Information Extraction (IE) is one of the interesting problems in machine learning. 
IE employes several statistical models like Maximum Entropy Markov Model (MEMM), Hidden Markov Model (HMM), 
Conditional Random Fields (CRF). Typically it builds a model which involves learning patterns from the training data, 
and applying it on test data. In real life applications, training and testing data may come from different domains. 
Hence a model must learn general patterns, and not specific to a particular domain.

In this project, we propose to design a model, based on Conditional Random Fields in such a manner 
that a model built using training sequences from a domain can be successfully applied on test sequences 
from other domain. In this report, we present preliminary work done in this direction.