Application invocation method and apparatus, computer device, and storage medium

Abstract

An application invocation method and apparatus, a computer device, and a storage medium. The application invocation method comprises: acquiring an application set, wherein the application set is a set of target applications corresponding to a target instruction and identified by a preset model on the basis of the target instruction, and the preset model is configured to receive the target instruction and, on the basis of the target instruction, determine and launch a plurality of target applications in the application set; acquiring the target instruction and, on the basis of the target instruction, determining execution steps for the plurality of target applications; and invoking the plurality of target applications, and executing the target instruction according to the execution steps.

Description

Application calling method, device, computer equipment and storage medium

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to Chinese Patent Application No. 202311797385.4 filed on December 25, 2023, the contents of which are incorporated herein by reference in their entirety. Technical Field

[0003] The present invention relates to the field of artificial intelligence technology, and in particular to a method, device, computer equipment and storage medium for calling an application program. Background Art

[0004] With the development of artificial intelligence (AI), its application scenarios are increasing, and the applications of AI are also increasing. The application process of AI includes obtaining tasks posted by users, semantic understanding of tasks, information integration, reasonable reasoning and planning, and finally calling external tools to complete tasks.

[0005] In the process of calling external tools, there are generally two ways, one is to directly call the relevant application programming interface (Application Programming Interface, API) of the interactive application, and the other is to simulate human actions and operate the interactive application interface. The applications mentioned in the present invention, unless otherwise specified, refer to such interactive applications. For example, when using the AI ​​big model to call social software to send messages, since the social software does not provide an application interface to the outside world, the AI ​​big model can only simulate people's click behavior and operate the social software to complete the sending of messages. However, the running speed of this method is greatly restricted. It is necessary to wait for the previous step to be completed before executing the next step, resulting in a long running time, a poor experience for users, and even completely unable to be applied to actual business scenarios.

[0006] Summary of the Invention

[0007] In view of this, the present invention provides a method, apparatus, computer device and storage medium for calling an application program to solve the problem of long running time.

[0008] In a first aspect, the present invention provides a method for calling an application, which is applied to an AI large model, and includes: obtaining an application set; wherein the application set is a set of target applications corresponding to a target instruction identified by a preset model according to the target instruction; the preset model is used to receive the target instruction, and to determine and start multiple target applications in the application set according to the target instruction; obtaining the target instruction, and determining the execution steps of multiple target applications according to the target instruction; calling multiple target applications, and executing the target instruction according to the execution steps.

[0009] The present invention obtains an application set, which is a set of target applications corresponding to a target instruction obtained by a preset model based on target instruction identification. That is, the preset model is used to identify and start the target application corresponding to the target instruction based on the target instruction, determine the execution steps of multiple target applications based on the target instruction, and execute the target instruction according to the execution steps. The present invention starts the target application corresponding to the target instruction in advance and executes the target instruction according to the execution steps. Compared with the related art of opening an application to obtain the execution result and then opening the next application for execution, there is no need to wait for each step to be completed, which reduces the time it takes to start the application, increases the running speed, and improves the user experience.

[0010] In an optional embodiment, determining the execution steps of multiple target applications based on the target instruction includes: identifying and decomposing the target instruction to obtain multiple sub-instructions, determining the execution order of multiple target applications corresponding to the multiple sub-instructions; and determining the execution steps of the multiple target applications based on the execution order.

[0011] The present invention identifies and decomposes the target instruction to obtain different sub-instructions corresponding to different target applications, which can realize parallel execution between different target applications, thereby reducing the execution time of the target instruction. The present invention determines the execution order of multiple target applications corresponding to multiple sub-instructions, determines the execution steps of multiple target applications according to the execution order, and confirms the correlation between each target application according to different sub-instructions. When there is no correlation between the sub-instructions, parallel sub-instructions can be set in the execution order, and the unrelated execution steps in the execution steps are set to parallel execution, thereby shortening the running time and improving the efficiency of the target instruction execution.

[0012] In some optional implementations, the preset model is a fine-tuning-specific small model and / or a rule engine.

[0013] In an optional embodiment, the application set is a set of target applications corresponding to the target instructions identified by the rule engine based on the target instructions; the rule engine is used to extract keywords of the target instructions, and to search the target user keyword database, the system dynamic keyword database, and the system static keyword database based on the keywords, and is also used to determine and start multiple target applications in the application set based on the search results.

