Process execution method and system

Abstract

The invention discloses a process execution method and system, relates to the technical field of business process management, and mainly aims to improve the flexibility of process execution. According to the main technical scheme, a process execution engine executes a preset business process according to a business process model. And when executing to the target decision node, generating a corresponding decision request event and issuing the decision request event to the event bus. And the rule evaluation engine monitors the event through a preset subscription mechanism, and determines a business rule which corresponds to the target decision node and is defined by a domain-specific language model in a preset rule base. Then, the engine obtains target real-time service data corresponding to the event from an external system, generates a decision suggestion of a target decision node according to the data and rules, and publishes the suggestion to an event bus, and the decision suggestion comprises an approval result, a next to-be-executed node or a node adjustment condition; and finally, the process execution engine executes an operation corresponding to the decision suggestion according to the execution interface.

Description

Technical Field

[0001] This invention relates to the field of business process management technology, and in particular to a process execution method and system. Background Technology

[0002] In the context of digital transformation of equipment operations, business process management systems play a crucial role. Current mainstream architectures generally adopt a design pattern that separates the process execution engine from the rule evaluation engine: the process execution engine advances business execution according to a pre-modeled fixed path; when the process reaches a decision node, it pauses execution, encapsulates the relevant business data into a structured data packet, and passes it to the rule evaluation engine via API calls; after receiving this data, the rule evaluation engine loads and executes the corresponding business rules, ultimately returning a decision result (such as "pass," "reject," or specifying the next node ID in the process). The process execution engine then drives the process to the next node based on this result, thus completing the business loop.

[0003] However, this loosely coupled architecture still heavily relies on the initial static modeling in practice. On the one hand, rule logic is often embedded in the system in hard-coded or configurable form, only responding passively when the process execution engine actively calls it, and outputting limited decision results within a preset range. On the other hand, the execution path followed by the process execution engine is fixed during the process model deployment phase, and its node sequence, branch conditions, and jump logic cannot be dynamically reconstructed at runtime according to real-time business needs. This fragmented mode of "rules only judge, process only executes" makes it difficult for the system to respond flexibly to temporary approval needs, severely restricting the flexibility of process execution. Summary of the Invention

[0004] In view of the above problems, the present invention provides a process execution method and system, the main purpose of which is to improve the flexibility of process execution.

[0005] To solve the above-mentioned technical problems, the present invention proposes the following solution: In a first aspect, the present invention provides a process execution method, the method comprising: The process execution engine is used to execute the corresponding preset business process according to the business request, and when the target decision node in the preset business process is executed, a corresponding decision request event is generated and published to the event bus. The preset business process is a process defined in advance according to the business process model. The rule evaluation engine uses a preset subscription mechanism to listen for decision request events in the event bus and determines the business rules corresponding to the target decision node in the preset rule base. The business rules are defined according to a domain-specific language model. The rule evaluation engine is used to obtain the target real-time business data corresponding to the decision request event from an external system. The rule evaluation engine generates decision suggestions for the target decision node based on the target real-time business data and the business rules, and publishes the decision suggestions to the event bus. The decision suggestions include the approval result of the target decision node, the next node to be executed in the preset business process, or the node adjustment status of the preset business process. Using the process execution engine, the operation corresponding to the decision suggestion is executed according to the execution interface.

[0006] Secondly, the present invention provides a process execution system, the system comprising: The process execution engine is used to execute the corresponding preset business process according to the business request, and when the target decision node in the preset business process is executed, the corresponding decision request event is generated and published to the event bus of the process execution system. The preset business process is a process defined in advance according to the business process model. The rule evaluation engine is used to listen for decision request events published by the process execution engine in the event bus through a preset subscription mechanism, and determine the business rules corresponding to the target decision node in the preset rule base. The business rules are defined according to a domain-specific language model. The rule evaluation engine is also used to obtain the target real-time business data corresponding to the decision request event from external systems; The rule evaluation engine is also used to generate decision suggestions for the target decision node based on the target real-time business data and the business rules, and publish the decision suggestions to the event bus. The decision suggestions include the approval result of the target decision node, the next node to be executed in the preset business process, or the node adjustment status of the preset business process. The process execution engine is also used to perform operations corresponding to decision recommendations published by the engine in the event bus based on the execution interface rules.

[0007] To achieve the above objectives, according to a third aspect of the present invention, a storage medium is provided, the storage medium including a stored program, wherein, when the program is executed, the device where the storage medium is located is controlled to execute the process execution method of the first aspect.

[0008] To achieve the above objectives, according to a fourth aspect of the present invention, a processor is provided for running a program, wherein the program executes the process execution method of the first aspect described above.

