Task execution method and device, electronic equipment and storage medium

Abstract

The present disclosure provides a task execution method and device, electronic equipment and storage medium, relates to the technical field of artificial intelligence, in particular to deep learning, natural language processing and large language model technology. The method comprises: performing semantic analysis on task information, determining a plurality of sub-tasks according to the analysis result; generating a task dependency graph according to the dependency relationship between the plurality of sub-tasks; generating a workflow configuration file according to the information of each sub-task node in the task dependency graph; and executing the plurality of sub-tasks according to the workflow configuration file to obtain the execution result of the task information. The present disclosure improves the logic, stability and efficiency of complex task execution.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, specifically deep learning, natural language processing and large language model technology, and in particular to a task execution method, apparatus, electronic device and storage medium. Background Technology

[0002] With the rapid development of artificial intelligence technology, intelligent agents are increasingly being used in various fields such as automated office work, data analysis, and intelligent operation and maintenance. The task information they need to process is also constantly evolving from single, simple instructions to complex, multi-step natural language requirements. To achieve efficient execution of complex tasks, intelligent agents generally break down complex task information and execute each sub-task according to reasonable logic. Summary of the Invention

[0003] This disclosure presents a task execution method, apparatus, electronic device, and storage medium.

[0004] According to a first aspect of this disclosure, a task execution method is provided, comprising: performing semantic parsing on task information, determining multiple subtasks based on the parsing results; generating a task dependency graph based on the dependencies between the multiple subtasks; generating a workflow configuration file based on the information of each subtask node in the task dependency graph; and executing the multiple subtasks based on the workflow configuration file to obtain the execution result of the task information.

[0005] According to a second aspect of this disclosure, a task execution apparatus is provided, comprising: a semantic parsing module configured to perform semantic parsing on task information and determine multiple subtasks based on the parsing results; a graph generation module configured to generate a task dependency graph based on the dependency relationships between the multiple subtasks; a file generation module configured to generate a workflow configuration file based on the information of each subtask node in the task dependency graph; and a task execution module configured to execute the multiple subtasks based on the workflow configuration file to obtain the execution results of the task information.

[0006] According to a third aspect of this disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a method as described in any implementation of the first aspect.

[0007] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions for causing a computer to perform a method as described in any implementation of the first aspect.

[0008] According to a fifth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method as described in any implementation of the first aspect.

[0009] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0010] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein: Figure 1 This is an exemplary system architecture diagram to which this disclosure can be applied; Figure 2 This is a flowchart of the first embodiment of the task execution method according to the present disclosure; Figure 3 This is a flowchart of a second embodiment of the task execution method according to the present disclosure; Figure 4 This is a flowchart of a third embodiment of the task execution method according to the present disclosure; Figure 5 This is a flowchart of the fourth embodiment of the task execution method according to the present disclosure; Figure 6 This is a flowchart of the fifth embodiment of the task execution method according to the present disclosure; Figure 7 This is a flowchart of the sixth embodiment of the task execution method according to the present disclosure; Figure 8 This is a flowchart of the seventh embodiment of the task execution method according to the present disclosure; Figure 9 This is a system flowchart of the task execution method according to this disclosure; Figure 10 This is a schematic diagram of a structure of an embodiment of the task execution apparatus according to the present disclosure; Figure 11 This is a block diagram of an electronic device used to implement the task execution method of the embodiments of this disclosure. Detailed Implementation

[0011] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0012] It should be noted that, unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other. This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0013] Figure 1 An exemplary system frame 100 is shown, to which embodiments of the task execution method or task execution apparatus of the present disclosure may be applied.

[0014] like Figure 1 As shown, system architecture 100 may include terminal devices 101, 102, and 103, a network 104, and a server 105. Network 104 serves as the medium for providing communication links between terminal devices 101, 102, and 103 and server 105. Network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0015] Users can use terminal devices 101, 102, and 103 to interact with server 105 via network 104 to receive or send messages, etc. Various applications for enabling information communication between the terminal devices 101, 102, and 103 and server 105 can be installed. These applications include cloud storage applications and instant messaging applications.

[0016] Terminal devices 101, 102, and 103 and server 105 can be either hardware or software. When terminal devices 101, 102, and 103 are hardware, they can be various electronic devices with displays, including but not limited to smartphones, tablets, laptops, and desktop computers. When terminal devices 101, 102, and 103 are software, they can be installed in the aforementioned electronic devices, and can be implemented as multiple software programs or software modules, or as a single software program or software module; no specific limitation is made here. When server 105 is hardware, it can be implemented as a distributed server cluster composed of multiple servers, or as a single server. When server 105 is software, it can be implemented as multiple software programs or software modules, or as a single software program or software module; no specific limitation is made here.

[0017] Server 105 can provide various services through built-in applications. For example, users can operate through applications on terminal devices 101, 102, and 103 and send task execution requests to server 105. Server 105 can receive and process these task execution requests, performing the following steps: semantic parsing of the task information; determining multiple subtasks based on the parsing results; generating a task dependency graph based on the dependencies between the subtasks; generating a workflow configuration file based on the information of each subtask node in the task dependency graph; and executing the multiple subtasks according to the workflow configuration file to obtain the execution results of the task information.

[0018] It should be noted that the task execution method provided in this embodiment is generally executed by server 105, and correspondingly, the task execution device is generally located in server 105.

[0019] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0020] Continue to refer to Figure 2 The diagram illustrates a flow 200 of a first embodiment of a task execution method according to the present disclosure. The task execution method includes the following steps: Step 201: Perform semantic parsing on the task information and determine multiple sub-tasks based on the parsing results.

[0021] In this embodiment, the execution body of the task execution method (e.g.) Figure 1 The server 105 shown will perform semantic parsing on the task information and determine multiple sub-tasks based on the parsing results.

