Method and system for calling native application service in intelligent Internet of Things operation platform

Abstract

The invention provides a method and system for calling native application services in an intelligent Internet of Things operation platform, and relates to the technical field of Internet of Things operation systems, and the method comprises the steps that an intelligent agent receives a user request, carries out semantic understanding through a semantic understanding model, obtains a target function demand, and sends the target function demand to a server; calling a tool management framework to retrieve in a tool management service component according to the target function requirement to obtain an optimal tool list; determining an optimal tool and a calling parameter set; and initiating a calling request according to the optimal tool and the calling parameter set to obtain a calling result. Through the method, the problems that in the prior art, due to the fact that service calling across applications and across devices cannot be achieved due to the fact that service calling is usually limited to the service range of the intelligent software, ecological development of intelligent software is limited can be solved, the most appropriate tool and service are automatically selected through semantic understanding of the intelligent agent and retrieval of the tool service assembly, and the service efficiency is improved. And cross-tool, cross-application and even cross-equipment task collaboration is completed, and the development of intelligent application ecology is promoted.

Description

Technical Field

[0001] The present application relates to the technical field of Internet of Things operating systems, and in particular to a method and system for calling native application services in an intelligent Internet of Things operating platform. Background Art

[0002] With the rapid development of artificial intelligence technology, especially the emergence of large language models, artificial intelligence has made significant progress in semantic understanding and human intent recognition, and the pace towards general artificial intelligence is accelerating. Large language models have been widely used in many fields due to their powerful dialogue understanding ability, ability to identify diverse human needs, and planning and task processing. However, although large cloud models provide powerful computing power, actual applications still face a series of problems such as high cloud call costs and high latency. With the improvement of terminal chip computing power and the development of intelligent operating systems, it has become possible to deploy large models on the terminal side, and has spawned a large number of native intelligent applications in the field of smart terminals.

[0003] At present, major mobile operating system manufacturers and application software companies are actively building their own native intelligent application ecosystems. As an important part of artificial intelligence applications, intelligent agent technology can realize the effective implementation of large models by combining the generalized understanding ability of large models with professional knowledge and tools. Intelligent agent technology can not only provide more accurate user demand analysis and task planning, but also improve the autonomous intelligence level of terminal devices and reduce dependence on the cloud. The emergence of intelligent agents provides innovative intelligent transformation solutions for various applications. However, in the existing intelligent software ecosystem, whether at the operating system or application software level, the businesses are still relatively fragmented, and the functions of intelligent agents are still limited to a single business scope. Most of them rely on their own closed ecosystems and lack cross-application and cross-device capability calls, making the functions and services of intelligent applications unable to communicate with each other, greatly limiting the comprehensive development of the intelligent software ecosystem. Especially in the era of intelligent interconnection of all things, cross-device and cross-platform collaboration capabilities are key factors in achieving the full implementation of intelligent agents.

[0004] To sum up, there are technical problems in the existing technologies, such as they are often limited to their own business scope and cannot realize cross-application and cross-device service calls, which restricts the development of the intelligent software ecosystem. Summary of the invention

[0005] The purpose of this application is to provide a method and system for calling native application services in an intelligent Internet of Things operating platform, so as to solve the technical problem that the prior art is often limited to its own business scope and cannot realize cross-application and cross-device service calls, which limits the development of the intelligent software ecosystem.

[0006] In view of the above problems, the present application provides a method and system for calling native application services in a smart Internet of Things operating platform.

[0007] In the first aspect, the present application provides a method for calling native application services in an intelligent Internet of Things operating platform, and the method for calling native application services in the intelligent Internet of Things operating platform is applied to an intelligent agent, wherein the method for calling native application services in the intelligent Internet of Things operating platform includes: the intelligent agent receives a user request, uses a semantic understanding model to perform semantic understanding on the user request, and obtains a target functional requirement; the intelligent agent calls a tool management framework according to the target functional requirement to search in a tool management service component to obtain an optimal tool list; the intelligent agent determines an optimal tool and a call parameter set based on the optimal tool list; the intelligent agent initiates a call request to the optimal tool stored in the tool management service component through the tool management framework according to the optimal tool and the call parameter set, and obtains a call result.

[0008] On the second aspect, the present application also provides another method for calling a native application service in a smart Internet of Things operating platform, which is applied to a server, wherein the method for calling a native application service in the smart Internet of Things operating platform includes: the server registers a tool and sets calling permissions in a tool management service component; the server receives a calling request through the tool management service component; the server parses the calling request, performs identity and permission authentication on the calling request according to the calling permissions, and returns a calling result.

