An outranking model for web service discovery

The web service discovery is the cornerstone of the service oriented architecture. To solve this issue, we usually leverage a matching model as well as the operation signature in order to minimize the residual errors. In this paper, we resolve this problem by combining a set of similarity measures through the use of a majority voting model called “outranking”. The Experimental evaluation confirms that this model performs better than the well-known Borda and all input similarity measures.