[0009] By employing the above technical solution, the present invention provides a process execution method and system. When the process execution engine executes a preset business process to the target decision node, it generates a corresponding decision request event and publishes it to the event bus. This avoids the blocking mode of the process execution engine pausing and waiting for synchronous API calls in traditional architectures, providing a non-intrusive triggering mechanism for rule evaluation. The rule evaluation engine listens for this event through a preset subscription mechanism and determines the business rule corresponding to the target decision node from a preset rule base. Subsequently, the rule evaluation engine actively obtains the target real-time business data corresponding to the decision request event from external systems, ensuring that the decision basis not only includes the internal state of the process but also incorporates the latest external data. This solution overcomes the decision-making lag caused by traditional solutions relying solely on static data snapshots transmitted by the process execution engine, by using departmental business environment information. Based on this, the rule evaluation engine generates decision suggestions according to the target real-time business data and the business rules, and publishes these suggestions to the event bus. These suggestions explicitly include the approval results of the target decision node, the next node to be executed in the preset business process, or adjustments to nodes in the preset business process, thus upgrading the rule output from a single approval conclusion to an executable process structure instruction. Finally, the process execution engine receives and executes the operation corresponding to the decision suggestion, enabling dynamic adjustment of the running process path without redeploying the system. Therefore, this solution effectively solves the lack of flexibility caused by the rigidity of process paths and the inability of rules to express path intervention intentions in existing loosely coupled architectures, significantly improving response efficiency.

[0010] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0011] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of a process execution method provided by an embodiment of the present invention is shown; Figure 2 A flowchart of another process execution method provided by an embodiment of the present invention is shown; Figure 3 This diagram illustrates a block diagram of a process execution system provided by an embodiment of the present invention. Figure 4 A block diagram of another process execution system provided by an embodiment of the present invention is shown. Detailed Implementation

[0012] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0013] The fundamental limitation of existing process execution architectures lies in the fact that when a process engine starts a process instance, it compiles its execution path into an internally closed and immutable execution plan. Furthermore, the engine does not provide any controlled, transaction-safe runtime interface, allowing external systems to structurally modify the execution plan. Under this constraint, the rule evaluation engine can only receive a flattened snapshot of business data when actively invoked by the process execution engine, and returns predefined enumeration values ​​(such as "pass," "reject," or an existing next node ID). It cannot perceive the complete process context, nor can it express any intention to intervene in the process topology.

[0014] This solution overcomes the above limitations through the following mechanism: First, when the process execution engine reaches the target decision node, it no longer directly calls the rule evaluation engine, but generates standardized decision request events and publishes them to the event bus. This transforms the originally closed process control point into an open event that can be subscribed to and responded to by external systems, so that rule evaluation is naturally integrated into the process lifecycle.

[0015] Secondly, the rule evaluation engine listens for the event through a preset subscription mechanism and dynamically determines the business rules corresponding to the target decision node from the preset rule base, thereby realizing the runtime binding of rules and process nodes and getting rid of the constraints of hard-coded association in traditional solutions. Furthermore, the rule evaluation engine proactively obtains real-time business data (such as real-time risk control status, supplier blacklist updates, etc.) corresponding to the decision request event from external systems to ensure that the decision-making basis not only includes internal process variables, but also incorporates the latest external business environment; Based on this, the rule evaluation engine generates structured decision suggestions based on the target real-time business data and matching business rules, and publishes the suggestions back to the event bus. The decision suggestions not only indicate the next node to be executed in the preset business process, but also clearly describe the node adjustment of the preset business process (such as inserting a new review node, skipping unnecessary steps, or redirecting the execution path). In essence, it is a process structure operation instruction that can be parsed and executed. Finally, the process execution engine listens to the decision suggestion and performs the corresponding operation through the execution interface. This achieves proactive intervention of the rules in the process topology without interrupting the current transaction or disrupting state consistency, greatly improving execution efficiency when there are changes in business conditions.

[0016] In summary, the process execution system designed in this invention automatically triggers a rule evaluation mechanism at key nodes. The rule evaluation engine, relying on real-time business data, process context, and a rule base, rapidly generates a composite decision containing approval conclusions and path control instructions. The process execution engine then flexibly executes the corresponding operations through a runtime execution interface. This completely overturns the rigidity, lag, and disconnect caused by the traditional "rules and processes separation" model, endowing business processes with real-time responsiveness, autonomous evolution, and secure controllable intelligent execution capabilities.

[0017] Next, we will combine Figure 1 This invention describes the process execution method provided by the present invention. This method is applied to a process execution system, which includes a process execution engine, a rule evaluation engine, and an event bus. The specific execution steps are as follows: Figure 1 As shown, it includes: 101. Utilize the process execution engine to execute the corresponding preset business process according to the business request, and when the target decision node in the preset business process is executed, generate the corresponding decision request event and publish it to the event bus.

[0018] When constructing a preset business process, the process execution system of this invention follows the following standardized process: First, it conducts a structured analysis of the end-to-end process for specific business scenarios such as equipment procurement, contract approval, and project initiation, accurately identifying key activities, decision points, role divisions, and data dependencies. Then, the analysis results are transformed into standard process elements conforming to the BPMN (Business Process Model and Notation) specification, such as start events, user tasks, exclusive gateways, and end events, forming a deployable BPMN process definition file.

[0019] During the process element configuration phase, runtime metadata is injected into each node, including configurable attributes (such as approval level), conditional expression placeholders, and an event listener registration mechanism. Specifically for tasks marked as "decision nodes," decision event triggers are bound to ensure that event generation is automatically triggered at runtime. The completed pre-built business processes are stored in the process repository, forming process assets ready for use.

[0020] During business applications, when a business user initiates a specific operation (such as submitting a radar equipment purchase request worth 800,000 yuan), the process execution engine receives the request, parses the content, extracts the business type (such as "purchase request") and key parameters (amount, supplier, project number, applicant department, etc.), retrieves the matching preset process from the process repository, and initializes the instance.

