Minimized Models and Grammar-Informed Initialization for Supertagging with Highly Ambiguous Lexicons

We combine two complementary ideas for learning supertaggers from highly ambiguous lexicons: grammar-informed tag transitions and models minimized via integer programming. Each strategy on its own greatly improves performance over basic expectation-maximization training with a bitag Hidden Markov Model, which we show on the CCGbank and CCG-TUT corpora. The strategies provide further error reductions when combined. We describe a new two-stage integer programming strategy that efficiently deals with the high degree of ambiguity on these datasets while obtaining the full effect of model minimization.

• Type: Journal
• Year: 2010
• Where: ACL
• Authors: Sujith Ravi, Jason Baldridge, Kevin Knight
• Keywords: ACL 2010, Basic Expectation-maximization Training, Computational Linguistics, Integer Programming, Two-stage Integer Programming