A semantic framework supporting business process variability using event logs

Large organizations often have multiple branches situated in different locations, each branch may collaborate and learn from other branches' experience. Their Business processes (BPs) share often similar business goals and are slightly different. These branches are eager to develop new process variants to satisfy new requirements. Process execution logs, so called process event logs, can be used to analyze requirement changing situations and efficiently develop BP variants. However, these logs often have heterogeneous data-sources which prevent an easy and dynamic interoperability between different branches. In this paper, we propose a semantic framework tackling this heterogeneity issue. This framework promotes the creation of a semantic knowledge base from process event logs. Using this knowledge base, we offer BP designers the means to discover suitable BP fragments to assist process variant modeling. We performed experiments on a large public dataset and experimental results show that our approach is feasible and accurate in realistic situations.