Task execution method and device, storage medium and program product

Through the deep learning model's intention analysis and dialogue interaction of user's natural language requests, the problem of low efficiency of IT service management technology is solved, automated and intelligent IT services are realized, and the accuracy and efficiency of task execution are improved.

CN118940767APending Publication Date: 2024-11-12PICC INFORMATION TECH CO LTD
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202410933140.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-11
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing IT service management technology is inefficient and error-prone, making it difficult to adapt to complex and personalized user needs.

Method used

The user's natural language request is analyzed through a deep learning model, and dialogues with the user based on the preliminary intention, determine the end user needs, and convert them into executable IT service tasks for automatic execution.

Benefits of technology

It improves the accuracy and efficiency of IT services, provides automated and intelligent IT services to adapt to complex and personalized user needs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940767A_ABST
    Figure CN118940767A_ABST
Patent Text Reader

Abstract

The embodiment of the invention provides a task execution method and device, a storage medium and a program product, and the method comprises the steps: obtaining a to-be-processed natural language request, carrying out the analysis of the natural language request based on a deep learning model, obtaining a preliminary intention, determining a target dialogue process corresponding to the preliminary intention according to a corresponding relation, and executing the target dialogue process. The corresponding relation comprises the corresponding relation between the preliminary intention and the dialogue process, and the target dialogue process comprises a plurality of system response templates corresponding to possible intentions related to the preliminary intention, obtaining corresponding dialogue content based on the target dialogue process, and determining a final user demand according to the dialogue content and the preliminary intention, the dialogue content comprises reply content which is input by the user and aims at questions in the system response template, determining a corresponding service task according to a final user demand, and executing the service task. According to the method of the embodiment, the accuracy and efficiency of task execution are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of IT service management, and in particular to a task execution method, device, storage medium, and program product. Background Art

[0002] In modern enterprises, IT service management (ITSM) is a key area of ​​supporting and managing computer infrastructure, software and information systems.

[0003] In related technologies, the response and execution of IT service tasks usually rely on manual operations and predefined processes.

[0004] However, in the process of implementing the present application, the inventors found that there are at least the following problems in the prior art: the existing methods are inefficient and prone to errors. Summary of the invention

[0005] The embodiments of the present application provide a task execution method, device, storage medium and program product to improve the accuracy and efficiency of task execution.

[0006] In a first aspect, an embodiment of the present application provides a task execution method, including:

[0007] Obtaining a natural language request to be processed, parsing the natural language request based on a deep learning model, and obtaining a preliminary intent;

[0008] Determine a target dialogue process corresponding to the preliminary intention according to the corresponding relationship; the corresponding relationship includes a corresponding relationship between the preliminary intention and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to possible intentions related to the preliminary intention;

[0009] Based on the target dialogue process, corresponding dialogue content is obtained, and the end user needs are determined according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the question in the system response template;

[0010] Determine corresponding service tasks according to the end-user requirements, and execute the service tasks.

[0011] In a possible design, obtaining corresponding dialogue content based on the target dialogue process, and determining the end user demand according to the dialogue content and the preliminary intention include:

[0012] Perform entity recognition on the conversation content to obtain key information;

[0013] Determine a preliminary service model according to the key information and the preliminary intention; the preliminary service model includes multiple user requirements;

[0014] Obtaining confirmation information of the user on the preliminary service request model;

[0015] According to the confirmation information, the preliminary service model is adjusted to obtain the end-user requirements.

[0016] In a possible design, adjusting the preliminary service model according to the confirmation information to obtain final user requirements includes:

[0017] If the final user's requirements include requirements of a preset type, verification information input by the user for the requirements of the preset type is obtained, so as to obtain corresponding permissions when executing the task corresponding to the requirements of the preset type.

[0018] In a possible design, determining a corresponding service task according to the end user demand and executing the service task includes:

[0019] Determining a plurality of executable tasks according to the end-user requirements;

[0020] Determining the execution order of the plurality of executable tasks according to the dependency relationship among the plurality of executable tasks;

[0021] executing a service task according to the execution order of the plurality of executable tasks and the plurality of executable tasks;

[0022] Determining, from a plurality of execution scripts, a target execution script matching each executable task in the service task based on a preset rule;

[0023] For each executable task, the parameters of the executable task are passed to the corresponding target execution script, and the parameters are verified; if the verification passes, the corresponding target execution script is executed.

[0024] In a possible design, determining a target execution script matching each executable task in the service task from a plurality of execution scripts based on a preset rule includes:

[0025] For each executable task, extract features of the executable task to obtain corresponding key features;

[0026] The key features include at least one of the following: operation type, operation object, and performance requirement; and according to the weights corresponding to the key features, calculating the matching degree between the executable task and each execution script;

[0027] Based on a preset decision rule, the execution script with the greatest matching degree among multiple execution scripts whose specific features in the key features meet preset conditions is determined as the target execution script corresponding to the executable task.

[0028] In a possible design, before parsing the natural language request based on the deep learning model to obtain the preliminary intent, the method further includes:

[0029] Acquire multiple historical natural language requests, and construct a training set according to the multiple historical natural language requests; the historical natural language requests include context information;

[0030] Based on the training set and the pre-built synonym library, the target model is trained to obtain a deep learning model.

