Reply to comment

Collaborative filtering is a popular research area that is widely applied by many websites today. Amazon, Netflix and Google all use techniques from this field to recommend movies, books, musics and so on, to their visitors. This talk introduces you to collaborative filtering and shows how to create a recommendation engine using Apache Mahout / Taste. We start by covering basic concepts and discuss algorithms, Taste's architecture and extension points. We explain Taste along with a web application built with Apache Wicket and Taste that demonstrates several features. Additionally, we discuss how to customize your recommender and evaluate different algorithms. Finally, we conclude with lessons learned and go into some of the challenges of building recommendation engines and how Taste can solve some of these, now or in the future.