Syntactic Parsing for Ranking-Based Coreference Resolution
Altaf Rahman and Vincent Ng.
Proceedings of the 5th International Joint Conference on Natural Language Processing, pp. 465-473, 2011.
Click here for the
PostScript or PDF
version. The talk slides are available here.
Abstract
Recent research efforts have led to the development of a state-of-the-art
supervised coreference model, the cluster-ranking model. However,
it is not clear whether the features
that have been shown to be useful when employed in traditional coreference
models will fare equally well in this new model.
Rather than merely re-evaluate them using the cluster-ranking model,
we examine two
interesting types of features derived
from syntactic parses, tree-based features and path-based features, and
discuss the challenges involved in employing them in the
cluster-ranking model.
Experimental
results on a set of Switchboard dialogues show their
effectiveness when used in combination with the
cluster-ranking model: using them to augment a baseline coreference feature
set results in a 8.6-11.7% reduction in relative error.
BibTeX entry
@InProceedings{Rahman+Ng:11d,
author = {Altaf Rahman and Vincent Ng},
title = {Syntactic Parsing for Ranking-Based Coreference Resolution},
booktitle = {Proceedings of the 5th International Joint Conference on Natural Language Processing},
pages = {465--473},
year = 2011
}