Extracting relevant learning objects using a semantic annotation method

Information research refers, in our context, to information retrieval to obtain further learning information from documents. However, automatic tools for learning information retrieval from these documents based on semantic tags are not yet effective. We propose here a model which aims at automatically annotating texts with semantic metadata. These metadata will allow us to index and extract learning objects from texts. This model is composed of two parts: the first part consists of a semantic annotation of learning objects according to their semantic categories (definition, example, exercise, etc.). The second part uses automatic semantic annotation which is generated by the first part to create a semantic inverted index able to find relevant learning objects for queries associated with semantic categories. We have implemented a system called SRIDOP, on the basis of the proposed model and we have verified its effectiveness.