Paper: Exploiting Rich Syntactic Information for Relationship Extraction from Biomedical Articles

ACL ID N07-2025
Title Exploiting Rich Syntactic Information for Relationship Extraction from Biomedical Articles
Venue Human Language Technologies
Session Short Paper
Year 2007
Authors

This paper proposes a ternary relation extraction method primarily based on rich syntactic information. We identify PROTEIN-ORGANISM-LOCATION re- lations in the text of biomedical articles. Different kernel functions are used with an SVM learner to integrate two sources of information from syntactic parse trees: (i) a large number of syntactic features that have been shown useful for Seman- tic Role Labeling (SRL) and applied here to the relation extraction task, and (ii) fea- tures from the entire parse tree using a treekernel. Ourexperimentsshowthatthe use of rich syntactic features significantly outperforms shallow word-based features. The best accuracy is obtained by combin- ing SRL features with tree kernels.