to the user terminal for display through the subroutine on the user terminal.

[0149] Specifically, after receiving an agent access request sent by a user terminal, if the time interval between the agent access time and the historical access time is greater than an interval threshold, the agent recommendation data or agent recommendation instructions are used as agent access data and returned to the user terminal; optionally, the user terminal displays the agent access homepage; if the time interval is less than or equal to the interval threshold, the historical interaction data or historical interaction instructions are returned to the user terminal; optionally, the user terminal displays an agent dialogue page including historical interaction data.

[0150] In addition to creating and accessing intelligent entities through the service access page, intelligent entities can also be created and accessed through the user career data display page corresponding to the service category in the credit authentication service. Specifically, except for the different initiation locations, the other processes are similar to the process of creating and accessing intelligent entities through the service access page. Please refer to the relevant content, and this embodiment will not be repeated here.

[0151] In the specific execution process, after a user creates an agent, the user can share the agent with other users. For example, a fitness coach can share the agent with fitness trainees so that the fitness trainees can view the knowledge and courses in the agent. In addition, the fitness trainees can also ask questions through the agent. Specifically, in an optional implementation manner provided in this embodiment, the agent receives an agent access request sent by a user terminal, determines an access category according to the access user identifier carried in the agent access request and the agent user identifier of the user to whom the agent belongs, generates agent access data corresponding to the access category, and returns it to the user terminal. Optionally, the access category includes a first access category and a second access category. Among them, the first access category includes the access category for a user to access their own agent, and the second access category includes the access category for other users to access the agent of the user to whom the agent belongs.

[0152] Specifically, in the process of a fitness coach accessing their own agent and a fitness trainee accessing the fitness coach's agent, the service quick entry may be different, and the others can be the same or configured differently according to actual needs. This embodiment does not make a limitation here.

[0153] During specific implementation, the agent accesses through a subroutine of the user terminal. To improve the user's perception of the subroutine for agent access, in an optional implementation manner provided in this embodiment, during the process of the user accessing the subroutine list, the server detects whether the user is in the agent according to the list access request for accessing the subroutine list sent by the user terminal. If not, it returns the subroutine access entry and / or service access entry of other subroutines to the user terminal. If so, it returns the agent access entry and the subroutine access entry of other subroutines to the user terminal to generate a subroutine list page based on the agent access entry and the subroutine access entry. Optionally, after the agent access entry is triggered, it enters the subroutine for agent access to perform agent access. It should be noted that if the subroutine for agent access is associated, the subroutine list page is generated according to the association category. That is, the agent access entry is configured under the corresponding association category.

[0154] In summary, in the process of creating an agent in cooperation with the server, the user terminal obtains the creation request submitted by the user based on the user occupation data corresponding to the service category for creating an agent of the service category in the application program and sends it to the server, so that the server creates an agent for the user under the service category according to the agent template corresponding to the service category. In this way, by creating an agent through the agent template, the creation efficiency of the agent is improved. In this process, the user terminal performs an interactive display of agent creation according to the interactive materials corresponding to the user category and the identity materials generated based on the user occupation data. In this way, the user's perception of the agent creation process is improved. If agent access data is detected, the subroutine for agent access is entered according to the agent access data, and the agent is interacted with through the agent access page displayed by the subroutine. In this way, an agent for the user under the service category is created to provide service data under the service category through the agent, improving the effectiveness and convenience of the user's access to and management of service data under the service category. In the process of creating an agent, the user occupation data corresponding to the service category authenticated by the user through the credit authentication service is also used as one of the creation parameters, improving the effectiveness of the created agent and the user's or the accessing user's trust in the agent.

[0155] The following takes the application of an agent creation method provided in this embodiment in the fitness intelligent assistant creation scenario as an example to further illustrate the agent creation method provided in this embodiment. As Figure 10 described, the agent creation method applied to the fitness intelligent assistant creation scenario specifically includes the following steps.

[0156] Step S1002, obtain the list access request for the user to access the subroutine list of the application program and send it to the server.

[0157] Step S1008, receive the service access entry of the intelligent assistant service and the subroutine access entries under at least one associated category returned by the server after the user logging in to the application program meets the admission conditions of the intelligent assistant service.

[0158] Step S1010, generate a subroutine list page according to the service access entry, at least one associated category, and the subroutine access entries under each associated category and display it.

[0159] Step S1012, obtain the service access request submitted by the user through the service access entry of the intelligent assistant service in the subroutine list page and send it to the server.

[0160] Step S1016, obtain the access request for creating and accessing the fitness intelligent assistant submitted by the user through the service access page and send it to the server.