[0021] When the process execution engine drives the instance to run, ordinary task nodes (such as "filling out an application form") execute built-in logic (such as service calls and notification sending). When the process progresses to a decision node (such as "risk assessment approval"), a pre-bound decision event trigger is activated, automatically generating a structured decision request event. This event contains three elements: the current process instance ID, the target decision node ID, and a complete snapshot of process variables (including business parameters and runtime state), and is asynchronously published through a unified event bus (implemented based on Apache Kafka or in-memory message channels).

[0022] At this point, the process execution engine pauses, waiting for the rule evaluation engine to return decision suggestions. The system has a built-in state coordinator responsible for maintaining data consistency between the process execution engine and the rule evaluation engine, ultimately building a closed-loop control system of "process-driven - rule-triggered - state coordination" to achieve flexible process management in dynamic business scenarios.

[0023] 102. Utilize the rule evaluation engine to listen for decision request events in the event bus through a preset subscription mechanism, and determine the business rules corresponding to the target decision node in the preset rule base.

[0024] In this invention, business rules can be formally defined in advance using a Domain-Specific Language (DSL), and a unified rule meta-model specifies the syntax structure of the rule as "when [condition], execute [action]". The condition part can reference multi-dimensional context information, including: Business data dimensions (such as contract amount, project risk level, customer type, product category) Organizational authority dimension (such as the applicant's department / position, the current approver's workload). External real-time input (such as compliance review results, credit scores, and real-time risk warning signals) The action section breaks through the limitations of traditional "pass / reject" responses, supporting instructions for adjusting workflow paths, such as: Explicitly specify the next node Dynamically insert nodes (e.g., "insert a manual review node after the current node"). Conditional redirection Using a domain-specific language (DSL) allows rules to be completely separated from the program code and stored in a separate, graphically manageable rule base. This enables rules to be maintained directly by business personnel. Rules can be easily modified through the rule management interface, achieving decoupling from program code deployment.

[0025] Subsequently, the rules are centrally stored in a version control system (such as Git) or a dedicated rule management system in the form of DSL script files, and structured management is achieved through metadata such as rule ID, name, business domain, trigger conditions, execution actions, priority, effective time, and activation status, supporting audit tracing and version rollback.

[0026] At the logical organization level, a rule-node association mapping table is maintained to clearly record the specific process node to which each rule applies and the type of triggering event (e.g., "Rule R001 is bound to the pre-entry event of the 'Final Review of Large Purchase' node in process P1"). This mapping mechanism enables the rule evaluation engine to quickly locate the matching rule by node ID when it receives a decision request event, avoiding a full scan and improving matching efficiency.

[0027] During operation, the rule evaluation engine, acting as a pre-defined subscriber to the event bus, continuously listens for decision request events. When the process execution engine publishes an event containing the target decision node ID and context information, the rule evaluation engine automatically wakes up and parses the event content to extract the target decision node identifier; then it queries the rule-node mapping table to filter out the bound business rules.

[0028] 103. Use the rule evaluation engine to obtain the target real-time business data corresponding to the decision request event in the external system.

[0029] 104. Utilize the rule evaluation engine to generate decision recommendations for target decision nodes based on real-time business data and business rules, and publish the decision recommendations to the event bus.

[0030] 105. Utilize the process execution engine to perform the operations corresponding to the decision suggestions based on the execution interface.

[0031] The decision recommendations include the approval results of the target decision nodes, the next node to be executed in the preset business process, or the node adjustment status of the preset business process.

[0032] In step 103, after receiving the decision request event, the rule evaluation engine does not immediately perform rule matching, but first pulls the target real-time business data related to the current process instance from one or more external business systems.

[0033] These external systems include, but are not limited to: compliance review systems (which return blacklist comparison results), human resources systems (which query the current workload or job level status of approvers), and time services (which determine whether it is a holiday).

[0034] This operation can be achieved through a pre-configured data source adapter. The adapter can construct query parameters based on the business parameters in the complete process variable snapshot carried in the decision request event, call the standard API of the external system (such as a RESTful interface or message queue), and obtain the response synchronously or asynchronously within a set timeout period.

[0035] This step is essential because the internal variables passed by the process engine are merely historical snapshots generated at process initiation or early stages, failing to reflect real-time changes in the external environment. Relying solely on these snapshots for decision-making could lead to approval conclusions that are out of touch with actual business realities (e.g., a supplier is blacklisted midway through the process, but the rules still allow it based on the old data). By proactively acquiring real-time business data from the target through business parameters, rule evaluation can be built upon a complete, dynamic, and multi-source integrated decision context, significantly improving business compliance and risk control capabilities.

[0036] After obtaining the target real-time business data, proceed to step 104: The rule evaluation engine generates decision recommendations for the target decision node based on the data and the identified applicable business rules, and publishes the recommendations to the event bus.

[0037] In practice, the rule evaluation engine first determines whether there are multiple business rules that match the current target decision node. If multiple matching rules exist, the final target business rule is determined based on a preset conflict resolution mechanism (such as priority ranking, specificity judgment, or explicit arbitration strategy). If only one rule matches, the DSL parser directly evaluates the condition part of that rule. When the condition is met, the corresponding "action" part is extracted and converted into a structured decision suggestion object.

