Web Service Network Analysis

A Web Service is a software component that is accessible on the Web. Web Services can collaborate to form composite services, which are called Mashups. Services and Mashups interconnect with each other, forming a complex network. This technological network evolves in a large scale without central control. As a new type of software network, there is an urgent need to study its topological properties. This work studies the Web Service Network that consists of 4255 primitive web services and mashups collected from ProgrammableWeb.com. We study various centralities of the network, including degree, betweenness, closeness and pagerank. We find that it is a scale-free network, whose degree distributions follow a power law. We also identify the top web services according to these centrality measures, and demonstrate the correlations between them.