[0161] Step S1020: Receive the user identification data and fitness qualification data returned by the server, and generate and display a creation reminder page based on the user identification data and the access identity material generated based on the user's occupation data.

[0162] Step S1022: Obtain the creation request for creating a fitness intelligent assistant submitted by the user on the creation reminder page in the application and send it to the server.

[0163] Step S1030: In response to the interactive display reminder, perform an interactive display for creating a fitness intelligent assistant based on the interactive material and the identity material.

[0164] Step S1034: Enter a subroutine for accessing the fitness intelligent assistant according to the fitness intelligent assistant access data returned by the server.

[0165] The following takes the application of an intelligent agent creation method provided in this embodiment in the training intelligent agent creation scenario as an example to further illustrate the intelligent agent creation method provided in this embodiment. As Figure 11 shown, the intelligent agent creation method applied to the training intelligent agent creation scenario specifically includes the following steps.

[0166] Step S1102: Obtain the training intelligent agent access request submitted by the user through the training certificate page of the credit authentication service in the application and send it to the server.

[0167] Step S1110: Receive the user identification data and training qualification data returned by the server, and generate and display a creation reminder page based on the user identification data and the access identity material generated based on the user's occupation data.

[0168] Step S1112: Obtain the creation request for creating a training intelligent agent submitted by the user on the creation reminder page in the application and send it to the server.

[0169] Step S1118: In response to the interactive display reminder, perform an interactive display for creating a training intelligent agent based on the interactive material and the identity material.

[0170] Step S1122: Enter a subroutine for accessing the training intelligent agent according to the training intelligent agent access data returned by the server.

[0171] One or more embodiments of an intelligent agent-based interaction processing method provided in this specification are as follows: Refer to Figure 13 , the intelligent agent-based interaction processing method provided in this embodiment can be applied to an intelligent agent, and the method specifically includes Step S1302 to Step S1306.

[0172] Step S1302: Obtain the interaction data entered by the user through the intelligent agent access page.

[0173] Optionally, the agent accesses the page through a subroutine for agent access of the agent; the agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent; the agent is created according to an agent template corresponding to the service category. The user in this embodiment includes the user to whom the agent belongs. In addition, the user can also be other accessing users other than the user to whom the agent belongs.

[0174] It should be noted that the agent in this embodiment includes the agent of the service category in the application program created by using the agent creation method provided in the above embodiment; therefore, the relevant content of the above embodiment can be referred to when reading this embodiment; the relevant process of this embodiment can also be referred to when reading the above embodiment.

[0175] It should also be noted that the agent created by using the agent creation method provided in the above embodiment can be an agent deployed on the user terminal. If it is an agent deployed on the user terminal, the response data is displayed by executing steps S1302 to S1306. In addition, it can also be an agent deployed on the server. If it is an agent deployed on the server, the agent is accessed on the user terminal through a subroutine; the subroutine includes the subroutine for agent access created during the agent creation process; in this case, the user terminal obtains the interaction data entered by the user through the agent access page displayed by the subroutine for agent access and sends it to the agent. The agent inputs the interaction data into the large language model in the agent to generate response data, obtains the response data and sends it to the subroutine, and the response data is received and displayed through the agent access page in the subroutine.

[0176] Step S1304, input the interaction data into the large language model in the agent to generate response data, and obtain the response data.

[0177] The above embodiment specifically describes the generation process of the response data. Whether the response data is generated on the user terminal or on the server side in this embodiment is similar to the generation process of the response data provided in the above embodiment, and the relevant content in the above embodiment can be referred to.

[0178] Step S1306, display the response data through the agent access page.

[0179] In this embodiment, the user can access the agent by themselves and submit interactive data. In addition, after creating an agent of a service category, the user can share the agent with the accessing users, that is, share it with other users. The accessing users can access the agent by entering the subroutine for agent access, view the knowledge and courses uploaded by the user through accessing the agent, and interact with the agent to answer the submitted question data. In addition, the accessing users can also purchase the courses uploaded by the user to the agent based on the agent access.

[0180] In the specific execution process of the user creating an agent, the user can submit an access request for agent creation access through the service access page of the user terminal, and then enter the creation data and submit the creation request through the creation reminder page in response to the access request; in addition, the user can also submit an access request for agent creation access through the user qualification data page corresponding to the service category in the credit authentication service displayed by the user terminal, and then enter the creation data and submit the creation request through the creation reminder page in response to the access request; it should be noted that the user can also directly submit a creation request for agent creation through the user qualification data page corresponding to the service category in the credit authentication service displayed by the user terminal, without obtaining the creation data through the creation reminder page, and use the application identification data of the user in the application as the creation data.