[0031] In one possible design, the method further includes:

[0032] After the task is completed, the execution result is fed back to the user;

[0033] Obtaining an optimization requirement for the natural language request input by a user with respect to the execution result;

[0034] If the optimization requirement indicates that the natural language request needs to be optimized, obtaining an optimized natural language request, parsing the optimized natural language request based on a deep learning model, and obtaining a new preliminary intent;

[0035] Determine a target dialogue process corresponding to the new preliminary intention according to the corresponding relationship; the corresponding relationship includes a corresponding relationship between the preliminary intention and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to possible intentions related to the preliminary intention;

[0036] Based on the target dialogue process, corresponding dialogue content is obtained, and the end user needs are determined according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the question in the system response template;

[0037] Determine corresponding service tasks according to the end-user requirements, and execute the service tasks.

[0038] In a second aspect, an embodiment of the present application provides a task execution device, including:

[0039] A parsing module, used to obtain a natural language request to be processed, parse the natural language request based on a deep learning model, and obtain a preliminary intent;

[0040] A determination module, configured to determine a target dialogue flow corresponding to the preliminary intention according to a corresponding relationship; the corresponding relationship includes a corresponding relationship between the preliminary intention and the dialogue flow; the target dialogue flow includes a plurality of system response templates corresponding to possible intentions related to the preliminary intention;

[0041] The determination module is further used to obtain corresponding dialogue content based on the target dialogue process, and determine the end user needs according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the question in the system response template;

[0042] The execution module is used to determine the corresponding service task according to the end user's demand and execute the service task.

[0043] In a third aspect, an embodiment of the present application provides a task execution device, including: at least one processor and a memory;

[0044] The memory stores computer-executable instructions;

[0045] The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the method described in the first aspect and various possible designs of the first aspect.

[0046] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the method described in the first aspect and various possible designs of the first aspect are implemented.

[0047] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the method described in the first aspect and various possible designs of the first aspect.

[0048] The task execution method, device, storage medium and program product provided in the present embodiment include obtaining a natural language request to be processed, parsing the natural language request based on a deep learning model, obtaining a preliminary intent, determining a target dialogue process corresponding to the preliminary intent based on a corresponding relationship, the corresponding relationship including a corresponding relationship between the preliminary intent and the dialogue process, the target dialogue process including a plurality of system response templates corresponding to possible intents related to the preliminary intent, obtaining the corresponding dialogue content based on the target dialogue process, determining the end-user needs based on the dialogue content and the preliminary intent, the dialogue content including the answer content input by the user to the question in the system response template, determining the corresponding service task based on the end-user needs, and executing the service task. The task execution method provided in the embodiment of the present application performs a preliminary intent parsing on the natural language request put forward by the user through a deep learning model, and then further dialogues with the user based on the preliminary intent to obtain more accurate user needs, and then converts the user needs into executable tasks for automatic execution, thereby providing users with automated and intelligent IT services, and improving the accuracy and efficiency of task execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0050] Figure 1 A schematic diagram of an application scenario of the task execution method provided in an embodiment of the present application;

[0051] Figure 2 Schematic diagram of the task execution method provided in the embodiment of the present application Figure 1 ;

[0052] Figure 3 Schematic diagram of the task execution method provided in the embodiment of the present application Figure 2 ;

[0053] Figure 4 A schematic diagram of the structure of a task execution device provided in an embodiment of the present application;

[0054] Figure 5 A schematic diagram of the hardware structure of the task execution device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0056] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0057] It should be noted that the task execution method, device, storage medium and program product provided in this application can be used in the field of IT service management technology, and can also be used in any field outside the field of IT service management technology. The application field of the task execution method, device, storage medium and program product provided in this application is not limited.

[0058] In modern enterprises, IT service management (ITSM) is a key area to support and manage computer infrastructure, software and information systems. As the enterprise environment becomes increasingly complex, users have put forward more and more diverse and personalized demands for IT services. Traditional IT service response methods usually rely on manual operations and predefined processes, which are not only time-consuming and error-prone, but also difficult to adapt to rapidly changing business needs and technical environments.

[0059] With the development of artificial intelligence and machine learning technologies, especially the progress of natural language processing (NLP), a new possibility has been provided, that is, to understand and process users' natural language requests in an automated way, thereby improving the efficiency of IT services and user satisfaction. In actual application scenarios, users may use very natural language, such as "I need more storage space" or "Please help me reset my email password" to express their IT needs. Therefore, how to provide users with automated and intelligent IT services based on artificial intelligence and machine learning technologies is a technical problem that needs to be solved urgently.

[0060] In order to solve the above technical problems, the inventors of this application have found that a deep learning model can be used to perform preliminary intent analysis on natural language requests made by users. In order to accurately understand the user's needs, further dialogue can be conducted with the user based on the preliminary intent to obtain more accurate user needs, and then the user needs can be converted into executable tasks for automatic execution. Based on this, an embodiment of this application provides a task execution method.

[0061] Figure 1 Schematic diagram of application scenarios of the task execution method provided in the embodiment of the present application. Figure 1 As shown, the terminal device 101 is used to receive a natural language request from a user, parse the natural language request based on a deep learning model, and obtain a preliminary intent; based on the corresponding relationship, determine the target dialogue process corresponding to the preliminary intent; the corresponding relationship includes the corresponding relationship between the preliminary intent and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to possible intents related to the preliminary intent; based on the target dialogue process, obtain the corresponding dialogue content, and determine the end-user needs according to the dialogue content and the preliminary intent; the dialogue content includes the answer content input by the user to the question in the system response template; determine the corresponding service task according to the end-user needs, and execute the service task.

