A method and apparatus for assisted generation of a workflow

Abstract

This invention discloses an auxiliary method and apparatus for workflow generation. By receiving a query request from a user terminal, including a user question and pending data, a structured query language is determined based on the query request. A large model is then used to obtain various task nodes based on the structured query language. The corresponding workflow generation agent is invoked according to the task type of each task node to generate a logical execution subgraph for each task node. A workflow topology is generated based on the logical execution subgraph and sent to the user terminal. Therefore, workflow topologies can be automatically generated based on user input, reducing the professional skills required of operators, thereby lowering the barrier to workflow design, shortening the cycle time, improving the efficiency and rapid response capability of workflow design, and enhancing the consistency and accuracy of the workflow.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method and apparatus for assisting in the generation of workflows. Background Technology

[0002] With the rapid development of enterprise digital transformation and artificial intelligence technology, many scenarios require the establishment of workflows to achieve efficient collaboration and automated execution of tasks.

[0003] Currently, the construction and parameter settings of related workflows mainly rely on manual processes. Common existing technologies include manual configuration based on fixed templates, or relying on domain experts to define and orchestrate workflow nodes, rules, and execution conditions one by one based on experience.

[0004] However, manual setup is inefficient and cannot quickly respond to changing scenario requirements. Moreover, manual setup requires operators to have high professional skills, resulting in high barriers to workflow design and long cycles. At the same time, it is difficult to guarantee the consistency and accuracy of configuration. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a workflow auxiliary generation method and apparatus that can automatically generate workflow topology based on user input, reduce the professional skills required of operators, thereby lowering the threshold for workflow design, shortening the cycle, improving the efficiency and rapid response capability of workflow design, and improving the consistency and accuracy of workflow.

[0006] In a first aspect, embodiments of the present invention provide an auxiliary method for generating workflows, the method comprising: Receive a query request sent by a user terminal, the query request including user questions and data to be processed; Determine the structured query language based on the query request; Each task node is obtained through the large model using the structured query language. Based on the node type of each task node, the corresponding workflow is invoked to generate an intelligent agent, thereby generating a logical execution subgraph corresponding to each task node. Generate workflow topology based on the logic execution subgraph; The workflow topology is sent to the user terminal.

[0007] In some embodiments, the data to be processed is raw data or a structured query language; The step of determining the structured query language based on the query request includes: In response to the fact that the data to be processed is raw data, the structured query language is generated based on the user question and the data to be processed; In response to the fact that the data to be processed is a Structured Query Language (SCL), the data to be processed is used as the SCL.

[0008] In some embodiments, the step of invoking the corresponding workflow generation agent according to the node type of each task node to generate a logical execution subgraph for each task node includes: Based on the node type of each task node, the corresponding workflow generation agent is invoked in parallel to generate the logical execution subgraph corresponding to each task node.

[0009] In some embodiments, the node type includes one or more of the following: data cleaning node, grouping and aggregation node, table join node, condition judgment node, calculated column node, and merge table node.

[0010] Secondly, embodiments of the present invention provide an auxiliary method for generating workflows, the method comprising: Obtain a query request input by the user, the query request including the user's question and the data to be processed; The query request is sent to the server, which is used to determine the structured query language based on the query request, obtain each task node through the large model based on the structured query language, call the corresponding workflow to generate an intelligent agent according to the node type of each task node, generate a logical execution subgraph corresponding to each task node, and generate a workflow topology based on the logical execution subgraph. The workflow topology is displayed through a graphical interface.

[0011] In some embodiments, the node type includes one or more of the following: data cleaning node, grouping and aggregation node, table join node, condition judgment node, calculated column node, and merge table node.

[0012] In some embodiments, the display interface includes a human-computer interaction area, a workflow display area, an operation area, a data area, and a preview area; The human-computer interaction area is used to obtain query information input by the user; The workflow display area is used to display the workflow topology; The operation area is used to operate on the workflow topology; The data area is used to display data; The preview area is used to display preview information.

[0013] In some embodiments, the workflow topology includes logical execution subgraphs corresponding to multiple task nodes and the connection relationships between the logical execution subgraphs; The method further includes: In response to the selection of the logic execution sub-graph, the corresponding configuration information and operation controls are displayed in the operation area; In response to receiving an operation instruction input by the user through the operation control, the operation instruction is executed.

[0014] In some embodiments, the method further includes: In response to the selection of the logic execution subgraph, corresponding preview information is displayed in the preview area, and the preview information includes fields.

[0015] In some embodiments, the method further includes: In response to receiving a run command from the user, the workflow corresponding to the workflow topology is executed.

[0016] Thirdly, embodiments of the present invention provide a workflow auxiliary generation device, the device comprising: A query request receiving unit is used to receive query requests sent by user terminals, wherein the query request includes user questions and data to be processed; A structured query language determination unit is used to determine the structured query language based on the query request; The task node acquisition unit is used to acquire each task node through the large model based on the structured query language. The logic execution subgraph generation unit is used to call the corresponding workflow generation agent according to the node type of each task node to generate the logic execution subgraph corresponding to each task node. A workflow topology generation unit is used to generate a workflow topology based on the logical execution subgraph. A workflow topology sending unit is used to send the workflow topology to the user terminal.

