Model-based workflow creation method, apparatus, device, medium, and product

Abstract

According to embodiments of the present disclosure, a model-based workflow creation method, apparatus, device, storage medium and program product are provided. The method comprises: in response to receiving a workflow editing request issued in a data object, determining, based on the workflow editing request, a current draft of the workflow and information of the data object, a workflow framework corresponding to the workflow by using at least one machine learning model, the workflow framework indicating respective node types of a plurality of nodes included in the workflow and a topology of the plurality of nodes, the current draft including a structured representation of a current state of the workflow; and determining, according to the workflow framework, respective configuration information of the plurality of nodes; and determining the workflow based on the workflow framework and the respective configuration information of the plurality of nodes. The above workflow creation process can be implemented by using a large model. For example, a large model can be used to create a workflow for a database table.

Description

TECHNICAL FIELD

[0001] Example embodiments of the present disclosure generally relate to the field of computers, and in particular, to a model-based workflow creation method, apparatus, electronic device, computer-readable storage medium, and computer program product. BACKGROUND

[0002] With the rapid development of machine learning and data analysis technologies, more and more office and management systems begin to introduce intelligent assistants to simplify the creation and execution of task processes for data objects. In scenarios such as task management, form processing, approval processes, etc., users often need to define operation processes in a structured manner, for example, configure trigger conditions, judgment logic, and processing actions in sequence. Especially in the case of complex process logic and large data scale, how to accurately generate an operation process with clear structure and complete configuration becomes a problem worthy of attention. SUMMARY

[0003] In a first aspect of the present disclosure, a model-based workflow creation method is provided. The method comprises: in response to receiving a workflow editing request issued in a data object, determining, based on the workflow editing request, a current draft of the workflow, and information of the data object, a workflow framework corresponding to the workflow by using at least one machine learning model, the workflow framework indicating respective node types of a plurality of nodes included in the workflow and a topology of the plurality of nodes, the current draft including a structured representation of a current state of the workflow; and determining, according to the workflow framework, respective configuration information of the plurality of nodes; and determining the workflow based on the workflow framework and the respective configuration information of the plurality of nodes.

[0004] In a second aspect of the present disclosure, an apparatus for model-based workflow creation is provided. The apparatus comprises: a workflow framework determination module configured to, in response to receiving a workflow editing request issued in a data object, determine, based on the workflow editing request, a current draft of the workflow, and information of the data object, a workflow framework corresponding to the workflow by using at least one machine learning model, the workflow framework indicating respective node types of a plurality of nodes included in the workflow and a topology of the plurality of nodes, the current draft including a structured representation of a current state of the workflow; a configuration information determination module configured to determine, according to the workflow framework, respective configuration information of the plurality of nodes; and a workflow determination module configured to determine the workflow based on the workflow framework and the respective configuration information of the plurality of nodes.

[0005] In a third aspect of the present disclosure, an electronic device is provided. The device comprises at least one processor; and at least one memory coupled to the at least one processor and storing instructions for execution by the at least one processor. The instructions, when executed by the at least one processing unit, cause the electronic device to perform the method of the first aspect.

[0006] In a fourth aspect of the disclosure, a computer readable storage medium is provided. The medium has stored thereon computer instructions that, when executed by a processor, implement the method of the first aspect.

[0007] In a fifth aspect of the disclosure, a computer program product is provided. The product includes a computer program, wherein the computer program, when executed by a processor, implements the method according to the first aspect of the disclosure.

[0008] It should be understood that nothing in the Summary is to be construed as a limitation on the scope of the embodiments of the disclosure. Other features, aspects, and advantages of the disclosure will become apparent from the following detailed description, figures and claims. BRIEF DESCRIPTION OF DRAWINGS

[0009] The above and other features, aspects and advantages of embodiments of the disclosure will become more apparent from the following detailed description, taken in conjunction with the accompanying drawings, in which like reference numerals denote like elements, and wherein:

[0010] Figure 1 a schematic diagram illustrating an example environment in which embodiments of the disclosure can be implemented;

[0011] Figure 2 a schematic diagram illustrating an example process of workflow creation according to some embodiments of the disclosure;

[0012] Figure 3A a schematic diagram illustrating an example interface for generating a workflow framework according to some embodiments of the disclosure;

[0013] Figure 3B a schematic diagram illustrating an example interface for determining configuration information according to some embodiments of the disclosure;

[0014] Figure 3C a schematic diagram illustrating an example interface for updating a workflow framework according to some embodiments of the disclosure;

[0015] Figure 4 a flowchart of a method for model-based workflow creation according to some embodiments of the disclosure;

[0016] Figure 5 a schematic structural block diagram of an apparatus for model-based workflow creation according to some embodiments of the disclosure; and

[0017] Figure 6 a block diagram of an electronic device in which one or more embodiments of the disclosure can be implemented. DETAILED DESCRIPTION

[0018] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the drawings, it is understood that the present disclosure can be implemented in various forms and should not be interpreted as being limited to the embodiments set forth herein, but rather, these embodiments are provided so as to more completely and thoroughly understand the present disclosure. It is understood that the drawings and embodiments of the present disclosure are for exemplary purposes only and are not intended to limit the scope of protection of the present disclosure.

[0019] In the description of embodiments of the present disclosure, the term "comprising" and its conjugations should be understood as open-ended, i.e., "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "an embodiment" should be understood as "at least one embodiment". The term "some embodiments" should be understood as "at least some embodiments". Other explicit or implicit definitions can also be included below.

[0020] In this document, unless explicitly stated, performing a step "in response to A" does not mean performing the step immediately after A, but can include one or more intermediate steps.

[0021] It can be understood that the data involved in the technical solutions of the present disclosure (including but not limited to the data itself, obtaining, using, storing or deleting of the data) should comply with the requirements of relevant laws and regulations and relevant provisions.

[0022] It can be understood that before using the technical solutions disclosed in the embodiments of the present disclosure, the type of information involved in the present disclosure, the scope of use, the use scenario, etc. should be informed to the relevant user and the authorization of the relevant user should be obtained through appropriate means, wherein the relevant user can include any type of right subject, such as an individual, an enterprise, or a group.

[0023] For example, in response to receiving a user's active request, a prompt message is sent to the relevant user to explicitly prompt the relevant user that the operation requested to be performed will require obtaining and using the information of the relevant user, so that the relevant user can voluntarily choose whether to provide information to the software or hardware such as electronic devices, application programs, servers or storage media that perform the operation of the technical solutions of the present disclosure according to the prompt message.

[0024] As an optional but non-limiting implementation manner, in response to receiving an active request of a relevant user, the manner of sending a prompt message to the relevant user can be, for example, a pop-up window manner, in which the prompt message can be presented in the form of text. In addition, the pop-up window can also carry selection controls for the user to select "agree" or "disagree" to provide information to the electronic device.

[0025] It can be understood that the above notification and user authorization obtaining process is only illustrative and does not limit the implementation of the present disclosure, and other ways that meet relevant laws and regulations can also be applied to the implementation of the present disclosure.

