A Scalable and Near-Optimal Conformance Checking Approach for Long Traces

The paper presents a novel sliding window method for efficient conformance checking of long process traces, reducing computational complexity by segmenting traces into smaller subtraces and using global information to guide alignment decisions, thereby addressing the scalability issues inherent in process mining for long event logs.

Abstract

Long traces and large event logs that originate from sensors and prediction models are becoming more common in our data-rich world. In such circumstances, conformance checking, a key task in process mining, can become computationally infeasible due to the exponential complexity of finding an optimal alignment. This paper introduces a novel sliding window approach to address these scalability challenges while preserving the interpretability of alignment-based methods. By breaking down traces into manageable subtraces and iteratively aligning each with the process model, our method significantly reduces the search space. The approach uses global information that captures structural properties of the trace and the process model to make informed alignment decisions, discarding unpromising alignments even if they are optimal for a local subtrace. This improves the overall accuracy of the results. Experimental evaluations demonstrate that the proposed method consistently finds optimal alignments in most cases and highlight its scalability. This is further supported by a theoretical complexity analysis, which shows the reduced growth of the search space compared to other common conformance checking methods. This work provides a valuable contribution towards efficient conformance checking for large-scale process mining applications.

Method

The authors used a novel sliding window methodology for conformance checking of long process traces. This involved dividing long traces into smaller, manageable subtraces and iteratively aligning each subtrace with the process model. By doing so, they significantly reduced the search space and computational complexity. The approach also incorporated the use of global information to inform alignment decisions, which improved the accuracy of the conformance checking. The authors supported their methodology with a theoretical complexity analysis and empirical evaluations using various datasets, including those with very long traces from food preparation datasets.

Main Finding

The authors discovered that their novel sliding window approach to conformance checking is highly effective for long process traces. They found that by segmenting long traces into subtraces and aligning them iteratively, they could significantly reduce the search space and computational complexity. The use of global information to guide alignment decisions further enhanced the accuracy of the results. Theoretical complexity analysis showed that the proposed method has superior scalability compared to traditional conformance checking techniques. Empirical evaluations using various datasets, including those with very long traces from food preparation datasets, demonstrated that the method consistently finds near-optimal alignments, even for traces with thousands of events, within reasonable time frames. This addresses a major challenge in process mining and contributes to the field by providing a scalable solution for long event logs.

Conclusion

The conclusion of the paper is that the authors have successfully developed a scalable and efficient sliding window approach for conformance checking in process mining, which is particularly effective for long traces generated by sensors and prediction models. The method's ability to segment long traces into subtraces and align them iteratively, while leveraging global information to guide alignment decisions, significantly reduces computational complexity and improves accuracy. Theoretical analysis and empirical evaluations have demonstrated the method's scalability and near-optimal performance, even for traces with thousands of events. This represents a valuable contribution to the field of process mining by addressing the challenge of conformance checking for long event logs.

Keywords

Conformance Checking, Long Traces, Sliding Window, Process Mining, Scalability, Event Logs, Alignment-Based Methods, Search Space Reduction, Global Information, Iterative Alignment, Theoretical Complexity Analysis, Empirical Evaluation, Near-Optimal Alignments, Large-Scale Process Mining Applications, Sensor Data, Prediction Models, Trace Segmentation, Subtrace Alignment, Process Model Integrity, Computational Efficiency, Alignment Accuracy, Hyperparameters, Synchronous Product, Cost Function, Optimal Alignment, Subtrace Model, Partial Optimal Alignment, Algorithmic Design, Sliding Window Mechanism, Iterative Alignment, Complexity Analysis, Empirical Evaluation, Related Work, Future Research Directions.