[0009] In the third aspect, the present application also provides a calling system for native application services in an intelligent Internet of Things operating platform applied to an intelligent body, which is used to execute the calling method for native application services in the intelligent Internet of Things operating platform as described in the first aspect, wherein the calling system for native application services in the intelligent Internet of Things operating platform includes: a demand acquisition module, the demand acquisition module is used for the intelligent body to receive user requests, use a semantic understanding model to semantically understand the user requests, and obtain target functional requirements; a tool retrieval module, the tool retrieval module is used for the intelligent body to call the tool management framework according to the target functional requirements to search in the tool management service component to obtain an optimal tool list; a tool determination module, the tool determination module is used for the intelligent body to determine the optimal tool and the calling parameter set based on the optimal tool list; a calling request module, the calling request module is used for the intelligent body to initiate a calling request to the optimal tool stored in the tool management service component through the tool management framework according to the optimal tool and the calling parameter set, and obtain a calling result.

[0010] In the fourth aspect, the present application also provides another calling system for native application services in an intelligent Internet of Things operating platform applied to a server, which is used to execute the calling method for native application services in the intelligent Internet of Things operating platform as described in the second aspect, wherein the calling system for native application services in the intelligent Internet of Things operating platform includes: a permission setting module, which is used by the server to register tools and set calling permissions in a tool management service component; a calling request receiving module, which is used by the server to receive calling requests through the tool management service component; and a calling result returning module, which is used by the server to parse the calling request, perform identity and permission authentication on the calling request according to the calling permissions, and return the calling result.

[0011] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0012] The intelligent agent receives the user request, uses the semantic understanding model to semantically understand the user request, and obtains the target function requirements; the intelligent agent calls the tool management framework according to the target function requirements to retrieve the tool management service component and obtain the optimal tool list; the intelligent agent determines the optimal tool and the call parameter set based on the optimal tool list; the intelligent agent initiates a call request to the optimal tool stored in the tool management service component through the tool management framework according to the optimal tool and the call parameter set, and obtains the call result. In other words, with the help of the semantic understanding model, the meaning of the user request is deeply understood, not just the surface text; combined with the optimization mechanism of the tool management framework, when facing complex requirements, it can dynamically select the appropriate tool, and automatically adjust the tool's call parameters according to the task requirements, cross the barriers between different applications, and realize cross-application and cross-end collaboration; automatically select the optimal solution from multiple tools, and automatically adjust the strategy according to the complexity of the task, which improves the autonomy of the intelligent agent, further broadens the application scenarios of the intelligent agent, and promotes the development of the intelligent application ecosystem.

[0013] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0015] Figure 1 A flowchart of a method for calling a native application service in an intelligent Internet of Things operating platform provided in an embodiment of the present application;

[0016] Figure 2 A flowchart of another method for calling a native application service in an intelligent Internet of Things operating platform provided in an embodiment of the present application;

[0017] Figure 3 A schematic diagram of the structure of a calling system for native application services in an intelligent Internet of Things operating platform provided in an embodiment of the present application;

[0018] Figure 4 A schematic diagram of the structure of a calling system for native application services in another intelligent Internet of Things operating platform provided in an embodiment of the present application.

[0019] Explanation of the reference numerals: requirement obtaining module 11 , tool searching module 12 , tool determining module 13 , calling request module 14 , authority setting module 15 , calling request receiving module 16 , calling result returning module 17 . DETAILED DESCRIPTION

[0020] This application solves the technical problem that the existing technology is often limited to its own business scope and cannot realize cross-application and cross-device service calls, which limits the development of the intelligent software ecosystem by providing a method and system for calling native application services in the intelligent Internet of Things operating platform. With the help of the semantic understanding model, the meaning of user requests can be deeply understood, not just the surface text; combined with the optimization mechanism of the tool management framework, when faced with complex needs, it can dynamically select the appropriate tool and automatically adjust the tool's calling parameters according to task requirements, crossing the barriers between different applications and realizing cross-application and cross-end collaboration; automatically selecting the optimal solution from multiple tools and automatically adjusting the strategy according to the complexity of the task, which improves the autonomy of the intelligent body, further broadens the application scenarios of the intelligent body, and promotes the development of the intelligent application ecosystem.

[0021] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0022] For example, please refer to the attached Figure 1 The present application provides a method for calling a native application service in an intelligent Internet of Things operating platform, wherein the method for calling a native application service in the intelligent Internet of Things operating platform is applied to an intelligent entity, and the method for calling a native application service in the intelligent Internet of Things operating platform specifically includes the following steps:

[0023] Step 1: The intelligent agent receives the user request and uses the semantic understanding model to perform semantic understanding on the user request to obtain the target function requirements.

[0024] Specifically, the agent receives user requests, and the user issues instructions or inquiries to the agent in some way (such as voice, text, etc.). An agent is a system that can perceive the environment, understand user needs and perform corresponding tasks. It is usually autonomous, responsive, adaptable and social, and can achieve goals together with users or other agents through interaction. Here it refers to an artificial intelligence system that can receive and process user requests and perform specific tasks. User requests refer to instructions or questions issued by users to the agent in some way (such as voice, text or touch). User requests can include multiple contents, such as a question, a command or a certain demand.

