Actions

Expert

Computational Linguistics and Information Processing

The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Expertsourcing

This project, a collaboration with Such Saria [1] at Johns Hopkins, is developing a new approach to text classification that combines concepts from high accuracy rule-based prediction, crowdsourcing annotation consistency, and ensemble learning in order to learn classifiers in the presence of sparse annotations.