Intelligent tool configuration information processing method, device, system and storage medium

Abstract

The present disclosure relates to a method, device, system, and storage medium for processing tool configuration information for an intelligent agent, and relates to the field of data processing technology, in order to at least solve the technical problem in the related art that the entire service program of the intelligent agent is shut down due to a change in the tool for a task, resulting in the inability to execute other tasks of the intelligent agent, thereby causing low efficiency in the execution of the intelligent agent's tasks and a poor user experience. The method comprises: when the intelligent agent perceives a task request from a first user, selecting a target tool and target configuration information of the target tool from a preset disk file that provides a service for the target task of the task request; the configuration information included in the preset disk file that represents the association mapping relationship between each task and each tool is modifiable; the target tool and target configuration information are loaded into a target service program that executes the target task, so as to call the loaded target service program to execute the target task.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to a method, device, system and storage medium for processing intelligent tool configuration information. Background Art

[0002] To meet the growing demand for intelligent services from financial users, intelligent agents based on large language models are becoming a trend. Currently, after the agent's service program is launched, it first identifies the user's input question intent and extracts the corresponding parameters to determine the tasks to be performed, following the agent's technical framework. It then calls the corresponding tools to execute each task and obtain the desired results. However, the tools used to support these tasks are typically pre-associated and fixed within the framework. That is, after the service program is launched, the multiple tools used to support each task are typically fixed. If any of the tools associated with each task change, the entire service program must be stopped, the changed tool modified, and then restarted to re-execute the tasks using the changed tool. However, shutting down the entire agent service program due to a tool change for one task prevents the agent from executing other tasks, resulting in low task execution efficiency and a poor user experience. Summary of the Invention

[0003] The present invention provides a method, device, system, and storage medium for processing agent tool configuration information to at least address the technical problem in related art where a change in the tool for a task causes the entire agent service program to be shut down, resulting in the inability to execute other tasks, thus causing low agent task execution efficiency and a poor user experience. The technical solution of the present invention is as follows:

[0004] According to the first aspect of an embodiment of the present invention, a method for processing intelligent agent tool configuration information is provided, which includes: when the intelligent agent perceives a task request from a first user, selecting a target tool that provides a service for a target task of the task request and target configuration information of the target tool from a preset disk file; the configuration information included in the preset disk file that characterizes the association mapping relationship between each task and each tool is modifiable; the target tool and the target configuration information are loaded into a target service program that executes the target task, so as to call the target service program after loading to execute the target task.

[0005] In one implementation, the target tool and target configuration information are loaded into the target service program of the target task, including: when it is detected that the target tool and the target configuration information of the target tool in the preset disk file are both in a configuration completed state, the target configuration information in the configuration completed state is loaded into the target service program.

[0006] In another implementation, the target tool and target configuration information are loaded into the target service program of the target task, including: when it is detected that the first target configuration information of at least one first target tool in the target tool is in a configuration operation state, the second target configuration information of at least one second target tool other than the first target tool in the multiple target tools is loaded into the target service program; when the first target configuration information of at least one first target tool is in a configuration completion state, the first target configuration information after the configuration completion state is loaded into the target service program.

[0007] In another implementation, after loading the target tool and target configuration information into the target service program that executes the target task, the method also includes: when it is detected that the target tool and target configuration information are reconfigured, re-initiating the task request when the reconfiguration of the target tool and target configuration information is completed, so as to re-execute the target task based on the target tool and target configuration information after the reconfiguration is completed.

[0008] In another implementation, a task corresponds to at least one preset disk file, a preset disk file corresponds one-to-one to a tool, and a task corresponds to at least one tool; a target tool and target configuration information of the target tool that provide services for the target task requested by the task are selected from the preset disk file, including: determining a target task that matches the task request; selecting a target disk file corresponding to the target task from multiple preset disk files according to the associated correspondence between the task and the preset disk file; determining the target tool based on an identifier representing the tool type in the target disk file; traversing the file information of the target disk file, and determining the file information including the preset identifier representing the tool configuration information as the target configuration information.

[0009] In another implementation, the method further includes: in response to a modification request from a second user to reconfigure the target tool of the target task, displaying a target configuration information page including the target tool and the target configuration information of the target tool; the target configuration information page includes a target control also used to reconfigure the target tool and the target configuration information; when it is detected that the target control indicates that the reconfiguration of the target tool and / or the target configuration information is completed, sending indication information indicating that the reconfiguration of the target tool and / or the target configuration information is completed.

[0010] In another implementation, the target tool and target configuration information are reconfigured, including one or more of the following: adding tools other than the target tool, deleting at least one tool from the target tool, modifying the target configuration information of the target tool, adding configuration information of tools other than the target tool, and deleting configuration information of at least one tool from the target tool.

[0011] According to the second aspect of an embodiment of the present invention, an intelligent agent tool configuration information processing device is provided, including: a perception unit, configured to select a target tool and target configuration information of the target tool that provide services for the target task of the task request from a preset disk file when the intelligent agent perceives the task request of the first user; the configuration information included in the preset disk file that represents the association mapping relationship between each task and each tool is modifiable; a loading unit, configured to load the target tool and target configuration information into a target service program that executes the target task, so as to call the loaded target service program to execute the target task.

