Hi all, I am implementing an algorithm using Spark. I have one million users. I need to compute the similarity between each pair of users using some user's attributes. For each user, I need to get top k most similar users. What is the best way to implement this?
Thanks.
