Automatic Topic(s) Identification from Learning Material: An Ontological Approach

The ability to judge the relevance of topics and related sources in information-rich environments is a key to success when scanning online learning environments. A Learner may be looking for learning materials explaining given topic or exercises on the topic. Any learning material may cover multiple topics related to multiple subject domains. This paper presents ontological approach for identifying major topics, covered in the learning material. Along with the topics, the subject and discipline to which those topics belongs to and relevance of the topic in the learning material as compared to other topics present in the same document are also discovered. Domain ontology is developed to retrieve the topics covered in the document. We present an evaluation against a manually categorized topics as well as author's judgment of relevance of the topics discovered by our system. Evaluation results show that the technique presented by us is effective in identifying topics and subtopics covered in a single learning document. This work is part of research conducted on developing a web service for automatic semantic annotation of learning material.