Information retrieval in educational structured documents adapted to learners needs

The web is increasingly moving towards structuring and to taking into consideration of semantics, particularly with XML and ontology. In addition, access to information requires the use of web tools for information retrieval (IR). Many methods from traditional IR were extended to structured documents. On the other hand, approaches have been proposed to regard specifically semantics in structured documents by using external semantic resources while collecting original documents on which it is necessary to provide semantic similarity measures in order to perform comparisons between concepts. Most previous approaches did not take into account the relations between concepts and are not tailored to the specific needs of learners. In this paper a semantic IR system of structured educational documents is proposed and adapted to the needs and preferences of the learner. This approach is based on a representation of the nodes of the document´s tree and of the query through the semantic vectors of those concepts. Tests show the feasibility of the proposed approach.