CKY parsing with independence constraints

Abstract

The CKY algorithm is an important component in many natural language parsers. We propose a novel type of constraint for context-free parsing called independence constraints. Based on the concept of independence between words, we show how these constraints can be used to reduce the work done in the CKY algorithm. We demonstrate a classifier which can be used to identify boundaries between independent words in a sentence using only surface features, and show that it can be used to speed up a CKY parser. We investigate the trade-off between speed and accuracy, and indicate directions for improvement.