[0062] In the specific implementation process, an IT service system is installed in the terminal device 101. When a user needs to request an IT service, a natural language request can be sent to the system. The terminal device 101 parses the natural language request based on a deep learning model to obtain a preliminary intention; the target dialogue process corresponding to the preliminary intention is determined by the dialogue manager according to the corresponding relationship; the corresponding relationship includes the corresponding relationship between the preliminary intention and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to possible intentions related to the preliminary intention; based on the target dialogue process, a dialogue interaction is conducted with the user to obtain the corresponding dialogue content, and the final user needs are determined according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the question in the system response template; the corresponding service task is determined according to the final user needs, and the service task is executed. The task execution method provided in the embodiment of the present application performs a preliminary intent analysis on the natural language request proposed by the user through a deep learning model, and then further dialogues with the user based on the preliminary intention to obtain more accurate user needs, and then converts the user needs into executable tasks for automatic execution, so as to provide users with automated and intelligent IT services, and improve the accuracy and efficiency of task execution.

[0063] It should be noted that Figure 1 The scenario diagram shown is only an example. The task execution method and scenario described in the embodiment of the present application are intended to more clearly illustrate the technical solution of the embodiment of the present application, and do not constitute a limitation on the technical solution provided in the embodiment of the present application. Ordinary technicians in this field can know that with the evolution of the system and the emergence of new business scenarios, the technical solution provided in the embodiment of the present application is also applicable to similar technical problems.

[0064] The technical solution of the present application is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0065] Figure 2 Schematic diagram of the task execution method provided in the embodiment of the present application Figure 1 .like Figure 2 As shown, the method includes:

[0066] 201. Obtain a natural language request to be processed, and parse the natural language request based on a deep learning model to obtain a preliminary intent.

[0067] The execution subject of this embodiment may be a terminal device or a server, for example Figure 1 The terminal device shown.

[0068] In this embodiment, the natural language request may be presented in the form of text, voice, etc.

[0069] Specifically, after the user inputs a natural language request, the terminal device can input the natural language request into a pre-trained deep learning model to parse and obtain the user's initial intention.

[0070] Considering the complexity of IT service requests and the diversity and uncertainty of user expressions, the system may encounter ambiguity or unclear context when parsing users' natural language requests, which may lead to misunderstanding of users' intentions and thus the execution of incorrect IT service tasks. In order to improve the accuracy of parsing the user's initial intentions. In the training of the deep learning model, contextual information is added to the training samples, and a synonym library is specially constructed to enrich the samples of the training set, so that the accuracy and robustness of the trained deep learning model are improved.

[0071] Exemplarily, the training process of the deep learning model includes: obtaining multiple historical natural language requests, and constructing a training set based on the multiple historical natural language requests; the historical natural language requests include context information; based on the training set and a pre-built synonym library, training the target model to obtain a deep learning model.

[0072] Specifically, the following steps may be included:

[0073] Step 1: Data collection and organization.

[0074] Collect data: Gather a large amount of user natural language request data, which usually comes from IT service history, chat logs, service desk interactions, etc.

[0075] Annotated data: Manually annotate the collected user natural language request data and assign one or more intent labels to each request. For example, "Please reset my password" can be labeled as the password reset intent.

[0076] Step 2: Data preprocessing.

[0077] Clean data: remove irrelevant information, such as non-text content, duplicate requests, etc.

[0078] Text normalization: Converting text to a uniform format, for example, converting to lowercase, correcting spelling errors, expanding abbreviations, etc.

[0079] Word segmentation: Split a sentence into lexical units, which is character segmentation or word segmentation for Chinese and word segmentation for English.

[0080] Step 3: Build a synonym database.

[0081] Synonym Identification: Identify and collect synonyms and near-synonyms in the IT service field.

[0082] Synonym building: Build a synonym library for use during training and later query expansion.

[0083] Step 4: Integrate professional terminology.

[0084] Terminology dictionary compilation: Compile a dictionary covering IT service professional terms.

[0085] Knowledge base integration: Merge professional terminology dictionaries with relevant knowledge base materials to provide reference for model training.

[0086] Step 5: Divide the data set.

[0087] Divide the data set: Randomly divide the labeled data set into training set, validation set and test set.

[0088] Step 6: Model training and fine-tuning.

[0089] Pre-trained model loading: Load a pre-trained target model, such as the GPT model.

[0090] Fine-tuning training: Use the training set to fine-tune the target model so that it learns the association between user requests and intent labels.

[0091] Optimize training parameters: Adjust the hyperparameters during fine-tuning (such as learning rate, batch size, training epochs, etc.) based on the performance on the validation set.

[0092] Step 7: Model evaluation.

[0093] Model Validation: Use the validation set to evaluate the performance of the fine-tuned model and check for overfitting.

[0094] Performance testing: Use the test set to evaluate the final performance of the model, focusing on indicators such as precision, recall, and F1 score.

[0095] Step 8: Synonym and context analysis optimization.

[0096] Context analysis: Introduce synonyms and context information during fine-tuning to improve the model’s ability to understand polysemous words.

[0097] Optimization feedback: Further optimize the synonym database and professional knowledge base based on the model's performance in handling polysemous words and professional terms.

[0098] Specific steps may include:

[0099] Step 8.1, construction and integration of synonym library. Word mapping: For commonly used terms and expressions in the IT service field, establish word mapping relationships. For example, "reset password" and "change password" are considered to have the same intention. Synonym collection: Systematically collect synonymous expressions, which may be through the knowledge of professionals or use existing natural language processing tools to assist in identification. Library construction: Create a synonym library to store all synonyms and their mapping relationships to ensure that the system can understand and process different expressions that different users may use.