[0014] In the present invention, in addition to the preset model that can identify the target instructions, the rule engine can also realize the recognition of the target instructions. Based on the extraction of keywords and multiple searches, multiple target applications are determined, so that the determined multiple target applications are more accurate and more in line with the target instructions.

[0015] In an optional embodiment, the execution steps include parallel execution steps and serial execution steps, wherein the parallel execution steps represent execution steps in which multiple execution steps are processed in parallel; the serial execution steps represent that there is a sequential relationship between the execution steps, and when the first execution step is completed, the second execution step is executed.

[0016] In the present invention, the execution steps include parallel execution steps and serial execution steps, that is, the execution steps include execution steps that can be processed in parallel and execution steps that are executed separately. Parallel processing improves the execution efficiency of target instructions, shortens the running time, and improves the user experience.

[0017] In an optional embodiment, after executing the target instruction according to the execution step, the method further includes: obtaining multiple execution results sent by multiple target applications, combining the multiple execution results to obtain a target execution result, and displaying the target execution result.

[0018] In the present invention, the target execution result is displayed so that the user can clearly know the execution status of the target instruction and ensure that it is executed according to the user's true intention.

[0019] In a second aspect, the present invention provides an application calling system, comprising: a preset model for receiving a target instruction, and for identifying an application set corresponding to the target instruction based on the target instruction, and for starting multiple target applications in the application set based on the target instruction; wherein the application set represents a set of target applications corresponding to the target instruction; a preset model for sending the application set to an AI big model; the AI ​​big model for obtaining the target instruction, and also for determining the execution steps of multiple target applications based on the target instruction; the AI ​​big model for calling multiple target applications and executing the target instruction according to the execution steps.

[0020] The present invention presets a model for receiving a target instruction, and for identifying an application set corresponding to the target instruction according to the target instruction, and for starting multiple target applications in the application set according to the target instruction; wherein the application set represents a set of target applications corresponding to the target instruction; the preset model is used to send the application set to the AI ​​big model, the AI ​​big model is used to obtain the target instruction, and is also used to determine the execution steps of multiple target applications according to the target instruction; the AI ​​big model is used to call multiple target applications and execute the target instruction according to the execution steps. The present invention starts the target application corresponding to the target instruction in advance and executes the target instruction according to the execution steps. Compared with the related art of opening an application to obtain the execution result and then opening the next application for execution, there is no need to wait for the completion of each step, which reduces the time it takes to start the application, increases the running speed, and improves the user experience.

[0021] In some optional implementations, the preset model is a fine-tuning-specific small model and / or a rule engine.

[0022] In some optional implementations, the rule engine receives a target instruction input by a target user and determines a target application based on the interaction among the keyword management module, the target user keyword database, the system dynamic keyword database, and the system static keyword database.

[0023] In an optional embodiment, the system also includes: an AI big model, which is also used to obtain multiple execution results sent by multiple target applications; the AI ​​big model is also used to combine multiple execution results to obtain a target execution result; and is used to display the target execution result.

[0024] In a third aspect, the present invention provides an application calling device, which is applied to an AI large model, and the device includes: an application set acquisition module, which is used to obtain an application set; wherein the application set is a set of target applications corresponding to the target instruction identified by a preset model according to the target instruction; the preset model is used to receive the target instruction, and to determine and start multiple target applications in the application set according to the target instruction; the target instruction acquisition module is used to obtain the target instruction and determine the execution steps of multiple target applications according to the target instruction; the target instruction execution module is used to call multiple target applications and execute the target instruction according to the execution steps.

[0025] In a fourth aspect, the present invention provides a computer device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the method for calling an application of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0026] In a fifth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute a method for calling an application program according to the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0028] FIG1 is a schematic diagram of a flow chart of an application calling method in a related art according to an embodiment of the present invention;

[0029] FIG2 is a schematic flow chart of a method for calling an application program according to an embodiment of the present invention;

[0030] FIG3 is a schematic diagram of a design example of a rule engine according to an embodiment of the present invention;

[0031] FIG4 is a flow chart of a rule engine execution method according to an embodiment of the present invention;

[0032] 5 is a schematic flow chart of a method for calling another application according to an embodiment of the present invention;

[0033] FIG6 is a schematic flow chart of another method for calling an application according to an embodiment of the present invention;