[0017] Fourthly, embodiments of the present invention provide a workflow auxiliary generation device, the device comprising: The query request acquisition unit is used to acquire the query request input by the user, which includes the user's question and the data to be processed. A query request sending unit is used to send the query request to the server. The server is used to determine a structured query language based on the query request, obtain each task node through the large model based on the structured query language, call the corresponding workflow to generate an intelligent agent according to the node type of each task node, generate a logical execution subgraph corresponding to each task node, and generate a workflow topology based on the logical execution subgraph. The workflow topology display unit is used to display the workflow topology through a display interface.

[0018] Fifthly, embodiments of the present invention provide an electronic device, including a memory and a processor, wherein the memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the methods as described in the first and second aspects.

[0019] In a sixth aspect, embodiments of the present invention provide a computer-readable storage medium having stored computer program instructions thereon, which, when executed by a processor, implement the methods described in the first and second aspects.

[0020] The technical solution of this invention receives a query request sent by a user terminal. The query request includes a user question and data to be processed. Based on the query request, a structured query language is determined. A large model is used to obtain various task nodes based on the structured query language. According to the task type of each task node, the corresponding workflow generation agent is invoked to generate a logical execution subgraph for each task node. Based on the logical execution subgraph, a workflow topology is generated and sent to the user terminal. Therefore, a workflow topology can be automatically generated based on user input, reducing the professional skills required of operators, thereby lowering the threshold for workflow design, shortening the cycle, improving the efficiency and rapid response capability of workflow design, and enhancing the consistency and accuracy of the workflow. Attached Figure Description

[0021] The above and other objects, features and advantages of the present invention will become clearer from the following description of embodiments of the invention with reference to the accompanying drawings, in which: Figure 1 This is a schematic diagram of a workflow auxiliary generation system according to an embodiment of the present invention; Figure 2 This is a flowchart of data interaction according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the display interface according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the display interface according to another embodiment of the present invention; Figure 5 This is a flowchart of data interaction according to another embodiment of the present invention; Figure 6 This is a flowchart of an auxiliary method for generating server workflow according to an embodiment of the present invention; Figure 7 This is a flowchart of an auxiliary method for generating the workflow of a user terminal according to an embodiment of the present invention; Figure 8 This is a schematic diagram of an auxiliary device for generating the server workflow according to an embodiment of the present invention; Figure 9 This is a schematic diagram of an auxiliary device for generating the workflow of a user terminal according to an embodiment of the present invention; Figure 10 This is a schematic diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0022] The present application is described below based on embodiments, but it is not limited to these embodiments. In the detailed description of the present application below, certain specific details are described in detail. Those skilled in the art can fully understand the present application without these details. To avoid obscuring the substance of the present application, well-known methods, processes, flows, elements, and circuits are not described in detail.

[0023] Furthermore, those skilled in the art should understand that the accompanying drawings provided herein are for illustrative purposes only and are not necessarily drawn to scale.

[0024] Unless the context explicitly requires it, words such as "including" or "contains" throughout the application should be interpreted as including rather than exclusive or exhaustive; that is, meaning "including but not limited to".

[0025] In the description of this application, it should be understood that the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0026] The solutions described in this specification and embodiments, if involving the processing of personal information, will be processed only on the premise of having a legal basis (such as obtaining the consent of the personal information subject, or being necessary for the performance of a contract), and will only be processed within the scope stipulated or agreed upon. A user's refusal to process personal information beyond what is necessary for basic functions will not affect the user's use of basic functions.

[0027] Figure 1 This is a schematic diagram of a workflow auxiliary generation system according to an embodiment of the present invention. Figure 1 As shown, the workflow auxiliary generation system of this embodiment includes a user terminal 1, a server 2, and a network 3. The user terminal 1 and the server 2 are connected via the network 3.

[0028] User terminal 1 is a terminal device used by the user, which can access the workflow generation tool through an APP, mini-program, or webpage, and then generate the required workflow through the workflow generation tool. User terminal 1 can be a mobile phone, tablet, laptop, desktop computer, or other dedicated data processing terminal.

[0029] Server 2 is used to provide workflow generation services, and it can be a single server or a server cluster consisting of multiple servers.

[0030] Network 3 is used for the exchange of information and / or data between user terminal 1 and server 2. Network 3 can be any type of wired or wireless network. In some embodiments, network 3 may include a wired network, wireless network, fiber optic network, telecommunications network, intranet, Internet, local area network (LAN), wide area network (WAN), wireless local area network (WLAN), metropolitan area network (MAN), wide area network (WAN), public switched telephone network (PSTN), Bluetooth network, ZigBee network, or near field communication (NFC) network, or any combination thereof.