[0181] It should be noted that in the page displaying the identity material or access identity material generated based on the user's occupation data provided in the above embodiment, the credit authentication service can be entered through the qualification data viewing control in the identity material or access identity material, and the corresponding user qualification data page can be displayed; Similarly, in the specific execution process of the user accessing the agent, the user can submit an agent access request for agent access through the service access page displayed by the user terminal. In addition, the user can also submit an agent access request for agent access through the user qualification data page corresponding to the service category in the credit authentication service displayed by the user terminal.

[0182] In the process of the accessing user accessing the agent, the user can access the user's agent by entering the subroutine for agent access in the form of a link, password, or in-app message. In addition, the agent access request for the user's agent can also be submitted on the service front page corresponding to the service category, and the subroutine for agent access can be entered in response to the access request to access the user's agent.

[0183] An embodiment of an agent creation device provided in this specification is as follows: In the above embodiment, an agent creation method is provided. Correspondingly, an agent creation device is also provided, which will be described below with reference to the accompanying drawings.

[0184] Reference Figure 14 , which shows a schematic diagram of an agent creation device embodiment provided in this embodiment.

[0185] Since the device embodiment corresponds to the method embodiment, the description is relatively simple. For the relevant parts, please refer to the corresponding description of the method embodiment provided above. The device embodiments described below are merely illustrative.

[0186] This embodiment provides an agent creation device that can run on a server. The device includes: A creation request receiving module 1402, configured to receive a creation request for creating an agent of a service category sent by a user terminal; the creation request is submitted based on user occupation data corresponding to the service category; An agent creation module 1404, configured to respond to the creation request and create an agent for the user under the service category according to an agent template corresponding to the service category; After detecting the agent creation, a running data return module 1406. The data return module 1406 is configured to return agent access data to the user terminal after detecting the agent creation; the user terminal enters a subroutine for agent access of the agent according to the agent access data.

[0187] Another embodiment of the agent creation device provided in this specification is as follows: In the above embodiment, another agent creation method is provided. Correspondingly, another agent creation device is also provided. The following is described with reference to the drawings.

[0188] Reference Figure 15 , which shows a schematic diagram of an agent creation device embodiment provided in this embodiment.

[0189] Since the device embodiment corresponds to the method embodiment, the description is relatively simple. For the relevant parts, please refer to the corresponding description of the method embodiment provided above. The device embodiments described below are merely illustrative.

[0190] This embodiment provides an agent creation device that can run on a user terminal. The device includes: A creation request sending module 1502, configured to obtain a creation request for creating an agent of a service category by a user and send it to a server; the creation request is submitted based on user occupation data corresponding to the service category; An interaction display module 1504, configured to perform an interaction display for the agent creation according to interaction materials corresponding to the user category and identity materials generated based on the user occupation data; The subroutine enters module 1506, which is configured to enter the subroutine for agent access of the agent according to the agent access data returned by the server after the agent is created; the agent is created according to the agent template corresponding to the service category.

[0191] An embodiment of an agent-based interaction processing device provided in this specification is as follows: In the above embodiment, an agent-based interaction processing method is provided. Correspondingly, an agent-based interaction processing device is also provided. The following will be described with reference to the accompanying drawings.

[0192] Refer to Figure 16 , which shows a schematic diagram of an embodiment of an agent-based interaction processing device provided in this embodiment.

[0193] Since the device embodiment corresponds to the method embodiment, the description is relatively simple. For the relevant parts, please refer to the corresponding description of the method embodiment provided above. The device embodiments described below are merely illustrative.

[0194] This embodiment provides an agent-based interaction processing device that can run on an agent. The device includes: An interaction data acquisition module 1602, which is configured to acquire interaction data entered by a user through an agent access page; the agent access page is displayed through the subroutine for agent access of the agent; the agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent; the agent is created according to the agent template corresponding to the service category. A response data generation module 1604, which is configured to input the interaction data into a large language model in the agent to generate response data and obtain the response data. A response data display module 1606, which is configured to display the response data through the agent access page.

[0195] An embodiment of a server provided in this specification is as follows: Corresponding to the above-described agent creation method, based on the same technical concept, one or more embodiments of this specification also provide a server, which is used to execute the above-provided agent creation method. Figure 17 It is a schematic diagram of the structure of a server provided in one or more embodiments of this specification.