[0034] FIG7 is a flow chart of a method for operating a dedicated small model or rule engine according to an embodiment of the present invention;

[0035] 8 is a schematic diagram of a process flow of an application calling system according to an embodiment of the present invention;

[0036] 9 is a structural block diagram of an application calling device according to an embodiment of the present invention;

[0037] FIG10 is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0038] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0039] As shown in Figure 1, it is a flow chart of the application calling method in the related technology. In the related technology, multiple target applications are called, and the target instructions are executed according to the execution steps. As shown in Figure 1, after receiving the instruction, the AI ​​large model control application operation control module first starts application A according to the instruction, and controls application A to execute the instruction to obtain the execution result a, starts application B according to the execution result a, and controls application B to execute the instruction to obtain the execution result b, starts application C according to the execution result b, and controls application C to execute the instruction to obtain the execution result c.

[0040] Therefore, in the related art, each application program can be started only after it is finished executing, resulting in a long running time and low efficiency.

[0041] The embodiment of the present invention provides a method for calling an application program, which determines and simultaneously starts multiple target application programs through a preset model, so as to shorten the running time and improve the user experience.

[0042] According to an embodiment of the present invention, an embodiment of a method for calling an application is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0043] In some embodiments, a method for calling an application is provided that can be used for an AI large model. FIG2 is a flow chart of the method for calling an application according to an embodiment of the present invention. As shown in FIG2 , the flow includes the following steps:

[0044] Step S201, obtain an application set; wherein the application set is a set of target applications corresponding to the target instruction identified by the preset model according to the target instruction; the preset model is used to receive the target instruction, and to determine and start multiple target applications in the application set according to the target instruction.

[0045] In some optional implementations, the preset model is a fine-tuning-specific small model and / or a rule engine.

[0046] The preset model may be a small model dedicated to fine tuning, a rule engine, or a small model dedicated to fine tuning and a rule engine.

[0047] The preset model is used to identify the target command and determine multiple target applications corresponding to the target command based on the target command.

[0048] Among them, the target instruction is the instruction to be executed input by the target user, and the input form can be voice input, text input, gesture input, etc. The target application is the application corresponding to the target instruction, that is, the target application is the application that needs to be run to execute the target instruction.

[0049] In the embodiment of the present invention, unless otherwise specified, the target applications are all interactive applications.

[0050] In the example, if the target instruction is "summarize the 10 hottest news today and send the summary content to L via email", the preset model will identify the target instruction and obtain the target application as a news application and an email client; generate an application set, which includes a news application and an email client; start the target application, that is, start the news application and the email client.

[0051] In some optional embodiments, the application set is a set of target applications corresponding to the target instructions identified by the rule engine based on the target instructions; the rule engine is used to extract keywords of the target instructions, and to search the target user keyword database, the system dynamic keyword database, and the system static keyword database based on the keywords, and is also used to determine and start multiple target applications in the application set based on the search results.

[0052] Among them, the rule engine is a component embedded in the application. The rules of the rule engine can be customized according to actual needs. In the embodiment of the present invention, the rule engine has the same function as the preset model, but the operation process is different. After receiving the target instruction, the rule engine extracts the keywords of the target instruction, searches the target user keyword database, the system dynamic keyword database and the system static keyword database based on the keywords, and determines the target application according to the search results.

[0053] In the rule engine, the target user keyword database refers to a database that stores commonly used keywords of target users who frequently use the AI ​​large model; the system dynamic keyword database refers to a database that stores keywords that change in real time; and the system static keyword database refers to a database that stores keywords that have not changed.

[0054] As shown in Figure 3, it is a schematic diagram of a design example of a rule engine. In Figure 3, a target instruction input by a target user is received, and in the keyword management module, the keywords of the target instruction are extracted, and the target user keyword database, the system dynamic keyword database, and the system static keyword database are searched based on the keywords to obtain the search results, and the target application is determined based on the search results.

[0055] In an embodiment of the present invention, as shown in FIG4 , there is a schematic diagram of a rule engine execution method, comprising: receiving a target instruction input by a target user according to the rule engine, and determining a target application according to the interaction between the target user keyword database, the system dynamic keyword database, and the system static keyword database of the keyword management module. As shown in FIG4 , in the example, for keywords such as “send a message,” “contact,” “notify,” and “say something,” the corresponding target application is a social application; for keywords such as “buy for me,” “place an order,” “price,” “promotion,” and “shopping live,” the corresponding target application is a shopping application; for keywords such as “hot keywords,” “economic keywords,” “political keywords,” and “celebrity keywords,” the corresponding target application is a news application; for keywords such as “search,” “calculate,” “time keywords,” and “weather keywords,” the corresponding target application is a tool application.

