Intelligent agent generation method and device and storage medium

Abstract

The invention discloses an agent generation method and device and a storage medium, and relates to the technical field of artificial intelligence, the agent generation method comprises the steps that an agent editing interface is displayed, and the agent editing interface comprises a basic setting area; in response to a configuration operation for the basic setting area, setting information of the target agent is determined; and according to the setting information, generating a target agent in the agent editing interface. According to the method, the user is allowed to directly perform visual configuration on the intelligent agent through the intelligent agent editing interface, the operation is convenient, and the intelligent agent generation efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to an intelligent agent generation method, device and storage medium. Background Art

[0002] With the rapid development of artificial intelligence technology, intelligent agents have gradually evolved from simple rule engines to advanced systems that can autonomously learn and adapt to complex environments, and have been widely used in various fields. Intelligent agent technology integrates a variety of advanced technologies such as machine learning, deep learning, and natural language processing, enabling intelligent agents to simulate human thinking and behavior and complete complex tasks.

[0003] At present, by training large models with massive data, they can learn rich knowledge and patterns, thereby generating high-quality intelligent applications. However, due to the complexity and nonlinear characteristics of large models, their decision-making process is often difficult to fully understand and explain, which increases the difficulty of model optimization and adjustment. Moreover, the process of generating intelligent applications based on large models is usually automated and lacks intuitiveness. This opacity hinders users from accurately anticipating and controlling the behavior of intelligent applications, resulting in the generated intelligent applications not necessarily meeting user needs, and reducing the efficiency of intelligent application generation.

[0004] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of this application is to provide an intelligent agent generation method, device and storage medium, aiming to solve the technical problem of how to efficiently generate intelligent agents.

[0006] To achieve the above objectives, the present application proposes a method for generating an intelligent agent, the method comprising: Displaying an agent editing interface, the agent editing interface comprising: a basic setting area; In response to the configuration operation on the basic setting area, determining setting information of the target agent; According to the setting information, the target agent is generated in the agent editing interface.

[0007] In one embodiment, the step of determining the setting information of the target agent in response to the configuration operation on the basic setting area includes: In response to a configuration operation on the basic setting area, determining functional description information of the target agent; Based on the functional description information, setting information of the target agent is generated.

[0008] In one embodiment, the setting information includes: an agent model, and the step of generating the setting information of the target agent based on the functional description information includes: Determining the functional requirements of the target agent according to the functional description information; According to the functional requirements, the agent model is determined from a preset model library.

[0009] In one embodiment, the step of determining the agent model from a preset model library according to the functional requirements includes: According to the functional requirement, determining at least one candidate model from the model library; Displaying model options corresponding to the candidate model in the basic setting area; In response to a trigger operation for the model option, a target model is determined from the candidate models, and the target model is set as the agent model.

[0010] In one embodiment, the agent editing interface includes: a function expansion area, and the step of generating the target agent in the agent editing interface according to the setting information includes: In response to a configuration operation on the function extension area, determining extension information of the target agent; The target agent is generated in the agent editing interface according to the setting information and the extended information.

[0011] In one embodiment, the function extension area includes a knowledge base option, and the extension information includes: an available knowledge base, and before the step of generating the target agent in the agent editing interface according to the setting information and the extension information, the step further includes: In response to a triggering operation on the knowledge base option, the available knowledge base is determined.

[0012] In one embodiment, the agent editing interface includes: an agent preview area, and after the step of generating the target agent, further includes: In response to a configuration operation on the agent preview area, determining input information; Generate answer information corresponding to the input information through the target agent, and display the answer information in the agent preview area; Determining the question-answer correlation between the input information and the answer information through an evaluation module for the target intelligent agent; In the case where the question-answer correlation does not exceed a preset correlation threshold, the setting information and / or the extended information is adjusted, and the step of generating the target agent in the agent editing interface according to the setting information and the extended information is executed.

[0013] In one embodiment, the agent preview area includes: an information source related to the generation of the answer information, and the step of adjusting the setting information and / or the extended information includes: In response to a trigger operation on the information source, displaying a configuration area corresponding to the information source; In response to a configuration operation on the configuration area, the setting information and / or the extended information is adjusted.

[0014] In addition, to achieve the above-mentioned purpose, the present application also proposes an intelligent agent generation device, the intelligent agent generation device comprising: A display module is used to display an agent editing interface, wherein the agent editing interface includes: a basic setting area; A determination module, configured to determine setting information of a target agent in response to a configuration operation on the basic setting area; A generation module is used to generate the target intelligent agent in the intelligent agent editing interface according to the setting information.

[0015] In addition, to achieve the above-mentioned purpose, the present application also proposes an intelligent agent generation device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the intelligent agent generation method described above.

