基于模拟退火算法的业务流程上下文提取方法

Process mining can extract human resource usage in business processes to support employee management and process improvement. Often, resources from the same organization exhibit similar characteristics in process execution, such as performing the same set of process activities or participating in the same type of case. This is the natural result of the division of labor within the organization. While these characteristics can be modeled in different process dimensions, this paper uses the concept of execution context to describe resource groups and proposes a method to automatically learn execution context from event logs. This method encodes the classification rules that define the execution context, defines the objective functions and constraints, configures the search parameters, uses the simulated annealing algorithm to get the classification rules, and then gets the execution context. This approach contributes to a more comprehensive understanding of resource behavior and organizational processes, and provides more in-depth support for business process management and resource optimization. After the evaluation of two real business process event log data, the performance of this method is superior in process context extraction, and it has high practicability.