I refer to the paper by called Frustratingly Easy Domain Adaptation (http://www.umiacs.umd.edu/~hal/docs/daume07easyadapt.pdf) where the feature space of both the source and target data are augmented and used as input to a standard learning algorithm. The code is just 10 lines but in perl. I want to port to python but don't understand what the input data looks like that must be processed. http://hal3.name/easyadapt.pl.gz
Thanks