[0012] In one implementation, the loading unit is specifically configured to: when it is detected that the target tool and the target configuration information of the target tool in the preset disk file are both in a configuration completed state, load the target configuration information in the configuration completed state into the target service program.

[0013] In another implementation, the loading unit is specifically configured to: when it is detected that the first target configuration information of at least one first target tool in the target tool is in a configuration operation state, load the second target configuration information of at least one second target tool other than the first target tool in the multiple target tools into the target service program; when the first target configuration information of at least one first target tool is in a configuration completion state, load the first target configuration information after the configuration completion state into the target service program.

[0014] In another implementation, the device is further configured to: upon detecting that the target tool and target configuration information have been reconfigured, re-initiate a task request when the reconfiguration of the target tool and target configuration information is completed, so as to re-execute the target task based on the target tool and target configuration information after the reconfiguration is completed.

[0015] In another implementation, a task corresponds to at least one preset disk file, a preset disk file corresponds one-to-one to a tool, and a task corresponds to at least one tool; the perception unit is specifically configured to: determine a target task that matches the task request; select a target disk file corresponding to the target task from multiple preset disk files according to the associated correspondence between the task and the preset disk file; determine the target tool based on the identifier representing the tool type in the target disk file; traverse the file information of the target disk file, and determine the file information including the preset identifier representing the tool configuration information as the target configuration information.

[0016] In another implementation, the device is further configured to: in response to a modification request from a second user to reconfigure the target tool of a target task, display a target configuration information page including the target tool and the target configuration information of the target tool; the target configuration information page includes a target control for reconfiguring the target tool and the target configuration information; and when it is detected that the target control indicates that the reconfiguration of the target tool and / or the target configuration information is completed, send indication information indicating that the reconfiguration of the target tool and / or the target configuration information is completed.

[0017] In another implementation, the target tool and target configuration information are reconfigured, including one or more of the following: adding tools other than the target tool, deleting at least one tool from the target tool, modifying the target configuration information of the target tool, adding configuration information of tools other than the target tool, and deleting configuration information of at least one tool from the target tool.

[0018] According to the third aspect of an embodiment of the present invention, an information processing system is provided, comprising an intelligent agent, the intelligent agent comprising a processor and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the intelligent agent tool configuration information processing method of the above-mentioned first aspect and any possible implementation method thereof.

[0019] According to the fourth aspect of an embodiment of the present invention, an electronic device is provided, comprising: a processor and a memory for storing processor executable instructions; wherein the processor is configured to execute executable instructions to implement the intelligent body tool configuration information processing method as described in the first aspect and any possible implementation method thereof.

[0020] According to the fifth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which instructions are stored. When the instructions in the computer-readable storage medium are executed by the processor of the intelligent body, the electronic device is enabled to execute the intelligent body tool configuration information processing method as in the first aspect.

[0021] According to the sixth aspect of the embodiment of the present disclosure, a computer program product is provided, which includes computer instructions. When the computer instructions are run on an electronic device, the electronic device executes the intelligent body tool configuration information processing method of the above-mentioned first aspect and any possible implementation method thereof.

[0022] The technical solution provided by the embodiment of the present invention brings at least the following beneficial effects: the tools associated with each task and the configuration information corresponding to the tools are pre-set in the preset disk file of the preset disk. For any task request, based on the preset disk file, the target tool and the target configuration information of the target tool required for the task request can be obtained in real time and dynamically. Based on this, the loading of the tools corresponding to each task is not a one-time static loading, but is dynamically loaded according to the task request. The tool loading process of each task is independent of each other and does not affect each other, so that when the tool of any task changes, only the configuration information of the changed tool needs to be reloaded to re-execute the corresponding task, and the execution of other tasks whose corresponding tools have not changed is not affected, thereby improving the task execution efficiency and task processing speed of the intelligent agent, thereby improving the user experience.

[0023] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute an improper limitation of the present disclosure.

[0025] Figure 1 is a schematic diagram of an information processing system according to an exemplary embodiment;

[0026] Figure 2 is a schematic diagram showing a process of processing configuration information of an intelligent agent tool according to an exemplary embodiment;

[0027] Figure 3 is a flow chart of a method for processing agent tool configuration information according to an exemplary embodiment;

[0028] Figure 4 is a block diagram of an intelligent tool configuration information processing device according to an exemplary embodiment;

[0029] Figure 5 is a schematic diagram of an intelligent agent according to an exemplary embodiment. DETAILED DESCRIPTION

[0030] In order to enable ordinary people in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0031] It should be noted that the terms "first," "second," and the like in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the numbers used in this manner are interchangeable where appropriate so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure as detailed in the appended claims.

[0032] Before giving a detailed introduction to the intelligent agent tool configuration information processing method provided in the embodiment of the present application, a brief introduction to the application scenarios and implementation architecture involved in the embodiment of the present application is first given.

[0033] First, a brief introduction to the application scenarios involved in this application is given.