[0022] Users can input task information on the terminal, which can be in text, voice, or other formats; this embodiment does not specifically limit this. The aforementioned execution entity first performs preprocessing and noise reduction on the task information to filter out invalid interjections, redundant modifiers, and irrelevant notes, corrects ambiguous expressions, and transforms non-declarative instructions into standardized task statements, thereby obtaining noise-free, semantically clear, and standardized information. Then, the aforementioned execution entity performs deep semantic analysis on the standardized information to identify the core execution intent of the task and extract key semantic elements such as the execution entity, core actions, operation objects, target results, and explicit constraints. It then mines the implicit execution requirements and implicit constraints of the task to obtain an analysis result containing complete semantic elements. Finally, the aforementioned execution entity, according to preset task decomposition rules and considering the complexity and execution logic of the task analysis result, decomposes the analysis result into multiple sub-tasks with independent execution boundaries, ensuring that all sub-tasks completely cover the core execution objective of the original task and that there is no significant semantic overlap between the sub-tasks.

[0023] Step 202: Generate a task dependency graph based on the dependencies between multiple subtasks.

[0024] In this embodiment, the execution entity generates a task dependency graph based on the dependencies between multiple subtasks. First, the execution entity determines the relationships between the input and output data of the multiple subtasks, systematically analyzing the input and output data items of each subtask to clarify the precise correspondence between the output and input data of each subtask, and marking the direction of data flow and associated data types, thus obtaining a subtask data association list. Then, the execution entity determines the dependencies between the execution order of the multiple subtasks, and, in conjunction with the overall task execution logic, determines the preceding and following execution nodes of each subtask to distinguish between sequential and parallel execution order constraints. The execution entity also identifies the starting subtask without preceding dependencies and the terminating subtask without following dependencies, thus obtaining a subtask execution order constraint list.

[0025] The aforementioned execution entity will treat each subtask as an independent node in the dependency graph and assign it a unique identifier. Based on the data association list and the execution order constraint list, it will determine the direction of the edges between nodes and distinguish the attributes of the edges according to the relationship type. It will perform acyclicity, integrity, and logical consistency checks on the constructed basic dependency graph and form the final task dependency graph after the checks pass.

[0026] Step 203: Generate a workflow configuration file based on the information of each subtask node in the task dependency graph.

[0027] In this embodiment, the aforementioned execution entity generates a workflow configuration file based on the information of each sub-task node in the task dependency graph. The execution entity then transforms each sub-task node in the task dependency graph into an executable workflow node according to preset transformation rules, and matches each sub-task node with an execution tool or capability module that the agent can invoke. It also supplements the node execution timeout threshold, input parameter format, output result specifications, and other engineering execution attributes, and assigns each workflow node a unique identifier and associates it with corresponding sub-task information, thereby obtaining a standardized set of executable workflow nodes. Then, the execution entity transforms the topological relationships between each sub-task node into workflow execution rules. Based on the pointing and attributes of the edges between nodes in the task dependency graph, it defines the pre-dependency triggering conditions, serial or parallel execution rules, and scheduling priorities of the workflow nodes, clarifies the execution judgment rules for condition-triggered nodes, and the connection logic between upstream and downstream nodes, thus obtaining a workflow execution rule set covering all nodes.

[0028] Finally, the aforementioned execution entity will integrate workflow nodes and execution rules to generate a workflow configuration file. That is, the execution entity will integrate the attribute information of the workflow node set and the constraint requirements of the execution rule set according to the logical association between node identifiers and execution rules, and perform integrity and consistency verification on the integrated content. After the integrated content passes the verification, it will be encapsulated in a standardized format to generate a structured workflow configuration file containing node configuration, execution rules, and node association relationships.

[0029] Step 204: Execute multiple subtasks according to the workflow configuration file to obtain the execution results of the task information.

[0030] In this embodiment, the aforementioned execution entity executes multiple subtasks according to the workflow configuration file, obtaining the execution results of the task information. The execution entity loads and parses the structured workflow configuration file, verifying its integrity and validity. After successful verification, it extracts core information such as workflow node configurations, execution rules, and node relationships, thereby obtaining a set of runtime execution instructions that can be directly invoked by the execution engine. The execution entity also automatically schedules and executes all workflow nodes according to the execution rules in the workflow configuration file, prioritizing the startup of nodes without pre-dependencies and monitoring the node execution status in real time. After the pre-dependency node completes execution and outputs valid results, it automatically triggers the execution of downstream dependent nodes, completing the orderly scheduling of serial and parallel nodes according to the rules, thereby ensuring that each node executes its corresponding subtask according to preset logic.

[0031] In addition, the aforementioned execution entity records the execution logs and output results of each workflow node, integrates the execution results of all subtasks according to the node relationships configured in the workflow, and thus obtains the initial execution result of the task information. The execution entity then verifies the completeness and validity of the initial execution result to confirm whether the result fully covers the execution objectives of the original task information and meets preset constraints. Finally, the execution entity standardizes the verified results to obtain the final execution result of the task information.

[0032] The task execution method disclosed herein first performs semantic parsing on task information to determine multiple subtasks based on the parsing results; then, a task dependency graph is generated based on the dependencies between the subtasks; next, a workflow configuration file is generated based on the information of each subtask node in the task dependency graph; finally, the multiple subtasks are executed according to the workflow configuration file to obtain the execution result of the task information. This method decomposes tasks into subtasks through semantic parsing, generates a task dependency graph through dependency modeling, and then transforms it into a structured workflow configuration file to drive execution. This achieves full-process standardization from natural language tasks to automated execution, making the task execution process parsable and controllable, avoiding the uncertainty of implicit reasoning, thereby improving the logic, stability, and efficiency of complex task execution, while reducing process maintenance costs.