[0038] This decision recommendation not only includes traditional approval results (such as "approved", "rejected", or "returned"), but more importantly, it clearly expresses control instructions at the process execution level. Specifically, it includes two types of content: first, the ID of the next node to be executed in the preset business process (e.g., "nextNode='legal review'"); second, instructions for dynamically adjusting the preset business process topology, such as "insert a new node 'risk control special review' after the current node", "skip the original 'department review' node", or "terminate the current branch and redirect to the 'emergency channel' subprocess".

[0039] These instructions are coded in a standardized, machine-readable format to ensure that the process execution engine can parse and execute them unambiguously. After generating decision recommendations, the rule evaluation engine encapsulates them into standard events (such as DecisionRecommendationEvent) and publishes them asynchronously via the event bus.

[0040] This step upgrades the output of rules from a single, flat business conclusion to structured control instructions with process semantics, so that rules are no longer limited to "pointing the way" (selecting existing branches), but truly have the ability to "build the road" (dynamically modify the execution path), thus providing a solid and executable semantic foundation for runtime process adaptive adjustment.

[0041] Subsequently, in step 105, the process execution engine can be used to execute the operation corresponding to the decision suggestion according to the execution interface. Specifically, as another subscriber to the event bus, after listening to the above decision suggestion, the process execution engine first parses its content and identifies the approval result and path adjustment instructions contained therein. Then, the process execution engine calls the runtime path modification interface provided by its kernel (this interface has been integrated into the execution engine's control plane during system initialization, and is the aforementioned execution interface) to safely modify the execution plan of the current process instance according to the instructions.

[0042] In other words, this invention implements a mechanism for dynamically modifying the execution plan in the kernel of the process execution engine, and encapsulates this mechanism as a runtime path modification interface. When a decision suggestion arrives, the engine can call this interface to transform "insert / skip / redirect" instructions into safe changes to the execution state of the current instance without compromising the integrity of the process or relying on model reconstruction.

[0043] Based on the above Figure 1As can be seen from the implementation, the method provided by this invention generates a corresponding decision request event and publishes it to the event bus when the process execution engine executes a preset business process to the target decision node. This avoids the blocking mode of the process execution engine pausing and waiting for synchronous API calls in the traditional architecture, and provides a non-intrusive triggering mechanism for rule evaluation. The rule evaluation engine listens for this event through a preset subscription mechanism and determines the business rule corresponding to the target decision node from the preset rule base. Subsequently, the rule evaluation engine actively obtains the target real-time business data corresponding to the decision request event from the external system to ensure that the decision basis not only includes the internal state of the process, but also integrates the latest external business. This approach leverages environmental information to overcome the decision-making lag caused by traditional solutions relying solely on static data snapshots from the process execution engine. Building upon this, the rule evaluation engine generates decision suggestions based on the target's real-time business data and the business rules, and publishes these suggestions to the event bus. These suggestions explicitly include the approval results of the target decision node, the next node to be executed in the preset business process, or adjustments to nodes in the preset business process, thus upgrading rule output from a single approval conclusion to an executable process structure instruction. Finally, the process execution engine receives and executes the operation corresponding to the decision suggestion, enabling dynamic adjustment of the running process path without redeploying the system. Therefore, this solution effectively addresses the lack of flexibility in existing loosely coupled architectures caused by rigid process paths and the inability of rules to express path intervention intentions, significantly improving response efficiency.

[0044] Furthermore, as a response to Figure 1 Further refinement and extension of the illustrated embodiments, this invention also provides another process execution method, such as... Figure 2 As shown, the specific steps are as follows: 201. Utilize the process execution engine to execute the corresponding preset business process according to the business request, and when the target decision node in the preset business process is executed, generate the corresponding decision request event and publish it to the event bus.

[0045] 202. Using the rule evaluation engine, listen to decision request events in the event bus through a preset subscription mechanism, and determine the business rules corresponding to the target decision node in the preset rule base.

[0046] 203. Use the rule evaluation engine to obtain the target real-time business data corresponding to the decision request event in the external system.

[0047] The implementation methods of steps 201-203 are the same as those of steps 101-103, and can achieve the same technical effect and solve the same technical problem, so they will not be repeated here.

[0048] 204. Utilize the rule evaluation engine to generate decision recommendations for the target decision nodes based on the target's real-time business data and business rules, and publish the decision recommendations to the event bus.

[0049] In this embodiment, after the rule evaluation engine completes the acquisition of the target real-time business data, if it finds that there are multiple business rules matching the target decision node, it must resolve the logical conflicts that may be caused by the concurrent triggering of multiple rules. Because the content of different business rules may contradict each other, potentially giving conflicting path instructions (such as one requiring "insert risk control node," and another requiring "use the fast track to skip review"), a unique target business rule must be selected based on a preset conflict arbitration strategy to ensure the certainty and executability of the decision result. Then, the rule evaluation engine generates decision suggestions for the target decision node based on the target real-time business data and the target business rules.

[0050] When determining the target business rules based on the preset conflict arbitration strategy, the preset priority value and number of constraints for each business rule can be obtained first. Since the priority reflects the importance of the business (for example, compliance rules are higher than efficiency rules), the rules should be sorted in descending order of priority, and the highest priority business rules should be selected as candidate rules.