[0016] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the intelligent agent generation method described above are implemented.

[0017] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the agent generation method described above.

[0018] One or more technical solutions proposed in the present application have at least the following technical effects: by displaying the intelligent agent editing interface, the intelligent agent editing interface includes: a basic setting area; and then in response to the configuration operation on the basic setting area, determining the setting information of the target intelligent agent; further based on the setting information, generating the target intelligent agent in the intelligent agent editing interface, thereby realizing the quick configuration and generation of the target intelligent agent; the user can directly perform visual editing on the various function settings of the intelligent agent through the intelligent agent editing interface, and generate a target intelligent agent that meets their actual needs, which is convenient to operate and improves the efficiency of intelligent agent generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 A flowchart of the first embodiment of the method for generating an intelligent agent of the present application is provided; Figure 2 Schematic diagram of the agent editing interface provided in Example 2 of this application Figure 1 ; Figure 3 Schematic diagram of the agent editing interface provided in Example 4 of this application Figure 2 ; Figure 4 This is a schematic diagram of the module structure of the intelligent agent generation device according to an embodiment of the present application; Figure 5 Schematic diagram of the device structure of the hardware operating environment involved in the agent generation method in the embodiment of the present application. DETAILED DESCRIPTION

[0022] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0023] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0024] In this embodiment, for the convenience of description, the following description is made with the terminal as the execution subject.

[0025] In the process of generating intelligent agent applications, from data input, model training to the final output of intelligent agent applications, there are often multiple complex conversion steps and internal calculations, but the entire process is usually opaque, making it difficult for external users or developers to intuitively understand the generation logic and debugging details of intelligent agent applications. In some application scenarios that require high transparency and explainability, such as financial risk control and medical diagnosis, the application of ordinary large models will be limited. Moreover, debugging tools and methods can often only be debugged for specific problems or specific models. It is difficult for ordinary users to make subtle adjustments to the model settings, etc., resulting in multiple adjustments required to generate intelligent agent applications that meet user needs, thereby reducing the generation efficiency of intelligent agent applications.

[0026] The present application provides a solution by displaying an intelligent agent editing interface, which includes: a basic setting area; and then determining the setting information of the target intelligent agent in response to the configuration operation on the basic setting area; further generating the target intelligent agent in the intelligent agent editing interface based on the setting information, thereby realizing the quick configuration and generation of the target intelligent agent; the user can directly perform visual editing on the various function settings of the intelligent agent through the intelligent agent editing interface, and generate a target intelligent agent that meets their actual needs, which is convenient to operate and improves the efficiency of intelligent agent generation. Furthermore, after the target intelligent agent is generated, the intelligent agent editing interface includes: an intelligent agent preview area, and the intelligent agent preview area also includes: a source area for displaying the information source related to the generation of the answer information. When the question-answer correlation between the input information received by the target intelligent agent and the generated answer information does not exceed the preset qualified threshold, in response to the trigger operation on the information source, the configuration area corresponding to the information source is displayed; and then in response to the configuration operation on the configuration area, the setting information is adjusted. By visualizing the information source, the corresponding configuration area is displayed after the trigger operation on the information source is detected, so that the user can adjust it conveniently, thereby generating a target intelligent agent that meets their actual needs. The visualized configuration makes the adjustment very convenient, which improves the quality of the intelligent agent's answers while ensuring the generation efficiency of the intelligent agent.

[0027] It should be noted that the execution subject of this embodiment may be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, etc. The following takes the terminal as the execution subject as an example to illustrate this embodiment and the following embodiments.

[0028] Based on this, the present application embodiment provides a method for generating an intelligent agent, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the agent generation method of the present application.

[0029] In this embodiment, the agent generation method includes steps S10 to S30: Step S10, displaying the agent editing interface, the agent editing interface includes: a basic setting area; In a feasible embodiment, in order to facilitate users to generate and / or edit intelligent agents, an intelligent agent editing interface is displayed to users; and the intelligent agent editing interface includes a basic setting area to provide users with an input area for relevant setting information of the intelligent agent.

[0030] Exemplarily, the basic setting area may include a name input area, where the user can directly edit or modify the name of the target agent; then, after detecting the configuration operation for the name input area, the configuration operation is determined as the agent name, completing the visual configuration of the agent name. There is no need to find the corresponding name field in the script file and then modify it, and the operation is intuitive and clear.

[0031] Exemplarily, the basic setting area may also include an icon upload area, and the user can determine the local image as the interactive icon of the target intelligent agent by triggering the icon upload area. For example, on a desktop operating system, the user can start the interaction with the target intelligent agent by clicking the icon. At the same time, during the interaction process, the intelligent agent icon can be added before the dialogue information to facilitate the user to distinguish between the input information and the answer information provided by the intelligent agent, thereby ensuring that the generated target intelligent agent can provide the user with a good visual experience.

