Learning algorithms for real-time text interpretation

This talk covers some fundamental primitives in text interpretation -- named entity disambiguation, topic labeling and extending knowledge bases -- and describes learning solutions for these primitives that support real-time interpretation and have scalable training.