Abstract
In the natural language processing (NLP) scenario, Brazilian Portuguese (and Portuguese in general) still suffers from the lack of good quality base tools (e.g. parsers) and resources (e.g. annotated corpora). Corpora annotated with semantic information is particularly scarce and is a very costly resource to be produced manually. In order to provide some help to mend that situation, this paper presents an automatic hyponymy identification method for Brazilian Portuguese texts. The proposed method uses lexical and syntactic data alongside common sense information obtained from the Brazilian Open Mind Common Sense project (OMCS-Br). The results obtained so far are compatible with previous work and encourage other directions for further research. © 2012 Springer-Verlag.
Author supplied keywords
Cite
CITATION STYLE
Taba, L. S., & De Medeiros Caseli, H. (2012). Automatic hyponymy identification from Brazilian Portuguese texts. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7243 LNAI, pp. 186–192). https://doi.org/10.1007/978-3-642-28885-2_21
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