[0100] Step 8.2: Enhance contextual understanding. Contextual annotation: When annotating data, not only the intent label should be annotated, but also the information that can reflect the contextual relationship, such as the content of the previous sentence and the next sentence. Context simulation: Construct specific contextual scenarios to simulate the user's language usage habits in different scenarios to enhance the model's ability to understand the context.

[0101] Step 8.3: Synonyms and context application in model fine-tuning. Synonym replacement: During the fine-tuning process of the model, randomly replace the words in the training data with their synonyms to increase the model's adaptability to different expressions. Context injection: When fine-tuning the model, ensure that the input contains enough context information so that the model can make intent judgments in a wider context.

[0102] Step 8.4: Synonym and context analysis of model performance: Performance monitoring: During the model evaluation phase, pay special attention to the model's handling of synonyms and ability to understand context. Error analysis: Analyze the model's errors in synonym and context understanding during prediction to identify model deficiencies.

[0103] Step 8.5, optimize the synonym library and context understanding: Feedback loop: Update the synonym library according to the performance monitoring results of the model, including adding missing synonyms or adjusting inaccurate mapping relationships. Context model improvement: Adjust the context annotation strategy according to the error analysis results, or redesign the context information input to the model.

[0104] Step 8.6, Final adjustment and verification of the model: Comprehensive adjustment: According to the optimization results of synonym and context analysis, adjust the fine-tuning strategy of the model, including data re-annotation, update of synonym replacement strategy, etc. Verification and testing: Perform model evaluation again to ensure that the optimization measures have improved the accuracy and robustness of the model. Continuous optimization: Continue to collect feedback in actual applications and regularly update the model to maintain its accuracy and timeliness.

[0105] Step 9: Result analysis and feedback.

[0106] Error analysis: Analyze the incorrect prediction results to find out the shortcomings of the model.

[0107] Feedback iteration: Based on the results of the error analysis, adjust the data annotation, synonym library, and knowledge base, and then repeat steps 5 to 8.

[0108] Step 10: Model deployment.

[0109] Deploy the model: Integrate the trained deep learning model into the IT service automation system.

[0110] 202. Determine a target dialogue process corresponding to the preliminary intention based on the corresponding relationship; the corresponding relationship includes a corresponding relationship between the preliminary intention and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to the possible intentions related to the preliminary intention.

[0111] 203. Based on the target dialogue process, corresponding dialogue content is obtained, and the end user needs are determined according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the questions in the system response template.

[0112] Specifically, after obtaining the initial intent, the system can communicate effectively with the user through the dialogue manager to clarify and refine the specific content of the service request. This step is crucial to ensure the accuracy of the service, especially when facing vague or incomplete requests.

[0113] In some embodiments, in order to improve accuracy, entity recognition technology can be used to extract key information from the conversation content. Specifically, based on the target conversation process, the corresponding conversation content is obtained, and the final user needs are determined according to the conversation content and the preliminary intention, which can include: performing entity recognition on the conversation content to obtain key information; determining a preliminary service model according to the key information and the preliminary intention; the preliminary service model includes multiple user needs; obtaining the user's confirmation information on the preliminary service request model; adjusting the preliminary service model according to the confirmation information to obtain the final user needs.

[0114] In some embodiments, considering that some types of demands may involve authority issues and require users to provide verification information, the user's verification information can be collected in advance to improve convenience. The preliminary service model is adjusted based on the confirmation information to obtain the final user demand, which may include: if the final user demand includes a preset type of demand, the verification information entered by the user for the preset type of demand is obtained to obtain the corresponding authority when performing the task corresponding to the preset type of demand.

[0115] In this embodiment, the determination of the target dialogue process and the guidance of the dialogue interaction can be completed by the dialogue manager (DM). Exemplarily, the specific steps may include:

[0116] Step 1: Start the dialogue manager, which is a software module used to control and manage the interaction process with the user. Its role is to guide the flow of dialogue according to the user's intention and the context of the dialogue, and provide users with accurate service request responses.

[0117] Step 2: The dialogue manager selects a predefined dialogue flow based on the user intent parsed in step 1. This flow contains a series of possible user intents and corresponding system response templates.

[0118] Step 3: Through the dialogue manager, the system asks the user a series of clarification questions to collect more specific information about the service request, such as the specific content of the request, priority, related time requirements, etc. This step is also called the disambiguation process.

[0119] Step 4: The user provides corresponding answers based on the system's clarification questions, and these answers are further processed by the dialogue manager. At this stage, the system can use entity recognition technology to extract key information from the user's answers, such as date, time, device name, etc.

[0120] Step 5: The dialogue manager combines the detailed information and original intent provided by the user to build a complete service request model. This model describes the user's needs in detail and converts them into actionable tasks.

[0121] Step 6: The system performs information verification, asking the user whether the information provided is correct and whether any details need to be supplemented or modified. This step ensures the accuracy of the service request.

[0122] Step 7: After the user confirms the information, the dialogue manager passes the user's complete service request to the back-end IT Service Automation System. If the user needs to modify the information, the dialogue manager will guide the user back to the relevant clarification step, such as step 3.

[0123] Step 8: Before delivering the service request, the dialog manager decides whether further user verification (such as two-factor authentication) or permission approval is required based on the service type and content.

