Sense based organization of descriptive data

In this paper we propose a new technique allowing to map documents´ keywords into relative distance space, which is based primarily on senses of these terms. We use WordNet ontology to retrieve multiple senses of keywords with the aim of generating multidimensional space for our data. The focus of this work is mainly on transferring the available ontology of keywords (and their senses) into multiple dimensions, where each dimension reflects approximation of a single, but broad context of important keywords appearing in our document repository. We have concentrated on discovery of appropriate similarity measurements and construction of data-driven dimensions. It benefits quality of generated dimensions and provides a clearer view of the whole document repository in low-dimensional, context-driven space.