[0026] As used herein, the term “model” can learn the relationship between the corresponding input and output from the training data, so that after the training is completed, the corresponding output can be generated for a given input. The generation of the model can be based on machine learning techniques. Deep learning is a machine learning algorithm that processes input and provides a corresponding output by using multiple layers of processing units. The neural network model is an example of a model based on deep learning. In this text, “model” can also be referred to as “machine learning model”, “learning model”, “machine learning network” or “learning network”, which are used interchangeably in this text. In some embodiments described below, the machine learning model can be a large language model or have an architecture based on a large language model. The machine learning model can also be a multi-modal model capable of processing multi-modal input (e.g., text input and visual input).

[0027] Figure 1 A schematic diagram of an example environment 100 in which embodiments of the present disclosure can be implemented is shown. In this example environment 100, a digital assistant 120 and a business component 125 are installed in a terminal device 110. A user 140 can interact with the digital assistant 120 and the business component 125 via the terminal device 110 and / or an attached device of the terminal device 110.

[0028] In some embodiments, the digital assistant 120 and the business component 125 can be downloaded and installed in the terminal device 110. In some embodiments, the digital assistant 120 and the business component 125 can also be accessed by other means, such as accessed through a webpage, etc. In some embodiments, the digital assistant 120 and the business component 125 can be pre-installed in the terminal device 110. Figure 1 In the environment 100, in response to the business component 125 being launched, the terminal device 110 can present an interface 150 of the digital assistant 120 and the business component 125.

[0029] The business component 125 includes, but is not limited to, one or more of the following: a chat business component (also referred to as an instant messaging business IM component), a document business component, an audio and video conference business component, a mail business component, a task business component, a calendar business component, an objective and key result (OKR) business component, etc. It can be understood that although the business component 125 is illustrated as a single component in the environment 100, the business component 125 can include multiple components in some embodiments. Figure 1A single service component is shown in the figure, but in practice multiple service components can be installed on the terminal device 110. The service components can be integrated on a multi-functional collaboration platform. In the case where multiple service components are installed on the terminal device 110, the multiple service components can be integrated on one or more multi-functional collaboration platforms. In a multi-functional collaboration platform, different service components can be launched as needed by a person to complete corresponding information processing, sharing, communication, etc. The service component 125 can provide a content entity 126. The content entity 126 can be a content instance created by the user 140 or other users on the service component 125. By way of example, depending on the type of the service component 125, the content entity 126 can be a document (e.g., a word document, a pdf document, a presentation, a table document, etc.), an email, a message (e.g., a conversation message on an instant messaging service component), a calendar, a schedule, a task, an audio, a video, an image, etc.

[0030] In some embodiments, the digital assistant 120 can be provided by a separate service component, or can be integrated in a certain service component 125 capable of providing a content entity. The service component for providing the client interface of the digital assistant can correspond to a single-function service component or a multi-functional collaboration platform, such as an office suite or other collaboration platform capable of integrating multiple components. It can be understood that, similar to the service component, although Figure 1 A single digital assistant is shown in the figure, but in practice there can be multiple digital assistants.

[0031] In some embodiments, the digital assistant 120 supports the use of plugins. Each plugin is capable of providing one or more functions of a service component. Such plugins include, but are not limited to, one or more of the following: a search plugin, a contact plugin, a message plugin, a document plugin, a table plugin, an email plugin, a calendar plugin, a schedule plugin, a task plugin, etc.

[0032] The digital assistant 120 can be a smart assistant for the user, with intelligent conversation and information processing capabilities. In embodiments of the present disclosure, the digital assistant 120 is used for interaction with the user 140 to assist the user 140 in using the terminal device or the service component. An interaction window with the digital assistant 120 can be presented in the client interface. In the interaction window, the user 140 can be able to converse with the digital assistant 120 by inputting natural language, pictures, audio files, video files, web page files, etc. to instruct the digital assistant to assist in completing various tasks, including operations on the content entity 126.

[0033] The digital assistant 120 can be invoked or woken up by appropriate means (e.g., a shortcut, a button, or a voice) to present an interaction window with the user 140. By selecting the digital assistant 120, an interaction window with the digital assistant 120 can be opened. The interaction window can include interface elements for information interaction, such as an input box, a message list, a message bubble, and the like. In some other embodiments, the digital assistant 120 can be invoked through an entry control or a menu provided in a page, or can be invoked by inputting a preset instruction.

[0034] In some embodiments, the terminal device 110 communicates with the server 130 to implement the provisioning of services of the digital assistant 120 and the service components 125. The terminal device 110 can be any type of mobile terminal, fixed terminal, or portable terminal including a mobile handset, a desktop computer, a laptop computer, a notebook computer, a netbook computer, a tablet computer, a media computer, a multimedia tablet, a personal communication system (PCS) device, a personal navigation device, a personal digital assistant (PDA), an audio / video player, a digital camera / camcorder, a television receiver, a radio broadcast receiver, an electronic book device, a game device, or any combination thereof, including accessories and peripherals of these devices, or any combination thereof. In some embodiments, the terminal device 110 can also support any type of interface to the user (such as a “wearable” circuit, etc.). The server 130 can be various types of computing systems / servers capable of providing computing power, including but not limited to mainframes, edge computing nodes, computing devices in a cloud environment, and the like.

[0035] It should be understood that the structure and function of the various elements in the environment 100 are described for illustrative purposes only, without implying any limitation on the scope of the present disclosure.

[0036] As briefly mentioned before, with the development of data analysis and machine learning techniques, more and more workflow processing systems start to explore ways to assist users in creating automated task processes in a more intelligent and low-threshold manner. In some business scenarios, users need to set up specific processing logic according to data records, such as triggering notifications, judging conditional branches, data summarization, etc. For this purpose, some platforms start to provide low-code or graphical interface workflow editing tools, allowing users to configure automated processes by dragging and dropping nodes, filling in parameters, etc.

[0037] However, the traditional workflow configuration method still has limitations in many aspects. On the one hand, for users who are not familiar with automation logic or are not good at abstract modeling, it is necessary to understand complex information such as process structure, node logic, parameter dependency, and the learning cost is high. On the other hand, the traditional workflow process is usually static and lacks context awareness. This way cannot actively perceive and guide users to configure the process based on the user's current data environment (such as the current business table being operated).

[0038] Therefore, according to an embodiment of the present disclosure, an improved scheme for workflow creation is provided. According to the scheme, in response to receiving a workflow editing request issued in a data object, based on the workflow editing request, the current draft of the workflow, and the information of the data object, at least one machine learning model is used to determine a workflow framework corresponding to the workflow. The workflow framework indicates the respective node types of the plurality of nodes included in the workflow and the topology structure of the plurality of nodes. The current draft includes a structured representation of the current state of the workflow. Further, according to the workflow framework, the respective configuration information of the plurality of nodes is determined. Still further, based on the workflow framework and the respective configuration information of the plurality of nodes, the workflow is determined.

[0039] In this way, the user can issue a workflow editing request for various data objects (for example, structured data objects, database tables). Based on the workflow editing request, the workflow draft, and the information of the data object, a workflow framework that matches the target requirements can be automatically constructed. On this basis, by further determining the configuration information of each node, a complete and executable target workflow is finally formed. In this way, the automation level of workflow creation and modification can be effectively improved, and the understanding and operation threshold of the user can be reduced. This support implements an intelligent workflow creation method for specific business data in the data environment, which can improve the efficiency, data adaptability, and structure explainability of workflow creation. For example, a large model can be used to implement workflow creation.