[0031] Specifically, when a user needs to generate a workflow, they access the workflow generation tool through an app, mini-program, or webpage, input a query request through the display interface provided by the workflow generation tool, and the user terminal sends the query request to the server. The server generates a workflow topology based on the query request and sends the workflow topology to the user terminal, thereby completing the workflow generation.

[0032] Figure 2 This is a flowchart of data interaction according to an embodiment of the present invention. Figure 2 The data interaction between the user terminal and the server is illustrated, specifically including the following steps: Step S101: Enter the display interface.

[0033] In this embodiment, the user uses a user terminal to access the display interface through a workflow generation tool provided by the server.

[0034] Figure 3 This is a schematic diagram of a display interface according to an embodiment of the present invention. For example... Figure 3 As shown, the display interface includes a workflow display area A, a human-computer interaction area B, a data area C, and an extended area D.

[0035] The workflow display area A is used to display the workflow topology. Specifically, the workflow display area A includes a first control sub-area A1, a second control sub-area A2, and a workflow display sub-area A3.

[0036] The first control sub-area A1 includes controls for workflow management, such as "Run," "Save," and "Publish." The "Run" control triggers the start of the current workflow; the "Save" control saves the current workflow for later editing or use; and the "Publish" control publishes the current workflow, allowing other authorized personnel to use it.

[0037] The second control sub-area A2 includes controls for workflow editing, such as ODPS (Open Data Processing Service) input tables, output tables, data cleaning, grouping and aggregation, table joins, and table merging.

[0038] The ODPS input table serves as a data source, specifying the source table to be read by subsequent workflow nodes.

[0039] The output table control allows users to edit the output table.

[0040] For controls such as data cleaning, grouping and aggregation, table joining, and table merging, users can drag and drop these controls to add corresponding task nodes in the workflow display sub-area A3.

[0041] The workflow display sub-area A3 is used to display the workflow topology.

[0042] The human-computer interaction area B is used to obtain query information input by the user. Specifically, the human-computer interaction area B includes a historical dialogue display sub-area B1 and an input box B2.

[0043] Users can enter a query request through input box B2. The query request includes the user's question and the data to be processed. To add data to be processed, users can drag data (such as tables, files, etc.) from data area C to input box B2, or users can enter the data to be processed into input box B2.

[0044] The historical dialogue display sub-area B1 is used to display information from the current historical dialogue, including user-inputted queries and server-returned responses. In some embodiments, the historical dialogue display sub-area B1 may also display prompts to guide the user through the operation process.

[0045] The data to be processed can be raw data or Structured Query Language (SQL). The raw data can be data from tables or other files. The Structured Query Language is SQL pre-edited by the user.

[0046] For example, if a user needs to link tables x1 and x2, they can drag these two tables to the workflow display sub-area A3, and then enter "link tables x1 and x2" in the input box. Here, the two tables are the data to be processed, and "link tables x1 and x2" is the user's question.

[0047] For example, when a user has finished editing the SQL and needs to convert it into a workflow, they can copy the SQL into the input box and enter "Convert this SQL into a workflow". In this case, the entered SQL represents the data to be processed, and "Convert this SQL into a workflow" represents the user's question.

[0048] Data area C is used to display data. Specifically, data that the user has processed in the past can be displayed in data area C for easy viewing. Users can also manually add data to data area C; the added data can be local or cloud-based.

[0049] Extended area D allows users to implement corresponding functions based on the actual workflow. For example, after a workflow topology is generated, if the logical execution subgraph in the workflow topology is selected by the user, extended area D can serve as both the operation area and the preview area.

[0050] Step S102: Obtain the query request.

[0051] In this embodiment, the user terminal obtains a query request input by the user, which includes the user's question and the data to be processed.

[0052] Step S103: Send a query request.

[0053] In this embodiment, the user terminal sends the query request to the server.

[0054] Step S104: Determine the structured query information.

[0055] In this embodiment, after receiving a query request from a user terminal, the server determines a structured query language based on the query request. Specifically, if the data to be processed is raw data, the structured query language is generated based on the user's question and the data to be processed. If the data to be processed is in the structured query language, the data to be processed is used as the structured query language.

[0056] To generate the structured query language based on the user's question and the data to be processed, the server uses a Large Language Model (LLM) to parse the user-input data (such as table structure) and the user's question (natural language question) to understand the user's query intent. For the data to be processed, the LLM can identify table names, field names, and data types (such as numeric, text, and date). For the user's question, the LLM can extract the user's query objective, such as performing statistics on xx data, filtering xx data, or merging xx data; simultaneously, the LLM can also extract and identify conditions (such as time range, status filtering), aggregation operations (such as summation, counting), sorting or grouping requirements, etc. Then, the vocabulary in the natural language is mapped to database concepts. The understood query intent is converted into an SQL logical structure. For example, first, the query type is determined, such as SELECT, UPDATE, INSERT, or DELETE, etc., and then a query logic tree is constructed, such as selecting fields, filtering conditions, aggregation and grouping, sorting and limiting, etc. Finally, the logical structure is converted into SQL statements that conform to database syntax.