[0034] To meet the growing demand for intelligent services from financial users, intelligent agents based on large language models are becoming a trend. Currently, after the agent's service program is launched, it first identifies the user's input question intent and extracts the corresponding parameters to determine the tasks to be performed, following the agent's technical framework. It then calls the corresponding tools to execute each task and obtain the desired results. However, the tools used to support these tasks are typically pre-associated and fixed within the framework. That is, after the service program is launched, the multiple tools used to support each task are typically fixed. If any of the tools associated with each task change, the entire service program must be stopped, the changed tool modified, and then restarted to re-execute the tasks using the changed tool. However, shutting down the entire agent service program due to a tool change for one task prevents the agent from executing other tasks, resulting in low task execution efficiency and a poor user experience.

[0035] The following describes a more detailed explanation of the aforementioned issues using a large language model-based intelligent agent technology framework developed by financial institutions such as UnionPay Business. This framework identifies the intent of user input questions, extracts the corresponding parameters, and invokes the appropriate tools to execute and obtain the results. For example, if a user enters the question "I want to check the weather in Shanghai," the framework will recognize the user's intent and extract the parameter "city" as "Shanghai." It will then invoke the weather query tool and obtain the weather in Shanghai. Therefore, in addition to the core intent recognition and parameter extraction modules, this intelligent agent also includes multiple tools for performing specific tasks. Currently, all supported tools are hard-coded into the framework. Adding, deleting, or modifying a tool requires stopping the intelligent agent framework service. UnionPay Business currently utilizes this intelligent agent framework to identify user intent and invoke tools to execute tasks across multiple services. Stopping the entire service and its tools to modify a single tool is clearly unreasonable. Therefore, a dynamic configuration and loading method for intelligent agent tools is required to ensure that adding, deleting, or modifying a tool does not affect other services and tools, thus ensuring system and business stability.

[0036] In response to the above problems, the present application provides a method for processing intelligent tool configuration information, in which the tools associated with each task and the configuration information corresponding to the tools are pre-set in a preset disk file on a preset disk. For any task request, based on the preset disk file, the target tool and the target configuration information of the target tool required for the task request can be obtained in real time and dynamically. Based on this, the loading of the tools corresponding to each task is not a one-time static loading, but is dynamically loaded according to the task request. The tool loading process of each task is independent of each other and does not affect each other, so that when the tool of any task changes, only the configuration information of the changed tool needs to be reloaded to re-execute the corresponding task, and the execution of other tasks whose corresponding tools have not changed is not affected, thereby improving the intelligent task execution efficiency and task processing speed, thereby improving the user experience.

[0037] Next, the implementation architecture involved in this application is briefly introduced below.

[0038] Figure 1 Schematic diagram of an information processing system 10 provided by the present disclosure. Figure 1 As shown, the information processing system includes at least one intelligent agent 101 and a user terminal 102. The intelligent agent 101 and the user terminal 102 can be connected via a wired network or a wireless network.

[0039] In some embodiments, agent 101 may be a robot. Different types of agents are used to perform tasks of different service types. The specific service types that agent 101 can perform are determined based on the specific business scenarios in which they are configured. Different business representations represent different service types for the tasks they perform.

[0040] In some embodiments, a user accesses the information processing system 10 on the user terminal 102 and sends a task request. The agent 101 receives the task request, traverses multiple preset disk files in a preset disk, selects a target disk file containing a target tool and target configuration information that provides a service for the target task of the task request, and loads the target tool and target configuration information in the target disk file into a target service program that executes the target task, so that the target service program can execute the target task.

[0041] At the same time, the user can also interact with the agent 101 on the agent 101 or through the user terminal 102 to modify the configuration information and tools in the preset disk file.

[0042] The above target disk file can be understood as a tool configuration file of the target tool.

[0043] In some embodiments, the agent 101 includes a database or is connected to a database, and can store association information of the association relationship between tasks and preset disk files in the database. The user terminal 102 can access the association information in the database through the server.

[0044] like Figure 2 As shown, the agent 101 provides corresponding services based on the agent framework service, that is, performs tasks corresponding to different service types.

[0045] The specific task execution process is as follows: the administrator or developer (i.e., the second user) modifies the tools and tool configuration information in the tool configuration file on the agent tool management page, and stores the modified tools and tool configuration information in the tool configuration file and identifies them based on the corresponding tasks. The financial user, i.e., the first user, issues a task request (such as a task request for a consulting task) through the question-and-answer system. After the agent perceives the task request, it determines the tools and tool configuration information required to execute the task based on the agent framework service. The tools required for the task and the tool configuration information required are loaded into the service program of the agent framework service to complete the execution of the task.

[0046] In the service program of the agent framework service, different tasks correspond to different service programs, and different tasks require different tools. Therefore, the agent framework service includes tool services representing multiple agents using different service programs, such as Tool Service 1 and Tool Service 2. Different tool services correspond to different business systems. For example, Tool Service 1 corresponds to Business System 1, and Tool Service 2 corresponds to Business System 2.

[0047] The following description is given for the Intelligent Body Tool Management page.

[0048] First, on this page, administrators can create and modify tool configuration files by filling in or modifying information such as the tool name, tool purpose, parameters, parameter description, and business implementation service address, allowing them to dynamically add and modify tools. Furthermore, administrators can dynamically delete tools, that is, delete tool configuration files. When administrators add, modify, and delete tools, the dynamically changed tool configuration files are synchronized to the intelligent agent framework service in real time.

