The Project:

The company Netflix allows customers to rent movies online for a flat monthly fee, rather than charging on a per movie basis. Customers leave feedback about the movies they watch in the form of ratings, which can range from 1 to 5 stars. Netflix then uses these ratings to make recommendations to their customers, tailoring them to suit each individual's taste.

In 2006, Netflix announced a competition to improve their existing recommendation system. They released a data set containing millions of ratings that contestants could use to design their systems, and they promised a $1,000,000 prize to the first contestants to beat Netflix's system by %10 or more.

Although the competition ended before we began our comps, we took the Netflix prize as a starting point. We pursued three goals in our project:

• Try solving the Netflix prize with several algorithms
• Find patterns in the data through the use of clustering techniques
• Visualize the data set

Use the links in the sidebar to explore the algorithms we implemented as part of our project.

The Team:

This Computer Science Senior Comps project was created by Kei Kubo, Peter Nelson, Erik Ruggles, James Sheridan, and Sam Tucker. We thank our advisor, Dave Musicant, for all of his help, and we hope you find the website informative.