[0033] Furthermore, the collection, storage, use, processing, transmission, provision, and disclosure of any type of information, such as user personal information, involved in the technical solutions disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0034] Continue to refer to Figure 3 , Figure 3 A flow 300 of a second embodiment of a task execution method according to the present disclosure is shown. The task execution method includes the following steps: Step 301: Perform preprocessing and noise reduction on the task information to obtain standardized information.

[0035] In this embodiment, the execution body of the task execution method (e.g.) Figure 1 The server 105 shown will format the task information, standardizing text encoding, line breaks, and punctuation to remove redundant content such as blank lines, invalid spaces, and duplicate characters, and to standardize the sentence structure of non-standard expressions. Then, the aforementioned execution entity will perform content noise reduction and filtering to remove interjections, redundant modifiers, and irrelevant notes and chatter from the task information, retaining only the core task description. Afterwards, the execution entity will correct vague and non-standard expressions, thus completing the colloquial and abbreviated task description into a complete written expression, standardizing the naming of professional terms and operation objects, and eliminating semantic ambiguity. Furthermore, the execution entity will extract the core task description and standardize its encapsulation, organizing the text according to a fixed structure of execution action + operation object + task goal, thereby obtaining standardized information that is semantically clear, free of redundancy, and unambiguous.

[0036] Step 302: Perform semantic parsing on the standardized information to obtain the parsing results.

[0037] In this embodiment, the aforementioned execution entity performs deep semantic analysis on standardized information based on a natural language processing model. First, the execution entity identifies the core execution intent of the task and extracts core semantic elements such as the execution entity, core operation actions, specific operation objects, and expected task goals. Second, the execution entity mines explicit execution constraints in the information and, in conjunction with the task scenario, derives implicit execution requirements, result output specifications, and operational boundaries. Finally, the execution entity performs structured sorting and integration of all extracted and mined semantic information to obtain a structured analysis result that is complete in elements, logically clear, and unambiguous.

[0038] Step 303: Divide the parsing results into multiple subtasks according to the preset task decomposition rules.

[0039] In this embodiment, the aforementioned execution entity, based on preset task decomposition rules and combined with the core execution intent, semantic elements, and execution constraints in the parsing results, decomposes the overall task hierarchically according to execution logic, operational steps, or functional modules, following the decomposition principle of "single task, independent execution, and clear boundaries." Then, the execution entity further clarifies the core actions, operational objects, execution requirements, and expected outputs of each subtask, ensuring that there is no semantic overlap or execution omission among the subtasks, and that the execution goals of all subtasks collectively cover all the requirements of the original overall task. Furthermore, the execution entity organizes and numbers the decomposed subtasks, forming a subtask list containing basic information and execution associations for each subtask, thus completing the decomposition and transformation of the overall task into multiple subtasks.

[0040] Step 304: Generate a task dependency graph based on the dependencies between multiple subtasks.

[0041] Step 305: Generate a workflow configuration file based on the information of each subtask node in the task dependency graph.

[0042] Step 306: Execute multiple subtasks according to the workflow configuration file to obtain the execution results of the task information.

[0043] Steps 304-306 are basically the same as steps 202-204 in the aforementioned embodiments. For specific implementation methods, please refer to the aforementioned description of steps 202-204, which will not be repeated here.

[0044] from Figure 3 It can be seen from this that, with Figure 2 Compared with the corresponding embodiments, the task execution method in this embodiment emphasizes the step of splitting into multiple sub-tasks. This method eliminates redundant ambiguity through preprocessing noise reduction, making the task information standardized and semantically clear. Then, it extracts complete core semantic elements through deep analysis, and finally decomposes them into sub-tasks with clear boundaries and no overlap or omissions according to rules, ensuring the logic and orderliness of the overall task execution.

[0045] Continue to refer to Figure 4 , Figure 4 A flow 400 of a third embodiment of a task execution method according to the present disclosure is shown. The task execution method includes the following steps: Step 401: Perform preprocessing and noise reduction on the task information to obtain standardized information.

[0046] Step 402: Perform semantic parsing on the standardized information to obtain the parsing results.

[0047] Step 403: Divide the parsing results into multiple subtasks according to the preset task decomposition rules.

[0048] Steps 401-403 are basically the same as steps 301-303 in the aforementioned embodiments. For specific implementation methods, please refer to the aforementioned description of steps 301-303, which will not be repeated here.

[0049] Step 404: Determine the relationship between the input and output data of multiple subtasks.

[0050] In this embodiment, the execution body of the task execution method (e.g.) Figure 1 The server 105 shown will determine the relationship between the input and output data of multiple subtasks, sort out the input data items and output data items of each subtask one by one, thereby clarifying the precise correspondence between the output data and input data of each subtask, and marking the direction of data flow and the associated data type, thereby obtaining a list of subtask data associations.

[0051] Step 405: Determine the dependencies between the execution order of multiple subtasks.

[0052] In this embodiment, the execution entity determines the dependencies between the execution order of multiple subtasks and, in conjunction with the overall task execution logic, identifies the preceding and following execution nodes of each subtask to distinguish between sequential and parallel execution order constraints. The execution entity also identifies the starting subtask without preceding dependencies and the ending subtask without following dependencies, thereby obtaining a list of subtask execution order constraints.

[0053] Step 406: Construct a task dependency graph based on associations and dependencies.

[0054] In this embodiment, the execution entity treats each subtask as an independent node in the dependency graph and assigns it a unique identifier. It determines the direction of the edges between nodes based on the data association list and the execution order constraint list, distinguishes the attributes of the edges according to the relationship type, performs acyclicity, integrity, and logical consistency checks on the constructed basic dependency graph, and forms the final task dependency graph after the checks pass.

[0055] Step 407: Generate a workflow configuration file based on the information of each subtask node in the task dependency graph.

[0056] Step 408: Execute multiple subtasks according to the workflow configuration file to obtain the execution results of the task information.

