Content filtering for SMS systems based on Bayesian classifier and word grouping

There are many researches about e-mail spam filters. However, only a few look at the issue for SMS (Short Message Service) systems. This is a result of the difficulty in having access to SMS platforms of mobile operators. Furthermore, the volume of spam to SMS systems has increased year after year. The main objective of this study is to propose the implementation of a content filter for SMS systems based on the Bayesian classifier and word grouping. In order to evaluate the performance of this filter, 120,000 messages, sent from a content provider that services mobile operators, were tested. The results demonstrated that the proposed filter reached correct spam index detection close to 100%.