no code implementations • 23 Jan 2014 • Christina Sauper, Regina Barzilay
We test our model on two tasks, joint aspect identification and sentiment analysis on a set of Yelp reviews and aspect identification alone on a set of medical summaries.
Sentiment Analysis