[0032] Exemplarily, the basic setting area may also include a description input area, where the user can input other relevant information of the target agent, such as the capabilities expected of the target agent, etc. The basic setting area may also include a role instruction input area, where the user can edit and determine the role of the target agent, including the main tasks or goals expected of the target agent. Through the description input area and the role instruction input area, the user can complete the configuration of the relevant settings of the target agent in a visual way, thereby improving the efficiency of agent generation.

[0033] Exemplarily, the basic setting area may include a model selection area for users to select a preferred large model and determine it as the agent model for the target agent to use for behavior decision-making, which provides the target agent with the implementation logic of the main functions. Users can directly select in the model selection area according to the functions they expect the target agent to achieve, and complete the configuration of the agent model.

[0034] Exemplarily, the agent editing interface may include a script editing jump control. After detecting a trigger operation on the control, the content of the current agent editing interface is saved and filled into the script editing interface; the user can modify the running code of the agent in the script editing interface, thereby further refining the settings of the agent.

[0035] Step S20, in response to the configuration operation on the basic setting area, determining the setting information of the target agent; It should be noted that an intelligent agent refers to a software entity that has the ability to make autonomous decisions, perceive the environment, and perform tasks. It can interact and operate in a specific environment according to preset rules or algorithms.

[0036] In a feasible embodiment, in order to configure the intelligent agent, the user can set basic function descriptions, operating logic and other information for the target intelligent agent through configuration operations on the basic setting area after displaying the intelligent agent editing interface; when the configuration operation on the basic setting area is detected, the setting information of the target intelligent agent is determined according to the configuration operation, thereby realizing the visual configuration of the intelligent agent settings and improving the efficiency of intelligent agent generation.

[0037] Exemplarily, configuration operations may include single-clicking, dragging and dropping, dialogue input, etc. Users can directly type in the text input box to complete the entry of setting information without having to compile a script file; users can also directly upload pictures by clicking on an icon without having to enter the storage path of the picture, thereby improving the convenience of intelligent agent generation.

[0038] Exemplarily, the basic setting area may include a first generation control. After detecting a trigger operation for the first generation control, the setting information of the target intelligent agent can be automatically configured according to preset rules or algorithms (such as the user's historical configuration habits), thereby improving the development efficiency of the intelligent agent.

[0039] Step S30, generating a target agent in the agent editing interface according to the setting information.

[0040] In a feasible embodiment, after it is detected that the user has completed the configuration in the basic setting area, the setting information configured by the user is parsed, the parameters and rules therein are extracted, and a complete instance of the target intelligent agent is generated based on these parameters and rules, and displayed in the intelligent agent editing interface for further editing and testing by the user.

[0041] Exemplarily, the acquired setting information is filled into the corresponding field of the background agent generation code, that is, the setting information is coupled with the agent's running code to ensure that the target agent can operate normally according to user requirements, including initializing agent parameters, defining the agent's interface functions, setting data flows, etc.

[0042] The present embodiment provides a method for generating an intelligent agent, by displaying an intelligent agent editing interface including a basic setting area, and generating setting information according to configuration operations on the basic setting area. The visual configuration reduces the workload of manual coding; then, a target intelligent agent is generated according to the setting information, thereby realizing the automatic application of the setting information and the automatic generation of the intelligent agent, thereby realizing the efficient generation of the intelligent agent.

[0043] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the first embodiment can be referred to the above description, and will not be repeated in detail. On this basis, the agent editing interface includes: a function expansion area, step S30 includes: Step S31, in response to the configuration operation for the function extension area, determining the extension information of the target agent; In a feasible embodiment, after displaying the function expansion area of ​​the intelligent agent editing interface, the user can further expand the configuration of the intelligent agent based on the settings of the function expansion area to enhance the user experience of the intelligent agent, including but not limited to configuring the knowledge base, opening remarks and recommended questions and other information.

[0044] Exemplarily, the function expansion area may include an opening remarks input area, and the user can intuitively input relevant information such as the capabilities and identity of the intelligent agent through configuration operations on the opening remarks input area to complete the configuration of the opening remarks of the target intelligent agent; it may also include a recommended question input area, and the recommended questions are used to provide users with common question recommendations before the conversation begins, guiding users to conduct effective interaction and information acquisition. Users can directly enter questions that the current intelligent agent is good at answering in the recommended question input area to determine them as recommended questions for the target intelligent agent, thereby realizing visual configuration of recommended questions with simple and convenient operation.

