Structure compilation: trading structure for features

Structured models often achieve excellent performance but can be slow at test time. We investigate structure compilation, where we replace structure with features, which are often computationally simpler but unfortunately statistically more complex. We analyze this tradeoff theoretically and empirically on three natural language processing tasks. We also introduce a simple method to transfer predictive power from structure to features via unlabeled data, while incurring a minimal statistical penalty.

• Type: Conference
• Year: 2008
• Where: ICML
• Authors: Dan Klein, Hal Daumé III, Percy Liang
• Keywords: ICML 2008, Language Processing Tasks, Machine Learning, Minimal Statistical Penalty, Structure Compilation