[0051] If multiple rules with the same priority exist, the rule with the most constraints is selected, as more specific constraints result in more precise application scenarios and lower generalization risk. If a unique rule still cannot be determined, a fallback strategy is introduced: by querying the historical decision database, the cumulative usage frequency of each candidate business rule and the number of business dimension types it covers (such as customer type, product category, region, etc.) are counted, and an adaptation coefficient is calculated (e.g., adaptation coefficient = cumulative usage frequency × number of business dimension types). This allows the rule that has shown greater stability and adaptability in historical practice to be selected as the target business rule. This mechanism effectively avoids decision-making deadlock caused by rule conflicts and ensures the robustness of the system in complex scenarios.

[0052] When using the rule evaluation engine to generate decision suggestions for target decision nodes based on target real-time business data and target business rules, the rule evaluation engine can verify whether each judgment condition of the rule can be effectively supported by the target real-time business data. Since some required fields may be missing or invalid, the system first performs a data coverage integrity check: if any judgment condition cannot be evaluated due to missing or invalid data, the original rule is not forcibly executed, but a preset processing strategy is adopted; the rule is only executed and decision suggestions are output when all conditions can be calculated based on the target real-time data. Decision suggestions include: whether the target decision node has been approved, the next node to be executed after approval, or unnecessary nodes that can be skipped.

[0053] Next, when the preset processing strategy is adopted, it does not simply reject or default to pass, but activates the intelligent completion mechanism: it evaluates the existing target real-time business data based on the preset risk model and calculates the comprehensive risk value; if the risk value is lower than the preset threshold, it means that the current scenario is generally controllable, and a new business rule can be dynamically constructed based on the rule generation template and the existing data. For example, it selects a structured template that matches the current scenario from the rule template library (e.g., "If [amount field] < [threshold] and [credit field] ∈ [good range], then [action]"), and automatically fills in the specific parameters based on the distribution characteristics and business constraints of the current real-time data (such as amount quantiles, credit rating mapping, etc.), thereby dynamically constructing a temporary but semantically complete new business rule.

[0054] Since the new rule is driven by target real-time business data and meets the low-risk premise, its output is business-reasonable. Subsequently, the target real-time business data is re-evaluated using the new rule to obtain the approval result, which, together with the required manual review node, constitutes the final decision recommendation. The reason for adding the manual review node is that the new rule has not been validated over a long period of time, and a post-confirmation mechanism is needed to achieve a risk closure loop, ensuring both efficiency and compliance.

[0055] Then, after generating new business rules, to improve the long-term intelligence level of the system, since the new rules have been verified to be effective in real-world scenarios, the "process node-rule" mapping table maintained in the preset rule base is retrieved, and the new business rule is bound to the current target decision node and written into the rule base. This operation allows similar requests in the future to directly reuse the rule, reducing manual intervention, realizing the self-evolution capability of the rule base, and forming a positive cycle of "execution → learning → optimization".

[0056] Finally, regardless of whether the decision recommendations originate from the original target business rules or dynamically generated new rules, the rule evaluation engine encapsulates them into structured decision recommendation objects. These objects include: the approval result (pass / reject) of the target decision node, the ID of the next node to be executed after approval, and path control instructions such as unnecessary nodes to be skipped. Because the subsequent process execution engine needs to unambiguously parse and execute these instructions, the recommendations must adopt a standardized, machine-readable format (such as event payloads defined by JSON Schema).

[0057] Once generated, the decision recommendation is encapsulated as a DecisionRecommendationEvent and published asynchronously via the event bus. This completely decouples decision generation from process execution, making rule evaluation an independent, pluggable, and auditable intelligent link in the process lifecycle, providing a reliable, secure, and traceable semantic foundation for the entire system to achieve runtime adaptive adjustment.

[0058] Of course, in addition to the above-mentioned automated processing mechanism, when the rule conditions are not covered by the target real-time business data and the risk assessment result exceeds the preset threshold (i.e., the conditions for automatically generating new rules are not met), an alarm or prompt message can be sent to the designated display terminal (such as the business rule management platform, the risk control personnel's workbench, or the approver's operation interface) to indicate that the current scenario lacks applicable rules or has high-risk uncertainty, and to attach relevant business context information (such as process instance ID, missing data items, risk factors, etc.).

[0059] The reason for introducing this manual intervention channel is that certain edge or highly sensitive scenarios are not suitable for automatic system decisions. Through this message, rule administrators can view the situation in real time and manually add or adjust business rules applicable to the target decision node in the rule management system. This not only fills the coverage gaps of automated strategies but also provides a human-machine collaborative feedback entry point for the continuous improvement of the rule base, further enhancing the system's adaptability and governance capabilities.

[0060] Of course, in addition to the methods mentioned above, when existing rules cannot effectively judge the target real-time business data (e.g., due to missing key external data, mismatched rule conditions, or risk values ​​exceeding the threshold for automatically generating new rules), the system can also adopt another more conservative and efficient processing strategy: instead of attempting to generate new business rules, the decision suggestion is directly set as "insert a manual review node before the current target decision node." This manual review node is handled by business personnel or risk control experts with the appropriate permissions; if the manual review is passed, the process skips the original target decision node and directly proceeds to its subsequent node; if the review fails, it terminates or is returned according to the rejection path.