[0049] Secondly, on this page, the administrator can configure the available tools according to business needs. For example, if there are all tools ABCD, the administrator can set the available ABC tools for business 1 and the available BCD tools for business 2.

[0050] The following is an explanation of the question-answering system.

[0051] A system that provides users with question input and returns answers. The system inputs user questions into the intelligent agent framework service center for intent recognition and tool call.

[0052] The following is a description of the intelligent agent framework service.

[0053] It is used to accept user questions from the question-answering system, load supported tools, identify user intent based on user questions, select the appropriate tool and extract the parameters required by the tool, and finally call the tool to obtain the results.

[0054] The following describes the tool configuration file.

[0055] Multiple python files, each python file is configured with the tool name, tool purpose, parameters, parameter description, business implementation service address and other information of the corresponding tool, and each python file format is fixed and is dynamically generated by the administrator on the intelligent tool management page.

[0056] The tool services for intelligent agents are described below.

[0057] In each tool configuration file, there is a required item, which is the business implementation service address. The specific business logic implementation corresponding to the service address is provided by each tool service in the intelligent tool service. For example, in the weather query tool, the corresponding tool service needs to implement the following logic process: use the city name A as a parameter, call the weather query interface, query the weather conditions of city A, and return the temperature and humidity information in json format.

[0058] The following is an explanation of the business system.

[0059] In the tool service of the intelligent body, when implementing a specific tool business logic, it may be necessary to connect to the corresponding business system. For example, in the specific implementation logic of the tool service of the intelligent body, it is necessary to connect to the company's attendance system. Then the attendance system is a business system.

[0060] In other embodiments, the agent 101 may be a single agent, or may be an agent cluster composed of multiple agents. In some implementations, the agent cluster may also be a distributed cluster. This application does not limit the specific implementation of the agent 101.

[0061] The above-mentioned user terminals 102 can all be understood as terminal devices. Terminal devices can be mobile phones, tablet computers, desktop computers, laptop computers, handheld computers, notebook computers, ultra-mobile personal computers (UMPCs), netbooks, as well as cellular phones, personal digital assistants (PDAs), augmented reality (AR) and virtual reality (VR) devices, etc., which can install and use content community applications (such as Kuaishou). The present disclosure does not impose any special restrictions on the specific form of the terminal device. It can interact with the user through one or more methods such as keyboard, touchpad, touch screen, remote control, voice interaction or handwriting device.

[0062] Optionally, the above Figure 1 In the agent tool configuration information processing system shown, the agent 101 can be connected to at least one user terminal 102. This application does not impose any restrictions on the number and type of terminal devices.

[0063] The method for processing intelligent tool configuration information provided in the embodiment of the present application can be applied to the aforementioned Figure 1 The information processing system in the implementation architecture shown can also be applied to a terminal device or electronic device such as an intelligent agent. For ease of understanding, the following describes the intelligent agent tool configuration information processing method provided by this application in detail with reference to the accompanying drawings.

[0064] Figure 3 is a flow chart of a method for processing agent tool configuration information according to an exemplary embodiment. Figure 3 As shown, the intelligent tool configuration information processing method includes the following steps.

[0065] S31, when the intelligent agent perceives the task request of the first user, it selects a target tool that provides a service for the target task of the task request and target configuration information of the target tool from a preset disk file.

[0066] The first user mentioned above may be an agent user.

[0067] The configuration information included in the preset disk file and representing the association and mapping relationship between various tasks and various tools is modifiable.

[0068] The above default disk files are stored in the default disk. The default disk and the service program memory are not in the same storage space.

[0069] In some embodiments, in order to improve the efficiency of obtaining configuration information of the tool, the preset disk file is set to correspond one-to-one with the tool. Since a task requires one or more tools to improve the service, a task is set to correspond to at least one tool. In this way, a task corresponds to at least one preset disk file.

[0070] Based on the association mapping relationship between tasks, preset disk files and tools in this embodiment, the selection of target tools and target configuration information from the preset disk files in the above step S31 can be specifically implemented through the following steps.

[0071] First, determine the target task that matches the task request.

[0072] The intelligent agent can automatically parse task requests.

[0073] Secondly, according to the correlation between the tasks and the preset disk files, a target disk file corresponding to the target task is selected from a plurality of preset disk files.

[0074] The various tools used by the task and tool-related information (such as configuration information) are stored in various preset disk files associated with the task.

[0075] Third, the target tool is determined based on the identifier representing the tool type in the target disk file.

[0076] Fourthly, the file information of the target disk file is traversed, and the file information including the preset identifier representing the tool configuration information in the file information of the target disk file is determined as the target configuration information.

[0077] In the above implementation, the mapping of a task to at least one tool is demonstrated by mapping a task to at least one preset disk file, where each preset disk file corresponds one-to-one with a tool. Each preset disk file stores only one tool and its configuration information. Simply identifying the preset disk file allows for more accurate mapping of the task to the tool and configuration information.

[0078] S32, loading the target tool and the target configuration information into the target service program that executes the target task, so as to call the loaded target service program to execute the target task.

[0079] In one embodiment, the target tool and target configuration information are stored in the form of a tool configuration file.

