Thesaurus-based index term extraction for agricultural documents

Abstract

This paper describes a new algorithm for automatically extracting index terms from documents relating to the domain of agriculture. The domain-specific Agrovoc thesaurus developed by the FAO is used both as a controlled vocabulary and as a knowledge base for semantic matching. The automatically assigned terms are evaluated against a manually indexed 200-item sample of the FAO’s document repository, and the performance of the new algorithm is compared with a state-of-the-art system for keyphrase extraction.