SMS Spam Detection Using Noncontent Features

Short Message Service text messages are indispensable, but they face a serious problem from spamming. This service-side solution uses graph data mining to distinguish spammers from nonspammers and detect spam without checking a message's contents.

Antispam Approaches

Researchers have developed various computational approaches-in particular, data mining methods-to detect email spam, and some have achieved a certain degree of success. Content-based approaches³ were among the first to be applied. In email spam filtering, for example, such methods consider content-based features that can be used for classification. A spam email often contains some indicative keywords, such as “free” or “awards,” or unusual distribution of punctuation marks and capital letters, such as “BUY!!” or “MONEY,”⁴ such that these keywords become important features that a machine-learning-based classification algorithm can use.