[0080] Optionally, because the target service program of the target task will be updated and changed based on different business needs or updates to the business system, in order to avoid interference and impact of the currently loaded target tools and target configuration information on the execution process of the subsequently updated target service program, the loaded target tools and target configuration information will be released after the target task is completed, thereby reducing the memory space occupied by the service program in the intelligent body.

[0081] In some embodiments, tool configuration information includes basic tool information and business tool configuration information. The basic tool information is stored in an agent tool basic information table corresponding to a preset disk file, and the business tool configuration information is stored in an agent business tool configuration information table corresponding to a preset disk file, thereby creating a tool based on these two tables. The preset disk file is also called a tool configuration file.

[0082] Through the above implementation, the tools associated with each task and the configuration information corresponding to the tools are pre-set in the preset disk file of the preset disk. For any task request, based on the preset disk file, the target tool and the target configuration information of the target tool required for the task request can be obtained in real time and dynamically. Based on this, the loading of the tools corresponding to each task is not a one-time static loading, but is dynamically loaded according to the task request. The tool loading process of each task is independent of each other and does not affect each other, so that when the tool of any task changes, only the configuration information of the changed tool needs to be reloaded to re-execute the corresponding task, and the execution of other tasks whose corresponding tools have not changed will not be affected, thereby improving the task execution efficiency and task processing speed of the intelligent agent, thereby improving the user experience.

[0083] As an implementation method, the target tool and target configuration information are loaded into the target service program, which is specifically implemented through the following steps: when it is detected that the target tool and the target configuration information of the target tool in the preset disk file are in a configuration completed state, the target configuration information in the configuration completed state is loaded into the target service program.

[0084] In this implementation, it is ensured that the loaded target tool and target configuration information are in a configuration complete state.

[0085] As another implementation method, loading the target tool and target configuration information into the target service program can also be specifically implemented through the following steps.

[0086] First, when it is detected that the first target configuration information of at least one first target tool in the target tool is in a configuration operation state, the second target configuration information of at least one second target tool other than the first target tool in the multiple target tools is loaded into the target service program.

[0087] Secondly, when the first target configuration information of at least one first target tool is in a configuration-completed state, the first target configuration information in the configuration-completed state is loaded into the target service program.

[0088] In this embodiment, when the configuration information and tools are loaded

[0089] As another embodiment, when it is detected that the target tool and target configuration information are reconfigured, the task request is re-initiated when the reconfiguration of the target tool and target configuration information is completed, so as to re-execute the target task based on the target tool and target configuration information after the reconfiguration is completed.

[0090] Optionally, during the process of loading the target tool and target configuration information into the target service program, if it is detected that the target tool and target configuration information are reconfigured, the loading of the target tool and target configuration information into the target service program is stopped, and when the reconfiguration of the target tool and target configuration information is completed, the target tool and target configuration information are reloaded into the target service program to ensure that the tools and configuration information based on the task execution process better meet the task requirements, thereby ensuring that the task execution results are more accurate.

[0091] Optionally, after the target tool and target configuration information are loaded into the target service program but the target task has not been completed, if it is detected that the target tool and target configuration information have been reconfigured, the execution of the target task is stopped. When the reconfiguration of the target tool and target configuration information is completed, the target tool and target configuration information are reloaded into the target service program, and the target task is re-executed based on the reconfigured target tool and target configuration information to ensure that the tools and configuration information based on the task execution process better meet the task requirements, thereby ensuring that the task execution results are more accurate.

[0092] As an embodiment, in response to a modification request from a second user to reconfigure a target tool for a target task, a target configuration information page including the target tool and the target configuration information of the target tool is displayed; the target configuration information page includes a target control for reconfiguring the target tool and the target configuration information; when it is detected that the target control indicates that the reconfiguration of the target tool and / or the target configuration information is completed, an indication information indicating that the reconfiguration of the target tool and / or the target configuration information is completed is sent.

[0093] In this embodiment, the configuration of the tool is implemented on the configuration information page, and the editing and modification of the tool and configuration information are achieved through controls.

[0094] The modification of the tool includes the modification of the tool name.

[0095] Optionally, the target tool and target configuration information are reconfigured, including one or more of the following: adding tools other than the target tool, deleting at least one tool from the target tool, modifying the target configuration information of the target tool, adding configuration information of tools other than the target tool, and deleting configuration information of at least one tool from the target tool.

[0096] Taking the intelligent tool management system as an example, the tool creation process is first explained as follows.

[0097] First, the administrator logs in to the intelligent tool management system, and the system obtains the administrator's user name and department. For example, the administrator's user name is guolinye and the department is the business operation center.

[0098] Second, the administrator enters the following information.

[0099] (1) The Chinese name of the product to which the tool belongs, such as Account Manager Q&A Assistant.

[0100] (2) The English name of the product, such as wdzs.

[0101] (3) The English name of the tool, such as get_weather.

[0102] (4) Description of the tool, for example: query the weather based on the city name.

[0103] (5) The English name of the parameter, the type of the parameter, and the description of the parameter, for example: English name of the parameter: city, type of the parameter: str, description of the parameter: the city to be queried. Special note: If not mentioned in the question, city_name must default to an empty string.

[0104] (6) Determine whether the configuration selection tool is valid.

