Coupled Semi-Supervised Learning for Information Extraction

We consider the problem of semi-supervised learning to extract categories (e.g., academic fields, athletes) and relations (e.g., PlaysSport (athlete, sport)) from web pages, starting with a handful of