[0045] Exemplarily, the function extension area may also include a second generation control. When a trigger operation for the generation control of the function extension area is detected, the extended information of the target agent is generated according to the setting information. For example, for the target agent of automobile sales, when its application name is clearly "auto sales" and the application description is "providing automobile information and car purchase plans for users", the knowledge base containing information such as automobile brands, models, performance parameters, prices, and market trends is determined as an available knowledge base; in addition, the extended information may also be generated according to the user's historical configuration data. By automatically generating the extended information of the target agent, the user's configuration time can be saved, thereby achieving efficient generation of agents.

[0046] Step S32, generating a target agent in the agent editing interface according to the setting information and the extended information.

[0047] For example, please refer to Figure 2, displays the intelligent agent editing interface, including a basic setting area 210, a function extension area 220 and a submission control 230, wherein the basic setting area 210 is used to provide the necessary information input for the generation of the intelligent agent, the function extension area 220 is used to implement the function extension configuration of the intelligent agent, and the submission control 230 is used to submit the generated configuration information (setting information and extension information) to the background to generate the target intelligent agent expected by the user. The basic setting area 210 includes a name input area 211, a model selection control 212, a description input area 213, a role instruction input area 214 and an icon upload area 215. The user can perform configuration operations on the basic setting area 210, such as inputting the name of the target intelligent agent, a description of its basic information and the functions it can achieve through the name input area 211, the description input area 213 and the role instruction input area 214 respectively, determine the icon of the target intelligent agent through the icon upload area 215, and determine the intelligent agent model used for behavioral decision-making of the target intelligent agent through the trigger operation of the model selection control 212, thereby realizing the visualization configuration and basic logical behavior configuration of the target intelligent agent; the basic setting area 210 also includes a first generation control 216 and a first save control 217. The user can realize the automatic generation of setting information through the first generation control 216, and convert the user's configuration operations on the basic setting area 210 into setting information and save it through the first save control 217. The function extension area 220 includes a knowledge base selection area 221, a knowledge base upload control 222, an opening remark input area 223 and a recommendation question input area 224. The user can configure the extended functions of the target intelligent agent by configuring the function extension area 220. These extended functions can enhance the user's interactive experience or enhance the performance of the target intelligent agent. The user can determine the available knowledge base of the target intelligent agent by triggering the knowledge base selection area 221, and provide decision support information for the behavior decision of the target intelligent agent. The user can also provide local related resource data to the target intelligent agent by configuring the knowledge base upload control 222. In addition, the user can configure the extended information of the target intelligent agent through the opening remark input area 223 and the recommendation position input area 224. After detecting the trigger operation of the second generation control 225 of the function area, the configuration of the extended information in the function extension area can be automatically completed according to the above-determined setting information, and the configuration operation of the function extension area can be converted into extended information through the second save control 226 of the function extension area, and saved. Thus, the user can complete the configuration of the intelligent agent through a visual interface, which is convenient to operate, thereby improving the efficiency of intelligent agent generation.

[0048] For example, the basic setting area and function extension area of ​​the agent editing interface can be as follows: Figure 2The display is displayed on the same page, and after detecting the trigger operation of the first save control for the basic setting area, the function expansion area can be jumped to display in the intelligent body editing interface. This embodiment does not make specific limitations on this.

[0049] In this embodiment, by responding to the configuration operation for the functional extension area, the extended information of the target intelligent agent is determined, and the visual configuration of the extended information is realized. The user can directly edit and modify the relevant information in the visual interface to complete the configuration of the extended function of the target intelligent agent; then, according to the set information and extended information, the target intelligent agent is generated in the intelligent agent editing interface. The visual operation is simple and convenient, which improves the efficiency of intelligent agent generation.

[0050] In a feasible implementation manner, the function extension area includes a knowledge base option, and the extension information includes: available knowledge bases, and before step S32, further includes: Step S310 , in response to a trigger operation on a knowledge base option, determining an available knowledge base.

[0051] In a feasible embodiment, a knowledge base option for setting the available knowledge base in the function expansion area is provided in the agent editing interface. After detecting a trigger operation for the knowledge base option, the preset knowledge base determined by the trigger operation is set as the available knowledge base of the target agent, so that the target agent can understand user questions and generate answer information by referring to the content in the available knowledge base, thereby providing users with accurate information and suggestions, wherein the available knowledge base is used to provide decision support information for the behavioral decisions of the target agent.

[0052] Exemplarily, the knowledge base option may be provided to the user through a pop-up window or a drop-down menu; triggering operations for the knowledge base option include single-clicking, double-clicking, dragging and dropping, and the like.

[0053] Exemplarily, the knowledge base options include knowledge storage systems that can be accessed and used by the target agent, such as documents, reports, emails, public databases, and professional databases, etc., from which one or more can be selected as available knowledge bases for the target agent.

