Recognizing Relation Expression between Named Entities based on Inherent and Context-dependent Features of Relational words

This paper proposes a supervised learn-ing method to recognize expressions that show a relation between two named en-tities, e.g., person, location, or organiza-tion. The method uses two novel fea-tures, 1) whether the candidate words in-herently express relations and 2) how the candidate words are influenced by the past relations of two entities. These features together with conventional syntactic and contextual features are organized as a tree structure and are fed into a boosting-based classification algorithm. Experimental re-sults show that the proposed method out-performs conventional methods.