Short Message Service (Sms) Spam Detection and Classification Using Naïve Bayes
Publish place: International Journal of Mechatronics, Electrical and Computer Technology، Vol: 11، Issue: 40
Publish Year: 1400
نوع سند: مقاله ژورنالی
زبان: English
View: 116
This Paper With 6 Page And PDF Format Ready To Download
- Certificate
- من نویسنده این مقاله هستم
این Paper در بخشهای موضوعی زیر دسته بندی شده است:
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
JR_IJMEC-11-40_002
تاریخ نمایه سازی: 28 تیر 1402
Abstract:
The dynamic nature of technology has caused an unprecedented technological and socio-economical development in everyday life. This development is making everyone to be highly vulnerable to diverse threats. Short Message Service (SMS) spam is one of such threats that affect the security of mobile devices. These spam attempts to deceive users into providing their private information which could later result in a security breach. The major problem of SMS spam is that attackers, hackers, email phishing, ransomware, etc., used it to exploit the victims. The urge to curb this has necessitated this work. Different models have been developed to detect SMS spam, some of these models include Support Vector Machine, Linear Classifier, Decision Trees, Random Forest, Logistic Regression, Naive Bayes, etc. However, most of these techniques have not addressed the point that focuses on SMS spam detection and classifies new SMS spam. The goal of this research is to develop a machine learning model for the detection and classification of new SMS spam. This paper presents a model for SMS spam detection and classification that employs the Naïve Bayes machine learning methodology. String to word vector feature extraction was used to extract the SMS Spam text file from the contents that were collected via UCL repository in its original form. At this point, the proposed system is set to perform data preprocessing, dataset feature extraction, and model training as well as model evolution. The model was learned based on an SMS dataset that consists of ۵۵۲۵ samples collected from an online resource and utilized effectively. The experimental results indicate classification accuracies of ۹۹.۴۲%, for correctly classified and ۰.۵۷% for incorrectly classified, respectively in the best cases.
Keywords:
Authors
Christine Bukcola Asaju
Department of Computer Science, Federal Polytechnic Idah, Idah, Kogi State, Nigeria
Ekuma James Ekorabon
Department of Computer Science, Federal Polytechnic Idah, Idah, Kogi State, Nigeria
Richard Ojochegbe Orah
College of Information and Communication Technology Salem University Lokoja Kogi State, Nigeria