BSL Corpus Project goes online
26 February 2012
Unique BSL resource becomes available to all online
An
important development has been made in a project that will be familiar
to many DCAL newsletter readers from updates over the last few years:
DCAL is pleased to announce that data from the British Sign Language
Corpus Project (BSLCP) can now be accessed by all, having gone live
on-line at the end of 2011.
BSL language recordings
Available at the BSL Corpus Project website,
this resource will continue to develop to become the first national
computerised and publicly accessible BSL corpus - that is a unique
collection of language recordings of British Sign Language. The
recordings will be of enormous benefit to students and teachers of BSL
and to sign language interpreters across the country, leading to
improved services for Deaf people that will better ensure their full
participation in society.
The BSLCP was funded from 2008 to
2011 by the Economic and Social Research Council (ESRC) and led by
staff at DCAL. Project partners are from Bangor University (Wales),
Heriot-Watt University (Scotland), Queens University Belfast (Northern
Ireland) and the University of Bristol (England).
The
collection of video recordings shows 249 Deaf men and women of
different ages and backgrounds conversing in BSL with each other in
pairs. They answer questions, tell stories, and show their signs for
102 key concepts. The filming took place in 8 cities across the UK to
reflect regional variation within BSL.
Anyone can watch the video clips under the data section of the website.
For general visitors, clicking on the image on the left is recommended.
Those with a research or teaching interest can access more information
via the image on the right. This takes visitors to UCL's CAVA website
where anyone can view or download clips, and where researchers can
register for a licence to access restricted data.
Wider implications
In
addition to practical applications in the UK, the web-based corpus
video data is set to contribute significantly to international
linguistics research. It will also be a valuable resource for people
with an interest in technology, particularly those working towards
automatic sign language recognition (the signed equivalent of voice
recognition) and the development of virtual signers i.e., signing
avatars.
Current BSLCP Director, Dr Kearsy Cormier, explains:
"We are very pleased that the BSL corpus video data are now freely
available worldwide; this was one of the main aims of the project, but
the work is by no means completed. In the future, annotations and
translations of the data will be made available online to bring this
resource closer to what we mean by a "corpus" today in linguistic
research. These annotations will allow anyone to search for specific
signs quickly and facilitate peer-reviews of claims about BSL structure
and use amongst researchers.
Another aim was to use the data
to study why BSL varies and how it is changing, and to investigate
frequency of BSL signs - that is to find out which signs are the most
common in conversation. These completed studies represent an important
first step towards a better understanding of variation and change and
lexical frequency in BSL."
Former BSLCP Project Director, Dr.
Adam Schembri (now based at LaTrobe University, Australia) explains
further: "We expect the BSLCP will contribute to wider research in the
field of linguistics worldwide. Internationally the BSL Corpus is one
of only a few large sign language corpus projects (along with projects
in Australia, The Netherlands and Germany) and it's the second to have
video data available online (after The Netherlands)."
Professor
Bencie Woll, DCAL Director, expands on the significance of the work:
"DCAL hopes the BSLCP video data will lead directly to improved sign
language teaching and improvements in training BSL teachers, sign
language interpreters and teachers of deaf children. But the BSLCP
findings have the potential for much broader impact. Already there are
follow-on projects in DCAL making use of the data, which are helping to
extend ongoing work on production, comprehension, processing,
acquisition and neural bases of BSL."
PHOTO: Dr Adam Schembri (right) and Dr Jordan Fenlon
show the BSL Corpus Project data to colleagues
and members of the Deaf community at the
DCAL Deaf Open Day, March 2010.