Dynamic Scoring-Based sequence alignment for process diagnostics

Abstract

Even though process-aware information systems are intensively utilized in the organizations, traditional process management paradigms majorly concentrate on the design and configuration phases. Instead of starting with a process design, process mining attempts to discover interesting patterns from process enactment namely event logs and extract business processes by distilling these event logs as knowledge base. One of the challenging issues in process mining domain is process diagnostics, which is complex and sometimes infeasible, especially when dealing with real-time, flexible and unstructured processes. In this aspect sequence alignment is applicable to find out common subsequences of activities in event logs that are found to recur within the process instances emphasizing some domain significance. In this study, we focus on a hybrid quantitative approach for performing process diagnostics, i.e. comparing the similarity among process models based on dominant behavior concept, confidence metric and Needleman-Wunsch algorithm with dynamic pay-off matrix.