Firefly-based business process optimization

Abstract

This paper presents a firefly-based method for business process optimization. Each artificial firefly has a candidate business process associated which is modeled as a causal matrix. The evolution of a candidate business process is achieved using genetic operators which aim to perform structural and resource allocation modifications. To establish whether a candidate business process is optimal, the cost and the execution time of the process are evaluated using a penalty-based fitness function. The firefly-based method has been evaluated on a set of generic scenarios and its performance has been analyzed using the fitness graph evolutionary measure.