This represents a proof of concept work on using stylometrics and machine learning to detect Ghostwriting or contract cheating .
The project uses works from project Gutenburg as material to train various machine learning algorithms to determine wether 2 pieces of text were written by the same author.
The results are promising with the most successful algorithm producing a 77% accuracy rate.