[0196] A server provided in this embodiment includes: As Figure 17As shown, servers can vary significantly due to differences in configuration or performance. A server may include one or more processors 1701 and a memory 1702. The memory 1702 may store one or more applications or data. Among them, the memory 1702 may be volatile storage or persistent storage. The applications stored in the memory 1702 may include one or more modules (not shown in the figure), and each module may include a series of computer-executable instructions in the server. Further, the processor 1701 may be configured to communicate with the memory 1702 and execute a series of computer-executable instructions in the memory 1702 on the server. The server may also include one or more power supplies 1703, one or more wired or wireless network interfaces 1704, one or more input / output interfaces 1705, etc.

[0197] In a specific embodiment, the server includes a memory and one or more programs. One or more of the programs are stored in the memory, and one or more of the programs may include one or more modules. Each module may include a series of computer-executable instructions in the server and is configured to be executed by one or more processors. The one or more programs include the following computer-executable instructions for: Receiving a creation request sent by a user terminal for creating an agent of a service category in an application; the creation request is submitted based on user occupation data corresponding to the service category; Responding to the creation request, creating an agent for the user under the service category according to an agent template corresponding to the service category; After detecting the creation of the agent, returning agent access data to the user terminal; the user terminal enters a subroutine for accessing the agent according to the agent access data.

[0198] An embodiment of a user terminal provided in this specification is as follows: Corresponding to the above-described agent creation method, based on the same technical concept, one or more embodiments of this specification also provide a user terminal. The user terminal is used to execute the above-provided agent creation method. Figure 18 It is a schematic structural diagram of a user terminal provided by one or more embodiments of this specification.

[0199] A user terminal provided in this embodiment includes: As Figure 18As shown, user terminals can vary significantly due to differences in configuration or performance. They can include one or more processors 1801 and a memory 1802. One or more application programs or data can be stored in the memory 1802. Among them, the memory 1802 can be transient storage or persistent storage. The application programs stored in the memory 1802 can include one or more modules (not shown in the figure), and each module can include a series of computer-executable instructions in the user terminal. Further, the processor 1801 can be configured to communicate with the memory 1802 and execute a series of computer-executable instructions in the memory 1802 on the user terminal. The user terminal can also include one or more power supplies 1803, one or more wired or wireless network interfaces 1804, one or more input / output interfaces 1805, one or more keyboards 1806, etc.

[0200] In a specific embodiment, the user terminal includes a memory and one or more programs. One or more of the programs are stored in the memory, and one or more of the programs can include one or more modules. Each module can include a series of computer-executable instructions in the user terminal and is configured to be executed by one or more processors. The one or more programs include the following computer-executable instructions: Obtain a creation request for creating an agent of a service category in an application program and send it to the server; the creation request is submitted based on user occupation data corresponding to the service category; Perform an interactive display for creating the agent based on interactive materials corresponding to the user category and identity materials generated based on the user occupation data; According to the agent access data returned by the server after the agent is created, enter a subroutine for accessing the agent; the agent is created according to an agent template corresponding to the service category.

[0201] An embodiment of an interactive processing device based on an agent provided in this specification is as follows: Corresponding to the above-described interactive processing method based on an agent, based on the same technical concept, one or more embodiments of this specification also provide an interactive processing device based on an agent. This interactive processing device based on an agent is used to execute the above-provided interactive processing method based on an agent. Figure 19 It is a schematic structural diagram of an interactive processing device based on an agent provided by one or more embodiments of this specification.

[0202] An interactive processing device based on an agent provided in this embodiment includes: As Figure 19As shown, agent-based interactive processing devices can vary significantly due to differences in configuration or performance. They can include one or more processors 1901 and a memory 1902. One or more application programs or data can be stored in the memory 1902. Among them, the memory 1902 can be transient storage or persistent storage. The application programs stored in the memory 1902 can include one or more modules (not shown in the figure), and each module can include a series of computer-executable instructions in the agent-based interactive processing device. Further, the processor 1901 can be configured to communicate with the memory 1902 and execute a series of computer-executable instructions in the memory 1902 on the agent-based interactive processing device. The agent-based interactive processing device can also include one or more power supplies 1903, one or more wired or wireless network interfaces 1904, one or more input / output interfaces 1905, one or more keyboards 1906, etc.

[0203] In a specific embodiment, the agent-based interactive processing device includes a memory and one or more programs. One or more of the programs are stored in the memory, and one or more of the programs can include one or more modules. Each module can include a series of computer-executable instructions in the agent-based interactive processing device and is configured to be executed by one or more processors. The one or more programs include the following computer-executable instructions for: Obtain interaction data entered by the user through the agent accessing the page; the agent access page is displayed through a subroutine for the agent access of the agent; the agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent; the agent is created according to an agent template corresponding to the service category; Input the interaction data into a large language model in the agent to generate response data and obtain the response data; Display the response data through the agent access page.

