Abstract—Abstract—There has been a paradigm shift in the design of software applications. Each offering is designed as a service that can easily integrate with other services. As a result, there are large numbers of web services that are constantly created and consumed. One of the key challenges is to create relevant sub-space domain for web services that can identify the best suitable service to perform a task. Web service discovery process addresses the problem of selecting the best service. In this paper, we propose semi-supervised service community discovery on heterogeneous evolving web services data. The semi-supervised knowledge about the current time step is incorporated into the heterogeneous evolving environment. The evolving changes in the web services are captured and service community is created based on their prior and current knowledge of the web services. Also, the heterogeneous model helps us to create a highly relevant evolving sub space taking into account the operations and services performed by the web services along with the terms used to define the web service simultaneously.
Index Terms—Semi-supervised, heterogeneous, co-clustering, evolving data, web services.
Pankaj Andhale is with Intuit Inc. Mountain View, CA 94043 USA (e-mail: pankaj_andhale@intuit.com). Manjeet Rege and Qi Yu are with College of Computing and Information Sciences at Rochester Institute of Technology, NY 14623 USA (e-mail: mr@cs.rit.edu; qi.yu@rit.edu).
Cite:Pankaj Andhale, Manjeet Rege, and Qi Yu, "Discovering Heterogeneous Evolving Web Service Communities Using Semi-Supervised Non Negative Matrix Factorization," International Journal of Machine Learning and Computing vol. 3, no. 2, pp. 201-205, 2013.