[0057] Step S105: Obtain the task node.

[0058] In this embodiment, after obtaining the structured query language, a large model is used to obtain each task node based on the structured query language. Specifically, the large model analyzes the obtained structured query language to identify the logical execution units within it, divides the task nodes based on these logical execution units, and determines the node type of each task node. The node types include one or more of the following: data cleaning nodes, grouping and aggregation nodes, table join nodes, conditional judgment nodes, calculated column nodes, and merged table nodes.

[0059] Specifically, as described above, the large model can identify logical execution units in Structured Query Language (SCL) and divide task nodes based on these units. For each identified task node, the node type can be determined by identifying its keywords. For example, if the logical execution unit corresponding to the task node includes "cleaning," the node type is a data cleaning node; if it includes "join," the node type is a table join node; if it includes "group_by," the node type is a grouping aggregation node; if it includes "cond_value," the node type is a conditional judgment node; if it includes "compute," the node type is a computed column node; and if it includes "union," the node type is a merge table node.

[0060] Step S106: Generate the logical execution sub-graph for each task node.

[0061] In this embodiment, after obtaining the node type of each task node, the corresponding workflow generation agent is invoked according to the node type of each task node to generate the logical execution subgraph corresponding to each task node. Specifically, the corresponding workflow generation agent is invoked in parallel according to the node type of each task node to generate the logical execution subgraph corresponding to each task node. That is to say, this embodiment of the invention pre-configures one or more workflow generation agents for each node type, and for each node, the workflow generation agent corresponding to the node type is invoked in parallel to generate the corresponding logical execution subgraph. By using a parallel approach, the efficiency of workflow generation can be improved.

[0062] Furthermore, when the large model identifies logical execution units in the structured query language and divides task nodes based on these units, it needs to generate sub-fragments of the structured query language corresponding to each task node. These sub-fragments are then provided to the workflow generation agent, enabling the agent to obtain the corresponding logical execution subgraph based on the sub-fragments. When generating sub-fragments for each task node, in addition to the task node's own information, it also needs to acquire information about its dependencies. That is, the large language model analyzes the dependencies between task nodes, acquires information about other task nodes that depend on the target task node when generating the sub-fragment for the target task node, extracts key information from these other task nodes, combines this key information with the target task node's own structured query language to form a sub-fragment, and then calls the workflow generation agent to generate the corresponding logical execution subgraph based on the sub-fragment. The key information about these other task nodes is the information needed to generate the logical execution subgraph for the target task node.

[0063] For example, for two adjacent task nodes in the same structured query language, if the parent task node generates a new field, the new field needs to be added to the sub-fragment when generating the corresponding sub-fragment of the child task node.

[0064] While existing technologies also utilize agents to generate workflow topologies, these typically involve the workflow generation agent directly generating the topology based on a complete structured query language. This approach results in a large number of tokens and lengthy suggestions within the large model, impacting accuracy and processing efficiency. In contrast, this invention pre-obtains the dependencies between nodes before inputting them into the workflow generation agent and generates sub-fragments. This reduces the number of tokens processed by the large model and results in shorter suggestions, improving both accuracy and efficiency. Furthermore, it allows for parallel execution, significantly enhancing the efficiency of workflow generation.

[0065] Step S107: Generate workflow topology.

[0066] In this embodiment, multiple logical execution subgraphs are connected according to the logical dependencies between them to obtain a workflow topology, wherein the workflow topology is a data processing flowchart.

[0067] Step S108: Send workflow topology.

[0068] In this embodiment, the server sends the generated workflow topology to the user terminal.

[0069] Step S109: Display the workflow topology.

[0070] In this embodiment, the user terminal displays the workflow topology through a display interface.

[0071] Figure 4 This is a schematic diagram of the display interface according to another embodiment of the present invention. It is related to... Figure 3 The main difference is: Figure 3 The display interface shown is the interface before the workflow topology is generated. Figure 4 The display interface is the one shown after the workflow topology is generated. The main difference between the two is the expanded area of ​​the workflow display sub-region A3; other parts are basically the same, and will not be described again in this embodiment of the invention.

[0072] exist Figure 4 In the illustrated embodiment, the workflow display sub-area A3 shows the workflow topology, such as... Figure 4 As shown, the workflow topology includes multiple logical execution subgraphs and the connections between them. Figure 4 The diagram shows three logical execution subgraphs in the workflow topology: data source subgraph P1, data cleaning subgraph P2, and grouping and aggregation subgraph P3.

[0073] Furthermore, in response to the selection of the logical execution subgraph, the corresponding configuration information and operation controls are displayed in the operation area E. The operation area E includes a step display sub-area E1 and an operation sub-area E2. Step display sub-area E1 displays the execution steps corresponding to the selected logical execution subgraph. Operation sub-area E2 displays the operation controls for the selected execution steps. For example, assuming the logical execution subgraph for data cleaning is selected, the execution steps for data cleaning are displayed in step display sub-area E1. Figure 4 The following example illustrates a single execution step (data filtering). When the execution step (data filtering) is selected, corresponding operation controls are displayed. Through these controls, users can adjust the data filtering rules, such as adding new rules, deleting existing rules, and changing the logical relationships (OR, AND, etc.) between rules. The user terminal responds to receiving the operation command input by the user through the operation controls by executing the operation command. Executing the operation command refers to updating the logical execution subgraph according to the user-input configuration.