[0124] Step 9. After completing the above steps, the dialogue manager will provide an interaction summary (InteractionSummary) to the user, including the main content of the service request and the subsequent processing flow or timetable.

[0125] 204. Determine a corresponding service task according to the end user's demand, and execute the service task.

[0126] Specifically, in order to automate IT service tasks, user requests can be converted into specific operations and workflows, and seamlessly integrated with existing IT infrastructure.

[0127] In some embodiments, determining the corresponding service task according to the end-user demand and executing the service task may include: determining multiple executable tasks according to the end-user demand; determining the execution order of the multiple executable tasks according to the dependency relationship between the multiple executable tasks; executing the service task according to the execution order of the multiple executable tasks and the multiple executable tasks; determining a target execution script that matches each executable task in the service task from multiple execution scripts based on preset rules; for each executable task, passing the parameters of the executable task to the corresponding target execution script, and verifying the parameters; if the verification passes, executing the corresponding target execution script.

[0128] In some embodiments, determining the target execution script that matches each executable task in the service task from multiple execution scripts based on preset rules may include: for each executable task, performing feature extraction on the executable task to obtain corresponding key features; the key features include at least one of the following: operation type, operation object, performance requirements; calculating the degree of match between the executable task and each execution script based on the weight corresponding to the key features; based on preset decision rules, determining the execution script with the greatest match among multiple execution scripts whose specific features in the key features meet preset conditions as the target execution script corresponding to the executable task.

[0129] Exemplarily, the task execution process may include the following steps:

[0130] Step 1: Task analysis and planning. In this step, the system uses the service request model built in the previous step to analyze and determine the specific IT service tasks to be performed. Task analysis includes converting key information in the service request model (such as intent, entities, etc.) into executable task instructions. The planning stage requires the system to determine the execution order of tasks, resource allocation, and dependency processing.

[0131] Step 2: Automation script matching and calling. Based on the parsing results, the system selects a predefined automation script or execution module that matches the service request. If there are multiple available scripts or modules, the system selects the most appropriate one based on previously defined rules or priorities.

[0132] Exemplarily, the automated script matching process in step 2 may specifically include step 2.1, execution module matching evaluation. Feature extraction: Extract key features from the detailed content of the service request, including operation type, target object, performance requirements, etc. Module feature mapping: Each predefined execution module should have a feature mapping table indicating the applicable features of the module (such as supported operations, operable objects, performance characteristics, etc.). Matching calculation: For each module, calculate the matching degree of its features with the service request features. This can be achieved through a set of algorithms, such as a weight-based scoring system, in which the matching degree of each feature is assigned different weights according to its importance.

[0133] An exemplary matching degree calculation process is as follows:

[0134] Assume that the matching degree is calculated based on the following characteristics and parameters: GPU performance (C), memory size (M), storage space (S), network bandwidth (B), and customer tolerance for latency (L);

[0135] Predefined execution modules (m i ) is the characteristic vector (V i =C i , M i , S i , B i , L i );

[0136] Customer demand vector (D = C d , M d , S d , B d , L d )

[0137] Each feature f i With a weight w j ,The weight is determined according to the business importance of the feature;

[0138] The matching score calculation formula is as follows:

[0139] S i =Σ j∈{C,M,S,B,L} w j ·match(v ij , d j ) (1)

[0140] Where: v ij Indicates execution module m i In the feature f j Performance on j represents the demand value of the corresponding feature in the customer demand vector D; match(v ij , d j ) represents the normalization function;

[0141] For CPU, memory, storage, and bandwidth:

[0142]

[0143] The minimum function is used here to ensure that when the resources of the virtual machine exceed the demand, the matching degree will not exceed 1;

[0144] For latency tolerance:

[0145]

[0146] If the delay exceeds the user's tolerance, the matching degree is 0, otherwise it is 1.

[0147] Step 2.2, execution module selection. Sorting: Sort all available predefined execution modules according to the calculation results of the matching degree. Decision rule application: Apply some decision rules to determine which module is most suitable for the current service request. These rules may include: Highest matching degree first: directly select the module with the highest matching score. Performance requirement satisfaction: Select the module with the highest matching degree on the premise of meeting the performance requirements. Minimum resource consumption: Select the module with the least resource consumption among modules with similar matching degrees. User preference: If the user expresses a preference in the service request, then select the module with the highest matching degree on the premise of meeting the preference. Conflict resolution: If there are multiple modules with equal matching, additional decision factors or manual intervention can be introduced to resolve the selection conflict.

[0148] Step 2.3, perform module verification. Pre-execution check: Before executing the selected module, a series of pre-checks are performed to ensure that the module can be executed normally in the current environment and context. Resource availability check: Confirm the availability of system resources (such as memory, CPU, etc.) to ensure the execution of the selected module. Dependency check: Ensure that all necessary dependencies have been met, such as required library files, data access permissions, etc.

[0149] Step 2.4, execution module preparation. Environment configuration: adjust or configure the execution environment according to the requirements of the selected module. Parameter binding: bind the parameters extracted from the detailed content of the service request to the input parameters of the selected module. Execution preview: an optional step to provide an execution preview to the requester or system administrator so that they can verify whether the operation to be executed is correct. Through the above steps, the system can select the most matching execution module from multiple predefined execution modules, while ensuring the logic and professionalism of the selection process, which is suitable for actual application scenarios.

[0150] Step 3: Parameter configuration and execution preparation. The system passes the parameters extracted from the service request model to the selected automation script or execution module. This step also verifies the validity of the parameters and performs necessary security checks to ensure data security and operational security during execution.

