Research on Geo-spatial Web Services Classification Based on Manifold Learning
Xie Mingxia · Wuhan Daxue xuebao. Xinxi kexue ban · 2013
The problems existed in the traditional methods of Web services classification areanalyzed,the concepts of manifold and manifold learning and the purpose of introducing themanifold learning into the Web services are described.The algorithm for the visualizationand 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 thedata manifold is unrolled.In order to improve the precision of classification,we gain themapping rule from the GWS to the 2Ddata and the initial number of clusters according to thevisualization of 2Dmapping data.The experimental results prove the validity of the improvedvisualization and classification algorithm for GWS proposed in this paper.