Standardisation method of noun phrases for automatic indexing
This work proposes and evaluates a method of standardisation ofnoun phrases in canonical terms. This procedure aims to contribute to thequalitative improvement of automatic indexing avoiding the terminologicaldispersion and preserving the keywords present within the noun phrases. Theresearch is expl...
| Autores: | , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2019 |
| País: | Brasil |
| Institución: | Universidade Federal do Rio Grande do Sul (UFRGS) |
| Repositorio: | Em Questão (Online) |
| Idioma: | portugués |
| OAI Identifier: | oai:seer.ufrgs.br:article/81901 |
| Acceso en línea: | https://seer.ufrgs.br/index.php/EmQuestao/article/view/81901 |
| Access Level: | acceso abierto |
| Palabra clave: | Indexação automática. Sintagmas nominais. Normalização de sintagmas nominais. Palavras-chave. Tesauro. Automatic indexing. Noun phrases. Standardisation of noun phrases. Keywords. Thesaurus. |
| Sumario: | This work proposes and evaluates a method of standardisation ofnoun phrases in canonical terms. This procedure aims to contribute to thequalitative improvement of automatic indexing avoiding the terminologicaldispersion and preserving the keywords present within the noun phrases. Theresearch is exploratory and empirical, based on bibliographic research and anexperiment in a corpus composed of scientific articles in Information Science.The proposed standardisation method contains rules and criteria that follow theconstraints of preserving the valid structure of the noun phrase and thekeywords. The method evaluation consists of the analysis of the presence ofterms of the Brazilian Thesaurus in Information Science (TBCI) in the nounphrases resulting from the application of the proposed rules and criteria. Themethod consists of two stages: the first consists of 85 rules to reduce the size ofthe noun phrases, and the second stage contains seven criteria responsible foreliminating unnecessary grammatical elements from the noun phrases. Theresults of the evaluation indicate that the proposed method allows theachievement of positive results, even with two criteria of the second stage notpresenting results for the corpus. It concludes that the application of the methodin automatic indexing system is feasible and brings good results. |
|---|