Enhancing Roget's Thesaurus with Semantic Tags1

In the 1950s Margaret Masterman was a pioneer of natural language processing (NLP) and artificial intelligence (AI). Her research is mainly published in technical reports and was to some degree forgotten until Wilks published an edition of her work (Masterman, 2005). One of her ideas was to use Roget's Thesaurus (RT) as a tool for representing what she called "semantic transformations" (and what would be called "conceptual structures" in modern terminology) for the purposes of machine translation (or "mechanical translation" in the terminology of her time). Recently, the relevance of her thesaurus research has been investigated with respect to modern applications and found to be highly relevant from a modern viewpoint (Priss & Old, 2009).