[0025] After receiving the request, the agent processes the user request through the semantic understanding model. The semantic understanding model is a model that uses natural language processing technology (NLP) to parse and understand the intention behind text or voice input, and converts the user's input into structured information that the system can understand. The goal is to not only understand the literal meaning, but also recognize context, emotion, intent and other information through deep analysis of the language. Through semantic analysis, the agent can extract the target functional requirements from the complex and natural user language, that is, the specific tasks or functions that the user wants the agent to perform. The target functional requirements refer to the specific functions or task requirements determined by the agent after understanding the user's request.

[0026] For example, a user sends a request to an agent through a voice assistant (such as Google Assistant, Siri, Alexa) or a smart device (such as a smartphone or home appliance), expressing the request through a simple voice command or text input. In order to understand the request, the agent uses a pre-trained semantic understanding model, which is usually built based on machine learning or deep learning technology, and can analyze the vocabulary, grammatical structure, and contextual information in the user's request to infer the user's intention. The semantic understanding model is usually based on deep learning natural language processing (NLP) technology. After processing by the semantic understanding model, the agent extracts the user's target functional requirements. Through the semantic understanding model, the agent can perform in-depth analysis of the user's natural language and understand the user's true intention, rather than just responding to the literal content.

[0027] Step 2: The intelligent agent calls the tool management framework according to the target function requirements to search in the tool management service component to obtain the optimal tool list.

[0028] Specifically, the agent identifies the user's specific needs based on the user's input (such as voice commands, text queries, etc.). The target functional requirements are submitted to the tool management framework, which stores various tools (such as weather query API, music player, etc.). The agent uses the framework to retrieve and select the most appropriate tool to complete the task. The tool management framework contains a tool library and a query mechanism, which can query the available tools in the tool management service component according to the target requirements. The tool management service component performs a search operation in the tool library to find all tools that can meet the target functional requirements. After retrieving possible tools, the tool management service component will further filter and sort these tools by evaluating multiple criteria such as tool performance, accuracy, response speed, and applicability. Finally, the tool management service component returns an optimal tool list, which contains all tools that meet the requirements and is sorted according to priority. For example, if the target functional requirement is to query the weather, in the scenario of querying the weather, the service component may retrieve tools related to weather queries, such as weather API, meteorological applications, etc., and put the weather API with the fastest response and highest accuracy at the first place in the list. Through the collaboration of the tool management framework and service components, the intelligent agent can quickly retrieve the most suitable tool, reduce the selection of invalid tools, and greatly improve the efficiency of task completion.

[0029] Step 3: The agent determines the optimal tool and call parameter set based on the optimal tool list.

[0030] Specifically, after obtaining the list of optimal tools, the agent needs to select the most suitable tool in the list according to the target requirements and the characteristics of the tool. It is necessary not only to consider whether the function of the tool meets the requirements, but also to consider its performance indicators, such as response speed, accuracy, stability, etc. The optimal tool is selected by analyzing the functions and characteristics of the tool. After selecting the optimal tool, the agent needs to build a call parameter set according to the input requirements of the tool (such as the interface specifications and format requirements of the tool). The call parameter set refers to the parameter set that the agent needs to pass when calling the optimal tool. Each tool requires specific data input to work properly. The call parameter set includes these necessary input data, usually including specific information required by the tool function, such as data provided by the user or data automatically collected by the system. Once the optimal tool is selected and the call parameter set is prepared, the agent will pass this information to the optimal tool through the tool management framework, initiate the actual tool call, execute the task and obtain the return result. By selecting the optimal tool and determining the precise call parameters, the agent can ensure the efficiency and accuracy of the task execution and complete the user request with the most appropriate tool with the least time consumption. Selecting the right tool based on the target functional requirements and accurately preparing the required call parameters can improve the success rate of task processing and reduce failures caused by improper tool selection or incorrect parameters. By selecting the tool with the best performance, the agent can optimize the use of system resources while ensuring task completion and avoid unnecessary calculations and data transmission.

[0031] Step 4: The intelligent agent initiates a call request to the optimal tool stored in the tool management service component through the tool management framework according to the optimal tool and the call parameter set, and obtains a call result.

[0032] Specifically, when the user makes a request, the agent selects the appropriate tool according to the target functional requirements and generates a call request through the tool management framework, which includes the required tool and the corresponding call parameter set. In the scenario of a single device, after the tool management service component receives the agent's call request, it does not need to transmit data through the network, but directly calls the local application service installed on the device to complete the task. The local application service executes the task and returns the result, such as the queried weather data. After receiving the result, the tool management service component passes it to the agent, and the agent will make corresponding feedback or actions based on these results.

[0033] Between multiple devices, the agent still initiates tool call requests based on user needs, specifies the tool to be used and the required call parameter set, and performs this operation through the tool management framework. If the tool capability is located on another device, the tool management service component will pass the call request to the optimal tool in the tool management service component on another device through a cross-end protocol or bus (including but not limited to distributed soft bus, custom protocol, TCP / IP and RPC protocol, etc.), and transmit it through the network to ensure service collaboration between different devices. Through the network protocol, the optimal tool in the tool management service component returns the result of the request to the agent, and the agent further performs tasks based on the result, which may display data, trigger other operations, etc.

