Predictive methods for Text mining
I will give a general overview of using prediction methods in text mining applications, including text categorization, information extraction, summarization, and question answering. I will then discuss some of the more advanced issues encountered in real applications such as structured and complicated output classification, the use of unlabeled data, modeling link structures, collective inference and community effect, and transfer learning under changing environment, etc.
Author: Tong Zhang, Yahoo! Research