[0040] Some example embodiments of the present disclosure will be described below with reference to the accompanying drawings. Figure 2A schematic diagram of an example process 200 of workflow creation is shown according to some embodiments of the present disclosure. Part or all of the process 200 can be implemented by the terminal device 110 or can be implemented by other devices, for example, can be implemented by other remote devices (in the terminal device or the service device) with computing capabilities. In the following, for ease of discussion, the example embodiments will be mainly described with respect to the terminal device 110. It should be understood that the actions described with respect to the terminal device 110 can be performed at least partially by other entities, for example, can be performed by the digital assistant 120, or can be performed by the terminal device 110 alone, by the server 130 alone, or both. For example, computationally intensive tasks, algorithm-related processing can be performed by the server 130. In the following description, different nodes or stages of creating or editing a workflow can utilize machine learning models. The machine learning models utilized by different nodes or stages can be the same or different. The machine learning models utilized can be large models. The large models can be large language models. Alternatively or additionally, the large models can be multi-modal models capable of processing multiple modalities of input (e.g., textual input, visual input).

[0041] As shown, the terminal device 110 can receive a workflow editing request 201 issued in a data object. A data object refers to a variety of forms of data entities for carrying business information. The data object can include structured data objects (e.g., data tables, database tables, etc.) and unstructured data objects. The data object can be presented in the interface 150 in the form of a table, a form, or a chart, etc. For example, the data object can include one or more multi-dimensional data tables with associated structure information. Each data object can contain a field structure and data records, the field structure including the number of fields, field names, field types, etc., and the data records corresponding to actual data rows in the table. Figure 2

[0042] In some embodiments, the data object can be associated with multiple functional components, for example, a form for data input, a dashboard for data analysis, etc. The dashboard can include statistical analysis results on the content of the data table and be displayed through various chart forms, such as bar charts, line charts, pie charts, funnel charts, etc. The user 140 can initiate a request for creating or modifying a workflow based on the data content in the process of operating these data objects, for example, viewing a table, editing a form, or analyzing a chart.

[0043] A workflow is a chain of execution logic composed of multiple functional nodes. These nodes are organized in a specific topological structure for processing business logic associated with data objects. The design, configuration, and triggering of a workflow can all rely on the information of one or more data objects. The core role of a workflow is to realize automated operations under data-driven.

[0044] ​In some embodiments, data objects can provide contextual information and data sources for workflow creation and configuration. The workflow's structure design and node configuration can be based on the field structure, field value types, and historical records in the data object. For example, when a user wants to set up a process to "notify the person in charge when the project status is delayed," the terminal device 110 can identify the "project status" field contained in the data object and generate corresponding decision and notification node configurations accordingly.

[0045] Alternatively or additionally, workflow execution can target data records within a data object. For example, a decision node might iterate through each record in a data table, triggering subsequent operations on rows that meet certain conditions. The execution results might also be written back to the original data object or its associated state updated.

[0046] In some embodiments, a data object can be mapped to multiple workflow instances. Different workflows can be applied to different fields, record sets, or business views of the data object, thereby supporting multi-dimensional automation capabilities.

[0047] In some embodiments, in response to receiving a workflow editing request 201 issued in a data object, the terminal device 110 can determine a workflow framework corresponding to the workflow based on the workflow editing request, the current draft of the workflow, and information about the data object. The determination of the workflow framework can utilize machine learning models, such as large language models or multimodal large models. The workflow framework indicates the corresponding node types and topological structure of the multiple nodes included in the workflow. The current draft includes a structured representation of the current state of the workflow. The workflow editing request 201 can be triggered by the user 140 while viewing a data object (such as a multidimensional table, form, or dashboard) by clicking a corresponding button, inputting natural language, or other means.

[0048] The current draft of a workflow refers to the structured intermediate state of the current workflow, used to record existing nodes, configurations, or previous user editing actions. The current draft can be organized graphically, for example, presented as a canvas in interface 150. It can include node types, connection relationships, and some generated or retained configuration fields. Based on this, terminal device 110 can determine the workflow framework corresponding to the current editing request. The workflow framework describes the overall structure of the workflow. The workflow framework (or workflow shell protocol) can indicate the node types of multiple nodes (such as trigger nodes, decision nodes, data processing nodes, etc.) and the connection relationships (i.e., topology) between multiple nodes. Through the workflow framework, the execution logic structure of the current workflow can be clearly defined, providing a basis for subsequent node-by-node configuration generation.

[0049] As an example, reference is made to Figure 2 At block 210, the terminal device 110 can perform requirement identification based on the workflow editing request 201. For example, the workflow editing request 201 can indicate that the user 140 wishes to create a new workflow, or to make structural adjustment or configuration modification to an existing workflow. If the workflow editing request 201 indicates to create a new workflow based on the current data object, the terminal device 110 can perform block 230.

[0050] At block 220, the terminal device 110 can generate a workflow framework. For example, the terminal device 110 can determine a blank state of the workflow as a current draft. This means that a basic draft can be initialized as a container for subsequent workflow construction before specific nodes or logical configurations are specified. The current draft can be used to hold the workflow framework and configuration content generated subsequently. Further, the terminal device 110 can generate the workflow framework based on the workflow editing request, the current draft of the workflow, and the information of the data object, using at least one machine learning model.

[0051] In some embodiments, the terminal device 110 can extract structural information (e.g., field names, field types, inter-table association relationships) and sample data from the data object (e.g., a multi-dimensional table) as important contextual information for generating the workflow framework. The above-mentioned structural information can help the model to infer the types of nodes that should be included in the process, such as generating a trigger node after identifying a time field or generating a judgment node after identifying a classification state.

[0052] In some embodiments, the terminal device 110 can determine intermediate protocol information corresponding to the current draft. The intermediate protocol information is used to describe the structured representation using a language adapted to the at least one machine learning model. The terminal device 110 can first perform intermediate protocol conversion on the current draft of the workflow, i.e., convert the existing structural information in the draft into an intermediate representation form suitable for processing by the language model.

[0053] In some examples, the intermediate protocol information can include natural language-like intermediate protocol information with strong readability and expressiveness. Such intermediate protocol information can be organized in a structured format, in which field values are expressed in a natural language manner, with accompanying annotation text indicating the meaning of each field or the business goal of each node. Such intermediate protocol information can help improve the model's understanding of the structural semantics.

[0054] Alternatively or additionally, to avoid information loss caused by structure compression, the intermediate protocol information can include structure-preserving intermediate protocol information. Such intermediate protocol information can maintain a structure hierarchy substantially consistent with the workflow draft, with only readability optimization or semantic replacement on partial fields (e.g., identifier, table name, field name). Such intermediate protocol information can not compress or re-interpret the original structure, and can preserve the connection relationship between nodes, nested structure, and complete configuration fields to the greatest extent. Therefore, such intermediate protocol information can be suitable for long process or complex structure workflow generation tasks.

