Characterizing SMS spam in a large cellular network via mining victim spam reports

摘要

In this paper a study of SMS messages in a large US based cellular carrier utilizing both customer reported SMS spam and network Call Detail Records (CDRs) is conducted to develop a comprehensive understanding of SMS spam in order to develop strategies and approaches to detect and control SMS spam activity. The analysis provides insights into content classification of spam campaigns as well as spam characteristics based on sending patterns, tenure and geolocation.