[0054] Exemplarily, the function extension area also includes a knowledge base upload control, by which the user can trigger the control to set the local data resources as the available knowledge base of the target intelligent agent, and facilitate the user to intuitively select the corresponding local resources through visual operations.

[0055] In this embodiment, by setting the knowledge base option, the user can configure the available knowledge base through visual operations, so that the intelligent agent can answer questions based on the content of the knowledge base. The configuration is convenient and fast, which improves the generation efficiency of the intelligent agent.

[0056] Based on the first embodiment and / or the second embodiment of the present application, in the third embodiment of the present application, the same or similar contents as those of the above-mentioned first and second embodiments can be referred to the above introduction, and will not be repeated in the following. On this basis, step S20 includes: Step A21, in response to the configuration operation on the basic setting area, determining the functional description information of the target agent; Step A22, generating setting information of the target agent based on the functional description information.

[0057] In a feasible embodiment, a user can provide functional description information of the target intelligent agent through the basic setting area of ​​the intelligent agent editing interface; then, after detecting the trigger operation of the first generation control of the basic setting area, the user's configuration operation is converted into the functional description information of the target intelligent agent, and the functional description information is parsed through natural language processing technology to generate specific setting information of the target intelligent agent.

[0058] Exemplarily, the user configures the agent name and role instructions in the basic setting area of ​​the agent editing interface, and completes the configuration of other contents of the basic setting area by clicking the first generation control in the basic setting area; then, after detecting the configuration operation for the basic setting area, the determined functional description information includes the agent name and role instructions, and after detecting the trigger operation for the first generation control in the basic setting area, an answer including the agent description, agent icon and agent model of the target agent is generated according to the agent name and role instructions, and the answer is filled back into the corresponding module in the basic setting area to obtain the setting information of the target agent. Subsequent users can also modify it manually.

[0059] In this embodiment, the user can conveniently configure the functions of the intelligent agent through the basic setting area, and generate the setting information of the intelligent agent based on simple functional description information. The configuration of the setting information of the intelligent agent can be completed without writing complex codes or instructions, thereby improving the efficiency of intelligent agent generation.

[0060] In a feasible implementation manner, the setting information includes: an agent model, and step A22 includes: Step A23, determining the functional requirements of the target agent according to the functional description information; Exemplarily, after a configuration operation for a basic setting area is detected, functional description information of a target intelligent agent is determined according to the configuration operation; and then functional requirements of the target intelligent agent are determined by extracting information from the functional description information.

[0061] Exemplarily, data preprocessing is performed on the function description information, including data cleaning, format conversion, feature extraction, etc.; based on the data preprocessing, the function description information is segmented, part-of-speech tagging, syntactic analysis, etc. are performed through natural language processing technology to extract key information; the key information is classified according to different functions and prioritized to obtain the functional requirements of the final target intelligent body.

[0062] Step A24, determining the intelligent agent model from a preset model library according to functional requirements.

[0063] It should be noted that the preset model library includes at least one model instance that can be used for behavioral decision-making of intelligent agents, such as ChatGPT (Chat Generative Pre-trained Transformer) or different versions of CodeLlama (Open Foundation Models for Code).

[0064] Exemplarily, the model instance with the highest matching degree can be determined as the intelligent agent model to be finally adopted based on the matching degree between each model instance in the model library and the functional requirements of the target intelligent agent; for example, in an image processing application scenario, the functional requirement of the target intelligent agent is to recognize cats in images, and the model instances in the model library include model A for face recognition, model B for animal recognition, and model C for license plate recognition. The matching degrees of each model instance with the functional requirement are 45%, 80%, and 20%, respectively. Finally, model B is selected as the intelligent agent model for the behavior decision-making of the target intelligent agent.

[0065] Exemplarily, the functional requirements include performance requirements, and the degree of match between the performance of each model instance and the performance requirements of the target intelligent agent is further considered. The performance of each model instance on the pre-training data set is compared with the expected performance level. If the performance of the model is better than or close to the expected level, the model is given a relatively high performance match.

[0066] In this implementation, clear functional requirements are generated based on the functional description information configured by the user in the basic setting area, and then the intelligent agent model used for behavioral decision-making is determined based on the functional requirements. That is, the user only needs to perform visual configuration on the basic setting area to complete the model selection. The operation is simple and convenient, which improves the efficiency of intelligent agent generation.

[0067] In a feasible implementation manner, step A24 includes: Step A25, determining at least one candidate model from the model library according to the functional requirements; Exemplarily, functional requirements may include one or more of utility requirements, response time requirements, computing resource requirements, and compatibility requirements. The best model instance in the model library for each requirement may be determined as a candidate model; or a preset number of model instances in the model library that perform well for a specific requirement may be determined as candidate models.