[0055] In some embodiments, the terminal device 110 can generate a workflow framework based on the intermediate protocol information, information of the data object, and the workflow editing request, using at least one machine learning model. For example, when the user 140 initiates an editing request of “create an automatic reminder process” in the multi-dimensional table, the terminal device 110 can convert the current draft state into intermediate protocol information. Alternatively or additionally, the terminal device 110 can extract the field structure and sample values in the current multi-dimensional table that are most relevant to the semantics of the editing request, such as the field name, field type, and example values of the “amount”, “responsible person”, “project status”, and other fields, and provide them to the model together with the intermediate protocol information. Further, the terminal device 110 can use the machine learning model to generate a process structure including trigger nodes, judgment nodes, and notification nodes in combination with the fields related to business logic in the current table (such as “responsible person”, “status”, “deadline”, etc.), thereby achieving automatic task modeling of the data object.

[0056] In some embodiments, the terminal device 110 can update the current workflow draft based on the generated workflow framework. For example, the terminal device 110 can translate the structure information, node types, and connection relationships between nodes of the multiple nodes described in the workflow framework back to a structured representation conforming to the format of the workflow draft, thereby forming an updated workflow draft. Further, the terminal device 110 can present the updated current draft. For example, after the draft is updated, the terminal device 110 can present the updated workflow draft in a drawing board area (or other user-visualized editing interface). The drawing board can display the types, connection directions, and initial description information of the nodes in the workflow in a graphical manner. Based on the workflow draft, the user can intuitively view the framework generation result and further edit, supplement, or adjust according to business needs.

[0057] In some embodiments, to invoke conciseness and reusability, the terminal device 110 can encapsulate the complete flow of generating a workflow framework from the workflow editing request 201 as one functional block. This functional block can be exposed as an atomic capability to the outside, called "shell protocol generation capability" or "workflow architecture generation capability". This functional block integrates the complete flow of converting the current draft to intermediate protocol information, preparing context parameters, calling a machine learning model and architecture generation, and converting the generated result back to the current draft. By utilizing the functional block encapsulating the atomic capability, the terminal device 110 can concisely complete the complex workflow framework planning process, improve the calling efficiency and support cross-scene reuse.

[0058] As an example, Figure 3A A schematic diagram of an example interface 300A for generating a workflow framework is shown in accordance with some embodiments of the present disclosure. As shown in Figure 3A As shown, the terminal device 110 can receive the workflow editing request 201 via an input box 301. The input box 301 can support natural language input. The user 140 can describe the automation logic he / she wants to achieve through text, voice, image, etc. The received workflow editing request 201 can be presented in the form of a dialogue bubble 302 in the user interface for the user to view the historical input content.

[0059] The terminal device 110 can perform requirement recognition on the natural language request and parse it in combination with the data object (such as a multi-dimensional table) associated with the current interface, thereby generating the framework structure of the workflow. For example, the user input is "generate a workflow to automatically send a notification, and if the transaction order amount reaches 10,000 yuan, specially remind that the order is a large order". The terminal device 110 can parse the natural language request into a structured intent and start the workflow framework generation process based on the corresponding fields (such as "order table" and "amount field") and data records in the multi-dimensional table.

[0060] Continuing to refer to Figure 3A, the terminal device 110 can invoke the function block encapsulating the framework generation capability to automatically execute generation of the workflow framework based on the received workflow editing request 201. During the execution, the terminal device 110 can feed back the current state to the user through the execution information area 304. The execution information area 304 can display information in the execution process, such as “creating process file”, “starting to create process node”, and the like. In addition, in order to improve the understandability of the user and the process traceability, the terminal device 110 can present a node list 305 included in the generated workflow framework in the dialog. Further, the terminal device 110 can show the node name and number contained in the current workflow in the node list 305, for example, “1. Trigger when there is a new order”, “2. Determine whether it is a large order”, “3. Send a large order reminder message”, and the like, to assist the user in understanding the workflow structure and the execution order.

[0061] During the generation of the workflow framework, the terminal device 110 can initialize a current draft object and present it in the form of a draft canvas 303 on the interface 300A. The draft canvas 303 is used to show the structured state of the workflow draft, wherein each node is presented in the form of a module, and the topological relationship thereof in the whole process is represented by a connection line. For example, the draft canvas 303 represents that the generated workflow framework includes a “trigger when there is a new order” node (trigger node) numbered 1, a “determine whether it is a large order” node (conditional judgment node) numbered 2, and two parallel subsequent processing nodes 3 and 4 (corresponding to the message reminder actions under the conditions of meeting and not meeting, respectively). These nodes constitute a basic process structure including a conditional branch.

[0062] Alternatively or additionally, the draft canvas 303 supports the user to adjust or select the node structure through interaction modes such as dragging and clicking, so as to trigger subsequent node configuration generation process or manual modification. The user can further edit the node content, supplement the configuration item, or add new operation instructions on the basis of the draft, and the terminal device 110 can automatically reconstruct the draft or continue to perfect the process structure in combination with the canvas state and the user input, to support interactive iterative generation and adjustment.

[0063] Referring back to Figure 2 After the generation of the workflow framework, the terminal device 110 can execute a block 225. In the block 225, the terminal device 110 can determine, according to the workflow framework, the corresponding configuration information of the plurality of nodes in the framework. In some embodiments, the workflow framework further indicates the corresponding configuration field and the corresponding description information of the plurality of nodes, and the description information indicates at least one operation of the corresponding node. The configuration field indicates the parameter of the at least one operation.

[0064] The description information is used to explain the operation logic that the corresponding node should perform in natural language or the like, helping to understand the process corresponding to each node in the workflow framework. The configuration field is used to indicate the parameters or execution conditions required by the node, which is the key structure to realize specific actions. For example, the description information of the judgment node 2 can be "determine whether the order amount in the order table is greater than 10000 yuan", and the corresponding configuration field can include reference fields (such as "order amount"), judgment conditions (such as "greater than") and threshold values (such as "10000") and the like.

[0065] In some embodiments, the configuration field can be associated with a specific field or data type in the data object. For example, if the "amount" field in the data object is a numerical field, the terminal device 110 can select this field as the source of the judgment condition and infer the available comparison operators and their applicable ranges.

[0066] In some embodiments, for a node in the plurality of nodes, the terminal device 110 can determine a parameter value of at least one operation corresponding to the node based at least on the description information of the node. Further, the terminal device 110 can add the parameter value to the configuration field of the node as part of the configuration information of the node. The terminal device 110 can complete the configuration field of each node in turn based on the description information of the node, the configuration field structure and the available fields in the data object, thereby gradually filling in the blank nodes to form a complete process.

[0067] As an example, Figure 3B A schematic diagram of an example interface 300B for determining configuration information according to some embodiments of the present disclosure is shown. As Figure 3B shown, the terminal device 110 can present a configuration control 311 for the node. The terminal device 110 can configure the control 311 to receive user input of the configuration information for the node. The configuration control 311 is used to display the description information of the node and the field to be configured, and allows the user to input the required parameter value through a drop-down menu, text input or field selection. Based on the description information of the node and the user input, the terminal device 110 can determine the parameter value. For example, the terminal device 110 can parse and determine the value of the corresponding configuration field based on the description information of the node and the user input, and write it to the data field corresponding to the node. Alternatively or additionally, the terminal device 110 can provide a combination operation of the node logic condition through the control 312, for example, "and (AND)", "or (OR)" and the like, thereby supporting the construction of multi-condition judgment logic.