[0204] An embodiment of a computer-readable storage medium provided in this specification is as follows: Corresponding to the above-described agent creation method, based on the same technical concept, one or more embodiments of this specification also provide a computer-readable storage medium.

[0205] The computer-readable storage medium provided in this embodiment is used to store computer-executable instructions, and the computer-executable instructions, when executed, implement the following process: Receive a creation request sent by a user terminal for creating an agent of a service category in an application; the creation request is submitted based on user occupation data corresponding to the service category; In response to the creation request, create an agent for the user under the service category according to an agent template corresponding to the service category; After detecting the creation of the agent, return agent access data to the user terminal; the user terminal enters a subroutine for accessing the agent according to the agent access data.

[0206] It should be noted that the embodiments of a computer-readable storage medium in this specification and the embodiments of an agent creation method in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the corresponding method described above, and the repeated parts will not be elaborated.

[0207] The second embodiment of the computer-readable storage medium provided in this specification is as follows: Corresponding to the above-described another agent creation method, based on the same technical concept, one or more embodiments of this specification also provide another computer-readable storage medium.

[0208] The computer-readable storage medium provided in this embodiment is used to store computer-executable instructions, and the computer-executable instructions, when executed, implement the following processes: Obtain a creation request for creating an agent of a service category in an application by a user and send it to a server; the creation request is submitted based on user occupation data corresponding to the service category; Perform an interactive display for the agent creation according to interactive materials corresponding to the user category and identity materials generated based on the user occupation data; According to the agent access data returned by the server after the agent is created, enter a subroutine for accessing the agent; the agent is created according to an agent template corresponding to the service category.

[0209] It should be noted that the embodiments of the second computer-readable storage medium in this specification and the embodiments of another agent creation method in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the corresponding method described above, and the repeated parts will not be elaborated.

[0210] The third embodiment of the computer-readable storage medium provided in this specification is as follows: Corresponding to the above-described an interactive processing method based on an agent, based on the same technical concept, one or more embodiments of this specification also provide a computer-readable storage medium.

[0211] The computer-readable storage medium provided in this embodiment is used to store computer-executable instructions, and when the computer-executable instructions are executed, the following processes are implemented: Obtain the interaction data entered by the user through the agent access page; the agent access page is displayed through a subroutine for agent access of the agent; the agent is created in response to a creation request submitted based on the user occupation data corresponding to the service category of the agent; the agent is created according to the agent template corresponding to the service category; Input the interaction data into the large language model in the agent to generate response data, and obtain the response data; Display the response data through the agent access page.

[0212] It should be noted that the embodiment of the third computer-readable storage medium in this specification and the embodiment of an agent-based interaction processing method in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the corresponding method described above, and the repeated parts will not be elaborated.

[0213] The following is an embodiment of the first computer program product provided in this specification: Corresponding to the first agent creation method described above, based on the same technical concept, one or more embodiments of this specification also provide a computer program product.

[0214] A computer program product includes a computer program / instructions, and when the computer program / instructions are executed by a processor, the following steps are implemented: Receive a creation request sent by a user terminal for creating an agent of a service category in an application program; the creation request is submitted based on the user occupation data corresponding to the service category; In response to the creation request, create an agent for the user under the service category according to the agent template corresponding to the service category; After detecting the creation of the agent, return agent access data to the user terminal; the user terminal enters a subroutine for agent access of the agent according to the agent access data.

[0215] It should be noted that the embodiment of the first computer program product in this specification and the embodiment of an agent creation method in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the corresponding method described above, and the repeated parts will not be elaborated.

[0216] The following is an embodiment of the second computer program product provided in this specification: Another agent creation method corresponding to the above description. Based on the same technical concept, one or more embodiments of this specification also provide a computer program product.

[0217] A computer program product, including computer programs / instructions, which when executed by a processor implement the following steps: Obtain a creation request for creating an agent of a service category by a user in an application and send it to a server; the creation request is submitted based on user occupation data corresponding to the service category; Perform an interactive display of the agent creation based on interactive materials corresponding to the user category and identity materials generated based on the user occupation data; According to the agent access data returned by the server after the agent is created, enter a subroutine for accessing the agent; the agent is created according to an agent template corresponding to the service category.

[0218] It should be noted that the embodiments of the second computer program product in this specification and the embodiments of another agent creation method in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the corresponding method described above, and the repeated parts will not be elaborated.

[0219] The following is an embodiment of the third computer program product provided in this specification: Corresponding to an agent-based interactive processing method described above. Based on the same technical concept, one or more embodiments of this specification also provide a computer program product.

