Title: Predicting new composition relations between web services via link analysis
Authors: Mingdong Tang; Fenfang Xie; Wei Liang; Yanmin Xia; Kuan-Ching Li
Addresses: School of Information Science and Technology, Guangdong University of Foreign Studies, Guangzhou, 510006, China ' School of Data and Computer Science, Sun Yat-sen University, Guangzhou, 510006, China ' School of Opto-Electronics and Communication Engineering, Xiamen University of Technology, Xiamen, 361024, China ' School of Computer Science and Engineering, Hunan University of Science and Technology, Xiangtan, 411201, China ' Department of Computer Science and Information Engineering (CSIE), Providence University, Taichung City, 43301, Taiwan
Abstract: With the wide application of service-oriented architecture (SOA) and service-oriented computing (SOC), the past decade has witnessed a rapid growth of the number of web services on the internet. Against this background, combining different web services to create new applications has attracted great interest from developers. However, according to latest statistics, only a small number of popular services are frequently used by developers and the use rates of most web services are rather low. To help service users discover appropriate web services and promote service compositions has thus become a significant need. In the following paper, we propose a link-based approach to predict new composition relations between web services. The approach is based on exploration of known composition relations and similarity relations among web services. To measure the composition or similarity degrees, several link-based methods are exploited, and two reasonable heuristic rules for integrating the existing composition and similarity relations for service composition prediction are developed. Case studies and experiments based on real web service datasets validated the proposed approach.
Keywords: service composition; composition relations; similarity relations; link prediction; web services; API; Mashup.
International Journal of Computational Science and Engineering, 2019 Vol.20 No.1, pp.88 - 101
Received: 20 Jan 2018
Accepted: 26 Mar 2018
Published online: 23 Oct 2019 *