[0068] Continuing to refer to Figure 3B, the interface 300B includes an intelligent filling control 313. In response to triggering of the intelligent filling control 313, the terminal device 110 can utilize a machine learning model to determine the parameter value based on the description information of the node, the current draft of the workflow, and the information of the data object. For example, the machine learning model can be utilized to determine the parameter value based on the description information, the current draft, and the information of the data object. The terminal device 110 can utilize the machine learning model to automatically identify the configuration requirement of the current node based on the description information of the current node, the draft state of the workflow, and the information of the data object (e.g., the structure and field meaning of the current data table), and infer to generate a reasonable parameter value. For example, for a node described as “determine whether the order is a large order”, the terminal device 110 can automatically identify the “order amount” field, set the judgment relationship as “greater than”, and fill in the threshold value “10000”. For another example, the process of determining the parameter value can be further decomposed. The machine learning model can be utilized to determine the configuration requirement for the node. Further, the parameter value can be determined based on the configuration requirement. Alternatively or additionally, the terminal device 110 can display the inference result as a recommended configuration for the user to confirm, adjust, or directly adopt, thereby realizing efficient automation of node configuration.

[0069] In some embodiments, determining the parameter value can be performed by calling a function block encapsulating the node configuration capability. For example, the process of converting the current draft into intermediate protocol information, preparing context parameters, utilizing the machine learning model to determine the parameter value of the node, and converting the parameter value generation result back to the current draft can be automatically performed by calling the corresponding function block.

[0070] With continued reference to Figure 3B , the interface 300B includes an adding control 314. In response to triggering of the adding control 314, the terminal device 110 can add a new node in the current workflow framework. For example, the user can select to add a functional node of the type of “sending notification”, “calling robot”, or “data update”. The terminal device 110 can automatically initialize the node type and connection relationship according to the context information (such as the selected insertion position, the type of the preceding node) while adding the node, and guide the user to complete the parameter configuration of the new node or call the intelligent filling function. Through the adding control 314, incremental expansion can be performed based on the generated workflow framework, so as to flexibly adjust the workflow logic to adapt to diversified business requirements. In some embodiments, adding a new node in the current workflow framework can be performed by calling a function block encapsulating the node configuration capability.

[0071] In some embodiments, the terminal device 110 can determine the configuration order of the plurality of nodes based on a depth-first search (DFS) strategy. DFS can ensure that when a node is configured, its pre-node (i.e., the input source that can affect the current node) has completed the configuration, thereby avoiding the situation of referencing undefined parameters or incomplete structure. In some examples, the terminal device 110 can analyze the dependency relationship between fields and nodes in the data object in real time, and recommend related fields when adding or modifying nodes to improve interaction efficiency and the accuracy of field configuration.

[0072] Further, the terminal device 110 can determine the configuration information of each node in the plurality of nodes by sequentially calling the function block for node configuration generation according to the configuration order. For example, the function block for node configuration generation can be encapsulated as an atomic capability, referred to as “configuration information determination capability” or “node configuration capability”. This atomic capability can include the complete execution process of converting the current draft to intermediate protocol information, preparing context parameters, determining the parameter values (such as field name, judgment condition, execution action, etc.) of the node using large language model inference or template rule, converting the configuration generation result back to the current draft, and writing the generation result to the configuration field of the node, thereby forming complete node configuration information. The terminal device 110 can execute this process node by node according to the DFS order until all nodes of the entire workflow complete the filling of the configuration parameter values.

[0073] Referring back to Figure 2 If the workflow editing request includes a workflow modification request, the process 200 proceeds to block 230. At block 230, the terminal device 110 can determine a modification plan for the workflow based on the workflow modification request and the current draft of the workflow using a first machine learning model. The first machine learning model can be a preset planning module or other machine learning model (e.g., a large language model). The modification plan is used to determine and plan the modification operation of the current workflow at the structure and configuration level.

[0074] In some examples, the modification plan can indicate whether the current draft needs to be adjusted in structure, i.e., whether the workflow framework needs to be modified. For example, if the user requests “replace the second step judgment node with a data aggregation node”, the modification plan can include operations such as deleting the original judgment node, inserting a new aggregation node, and reconnecting the upstream and downstream nodes. Alternatively or additionally, the modification plan can contain local modification instructions that do not involve the framework structure, for example, only modifying the configuration parameters of a certain node.

[0075] At block 231, the terminal device 110 can determine whether the modification plan indicates modification to the workflow framework, i.e., whether the modification plan indicates structural modification to the current workflow framework. If the modification plan indicates modification to the workflow framework, the process 200 can proceed to block 232. At block 232, the terminal device 110 can update the workflow framework corresponding to the workflow based on the workflow editing request, the current draft of the workflow, and the information of the data object, by utilizing a second machine learning model. The second machine learning model can be the same or different from the first machine learning model used to determine the modification plan. Embodiments of the present disclosure are not limited in this regard. For example, the process of updating the workflow framework can be performed by invoking a function block encapsulating the capability of generating a workflow framework, converting the current draft into intermediate protocol information, preparing context parameters, generating the framework of the workflow by utilizing the second machine learning model, and converting the framework generation result back to the current draft.

[0076] Based on the updated workflow framework, the terminal device 110 can further determine the configuration information of each node, for completing the parameter configuration of the newly added node or updating the context dependency of the affected node. In this way, the consistency of the workflow in terms of logical structure and execution parameters can be ensured. The determination of the configuration information of each node can be performed by invoking the aforementioned function block encapsulating the capability of configuring a node.

[0077] In some embodiments, if the modification plan only indicates adjustment to the configuration information of one or more nodes (e.g., a first node in the plurality of nodes) without involving structural changes, the process 200 can proceed to block 235. At block 235, the terminal device 110 can update the configuration information of the first node based on the workflow framework. The process of updating the configuration information of the first node can be performed by invoking the aforementioned function block encapsulating the capability of configuring a node.

[0078] As an example, Figure 3C A schematic diagram of an example interface 300C for updating a workflow framework is shown, according to some embodiments of the present disclosure. As Figure 3C As shown, the terminal device 110 can receive the workflow editing request 201 via the input box 301. The workflow editing request 201 can include the modification intention of the user 140 to the current workflow structure or configuration, e.g., “modify the judgment value to 15000” or “replace the execution operation of a certain node”. The terminal device 110 can perform requirement identification on the workflow editing request 201. In response to the workflow editing request 201 indicating modification to the current workflow (e.g., workflow A), the terminal device 110 can parse the data object (e.g., a multi-dimensional table) associated with the current interface, thereby updating the framework structure of the workflow.