[0151] Step 4: Task execution. Execute the selected automation script or module. This may include calling APIs, performing database operations, adjusting system settings, deploying software updates, etc.

[0152] Step 5: Progress monitoring and exception handling. During the execution of the task, the system will monitor its progress and status in real time. For any exceptions or errors that occur, the system will take predefined exception handling measures, which may include retrying, rolling back, or sending alerts to IT support personnel.

[0153] Step 6: Result verification. After the task is completed, the system will verify whether the result is as expected. This may include checking the integrity of the output data, the correctness of the system status, and the satisfaction of the user request.

[0154] Step 7: Logging and auditing. The system will record key activities and results during the execution process for future auditing and analysis. The log should include information such as execution time, executor (automated system), and task results.

[0155] Step 8. Prepare feedback. Prepare feedback information that users can understand based on the execution results. If the task succeeds, a success message and any requested output data will be provided. If the task fails or partially succeeds, error information, the reason for the failure, and possible solutions or next steps will be provided.

[0156] The task execution method provided in this embodiment performs preliminary intent analysis on the natural language request proposed by the user through a deep learning model, and then conducts further dialogue with the user based on the preliminary intent to obtain more accurate user needs, and then converts the user needs into executable tasks for automatic execution, thereby providing users with automated and intelligent IT services and improving the accuracy and efficiency of task execution.

[0157] Figure 3 Schematic diagram of the task execution method provided in the embodiment of the present application Figure 2 .like Figure 3 As shown, based on the above embodiment, for example Figure 2 On the basis of the embodiment shown, this embodiment adds a detailed description of the feedback process after the service task is completed. The method includes:

[0158] 301. Obtain a natural language request to be processed, and parse the natural language request based on a deep learning model to obtain a preliminary intent.

[0159] 302. Determine a target dialogue process corresponding to the preliminary intention based on the corresponding relationship; the corresponding relationship includes a corresponding relationship between the preliminary intention and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to possible intentions related to the preliminary intention.

[0160] 303. Based on the target dialogue process, corresponding dialogue content is obtained, and the end user needs are determined according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the questions in the system response template.

[0161] 304. Determine a corresponding service task according to the end user's demand, and execute the service task.

[0162] Steps 301 to 304 in this embodiment are similar to steps 201 to 204 in the above embodiment, and are not described again here.

[0163] 305. After the task is completed, the execution result is fed back to the user.

[0164] 306. Obtaining an optimization requirement for the natural language request input by the user with respect to the execution result;

[0165] 307. Determine whether the optimization requirement indicates that the natural language request needs to be optimized. If so, return to step 301; if not, end the process.

[0166] Specifically, after the IT service task is completed, the execution result feedback is provided, and the user determines whether the natural language request needs to be optimized based on the execution result feedback. If optimization is performed, the process continues to return to the previous steps of preliminary intent recognition, user demand determination, and task execution until the user no longer needs to optimize the natural language request.

[0167] Exemplarily, the following steps may be specifically included:

[0168] Step 1: Collect the execution results of the execution module, and convert the collected execution results into natural language descriptions that are easy for users to understand through the natural language generation module.

[0169] Step 2: Feedback the result converted into natural language description to the user through the user interface, and the user determines the subsequent operation based on the feedback information; if the feedback information does not meet expectations, the user needs to optimize the natural language request.

[0170] Step 3: When the user determines that the natural language request needs to be optimized, return to step 1 to step 4 until the natural language request does not need to be optimized.

[0171] The task execution method provided in this embodiment determines the purpose of the user request through context understanding, thereby converting the natural language request into a machine-understandable intent representation; then based on the identified intent, interacts with the user through the dialogue manager to determine the detailed content of the service request, so that the understood intent can be determined to be the correct representation of the user, avoiding the possibility of system misunderstanding, and improving the accuracy of the service request; then, according to the user intent and the detailed content determined by the dialogue, the corresponding IT service task is executed, and feedback is provided after the execution is completed; this greatly improves the accuracy of processing requests, reduces labor costs, and improves user satisfaction.

[0172] Figure 4 This is a schematic diagram of the structure of the task execution device provided in the embodiment of the present application. Figure 4 As shown, the task execution device 40 includes: a parsing module 401 , a determining module 402 and an executing module 403 .

[0173] The parsing module 401 is used to obtain a natural language request to be processed, and parse the natural language request based on a deep learning model to obtain a preliminary intent;

[0174] A determination module 402 is used to determine a target dialogue process corresponding to the preliminary intention according to the corresponding relationship; the corresponding relationship includes a corresponding relationship between the preliminary intention and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to possible intentions related to the preliminary intention;

[0175] The determination module 402 is further used to obtain corresponding dialogue content based on the target dialogue process, and determine the end user needs according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the question in the system response template;

[0176] The execution module 403 is used to determine the corresponding service task according to the end user's demand and execute the service task.

[0177] The task execution device provided in the embodiment of the present application performs preliminary intent analysis on the natural language request put forward by the user through a deep learning model, and then conducts further dialogue with the user based on the preliminary intent to obtain more accurate user needs, and then converts the user needs into executable tasks for automatic execution, thereby providing users with automated and intelligent IT services and improving the accuracy and efficiency of task execution.

