Automatic generation of hierarchical taxonomies from free text using linguistic algorithms

Abstract

This paper presents a technique, based on linguistic algorithms, to construct hierarchical taxonomies from free text. These hierarchies, as well as other relationships, are extracted from free text by identifying verbal structures with semantic meaning. The created taxonomies can be used as kernel for complete domain representation of a particular knowledge area. The domain representation could allow software architects to reuse Information. In order to test the algorithms, an example domain is presented to exemplify the use of the technique.