[0034] On a single device, the optimal tool in the tool management service component can directly call the local application, avoiding network transmission delays and bandwidth consumption, and can quickly respond to user requests and provide a smooth user experience. In multi-device scenarios, the agent can coordinate operations between different devices through network protocols. The tool management service component can transmit requests and results between multiple devices, achieve efficient collaboration between devices, and expand the functionality and flexibility of the system. Through the tool management framework, the agent and the tool management service component can simplify the calling process. Whether in a single device or multi-device scenario, the agent only needs to focus on the target task, and the framework is responsible for coordinating the selection and calling of tools, thereby reducing the complexity of the system.

[0035] Furthermore, step 2 of this application includes:

[0036] The intelligent agent calls the tool management framework according to the target function requirements to search in the tool management service component to obtain a search tool list; the tool management service component prioritizes the search tool list to obtain the optimal tool list.

[0037] Specifically, the agent searches for tools that can perform the task according to the target functional requirements. The tool management framework is a software architecture that manages and coordinates different tools and services to ensure that the agent can select appropriate tools from the tool library to perform tasks according to requirements. It provides a unified interface that allows the agent to access various tools or services and select the most suitable tool according to task requirements. The tool management service component is a core part of the system architecture. It is responsible for managing the registration, maintenance, call and other related functions of various tools and services. By providing a unified service interface, different tools and services can be efficiently called by agents or applications. The tool management service component usually handles operations such as tool retrieval, sorting and execution. In the tool management service component, various tools that can be called by the agent are stored. The component returns a list of tools according to the agent's needs and performs sorting optimization.

[0038] The tool management framework searches the tool management service component and returns a list of all available tools. This list is a collection of all tools that meet the target functional requirements. The search tool list is a collection of candidate tools that the agent retrieves from the tool management service component based on the target functional requirements, and contains all tools that may meet the requirements. The tool management service component will sort the retrieved tool list, usually based on multiple factors for priority judgment. For example, some tools may have faster response time, more accurate functions, or lower costs. Priority sorting is the process of sorting the retrieved tool list. The basis for sorting can be multiple factors, such as tool efficiency, accuracy, resource consumption, availability, etc., to ensure that the selected tool is most suitable for the current task. Through sorting, the agent can find the best tool to ensure that the task can be completed in the most efficient way.

[0039] The sorted tool list is the optimal tool list, showing the order of tools sorted according to requirements and evaluation factors. The agent will select the best tool in the list to perform the task. By sorting the tool priorities, the agent can efficiently select the most appropriate tool, reduce unnecessary tool calls, and optimize the task execution process. By intelligently retrieving and sorting tools, the agent can accurately select the tool that best meets the requirements to ensure the quality and efficiency of the task. The agent can quickly adapt to different application scenarios and select the most appropriate tool to perform the task, ensuring the efficient operation of the system.

[0040] Further, step 4 of this application includes:

[0041] When there is a single device, the agent initiates a call request to the tool management service component through the tool management framework, and the tool management service component directly makes a local application service call to obtain the call result.

[0042] When there are at least two devices, the intelligent agent initiates a call request to the tool management service component through the tool management framework, and the tool management service component makes the call through the network protocol to obtain the call result.

[0043] Specifically, when there is only one device, the agent will initiate a call request through the tool management framework based on the user's request and needs. A single device means that there is only one device in the system that is working or performing tasks. In this scenario, all tasks and services are performed on this device, and there is no interaction or communication between multiple devices. The user makes a request through interaction with the agent (such as voice commands, text input, etc.). The agent first understands the user's request content and determines the appropriate tool to complete the task. The request is initiated through the tool management framework, which is an intermediary responsible for matching the agent's needs with the appropriate tool. In the scenario of a single device, after receiving the request, the tool management service component checks whether there is a suitable local application that can handle the task. The tool management service component will directly call the local application service installed on the device to perform the task without accessing the cloud service through the network.

[0044] Local application service calls refer to the agent calling existing applications or services on the device to complete tasks. In the scenario of a single device, the agent initiates a request to the tool management service component through the tool management framework, which directly calls the local application or service on the device instead of requesting an external server or cloud service through the network. The local application executes the task and returns the result, which can be the data requested by the user (such as weather information, application status, etc.), or other feedback information after the application is executed. Once the agent receives the results, it presents these results in a user-friendly manner. The call result is the return data or feedback after the tool executes the task, such as the query result, the success or failure of the operation, the output of the execution, etc. The agent continues to process the task or displays the result to the user based on the call result.

[0045] Exemplarily, a user issues a weather query command on a mobile phone. The agent recognizes the user's voice or text input, understands it as a weather query demand, and initiates a weather query request to the tool management service component through the tool management framework. The framework will pass the corresponding call parameters, such as the city name (Shanghai) and date (tomorrow). The tool management service component searches for the weather application installed in the device (such as the weather app), directly calls the local application, passes the required parameters, and requests weather data. The weather app returns the weather information in Shanghai tomorrow, and the tool management service component receives this data and returns it to the agent. The query results are presented to the user through text or voice: The weather in Shanghai tomorrow will be sunny, and the temperature will be between 18 and 25 degrees.

