Improving process models by discovering decision points

  • Authors:

  • Sharmila Subramaniam;Vana Kalogeraki;Dimitrios Gunopulos;Fabio Casati;Malu Castellanos;Umeshwar Dayal;Mehmet Sayal

  • Affiliations:

  • Computer Science and Engineering Department, University of California, Riverside, CA 92521, USA;Computer Science and Engineering Department, University of California, Riverside, CA 92521, USA;Computer Science and Engineering Department, University of California, Riverside, CA 92521, USA;Department of Information and Communication Technology, University of Trento, via sommarive 14, 38050 Povo (Trento), Italy;HP Labs, 1501 Page Mill Road, Palo Alto, CA 94304, USA;HP Labs, 1501 Page Mill Road, Palo Alto, CA 94304, USA;HP Labs, 1501 Page Mill Road, Palo Alto, CA 94304, USA

  • Venue:
  • Information Systems
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Workflow management systems (WfMS) are widely used by business enterprises as tools for administrating, automating and scheduling the business process activities with the available resources. Since the control flow specifications of workflows are manually designed, they entail assumptions and errors, leading to inaccurate workflow models. Decision points, the XOR nodes in a workflow graph model, determine the path chosen toward completion of any process invocation. In this work, we show that positioning the decision points at their earliest points can improve process efficiency by decreasing their uncertainties and identifying redundant activities. We present novel techniques to discover the earliest positions by analyzing workflow logs and to transform the model graph. The experimental results show that the transformed model is more efficient with respect to its average execution time and uncertainty, when compared to the original model.