engleski

Enhanced Thesaurus Terms Extraction for Document Indexing

In this paper we present an enhanced method for the thesaurus term extraction regarded as the main support to a semi-automatic indexing system. The enhancement is achieved by neutralising the efect of language morphology applying lemmatisation on both the text and the thesaurus, and by implementing an effcient recursive algorithm for term extraction. Formal definition and statistical evaluation of the experimental results of the proposed method for thesaurus term extraction are given. The need for disambiguation methods and the efect of lemmatisation in the realm of thesaurus term extraction are discussed.

Information retrieval; term extraction; NLP; lemmatisation; Eurovoc

nije evidentirano