Towards Heuristic Web Services Composition Using Immune Algorithm

Abstract

One of the main benefits of web services is the dynamic composability, however how to achieve this is one of the current research challenges. Web service composition has been studied and, amongst other methods, the use of natural computing methods has been proposed previously. In this paper, we address the need for a fast response when computing the most suitable sequence of services. In particular, we propose a novel heuristic immune algorithm with an efficient encoding and mutation method. The algorithm involves two steps: an immune selection operation, which is maintaining antibody population diversity and a clonal selection. The use of a vaccine during the evolution provides heuristic information that accelerates the convergence. Our experimental results illustrate that the proposed heuristic immune algorithm is very effective in improving the convergence speed.