posted on 2015-05-08, 14:39authored byJ. Xu, Stephan Reiff-Marganiec
The introduction of of web services has led to web service composition being a focus of many researchers. Composing web services using workflows is seen as the most realistic method from an industrial viewpoint. Amongst other method, the use of natural computing methods has been proposed previously to automate web service composition. The need for a fast response when computing the most suitable sequence of services is addressed in this paper. 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 the 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. We also provide a schema analysis for this method.
Funding
This work is supported by the Fundamental Research Funds for the Central Universities (No.14CX06007A, No.13CX06009A).
History
Citation
Chinese Journal of Electronics, 2014, 23 (3), pp. 579-585 (7)
Author affiliation
/Organisation/COLLEGE OF SCIENCE AND ENGINEERING/Department of Computer Science
Version
AM (Accepted Manuscript)
Published in
Chinese Journal of Electronics
Publisher
Chinese Institute of Electronics, Tsinghua University, Peking University, Institute of Semiconductors of Chinese Academy of Sciences, University of Electronic Science and Technology of China, Xidian University, the Chinese University of Hong Kong, Sh