Preserving the Value of SMS Texting:A Survey on Mobile SMS Spam Classification Techniques and Algorithms

Text classification concept can be realized in spam filtering for Email/SMS domain. There had been a variety of classification strategies used for email categorization including both supervised and unsupervised. SMS spam filtering is a new task comparatively to email spam filtering which inherits many issues and solutions from email spam filtering. The increase in the mobile phone users has lead to a substantial growth of SMS spam messages. Due to increase in use of Short Message Service (SMS) over mobile phones because of cheap and bulk pre-pay sms packages, there has been a burst of spam SMSes. Along with the range of classification algorithms the use of SMS features is also varied accordingly from non-content based to content based. In this survey paper the previous attempts in SMS spam domain and recent developments in SMS spam filtering have been summarized that contributes in preserving the value of SMS by reducing the spamminess of the SMS. Some of the algorithms that contributed in SMS spam filtering and their comparison is discussed. Further SVM based spam filtering system is proposed, that classifies spam and non-spam SMSes. This filtering system can further classify non spam SMSes in predefined categories that can be viewed in different tabs for comparatively easier and faster user access to a particular SMS.