[0056] In the example, in the target user keyword database, "send a message" means that for the target user, the application used is the social software V; in the system dynamic keyword database, "news hot spots" refers to the hot spots on the news software S; in the system static keyword database, "set an alarm" means setting an alarm on the alarm software.

[0057] In the embodiment of the present invention, in addition to the preset model being able to identify the target instructions, the rule engine can also realize the recognition of the target instructions, and determine multiple target applications based on the extraction of keywords and multiple searches, so that the determined multiple target applications are more accurate and more in line with the target instructions.

[0058] Step S202: Obtain target instructions, and determine execution steps of multiple target applications according to the target instructions.

[0059] Among them, the method of obtaining the target instruction can be that the user sends the target instruction in the form of voice to the AI ​​large model, or the target user types the target instruction in the form of text, etc.

[0060] In some optional implementations, the target instruction is identified and decomposed to obtain multiple sub-instructions, and the execution order of multiple target applications corresponding to the multiple sub-instructions is determined; based on the execution order, the execution steps of the multiple target applications are determined.

[0061] In the example, for the target instruction "Summarize the 10 hottest news today and send the summary to L via email", the target instruction is decomposed to obtain the first sub-instruction, running the news application; the second sub-instruction, summarizing the content of the first 10 news; the third sub-instruction, opening the email client; the fourth sub-instruction, entering L's email address in the address bar of the email client; the fifth sub-instruction, pasting the content summarized in the second sub-instruction into the content bar of the email client and sending it; then the execution order of the target application is to start the news application and the email client at the same time; then the execution steps of the target application are: the first sub-instruction and the third sub-instruction are executed in parallel; the second sub-instruction and the fourth sub-instruction are executed in parallel; and finally, the fifth sub-instruction is executed.

[0062] In some optional embodiments, the execution steps include parallel execution steps and serial execution steps, wherein the parallel execution steps represent execution steps in which multiple execution steps are processed in parallel; the serial execution steps represent that there is a sequential relationship between the execution steps, and when the first execution step is completed, the second execution step is executed.

[0063] The order relationship between the first execution step and the second execution step is that the first execution step precedes the second execution step.

[0064] In some optional implementations, there is no dependency between parallel execution steps, but there is a dependency between serial execution steps. The dependency indicates the order of execution steps. In an embodiment of the present invention, when there is a dependency between execution steps, the content of the dependent execution step is filled into the preset position to realize the setting of the dependency.

[0065] In the example, running the news application and opening the email client can be executed in parallel, so running the news application and opening the email client are parallel execution steps, while pasting the content summarized in the second sub-instruction into the content bar of the email client and sending it can only be executed separately, so pasting the content summarized in the second sub-instruction into the content bar of the email client and sending it is a serial execution step.

[0066] In an embodiment of the present invention, the execution steps include parallel execution steps and serial execution steps, that is, the execution steps include execution steps that can be processed in parallel and execution steps that are executed separately. Parallel processing improves the execution efficiency of target instructions, shortens the running time, and improves the user experience.

[0067] Step S203: calling multiple target applications and executing target instructions according to the execution steps.

[0068] In the embodiment of the present invention, the target execution result is displayed so that the user can clearly know the execution status of the target instruction and ensure that it is executed according to the user's intention.

[0069] The application calling method provided in this embodiment obtains an application set. The application set is a set of target applications corresponding to the target instruction obtained by a preset model based on the target instruction. That is, the preset model is used to identify and start the target application corresponding to the target instruction based on the target instruction, determine the execution steps of multiple target applications based on the target instruction, and execute the target instruction according to the execution steps. The embodiment of the present invention starts the target application corresponding to the target instruction in advance and executes the target instruction according to the execution steps. Compared with the related art of opening an application to obtain the execution result and then opening the next application for execution, there is no need to wait for the completion of each step, which reduces the time it takes to start the application, increases the running speed, and improves the user experience.