[0105] (7) Whether the configuration tool requires user information to be authenticated in the question. If the tool requires user information to be authenticated, the system will check whether the user's personal information, such as username, work number, position, and name, is involved when the user asks the question. For example, a vacation balance inquiry tool will require user information to be authenticated in the question, because the user usually only inquires about his or her own vacation balance. So if user A enters the question "I want to check B's vacation balance", the system will first extract and verify the personal information. If it finds that A wants to check B, the system will not return the result, but will give a prompt and warning. For example, a weather inquiry tool will not require user information to be authenticated in the question, because the tool does not contain sensitive information.

[0106] (8)Configure the tool service layer address.

[0107] (9) Configuration tool sources, such as Business Operations Center_Customer Manager Q&A Assistant_Weather Query Tool.

[0108] Third, write the newly added tool information into the agent_base_info table.

[0109] Fourth, the service can be updated in real time based on the content of the configuration file. The tool configuration information filled in by the administrator is used as a parameter to call the tool configuration file generation interface to generate the tool configuration file.

[0110] For example, the tool configuration file ywyyzx_wdzs_get_weather.py generated by the weather query tool is as follows:

[0111]

[0112] Among them, ywyyzx_wdzs_get_weather is the full English name of the tool, and the format is "department_product_tool name".

[0113] In the above tool configuration file, except for the user configuration information which is dynamically input, the other formats are fixed, thus ensuring the format uniformity of the tool configuration file.

[0114] Fifth, although the tool configuration files have been generated, the specific business implementation has not yet been done. Only the service address for the business implementation has been configured. Therefore, this step requires developing code according to the interface specifications of each business system to implement the tool logic. For example, the weather query tool needs to obtain city parameters, call the specific weather query interface, and encapsulate the results in JSON format to ensure that the service at the aforementioned service address is accessible.

[0115] In the above implementation, the generation of tool configuration files and specific intelligent tool services are separated, which has the following advantages: on the one hand, it ensures the uniformity of tool configuration files and facilitates the maintenance of tool configuration file templates; on the other hand, general tool configuration is operated by business personnel, and specific intelligent tool services are generally developed by technical personnel, which not only ensures the independence of the work of the two, but also enables the work of the two to be carried out in parallel.

[0116] Furthermore, secondly, the tool modification process is explained as follows.

[0117] When an administrator modifies a tool, they open the edit page for the specific tool. After submitting the modified information, the field content of the corresponding tool in the agent_base_info table of the database will be changed. At the same time, the tool configuration file will be regenerated, just like when creating a tool.

[0118] Furthermore, again, the tool deletion process is described as follows.

[0119] The administrator selects the tool to be deleted and clicks Delete. The system will delete the record of the corresponding tool in the agent_base_info table of the database and also delete the tool configuration file.

[0120] Furthermore, the process of binding services and tools is described as follows, where services can be understood as tasks.

[0121] Although the system is configured with several tools, most of these tools are configured according to specific businesses, except for a few general functions, such as date query tools. In the question-and-answer system, the robots set up are separated according to specific businesses. Therefore, it is necessary to bind certain tools in the system with business robots according to business needs. The detailed process is as follows.

[0122] First, administrator A logs in to the agent tool management system, and the system obtains the administrator's username;

[0123] Second, open a robot editing page, select the tool(s) you want to connect to from the drop-down box, and submit an access application to the creator of each tool.

[0124] Third, administrator B who created the tool reviews the application for access to the tool.

[0125] Fourth, the tools that have passed the review will be written to the agent_tool_config table. For example, if the business operation robot Robot_A is configured with the above-mentioned weather query tool, the key fields to be inserted are: dept: ywyyzx_Robot_A, tool_name: ywyyzx_wdzs_get_weather, creator: A.

[0126] Furthermore, the dynamic loading and usage process of the tool is described below.

[0127] The above tools and configuration information can be used to create, modify, and delete tools at any time according to application requirements. That is, the data in the agent_base_info table and the agent_tool_config table, as well as the corresponding tool configuration files, can be created, modified, and deleted at any time. Therefore, if the tool is to work dynamically, the above table data and configuration files must be read each time the intelligent agent framework is called. The specific process is as follows.

[0128] First, the user selects robot R in the question-answering system and enters a question.

[0129] Second, the system reads and parses all tool configuration files. If a function in the file has the @tool annotation, it indicates that the function is a tool configuration. For example, the @tool annotation appears above the ywyyzx_wdzs_get_weather function created by the aforementioned tool, indicating that the ywyyzx_wdzs_get_weather function is a tool configuration. All functions annotated with @tool are registered as function objects, and a dictionary is constructed with the function name (the tool's full English name) as the primary key. This ensures that the function object can be retrieved and executed based on the function name (the tool's full English name). This implements dynamic tool loading. For example, the stored dictionary might look like this: {"Tool A":"Function A Object", "Tool B":"Function B Object", ...}.

[0130] Third, the system reads the tools configured under the robot R from the agent_tool_config table;

[0131] Fourthly, only the tool contents configured by the robot R in the second dictionary are retained, and the tools that are not configured by the robot are deleted.

