Research on Geo\|spatial Web Services Classification Based on Manifold Learning

The problems existed in the traditional methods of Web services classification are analyzed, the concepts of manifold and manifold learning and the purpose of introducing the manifold learning into the Web services are described. The algorithm for the visualization and classification of geo\|spatial Web services(GWS) based on manifold learning is proposed.During the process of dimension reduction, the similarity between GWS is preserved and the data manifold is unrolled. In order to improve the precision of classification, we gain the mapping rule from the GWS to the 2D data and the initial number of clusters according to the visualization of 2D mapping data. The experimental results prove the validity of the improved visualization and classification algorithm for GWS proposed in this paper.