[0074] Furthermore, in response to the selection of the logic execution subgraph, corresponding preview information is displayed in the preview area F, the preview information including fields.

[0075] Furthermore, after the workflow is generated, the user can click the run control. The user terminal responds by receiving the run command input by the user, executes the workflow corresponding to the workflow topology, and generates the corresponding execution result.

[0076] This invention receives query requests from user terminals, including user questions and pending data. Based on the query requests, a structured query language is determined. A large model then uses this structured query language to obtain various task nodes. Based on the task type of each task node, the corresponding workflow generation agent is invoked to generate a logical execution subgraph for each task node. A workflow topology is then generated based on the logical execution subgraph and sent to the user terminal. This allows for automatic generation of workflow topologies based on user input, reducing the professional skills required of operators, lowering the barrier to workflow design, shortening the cycle time, improving workflow design efficiency and responsiveness, and enhancing workflow consistency and accuracy.

[0077] In some embodiments, the server can be further divided into a message management module, an SQL generation agent, and a workflow generation agent.

[0078] The message management module is used to manage the receiving, sending, and forwarding of messages.

[0079] The SQL generation agent is used to generate structured query language (SQL) based on the user's natural language. Specifically, it possesses NLP2SQL (Natural Language to SQL) capabilities. NLP2SQL refers to the ability to convert human-generated questions in natural language into structured query language (SQL) using natural language processing technology. This lowers the barrier to database queries, allowing non-expert users to easily access data. The SQL generation agent can parse the user's query intent to complete tasks such as entity recognition (table names, field names, condition values), intent classification (query type: selection, aggregation, join, etc.), and semantic parsing. It then generates the SQL structure, including SELECT clauses (determining query fields), WHERE clauses (condition extraction), JOIN clauses (table join identification), and GROUP BY / ORDER BY clauses (aggregation and sorting identification).

[0080] The workflow generation agent is used to generate logical execution subgraphs based on Structured Query Language (SQL). Specifically, the workflow generation agent has SQL2Flow (SQL to Flow, Structured Query Language to Data Flow Diagram) capabilities. SQL2Flow is the ability to automatically convert SQL code into data flow diagrams, improving SQL readability, maintainability, and team collaboration efficiency. It can transform text code (such as multi-table JOINs, subqueries, and CTEs) into intuitive graphics, clearly showing the source, transformation steps, and destination of data.

[0081] Figure 5 This is a flowchart of data interaction according to another embodiment of the present invention. Figure 5 The data interaction between the user terminal, message management module, SQL generation agent, and workflow generation agent is illustrated, specifically including the following steps: Step S201: Enter the display interface.

[0082] Step S202: Obtain the query request.

[0083] The steps S201-S202 described above can be referred to as steps S101-S102 described above, and will not be repeated here in the embodiments of the present invention.

[0084] Step S203: Send a query request.

[0085] In this embodiment, the user terminal sends a query request to the server, and the server's message management module receives the query request.

[0086] Step S204: Forward the query request.

[0087] In this embodiment, the message management module forwards the query request to the SQL generation agent according to the predetermined message management rules.

[0088] Step S205: Determine the structured query information.

[0089] In this embodiment, after receiving a query request from a user terminal, the SQL generation agent determines a structured query language based on the query request. Specifically, if the data to be processed is raw data, the structured query language is generated based on the user's question and the data to be processed. If the data to be processed is in the structured query language, the data to be processed is used as the structured query language.

[0090] To generate the structured query language based on the user question and the data to be processed, the SQL generation agent uses NLP2SQL (Natural Language to SQL) capability to convert the query information into the structured query language SQL.

[0091] Step S206: Obtain the task node.

[0092] In this embodiment, the SQL generation agent obtains each task node through the large model based on the structured query language, and determines the node type of each task node.

[0093] Step S207: Invoke the workflow to generate an intelligent agent.

[0094] In this embodiment, one or more workflow generation agents are pre-configured for each node type. The SQL generation agent calls the corresponding workflow generation agent in parallel according to the node type of each task node.

[0095] Step S208: Return to the logic execution subgraph.

[0096] In this embodiment, the SQL generation agent calls the workflow generation agent to generate the corresponding logical execution subgraph based on the logical execution unit corresponding to the task node, and sends the logical execution subgraph to the SQL generation agent.

[0097] Step S209: Generate workflow topology.

[0098] In this embodiment, the SQL generation agent connects multiple logical execution subgraphs according to the logical dependencies between them to obtain a workflow topology, wherein the workflow topology is a data processing flowchart.

[0099] Step S210: Send workflow topology.

