A Metaheuristic Search Framework to Derive Cancer Care Services from Business Process Models

Abstract

Cancer Care involves not only handling patients' medical or physical needs but also other services to facilitate patient needs which are underpinned by appropriate software systems that assist in patient care processes. The Service-Oriented Architecture (SOA) model of computing has become widely adopted and can provide efficient and agile business solutions in the face of rapid changes to business requirements. Instead of adopting a more traditional way of building an IT system for Cancer Care by rigidly piecing together a collection of hardware, software and networking, SOA offers the opportunity to build the IT systems in an increasingly flexible and reconfigurable way. However, current service identification methods can suffer from shortcomings such as a lack of computational support, and not being able to address all the necessary activities of the service identification. To address these shortcomings, this paper presents a comprehensive metaheuristic search framework for deriving SOA-based services applied to Cancer Care business process models. This framework is evaluated using both quantitative and qualitative methods with the help of domain experts at King Hussein Cancer Centre (KHCC), Jordan. Evaluation by domain experts confirmed that the resulting services are feasible (i.e., valid services that can be practically applied for real-life projects) that the domain experts might not have arrived at manually. Statistical analysis shows candidate services produced by the search-based framework are superior to the services produced manually by domain experts at KHCC with respect to metrics for coupling and cohesion.