Automatic Service Composition in Parallel with MapReduce

HUANG Long-tao,DENG Shui-guang,DAI Kang,LI Ying,YIN Jian-wei
DOI: https://doi.org/10.3969/j.issn.0372-2112.2012.07.018
2012-01-01
Abstract:How to compose services automatically and efficiently is a difficult issue,especially for a large number of services.Traditional methods based on single-computation usually come to inefficiency due to the explosion of the planning and searching space when the number of services grows too much.Based on the MapReduce framework,this paper proposes an automatic service composition method based on depth-first searching for the Top-k Qos service composition issue.The result from a serial of experiments indicates that the method can satisfy composition requirements quickly and efficiently even with a large-scale service repository.
What problem does this paper attempt to address?