[0061] This strategy differs fundamentally from the aforementioned approach of "first generating new rules, then outputting approval results based on the new rules and adding a manual review step": Directly inserting human review is a pre-emptive interception mechanism that is suitable for situations where the rule system is not yet able to cover the current scenario. Its core idea is "when uncertain, let humans judge", avoiding the system from making any automated inferences without evidence, thereby maximizing compliance and security. Generating new rules and then reviewing them afterward is an exploratory automation mechanism. It attempts to generate temporary rules using a data-driven approach only when the risks are controllable, and then uses subsequent manual review to achieve "trial before verification". The aim is to improve processing efficiency while accumulating rule samples.

[0062] Therefore, the two strategies are designed for high-risk, unknown scenarios and low-risk, explorable scenarios, respectively, together forming a layered, flexible decision-making safety net that balances security and efficiency. By dynamically selecting the appropriate strategy, the system can prevent automation from overstepping its limits while gradually expanding the boundaries of intelligence, achieving robust evolution under human-machine collaboration.

[0063] 205. Utilize the process execution engine to perform the operations corresponding to the decision suggestions based on the execution interface.

[0064] The specific implementation method of this step is the same as that of step 105, and it can achieve the same technical effect and solve the same technical problem. Therefore, it will not be repeated here.

[0065] Next, a summary example of the above-described method of the present invention will be provided, focusing on specific fields to which the present invention applies: If a company initiates a procurement request for an equipment, intending to purchase a CNC machine tool worth 8.5 million yuan, and the application is submitted by the manufacturing technology department, with the supplier being "XXX Company" and the associated project number being "XXX-2025-XXX123".

[0066] Upon receiving the business request, the process execution engine identifies it as belonging to the "Major Capital Expenditures" category, loads the corresponding business process model—"Major Capital Expenditures Approval Process V4.1"—from the preset process asset library, and creates a process instance. This process model contains multiple task nodes, among which "Large Expenditures Risk Assessment Approval" is marked as the target decision node.

[0067] When the process reaches the target decision node, a structured decision request event is automatically generated and published to the event bus. This event contains a unique identifier for the current process instance, the identifier of the target decision node, and a complete snapshot of process variables, including key business parameters such as purchase amount, supplier code, applicant's department, and project number.

[0068] As a subscriber to the event bus, the rule evaluation engine, upon receiving the decision request event, first queries the preset rule-node mapping table based on the target decision node identifier and finds two business rules bound to that node: The first rule stipulates that if the purchase amount exceeds 8 million yuan and the supplier has a good credit record, it can directly enter the final review stage by the CFO. The second rule stipulates that if the purchase amount exceeds 7 million yuan and the supplier has recently been included in the special attention list of the external compliance system, an additional risk control review process must be inserted before the current node.

[0069] To accurately determine which rule applies, the rule evaluation engine proactively calls external system interfaces to obtain real-time business data related to the supplier. The results show that although the supplier has a high historical credit score, it has been placed on the "special attention list" by the compliance review system due to its recent involvement in multiple contract performance disputes.

[0070] Based on the target real-time business data, the rule evaluation engine determines that the applicable conditions of the first rule are no longer met, while the triggering conditions of the second rule have been satisfied. The second rule is then designated as the target business rule.

[0071] Subsequently, the rule evaluation engine generates decision suggestions based on the rule: it is recommended to dynamically insert a manual review node called "additional risk control review" before the current "large expenditure risk assessment approval" node; if the manual review is approved, the original risk assessment node will be skipped and the process will proceed directly to the subsequent approval stage; if it is not approved, the process will be terminated and the application will be returned to the applicant.

[0072] The decision recommendation is encapsulated as a standard event and published to the event bus. Upon receiving it, the process execution engine calls the runtime path modification interface provided by its kernel to safely insert the "additional risk control review" node into the current execution path without interrupting the process instance or redeploying the process model, and then pauses automatic progress, awaiting the result of manual processing.

[0073] As can be seen from the above, this invention achieves decoupling of processes and rules through an event-driven mechanism; enhances the timeliness and accuracy of decision-making context by utilizing external real-time data; and supports dynamic adjustment of process paths during runtime, including operations such as inserting nodes and skipping nodes; and ensures the reliability and compliance of decisions through priority, risk assessment, and manual fallback mechanisms when rules conflict or data is incomplete.

[0074] Furthermore, as a response to the above Figure 1 In addition to the implementation of the method shown, this embodiment of the invention also provides a process execution system for the above-described process. Figure 1 The method shown is implemented accordingly. This device embodiment corresponds to the aforementioned method embodiment. For ease of reading, this system embodiment will not repeat the details of the aforementioned method embodiments one by one, but it should be clear that the system in this embodiment can implement all the contents of the aforementioned method embodiments. Figure 3 As shown, the system includes: The process execution engine 301 is used to execute the corresponding preset business process according to the business request, and when the target decision node in the preset business process is executed, the corresponding decision request event is generated and published to the event bus of the process execution system. The preset business process is a process defined in advance according to the business process model. The rule evaluation engine 302 is used to listen to the decision request event published in the event bus by the process execution engine 301 through a preset subscription mechanism, and determine the business rule corresponding to the target decision node in the preset rule base. The business rule is defined according to the domain-specific language model. The rule evaluation engine 302 is also used to obtain the target real-time business data corresponding to the decision request event in an external system; The rule evaluation engine 302 is also used to generate decision suggestions for the target decision node based on the target real-time business data and the business rules, and publish the decision suggestions to the event bus. The decision suggestions include the approval result of the target decision node, the next node to be executed in the preset business process, or the node adjustment status of the preset business process. The process execution engine 301 is also used to perform operations corresponding to the decision suggestions published in the event bus by the execution rule evaluation engine 302 according to the execution interface.

