Optimized One Dimensional-Ternary Pattern (1D-TP) for SMS Spam Feature Extraction

International Journal of Research and Innovation in Applied Science (IJRIAS) | Volume V, Issue III, January 2020 | ISSN 2454–6186

Optimized One Dimensional-Ternary Pattern (1D-TP) for SMS Spam Feature Extraction

 Oluwakemi Christiana Abikoye¹, Fatimoh Abidemi Taofeek-Ibrahim², Taye Oladele Aro^3
1Department of Computer Science, University of Ilorin, Ilorin
²Department of Computer Science, Federal Polytechnic Offa, Kwara, Nigeria
³Department of Mathematical and Computing Sciences, KolaDaisi University, Ibadan, Oyo State, Nigeria

Abstract: Short Message Service (SMS) has become an important form of the mobile communication channel; their popularity is attributed to several conditions including low-cost sending, simple delivery mode and convenient usage. Feature extraction phase has been identified by researchers to be one of the major steps in the Spam SMS detection system. The extraction of features in SMS involves a process of reduction of an initial set of raw features into more manageable forms for processing. This paper employed a new One Dimensional Ternary Pattern (1D-TP) for SMS feature extraction while the simulated annealing was applied to optimize the extracted features. Experimental results showed a better objective function

Keywords: Feature Extraction, Short Message Service, Spam SMS, One Dimensional Ternary Pattern,

I. INTRODUCTION

Short Message Service (SMS) technology evolved out of the GSM Communications standard, an internationally accepted cell phone network specification (Geertsema, Hyman, & van Deventer, 2011). SMS remains the most powerful tool in terms of communication especially for mobile users (Reaves, Blue, Tian, Traynor, & Butler, 2016). It does not limit anyone regardless of high-or low-end mobile phones for as long as they can receive and send messages anytime, anywhere (Shaw & Bosworth, 2012). In a fragmented mobile world of multiple devices, operating systems and service providers, messaging remains the one constant that offers a singular ubiquitous channel through which all end users can communicate with each other (Downer, Meara, Da Costa, & Sethuraman, 2006). SMS grows beyond traditional texting and is now being used in different authentication domains such as mobile banking, one-time password delivery, information retrieval systems, smartphone configuration, Over-The-Air (OTA) configuration and social web site alerts (Mizuki, Matsumoto, Uemura, & Kichimi, 2013).
There is still a major threat of SMS Spam in the mobile communication world despite the different benefits associated with SMS (Abdulhamid, Shafie, Latiff, Chiroma, & Osho, 2017). SMS Spam commonly refers to the unsolicited and unwanted SMS usually conveyed to a large number of recipients (Mahmoud & Mahfouz, 2012). Spam is usually sent in bulk for commercial or other purposes and indiscriminately (Chaudhari, Jayvala, & Vinitashah, 2016)

Read More