Hi guys, i'm going ahead with some tests with Hadoop and Mahout. Right now my goal is to use SVD to recommend on my Movielens data set, but i have some questions:
1) i have a .txt movielens data set right now, composed as UserID, itemID, PreferenceValue. How can i convert this to use SVD? I think i need to create a matrix, is there some tool or class that i can use? 2) the final goal is using DistributedLanczosSolver with Hadoop, are there some suggestions or info that i need to know about this one? Thank you! regards, ste
