SMS Spam Filtering based on Text Classification and Expert System

Bozan Y. S., Coban Ö., Ozyer G. T., ÖZYER B.

23nd Signal Processing and Communications Applications Conference (SIU), Malatya, Türkiye, 16 - 19 Mayıs 2015, ss.2345-2348 identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/siu.2015.7130350
  • Basıldığı Şehir: Malatya
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.2345-2348
  • Anahtar Kelimeler: Spam SMS, Bayes, SVM, k-NN, expert systems, mobile system software, android
  • Atatürk Üniversitesi Adresli: Evet

Özet

Even though short message service(SMS) is gradually being replaced by social network sites' messaging systems, it still is one of the most widely used communication systems. For reasons such as well established e-mail filter systems, cheap SMS bundles and ineffective spam solutions for message services, SMS is one of the services which is widely used by advertising companies. In this study, expert system and data classification methods are proposed for SMS spam problem and a prototype software for Jelly Bean version of Android operating system. SVM, bayesian classification and k-NN classification methods applied and 98.61%, 97.55% and 93.35% successful results were obtained.