[0075] Furthermore, as a response to the above Figure 2 In addition to the implementation of the method shown, this embodiment of the invention also provides another process execution system for the above-described process. Figure 2 The system is implemented using the method shown. This system embodiment corresponds to the aforementioned method embodiment. For ease of reading, this system embodiment will not repeat the details of the aforementioned method embodiment, but it should be understood that the system in this embodiment can implement all the contents of the aforementioned method embodiment. For example... Figure 4 As shown, the system includes: The process execution engine 301 is used to execute the corresponding preset business process according to the business request, and when the target decision node in the preset business process is executed, the corresponding decision request event is generated and published to the event bus of the process execution system. The preset business process is a process defined in advance according to the business process model. The rule evaluation engine 302 is used to listen to the decision request event published in the event bus by the process execution engine 301 through a preset subscription mechanism, and determine the business rule corresponding to the target decision node in the preset rule base. The business rule is defined according to the domain-specific language model. The rule evaluation engine 302 is also used to obtain the target real-time business data corresponding to the decision request event in an external system; The rule evaluation engine 302 is also used to generate decision suggestions for the target decision node based on the target real-time business data and the business rules, and publish the decision suggestions to the event bus. The decision suggestions include the approval result of the target decision node, the next node to be executed in the preset business process, or the node adjustment status of the preset business process. The process execution engine 301 is also used to perform operations corresponding to the decision suggestions published in the event bus by the execution rule evaluation engine 302 according to the execution interface.

[0076] In one optional implementation, when the rule evaluation engine 302 generates decision suggestions for the target decision node based on the target real-time business data and the business rules, it includes: The rule arbitration module 3021 is used to select the target business rule from the multiple business rules according to a preset conflict arbitration strategy if there are multiple business rules. The decision recommendation module 3022 is used to generate decision recommendations for the target decision node based on the target real-time business data and target business rules using the rule evaluation engine.

[0077] In one optional implementation, the suggestion decision module 3022 is specifically used for: Determine whether the judgment conditions in the target business rule are supported by the target real-time business data; If not, then a decision suggestion for the target decision node is generated according to the preset processing strategy; If so, the target business rules are executed based on the target real-time business data, and decision suggestions are output. The decision suggestions include: whether the target decision node has been approved, and whether the next node to be executed or a non-essential node to be skipped after approval.

[0078] In one optional implementation, when the decision-making module 3022 generates decision suggestions for the target decision node according to a preset processing strategy, it is specifically used for: The target real-time business data is evaluated based on a preset risk model to obtain a risk value; If the risk value is less than the preset risk threshold, then based on the target real-time business data and the preset rule generation template, a new business rule for evaluating the target decision node is generated. The new business rules are used to evaluate the target's real-time business data to obtain the approval result for the target decision node; The approval result and the manual review node to be inserted are jointly determined as the decision suggestion of the target decision node, wherein the manual review node is used to confirm the approval result after the fact.

[0079] In an optional implementation, after obtaining the new business rule, the rule evaluation engine 302 is further configured to: Obtain the mapping table between process nodes and rules in the preset rule base; The correspondence between the target decision node and the business rule in the preset rule base is updated according to the new business rule.

[0080] In one optional implementation, the rule arbitration module 3021 is specifically used for: Obtain the preset priority value of each business rule and the number of constraints in each business rule; Sort the preset priority values ​​corresponding to multiple business rules to obtain the priority sorting result; Select the business rule with the highest preset priority value from the priority ranking results as the candidate business rule; If there are multiple candidate business rules, the candidate business rule with the most constraints is selected as the target business rule. If there are still two or more candidate business rules with the same number of constraints, the target business rule is determined by using multiple candidate business rules according to the rule fallback strategy.

[0081] In one optional implementation, when the rule arbitration module 3021 determines the target business rule using multiple candidate business rules according to the rule fallback strategy, it is specifically used for: In the historical decision database, the cumulative number of times each candidate business rule has been used is recorded; Obtain the number of business dimension types covered by each candidate business rule; The adaptation coefficient of each candidate business rule is determined based on the cumulative number of uses and the number of business dimension types. The candidate business rule with the highest adaptation coefficient is determined as the target business rule.

[0082] Furthermore, embodiments of the present invention also provide a storage medium for storing a computer program, wherein the computer program, when running, controls the device where the storage medium is located to execute the above-described... Figure 1-2 The method described in [the document / article].

[0083] Furthermore, embodiments of the present invention also provide a processor for running a program, wherein the program executes the above-described... Figure 1-2 The method described in [the document / article].

[0084] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0085] It is understood that the relevant features in the above methods and apparatus can be referenced interchangeably. Furthermore, the terms "first," "second," etc., in the above embodiments are used to distinguish between embodiments and do not represent the superiority or inferiority of any particular embodiment.

