Scaling machine learning with Apache Mahout by Isabel Drost

Today it is easy and comparably cheap to buy hardware capable of
storing terrabytes of data. Now people need a means to process and
analyze that data.

The talk gives a brief overview of Apache Mahout, a project with the
goal of building highly scalable machine learning algorithms: Scalable
here means commercially friendly licensed, vibrant community supporting
its users and naturally scalable to large amounts of data. The
presentation first motivates the need for machine learning techniques
when building great applications, shows which tasks can be solved with
Mahout so far and closes with an overview on coming developments in
Mahout.