[Corpora-List] Using version control software in corpus construction

Geoffrey Sampson grs2 at sussex.ac.uk
Mon Mar 29 11:49:48 UTC 2010


I routinely did this (it seemed crucial) when I and my researchers were
developing various annotated corpora such as SUSANNE and her sisters.  We
used a Unix version-control utility called SCCS if I remember correctly. 
Even though these corpora are relatively small, it would have been
well-nigh inconceivable, I believe, to have reliably maintained their
integrity keeping check on changes just manually.  I can't say I recall any
special points to look out for or beware, I'm afraid.  And in particular, I
don't know whether SCCS or similar systems, which were primarily intended
for controlling versions of program code, would fail to scale up
successfully to much larger corpora such as the entire BNC.

Geoff Sampson


_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora



More information about the Corpora mailing list