[0079] Continuing to refer to Figure 3CDuring the process of modifying the workflow framework, the terminal device 110 can feed back the current state to the user by performing the progress area 321. The execution progress area 321 can show the phased progress of the current structure update, for example, the nodes affected by this modification, the changes and the expected results. In this way, the terminal device 110 can improve the transparency of the system and the controllability of the user, enabling the user to clearly understand the impact of the editing request on the overall structure of the workflow, effectively supporting multiple rounds of adjustment and verification of complex processes.

[0080] With continued reference to Figure 3C The interface 300C can include a control control 322 for confirming or canceling the update operation. For example, the control control 322 can include an “adopt” button and an “abandon” button, respectively, for confirming the application of the current modification or canceling this round of modification operation. When the user 140 clicks the “adopt” button, the terminal device 110 can write the updated workflow framework and its configuration information into the current draft, and refresh the display area to make the new structure effective. If the user 140 chooses “abandon”, the terminal device 110 can undo this round of modification, keeping the current draft of the workflow unchanged.

[0081] In some embodiments, the terminal device 110 can highlight the nodes in the draft canvas 303 that have been completed configured, or prompt the parts that have not been configured yet, to guide the user to supplement. The user 140 can further proofread, modify or save the workflow on this basis.

[0082] With reference back to Figure 2 After determining the configuration information of each node in the workflow framework, the terminal device 110 can determine the workflow 202 based on the workflow framework and the corresponding configuration information of the plurality of nodes. For example, the terminal device 110 can fuse the type, connection relationship of each node in the framework structure and its corresponding configuration field to construct a complete process definition with execution semantics. The process definition can include trigger conditions, judgment logic, data processing methods, message notification content, etc., and be stored as a process entity that can be executed in a structured data format.

[0083] In some embodiments, the terminal device 110 can present the determined complete workflow in the draft canvas 303, including the type, number, connection direction of all nodes and the configured core parameters. The terminal device 110 can support exporting the determined complete workflow into a standard format, or submitting it to a workflow engine to support subsequent automated operation.

[0084] In some embodiments, the terminal device 110 can automatically generate summary information of the workflow based on the workflow framework corresponding to the workflow, the respective configuration information of the plurality of nodes included in the workflow, and the historical input context of the user after generating the complete workflow. The summary information is used to summarize the business objectives and execution logic of the current workflow, and can include trigger conditions, key decision nodes, core processing steps, and the like.

[0085] In some embodiments, the terminal device 110 can determine the title of the workflow based on the summary information. The title is used to briefly identify the core function or scenario of the workflow. The terminal device 110 can construct the title from the keywords extracted from the summary information or automatically generate the title in combination with user input. Further, the terminal device 110 can synchronously present the generated summary information and title in the interface of the data object. For example, the title can be displayed in the upper area of the multi-dimensional table interface as the identification name of the workflow. The summary information can be presented in the conversation area or the workflow list to assist the user in quickly understanding the function and applicable scenario of the current process. Through the generation and display of the title and the summary, not only the readability and interpretability of the workflow display are enhanced, but also the subsequent process management and retrieval of the user are facilitated.

[0086] In some embodiments, the terminal device 110 can detect a situation where the workflow associated with the current interface is inconsistent with the input workflow editing request 201. In other words, the user can make a modification or creation demand for workflow B in the editing interface of workflow A. For example, the terminal device 110 can determine whether the request can be directly configured on the workflow corresponding to the current draft based on the target expression in the editing request. If it cannot be determined or is determined to be mismatched, the terminal device 110 can further confirm the intention of the user 140. For example, through the interface prompt “Do you want to apply this configuration to workflow B?” or guiding the user 140 to select the target workflow in the multiple workflow candidates.

[0087] Alternatively or additionally, the terminal device 110 can determine whether to switch the current workflow in combination with at least one other workflow recently operated by the user (for example, the last opened workflow file), the editing request content currently input by the user, and the workflow features (such as the name, process type, and the like) pointed to by the user. Ultimately, the terminal device 110 can determine an explicit current workflow, and use the information as the input of the subsequent “create workflow” or “modify workflow” process. In this way, it can be ensured that the subsequent structure generation or parameter configuration operation is located on the correct workflow file, thereby avoiding mismodification or semantic conflict.

[0088] In summary, according to embodiments of the present disclosure, by generating a clear structure and complete configuration workflow automatically based on the structural information and content data related to the data object, efficient transformation from business data to automated process is achieved. In addition, by dividing the workflow generation into two stages of framework generation and configuration completion, and introducing an intermediate protocol to bridge natural language and structured logic, the understanding ability of large models on the context of data objects and the generation accuracy are improved.

[0089] Figure 4 A flowchart of a method 400 for model-based workflow creation is shown according to some embodiments of the present disclosure. The method 400 can be implemented in any device. For example, the method 400 can be implemented at the terminal device 110, or at the server 130, or can be completed by the terminal device 110 and the server 130 in coordination. The following describes the method 400 with reference to the terminal device 110 and the server 130. Figure 1 The method 400 is described.

[0090] At block 410, in response to receiving a workflow editing request issued in a data object, the terminal device 110 determines, based on the workflow editing request, a current draft of the workflow, and information of the data object, a workflow framework corresponding to the workflow using at least one machine learning model, the workflow framework indicating respective node types of a plurality of nodes included in the workflow and a topology of the plurality of nodes, the current draft including a structured representation of a current state of the workflow.

[0091] At block 420, the terminal device 110 determines, according to the workflow framework, respective configuration information of the plurality of nodes.

[0092] At block 430, the terminal device 110 determines the workflow based on the workflow framework and the respective configuration information of the plurality of nodes.

[0093] In some embodiments, the method 400 further includes, in response to determining that the workflow editing request includes a workflow creation request, determining a structured representation of a blank state of the workflow as the current draft.

[0094] In some embodiments, determining the workflow framework corresponding to the workflow includes determining intermediate protocol information corresponding to the current draft, the intermediate protocol information being used to describe the structured representation using a language adapted to the at least one machine learning model; and generating the workflow framework based on the intermediate protocol information, the information of the data object, and the workflow editing request using the at least one machine learning model.

[0095] In some embodiments, the method 400 further includes updating the current draft of the workflow based on the workflow framework; and presenting the updated current draft of the workflow.

[0096] In some embodiments, the workflow framework further indicates respective configuration fields and respective description information of the plurality of nodes, the description information indicating at least one operation of the respective node, the configuration field indicating parameters of the at least one operation, and determining the respective configuration information of the plurality of nodes comprises: for a node in the plurality of nodes, determining, based at least on the description information of the node, a parameter value of the at least one operation corresponding to the node; and adding the parameter value to the configuration field of the node as part of the configuration information of the node.

[0097] In some embodiments, determining the parameter value of the at least one operation corresponding to the node comprises: receiving a user input of the configuration information of the node; and determining the parameter value based on the description information of the node and the user input.

[0098] In some embodiments, determining the parameter value of the at least one operation corresponding to the node comprises: utilizing a machine learning model to determine the parameter value based on the description information of the node, the current draft of the workflow, and the information of the data object.

[0099] In some embodiments, determining the respective configuration information of the plurality of nodes comprises: determining a configuration order of the plurality of nodes based on a depth-first traversal strategy; and determining the configuration information of each node in the plurality of nodes respectively by invoking function blocks for node configuration generation in sequence according to the configuration order.