[0070] In some embodiments, a method for calling an application is provided that can be used for the aforementioned AI large model. FIG5 is a flow chart of another method for calling an application according to an embodiment of the present invention. As shown in FIG5 , the flow includes the following steps:

[0071] Step S501: Obtain an application set. The application set is a set of target applications corresponding to a target instruction identified by a preset model based on the target instruction. The preset model is used to receive the target instruction and, based on the target instruction, determine and launch multiple target applications in the application set. For details, please refer to step S201 of the embodiment shown in FIG. 2 , and will not be repeated here.

[0072] Step S502: Obtain target instructions, and determine execution steps of multiple target applications according to the target instructions. Please refer to step S202 of the embodiment shown in FIG2 for details, which will not be repeated here.

[0073] In some optional implementations, the above step S502 includes:

[0074] Step S5021 : Identify and decompose the target instruction to obtain multiple sub-instructions, and determine the execution order of multiple target application programs corresponding to the multiple sub-instructions.

[0075] Step S5022: Determine the execution steps of the multiple target applications according to the execution order.

[0076] Step S503: Call multiple target applications and execute target instructions according to the execution steps. Please refer to step S203 of the embodiment shown in FIG2 for details, which will not be repeated here.

[0077] Step S504: Acquire multiple execution results sent by multiple target applications, combine the multiple execution results to obtain a target execution result, and display the target execution result.

[0078] The present embodiment provides another application calling method, which identifies and decomposes the target instruction to obtain different sub-instructions corresponding to different target applications, and can realize parallel execution between different target applications, thereby reducing the execution time of the target instruction. The present invention determines the execution order of multiple target applications corresponding to multiple sub-instructions, determines the execution steps of multiple target applications according to the execution order, and confirms the correlation between the target applications according to different sub-instructions. When there is no correlation between the sub-instructions, parallel sub-instructions can be set in the execution order, and the unrelated execution steps in the execution steps are set to parallel execution, thereby shortening the running time and improving the efficiency of the target instruction execution.

[0079] In some embodiments, a method for calling an application is provided, which can be used for the above-mentioned AI large model. FIG6 is a flowchart of another method for calling an application according to an embodiment of the present invention. As shown in FIG6 , the method includes:

[0080] Obtain the target instruction, determine and start application A and application B corresponding to the target instruction based on a dedicated small model or rule engine; control the application operation control module according to the AI ​​large model to decompose the target instruction into a first sub-instruction, a second sub-instruction, a third sub-instruction and a fourth sub-instruction; control application A to execute the third sub-instruction and the fourth sub-instruction, and control application B to execute the first sub-instruction and the second sub-instruction; thereby obtaining the third execution result and the fourth execution result of application A and the second execution result and the first execution result of application B.

[0081] In Figure 6, the second sub-instruction and the fourth sub-instruction represented by bold lines are executed in parallel, and the first sub-instruction and the third sub-instruction represented by thin lines are executed in parallel. The execution order of the embodiment of the present invention is to first execute the first sub-instruction and the third sub-instruction represented by the thin lines, and then execute the second sub-instruction and the fourth sub-instruction represented by the bold lines; the bold dotted lines represent the second execution results and the fourth execution results corresponding to the second sub-instruction and the fourth sub-instruction, and the thin dotted lines represent the first execution results and the third execution results corresponding to the first sub-instruction and the third sub-instruction.

[0082] In this embodiment of the present invention, the first and third sub-instructions are executed first, wherein the first and third sub-instructions are executed in parallel, and application A corresponding to the first sub-instruction and application B corresponding to the third sub-instruction run in parallel. Then, the second and fourth sub-instructions are executed, wherein the second and fourth sub-instructions are executed in parallel, and application A corresponding to the fourth sub-instruction and application B corresponding to the second sub-instruction run in parallel. In this embodiment of the present invention, application A and application B are executed simultaneously according to the sub-instructions, without having to wait for one application to complete execution before executing the other. This increases the degree of parallelism, reduces runtime, and improves the execution efficiency of the target instructions.

[0083] In an embodiment of the present invention, a dedicated small model or rule engine is used to determine and start an application. In the example, as shown in Figure 7, which is a schematic diagram of the operation method of the dedicated small model or rule engine, the dedicated small model or rule engine determines multiple applications corresponding to the target instruction based on the target instruction. It can prepare the operating environment of multiple applications for application A to application N, and start multiple applications. The specially trained small model of the application operation control module controls multiple applications in parallel and executes the target instruction.

