Dependency Identification Engine for Process Mining

The patent segments the process model into two distinct components: a structural component (directly-follows graph showing sequence of events) and a semantic component (annotations representing dependencies between choices). This segmentation allows the model to visualize process structure while separately capturing dependency information that would otherwise be lost, resolving the contradiction between information preservation and model complexity.

The patent introduces an intermediary annotation layer that connects choices across different parts of the process without requiring direct structural modification. These annotations act as mediators that encode dependency relationships (such as whether one choice influences future choices) without complicating the underlying process structure, thus preserving dependency information while maintaining model simplicity.

The patent applies preliminary action by pre-defining a set of standardized annotation types that represent common dependency relationships between choices. These annotations are prepared in advance and can be systematically applied to process models, enabling accurate capture of semantic relationships without requiring complex ad-hoc modeling for each dependency scenario, thus improving reliability while controlling complexity.

The patent changes the parameter representation by adding dependency annotations as additional attributes to process model elements. Rather than fundamentally restructuring the process model to capture dependencies, the solution modifies existing parameters by appending semantic information (annotations) that indicate whether and how choices influence each other, thereby improving process understanding accuracy without significantly increasing model complexity.