2025/11/03 by Mara Moita, Paulo Vaz de Carvalho, Helena Carmo +1 · 1 voice
Psychology · Computer Science · Arts and Humanities · #Hearing Impairment and Communication #Hand Gesture Recognition Systems #Subtitles and Audiovisual Media
paper · doi:10.16995/jpl.23006
openalex publication_date 2025/11/03 · openalex created_date 2025/11/03 · openalex updated_date 2026/07/31
This is an accepted article with a DOI pre-assigned that is not yet published.The Portuguese Sign Languages (LGP) studies developed to date were based on elicited data collections from small groups of native signers and mainly depended on researchers' intuition, not representing the linguistic phenomena that characterize this language and not providing robust linguistic descriptions to develop resources for deaf education and communication in its basic needs. The lack of a structured data collection representing the natural use of sign language limits linguistic understanding, language preservation, educational opportunities, accessibility, and technological advancement while also threatening the cultural heritage of the Deaf community. To have a well-documented LGP corpus to ensure that LGP thrives and continues to be recognized as a legitimate and fully developed language, we built the LGP Reference Corpus. In this article, we describe the construction of the first machine-readable digital reference corpus for LGP, to which we listed, digitalized, archived, and annotated signed data along with the LGP videos and associated metadata collected between 1992 and 2019. The LGP Reference Corpus is the first annotated LGP corpus with diachronic variation.