[0084] A specially trained small model is a machine learning model used for a specific task. Typically, when training a small model, sufficient domain-specific samples are given. These samples are typically question-and-answer samples, meaning questions and answers are given. By learning from these samples, the model learns reasoning skills in that domain.

[0085] In the implementation manner of the present invention, the specially trained small model is only used in specific scenarios, so the inference computing power will be greatly reduced, and the inference cost will be greatly reduced; the specially trained small model has a very fast response time, because the number of parameters is small and there is a special training process, so the response speed is even faster.

[0086] In some embodiments, an application calling system is provided. FIG8 is a flow chart of the application calling system according to an embodiment of the present invention. As shown in FIG8 , the flow chart includes the following steps:

[0087] Step S801: A preset model is configured to receive a target instruction, identify an application set corresponding to the target instruction based on the target instruction, and launch multiple target applications in the application set based on the target instruction; wherein the application set represents a set of target applications corresponding to the target instruction.

[0088] Step S802: preset a model for sending the application set to the AI ​​big model.

[0089] Step S803: The AI ​​large model is used to obtain target instructions and also to determine the execution steps of multiple target applications based on the target instructions.

[0090] Step S804: The AI ​​large model is used to call multiple target applications and execute target instructions according to the execution steps.

[0091] In some optional implementations, the preset model is a fine-tuning-specific small model and / or a rule engine.

[0092] In some optional embodiments, the rule engine receives a target instruction input by a target user and determines a target application based on the interaction among the keyword management module, the target user keyword database, the system dynamic keyword database, and the system static keyword database.

[0093] In some optional embodiments, the above system also includes: an AI big model, which is also used to obtain multiple execution results sent by multiple target applications; the AI ​​big model is also used to combine multiple execution results to obtain target execution results; and is used to display the target execution results.

[0094] In an embodiment of the present invention, a preset model is used to receive a target instruction, and to identify an application set corresponding to the target instruction according to the target instruction, and to start multiple target applications in the application set according to the target instruction; wherein the application set represents a set of target applications corresponding to the target instruction; the preset model is used to send the application set to the AI ​​big model, the AI ​​big model is used to obtain the target instruction, and is also used to determine the execution steps of multiple target applications according to the target instruction; the AI ​​big model is used to call multiple target applications and execute the target instruction according to the execution steps. The present invention starts the target application corresponding to the target instruction in advance and executes the target instruction according to the execution steps. Compared with the related art of opening an application to obtain the execution result and then opening the next application for execution, there is no need to wait for the completion of each step, which reduces the time it takes to start the application, increases the running speed, and improves the user experience.

[0095] In some embodiments, a device for calling an application program is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments. Details that have already been described will not be repeated here. As used below, the term "module" may refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0096] This embodiment provides a device for calling an application program, as shown in FIG9 , including:

[0097] The application set acquisition module 901 is used to acquire an application set; wherein the application set is a set of target applications corresponding to the target instruction identified by the preset model according to the target instruction; the preset model is used to receive the target instruction, and to determine and start multiple target applications in the application set according to the target instruction.

[0098] The target instruction acquisition module 902 is used to acquire target instructions and determine execution steps of multiple target application programs according to the target instructions.

[0099] The target instruction execution module 903 is used to call multiple target application programs and execute target instructions according to the execution steps.

[0100] In some optional implementations, the target instruction acquisition module 902 includes:

[0101] The execution order determination unit is used to identify and decompose the target instruction to obtain multiple sub-instructions, and determine the execution order of multiple target application programs corresponding to the multiple sub-instructions.

[0102] The execution step determination unit is used to determine the execution steps of multiple target application programs according to the execution order.

[0103] In some optional embodiments, the above device further includes:

[0104] The execution result display module is used to obtain multiple execution results sent by multiple target applications, combine the multiple execution results, obtain the target execution result, and display the target execution result.

[0105] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0106] In some embodiments, the application calling device is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0107] An embodiment of the present invention further provides a computer device having the application calling device shown in FIG. 9 .

[0108] Please refer to Figure 10, which is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention. As shown in Figure 10, the computer device includes: one or more processors 1010, a memory 1020, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in or on the memory for displaying graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used with multiple memories and multiple storage devices. Similarly, multiple computer devices can be connected, with each device providing some of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 10 uses a single processor 1010 as an example.