[0100] In some embodiments, determining the workflow framework corresponding to the workflow comprises: in response to determining that the workflow editing request comprises a workflow modification request, determining, based on the workflow modification request and the current draft of the workflow, a modification plan for the workflow by utilizing a first machine learning model; and in response to the modification plan indicating a modification to the workflow framework, updating the workflow framework corresponding to the workflow based on the workflow editing request, the current draft of the workflow, and the information of the data object by utilizing a second machine learning model.

[0101] In some embodiments, determining the workflow framework corresponding to the workflow further comprises: in response to the modification plan indicating a modification to configuration information of a first node in the plurality of nodes, updating the configuration information of the first node based on the workflow framework.

[0102] In some embodiments, the method 400 further comprises: determining summary information of the workflow based on the workflow framework corresponding to the workflow, the respective configuration information of the plurality of nodes included in the workflow, and the historical input context; determining a title of the workflow based on the summary information; and presenting the summary information and the title in the interface of the data object.

[0103] Embodiments of the present disclosure also provide a corresponding apparatus for implementing the above method or process. Figure 5A schematic structural block diagram of an apparatus 500 for model-based workflow creation is shown in accordance with some embodiments of the present disclosure. The apparatus 500 may, for example, be implemented in or included in the terminal device 110, or the server 130, or partially implemented in the terminal device 110 and partially implemented in the server 130. Various modules / components in the apparatus 500 can be implemented by hardware, software, firmware, or any combination thereof.

[0104] As shown, the apparatus 500 includes a workflow framework determination module 510 configured to, in response to receiving a workflow edit request issued in a data object, determine, with at least one machine learning model, a workflow framework corresponding to the workflow based on the workflow edit request, a current draft of the workflow, and information of the data object, the workflow framework indicating respective node types of a plurality of nodes included in the workflow and a topology of the plurality of nodes, the current draft including a structured representation of a current state of the workflow; a configuration information determination module 520 configured to determine, according to the workflow framework, respective configuration information of the plurality of nodes; and a workflow determination module 530 configured to determine the workflow based on the workflow framework and the respective configuration information of the plurality of nodes. Figure 5

[0105] In some embodiments, the apparatus 500 further includes a current draft determination module configured to, in response to determining that the workflow edit request includes a workflow creation request, determine a structured representation of a blank state of the workflow as the current draft.

[0106] In some embodiments, the workflow framework determination module 510 is further configured to determine intermediate protocol information corresponding to the current draft, the intermediate protocol information being used to describe the structured representation using a language adapted to the at least one machine learning model; and generate the workflow framework with the at least one machine learning model based on the intermediate protocol information, the information of the data object, and the workflow edit request.

[0107] In some embodiments, the current draft determination module is further configured to update the current draft of the workflow based on the workflow framework; and present the updated current draft of the workflow.

[0108] In some embodiments, the workflow framework further indicates respective configuration fields of the plurality of nodes and respective description information, the description information indicating at least one operation of a respective node, the configuration field indicating a parameter of the at least one operation, and the configuration information determination module 520 is further configured to, for a node of the plurality of nodes, determine, based on at least the description information of the node, a parameter value of the at least one operation corresponding to the node; and add the parameter value to the configuration field of the node as part of the configuration information of the node.

[0109] ​In some embodiments, the configuration information determining module 520 is further configured to: receive a user input of the configuration information for the node; and determine the parameter value based on the description information of the node and the user input.

[0110] In some embodiments, the configuration information determining module 520 is further configured to: determine the parameter value based on the description information of the node, the current draft of the workflow and the information of the data object, by utilizing a machine learning model.

[0111] In some embodiments, the configuration information determining module 520 is further configured to: determine a configuration order of the plurality of nodes based on a depth-first traversal strategy; and determine the configuration information of each node of the plurality of nodes respectively by invoking the function block for the node configuration generation in sequence according to the configuration order.

[0112] In some embodiments, the workflow framework determining module 510 is further configured to: in response to determining that the workflow editing request comprises a workflow modification request, determine a modification plan for the workflow based on the workflow modification request and the current draft of the workflow, by utilizing a first machine learning model; and in response to the modification plan indicating a modification to the workflow framework, update the workflow framework corresponding to the workflow based on the workflow editing request, the current draft of the workflow and the information of the data object, by utilizing a second machine learning model.

[0113] In some embodiments, the workflow framework determining module 510 is further configured to: in response to the modification plan indicating a modification to the configuration information of a first node of the plurality of nodes, update the configuration information of the first node based on the workflow framework.

[0114] In some embodiments, the apparatus 500 further comprises an information determining module configured to: determine summary information of the workflow based on the workflow framework corresponding to the workflow, the respective configuration information of the plurality of nodes included in the workflow and the historical input context; determine a title of the workflow based on the summary information; and present the summary information and the title in an interface of the data object.

[0115] The units and / or modules included in the apparatus 500 can be implemented utilizing various means, including software, hardware, and / or firmware. In some embodiments, one or more units and / or modules can be implemented using software and / or firmware, e.g., machine-executable instructions stored on a storage medium. In addition to or alternatively, some or all of the units and / or modules in the apparatus 500 can be implemented at least partially by one or more hardware logic components. As an example and not by way of limitation, example types of hardware logic components that can be used include Field- programmable Gate Arrays (FPGAs), Application-specific Integrated Circuits (ASICs), Application-specific Standard Products (ASSPs), System-on-a-Chip (SOCs), Complex Programmable Logic Devices (CPLDs), etc.

[0116] Figure 6 A block diagram of an electronic device 600 in which one or more embodiments of the disclosure can be implemented is shown. The electronic device 600 may, for example, be used to implement a terminal device 110 as shown in Figure 1 It should be understood that the electronic device 600 shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments described herein. Figure 6

[0117] Referring to Figure 6 , the electronic device 600 is in the form of a general electronic device. Components of the electronic device 600 can include, but are not limited to, one or more processors or processor(s) 610, memory 620, storage 630, one or more communication units 640, one or more input devices 650, and one or more output devices 660. The processor(s) 610 can be a real or virtual processor and capable of performing various processing according to programs stored in the memory 620. In a multi-processor system, multiple processing units perform computer-executable instructions in parallel to improve parallel processing capabilities of the electronic device 600.

[0118] The electronic device 600 typically includes a number of computer storage media. Such media can be any available media that is accessible by the electronic device 600 and includes both volatile and non-volatile media, removable and non-removable media. The memory 620 can be a volatile memory (e.g., registers, cache, random access memory (RAM)), a non-volatile memory (e.g., read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory), or some combination thereof. The storage 630 can be a removable or non-removable media, and can include machine-readable media, such as a flash drive, a magnetic disk, or any other medium that can be used to store information and / or data and that can be accessed by the electronic device 600.

[0119] ​Electronic device 600 may further include additional removable / non-removable, volatile / non-volatile storage media. Although not explicitly stated... Figure 6 As shown, disk drives for reading from or writing to removable, non-volatile disks (e.g., "floppy disks") and optical disk drives for reading from or writing to removable, non-volatile optical disks can be provided. In these cases, each drive can be connected to a bus (not shown) via one or more data media interfaces. Memory 620 may include computer program product 625 having one or more program modules configured to perform various methods or actions of various embodiments of this disclosure.

