I want to apply feature selection based on information gain: I have many features many of which are redundant. I am planning on selecting a feature and then iteratively add features that 'add the more information'. Is there any relevant paper I should read to get familiar with such methods?
If many of your features are redundant I would propose performing PCA. This will eliminate the redundancies while retaining the variance in the data. You can find numerous example/explanations of this with a google search. Also, consider reading this post (Basic question on PCA: can it be used to divide original features into groups?)
Here is an excellent text that covers the area of feature selection (and many others): http://www.stanford.edu/~hastie/local.ftp/Springer/OLD/ESLII_print4.pdf
A quick google search also shows numerous papers, for example http://www.ijitee.org/attachments/File/v2i2/B0352012213.pdf