[0057] Steps 407-408 are basically the same as steps 203-204 in the aforementioned embodiments. For specific implementation methods, please refer to the aforementioned description of steps 203-204, which will not be repeated here.

[0058] from Figure 4 It can be seen from this that, with Figure 3 Compared to the corresponding embodiments, the task execution method in this embodiment emphasizes the step of generating a task dependency graph. By mapping subtasks as nodes and defining dependencies as edges through the association of data input and output between subtasks and the dependency relationship of execution order, a task dependency graph is constructed, thereby realizing the visualization and structured modeling of the relationship between subtasks and clearly presenting the task execution logic and data flow path.

[0059] Continue to refer to Figure 5 , Figure 5 A flow 500 of a fourth embodiment of a task execution method according to the present disclosure is shown. The task execution method includes the following steps: Step 501: Perform preprocessing and noise reduction on the task information to obtain standardized information.

[0060] Step 502: Perform semantic parsing on the standardized information to obtain the parsing results.

[0061] Step 503: Divide the parsing results into multiple subtasks according to the preset task decomposition rules.

[0062] Step 504: Determine the relationships between the input and output data of multiple subtasks.

[0063] Step 505: Determine the dependencies between the execution order of multiple subtasks.

[0064] Steps 501-505 are basically the same as steps 401-405 in the previous embodiment. For specific implementation methods, please refer to the above description of steps 401-405, which will not be repeated here.

[0065] Step 506: Treat the subtasks as nodes in the basic dependency graph; determine the direction of the edges in the basic dependency graph based on the association and dependency relationships.

[0066] In this embodiment, the execution body of the task execution method (e.g.) Figure 1The server 105 shown treats subtasks as nodes in the basic dependency graph, assigning each subtask a unique identifier and making it an independent node in the graph, while fully recording the core information of each subtask. Furthermore, the execution entity determines the direction of edges in the basic dependency graph based on associations and dependencies. According to the input-output data relationships between subtasks, it points the subtask node providing output data to the subtask node receiving that data as input. Combined with the execution order dependency, it points the preceding subtask node to the following subtask node, clarifying the unidirectional pointing relationship between all nodes.

[0067] Step 507: Perform multi-dimensional verification of the logic of each node and edge in the basic dependency graph to obtain the verified task dependency graph.

[0068] In this embodiment, the execution entity performs node integrity verification on the basic dependency graph to confirm that all subtasks are mapped to nodes without missing or duplicate nodes. The execution entity also performs edge pointing logic verification to check whether the edge direction is consistent with the data association and execution order dependency relationship, and whether there are any reverse or incorrect pointings. Based on this, the execution entity also performs a loop closure verification to check for any circular dependency nodes in the graph, ensuring there are no execution logic loops. Finally, the execution entity completes association validity verification to verify whether the edges between nodes correspond to the actual subtask dependencies, and whether there are any invalid association edges; and after correcting any problems found in the verification, a logically rigorous and accurately verified task dependency graph is obtained.

[0069] Step 508: Generate a workflow configuration file based on the information of each subtask node in the task dependency graph.

[0070] Step 509: Execute multiple subtasks according to the workflow configuration file to obtain the execution results of the task information.

[0071] Steps 508-509 are basically the same as steps 203-204 in the aforementioned embodiments. For specific implementation methods, please refer to the aforementioned description of steps 203-204, which will not be repeated here.

[0072] from Figure 5 It can be seen from this that, with Figure 4 Compared to the corresponding embodiments, the task execution method in this embodiment emphasizes the step of constructing a task dependency graph. By mapping subtasks to dependency graph nodes and defining the direction of edges based on data and sequential dependencies, and then correcting problems through multi-dimensional logical verification, a compliant task dependency graph is finally formed. This achieves structured and visual modeling of subtask relationships, ensuring that the dependency logic is error-free and without closed loops, thereby improving the accuracy, rigor, and standardization of subtask dependency relationship modeling.

[0073] Continue to refer to Figure 6 , Figure 6 A flow 600 of a fifth embodiment of a task execution method according to the present disclosure is shown. The task execution method includes the following steps: Step 601: Perform semantic parsing on the task information and determine multiple sub-tasks based on the parsing results.

[0074] Step 602: Generate a task dependency graph based on the dependencies between multiple subtasks.

[0075] Steps 601-602 are basically the same as steps 201-202 in the aforementioned embodiments. For specific implementation methods, please refer to the aforementioned description of steps 201-202, which will not be repeated here.

[0076] Step 603: Convert each subtask node in the task dependency graph into an executable workflow node according to the preset conversion rules.

[0077] In this embodiment, the execution body of the task execution method (e.g.) Figure 1 The server 105 shown extracts basic information such as the unique identifier, core execution requirements, and input / output data specifications of each sub-task node according to preset node transformation rules, and maps them to workflow node attributes. The execution entity also matches suitable execution tools, capability modules, and runtime resources for each workflow node, while supplementing the settings with engineered execution configurations such as node execution timeout thresholds, parameter validation rules, and result output formats. Finally, the execution entity synchronizes the association information of the sub-task nodes to the attribute fields of the corresponding workflow nodes, assigning each workflow node standardized execution attributes that can be recognized and invoked by the execution engine, thereby obtaining executable workflow nodes with independent execution capabilities and complete attributes.

[0078] Step 604: Transform the topological relationships between each subtask node into workflow execution rules.

[0079] In this embodiment, the execution entity extracts the topological relationships between subtask nodes in the task dependency graph. Based on the pointers and attributes of node edges, it transforms these relationships into prerequisite dependency triggering rules for workflow nodes, while also defining the prerequisite node completion conditions for downstream node initiation. The execution entity then combines the serial and parallel relationship characteristics in the topological relationships to formulate sequential execution rules for nodes and resource allocation and synchronization scheduling rules for parallel execution. Simultaneously, the execution entity transforms these rules into node scheduling priority rules, execution judgment rules for condition-triggered nodes, and data flow connection rules for upstream and downstream nodes. Through integration, a complete workflow execution rule system that can be recognized by the execution engine is obtained.