[0109] Processor 1010 may be a central processing unit, a network processor, or a combination thereof. Processor 1010 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CPLD), a field programmable gate array (FPGA), a general purpose array logic (GAL), or any combination thereof.

[0110] The memory 1020 stores instructions that can be executed by at least one processor 1010, so that the at least one processor 1010 executes the method shown in the above embodiment.

[0111] The memory 1020 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of the computer device, etc. In addition, the memory 1020 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 1020 may optionally include a memory remotely located relative to the processor 1010, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0112] The memory 1020 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 1020 may also include a combination of the above types of memory.

[0113] The computer device further includes a communication interface 1030 for the computer device to communicate with other devices or a communication network.

[0114] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0115] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for invoking an application, applied to an AI large model, the method comprising: Obtaining an application set; wherein, the application set is a set of target applications corresponding to a target instruction identified by a preset model according to the target instruction; the preset model is used to receive the target instruction and to determine and start multiple target applications in the application set according to the target instruction; Obtaining the target instruction and determining execution steps of the multiple target applications according to the target instruction; and Invoking the multiple target applications and executing the target instruction according to the execution steps.

2. The method according to claim 1, wherein The determining the execution steps of the multiple target applications according to the target instruction includes: Identifying and decomposing the target instruction to obtain multiple sub-instructions, and determining an execution order of the multiple target applications corresponding to the multiple sub-instructions; and Determining the execution steps of the multiple target applications according to the execution order.

3. The method according to claim 1 or 2, wherein The preset model is a fine-tuned dedicated small model and / or a rule engine.

4. According to the method of claim 3, wherein, The application set is a set of target applications corresponding to the target instruction identified by the rule engine according to the target instruction; the rule engine is used to extract keywords of the target instruction, and to retrieve a target user keyword database, a system dynamic keyword database, and a system static keyword database according to the keywords, and is further used to determine and start multiple target applications in the application set according to the retrieval result.

5. According to the method of claim 1 or 2, wherein, The execution steps include parallel execution steps and serial execution steps, wherein the parallel execution steps represent execution steps in which multiple execution steps are processed in parallel; the serial execution steps represent that there is an order relationship between the execution steps, and when the first execution step is executed, the second execution step is executed.

6. The method according to claim 1 or 2, wherein After executing the target instruction according to the execution steps, the method further comprises: Obtaining multiple execution results sent by the multiple target applications, combining the multiple execution results to obtain a target execution result, and presenting the target execution result.

7. An application invocation system, the system comprising: A preset model, configured to receive a target instruction, and to identify an application set corresponding to the target instruction according to the target instruction, and to start multiple target applications in the application set according to the target instruction; wherein, the application set represents a set of target applications corresponding to the target instruction; The preset model is configured to send the application set to an AI large model; An AI large model, configured to obtain the target instruction, and to determine execution steps of the multiple target applications according to the target instruction; and The AI large model is configured to invoke the multiple target applications and execute the target instruction according to the execution steps.

8. The system according to claim 6, wherein The preset model is a fine-tuned dedicated small model and / or a rule engine.

9. The system according to claim 8, wherein, The rule engine receives a target instruction input by a target user, and determines a target application program according to the interaction among a keyword management module, a target user keyword database, a system dynamic keyword database, and a system static keyword database.

10. The system according to claim 6, wherein, The system further includes: The AI large model is further configured to obtain multiple execution results sent by the multiple target application programs; and The AI large model is further configured to combine the multiple execution results to obtain a target execution result; and is configured to display the target execution result.

11. An apparatus for invoking an application program, applied to an AI large model, the apparatus includes: An application program set acquisition module, configured to acquire an application program set; wherein, the application program set is a set of target application programs corresponding to the target instruction recognized by a preset model according to the target instruction; the preset model is configured to receive the target instruction, and is configured to determine and start multiple target application programs in the application program set according to the target instruction; A target instruction acquisition module, configured to acquire the target instruction, and determine execution steps of the multiple target application programs according to the target instruction; and A target instruction execution module, configured to invoke the multiple target application programs, and execute the target instruction according to the execution steps.

12. A computer device, including: A memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to execute the application program invocation method according to any one of claims 1 to 6.

13. A computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to cause a computer to execute the application program invocation method according to any one of claims 1 to 6.

Images