Kenji Sagae: “Self-Training without Reranking for Parser Domain Adaptation and Its Impact on Semantic Role Labelling”

July 15, 2010 | Uppsala University, Uppsala, Sweden

Speaker: Kenji Sagae, ICT
Host: The Domain Adaptation in Natural Language Processing (DANLP 2010)

Sagae compares self-training with and without reranking for parser domain adaptation and examines the impact of syntactic parser adaptation on a semantic role labeling system. Although self-training without reranking has been found not to improve in-domain accuracy for parsers trained on the WSJ Penn Treebank, Sagae shows that it is surprisingly effective for parser domain adaptation. Sagae also shows that simple self-training of a syntactic parser improves out-of-domain accuracy of a semantic role labeler.