[0080] Step 605: Integrate workflow nodes and execution rules to generate workflow configuration files.

[0081] In this embodiment, the execution entity, following a preset structured configuration specification, integrates the complete attribute information of all executable workflow nodes with the workflow execution rule system, establishing a mapping relationship between node attributes and corresponding execution rules based on the node's unique identifier, thus obtaining an initial workflow configuration file. Then, the execution entity performs integrity and consistency checks on the initial file, verifying that node information is complete, rules and node topology relationships are conflict-free, and configuration items are logically consistent. After successful verification, the file is encapsulated in a standardized format to generate a structured workflow configuration file containing node configurations, rule mappings, and relationships, which can be directly parsed and invoked by the execution engine.

[0082] Step 606: Execute multiple subtasks according to the workflow configuration file to obtain the execution results of the task information.

[0083] Step 606 is basically the same as step 204 in the aforementioned embodiment. For the specific implementation method, please refer to the aforementioned description of step 204, which will not be repeated here.

[0084] from Figure 6 It can be seen from this that, with Figure 2 Compared to the corresponding embodiments, the task execution method in this embodiment emphasizes the step of generating a workflow configuration file. By converting the task dependency graph into executable workflow nodes and scheduling rules, and then integrating them to generate a standardized configuration file, the method achieves automatic conversion from logical topology to executable process, improves workflow construction efficiency and standardization, reduces manual configuration costs and error rates, and makes the execution logic clear and controllable.

[0085] Continue to refer to Figure 7 , Figure 7 A flow 700 of a sixth embodiment of a task execution method according to the present disclosure is shown. The task execution method includes the following steps: Step 701: Perform semantic parsing on the task information and determine multiple sub-tasks based on the parsing results.

[0086] Step 702: Generate a task dependency graph based on the dependencies between multiple subtasks.

[0087] Step 703: Convert each subtask node in the task dependency graph into an executable workflow node according to the preset conversion rules.

[0088] Step 704: Transform the topological relationships between each subtask node into workflow execution rules.

[0089] Steps 701-704 are basically the same as steps 601-604 in the previous embodiment. For specific implementation methods, please refer to the above description of steps 601-604, which will not be repeated here.

[0090] Step 705: Generate data flow paths between workflow nodes based on the data flow direction of each directed edge in the task dependency graph.

[0091] In this embodiment, the execution body of the task execution method (e.g.) Figure 1 The server 105 shown traverses all directed edges in the task dependency graph, extracts the data source, data type, and transmission format between upstream and downstream nodes, and clarifies the data flow direction and content corresponding to each edge. Then, according to the node execution order and dependency relationship, the discrete data flows are connected to form a continuous link. The data aggregation and distribution logic of parallel nodes and branch nodes is sorted out, the data merging rules and routing strategies are determined, and finally a structured data flow path covering the entire workflow and recognizable by the execution engine is generated.

[0092] Step 706: Integrate workflow nodes, data flow paths, and execution rules to generate a workflow configuration file.

[0093] In this embodiment, the aforementioned execution entity will structurally associate workflow node attributes, data flow paths between nodes, and execution scheduling rules according to a preset configuration format, establishing a one-to-one mapping relationship between nodes, data, and rules. The execution entity will perform integrity and consistency checks on the configuration content to ensure that there are no conflicts between data flow and execution order, and that parameters match correctly. After successful verification, the execution entity will encapsulate the configuration in a standard format to generate a workflow configuration file that can be directly loaded and run by the scheduling engine.

[0094] In some optional implementations of this embodiment, step 706 includes: integrating workflow nodes, data flow paths, and execution rules according to preset logical associations to obtain an initial configuration file; in response to determining that the integrity and consistency of the initial configuration file have passed verification, encapsulating the initial configuration file in a standardized format to obtain a structured workflow configuration file.

[0095] In this implementation, if the completeness and consistency of the initial configuration file pass the verification, the execution entity will encapsulate the initial configuration file in a standardized format to obtain a structured workflow configuration file. Specifically, the execution entity will bind and organize workflow node attributes, data flow paths, and execution rules at the field level and hierarchically according to a preset mapping relationship, thereby obtaining a structurally complete initial configuration file. Then, the execution entity will perform a completeness verification on the initial configuration file to ensure that there are no missing or redundant nodes, data flows, or rules. At the same time, the execution entity will also perform a consistency verification to verify that there are no conflicts between the data flow direction and execution order, and that the parameters match correctly. After all verifications pass, the execution entity will encapsulate the configuration file according to the specified coding specifications and format standards, thereby generating a structured workflow configuration file that can be directly parsed and loaded by the execution engine.

[0096] By integrating workflow nodes, data flow paths, and execution rules through preset logic, an initial configuration file is formed. After integrity and consistency checks to eliminate conflicts and omissions, it is then standardized and encapsulated into a structured configuration file, thereby improving the standardization and reliability of configuration and reducing the risk of execution anomalies.

[0097] Step 707: Execute multiple subtasks according to the workflow configuration file to obtain the execution results of the task information.

[0098] Step 707 is basically the same as step 204 in the aforementioned embodiment. For the specific implementation method, please refer to the aforementioned description of step 204, which will not be repeated here.

[0099] from Figure 7 It can be seen from this that, with Figure 6 Compared to the corresponding embodiments, the task execution method in this embodiment emphasizes the step of generating data flow paths, thereby automatically generating data flow paths based on the directed edges of the task dependency graph. It integrates nodes, data flows, and execution rules into a configuration file, realizing unified modeling of task logic and data flow, improving configuration accuracy, reducing manual definition errors, and making workflow execution traceable and verifiable.

