Constraint-Driven Rank-Based Learning for Information Extraction

Most learning algorithms for undirected graphical models require complete inference over at least one instance before parameter updates can be made. SampleRank is a rankbased learning framework that alleviates this problem by updating the parameters during inference. Most semi-supervised learning algorithms also perform full inference on at least one instance before each parameter update. We extend SampleRank to semi-supervised learning in order to circumvent this computational bottleneck. Different approaches to incorporate unlabeled data and prior knowledge into this framework are explored. When evaluated on a standard information extraction dataset, our method significantly outperforms the supervised method, and matches results of a competing state-of-the-art semi-supervised learning approach.

• Type: Journal
• Year: 2010
• Where: NAACL
• Authors: Sameer Singh, Limin Yao, Sebastian Riedel, Andrew McCallum
• Keywords: Computational Linguistics, Inference, NAACL 2010, Parameter, Semi-supervised Learning