[0068] Step A26, displaying model options corresponding to the candidate models in the basic setting area; Step A27, in response to the trigger operation for the model option, determine the target model from the candidate models, and set the target model as the agent model.

[0069] In a feasible embodiment, a model option for setting an agent model is provided in the agent editing page. When the user performs a trigger operation on the model option, at least one candidate model can be provided to the user for selection; then, through the user's selection operation on the candidate model, the target model is determined, and the target model is set as the agent model of the target agent, so that the target agent makes behavioral decisions through the agent model. The candidate options can be displayed in the form of a pop-up window or a drop-down menu, which is not specifically limited in this embodiment.

[0070] For example, please refer to Figure 2 A model option 212 for setting the agent model is provided in the basic setting area 210; when the user triggers the option by clicking on the model option 212, at least one candidate model provided to the user is displayed through a drop-down menu; the user can further select the candidate model by clicking on the candidate model in the drop-down menu, and then determine the target model and configure the agent model for the target agent.

[0071] In this embodiment, the basic setting area of ​​the intelligent agent editing interface includes a model option. The user can trigger the model option to directly determine the target model from at least one provided candidate model and set the target model as the intelligent agent model, thereby realizing the visual configuration of the intelligent agent model, facilitating operation, and improving the generation efficiency of the intelligent agent.

[0072] Based on the first embodiment, the second embodiment and / or the third embodiment of the present application, in the fourth embodiment of the present application, the same or similar contents as those in the first embodiment, the second embodiment and the third embodiment can be referred to the above introduction, and will not be repeated in the following. On this basis, the intelligent agent editing interface includes: an intelligent agent preview area, after step S30, also includes: Step S40, determining input information in response to a configuration operation on the agent preview area; Exemplarily, the agent preview area may include a recommended question option, and the user may determine the input information by clicking the recommended question option. After detecting a trigger operation for the recommended question option, the input information is determined according to the trigger operation.

[0073] Exemplarily, the agent preview area may include a dialogue input box, and after an input operation to the dialogue input box is detected, the user's input information is determined based on the input operation.

[0074] Step S50, generating answer information corresponding to the input information through the target agent, and displaying the answer information in the agent preview area; By previewing the agent's response during the development phase, we can ensure that its behavior meets expectations, making it easier for users to adjust and optimize the agent's related configurations in a timely manner.

[0075] Step S60, determining the question-answer correlation between the input information and the answer information through an evaluation module for the target intelligent agent; It should be noted that the evaluation module uses other models or algorithms to calculate the correlation between input information and answer information, such as the double tower model, cross attention model, etc. The higher the value, the more accurate the answer information provided by the target agent.

[0076] Step S70, when the question-answer correlation does not exceed the preset correlation threshold, adjust the setting information and / or extended information, and execute the step of generating the target intelligent agent in the intelligent agent editing interface according to the setting information and extended information.

[0077] It should be noted that the qualified threshold is a preset numerical standard, which can be set directly by the user or generated and adjusted by the evaluation module. For example, it can be obtained by training the evaluation module with a preset question and answer library, and adjusted according to user operations in the actual process. For example, after the evaluation module determines that it is qualified, it still detects the user's configuration operations on the setting information and / or extended information, and appropriately raises the qualified threshold.

[0078] Exemplarily, when the question-answer correlation does not exceed the preset qualified threshold, the setting information can be automatically further constrained and refined according to the function description information. For example, when the function description information is "accurately answer the vehicle's steering wheel related configuration", the setting information of the target intelligent agent "providing customers with comprehensive configuration information of steering wheels of different models" can be further added with constraints "including but not limited to steering wheel material, size, etc."; the available knowledge base of the target intelligent agent can also be adjusted according to the function description information, such as eliminating data on other configuration information such as vehicle tires in the knowledge base, and adding data resources on steering wheel material, size, etc.; the application model of the intelligent application body can also be adjusted, such as replacing a model with a large number of parameters, so as to accurately understand the input information and the content of the knowledge base; as to how to make the adjustment, this embodiment does not impose specific restrictions on this.

[0079] In this embodiment, the correlation between the input information and the answer information is determined to evaluate the understanding ability and answer quality of the generated intelligent agent, so that the relevant settings of the target intelligent agent can be adjusted in time, thereby ensuring that the published target intelligent agent can accurately understand and respond to user questions.

