This demonstration proposes StreamRec, a novel approach to building recommender systems that leverages a stream processing system capable of handling an end-to-end recommendation process in order to produce real-time recommendations. We demonstrate several popular collaborative filtering recommendation methods within StreamRec by providing an application scenario that uses StreamRec as the underlying recommendation engine.

PDF file

In  ACM SIGMOD International Conference on Management of Data (SIGMOD 2011)

Publisher  ACM SIGMOD

Details