[0220] A computer program product, including computer programs / instructions, which when executed by a processor implement the following steps: Obtain interactive data entered by a user through an agent access page; the agent access page is displayed through a subroutine for accessing the agent; the agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent; the agent is created according to an agent template corresponding to the service category; Input the interactive data into a large language model in the agent to generate response data and obtain the response data; Display the response data through the agent access page.

[0221] It should be noted that the embodiments of the third computer program product in this specification and the embodiments of an agent-based interactive processing method in this specification are based on the same inventive concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the corresponding method described above, and the repeated parts will not be elaborated.

[0222] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. The key point of each embodiment is to illustrate the differences from other embodiments. For example, the apparatus embodiment, the device embodiment, and the computer-readable storage medium embodiment are all similar to the method embodiment, so the description is relatively simple. To read the relevant content in the apparatus embodiment, the device embodiment, and the computer-readable storage medium embodiment, please refer to the corresponding description in the method embodiment.

[0223] The specific embodiments of this specification are described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0224] In the 1930s, it was obvious to distinguish whether an improvement to a technology was a hardware improvement (e.g., improvement to circuit structures such as diodes, transistors, switches, etc.) or a software improvement (improvement to method flows). However, with the development of technology, many method flow improvements today can be regarded as direct improvements to hardware circuit structures. Almost all designers obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that an improvement to a method flow cannot be implemented with a hardware entity module. 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. The designer can program by himself to "integrate" a digital system on a PLD, without having to ask a chip manufacturer to design and fabricate a dedicated integrated circuit chip. Moreover, nowadays, instead of manually fabricating integrated circuit chips, this programming is mostly implemented using "logic compiler" software, which is similar to the software compiler used in program development and writing. The original code before compilation also has to be written in a specific programming language, which is called a Hardware Description Language (HDL). And there is not only one type of HDL, but many types, 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, RHDL (Ruby Hardware Description Language), etc. Currently, the most commonly used are VHDL (Very-High-Speed Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also be aware that by simply making a little logical programming of the method flow with the above-mentioned several hardware description languages and programming it into an integrated circuit, it is easy to obtain the hardware circuit that implements the logical method flow.

[0225] The controller can be implemented in any suitable manner. For example, the controller can take the form of, for example, a microprocessor or a processor and a computer-readable medium storing computer-readable program code (such as software or firmware) executable by the (micro)processor, logic gates, switches, an application specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of the controller include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art also know that, in addition to implementing the controller in the form of pure computer-readable program code, it is entirely possible to make the controller implement the same function in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, such a controller can be considered a hardware component, and the devices included therein for implementing various functions can also be regarded as the structures within the hardware component. Or even, the devices for implementing various functions can be regarded as either software modules for implementing the method or structures within the hardware component.

[0226] The systems, devices, modules, or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0227] For the convenience of description, when describing the above devices, they are described separately as various units according to their functions. Of course, when implementing the embodiments of this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0228] Those skilled in the art should understand that one or more embodiments of this specification can be provided as a method, a system, or a computer program product. Therefore, one or more embodiments of this specification can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, this specification can take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program code.

[0229] This specification is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the specification. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device generate means for implementing the functions specified in one flow Figure One one flow or multiple flows and / or blocks Figure One or multiple blocks.

[0230] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in one flow Figure One one flow or multiple flows and / or blocks Figure One or multiple blocks.

[0231] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one flow Figure One one flow or multiple flows and / or blocks Figure One or multiple blocks.

[0232] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0233] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0234] Computer-readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. 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 technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.

[0235] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of further restrictions, the elements defined by the sentence "includes at least one ..." do not exclude the presence of other identical elements in the process, method, commodity or device including the elements.

[0236] One or more embodiments of the present specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data categories. One or more embodiments of the present specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0237] The above description is only an embodiment of this document and is not intended to limit this document. For those skilled in the art, this document may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this document should be included in the scope of the claims of this document.

Claims

1. A method for creating an intelligent agent, applied to a server, the method comprising: Receiving a creation request sent by a user terminal for creating an intelligent agent of a service category in an application program; The creation request is submitted based on the user occupation data corresponding to the service category; In response to the creation request, creating an agent of the user under the service category according to an agent template corresponding to the service category; After detecting the creation of the agent, returning agent access data to the user terminal; The user terminal enters a subroutine for performing an agent access to the agent according to the agent access data.

2. The agent creation method according to claim 1, after the step of returning agent access data to the user terminal after detecting the creation of the agent, further comprising: Receiving interaction data sent by the user terminal; The interactive data is entered through the agent access page of the subprogram; The interaction data is input into the large language model in the agent to generate response data, and the response data is obtained and returned to the user terminal.

