Recent systems for semantic role labeling are very dependent on the specific predicates and corpora on which they are trained, but labeling new data is expensive. We study which features and classifiers are best able to generalize to unseen predicates from new semantic frames. We find that automatically derived cluster information is especially helpful in this setting, and that a relatively simple a posteriori classifier outperforms Maximum Entropy. This material is based upon work supported by from the National Institute of Health grant number 2ROERR0928309, a National Science Foundation grant number #5-41984-A and a National Science Foundation grant number E1A0080124 . Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of above named organizations.
[1]
Daniel Jurafsky,et al.
Automatic Labeling of Semantic Roles
,
2002,
CL.
[2]
Naftali Tishby,et al.
Distributional Clustering of English Words
,
1993,
ACL.
[3]
Hae-Chang Rim,et al.
Semantic Role Labeling using Maximum Entropy Model
,
2004,
CoNLL.
[4]
Daniel Jurafsky,et al.
Support Vector Learning for Semantic Argument Classification
,
2005,
Machine Learning.
[5]
Michael Collins,et al.
Three Generative, Lexicalised Models for Statistical Parsing
,
1997,
ACL.
[6]
Gemma Boleda,et al.
The Influence of Argument Structure on Semantic Role Assignment
,
2004,
EMNLP.