0
votes

Have any one know recommendation algorithm in mahout for finding support and confidence from given datasets of User id , Item id and Item preference value..And based on that information the algorithm recommends items...

1
"finding support and confidence" of what? Are you looking at 'simple' (Statistics 101) confidence intervals based on a set of data? What is your end goal?warren
I think he is trying to do associative rule learning. en.wikipedia.org/wiki/Association_rule_learning#Useful_ConceptsRob Neuhaus
Confidence and support come from achieving goals that you set yourself, and from having a small number of close personal relationships with people you care about and with whom you can be yourself. Admittedly, these approaches have appalling time complexity :-Pj_random_hacker

1 Answers

2
votes

The concepts of support and confidence do not quite apply to item-based recommender algorithms. So in that sense the answer is 'no'.

I can guess what you may have in mind, which is related but different. Please clarify what exactly you want to do or extract in your question if so.