If you're the author, I'm not sure that the definition of recall@k is correct.<p>>Recall@k is a metric evaluate how many items, in the recommendation list, are relevant (hit) in the ground-truth data.
My collection of usefull links about recommender systems (dirty) <a href="https://gist.github.com/tandav/a2f87e91cb5c441c57657cceb788c86c" rel="nofollow">https://gist.github.com/tandav/a2f87e91cb5c441c57657cceb788c...</a>
Sadly but evaluation metrics are only implemented in pyspark.mllib (RDD API) but not in pyspark.ml (Dataframe API)<p>Also worth mention about implicit feedback