3. According to the intelligent agent creation method according to claim 1, the subroutine for accessing the intelligent agent is created using the subroutine creation template of the application.

4. The method for creating an agent according to claim 1, before the step of receiving a creation request sent by a user terminal to create an agent of a service category in an application program is executed, further comprising: Reading the user identification data of the application and the user occupation data according to the access request for creating an agent of the service category sent by the user terminal; The user identification data and the user occupation data are returned to the user terminal so as to create a reminder page display according to the user identification data and the user occupation data.

5. The agent creation method according to claim 4, before the step of reading the user identification data and the user occupation data of the application program according to the access request for agent creation access of the service category sent by the user terminal, further comprises: Detecting whether the user meets the access condition of the agent service according to the list access request for accessing the subroutine list sent by the user terminal; If so, the service access portal of the intelligent agent service is returned to the user terminal to generate a sub-program list page based on the service access portal and the sub-program access portal.

6. The method for creating an agent according to claim 5, after the operation of returning the service access entry of the agent service to the user terminal is executed, further comprising: receiving a service access request for the agent service sent by the user terminal; The service access request is submitted based on the service access entry in the subprogram list page; The first service category of the user's created agent and the second service category of the user's uncreated agent are read and returned to the user terminal to display a service access page.

7. The method for creating an agent according to claim 1, before the step of creating an agent for a user under the service category according to the agent template corresponding to the service category is performed, the method further comprises: Performing data detection on the creation data carried in the creation request; If the detection is successful, the user occupation data and the user category included in the creation data are returned to the user terminal for interactive display of the agent creation.

8. The agent creation method according to claim 1, further comprising: Reading the historical access time of the user to the agent according to the agent access request to the agent sent by the user terminal; Perform access time interval detection according to the historical access time and the agent access time carried in the agent access request; The corresponding agent access data is generated according to the detection result and returned to the user terminal so as to be displayed on the user terminal through the subroutine.

9. The agent creation method according to claim 1, further comprising: receiving an agent access request for the agent sent by the user terminal; Determine the access category according to the access user identifier carried in the agent access request and the agent user identifier of the user to which the agent belongs; Generate agent access data corresponding to the access category and return it to the user terminal.

10. The agent creation method according to claim 1, further comprising: Detecting whether the user has an intelligent agent according to a list access request for accessing a subroutine list sent by the user terminal; If so, the agent access entrance and the subprogram access entrance are returned to the user terminal to generate a subprogram list page based on the agent access entrance and the subprogram access entrance; after the agent access entrance is triggered, the subprogram is entered to perform agent access.

11. The method for creating an agent according to claim 2, wherein the step of inputting the interaction data into the large language model in the agent to generate response data and obtaining the response data further comprises: Inputting the response data and the detection prompt data into the large language model to perform response quality detection to obtain a response quality detection result; The large language model is trained according to the response quality detection result, and the interaction data is input into the trained large language model to generate secondary response data.

12. The method for creating an intelligent agent according to claim 2, wherein the generating of the response data comprises: Performing intent recognition on the interaction data through an intent recognition module to obtain an interaction intent; Determine a data generation module corresponding to the interaction intention, and generate initial response data for the interaction data through the data generation module to obtain initial response data; The interaction data and the initial response data are summarized by a data summarizing module to obtain the response data.

13. The agent creation method according to claim 12, after the data generation module operation of determining the interaction intention is executed, further comprising: If it is detected that the data generation module is empty, the corresponding data generation module configuration is performed according to the interaction intention.

14. The method for creating an intelligent agent according to claim 12, wherein the generating of the initial response data comprises: Performing a knowledge data query in a preset knowledge base according to the intention data included in the interaction data to obtain first knowledge data, and / or performing a knowledge data query in a data set according to the intention data to obtain second knowledge data; Inputting the first knowledge data, the second knowledge data and the third knowledge data included in the interaction data into a summarizing submodule for data summarization to obtain the initial response data; or, Performing a reply data query in a database according to the question data contained in the interactive data; If the query is successful, the queried data is determined as the initial response data; If the query fails, the question data is input into the question-answer submodule to generate an answer to obtain the initial response data.

15. A method for creating an intelligent agent, applied to a user terminal, the method comprising: Obtain the creation request of the user for creating an intelligent agent of the service category in the application and send it to the server; The creation request is submitted based on the user occupation data corresponding to the service category; Performing interactive display of the agent creation according to the interactive material corresponding to the user category and the identity material generated based on the user occupation data; Entering a subroutine for performing agent access to the agent according to agent access data returned by the server after the agent is created; The agent is created according to the agent template corresponding to the service category.

