we were working with stand-off annotation since we had issues with
overlapping markup (dealing with multiple independant annotations) in a
project two years ago.
Our approach is quite different to the one shown in the paper you cited
by Piotr Bański (Balisage 2010): we use a wrapper markup language that
handles all the stand-off aspects (e.g., the relation between character
position and markup) and incorporates the original inline annotation(s)
with only slight changes (cf.
That means that we don't go the TEI route but we are able to import (and
export) TEI markup.
The key difficulty is -- in my humble opinion -- to maintain a stable
annotation base (primary data/the text that is annotated) without
changing character positions by introducing markup. You may find some
information about our solution in the second paper cited above.
More examples can be found at http://www.xstandoff.net.
If you have any questions please let me know.
Maik Stührenberg, M.A.
Fakultät für Linguistik und Literaturwissenschaft
Telefon: +49 (0)521/106-2534
E-Mail: [log in to unmask]