We present a new kernel method for extracting semantic relations between entities in natural language text, based on a generalization of subsequence kernels. This kernel uses three types of subsequence patterns that are typically employed in natural language to assert relationships between two entities. Experiments on extracting protein interactions from biomedical corpora and top-level relations from newspaper corpora demonstrate the advantages of this approach.

PDF, PS

In Advances in Neural Information Processing Systems, Vol. 18: Proceedings of the 2005 Conference (NIPS), Y. Weiss, B. Schoelkopf, J. Platt (Eds.) 2006.

@inproceedings{bunescu:nips05, title={Subsequence Kernels for Relation Extraction}, author={Razvan Bunescu and Raymond J. Mooney}, booktitle={Advances in Neural Information Processing Systems, Vol. 18: Proceedings of the 2005 Conference (NIPS)}, editor={Y. Weiss and B. Schoelkopf and J. Platt}, url="http://www.cs.utexas.edu/users/ai-lab?bunescu:conll05", year={2006} }