Genetic algorithm based approach to personal worklist resource scheduling

In workflow management systems (WFMSs), appropriate consideration of applying scheduling techniques to manage actors' personal worklists is essential for successful implementation of workflow technology. Mainly, the attention of existing workflow scheduling has focused on the process perspective. As a result, issues associated with personal worklist's perspective, i.e., worklists that contain actors' to-do activity instances, have been largely neglected. Given this motivation, this paper for the first time, investigates issues in personal worklist scheduling under dynamic workflow environment. Towards these issues, the paper proposes a novel genetic algorithm to optimize the personal worklist management. This algorithm recommends for each actor a feasible worklist that will ensure the worklist's activity instances' successful execution while minimizing the total overtime costs for all personal worklists. Through comparing with other well-known workflow scheduling algorithms, the paper evaluates the effectiveness of the proposed genetic algorithm with a specific example and a simulation experiment. © 2012 ISCAS.