[0086] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0087] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0088] In addition, the memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0089] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0090] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will 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 program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0091] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0092] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0093] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0094] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0095] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0096] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0097] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0098] The above are merely embodiments of this application and are not intended to limit the scope of 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 scope of the claims of this application.

Claims

1. A process execution method, characterized in that, The method is applied to a process execution system, which includes a process execution engine, a rule evaluation engine, and an event bus. The method includes: The process execution engine is used to execute the corresponding preset business process according to the business request, and when the target decision node in the preset business process is executed, a corresponding decision request event is generated and published to the event bus. The preset business process is a process defined in advance according to the business process model. The rule evaluation engine uses a preset subscription mechanism to listen for decision request events in the event bus and determines the business rules corresponding to the target decision node in the preset rule base. The business rules are defined according to a domain-specific language model. The rule evaluation engine is used to obtain the target real-time business data corresponding to the decision request event from an external system. The rule evaluation engine generates decision suggestions for the target decision node based on the target real-time business data and the business rules, and publishes the decision suggestions to the event bus. The decision suggestions include the approval result of the target decision node, the next node to be executed in the preset business process, or the node adjustment status of the preset business process. Using the process execution engine, the operation corresponding to the decision suggestion is executed according to the execution interface.

2. The method according to claim 1, characterized in that, The rule evaluation engine generates decision recommendations for the target decision node based on the target's real-time business data and the business rules, including: If there are multiple business rules, the target business rule is selected from among the multiple business rules according to the preset conflict arbitration strategy; The rule evaluation engine is used to generate decision recommendations for the target decision node based on the target real-time business data and target business rules.

3. The method according to claim 2, characterized in that, The rule evaluation engine generates decision recommendations for the target decision node based on the target real-time business data and target business rules, including: Determine whether the judgment conditions in the target business rule are supported by the target real-time business data; If not, then a decision suggestion for the target decision node is generated according to the preset processing strategy; If so, the target business rules are executed based on the target real-time business data, and decision suggestions are output. The decision suggestions include: whether the target decision node has been approved, and whether the next node to be executed or a non-essential node to be skipped after approval.

4. The method according to claim 3, characterized in that, Decision suggestions for the target decision node are generated according to a preset processing strategy, including: The target real-time business data is evaluated based on a preset risk model to obtain a risk value; If the risk value is less than the preset risk threshold, then based on the target real-time business data and the preset rule generation template, a new business rule for evaluating the target decision node is generated. The new business rules are used to evaluate the target's real-time business data to obtain the approval result for the target decision node; The approval result and the manual review node to be inserted are jointly determined as the decision suggestion of the target decision node, wherein the manual review node is used to confirm the approval result after the fact.

5. The method according to claim 4, characterized in that, After obtaining the new business rules, the method further includes: Obtain the mapping table between process nodes and rules in the preset rule base; The correspondence between the target decision node and the business rule in the preset rule base is updated according to the new business rule.

6. The method according to claim 2, characterized in that, Based on a preset conflict arbitration strategy, the target business rule is selected from multiple business rules, including: Obtain the preset priority value of each business rule and the number of constraints in each business rule; Sort the preset priority values ​​corresponding to multiple business rules to obtain the priority sorting result; Select the business rule with the highest preset priority value from the priority ranking results as the candidate business rule; If there are multiple candidate business rules, the candidate business rule with the most constraints is selected as the target business rule. If there are still two or more candidate business rules with the same number of constraints, the target business rule is determined by using multiple candidate business rules according to the rule fallback strategy.

7. The method according to claim 6, characterized in that, Based on the rule-based fallback strategy, the target business rule is determined using multiple candidate business rules, including: In the historical decision database, the cumulative number of times each candidate business rule has been used is recorded; Obtain the number of business dimension types covered by each candidate business rule; The adaptation coefficient of each candidate business rule is determined based on the cumulative number of uses and the number of business dimension types. The candidate business rule with the highest adaptation coefficient is determined as the target business rule.

8. A process execution system, characterized in that, The system includes: The process execution engine is used to execute the corresponding preset business process according to the business request, and when the target decision node in the preset business process is executed, the corresponding decision request event is generated and published to the event bus of the process execution system. The preset business process is a process defined in advance according to the business process model. The rule evaluation engine is used to listen for decision request events published by the process execution engine in the event bus through a preset subscription mechanism, and determine the business rules corresponding to the target decision node in the preset rule base. The business rules are defined according to a domain-specific language model. The rule evaluation engine is also used to obtain the target real-time business data corresponding to the decision request event from external systems; The rule evaluation engine is also used to generate decision suggestions for the target decision node based on the target real-time business data and the business rules, and publish the decision suggestions to the event bus. The decision suggestions include the approval result of the target decision node, the next node to be executed in the preset business process, or the node adjustment status of the preset business process. The process execution engine is also used to perform operations corresponding to decision recommendations published by the engine in the event bus based on the execution interface rules.

9. A storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, it controls the device where the storage medium is located to perform the process execution method as described in any one of claims 1 to 7.

10. A processor, characterized in that, The processor is used to run a program, wherein the program executes the process execution method as described in any one of claims 1 to 7.

Images