Hand-checked Eve corpus

[Brian MacWhinney]

Dear Rui,
    The manual needs some updating in these regards.  Currently, we have two training sets.  The Wright-train is from adults doing picture descriptions and narratives and the Eve-train includes all 20 of the Eve files from Brown 1973.  The Eve files were very much out of date, but I brought them back up to date just now.  In general, tagging of child utterances, particularly the shortest ones, is never going to be determinate.  However, using TRNFIX, the total number of words differing between the %trn and the %mor lines is not too large and mostly is for relatively uninteresting ambiguities.  We are developing some POSTMORTEM rules to fix some of these remaining things, such as the treatment of “right” when it appears alone.
    There is now no difference between the Eve corpus in the ENG MOR grammar and the one in the database, but make sure you get new versions of ENG MOR for your work.

—Brian MacWhinney

> On Jan 9, 2015, at 8:00 PM, Rui Huang <huang3740 at gmail.com> wrote:
> 
> Hi Chibolts,
> 
>  The manual says that the first 15 files of Eve corpus is hand-checked. I would like to use it as testing data in an evaluation task. There are two version of Eve corpus can be found: one is from MOR Grammar, the other one is from transcript database. When running 'trnsfix' command on both files, the one from database shows several difference between %trn and %mor tier, but none in the script from Grammar file. A gut feeling that the grammar one should be the gold-standard training data, but not sure. Do you have some idea on finding the hand-checked Eve corpus?
> 
> Thanks for your time,
> 
> Rui Huang
> Linguistics student
> Graduate Center, City university of New York
> 
> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups "chibolts" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to chibolts+unsubscribe at googlegroups.com <mailto:chibolts+unsubscribe at googlegroups.com>.
> To post to this group, send email to chibolts at googlegroups.com <mailto:chibolts at googlegroups.com>.
> To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/495ccd77-c35f-49fd-9115-50f14a5172e1%40googlegroups.com <https://groups.google.com/d/msgid/chibolts/495ccd77-c35f-49fd-9115-50f14a5172e1%40googlegroups.com?utm_medium=email&utm_source=footer>.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.

-- 
You received this message because you are subscribed to the Google Groups "chibolts" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chibolts+unsubscribe at googlegroups.com.
To post to this group, send email to chibolts at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/BA33A41F-E07F-4D40-9191-16F5246AA12F%40cmu.edu.
For more options, visit https://groups.google.com/d/optout.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/chibolts/attachments/20150109/a6172a57/attachment.htm>