SV: Digital Texts: Tagging

Lars Martin Fosse lmfosse at ONLINE.NO
Tue Nov 16 09:35:25 UTC 1999


Mahoney, Richard B [SMTP:rbm49 at STUDENT.CANTERBURY.AC.NZ] skrev 16. november
1999 05:41:
>
> Over the last week or so I've been fumbling with the processes of a couple
> of Textual Analysis programmes: "TACT" and "Lexa".  From what I can gather
> so far it seems that it is probably worth persevering.
>
> What is clear though is that neither programme is going to make up for my
> ignorance - much as I'd like them to.  Before I can do anything, I'm going
> to have to "tag" or "mark-up" a digital text.  Here I'm in need of
> assistance.
>
> I've read that Sidney Greenbaum has developed a tagging system for modern
> English grammar for University College, London.  What I would like to
> have access to - if such a thing exists - is a similar scheme for
> Classical Sanskrit and Tibetan.
>
> Have any readers been over this ground before?  And if so, would any be so
> good as to give me some much needed advice?

I used Lexa when I did my Ph.D., and I give a description of how I used it for
tagging in my thesis. I found that Lexa through its search-and-replace function
gives you a primitive tagger. I was able to tag quite a lot of text
automatically, but it had to be cleaned up manually, and certain things had to
be entered manually.

Basic principle: Lexa allows you to search for parts of words, which allows you
create search and replace lists looking like this:

Say you want to replace "anyoldword" with "anotherandcompletelynewword" and so
on. You can them produce a list file looking like this:

word<TAB><TAB>word
etc.

This principle allows you to tag case endings, e.g.

CLASS_n: INSTRUMENTAL
ena<TAB><TAB>ena_INSTRUMENTAL
e.na<TAB><TAB>e.na_INSTRUMENTAL

etc.

or:

CLASS_n: PTCL
hy<TAB><TAB>hy_PTCL
hi<TAB><TAB>hi_PTCL

where ptcl = particle.

Words that are tagged in the wrong way frequently you may have to collect in a
separate listfile and tag them as complete words.

You will find a more detailed description in the appendices to my thesis: The
 Crux of Chronology in Sanskrit Literature. Statistics and Indology. A Study
of Method. Scandinavian Univesity Press, Oslo 1997.

Good luck!

Lars Martin Fosse


Dr. art. Lars Martin Fosse
Haugerudvn. 76, Leil. 114,
0674 Oslo
Norway
Phone/Fax: +47 22 32 12 19
Email: lmfosse at online.no





More information about the INDOLOGY mailing list