Performance evaluation of the novel DSS based on humanware service

Currently, service-oriented architecture (SOA) is implemented by using web service technology. SOA-based decision support systems (DSS) can solve structured problems, but cannot solve semi-(un-)structured problems satisfactorily. To enhance the capability of SOA-based DSS, we present the innovative idea of “human as a service” and the concept of Humanware Service. A novel DSS based on Humanware Service is established. The proposed DSS integrates human intelligence and computer intelligence. Each activity of a problem solving process is implemented by Software or Humanware Service. A Generalized Stochastic Petri Net (GSPN) model is proposed to describe the service invocation process of an activity. And the response time, reliability and availability of an activity are able to be figured out by means of the model. Furthermore, a reduced model of an activity is presented, which is useful to evaluate the performance of the problem solving process. At last, an illustrative example shows the efficiency of the proposed DSS.