This is a very basic recommendaion engine. Data set used: REUTERS data set I have also used Apache Mahout Item clustering is done via MinHash algorithm from Mahout. User clustering is custom written and I will try to figure out how to use Mahout. Recommendation engine is also very basic and custom written. Not sure how to use the Taste library from Mahout. This is a very basic implementation done over a weekend. Feel free to comment. There will be lots of improvemts in the future!!