SyntaLex
The SyntaLex system extends the
Duluth systems from Senseval-2, with part
of speech and syntactic features. It is a supervised learning approach
that carries out lexical sample disambiguation. Find more information in
the Syntalex README.
The part of speech features come from the Brill Tagger, and the parse
features come from the Collin's Parser. You can find tools that extract
the features from these tools here.
POS tagged and parsed data
The following sense-tagged corpora have been created with the above
mentioned tools, and are therefore POS tagged and parsed. They can be
used with SyntaLex, and were used to create the results found in the MS
thesis mentioned below.
Related Publications
-
Complementarity of Lexical and Simple Syntactic Features: The Syntalex
Approach to Senseval-3
(Mohammad and Pedersen) - Appears in the Proceedings of the
Third International Workshop on the Evaluation of Systems
for the Semantic Analysis of Text (Senseval-3), July 25-26, 2004,
Barcelona, Spain.
-
Combining Lexical and Syntactic Features for Supervised Word Sense
Disambiguation (Mohammad and Pedersen) - Appears in the Proceedings
of the Conference on Computational Natural Language Learning (CoNLL), May
6-7, 2004, Boston, MA
-
Combining Lexical and Syntactic Features for Supervised Word Sense
Disambiguation (Mohammad) - MS Thesis, University of Minnesota,
Duluth, August, 2003.
By:
Ted Pedersen
- tpederse AT d umn edu