Object Centric Process Mining Algorithm for Multi-Entity Data Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional process mining technologies flatten multi-dimensional execution data into a single dimension by selecting only one entity for analysis, leading to incomplete, inaccurate, and duplicate data, which results in incorrect metrics and loss of behavioral insights.
Innovation Solution
An end-to-end object-centric process mining algorithm that considers all entities of a process to generate a process graph, by creating object networks, determining transitions between events, and weighting edges based on frequency counts, thereby preserving the multi-dimensional nature of the data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If only one entity is selected for analysis to generate the event log, then the analysis process is simplified, but the data becomes incomplete, inaccurate, and loses multi-dimensional insights
Solution Approach 1:
The patent segments the process analysis into multiple entity-specific views while maintaining an integrated event log. Each entity (order, delivery, invoice) is analyzed separately to preserve its unique behavioral patterns, then combined to form a comprehensive process model that retains multi-dimensional information without overwhelming complexity
Solution Approach 2:
The patent adds entity dimensionality to the traditional event log by incorporating entity-specific attributes and relationships. Instead of flattening all data into a single dimension, the system maintains multi-dimensional structure through entity contexts, allowing comprehensive analysis while managing complexity through structured organization
2Productivity
If only one entity is selected for analysis, then the processing is faster, but incorrect metrics and duplicate data are generated
Solution Approach 1:
The system implements feedback mechanisms where entity-specific analyses inform and correct the overall process metrics. By continuously validating events against multiple entity contexts, the system identifies and corrects metric inaccuracies and duplicates, ensuring precise measurements while maintaining efficient processing through iterative refinement
Solution Approach 2:
The patent performs preliminary entity validation and relationship mapping before generating the event log. By pre-processing entity data to establish correct relationships and filters, the system avoids generating incorrect metrics and duplicates during main processing, thereby maintaining both speed and accuracy
3Loss of information
If all entities are considered for process mining, then complete and accurate insights are achieved, but the complexity of generating the process graph increases
Solution Approach 1:
The patent segments the process graph generation into entity-specific sub-graphs that are then integrated. Each entity contributes its relevant processes and relationships to the overall process model, allowing complete coverage of all entities while managing complexity through modular construction and systematic integration
Data Source
AI summary
Systems and methods for object centric process mining are provided. Execution data of a process having a plurality of entities is received. A plurality of object networks representing relationships between objects of the plurality of entities are generated based on the execution data. A set of transitions is determined for each of the plurality of object networks. A process graph of execution of the process is generated based on the sets of transitions. The process graph is output.