Querying Process Models Based on the Temporal Relations between Tasks

Abstract

As business process management technology matures, organisations accumulate hundreds, even thousands of models whose management poses real challenges. One of these challenges concerns model retrieval where support should be provided for the formulation and efficient evaluation of business process model queries. Recent years, queries based on behavioral information attracted more and more attentions both in the academic and industrial fields. However, there is not a widely accepted formal language to express the behavioral requirements. And the efficiency of the behavioral queries is another challenge when the process models contains arbitrary loop and/or extra-large parallel structures. The complete finite prefix (CFP) has been widely used to express the partial-order semantics of a Petri net. In this paper, we first extend CFP by connecting the cut-off events to continuation events, we call it temporal-order preserving complete finite prefix (TPCFP), where all the temporal relations between tasks can be extracted. Then we employ linear temporal logic formulae as the behavioral query language, whose formal semantics are defined over the TPCFP. The related algorithms are implemented in BeehiveZ. Experimental results are investigated with real-life processes and artificial processes in detail.