Julia Krivanek and Detmar Meurers
Proceedings of the Int. Conference on Dependency Linguistics (Depling 2011). 310-317.
We explore the performance of two dependency parsing approaches, the rule-based WCDG approach (Foth and Menzel 2006) and the data-driven dependency parser MaltParser (Nivre et al 2007) on texts written by language learners.
We show that WCDG outperforms MaltParser in identifying the main functor-argument relations, whereas MaltParser is more successful than WCDG in establishing optional, adjunct dependency relations. This can be interpreted as a tradeoff between the rich, hand-crafted lexical resources capturing obligatory argument relations in WCDG and the ability of a data-driven parser to identify optional, adjunct relations based on the linguistic and world knowledge encoded in the gold-standard training corpora.
Electronically available:
Note: The electronic versions of the publications linked on this page are the last versions I had the copyright for. Where a publisher copyedited and/or typeset the papers, the electronic copies linked here are NOT identical to the officially published version, which should be used for any quotes, references to page numbers, etc.
Bibtex entry:
@InProceedings{Krivanek.Meurers-11,
author = {Julia Krivanek and Detmar Meurers},
title = {Comparing Rule-Based and Data-Driven Dependency Parsing
of Learner Language},
booktitle = {Proceedings of the Int. Conference on Dependency Linguistics
(Depling 2011)},
address = {Barcelona},
pages = {310--317},
year = {2011},
url = {http://purl.org/dm/papers/krivanek-meurers-11.html}
}