PAIVA, F. A. S. ; BARBOSA, P. A. ; FARIA, P. ; DE MARTINO, J. M. Developing a rule‑based Brazilian Portuguese‑to‑Libras machine translation system. In: FINATTO, Maria José Bocorny; REBECHI, Rozane Rodrigues; SARMENTO, Simone; BOCORNY, Ana Eliza Pereira (Orgs.). Linguística de Corpus: Perspectivas. Porto Alegre: Instituto de Letras da UFRGS, 2018, p. 127-154.

Abstract: : Sign Language Machine Translation is a way that can bring, for deaf people, more education and accessibility to information. This paper presents an ongoing rule‑based automatic translation system from Brazilian Portuguese (BP) to Brazilian Sign Language (Libras). Our approach is based on the analysis of a BP‑Libras parallel corpus composed of the content of a science textbook. We describe a methodology to formalize the morphosyntactic and semantic rules that translate BP text to Libras glosses. Each gloss will be used to control a signing avatar. As a result, we aim to provide a way for deaf children to study by using signed material in Libras, providing more technological resources for students to improve their learning.

Keywords: Machine Translation. Rule‐Based. Brazilian Portuguese. Libras.