Extracting and verifying hyponymy relations based on multiple patterns and features

Abstract

Hyponymy relations play a crucial role in various natural language processing systems. Automatic acquisition and verification of hyponymy relations is a basic problem in knowledge acquisition from text. We present a method that acquires and verifies hyponymy relations based on multiple patterns and features. It initially obtains a set of removable patterns using Chinese lexico-syntactic patterns. Then the concepts of constituting hyponymy relation are acquired with removable patterns. Finally, hyponymy relations are verified with space structure features, semantic features and context features. Experimental results demonstrate good performance of the method. © 2008 Springer-Verlag Berlin Heidelberg.