Semi-automated Semantic Annotation of Learning Resources by Identifying Layout Features

Abstract. It is now widely accepted that any kind of digital content must be somehow semantically annotated to be intelligently used by computer programs. Annotations can be metadata, descriptions, etc. When dealing with learning, most systems require the author to manually annotate resources so that the system can deploy a navigation strategy, an adaptive behavior etc. However this task is very problematic, and often reveals to be an overwhelming enterprise. In this paper we propose a methodology, based on the reuse of existing pedagogical documents to achieve a semi-automated extraction of semantic annotations by identifying semantic information contained in the layout. We are applying this methodology in the design of a Web Based Learning System