Analysis Of Mashup-oriented Web API Ecological Networks

With the development of the Internet and technology of Web services,more and more software services providing application programming interface(API)for peripheral equipment start to appear on the Internet.Developers can develop rapidly various kinds of applications by calling and synthetizing opening services and API on the Internet.Due to value-added effect brought by opening services and API,more and more companies release their own Web API and provide all kinds of information and data services,such as Google and Baidu.Meanwhile,Mashup applications of various Web API become popularly and attracting a wide range of attention of developers.On the Internet,Mashup applications and Web API have formed a kind of ecological system.There are many characteristics in the Web API ecological system,such as large scale,complex of relationship and dynamics.At present,we haven't understood deeply about its internal structure and evolution mechanism.In order to reveal the structure characteristic of Web API,analyzing the influence and relationships of different Web API and helping users to find appropriate Web API quickly.In this paper,we have researched intense on the Web API ecological system.The main contributions of this paper can be summarized as follows:1)Base on the Web API and Mashup data information from Programmable Web.com,we have proposed an approach to construct a collaborative relationship network and similar relationship network of Web API.We have also analyzed the basic structures characteristic of these two relationship networks,such as the number of nodes,the number of edges,average degree,degree distribution,aggregation coefficient,match and so on.Besides,we have analyzed the strong Web API pair of collaborative relationship network and similar relationship network of Web API.2)Based on the collaborative relationship network and similar relationship network of Web API,we have infered a cooperation relationship network and collaborative relationship network of Web API providers.At the same time,we have analyzed the structures of two types of relationship networks.3)As for the collaborative relationship network of Web API and relationship network of Web API providers,we have used the K-core decomposing technique to analyzethe level structures of two types of relationship networks and identifying the leve of different Web API.The work of this paper can not only comprehend the structures and characteristic of the Web API ecological system better,but also provide a support for improving Web API discovery,recommendation and composition in using its structural knowledge and features.