[0178] In some embodiments, when the determination module 402 obtains the corresponding dialogue content based on the target dialogue process and determines the end user's needs according to the dialogue content and the preliminary intention, it is specifically used to:

[0179] Perform entity recognition on the conversation content to obtain key information;

[0180] Determine a preliminary service model according to the key information and the preliminary intention; the preliminary service model includes multiple user requirements;

[0181] Obtaining confirmation information of the user on the preliminary service request model;

[0182] According to the confirmation information, the preliminary service model is adjusted to obtain the end-user requirements.

[0183] In some embodiments, when the determination module 402 adjusts the preliminary service model according to the confirmation information to obtain the final user demand, it is specifically used to:

[0184] If the final user's requirements include requirements of a preset type, verification information input by the user for the requirements of the preset type is obtained, so as to obtain corresponding permissions when executing the task corresponding to the requirements of the preset type.

[0185] In some embodiments, when the execution module 403 determines the corresponding service task according to the end user's demand and executes the service task, it is specifically used to:

[0186] Determining a plurality of executable tasks according to the end-user requirements;

[0187] Determining the execution order of the plurality of executable tasks according to the dependency relationship among the plurality of executable tasks;

[0188] executing a service task according to the execution order of the plurality of executable tasks and the plurality of executable tasks;

[0189] Determining, from a plurality of execution scripts, a target execution script matching each executable task in the service task based on a preset rule;

[0190] For each executable task, the parameters of the executable task are passed to the corresponding target execution script, and the parameters are verified; if the verification passes, the corresponding target execution script is executed.

[0191] In some embodiments, when the execution module 403 determines the target execution script matching each executable task in the service task from multiple execution scripts based on a preset rule, it is specifically used to:

[0192] For each executable task, extract features of the executable task to obtain corresponding key features;

[0193] The key features include at least one of the following: operation type, operation object, and performance requirement; and according to the weights corresponding to the key features, calculating the matching degree between the executable task and each execution script;

[0194] Based on a preset decision rule, the execution script with the greatest matching degree among multiple execution scripts whose specific features in the key features meet preset conditions is determined as the target execution script corresponding to the executable task.

[0195] In some embodiments, the parsing module 401 is further configured to:

[0196] Acquire multiple historical natural language requests, and construct a training set according to the multiple historical natural language requests; the historical natural language requests include context information;

[0197] Based on the training set and the pre-built synonym library, the target model is trained to obtain a deep learning model.

[0198] In some embodiments, the execution module 403 is further configured to:

[0199] After the task is completed, the execution result is fed back to the user;

[0200] Obtaining an optimization requirement for the natural language request input by a user with respect to the execution result;

[0201] If the optimization requirement indicates that the natural language request needs to be optimized, obtaining an optimized natural language request, parsing the optimized natural language request based on a deep learning model, and obtaining a new preliminary intent;

[0202] Determine a target dialogue process corresponding to the new preliminary intention according to the corresponding relationship; the corresponding relationship includes a corresponding relationship between the preliminary intention and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to possible intentions related to the preliminary intention;

[0203] Based on the target dialogue process, corresponding dialogue content is obtained, and the end user needs are determined according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the question in the system response template;

[0204] Determine corresponding service tasks according to the end-user requirements, and execute the service tasks.

[0205] The task execution device provided in the embodiment of the present application can be used to execute the above-mentioned method embodiment. Its implementation principle and technical effects are similar, and this embodiment will not be repeated here.

[0206] Figure 5 A schematic diagram of the hardware structure of a task execution device provided in an embodiment of the present application, wherein the device may be a computer, a message sending and receiving device, a tablet device, a server, etc.

[0207] The device 50 may include one or more of the following components: a processing component 501 , a memory 502 , a power component 503 , a multimedia component 504 , an audio component 505 , an input / output (I / O) interface 506 , a sensor component 507 , and a communication component 508 .

[0208] The processing component 501 generally controls the overall operation of the device 50, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 501 may include one or more processors 509 to execute instructions to complete all or part of the steps of the above-mentioned method. In addition, the processing component 501 may include one or more modules to facilitate the interaction between the processing component 501 and other components. For example, the processing component 501 may include a multimedia module to facilitate the interaction between the multimedia component 504 and the processing component 501.

[0209] The memory 502 is configured to store various types of data to support operations on the device 50. Examples of such data include instructions for any application or method operating on the device 50, contact data, phone book data, messages, pictures, videos, etc. The memory 502 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0210] The power supply component 503 provides power to the various components of the device 50. The power supply component 503 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device 50.

[0211] The multimedia component 504 includes a screen that provides an output interface between the device 50 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 504 includes a front camera and / or a rear camera. When the device 50 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and the rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

[0212] The audio component 505 is configured to output and / or input audio signals. For example, the audio component 505 includes a microphone (MIC), and when the device 50 is in an operating mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 502 or sent via the communication component 508. In some embodiments, the audio component 505 also includes a speaker for outputting audio signals.

[0213] I / O interface 506 provides an interface between processing component 501 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: home button, volume button, start button, and lock button.

[0214] The sensor assembly 507 includes one or more sensors for providing various aspects of status assessment for the device 50. For example, the sensor assembly 507 can detect the open / closed state of the device 50, the relative positioning of components, such as the display and keypad of the device 50, and the sensor assembly 507 can also detect the position change of the device 50 or a component of the device 50, the presence or absence of user contact with the device 50, the orientation or acceleration / deceleration of the device 50, and the temperature change of the device 50. The sensor assembly 507 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 507 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 507 may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0215] The communication component 508 is configured to facilitate wired or wireless communication between the device 50 and other devices. The device 50 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 508 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 508 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0216] In an exemplary embodiment, the device 50 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above methods.

