한국수어 말뭉치 구축 모형 연구
Korean Sign Language Corpus
0
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
The Korean Sign Language (KSL) is the language of the Deaf in the Republic of Korea. KSL is a visual-spatial language and it’s branch of research is rather young, but due to the Korean Sign Language Act (2016) the interest and need of deeper understanding of the language has risen. This paper describes how the first KSL-Corpus has been build and how the data has been collected and transcribed. 60 deaf native and near-native signers from the area of Seoul were filmed in pairs. The deaf informants were asked to deal with 13 tasks, which included open conversation or discussion on various topics, retelling of a picture story or movie clip. Each session of this naturalistic, controlled and elicited signed language sample has a length of about three hours that means the complete recordings contain about 90 hours of sign language data. Almost 12 hours of this sign language data has been annotated in ELAN, a professional tool for the creation of complex annotations on video and audio resources developed by the Max-Planck-Institute of Psycholinguistics in the Netherlands. The time-aligned annotation in ELAN is meant to be a basic transcription of the KSL data that means the focus of the transcription was to identify sign units in a consistent way. This paper also addresses the need of a systematic database which administers the annotated glosses and reveals the concrete strategy to develop such tool.
Abstract
The Korean Sign Language (KSL) is the language of the Deaf in the Republic of Korea. KSL is a visual-spatial language and it’s branch of research is rather young, but due to the Korean Sign Language Act (2016) the interest and need of deeper understanding of the language has risen. This paper describes how the first KSL-Corpus has been build and how the data has been collected and transcribed. 60 deaf native and near-native signers from the area of Seoul were filmed in pairs. The deaf informants were asked to deal with 13 tasks, which included open conversation or discussion on various topics, retelling of a picture story or movie clip. Each session of this naturalistic, controlled and elicited signed language sample has a length of about three hours that means the complete recordings contain about 90 hours of sign language data. Almost 12 hours of this sign language data has been annotated in ELAN, a professional tool for the creation of complex annotations on video and audio resources developed by the Max-Planck-Institute of Psycholinguistics in the Netherlands. The time-aligned annotation in ELAN is meant to be a basic transcription of the KSL data that means the focus of the transcription was to identify sign units in a consistent way. This paper also addresses the need of a systematic database which administers the annotated glosses and reveals the concrete strategy to develop such tool.
참고문헌 (9)
- 국립국어원, "한국수어사전" (2007), 국립국어원
- 한국수어 말뭉치 기반 조사 및 시범 구축, "국립국어원", ~ (2015)
- 한국수어 발전을 위한 기본 계획 수립 연구, "국립국어원", ~ (2014)
- 한국수어 말뭉치 구축을 위한 기초연구, "국립국어원", ~ (2014)
- 홍성은, "독일수어 코퍼스 프로젝트의 유도 방법", 제12회 한국수어학회 (2010)
- Johnston, Trevor, "Corpus linguistics and signed languages：no lemmata, no corpus", Proceedings of the 3rd Workshop on the Representation and Processing of Sign Languages：Construction and Exploitation of Sign Language Corpora, pp.82~87 (2008)
- Lucas Ceil, "What’s Your Sign for PIZZA? An Introduction to Variation in American Sign Language" (2003), Gallaudet University Press
- McCaskill Carolyn, "The Hidden Treasure of Black ASL：Its History and Structure" (2011), Gallaudet University Press
- Woodbury Tony, "Archives and audiences：Toward making endangered language documentations people can read, use, understand, and admire", Proceedings of Endangered Languages Achive Workshop on Language Documentation and Archiving, pp.11~20 (2011), School of Oriental and African Studies