Flexibility Enhancements in BPM by applying Executable Product Models and Intelligent Agents

Abstract

In this paper we present an alternative approach to model and execute business processes. The approach combines a compact model with an intelligent control flow mechanism. Instead of using a pre-designed process model that is executed dur- ing runtime by a workflow engine, we use a special model called executable product model (EPM) that is executed by a multi-agent system. The EPM provides a compact representation of the set of possible execution paths of a business process by defining information dependencies instead of the order of activities. In our approach the ap- plication of intelligent agents takes advantage of the flexibility provided by the EPM. Relational reinforcement learning (RRL) with a genetic algorithm (GA) is applied for managing the control flow. In experiments we show that business processes can be executed successfully with our approach and that the application of machine learning leads to significant performance gains.