[0046] At least two devices means that there are two or more devices in the system working or performing tasks. In this scenario, the task needs to be completed collaboratively across devices, that is, different devices need to communicate with each other to share data or perform tasks. According to the target task, the agent selects the appropriate tool to execute the request through the tool management framework. Unlike the single-device scenario, in the multi-device scenario, the agent needs to select and coordinate the tools of multiple devices to complete the task together. After receiving the agent's request, the tool management service component determines which device to call the tool on according to the task requirements. Due to the multi-device scenario, the tool management service component cannot complete the task directly through local calls, but communicates with another device through a network protocol. For example, the agent initiates a request on a smartphone to control a smart home device (such as a smart light), and the tool management service component sends the control command to the smart home device through a cross-end protocol or bus.

[0047] After receiving the call request, the target device (such as a smart home device) performs the corresponding operation through the network protocol and returns the result. In a multi-device scenario, the tool management service component sends the request from one device to another through the network protocol and transmits the data through the network. For example, if it is to control a smart light, after the smart home device performs the light-on operation, it will return the result (such as feedback on the success or failure of the operation) to the tool management service component through the network. Once the tool management service component obtains all the required data or execution results, the agent integrates these results and presents the response to the user.

[0048] For example, suppose a user issues two requests in a smart home scenario: turn on the lights in the living room and check tomorrow's weather. The agent first analyzes and understands the user's needs: control home devices and check the weather. Through the tool management framework, two requests are sent to the tool management service component: one is to request a tool to control the lights in the living room (which may exist on the smart home device), and the other is to request a tool to check the weather (which may be on a smartphone or in the cloud). The tool management service component first determines whether the request needs to be called across devices. The request to control the lights will be sent to the smart home device, while the request to check the weather will be sent to the mobile phone or the cloud. After receiving the control command, the smart home device executes the light-on operation and returns a successful feedback. The weather query service (such as the weather application on the mobile phone or the cloud weather interface) executes the query and returns the weather information. After receiving all the results, the agent integrates and feeds back to the user through voice or text: the lights in the living room have been turned on, tomorrow's weather will be sunny, and the temperature will be 18-25 degrees. Through the cross-end protocol, the tool management service component can coordinate the operations between different devices, so that different devices can be combined to complete complex tasks. It only needs to adapt new devices or applications through the tool management framework without reconstructing the entire system. In a multi-device environment, the tool management service component can dynamically select devices according to task requirements, fully utilize the distributed characteristics of the devices, and improve overall execution efficiency.

[0049] Furthermore, the present application also includes the following steps:

[0050] The tool management service component is used to provide interface specifications, wherein the interface specifications include interface category names, interface prompt word descriptions, and input and output formats.

[0051] Specifically, the tool management service component is a core component of the tool management framework. It is responsible for the management, retrieval, and sorting of tools. It provides interfaces for interacting with agents and other modules, helping agents select and call different tools. The tool management service component is a bridge that connects tools with other parts of the system. The interface specification is a set of rules and standards that define how to interact with tools, including the tool calling method, input and output formats, and other necessary information. The interface specification provides tool users (such as agents) with detailed guidelines for operating the tool, ensuring the standardization and consistency of the calling process. The tool management service component provides an interface specification for each tool registered in the tool library, defining how to interact with the tool to ensure that the agent can use the tool correctly according to the predetermined rules. It contains multiple elements, the most important of which are the interface category name, interface prompt word description, and input and output format.

[0052] Each tool is classified into a specific category, which is identified by the interface category name. The interface category name refers to the service type or function classification provided by the tool. Through the interface category name, different types of tools can be organized to facilitate the agent to select the appropriate tool according to the needs. The interface prompt description is the description document provided by the tool management service component for each tool to help the caller understand the function, input and output requirements and usage of the tool. It usually includes what the tool can do, what input parameters are required, what format the returned results are, etc., which helps the agent or developer quickly understand how to use the tool and reduce the possibility of incorrect use. Each tool defines the input and output data formats to ensure that the data passed when calling the tool can be processed correctly and the returned data can be correctly parsed by the agent. In order to ensure the correct calling of the tool, the input data and output data must meet the predetermined format requirements of the tool. The input format may include strings, numbers, JSON objects, etc., and the output format is usually the data format of the tool processing results, such as JSON, XML, etc.

[0053] The tool management service component defines a category for each tool. When the agent receives a user request, it determines the category of the tool according to the requirements and selects the tool of the corresponding category from the tool management framework. The tool management service component provides a clear interface prompt word description for each tool to help the agent understand how to use the tool. It is usually stored in the form of a document and can be automatically displayed to the agent developer or system when the agent needs to call the tool. The tool management service component ensures that all tools follow certain data format standards. The format of input data, the request body of tool call, the format of output data, etc., all need to be unified according to the interface specification. When the agent initiates a tool call, it will format the parameters into the correct form according to the interface specification to ensure that the tool can process it correctly. Through the unified interface specification, the interface prompt word description and the clear provisions of the input and output formats, it is ensured that the agent can call the tool correctly and in a standardized manner, execute user requirements and return results, which greatly improves the efficiency and accuracy of the interaction between the agent and the tool.