[0080] In a feasible implementation, the agent preview area includes: information sources related to the generation of answer information, and the step of adjusting the setting information and / or the extended information in step S70 includes: Step S71, in response to a trigger operation on an information source, displaying a configuration area corresponding to the information source; In a feasible embodiment, after displaying the source area, the user can display the configuration area associated with the information source through a trigger operation on the information source, so that the user can make further settings or adjustments; after detecting the trigger operation on the information source, the position of the configuration area of ​​the information source is determined according to the relevant identification information of the information source, and the configuration area corresponding to the information source is automatically jumped to display according to the position; wherein the trigger operation may include buttons, gestures, single-clicks, double-clicks, etc.

[0081] For example, if the basic setting area, function extension area and intelligent body preview area of ​​the intelligent body editing interface are displayed on the same page, area jumping within the page can be achieved through CSS (Cascading Style Sheets) positioning jump, HTML (HyperText Markup Language) anchor jump and other methods.

[0082] Exemplarily, if the basic setting area, function extension area and agent preview area of ​​the agent editing interface are displayed in pages, page jump is achieved by adding a uniform resource locator to the identification information of the information source.

[0083] It can be understood that the visual information source display and automatic jump reduce the user's manual screening of useful information and positioning of configuration areas, making the operation convenient and improving the efficiency of agent generation.

[0084] Step S72: adjusting the setting information and / or the extended information in response to the configuration operation on the configuration area.

[0085] In one feasible embodiment, after displaying the configuration area corresponding to the information source clicked by the user, the user can edit the setting information or extended information in the configuration area according to the content of the answer information; and then, when a configuration operation for the configuration area is detected, the setting information or extended information is further modified according to the configuration operation, thereby generating a target intelligent entity that meets the user's needs.

[0086] Exemplarily, the user can replace the setting information and / or extended information with the optimized content generated by the evaluation module. The evaluation module can evaluate the correlation between the input information and each information source, and generate corresponding optimized content for information sources with low correlation based on the input information, and after detecting the user's trigger operation on the corresponding information source, display the configuration area corresponding to the information source, and display the corresponding optimized content in the configuration area, and then after detecting the trigger operation on the optimized content, replace the setting information or extended information corresponding to the configuration area with the optimized content, thereby realizing the visual configuration of the setting information and / or extended information, facilitating user operation, and improving the efficiency of intelligent agent generation.

[0087] Exemplarily, the user may also adjust the setting information and / or extended information according to the generation process and results of the answer information, such as increasing the available knowledge base when the answer information is incomplete; replacing the available knowledge base with a knowledge base related to the input information when the answer information is completely unrelated to the input information, or replacing the intelligent agent model with a model that performs better in understanding natural language; and replacing the intelligent agent model with a model with a larger number of parameters when the answer information is generated slowly. Generally, the larger the number of parameters, the better the model performance.

[0088] For example, please refer to Figure 3 The agent editing page includes a basic setting area 210, a function extension area 220, a submission control 230 and an agent preview area 240, wherein the basic setting area 210, the function extension area 220 and the submission control 230 are Figure 2The configuration in the agent preview area 240 is the same as that in the example above. The agent preview area 240 is used to display the generated target agent. The agent preview area displays an input information display box 241, an answer information display box 242, a source area 243, and a dialogue input box 244. The input information display box 241 and the answer information display box 242 are used to display the user's question and the answer provided by the target agent to the question, respectively. The question-answer format is convenient for users to understand and read; the dialogue input box 244 is used for users to input information and interact with the target agent; the source area 243 displays an information source 245 related to the generation of the answer information. After detecting a trigger operation on the information source 245, it will automatically jump to display the configuration operation corresponding to the information source, which is convenient for users to configure or modify it, thereby improving the efficiency of agent generation.

[0089] In this embodiment, by responding to the trigger operation and displaying the configuration area, the user can directly click on the information source to automatically locate the configuration area, thereby directly modifying the setting information or extended information corresponding to the configuration area, thereby improving the efficiency of intelligent agent generation.

[0090] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the target agent generation method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0091] The present application also provides an intelligent agent generation device, please refer to Figure 4 , the agent generating device comprises: Display module 10, used to display the agent editing interface, the agent editing interface includes: basic setting area; A determination module 20, for determining setting information of a target agent in response to a configuration operation on a basic setting area; The generation module 30 is used to generate a target agent in the agent editing interface according to the setting information.

[0092] The intelligent agent generation device provided in the embodiment of the present application adopts the intelligent agent generation method in the above embodiment, which can solve the technical problem of how to efficiently generate intelligent agents. Compared with the prior art, the beneficial effects of the intelligent agent generation device provided in the present application are the same as the beneficial effects of the intelligent agent generation method provided in the above embodiment, and the other technical features in the intelligent agent generation device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0093] An embodiment of the present application provides an intelligent agent generation device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the intelligent agent generation method in the above-mentioned embodiment one.

