A new approach to filtering spam SMS: Motif Patterns

Abstract

Along with the widespread of every technology, it comes with many problems. Mobile Short Message Service (SMS), which is widely used in mobile technologies, has brought many problems. The most important problem of SMS is unwanted messages named spam that are spread on the mobile network. Spam messages prevent mobile traffic and keep people busy unnecessarily. In this study to filter SMS spam, a novel feature extraction method, motif pattern method, is proposed, which uses forms that composed of comparision on UTF-8 codes of characters. In the proposed motif pattern method, the appearance of the values entered into a window size (PB) defined on the unicodes of SMS is considered as a motif pattern. The frequencies of these motifs in the SMS are used as the feature vector. The motif types depend on the specified PB. Three benchmark datasets were used to test the motif pattern method. The success rate was 93.76%, 90.07% and 94.29%, respectively, for three sets of data. According to the observed results, it is seen that the proposed method is a successful feature extraction method from SMS messages in spam filtering. It is also thought that the motif method can be used in other text mining, natural language processing fields.

Authors and Affiliations

Yılmaz Kaya, Cüneyt Özdemir

Keywords

Related Articles

A Survey of Hyper-parameter Optimization Methods in Convolutional Neural Networks

Convolutional neural networks (CNN) are special types of multi-layer artificial neural networks in which convolution method is used instead of matrix multiplication in at least one of its layers. Although satisfactory r...

Design of The High Efficiency Power Factor Correction Circuit for Power Supply

Designing power factor correction circuits for switched power supplies has become important in recent years in terms of efficient use of energy. Power factor correction techniques play a significant role in high power de...

Designing an Assistant System Encouraging Ergonomic Computer Usage

Today, people of almost every age group are users of computers and computer aided systems. Technology makes our life easier, but it can also threaten our health. In recent years, one of the main causes of the proliferati...

Design and Control of Multi – Input Multi – Output DC-DC Converter for Neutral Point Clamped Inverters

In this study, multi-input multi-output DC/DC converter topology is presented for both multi-source operation and voltage unbalancing of Neutral Point Clamped (NPC) inverters. Multi-source operation is provided with mult...

Effect of Graphene Nanoplatelets Reinforcement on the Microstructure and Mechanical Properties of AlSi10Mg Alloy

In this study, the effect of reinforcement of graphene nanoplatelets (GNPs that consist of a few graphene layers with a thickness of less than 100 nm and have extraordinary mechanical properties) on the microstructure an...

Download PDF file
  • EP ID EP490326
  • DOI 10.29109/http-gujsc-gazi-edu-tr.372880
  • Views 100
  • Downloads 0

How To Cite

Yılmaz Kaya, Cüneyt Özdemir (2018). A new approach to filtering spam SMS: Motif Patterns. Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji, 6(2), 436-450. https://europub.co.uk/articles/-A-490326