A Method for Stopping Active Learning Based on Stabilizing Predictions and the Need for User-Adjustable Stopping.

Michael Bloodgood, K. Vijay-Shanker

Browse the full CoNLL paper archive.