update
Brian MacWhinney
It seems to me time for a bit of an update on CA-related
TalkBank developments.
1. Over the last two months, some of our emphasis here at CMU has
been on making sure that media are
available for all the TalkBank corpora and that all files with media
can be accessed through CLAN's WebData
facility. Since some connections are too slow for video playback, we
are making sure that all files also have
MP3 audio available. In order to playback conversations using MP3,
you should tell QuickTime that you have a
slow modem connection (56K) and then QuickTime will automatically
default to MP3 audio playback from transcripts.
2. On the corpus front, you may note that we have added back the ESF
English-Italian and English-Punjabi corpora.
We also have nearly complete transcriptions with video for the Carmen
Curtis geometry classroom sessions from Rich
Lehrer. For the Supreme Court, we have linked audio now for about 20
years. We continue to consolidate the Jefferson CA corpus, with some
of the material open and some available from Johannes.
3. Corpora currently in development include family and group
conversations from Susan Ervin-Tripp and gesture data
from Sotaro Kita and colleagues.
4. In the AphasiaBank area, we will have a meeting of the Aphasia
consortium at CMU on November 30. People in this
group are committed to filming and transcribing a fairly complete
conversational corpus from patients.
5. In the CLAN program area, we are focusing on the construction of
training videos for beginning users. We are also providing and
improving interoperability between CLAN, EXMARaLDA, Praat, OpenSHAPA,
and ELAN.
Please feel free to post questions about data, ideas for new corpora,
and requests for program modifications.
-- Brian MacWhinney, CMU