Presentation describing my implementation of a text prediction model built from a corpora of Twitter, blog, and news content. Several models and concepts were tested, and the Simple Back-Off 4-gram model was chosen for implementation due to its speed and accuracy.

A Presentation