16. The method for creating an agent according to claim 15, after the execution of the subroutine step of entering into agent access of the agent according to the agent access data returned by the server after the agent is created, further comprising: Generate access identity material according to user occupation data carried by the agent access data; An agent access page is generated and displayed based on the access identity material, the interaction material and the interaction reminder data carried by the agent access data.

17. The method for creating an agent according to claim 15, before the step of obtaining a user's creation request for creating an agent of a service category in an application and sending the request to a server, further comprising: Obtaining an access request submitted by the user for creating an agent of the service category and sending the request to the server; The access request is submitted based on the service access page of the agent service; Receiving the user identification data and the user occupation data returned by the server, and generating and displaying a creation reminder page according to the user identification data and the access identity material generated based on the user occupation data; Wherein, the creation request is submitted based on the creation reminder page.

18. An agent-based interaction processing method, applied to an agent, the method comprising: Obtain the interaction data entered by the user through the agent access page; The agent access page is displayed by a subroutine for performing agent access of the agent; The agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent; The agent is created according to the agent template corresponding to the service category; Inputting the interaction data into a large language model in the agent to generate response data, thereby obtaining response data; The response data is displayed through the agent access page.

19. An agent creation device, running on a server, comprising: A creation request receiving module, configured to receive a creation request sent by a user terminal for creating an agent of a service category in an application program; The creation request is submitted based on the user occupation data corresponding to the service category; An agent creation module, configured to create an agent of the user under the service category according to an agent template corresponding to the service category in response to the creation request; After detecting the creation of the intelligent agent, the data return module is run, and the data return module is configured to return intelligent agent access data to the user terminal after detecting the creation of the intelligent agent; the user terminal enters the subroutine for performing intelligent agent access to the intelligent agent according to the intelligent agent access data.

20. An intelligent agent creation device, running on a user terminal, comprising: A creation request sending module is configured to obtain a creation request of a user for creating an intelligent agent of a service category in an application and send it to a server; The creation request is submitted based on the user occupation data corresponding to the service category; An interactive display module, configured to perform interactive display of the agent creation according to the interactive material corresponding to the user category and the identity material generated based on the user occupation data; A subroutine entry module, configured to enter a subroutine for performing agent access to the agent according to agent access data returned by the server after the agent is created; The agent is created according to the agent template corresponding to the service category.

21. An agent-based interactive processing device, running on an agent, comprising: An interaction data acquisition module is configured to acquire interaction data entered by a user through an agent access page; The agent access page is displayed by a subroutine for performing agent access of the agent; The agent is created in response to a creation request submitted based on user occupation data corresponding to the service category of the agent; The agent is created according to the agent template corresponding to the service category; A response data generation module is configured to input the interaction data into a large language model in the agent to generate response data and obtain response data; The response data display module is configured to display the response data through the agent access page.

22. A server, comprising: processor; and a memory configured to store computer executable instructions that, when executed, cause the processor to: Receiving a creation request from a user terminal for creating an agent of a service category in an application program; the creation request is submitted based on user occupation data corresponding to the service category; In response to the creation request, creating an agent of the user under the service category according to an agent template corresponding to the service category; After detecting the creation of the agent, returning agent access data to the user terminal; The user terminal enters a subroutine for performing an agent access to the agent according to the agent access data.

23. A user terminal, comprising: processor; and a memory configured to store computer executable instructions that, when executed, cause the processor to: Obtaining a creation request from a user for creating an agent of a service category in an application and sending the request to a server; the creation request is submitted based on user occupation data corresponding to the service category; Performing interactive display of the agent creation according to the interactive material corresponding to the user category and the identity material generated based on the user occupation data; Entering a subroutine for performing agent access to the agent according to agent access data returned by the server after the agent is created; The agent is created according to the agent template corresponding to the service category.

24. An agent-based interactive processing device, comprising: processor; and a memory configured to store computer executable instructions that, when executed, cause the processor to: Acquiring interactive data entered by a user through an agent access page; the agent access page is displayed by a subroutine for performing agent access of the agent; the agent is created in response to a creation request submitted based on user occupation data corresponding to a service category of the agent; The agent is created according to the agent template corresponding to the service category; Inputting the interaction data into a large language model in the agent to generate response data, thereby obtaining response data; The response data is displayed through the agent access page.

25. A computer-readable storage medium for storing computer-executable instructions, which, when executed, implement the steps of the method of claim 1, claim 15 or claim 18.

Images