[0100] In this embodiment, the SQL generation agent sends the generated workflow topology to the message management module.

[0101] Step S211: Send workflow topology.

[0102] In this embodiment, the message management module forwards the workflow topology to the user terminal.

[0103] Step S212: Display the workflow topology.

[0104] In this embodiment, the user terminal displays the workflow topology through a display interface.

[0105] This invention receives query requests from user terminals, including user questions and pending data. Based on the query requests, a structured query language is determined. A large model then uses this structured query language to obtain various task nodes. Based on the task type of each task node, the corresponding workflow generation agent is invoked to generate a logical execution subgraph for each task node. A workflow topology is then generated based on the logical execution subgraph and sent to the user terminal. This allows for automatic generation of workflow topologies based on user input, reducing the professional skills required of operators, lowering the barrier to workflow design, shortening the cycle time, improving workflow design efficiency and responsiveness, and enhancing workflow consistency and accuracy.

[0106] Figure 6 This is a flowchart of an auxiliary method for generating server workflows according to an embodiment of the present invention. For example... Figure 6 As shown, the auxiliary generation method for server workflow in this embodiment of the invention includes the following steps: Step S310: Receive a query request sent by the user terminal, the query request including user questions and data to be processed.

[0107] Step S320: Determine the structured query language based on the query request.

[0108] Step S330: Obtain each task node through the large model using the structured query language.

[0109] Step S340: Based on the node type of each task node, call the corresponding workflow to generate an intelligent agent to generate a logical execution subgraph for each task node.

[0110] Step S350: Generate workflow topology based on the logic execution subgraph.

[0111] Step S360: Send the workflow topology to the user terminal.

[0112] In some embodiments, the data to be processed is raw data or a structured query language; The step of determining the structured query language based on the query request includes: In response to the fact that the data to be processed is raw data, the structured query language is generated based on the user question and the data to be processed; In response to the fact that the data to be processed is a Structured Query Language (SCL), the data to be processed is used as the SCL.

[0113] In some embodiments, the step of invoking the corresponding workflow generation agent according to the node type of each task node to generate a logical execution subgraph for each task node includes: Based on the node type of each task node, the corresponding workflow generation agent is invoked in parallel to generate the logical execution subgraph corresponding to each task node.

[0114] In some embodiments, the node type includes one or more of the following: data cleaning node, grouping and aggregation node, table join node, condition judgment node, calculated column node, and merge table node.

[0115] This invention receives query requests from user terminals, including user questions and pending data. Based on the query requests, a structured query language is determined. A large model then uses this structured query language to obtain various task nodes. Based on the task type of each task node, the corresponding workflow generation agent is invoked to generate a logical execution subgraph for each task node. A workflow topology is then generated based on the logical execution subgraph and sent to the user terminal. This allows for automatic generation of workflow topologies based on user input, reducing the professional skills required of operators, lowering the barrier to workflow design, shortening the cycle time, improving workflow design efficiency and responsiveness, and enhancing workflow consistency and accuracy.

[0116] Figure 7 This is a flowchart of an auxiliary method for generating the workflow of a user terminal according to an embodiment of the present invention. For example... Figure 7 As shown, the workflow auxiliary generation method for user terminals in this embodiment of the invention includes the following steps: Step S410: Obtain the query request input by the user, which includes the user's question and the data to be processed.

[0117] Step S420: Send the query request to the server. The server is used to determine the structured query language according to the query request, obtain each task node through the large model according to the structured query language, call the corresponding workflow to generate an intelligent agent according to the node type of each task node, generate a logical execution subgraph corresponding to each task node, and generate a workflow topology according to the logical execution subgraph.

[0118] Step S430: Display the workflow topology through the display interface.

[0119] In some embodiments, the node type includes one or more of the following: data cleaning node, grouping and aggregation node, table join node, condition judgment node, calculated column node, and merge table node.

[0120] In some embodiments, the display interface includes a human-computer interaction area, a workflow display area, an operation area, a data area, and a preview area; The human-computer interaction area is used to obtain query information input by the user; The workflow display area is used to display the workflow topology; The operation area is used to operate on the workflow topology; The data area is used to display data; The preview area is used to display preview information.

[0121] In some embodiments, the workflow topology includes logical execution subgraphs corresponding to multiple task nodes and the connection relationships between the logical execution subgraphs; The method further includes: In response to the selection of the logic execution sub-graph, the corresponding configuration information and operation controls are displayed in the operation area; In response to receiving an operation instruction input by the user through the operation control, the operation instruction is executed.

[0122] In some embodiments, the method further includes: In response to the selection of the logic execution subgraph, corresponding preview information is displayed in the preview area, and the preview information includes fields.

[0123] In some embodiments, the method further includes: In response to receiving a run command from the user, the workflow corresponding to the workflow topology is executed.

