Comparative Analysis between Notations to Classify Named Entities using Conditional Random Fields

  • Daniela Oliviera F. do Amaral PUCRS
  • Maiki Buffet PUCRS
  • Renata Vieira PUCRS

Resumo


Conditional Random Fields (CRF) is a probabilistic Machine Learning (ML) method based on structured prediction. It has been applied in several areas, such as Natural Language Processing (NLP), image processing, computer vision, and bioinformatics. In this paper we analyse two different notations for identifying the words that compose a Named Entity (NE): BILOU and IO. We found out that IO notation presents better results in F-measure than BILOU notation in all categories of HAREM corpus.

Publicado
04/11/2015
AMARAL, Daniela Oliviera F. do; BUFFET, Maiki; VIEIRA, Renata. Comparative Analysis between Notations to Classify Named Entities using Conditional Random Fields. In: SIMPÓSIO BRASILEIRO DE TECNOLOGIA DA INFORMAÇÃO E DA LINGUAGEM HUMANA (STIL), 1. , 2015, Natal/RN. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2015 . p. 27-31.