1 2 3 
A rate-distortion one-class model and its applications to clustering (Crammer, Koby, Talukdar, Partha Pratim, Pereira, Fernando), 2008. [pdf]

Lightly-Supervised Attribute Extraction for Web Search (Bellare, Kedar, Talukdar, Partha Pratim, Kumaran, Giridhar, Pereira, Fernando, Liberman, Mark, McCallum, Andrew, Dredze, Mark) In The NIPS 2007 Workshop on Machine Learning for Web Search, 2008. [pdf]

2007
Transductive Structured Classification through Constrained Min-Cuts (Ganchev, Kuzman, Pereira, Fernando) In Proceedings of the Second Workshop on TextGraphs: Graph-Based Algorithms for Natural Language Processing, 2007. [pdf]

Penn/UMass/CHOP Biocreative II systems (Ganchev, Kuzman, Crammer, Koby, Pereira, Fernando, Mann, Gideon, Bellare, Kedar, McCallum, Andrew, Carroll, Steven, Jin, Yang, White, Peter) In Proceedings of the Second BioCreative Challenge Evaluation Workshop, 2007. [pdf]

Semi-Automated Named Entity Annotation (Ganchev, Kuzman, Pereira, Fernando, Mandel, Mark, Carroll, Steven, White, Peter) In Proceedings of the Linguistic Annotation Workshop, 2007. [pdf]

Frustratingly Hard Domain Adaptation for Dependency Parsing (Dredze, Mark, Blitzer, John, Talukdar, Partha Pratim, Ganchev, Kuzman, Graca, Joao, Pereira, Fernando) In Proceedings of the CoNLL Shared Task Session of EMNLP-CoNLL, 2007. [pdf]

Biographies, Bollywood, Boom-boxes and Blenders: Domain Adaptation for Sentiment Classification (Blitzer, John, Dredze, Mark, Pereira, Fernando) In Association for Computational Linguistics (ACL), 2007. [pdf]

Lightly-Supervised Attribute Extraction (Bellare, Kedar, Talukdar, Partha Pratim, Kumaran, G., Pereira, Fernando, Liberman, Mark, McCallum, Andrew, Dredze, Mark), 2007. [pdf]

2006
A context pattern induction method for named entity extraction (Talukdar, Partha Pratim, Brants, Thorsten, Liberman, Mark, Pereira, Fernando), 2006. [pdf]

``Sorry, I Forgot the Attachment:'' Email Attachment Prediction (Dredze, Mark, Blitzer, John, Pereira, Fernando) In Conference on Email and Anti-Spam (CEAS), 2006. [pdf]