[0100] Continue to refer to Figure 8 This illustrates a flow 800 of a seventh embodiment of a task execution method according to the present disclosure. The task execution method includes the following steps: Step 801: Perform semantic parsing on the task information and determine multiple sub-tasks based on the parsing results.

[0101] Step 802: Generate a task dependency graph based on the dependencies between multiple subtasks.

[0102] Step 803: Convert each subtask node in the task dependency graph into an executable workflow node according to the preset conversion rules.

[0103] Step 804: Transform the topological relationships between each subtask node into workflow execution rules.

[0104] Step 805: Integrate workflow nodes and execution rules to generate workflow configuration files.

[0105] Steps 801-805 are basically the same as steps 601-605 in the previous embodiment. For specific implementation methods, please refer to the above description of steps 601-605, which will not be repeated here.

[0106] Step 806: Schedule subtask nodes sequentially according to the execution rules in the workflow configuration file.

[0107] In this embodiment, the execution body of the task execution method (e.g.) Figure 1 Clients 101, 102, and 103 (shown) load and parse the workflow configuration file, extracting node dependencies, execution order, and triggering conditions. The aforementioned execution entities prioritize starting nodes without prerequisite dependencies and monitor their execution status in real time. After a preceding node completes and outputs valid data, the execution entities automatically trigger downstream nodes according to the configured rules, sequentially scheduling serial, branch, and convergence nodes to ensure stable and orderly execution according to preset logic.

[0108] Step 807: Integrate the output results of each subtask node to obtain the execution result of the task information.

[0109] In this embodiment, the aforementioned execution entity collects valid output data from all subtask nodes according to the result aggregation rules and data structure requirements in the workflow configuration file. The execution entity also verifies, deduplicates, and standardizes the format of the output results, and handles the result merging logic for branches and parallel nodes. Afterward, the execution entity assembles and structures the results according to the original task objectives, thereby obtaining a complete, compliant, and directly usable final execution result.

[0110] By scheduling subtask nodes in an orderly manner according to the rules in the configuration file, ensuring that the execution order strictly matches the data flow, and then uniformly integrating the output results, the entire process is automated, which improves the stability, consistency and accuracy of task execution, reduces manual intervention, ensures the completeness and reliability of the final result, and thus improves the overall task processing efficiency.

[0111] In some optional implementations of this embodiment, step 807 includes: integrating the output results of each subtask node according to the logical relationship of the subtask nodes to obtain an initial result; verifying and standardizing the initial result to obtain the processed execution result.

[0112] In this implementation, the aforementioned execution entity integrates the outputs of each subtask node according to their logical relationships to obtain an initial result. Then, the execution entity verifies and standardizes the initial result to obtain a processed execution result. Specifically, based on the execution logic and data dependencies between subtasks, the execution entity systematically collects, merges, and concatenates the outputs of each node to form the initial result. The initial result is then verified for completeness and accuracy, and missing, conflicting, and abnormal data is checked. Furthermore, it undergoes format standardization, field regularization, and content normalization to ultimately generate a reliable and usable standardized execution result.

[0113] By organizing the output in an orderly manner according to the logical relationship of subtasks, an initial result is formed. Then, through verification and standardization, anomalies are eliminated and the format is unified, thereby ensuring that the execution result is complete, accurate, standardized and usable, avoiding data chaos and errors, and improving the reliability and readability of the final output.

[0114] Step 808: Monitor the execution status of each subtask node.

[0115] In this embodiment, the aforementioned execution entity collects and aggregates the operation logs, execution duration, return codes, and status identifiers of each node in real time. Then, a heartbeat detection mechanism is established to sense the node's operational activity. The execution entity intelligently identifies abnormal scenarios such as interruptions, timeouts, and errors, triggering early warnings and automatic retry processes. This allows for a visual dashboard that intuitively presents the node topology status, supporting multi-dimensional traceability and problem localization, ensuring full-link visibility, controllability, and traceability.

[0116] Step 809: In response to determining that the execution status is abnormal, perform exception handling according to the exception handling rules in the workflow configuration file.

[0117] In this embodiment, if the execution status is determined to be abnormal, the aforementioned execution entity will identify abnormal types such as subtask node timeouts, execution failures, and data anomalies in real time, and match them with the preset abnormal handling strategies in the workflow configuration. The aforementioned execution entity will perform operations such as automatic retries, skipping abnormal nodes, triggering branch rollbacks, or terminating the process according to the rules. At the same time, it will record abnormal details and processing logs, synchronously update the process status, and output abnormal information.

[0118] from Figure 8 It can be seen from this that, with Figure 6Compared with the corresponding embodiments, the task execution method in this embodiment emphasizes the steps of generating execution results and handling exceptions. By monitoring the execution status of sub-task nodes in real time, exceptions are identified in a timely manner and automatically handled according to the configured rules, thus achieving full-link manageability and controllability of the process. This improves the stability and fault tolerance of task execution, reduces manual intervention, lowers the risk of process interruption caused by exceptions, and ensures reliable workflow closure and result integrity.

[0119] Continue to refer to Figure 9 , Figure 9 A system flow of a task execution method according to the present disclosure is shown, which includes: a task understanding and workflow construction stage (stage one), a structured workflow generation stage (stage two), and an execution control stage based on the structured workflow (stage three). The stages are interconnected and together constitute a complete technical solution from task input to controllable execution.

[0120] Phase 1: Task Understanding and Workflow Construction

[0121] 1. Receive high-level task objectives: The system will receive task objectives input by the user or the upper-level system.

[0122] 2. Task Objective Parsing and Decomposition: The system performs semantic parsing on the received task objectives, breaking down complex tasks into several sub-task units with clear semantic boundaries. This step transforms the unstructured task description into task elements that can be further modeled.

