Approach to data representation in information learning systems

Information value depends on previous fund of knowledge - thesaurus, so it is necessary to construct educational reference materials representation taking into account measure of adjacency between content fragments, choosing depth of studying, etc. For automation this process proposes to index content fragments on the basis of base vector, which constructs in automated mode on user demand. As basis for creation of indexes the vector of terms, relevant to the inquiry thesaurus is chosen. The inquiry contains the list of terms (or topics), that learning plans to master, automatically expanded with semantic close terms. On the basis of indexes the measure of adjacency between content fragments is defined, and applied clustering in one of this way: by means of dendrogram designing or by means self-organizing maps. Procedure of the automated creation of individual learning trajectories is based on the analysis of the semantic network formed on the basis of indexation results and matrix of adjacency between content fragments. Usage of the automated procedures of indexation and clustering of educational content fragments allows us to present in an evident look relations between these fragments, their thematic adjacency and to construct the individual learning trajectory meeting individual user needs.