This is my playground for using Mahout (the machine learning and recommendation library) with Clojure.

An example ratings.csv is included in the ./data directory. It contains CSV records of: userid,itemid,rating.

user> (def r (user-recommender (file-model "./data/ratings.csv")))
user> (recommend r 1)
({:item 105, :value 4.0} {:item 104, :value 3.1406214})

user> (estimate-user-preference r 1 104)
3.1406214
user> (estimate-user-preference r 1 107)
NaN

user> (evaluate #(item-recommender % (tanimoto %)) m)
1.4168497562408446

• Build DataModel implementations that back onto Cassandra/MongoDB
• Build part-distributed/pre-computed and part-live recommendation system

Copyright (c) 2011 Paul Ingles

Distributed under the Eclipse Public License, the same as Clojure.