Quality improvement in web services using function replication

Service identification is the first and the most important step in service-oriented development and improving the quality of identified services is an important issue. In this paper, a method for replicating functions in identified services is proposed. In this method, after clustering the functions and defining each cluster as a service, the quality of each service is calculated using Turbo-MQ fitness function. Then, functions that could increase the value of fitness function are identified and replicated in the target cluster. Function replication might cause new inter- and intra-cluster relations; also, creating a replica for a function has extra cost in terms of implementation and maintenance. Therefore, after considering these challenges, a useful function replication will increase service quality and service call efficiency in a whole system. This method can be added to all business process-based service identification methods, and improve the accuracy of these methods.