A Composite Kernel to Extract Relations with both Flat and Structured Features

This paper proposes a novel composite kernel for relation extraction. The composite kernel consists of two individual kernels: an entity kernel that allows for entity-related features and a convolution parse tree kernel that models syntactic information of relation examples. The motivation of our method is to fully utilize the nice properties of kernel methods to explore and combine diverse features for relation extraction. Our study illustrates that the composite kernel can capture both flat and structured features effectively, and can also easily scale to include more features. Evaluation on the ACE corpus shows that our method outperforms the previous best-reported method. It also shows that due to the effective exploration of the syntactic features the sole parse tree kernel significantly outperforms the previous two dependency kernels by 16 in F-measure on the ACE 2003 corpus.