An Approach for Automatic Discovery, Composition and Invocation of Semantics Web Services for Data Mining

Abstract

Nowadays, several educational institutions make use of e-Learning environments and other technologies to support the teaching and learning process. As a consequence, a large amount of data is generated from the many interactions of students, tutors, teachers and other actors involved in these environments. These data can be a great and important source of information, however, analyzing them is a complex and expensive task. One way to analyze such data properly is to apply Educational Data Mining (EDM) techniques, and thus to use the information obtained in decision making support. There are, however, several challenges in the application of mining in educational data. In particular, the integration challenge is complex because it involves different tools developed in different programming languages. Thus, we propose an approach for automatic discovery, composition and invocation of Semantic Web Services (SWS) for data mining based on a new semantic model. With this, we hope to contribute to a greater flexibility in the integration between data mining tools and e-Learning environments. In order to evaluate, we adopted a scenario-based method to evaluate quality attributes of performance and reliability of the proposed solution in these scenarios.