Recognising Complex Prepositions Prep+N+Prep as Negative Patterns in Automatic Term Extraction from Texts

Claudia OliveiraMilena GarrãoLuiz Alexandre Amaral

This work is a study of the delimitation of complex prepositions (CP) as lexical units, items of a computational lexicon that includes compounds and phrases. In addition, given the utmost importance of spotting noun phrases (NP) in document retrieval systems, parsing prepositional structures such as "Prep1 N Prep2 X" prevents the fragment "N Prep2 X" from being detected as a noun phrase, i.e. the prepositional structure is a negative pattern to be used in the extraction of noun phrases from texts. We analyse the distinct attempts at the description of the class of CPs, in order to establish criteria that could be formalised and therefore made available in a computer system.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: