Discovering the best web service: A neural network-based solution
Source of Publication
Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics
Differentiating between Web services that share similar functionalities is becoming a major challenge into the discovery of Web services. In this paper we propose a framework for enabling the efficient discovery of Web services using Artificial Neural Networks (ANN) best known for their generalization capabilities. The core of this framework is applying a novel neural network model to Web services to determine suitable Web services based on the notion of the Quality of Web Service (QWS). The main concept of QWS is to assess a Web service's behaviour and ability to deliver the requested functionality. Through the aggregation of QWS for Web services, the neural network is capable of identifying those services that belong to a variety of class objects. The overall performance of the proposed method shows a 95% success rate for discovering Web services of interest. To test the robustness and effectiveness of the neural network algorithm, some of the QWS features were excluded from the training set and results showed a significant impact in the overall performance of the system. Hence, discovering Web services through a wide selection of quality attributes can considerably be influenced with the selection of QWS features used to provide an overall assessment of Web services. ©2009 IEEE.
Al-Masri, Eyhab and Mahmoud, Qusay H., "Discovering the best web service: A neural network-based solution" (2009). Scopus Indexed Articles. 2237.