Un enfoque gramatical para la extracción de términos índice

The extraction of the keywords that characterize a document in a given collection is one of the most important components of an Information Retrieval system. In this article, we propose to apply shallow parsing, implemented by means of cascades of finite-state transducers, to extract complex index terms based on an approximated grammar of Spanish. The coverage of the grammar is small but it allows us to extract the words involved in the most relevant syntactic dependencies. The effectiveness of the extracted index terms has been evaluated in the CLEF collection of Spanish texts.