[0123] 3. Subtask Capability Mapping and Constraint Identification: For each subtask obtained from the decomposition, the system identifies the types of intelligent agent capabilities required and the potential tool calls or external resource access needs, and simultaneously identifies the sequential relationships and dependency constraints between subtasks. This step is used to clarify the functional roles of each node in the subsequent workflow.

[0124] 4. Preliminary modeling of task dependencies: Based on the data input-output relationships and logical dependencies between subtasks, the system constructs a preliminary task dependency graph, providing a constraint basis for the generation of structured workflows.

[0125] Phase Two: Structured Workflow Generation

[0126] 1. Generate workflow nodes: The system maps each subtask to an independent execution node in the workflow. Each node corresponds to a clear execution goal, input parameters, and output result description. This step is used to form the basic execution unit of the workflow.

[0127] 2. Determine Node Order and Dependencies: Based on the task dependency modeling results, the system automatically determines the execution order and dependencies between workflow nodes, clarifying which nodes need to wait for the output of the preceding node as input parameters. This step is used to avoid uncertain behavior caused by implicit dependencies during execution.

[0128] 3. Parameter Dependency and Data Flow Modeling: The system explicitly models the parameter transmission relationship between nodes to form a complete data flow path description, so that the workflow not only describes "what to do", but also clarifies "where the data comes from and where it flows to".

[0129] Phase 3: Execution Control Based on Structured Workflow

[0130] 1. Workflow-driven execution scheduling: The system schedules each execution node according to the structured workflow configuration, ensuring that the node is only triggered to execute when its dependency conditions are met.

[0131] 2. Node-level execution and result backfilling: When an agent executes a single workflow node, it processes only the input parameters and execution target defined for that node, and the execution result is backfilled to subsequent nodes according to the predefined data flow relationship.

[0132] 3. Execution Process Control and Exception Handling: When a node fails to execute or outputs an abnormal result, the system can interrupt, retry, or adjust the execution process based on the workflow structure to prevent errors from spreading in the execution chain. This step makes the overall execution process more stable and controllable.

[0133] Further reference Figure 10 As an implementation of the methods shown in the above figures, this disclosure provides an embodiment of a task execution device, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0134] like Figure 10 As shown, the task execution device 1000 of this embodiment includes: a semantic parsing module 1001, a graph generation module 1002, a file generation module 1003, and a task execution module 1004. The semantic parsing module 1001 is configured to perform semantic parsing on task information and determine multiple subtasks based on the parsing results; the graph generation module 1002 is configured to generate a task dependency graph based on the dependencies between the multiple subtasks; the file generation module 1003 is configured to generate a workflow configuration file based on the information of each subtask node in the task dependency graph; and the task execution module 1004 is configured to execute multiple subtasks according to the workflow configuration file to obtain the execution results of the task information.

[0135] In this embodiment, the specific processing of the semantic parsing module 1001, the graph generation module 1002, the file generation block 1003, and the task execution module 1004 in the task execution device 1000, and the resulting technical effects, can be found in reference to [reference needed]. Figure 2 The relevant descriptions of steps 201-204 in the corresponding embodiments will not be repeated here.

[0136] In some optional implementations of this embodiment, the semantic parsing module 1001 is further configured to: preprocess and denoise the task information to obtain standardized information; perform semantic parsing on the standardized information to obtain parsing results; and divide the parsing results into multiple sub-tasks according to preset task decomposition rules.

[0137] In some optional implementations of this embodiment, the graph generation module 1002 includes: an association determination submodule, configured to determine the association relationship between the input and output data of multiple subtasks; a dependency determination submodule, configured to determine the dependency relationship between the execution order of multiple subtasks; and a graph construction submodule, configured to construct a task dependency graph based on the association relationship and the dependency relationship.

[0138] In some optional implementations of this embodiment, the graph construction submodule is further configured to: treat subtasks as nodes in the basic dependency graph; determine the direction of edges in the basic dependency graph based on associations and dependencies; and perform multi-dimensional verification of the logic of each node and edge in the basic dependency graph to obtain the verified task dependency graph.

[0139] In some optional implementations of this embodiment, the file generation block 1003 includes: a node conversion submodule, configured to convert each subtask node in the task dependency graph into an executable workflow node according to a preset conversion rule; a rule conversion submodule, configured to convert the topological relationship between each subtask node into the execution rule of the workflow; and an integration submodule, configured to integrate the workflow nodes and the execution rule to generate a workflow configuration file.

[0140] In some optional implementations of this embodiment, the task execution device 1000 further includes: a path determination submodule, configured to generate data flow paths between workflow nodes based on the data flow direction of each directed edge in the task dependency graph; and an integration submodule including: an integration unit, configured to integrate workflow nodes, data flow paths, and execution rules to generate a workflow configuration file.

[0141] In some optional implementations of this embodiment, the integration unit is further configured to: integrate workflow nodes, data flow paths and execution rules according to preset logical associations to obtain an initial configuration file; in response to determining that the integrity and consistency of the initial configuration file have passed the verification, encapsulate the initial configuration file in a standardized format to obtain a structured workflow configuration file.

[0142] In some optional implementations of this embodiment, the task execution module 1004 includes: a scheduling submodule, configured to schedule subtask nodes sequentially according to the execution rules in the workflow configuration file; and a result generation submodule, configured to integrate the output results of each subtask node to obtain the execution result of the task information.

[0143] In some optional implementations of this embodiment, the result generation submodule is further configured to: integrate the output results of each subtask node according to the logical relationship of the subtask nodes to obtain the initial result; and perform verification and standardization processing on the initial result to obtain the processed execution result.

