Service Composition Design Pattern for Autonomic Computing Systems Using Association Rule Based Learning and Service-Oriented Architecture

In this paper we will compose the design patterns which will satisfy the properties of autonomic computingsystem: for the Decision-Making phase we will introduce Case-Based Reasoning design pattern, and forReconfiguration phase we will introduce Reactor design pattern. The most important proposal in ourcomposite design pattern is that we will use the Association Rule Learning method of Data Mining to learnabout new services that can be added along with the requested service to make the service as a dynamiccomposition of two or more services. Then we will include the new service as an aspectual feature modulecode without interrupting the user.As far as we know, there are no studies on composition of designpatterns and pattern languages for autonomic computing domain. We will authenticate our work by asimple case study work. A simple Class and Sequence diagrams are depicted.