Hi Pranay, I think you're maybe mixing up two different things : You partition the movie ratings into movies that were liked and disliked. But if you want to do recommendations on purchase data, you usually only know which items were sold and that means you can only be sure what items the users like. If someone has not bought something, this does not necessarily have to mean he doesn't like it, it may just mean that he has not yet seen it.
I guess that's why Sean proposed to map all ratings to 1. -sebastian 2010/6/11 pranay venkata <[email protected]> > Hi, > > I'm a newbie to mahout.My aim is to produce recommendations on binary user > purchased data.So i applied item-item similarity model in computing top N > recommendations for movie lens data assuming 1-3 ratings as a 0 and 4-5 > ratings as a 1.Then i tried evaluating my recommendations with the ratings > in the test-data but hardly there have been two or three matches from my > top > 20 recommendations to the top rated items in test data and no match for > most > users. > > So are my recommendations totally bad by nature or do i need to go for a > different measure for evaluating my recommendations ? > > Please help me ! Thanks in advance. > > Pranay, 2nd yr ,UG student. >
