What do you mean by "other types of information"? As far as I know,
automatic term extraction methods are based on features that can be
extracted/calculated automatically from texts. So in the begining, we can
only calculate statistical scores, and then we can tag/shallow-parse
(extract linguistic information). If we can go further, say can deep-parse,
or extract semantic or discource information automatically, then new
approaches of automatic term extraction will be introduced.
You can find some discussion about using other information in automatic
terminology processing in:
Maynard, D. and S. Ananiadou, 1999. “Identifying Contextual Information for
Multi-Word Term Extraction”. Proceedings of the Fifth International Congress
on Terminology and Knowledge Engineering (TKE'99), 212-221. Vienna, Austria.
Meyer, I. 2001. “Extracting knowledge-rich contexts for terminography”. In
Recent Advances in Computational Terminology ed. by D. Bourigault, C.
Jacquemin and M. C L’Homme. Amsterdam, John Benjamins.
Morin, E. 1999. “Automatic acquisition of semantic relations between terms
from technical corpora”. Proceedings of the Fifth International Congress on
Terminology and Knowledge Engineering (TKE'99), 268-278. Vienna, Austria.
Hope this help,
Le An Ha
From: firstname.lastname@example.org [mailto:email@example.com]On
Behalf Of KERREMANS, Koen
Sent: 22 October 2002 09:14
Subject: [Corpora-List] alternative methods/approaches in automatic term
The possibilities of automatic term extraction are usually explored using
statistical information, linguistic information or a combination of both.
Does anyone know of "alternative methods/approaches", i.e.
methods/approaches that make use of other types of information (besides
statistical or linguistic information)?
This archive was generated by hypermail 2b29 : Wed Oct 23 2002 - 09:14:00 MET DST