[0054] In summary, the method for calling native application services in the intelligent Internet of Things operating platform provided by this application has the following technical effects:

[0055] The intelligent agent receives user requests, uses the semantic understanding model to semantically understand the user requests, and obtains the target functional requirements; the intelligent agent calls the tool management framework according to the target functional requirements to search in the tool management service component to obtain the optimal tool list; the intelligent agent determines the optimal tool and the call parameter set based on the optimal tool list; the intelligent agent initiates a call request to the tool management service component through the tool management framework according to the optimal tool and the call parameter set to obtain the call result. In other words, with the help of the semantic understanding model, the meaning of the user request is deeply understood, not just the surface text; combined with the optimization mechanism of the tool management framework, when facing complex requirements, it can dynamically select the appropriate tool, and automatically adjust the tool's call parameters according to the task requirements, cross the barriers between different applications, and realize cross-application and cross-end collaboration; automatically select the optimal solution from multiple tools, and automatically adjust the strategy according to the complexity of the task, which improves the autonomy of the intelligent agent, further broadens the application scenarios of the intelligent agent, and promotes the development of the intelligent application ecosystem.

[0056] For Example 2, please refer to the attached Figure 2 The present application also provides a method for calling a native application service in an intelligent Internet of Things operating platform, wherein the method for calling a native application service in the intelligent Internet of Things operating platform is applied to a server, and the method for calling a native application service in the intelligent Internet of Things operating platform specifically includes the following steps:

[0057] The server registers the tool and sets the calling authority in the tool management service component; the server receives the calling request through the tool management service component; the server parses the calling request, performs identity and authority authentication on the calling request according to the calling authority, and returns the calling result.

[0058] Specifically, the server refers to a computer system or software module that responds to and processes requests from the client, and is responsible for managing operations such as registration, permission setting, request parsing, and identity authentication of tool services. The server provides a tool registration interface through the tool management service component, allowing different applications or systems to register their own tools (such as smart home control, weather query, etc.) to the server. Each tool needs to provide necessary configuration information, such as interface definition, function description, input and output format, call restrictions, etc. The server sets call permissions in the tool management service component to ensure that only authorized users or devices can use registered tools. Permission settings include defining which roles, users, or devices can call which tools, whether identity authentication is required, and whether to restrict certain operations (such as access frequency restrictions or call restrictions within a specific time period). Permissions can be set based on user identity, device type, call frequency, etc. For example, a family member may have permission to control home devices, while a visitor does not.

[0059] The tool management service component is the core module responsible for tool management and service calls. It manages tasks such as tool registration, permission setting, request reception and processing, provides tool calling interfaces for agents or users, and is responsible for processing and coordinating the use of tools. Tool registration refers to registering available tools (such as smart home control tools, weather query tools, etc.) to the tool management service component for subsequent calls. It usually includes basic information about the tool (such as tool name, function description, input and output format, interface, etc.) and tool usage permissions. Call permission setting refers to defining access control rules for tool calls, which stipulates which users or devices can call which tools and how to use these tools. Permission settings ensure that only legitimate users or devices can call tools to avoid abuse or illegal access.

[0060] When an agent or user makes a request, the server receives the request through the tool management service component, including the name of the requested tool, input parameters, etc. A call request is a request sent by an agent or other system to the server to perform a certain operation (such as calling a tool). After receiving the request, the server needs to parse the request content to confirm which tool is requested and the required parameters. After parsing the request, identity and permission authentication will be performed. First, the server verifies whether the request comes from a legitimate user or device, and confirms it based on the identity information of the requester (such as user name, device ID, IP address, etc.). Secondly, according to the call permission setting, it determines whether the requester has the permission to call the tool. If the user does not have permission, the server may return an error message of access denied; if the permission is passed, the server will continue to process the request. Identity and permission authentication verifies the identity of the requester and checks whether it has the permission to call a specific tool. The authentication process ensures that the requester is legitimate and has the permission to perform the corresponding operation.

[0061] Furthermore, the present application also includes the following steps:

[0062] When the identity and authority authentication is passed, the server performs data format conversion according to the parsing result of the call request to obtain the output result; the server packages the output result according to the interface requirements and returns the call result.

[0063] Specifically, first, the server verifies the identity and permissions of the requester. If the identity authentication is successful and the requester has the permission to call a specific tool, the server will continue to execute the subsequent processing steps. Suppose the user requests to call a smart home control tool to turn on the lights at home, and the user has been authenticated as a family member and has this permission. Next, the server will parse the request to understand the specific operation the user wants to perform. For example, when the user requests to turn on the lights in the living room, the parsing result will include: tool name (smart home control), operation type (turn on the lights), and input parameters (living room lights). The parsing result is the basis for the server to process the request. The server needs to convert the requested data or operation parameters into the format required by the target tool or service. Data format conversion refers to converting the data in the request or the data generated after the server processes it into a format that meets the requirements of the target tool or system interface.

