Project: Intention Identification

This project studied a novel problem of identifying intention from text data.

In (Chen et al., NAACL 2013), we proposed to study the problem of identifying intention posts in online discussion forums. To the best of our knowledge, our paper is the first reported study of this problem. Our research found that this problem is particularly suited to transfer learning because in different domains, people express the same intention in similar ways. We then proposed a new transfer learning method (called Co-Class) which, unlike a general transfer learning algorithm, exploits several special characteristics of the problem to produce more accurate classifiers.

Publications


Resources
Datasets (NAACL 2013)