[0124] This invention receives query requests from user terminals, including user questions and pending data. Based on the query requests, a structured query language is determined. A large model then uses this structured query language to obtain various task nodes. Based on the task type of each task node, the corresponding workflow generation agent is invoked to generate a logical execution subgraph for each task node. A workflow topology is then generated based on the logical execution subgraph and sent to the user terminal. This allows for automatic generation of workflow topologies based on user input, reducing the professional skills required of operators, lowering the barrier to workflow design, shortening the cycle time, improving workflow design efficiency and responsiveness, and enhancing workflow consistency and accuracy.

[0125] Figure 8 This is a schematic diagram of an auxiliary device for generating server workflows according to an embodiment of the present invention. Figure 8 As shown, the auxiliary workflow generation device for a server in this embodiment of the invention includes a query request receiving unit 81, a structured query language determining unit 82, a task node acquiring unit 83, a logical execution subgraph generating unit 84, a workflow topology generating unit 85, and a workflow topology sending unit 86. The query request receiving unit 81 receives query requests sent by user terminals, the query requests including user questions and data to be processed. The structured query language determining unit 82 determines a structured query language based on the query request. The task node acquiring unit 83 acquires each task node using a large model based on the structured query language. The logical execution subgraph generating unit 84 calls the corresponding workflow generation agent according to the node type of each task node to generate a logical execution subgraph corresponding to each task node. The workflow topology generating unit 85 generates a workflow topology based on the logical execution subgraph. The workflow topology sending unit 86 sends the workflow topology to the user terminal.

[0126] This invention receives query requests from user terminals, including user questions and pending data. Based on the query requests, a structured query language is determined. A large model then uses this structured query language to obtain various task nodes. Based on the task type of each task node, the corresponding workflow generation agent is invoked to generate a logical execution subgraph for each task node. A workflow topology is then generated based on the logical execution subgraph and sent to the user terminal. This allows for automatic generation of workflow topologies based on user input, reducing the professional skills required of operators, lowering the barrier to workflow design, shortening the cycle time, improving workflow design efficiency and responsiveness, and enhancing workflow consistency and accuracy.

[0127] Figure 9 This is a schematic diagram of an auxiliary device for generating the workflow of a user terminal according to an embodiment of the present invention. Figure 9As shown, the workflow auxiliary generation device for a user terminal in this embodiment of the invention includes a query request acquisition unit 91, a query request sending unit 92, and a workflow topology display unit 93. The query request acquisition unit 91 acquires a query request input by the user, which includes a user question and data to be processed. The query request sending unit 92 sends the query request to a server. The server determines a structured query language based on the query request, acquires various task nodes using a large model based on the structured query language, calls the corresponding workflow generation agent based on the node type of each task node to generate a logical execution subgraph corresponding to each task node, and generates a workflow topology based on the logical execution subgraph. The workflow topology display unit 93 displays the workflow topology through a display interface.

[0128] This invention receives query requests from user terminals, including user questions and pending data. Based on the query requests, a structured query language is determined. A large model then uses this structured query language to obtain various task nodes. Based on the task type of each task node, the corresponding workflow generation agent is invoked to generate a logical execution subgraph for each task node. A workflow topology is then generated based on the logical execution subgraph and sent to the user terminal. This allows for automatic generation of workflow topologies based on user input, reducing the professional skills required of operators, lowering the barrier to workflow design, shortening the cycle time, improving workflow design efficiency and responsiveness, and enhancing workflow consistency and accuracy.

[0129] Figure 10 This is a schematic diagram of an electronic device according to an embodiment of the present invention. In this embodiment, the electronic device 10 includes a server, a terminal, etc. Figure 10 As shown, the electronic device 10 includes at least one processor 101; a memory 102 communicatively connected to at least one processor 101; and a communication component 103 communicatively connected to a scanning device, wherein the communication component 103 receives and transmits data under the control of the processor 101; wherein the memory 102 stores instructions executable by at least one processor 101, the instructions being executed by at least one processor 101 to implement the above-described workflow auxiliary generation method.

[0130] Specifically, the electronic device includes: one or more processors 101 and a memory 102. Figure 10 Taking a processor 101 as an example, the processor 101 and the memory 102 can be connected via a bus or other means. Figure 10Taking a bus connection as an example, memory 102, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. Processor 101 executes various functional applications and data processing of the device by running the non-volatile software programs, instructions, and modules stored in memory 102, thereby realizing the above-mentioned workflow auxiliary generation method.

[0131] Memory 102 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and applications required for at least one function; the data storage area may store an option list, etc. Furthermore, memory 102 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 102 may optionally include memory remotely located relative to processor 101, and these remote memories may be connected to external devices via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0132] One or more modules are stored in memory 102 and, when executed by one or more processors 101, execute the workflow auxiliary generation method in any of the above method embodiments.

[0133] The above-mentioned products can perform the methods provided in the embodiments of this application, and have the corresponding functional modules and beneficial effects of performing the methods. For technical details not described in detail in this embodiment, please refer to the methods provided in the embodiments of this application.