[0064] Between multiple systems, data may need to be converted from one format to another in order to be compatible with different systems or devices. For example, a smart home control tool may require receiving a specific data structure, such as converting string format data into JSON format or other protocol data format that meets the tool interface requirements. Therefore, the server must convert the input parameters according to the tool interface requirements. After the server completes the data format conversion and issues a call request to the corresponding tool, the tool will perform the operation and return the result. It can also process the output result, such as generating detailed feedback or further data based on the tool's return information.

[0065] The output results need to be packaged according to the interface requirements of the tool, that is, the output data needs to be organized into a standard response format so that it can be correctly returned to the requester. Interface requirements refer to the input and output standards and protocols defined by the tool or service. Each interface has a specific input parameter format and output result format, which must be followed to ensure correct communication between systems. Packaging refers to organizing the output results into a complete response package according to the requirements of the interface specification, usually including data, status code, error information, etc., so as to be returned to the requester. After packaging, the results will be encapsulated into a data format that conforms to the network protocol or other communication protocol. The process of packaging the output results according to the interface requirements may include: adding a status code to the response data, packaging the return data, and placing the status information, output results and additional data in a unified data structure.

[0066] The packaged call result will be returned to the agent or user, usually including the status of whether the operation is successful, execution feedback (such as success or failure message) and related data (such as status information after execution). For example, user A passes the identity authentication and is authenticated as a family member with the authority to control home devices. The server parses the user's request, identifies the requested operation to turn on the lights in the living room, and determines that the required tool is the smart home control tool. The user's request parameter is to turn on the lights in the living room, but the interface of the smart home control tool requires data in JSON format, so the server converts the request data into the format required by the tool interface. The smart home tool performs the operation according to the request and returns the result: the light has been successfully turned on. The server packages the output result into a format that meets the interface requirements and returns it to the user, indicating that the operation is successful and notifying the user that the light has been turned on. Through data format conversion and packaging of interface requirements, data compatibility between different systems and tools is ensured, avoiding call errors or unparseable problems caused by inconsistent formats. Through data format conversion and packaging, the server enables different devices and applications to call and be compatible with each other. Whether it is controlling smart home devices or querying weather information, the server can handle it uniformly and ensure the consistency and scalability of the interface. The output results required by the interface are packaged so that the caller only needs to focus on the call results without having to worry about the internal format and data details, making the system more modular and easier to expand and maintain.

[0067] Embodiment 3, based on the same inventive concept as the method for calling native application services in the smart Internet of Things operating platform in the aforementioned embodiment 1, the present application also provides a system for calling native application services in the smart Internet of Things operating platform, please refer to the attached Figure 3 , the calling system of the native application service in the intelligent Internet of Things operating platform includes:

[0068] The demand acquisition module 11 is used for the intelligent agent to receive user requests, use the semantic understanding model to perform semantic understanding on the user requests, and obtain the target function requirements.

[0069] The tool retrieval module 12 is used by the intelligent agent to call the tool management framework according to the target function requirements to search in the tool management service component to obtain the optimal tool list.

[0070] The tool determination module 13 is used by the agent to determine the optimal tool and a set of calling parameters based on the optimal tool list.

[0071] The call request module 14 is used by the intelligent agent to initiate a call request to the optimal tool stored in the tool management service component through the tool management framework according to the optimal tool and the call parameter set, and obtain a call result.

[0072] Furthermore, the tool retrieval module 12 in the calling system of the native application service in the intelligent Internet of Things operating platform is also used for:

[0073] The intelligent agent calls the tool management framework according to the target function requirements to search in the tool management service component to obtain a search tool list; the tool management service component prioritizes the search tool list to obtain the optimal tool list.

[0074] Furthermore, the calling request module 14 in the calling system of the native application service in the intelligent Internet of Things operating platform is also used for:

[0075] When there is a single device, the agent initiates a call request to the tool management service component through the tool management framework, and the tool management service component directly makes a local application service call to obtain the call result.

[0076] Furthermore, the calling request module 14 in the calling system of the native application service in the intelligent Internet of Things operating platform is also used for:

[0077] When there are at least two devices, the intelligent agent initiates a call request to the tool management service component through the tool management framework, and the tool management service component makes the call through the network protocol to obtain the call result.

[0078] Furthermore, the calling request module 14 in the calling system of the native application service in the intelligent Internet of Things operating platform is also used for:

[0079] The tool management service component is used to provide interface specifications, wherein the interface specifications include interface category names, interface prompt word descriptions, and input and output formats.

[0080] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1The calling method and specific examples of the native application service in the intelligent Internet of Things operating platform in the first embodiment are also applicable to the calling system of the native application service in the intelligent Internet of Things operating platform in this embodiment. Through the above detailed description of the calling method of the native application service in the intelligent Internet of Things operating platform, those skilled in the art can clearly know the calling system of the native application service in the intelligent Internet of Things operating platform in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0081] Embodiment 4, based on the same inventive concept as the method for calling native application services in the smart Internet of Things operating platform applied to the server in the aforementioned embodiment 2, this application also provides a system for calling native application services in the smart Internet of Things operating platform, please refer to the attached Figure 4 , the calling system of the native application service in the intelligent Internet of Things operating platform includes:

[0082] The permission setting module 15 is used by the server to register tools and set calling permissions in the tool management service component.