[0094] Reference below Figure 5 , which shows a schematic diagram of the structure of an agent generation device suitable for implementing the embodiment of the present application. The agent generation device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The intelligent agent generation device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0095] like Figure 5As shown, the agent generation device may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the agent generation device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the agent generation device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows an agent generation device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0096] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0097] The intelligent agent generation device provided in the embodiment of the present application adopts the intelligent agent generation method in the above embodiment, which can solve the technical problem of how to efficiently generate intelligent agents. Compared with the prior art, the beneficial effects of the intelligent agent generation device provided in the present application are the same as the beneficial effects of the intelligent agent generation method provided in the above embodiment, and the other technical features in the intelligent agent generation device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0098] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0099] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0100] An embodiment of the present application provides a computer-readable storage medium having computer-readable program instructions (ie, a computer program) stored thereon, wherein the computer-readable program instructions are used to execute the agent generation method in the above-mentioned embodiment.

[0101] The computer-readable storage medium provided in the embodiment of the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0102] The computer-readable storage medium may be included in the intelligent agent generation device; or it may exist independently without being assembled into the intelligent agent generation device.

[0103] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the intelligent agent generation device, the intelligent agent generation device: displays an intelligent agent editing interface, and the intelligent agent editing interface includes: a basic setting area; determines the setting information of the target intelligent agent in response to the configuration operation on the basic setting area; and generates the target intelligent agent in the intelligent agent editing interface according to the setting information.

[0104] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0105] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0106] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0107] The readable storage medium provided in the embodiment of the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned agent generation method, and can solve the technical problem of how to efficiently generate an agent. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in the present application are the same as the beneficial effects of the agent generation method provided in the above-mentioned embodiment, and will not be repeated here.

[0108] An embodiment of the present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned agent generation method when executed by a processor.

[0109] The computer program product provided in the embodiment of the present application can solve the technical problem of how to efficiently generate an intelligent agent. Compared with the prior art, the beneficial effects of the computer program product provided in the present application are the same as the beneficial effects of the intelligent agent generation method provided in the above embodiment, which will not be repeated here.

[0110] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for generating an intelligent agent, characterized in that: The method comprises: Displaying an agent editing interface, the agent editing interface comprising: a basic setting area; In response to the configuration operation on the basic setting area, determining setting information of the target agent; According to the setting information, the target agent is generated in the agent editing interface.

2. The method according to claim 1, characterized in that The step of determining the setting information of the target agent in response to the configuration operation on the basic setting area includes: In response to a configuration operation on the basic setting area, determining functional description information of the target agent; Based on the functional description information, setting information of the target agent is generated.

3. The method according to claim 2, characterized in that The setting information includes: an agent model, and the step of generating the setting information of the target agent based on the functional description information includes: Determining the functional requirements of the target agent according to the functional description information; According to the functional requirements, the agent model is determined from a preset model library.

4. The method according to claim 3, characterized in that The step of determining the agent model from a preset model library according to the functional requirements includes: According to the functional requirement, determining at least one candidate model from the model library; Displaying model options corresponding to the candidate model in the basic setting area; In response to a trigger operation for the model option, a target model is determined from the candidate models, and the target model is set as the agent model.

5. The method according to claim 1, characterized in that The agent editing interface includes: a function extension area, and the step of generating the target agent in the agent editing interface according to the setting information includes: In response to a configuration operation on the function extension area, determining extension information of the target agent; The target agent is generated in the agent editing interface according to the setting information and the extended information.

6. The method according to claim 5, characterized in that The function extension area includes a knowledge base option, and the extension information includes: an available knowledge base, and before the step of generating the target agent in the agent editing interface according to the setting information and the extension information, the step further includes: In response to a triggering operation on the knowledge base option, the available knowledge base is determined.

7. The method according to claim 5, characterized in that The agent editing interface includes: an agent preview area, and after the step of generating the target agent, further includes: In response to a configuration operation on the agent preview area, determining input information; Generate answer information corresponding to the input information through the target agent, and display the answer information in the agent preview area; Determining the question-answer correlation between the input information and the answer information through an evaluation module for the target intelligent agent; In the case where the question-answer correlation does not exceed a preset correlation threshold, the setting information and / or the extended information is adjusted, and the step of generating the target agent in the agent editing interface according to the setting information and the extended information is executed.

8. The method according to claim 7, characterized in that The agent preview area includes: information sources related to the generation of the answer information, and the step of adjusting the setting information and / or the extended information includes: In response to a trigger operation on the information source, displaying a configuration area corresponding to the information source; In response to a configuration operation on the configuration area, the setting information and / or the extended information is adjusted.

9. An intelligent agent generation device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the agent generation method according to any one of claims 1 to 8.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the agent generation method according to any one of claims 1 to 8 are implemented.

Images