[0120] The communication unit 640 enables communication with other electronic devices via a communication medium. Additionally, the functionality of the components of the electronic device 600 can be implemented using a single computing cluster or multiple computing machines capable of communicating via communication connections. Therefore, the electronic device 600 can operate in a networked environment using logical connections to one or more other servers, network personal computers (PCs), or another network node.

[0121] Input device 650 can be one or more input devices, such as a mouse, keyboard, trackball, etc. Output device 660 can be one or more output devices, such as a monitor, speaker, printer, etc. Electronic device 600 can also communicate with one or more external devices (not shown) via communication unit 640 as needed. These external devices include storage devices, display devices, etc., and can communicate with one or more devices that enable user interaction with electronic device 600, or with any device that enables electronic device 600 to communicate with one or more other electronic devices (e.g., network card, modem, etc.). Such communication can be performed via input / output (I / O) interface (not shown).

[0122] According to an exemplary implementation of this disclosure, a computer-readable storage medium is provided that stores computer-executable instructions thereon, wherein the computer-executable instructions are executed by a processor to implement the methods described above. According to an exemplary implementation of this disclosure, a computer program product is also provided, which is tangibly stored on a non-transitory computer-readable medium and includes computer-executable instructions, which are executed by a processor to implement the methods described above.

[0123] Various aspects of this disclosure are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatuses, devices, and computer program products implemented according to this disclosure. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0124] The computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0125] The computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0126] The computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0127] implementations of the present disclosure have been described above, the description is illustrative only and not restrictive ones, and is not limited to the disclosed implementations. Numerous modifications and variations will become apparent to those skilled in the art in light of the above teachings. It is therefore to be understood that within the scope of the disclosed implementations, modifications can be made by one of ordinary skill in the art without departing from the scope and spirit of the disclosed implementations. Determination of the terms used herein is intended to best explain the principles of the implementations, the practical application, or the improvement over the technology in the art, or to enable other skilled persons in the art to understand the various implementations disclosed herein.

Claims

1. A model-based workflow creation method, comprising: in response to receiving a workflow editing request issued in a data object, determining, based on the workflow editing request, a current draft of a workflow, and information of the data object, a workflow framework corresponding to the workflow, the workflow framework indicating respective node types of a plurality of nodes included in the workflow and a topology of the plurality of nodes, the current draft comprising a structured representation of a current state of the workflow, the information of the data object comprising structural information and sample data extracted from the data object, wherein determining the workflow framework corresponding to the workflow comprises: determining, by performing an intermediate protocol conversion on the current draft, intermediate protocol information corresponding to the current draft, the intermediate protocol information being used to describe the structured representation using a language adapted to the at least one machine learning model, wherein structural information already in the current draft is transformed into an intermediate representation form suitable for processing by the at least one machine learning model; and generating, based on the intermediate protocol information, the information of the data object, and the workflow editing request, the workflow framework using the at least one machine learning model; determining, according to the workflow framework, respective configuration information of the plurality of nodes, the configuration information comprising at least values of configuration fields of the nodes, the configuration fields being associated with fields or data types in the data object; and determining the workflow based on the workflow framework and the respective configuration information of the plurality of nodes.

2. The method of claim 1, further comprising: in response to determining that the workflow editing request comprises a workflow creation request, determining a structured representation of a blank state of a workflow as the current draft.

3. The method of claim 1, further comprising: updating the current draft of the workflow based on the workflow framework; and presenting the updated current draft of the workflow.

4. The method of claim 1, wherein the workflow framework further indicates respective configuration fields and respective description information of the plurality of nodes, the description information indicating at least one operation of a respective node, the configuration fields indicating parameters of the at least one operation, and determining the respective configuration information of the plurality of nodes comprises: for a node of the plurality of nodes, determining, based at least on the description information of the node, parameter values of the at least one operation corresponding to the node; and adding the parameter values to the configuration fields of the node as part of the configuration information of the node.

5. The method of claim 4, wherein determining the parameter values of the at least one operation corresponding to the node comprises: receiving user input for the configuration information of the node; and determining the parameter values based on the description information of the node and the user input.

6. The method of claim 4, wherein determining the parameter values of the at least one operation corresponding to the node comprises: ​ ​ determining, using a machine learning model, the parameter value based on the description information of the node, a current draft of the workflow, and information of the data object.

7. The method of claim 1, wherein determining the respective configuration information of the plurality of nodes comprises: determining a configuration order of the plurality of nodes based on a depth-first traversal strategy; and determining the configuration information of each node of the plurality of nodes respectively by invoking function blocks for node configuration generation in sequence according to the configuration order.

8. The method of claim 1, wherein determining a workflow framework corresponding to the workflow comprises: in response to determining that the workflow editing request comprises a workflow modification request, determining, using a first machine learning model, a modification plan for the workflow based on the workflow modification request and a current draft of the workflow; and in response to the modification plan indicating a modification to the workflow framework, updating, using a second machine learning model, the workflow framework corresponding to the workflow based on the workflow editing request, the current draft of the workflow, and information of the data object.

9. The method of claim 8, further comprising: in response to the modification plan indicating a modification to configuration information of a first node of the plurality of nodes, updating the configuration information of the first node based on the workflow framework.

10. The method of claim 1, further comprising: determining summary information of the workflow based on a workflow framework corresponding to the workflow, respective configuration information of a plurality of nodes included in the workflow, and historical input context; determining a title of the workflow based on the summary information; and presenting the summary information and the title in an interface of the data object.

11. An apparatus for model-based workflow creation, comprising: a workflow framework determination module configured to, in response to receiving a workflow editing request issued in a data object, determine, using at least one machine learning model, a workflow framework corresponding to the workflow based on the workflow editing request, a current draft of the workflow, and information of the data object, the workflow framework indicating respective node types of a plurality of nodes included in the workflow and a topology of the plurality of nodes, the current draft comprising a structured representation of a current state of the workflow, the information of the data object comprising structural information and sample data extracted from the data object, wherein determining the workflow framework corresponding to the workflow comprises: determining intermediate protocol information corresponding to the current draft, the intermediate protocol information being used to describe the structured representation using a language adapted to the at least one machine learning model; and generating, using the at least one machine learning model, the workflow framework based on the intermediate protocol information, the information of the data object, and the workflow editing request. ​ a configuration information determining module configured to determine respective configuration information of the plurality of nodes according to the workflow framework, the configuration information comprising at least values of configuration fields of the nodes, the configuration fields being associated with fields or data types in the data object; and a workflow determining module configured to determine the workflow based on the workflow framework and the respective configuration information of the plurality of nodes.

12. An electronic device, comprising: at least one processor; and at least one memory coupled to the at least one processor and storing instructions for execution by the at least one processor, the instructions, when executed by the at least one processor, causing the electronic device to perform the method according to any one of claims 1-10.

13. A computer-readable storage medium having computer-executable instructions stored thereon that are executable by a processor to implement the method according to any one of claims 1-10.

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

Images