[0132] Fifth, for the remaining tools, i.e. the configured tools, the tool name, tool purpose, parameters, parameter description and other information are read according to the tool configuration file, the system prompt words are encapsulated, and the user questions are input into the large model to identify the user intention, select the appropriate tool and extract the parameters required by the tool.

[0133] Sixth, according to the full English name of the selected tool, the function object is taken out from the second dictionary, the extracted function parameters are in json format, the function object is called for execution, and the execution result is obtained.

[0134] For example, if the big model selects the weather query tool, then the parameter {"city_name":"Shanghai"} is passed to the ywyyzx_wdzs_get_weather function object for execution. In fact, the service address of this function is still called internally to implement the service and get the result.

[0135] Seventh, the tool execution results and user questions are input into the big model to summarize the answers, and the summarized answers are returned to the question-answering system to be displayed to the user.

[0136] Through the above implementation method, the tool configuration and the specific intelligent tool implementation service are separated, which ensures the uniformity of the tool configuration file and facilitates the maintenance of the tool configuration file template; at the same time, the general tool configuration is operated by business personnel, and the specific intelligent tool service is generally developed by technical personnel, which not only ensures the independence of the work of the two, but also enables the work of the two to be carried out in parallel.

[0137] In order to realize the above functions, the intelligent tool configuration information processing device includes hardware structures and / or software modules corresponding to the execution of each function. It should be easily appreciated by those skilled in the art that, in combination with the algorithm steps of each example described in the embodiments disclosed herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0138] The present disclosure also provides a Figure 4 An intelligent tool configuration information processing device 40 is shown, which includes: a perception unit 401, a loading unit 402 and a display unit 403.

[0139] The perception unit 401 is configured to select a target tool and target configuration information of the target tool that provide services for the target task of the task request from a preset disk file when the intelligent agent perceives the task request of the first user; the configuration information included in the preset disk file that represents the association mapping relationship between each task and each tool is modifiable.

[0140] The loading unit 402 is configured to load the target tool and target configuration information into the target service program that executes the target task, so as to call the loaded target service program to execute the target task.

[0141] In one implementation, the loading unit 402 is specifically configured to: when it is detected that the target tool and the target configuration information of the target tool in the preset disk file are both in the configuration completed state, load the target configuration information in the configuration completed state into the target service program.

[0142] In another implementation, the loading unit 402 is specifically configured to: when it is detected that the first target configuration information of at least one first target tool in the target tool is in a configuration operation state, load the second target configuration information of at least one second target tool other than the first target tool in the multiple target tools into the target service program; when the first target configuration information of at least one first target tool is in a configuration completion state, load the first target configuration information after the configuration completion state into the target service program.

[0143] In another implementation, the device is further configured to: upon detecting that the target tool and target configuration information have been reconfigured, re-initiate a task request when the reconfiguration of the target tool and target configuration information is completed, so as to re-execute the target task based on the target tool and target configuration information after the reconfiguration is completed.

[0144] In another implementation, a task corresponds to at least one preset disk file, a preset disk file corresponds one-to-one to a tool, and a task corresponds to at least one tool; the perception unit 401 is specifically configured to: determine a target task that matches the task request; select a target disk file corresponding to the target task from multiple preset disk files according to the associated correspondence between the task and the preset disk file; determine the target tool based on the identifier representing the tool type in the target disk file; traverse the file information of the target disk file, and determine the file information including the preset identifier representing the tool configuration information as the target configuration information.

[0145] In another implementation, the device also includes a display unit 403, which is configured to: in response to a modification request from a second user to reconfigure the target tool of the target task, display a target configuration information page including the target tool and the target configuration information of the target tool; the target configuration information page includes a target control for reconfiguring the target tool and the target configuration information; when it is detected that the target control indicates that the reconfiguration of the target tool and / or the target configuration information is completed, send indication information indicating that the reconfiguration of the target tool and / or the target configuration information is completed.

[0146] In another implementation, the target tool and target configuration information are reconfigured, including one or more of the following: adding tools other than the target tool, deleting at least one tool from the target tool, modifying the target configuration information of the target tool, adding configuration information of tools other than the target tool, and deleting configuration information of at least one tool from the target tool.

[0147] Regarding the device in the above embodiment, the specific manner in which each unit module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0148] Figure 5 This is a schematic diagram of an intelligent agent provided by this application. Figure 5 The agent 50 may include at least one processor 501 and a memory 503 for storing processor executable instructions. The processor 501 is configured to execute instructions in the memory 503 to implement the agent tool configuration information processing method in the following embodiment.

[0149] In addition, the intelligent agent 50 may further include a communication bus 502 , at least one communication interface 504 , an input device 506 , and an output device 505 .

[0150] The processor 501 may be a central processing unit (CPU), a microprocessor unit, an ASIC, or one or more integrated circuits for controlling the execution of the program of the present application.

[0151] The communication bus 502 may include a pathway for transmitting information between the aforementioned components.

[0152] The communication interface 504 uses any transceiver or other device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area network (WLAN), etc.

[0153] The input device 506 is used to receive input signals and the output device 505 is used to output signals.

[0154] The memory 503 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, an optical disc storage (including a compressed optical disc, a laser disc, an optical disc, a digital versatile disc, a Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program codes in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may be independent and connected to the processing unit via a bus. The memory may also be integrated with the processing unit.

