Classifying Swahili Smishing Attacks for Mobile Money Users: A Machine-Learning Approach

Aim:

   Predict the Swahili smishing attack. Mobile money platform evolution could be attributed to the bureaucracy of owning a bank account, phishing attack on mobile user .This study proposes a machine-learning based model to classify Swahili phishing text messages targeting mobile money users.

Abstract:

     Due to the massive adoption of mobile money in Sub-Saharan countries, the global transaction value of mobile money exceeded $2 billion in 2021. Projections show transaction values will exceed $3 billion by the end of 2022, and Sub-Saharan Africa contributes half of the daily transactions. SMS (Short Message Service) phishing cost corporations and individuals millions of dollars annually. Spammers use Smishing (SMS Phishing) messages to trick a mobile money user into sending electronic cash to an unintended mobile wallet. Though Smishing is an incarnation of phishing, they differ in the information available and attack strategy. As a result, detecting Smishing becomes difficult. Numerous models and techniques to detect Smishing attacks have been introduced for high-resource languages, yet few target low-resource languages such as Swahili. This study proposes a machine-learning based model to classify Swahili Smishing text messages targeting mobile money users.

 Synopsis:

      Spam filtering has caught the interest of various researchers around the globe due to the unprecedented increase in spam message flow on networks. The proposed work spans from detecting spam, ham on social networks.

 Existing system:

     Over the years, mobile company operators have employed various ways to detect malicious text messages with little success. There are set of rules against every SMS going through an SMS gateway. Blacklist and white list techniques have also been employed to no available, because attackers keep on changing mobile numbers every now and then. Furthermore, blacklist and white list datasets are incapable of detecting zero-hour attacks and quickly become overpopulated and obsolete. User awareness programs on security good practice have not produced the desired results and are unlikely to reduce this vulnerability to zero.

Problem Definition:

       The Problem is mainly caused by the overconfidence of users, a belief that those who fall for social engineering attacks are idiots, and rapidly changing attack vectors.

Proposed System:

     Innspired by advancements in machine-learning techniques coupled with promising results obtained in message classification. This study proposes a machine-learning based model to classify Smishing text messages targeting mobile money users. Machine-learning techniques are advantageous to other techniques as they can detect both known malware and obfuscated malware. The contributions of this study, organized and carried out under a real-world Smishing dataset collected from mobile money users. The proposed model would save mobile money users from financial losses they incur as a result of social engineering attacks that keep on utilizing local dialects that are less studied.

 Advantage:

          Smishing messages targeting mobile money users use words in a well-orchestrated pattern and a mobile number to receive electronic money from a victim. The proposed model has a high accuracy score compared to general Smishing detection models. A lower accuracy for the dataset can be attributed to the fact that the formation of words and sentences in the language is very different from other well studied languages such as English, which has been extensively used by other researchers.