摘要

Data-aware processes play a crucial role in various IT systems, including requirement elicitation, domain analysis, software design, and system execution. Due to frequent changes in business environments and continual internal adjustments of enterprises, data-aware processes are increasingly evolved into multiple process variants. The detection of differences between variants can be related to process mapping, process integration, or process substitution. A critical step of the procedure is to investigate the data-aware process consistency. Unfortunately, existing studies only provide a simple "yes" or "no" answer or look for an answer purely from the control flow perspective. The objective of this paper is to propose a systematic solution for effective measurement of consistency between data-aware processes. First, we identify essential activity constraints which reside in data-aware processes. Then, we introduce a novel concept of activity constraint graph (ACG) and propose an algorithm for constructing ACGs. Finally, we use ACGs to measure the data-aware process consistency on a scale from 0 to 1. Our technique has been implemented in a prototype tool, and extensive experiments using both real and synthetic datasets are conducted to evaluate the accuracy, distribution of consistency degrees, and capacity of difference detection of our approach. Results show that our approach is more accurate, generates a finer distribution of consistency degrees, and detects differences more effectively than other state-of-the-art approaches.

全文