Using Personality Recognition Techniques to Improve Bayesian Spam Filtering

Millions of users per day are affected by unsolicited email campaigns. During the last years several techniques to detect spam have been developed, achieving specially good results using machine learning algorithms. In this work we provide a baseline for a new spam filtering method. Carrying out this research we validate our hypothesis that personality recognition techniques can help in Bayesian spam filtering. We add the personality feature to each email using personality recognition techniques, and then we compare Bayesian spam filters with and without personality in terms of accuracy. In a second experiment we combine personality and polarity features of each message and we compare all the results. At the end, the top ten Bayesian filtering classifiers have been improved, reaching to a 99.24% of accuracy, reducing also the false positive number.