[0217] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 502 including instructions, and the instructions can be executed by a processor 509 of the device 50 to perform the above method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0218] The computer-readable storage medium mentioned above can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special-purpose computer.

[0219] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). Of course, the processor and the readable storage medium can also exist in the device as discrete components.

[0220] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0221] An embodiment of the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the task execution method executed by the task execution device as described above.

[0222] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A task execution method, characterized in that: include: Obtaining a natural language request to be processed, parsing the natural language request based on a deep learning model, and obtaining a preliminary intent; Determine the target dialogue flow corresponding to the preliminary intention according to the corresponding relationship; The correspondence relationship includes the correspondence relationship between the preliminary intention and the dialogue flow; The target dialogue process includes a plurality of system response templates corresponding to possible intentions related to the preliminary intention; Based on the target dialogue process, corresponding dialogue content is obtained, and the end user needs are determined according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the question in the system response template; Determine corresponding service tasks according to the end-user requirements, and execute the service tasks.

2. The method according to claim 1, characterized in that The step of acquiring corresponding dialogue content based on the target dialogue process and determining the end user needs according to the dialogue content and the preliminary intention includes: Perform entity recognition on the conversation content to obtain key information; Determine a preliminary service model according to the key information and the preliminary intention; the preliminary service model includes multiple user requirements; Obtaining confirmation information of the user on the preliminary service request model; According to the confirmation information, the preliminary service model is adjusted to obtain the end-user requirements.

3. The method according to claim 2, characterized in that The step of adjusting the preliminary service model according to the confirmation information to obtain final user requirements includes: If the final user's requirements include requirements of a preset type, verification information input by the user for the requirements of the preset type is obtained, so as to obtain corresponding permissions when executing the task corresponding to the requirements of the preset type.

4. The method according to claim 1, characterized in that: Determining the corresponding service task according to the end user's needs and executing the service task includes: Determining a plurality of executable tasks according to the end-user requirements; Determining the execution order of the plurality of executable tasks according to the dependency relationship among the plurality of executable tasks; executing a service task according to the execution order of the plurality of executable tasks and the plurality of executable tasks; Determining, from a plurality of execution scripts, a target execution script matching each executable task in the service task based on a preset rule; For each executable task, the parameters of the executable task are passed to the corresponding target execution script, and the parameters are verified; if the verification passes, the corresponding target execution script is executed.

5. The method according to claim 4, characterized in that The step of determining, from a plurality of execution scripts, a target execution script matching each executable task in the service task based on a preset rule comprises: For each executable task, extract features of the executable task to obtain corresponding key features; The key features include at least one of the following: operation type, operation object, and performance requirement; and according to the weights corresponding to the key features, calculating the matching degree between the executable task and each execution script; Based on a preset decision rule, the execution script with the greatest matching degree among multiple execution scripts whose specific features in the key features meet preset conditions is determined as the target execution script corresponding to the executable task.

6. The method according to any one of claims 1 to 5, characterized in that: Before parsing the natural language request based on the deep learning model to obtain the preliminary intent, the method further includes: Acquire multiple historical natural language requests, and construct a training set according to the multiple historical natural language requests; the historical natural language requests include context information; Based on the training set and the pre-built synonym library, the target model is trained to obtain a deep learning model.

7. The method according to any one of claims 1 to 5, characterized in that: The method further comprises: After the task is completed, the execution result is fed back to the user; Obtaining an optimization requirement for the natural language request input by a user with respect to the execution result; If the optimization requirement indicates that the natural language request needs to be optimized, obtaining an optimized natural language request, parsing the optimized natural language request based on a deep learning model, and obtaining a new preliminary intent; Determine a target dialogue process corresponding to the new preliminary intention according to the corresponding relationship; the corresponding relationship includes a corresponding relationship between the preliminary intention and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to possible intentions related to the preliminary intention; Based on the target dialogue process, corresponding dialogue content is obtained, and the end user needs are determined according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the question in the system response template; Determine corresponding service tasks according to the end-user requirements, and execute the service tasks.

8. A task execution device, characterized in that: include: A parsing module, used to obtain a natural language request to be processed, parse the natural language request based on a deep learning model, and obtain a preliminary intent; A determination module, used to determine a target dialogue flow corresponding to the preliminary intention according to the corresponding relationship; The corresponding relationship includes the corresponding relationship between the preliminary intention and the dialogue process; the target dialogue process includes a plurality of system response templates corresponding to the possible intentions related to the preliminary intention; The determination module is further used to obtain corresponding dialogue content based on the target dialogue process, and determine the end user needs according to the dialogue content and the preliminary intention; the dialogue content includes the answer content input by the user to the question in the system response template; The execution module is used to determine the corresponding service task according to the end user's demand and execute the service task.

9. A task execution device, characterized in that: include: at least one processor and memory; The memory stores computer-executable instructions; The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the task execution method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions. When the processor executes the computer-executable instructions, the task execution method according to any one of claims 1 to 7 is implemented.

11. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the task execution method according to any one of claims 1 to 7 is implemented.

Citation Information

Cited By

  • Task processing method and device of vehicle-mounted system, electronic equipment, medium and product

    CN119336419A

  • Dialogue generation method and device based on application tool, electronic equipment and medium

    CN119621224A

  • Dialogue generation method and device based on application tool, electronic device, and medium

    CN119621224B