Towards a Multi-label Classification of Open Educational Resources

Nowadays, there are a lot of online repositories containing thousands of very useful educational resources for the educational community. To take full advantage of these resources requires a simple, direct and effective access to those resources that are of interest, therefore, it is necessary that those resources are ordered or ranked based on some criteria?-that is to say, they have to be classified. Classification is usually done manually by the resource provider, which directly implies a main problem: the time spent categorising resources. In this paper we propose a solution to this problem through the design and implementation of a multi-label classifier that enables the automatic classification of a set of educational resources in their most suitable category or categories, thus eliminating the need to manually perform this classification. Evaluation results show that the performance of the OER classifier is comparable to classification of a de-facto standard corpus: OHSUMED.