[0083] The call request receiving module 16 is used for the server to receive the call request through the tool management service component.

[0084] The call result returning module 17 is used by the server to parse the call request, perform identity and authority authentication on the call request according to the call authority, and return the call result.

[0085] Furthermore, the calling result returning module 17 in the calling system of the native application service in the intelligent Internet of Things operating platform is also used for:

[0086] When the identity and authority authentication is passed, the server performs data format conversion according to the parsing result of the call request to obtain the output result; the server packages the output result according to the interface requirements and returns the call result.

[0087] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 2The calling method and specific examples of the native application service in the intelligent Internet of Things operating platform in the second embodiment are also applicable to the calling system of the native application service in the intelligent Internet of Things operating platform in this embodiment. Through the above detailed description of the calling method of the native application service in the intelligent Internet of Things operating platform, those skilled in the art can clearly know the calling system of the native application service in the intelligent Internet of Things operating platform in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0088] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0089] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A method for calling native application services in an intelligent Internet of Things operating platform, characterized in that: Applied to intelligent entities, the calling method of the native application service in the intelligent Internet of Things operating platform includes: The intelligent agent receives the user request, uses the semantic understanding model to perform semantic understanding on the user request, and obtains the target function requirements; The intelligent agent calls the tool management framework according to the target function requirements to search in the tool management service component to obtain an optimal tool list; The agent determines an optimal tool and a call parameter set based on the optimal tool list; The intelligent agent initiates a call request to the optimal tool stored in the tool management service component through the tool management framework according to the optimal tool and the call parameter set, and obtains a call result.

2. The method for calling native application services in the intelligent Internet of Things operating platform according to claim 1, characterized in that: include: The intelligent agent calls the tool management framework according to the target function requirement to search in the tool management service component to obtain a search tool list; The tool management service component prioritizes the search tool list to obtain the optimal tool list.

3. The method for calling native application services in the intelligent Internet of Things operating platform according to claim 1, characterized in that: include: When there is a single device, the agent initiates a call request to the tool management service component through the tool management framework, and the tool management service component directly makes a local application service call to obtain the call result.

4. The method for calling native application services in the intelligent Internet of Things operating platform according to claim 1, characterized in that: include: When there are at least two devices, the intelligent agent initiates a call request to the tool management service component through the tool management framework, and the tool management service component makes the call through the network protocol to obtain the call result.

5. The method for calling native application services in the intelligent Internet of Things operating platform according to claim 1, characterized in that: The tool management service component is used to provide interface specifications, wherein the interface specifications include interface category names, interface prompt word descriptions, and input and output formats.

6. A method for calling native application services in an intelligent Internet of Things operating platform, characterized in that: Applied to the server side, the calling method of the native application service in the intelligent Internet of Things operating platform includes: The server registers the tool and sets the calling authority in the tool management service component; The server receives a call request through the tool management service component; The server parses the call request, performs identity and authority authentication on the call request according to the call authority, and returns the call result.

7. The method for calling native application services in the intelligent Internet of Things operating platform according to claim 6, characterized in that: include: When the identity and authority authentication is passed, the server performs data format conversion according to the parsing result of the call request to obtain the output result; The server packages the output results according to the interface requirements and returns the call results.

8. A calling system for native application services in an intelligent Internet of Things operating platform, characterized in that: Steps for implementing the method for calling the native application service in the intelligent Internet of Things operating platform as described in any one of claims 1 to 5, wherein the calling system for the native application service in the intelligent Internet of Things operating platform comprises: A demand acquisition module, wherein the demand acquisition module is used for the intelligent agent to receive user requests, perform semantic understanding on the user requests using a semantic understanding model, and obtain target function requirements; A tool retrieval module, wherein the tool retrieval module is used by the agent to call the tool management framework according to the target function requirements to search in the tool management service component to obtain an optimal tool list; A tool determination module, wherein the tool determination module is used by the agent to determine an optimal tool and a call parameter set based on the optimal tool list; A call request module is used by the intelligent agent to initiate a call request to the optimal tool stored in the tool management service component through the tool management framework according to the optimal tool and the call parameter set, so as to obtain a call result.

9. A calling system for native application services in an intelligent Internet of Things operating platform, characterized in that: Steps for implementing the method for calling the native application service in the intelligent Internet of Things operating platform as described in any one of claims 6 to 7, wherein the calling system for the native application service in the intelligent Internet of Things operating platform comprises: A permission setting module, which is used by the server to register tools and set calling permissions in the tool management service component; A call request receiving module, wherein the call request receiving module is used for the server to receive a call request through the tool management service component; The call result returning module is used by the server to parse the call request, perform identity and authority authentication on the call request according to the call authority, and return the call result.

Images