A Method For Optimizing Top-K Composite Services Towards Preference-Aware Service Dominance
Shaoqian Zhang,Wenmin Lin,Wanchun Dou,Jinjun Chen
DOI: https://doi.org/10.4018/jwsr.2013040104
2013-01-01
International Journal of Web Services Research
Abstract:Web service composition allows users to create value-added composite Web services on existent services, where top-k composite services are helpful for users to find a satisfying composite service efficiently. However, with an increasing number of Web services and users' various composition preferences, computing top-k composite services dynamically for different users is difficult. In view of this challenge, an optimization method for top-k composite services is proposed, based on a preference-aware service dominance relationship. Concretely speaking, firstly, user preferences are modeled with the preference-aware service dominance. Then, in local service selection, a multi-index based algorithm is proposed, named Multi-Index, for computing candidate services of each task dynamically. After that, in global optimization, combined with a service lattice, top-k composite services are selected under a dominant number-aware service ranking. A case study is also presented for illustrating the authors' solution. At last, an experiment was conducted to verify the proposed method.