[0144] In some optional implementations of this embodiment, the task execution device 1000 further includes: a monitoring module configured to monitor the execution status of each sub-task node; and a processing module configured to perform exception processing according to the exception handling rules in the workflow configuration file in response to determining that the execution status is an abnormal status.

[0145] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0146] Figure 11 A schematic block diagram of an example electronic device 1100 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0147] Figure 11A schematic block diagram of an example electronic device 1100 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0148] like Figure 11 As shown, device 1100 includes a computing unit 1101, which can perform various appropriate actions and processes according to a computer program stored in read-only memory (ROM) 1102 or a computer program loaded into random access memory (RAM) 1103 from storage unit 1108. The RAM 1103 may also store various programs and data required for the operation of device 1100. The computing unit 1101, ROM 1102, and RAM 1103 are interconnected via bus 1104. Input / output (I / O) interface 1105 is also connected to bus 1104.

[0149] Multiple components in device 1100 are connected to I / O interface 1105, including: input unit 1106, such as keyboard, mouse, etc.; output unit 1107, such as various types of monitors, speakers, etc.; storage unit 1108, such as disk, optical disk, etc.; and communication unit 1109, such as network card, modem, wireless transceiver, etc. Communication unit 1109 allows device 1100 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0150] The computing unit 1101 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1101 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1101 performs the various methods and processes described above, such as task execution methods. For example, in some embodiments, the task execution method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 1108. In some embodiments, part or all of the computer program may be loaded and / or installed on device 1100 via ROM 1102 and / or communication unit 1109. When the computer program is loaded into RAM 1103 and executed by the computing unit 1101, one or more steps of the task execution method described above may be performed. Alternatively, in other embodiments, the computing unit 1101 may be configured to perform the task execution method by any other suitable means (e.g., by means of firmware).

[0151] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0152] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0153] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0154] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0155] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0156] Cloud computing refers to a technological system that enables access to elastic and scalable shared physical or virtual resources via a network. These resources can include servers, operating systems, networks, software, and storage devices, and can be deployed and managed in an on-demand, self-service manner. Cloud computing technology can provide efficient and powerful data processing capabilities for applications such as artificial intelligence and blockchain, as well as for model training.

[0157] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and Virtual Private Server (VPS) services, such as high management difficulty and weak business scalability.

[0158] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0159] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A task execution method, comprising: Perform semantic parsing on the task information, and determine multiple sub-tasks based on the parsing results; Generate a task dependency graph based on the dependencies between the multiple subtasks; Generate a workflow configuration file based on the information of each subtask node in the task dependency graph; The multiple subtasks are executed according to the workflow configuration file to obtain the execution results of the task information.

2. The method according to claim 1, wherein, The process involves semantic parsing of the task information and determining multiple sub-tasks based on the parsing results, including: The task information is preprocessed and noise reduction is performed to obtain standardized information; The standardized information is semantically parsed to obtain the parsing results; The parsing results are divided into multiple sub-tasks according to the preset task decomposition rules.

3. The method according to claim 1, wherein, The step of generating a task dependency graph based on the dependencies between the multiple subtasks includes: Determine the correlation between the input and output data of the multiple subtasks; Determine the dependencies between the execution order of the multiple subtasks; The task dependency graph is constructed based on the associations and dependencies.

4. The method according to claim 3, wherein, The step of constructing the task dependency graph based on the association and the dependency includes: Treat the subtasks as nodes in the base dependency graph; The direction of the edges in the basic dependency graph is determined based on the association and dependency relationships; The logic of each node and edge in the basic dependency graph is validated in multiple dimensions to obtain the validated task dependency graph.

5. The method according to claim 1, wherein, The step of generating a workflow configuration file based on the information of each subtask node in the task dependency graph includes: According to the preset transformation rules, each subtask node in the task dependency graph is transformed into an executable workflow node; The topological relationships between the sub-task nodes are transformed into workflow execution rules; The workflow nodes and execution rules are integrated to generate the workflow configuration file.

6. The method according to claim 5, further comprising: Data flow paths between workflow nodes are generated based on the data flow direction of each directed edge in the task dependency graph; as well as The process of integrating the workflow nodes and the execution rules to generate the workflow configuration file includes: The workflow node, the data flow path, and the execution rule are integrated to generate the workflow configuration file.

7. The method according to claim 6, wherein, The process of integrating the workflow nodes, the data flow paths, and the execution rules to generate the workflow configuration file includes: The workflow nodes, data flow paths, and execution rules are integrated according to a preset logical association to obtain an initial configuration file; In response to the determination that the integrity and consistency of the initial configuration file have passed the verification, the initial configuration file is encapsulated in a standardized format to obtain a structured workflow configuration file.

8. The method according to claim 5, wherein, The execution result of executing the multiple subtasks according to the workflow configuration file to obtain the task information includes: The subtask nodes are scheduled sequentially according to the execution rules in the workflow configuration file; The output results of each subtask node are integrated to obtain the execution result of the task information.

9. The method according to claim 8, wherein, The execution result of integrating the outputs of each subtask node to obtain the task information includes: Based on the logical relationship between the sub-task nodes, the output results of each sub-task node are integrated to obtain the initial result; The initial result is verified and standardized to obtain the processed execution result.

10. The method according to any one of claims 1-9, further comprising: Monitor the execution status of each subtask node; In response to determining that the execution state is an abnormal state, the abnormality is handled in accordance with the abnormality handling rules in the workflow configuration file.

11. A task execution device, comprising: The semantic parsing module is configured to perform semantic parsing on task information and determine multiple subtasks based on the parsing results. The graph generation module is configured to generate a task dependency graph based on the dependencies between the multiple subtasks; The file generation module is configured to generate a workflow configuration file based on the information of each subtask node in the task dependency graph. The task execution module is configured to execute the multiple subtasks according to the workflow configuration file and obtain the execution result of the task information.

12. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-10.

13. A non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-10.

14. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-10.

Images