[0155] The memory 503 is used to store instructions for executing the solution of the present application, and the execution is controlled by the processor 501. The processor 501 is used to execute the instructions stored in the memory 503, thereby realizing the functions of the method of the present application.

[0156] In a specific implementation, as an embodiment, the processor 501 may include one or more CPUs, such as Figure 5 CPU0 and CPU1 in.

[0157] In a specific implementation, as an embodiment, the intelligent agent 50 may include multiple processors, such as Figure 5 1 and 507. Each of these processors may be a single-CPU processor or a multi-CPU processor. A processor herein may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0158] The agent Figure 5 The system includes a processor 501 and a memory 503 for storing executable instructions for the processor 501. The processor 501 is configured to execute the executable instructions to implement the agent tool configuration information processing method according to any of the above possible implementations. The methods can achieve the same technical effects and are not described here in detail to avoid repetition.

[0159] The present application also provides a computer-readable storage medium. When the instructions in the computer-readable storage medium are executed by a processor of an intelligent tool configuration information processing device or electronic device, the intelligent tool configuration information processing device or electronic device can perform the intelligent tool configuration information processing method of any of the possible embodiments described above. The same technical effects can be achieved, and to avoid repetition, they are not further described here.

[0160] The present application also provides a computer program product, including a computer program or instructions, which is executed by a processor to implement the method for processing agent tool configuration information according to any of the above-described possible implementations. The method achieves the same technical effects and is not further described here to avoid repetition.

[0161] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0162] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for processing agent tool configuration information, characterized in that: The method comprises: When the agent perceives a task request from the first user, the agent selects a target tool that provides a service for a target task of the task request and target configuration information of the target tool from a preset disk file; the configuration information representing the association mapping relationship between each task and each tool included in the preset disk file is modifiable; The target tool and the target configuration information are loaded into a target service program that executes the target task, so as to call the loaded target service program to execute the target task.

2. The method according to claim 1, characterized in that The step of loading the target tool and the target configuration information into the target service program of the target task includes: When it is detected that the target tool and the target configuration information of the target tool in the preset disk file are both in a configuration-completed state, the target configuration information in the configuration-completed state is loaded into the target service program.

3. The method according to claim 1, characterized in that The step of loading the target tool and the target configuration information into the target service program of the target task includes: When detecting that first target configuration information of at least one first target tool in the target tool is in a configuration operation state, loading second target configuration information of a second target tool other than the at least one first target tool in the target tool into the target service program; When the first target configuration information of the at least one first target tool is in a configuration-completed state, the first target configuration information in the configuration-completed state is loaded into the target service program.

4. The method according to claim 1, wherein After loading the target tool and the target configuration information into the target service program that executes the target task, the method further includes: In the case where it is detected that the target tool and the target configuration information are reconfigured, the task request is re-initiated when the reconfiguration of the target tool and the target configuration information is completed, so as to re-execute the target task based on the target tool and the target configuration information after the reconfiguration is completed.

5. The method according to any one of claims 1 to 4, characterized in that One task corresponds to at least one preset disk file, one preset disk file corresponds to one tool, and one task corresponds to at least one tool; The step of selecting a target tool and target configuration information of the target tool to provide a service for the target task of the task request from a preset disk file includes: Determining a target task that matches the task request; According to the corresponding relationship between the tasks and the preset disk files, a target disk file corresponding to the target task is selected from the plurality of preset disk files; Determine the target tool based on an identifier representing the tool type in the target disk file; The file information of the target disk file is traversed, and the file information including the preset identifier representing the tool configuration information is determined as the target configuration information.

6. The method according to any one of claims 1 to 4, characterized in that The method further comprises: In response to a modification request from a second user to reconfigure a target tool for the target task, displaying a target configuration information page including the target tool and the target configuration information of the target tool; the target configuration information page including a target control further used to reconfigure the target tool and the target configuration information; When it is detected that the target control indicates that the reconfiguration of the target tool and / or the target configuration information is completed, indication information indicating that the reconfiguration of the target tool and / or the target configuration information is completed is sent.

7. The method according to claim 6, characterized in that The reconfiguration of the target tool and the target configuration information includes one or more of the following: adding other tools besides the target tool, deleting at least one tool in the target tool, modifying the target configuration information of the target tool, adding configuration information of other tools besides the target tool and deleting configuration information of at least one tool in the target tool.

8. An intelligent tool configuration information processing device, characterized in that: The device comprises: The sensing unit is configured to, when the agent senses a task request from the first user, select a target tool that provides a service for a target task of the task request and target configuration information of the target tool from a preset disk file; the configuration information representing the association mapping relationship between each task and each tool included in the preset disk file is modifiable; The loading unit is configured to load the target tool and the target configuration information into a target service program for executing the target task, so as to call the loaded target service program to execute the target task.

9. An information processing system, characterized in that: include: An agent and a user terminal, wherein the agent is configured to execute the agent tool configuration information processing method according to any one of claims 1 to 7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions in the computer-readable storage medium are executed by the processor of the intelligent agent, the intelligent agent is enabled to execute the intelligent agent tool configuration information processing method as described in any one of claims 1 to 7.

Images