data is available at: jmcauley.ucsd.edu/data/amazon/links.html
(1) Figure out how to read in the data and start playing with it. Start with the 5-core clothing dataset, or maybe even cut it down to a smaller subset.
(2) Take a look at my lecture notes here: http://cseweb.ucsd.edu/classes/fa15/cse255-a/ In particular weeks 1 and 4 are relevant to what we're trying to do, beyond that we'll have to start looking at papers.
(3) Some good initial papers to look at can be found at the end of the week 4 slides.
(4) Set up a machine learning pipeline to evaluate and compare methods.
(5) Once we reach that point (which I think is already non-trivial!) then we'll start thinking how we can extend the basic methods and implement more complicated techniques.