[0134] This invention receives query requests from user terminals, including user questions and pending data. Based on the query requests, a structured query language is determined. A large model then uses this structured query language to obtain various task nodes. Based on the task type of each task node, the corresponding workflow generation agent is invoked to generate a logical execution subgraph for each task node. A workflow topology is then generated based on the logical execution subgraph and sent to the user terminal. This allows for automatic generation of workflow topologies based on user input, reducing the professional skills required of operators, lowering the barrier to workflow design, shortening the cycle time, improving workflow design efficiency and responsiveness, and enhancing workflow consistency and accuracy.

[0135] Another embodiment of the present invention relates to a non-volatile storage medium for storing a computer-readable program for use by a computer to execute some or all of the above-described method embodiments.

[0136] That is, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0137] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A workflow-aided generation method, characterized in that, The method includes: Receive a query request sent by a user terminal, the query request including user questions and data to be processed; Determine the structured query language based on the query request; Each task node is obtained through the large model using the structured query language. Based on the node type of each task node, the corresponding workflow is invoked to generate an intelligent agent, thereby generating a logical execution subgraph corresponding to each task node. Generate workflow topology based on the logic execution subgraph; The workflow topology is sent to the user terminal.

2. The method according to claim 1, characterized in that, The data to be processed is raw data or structured query language; The step of determining the structured query language based on the query request includes: In response to the fact that the data to be processed is raw data, the structured query language is generated based on the user question and the data to be processed; In response to the fact that the data to be processed is a Structured Query Language (SCL), the data to be processed is used as the SCL.

3. The method according to claim 1, characterized in that, The step of calling the corresponding workflow to generate an intelligent agent based on the node type of each task node, in order to generate the logical execution subgraph corresponding to each task node, includes: Based on the node type of each task node, the corresponding workflow generation agent is invoked in parallel to generate the logical execution subgraph corresponding to each task node.

4. The method according to claim 1, characterized in that, The node types include one or more of the following: data cleaning node, grouping and aggregation node, table join node, condition judgment node, calculated column node, and merge table node.

5. A workflow-aided generation method, characterized in that, The method includes: Obtain a query request input by the user, the query request including the user's question and the data to be processed; The query request is sent to the server, which is used to determine the structured query language based on the query request, obtain each task node through the large model based on the structured query language, call the corresponding workflow to generate an intelligent agent according to the node type of each task node, generate a logical execution subgraph corresponding to each task node, and generate a workflow topology based on the logical execution subgraph. The workflow topology is displayed through a graphical interface.

6. The method according to claim 5, characterized in that, The node types include one or more of the following: data cleaning node, grouping and aggregation node, table join node, condition judgment node, calculated column node, and merge table node.

7. The method according to claim 5, characterized in that, The display interface includes a human-computer interaction area, a workflow display area, an operation area, a data area, and a preview area; The human-computer interaction area is used to obtain query information input by the user; The workflow display area is used to display the workflow topology; The operation area is used to operate on the workflow topology; The data area is used to display data; The preview area is used to display preview information.

8. The method according to claim 7, characterized in that, The workflow topology includes logical execution subgraphs corresponding to multiple task nodes and the connection relationships between each logical execution subgraph. The method further includes: In response to the selection of the logic execution sub-graph, the corresponding configuration information and operation controls are displayed in the operation area; In response to receiving an operation instruction input by the user through the operation control, the operation instruction is executed.

9. The method according to claim 8, characterized in that, The method further includes: In response to the selection of the logic execution subgraph, corresponding preview information is displayed in the preview area, and the preview information includes fields.

10. The method according to claim 5, characterized in that, The method further includes: In response to receiving a run command from the user, the workflow corresponding to the workflow topology is executed.

11. A workflow auxiliary generation device, characterized in that, The device includes: A query request receiving unit is used to receive query requests sent by user terminals, wherein the query request includes user questions and data to be processed; A structured query language determination unit is used to determine the structured query language based on the query request; The task node acquisition unit is used to acquire each task node through the large model based on the structured query language. The logic execution subgraph generation unit is used to call the corresponding workflow generation agent according to the node type of each task node to generate the logic execution subgraph corresponding to each task node. A workflow topology generation unit is used to generate a workflow topology based on the logical execution subgraph. A workflow topology sending unit is used to send the workflow topology to the user terminal.

12. A workflow auxiliary generation device, characterized in that, The device includes: The query request acquisition unit is used to acquire the query request input by the user, which includes the user's question and the data to be processed. A query request sending unit is used to send the query request to the server. The server is used to determine a structured query language based on the query request, obtain each task node through the large model based on the structured query language, call the corresponding workflow to generate an intelligent agent according to the node type of each task node, generate a logical execution subgraph corresponding to each task node, and generate a workflow topology based on the logical execution subgraph. The workflow topology display unit is used to display the workflow topology through a display interface.

13. An electronic device comprising a memory and a processor, characterized in that, The memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method as described in any one of claims 1-10.

14. A computer-readable storage medium storing computer program instructions thereon, characterized in that, The computer program instructions, when executed by a processor, implement the method as described in any one of claims 1-10.

Images