Generally in recommendation engines, user’s past history on engagements with different items is a key input. However, in many situations in an enterprise’s business cycle, it is necessary to generate recommendations based on user activity in real time.
In this Big Data Cloud’s meetup on April 3, 2014, we discussed how to decipher real time click streams into meaningful recommendations in real time.
Pranab Ghosh discussed the real time recommendations feature of Sifarish, which is an open source project built on Hadoop, Storm and Redis.
Sifarish is a recommendation engine that does content based recommendation as well as social collaborative filtering based recommendation.
You can check out & download the complete presentation here: