"New Model Provides Smishing Protection in Swahili"

Smishing (SMS phishing) extends scams to mobile devices by sending text messages impersonating organizations, such as banks, in order to obtain victims' personal information. Smishing is a major cybersecurity concern in Africa, which has the world's highest rate of mobile banking. Since electronic funds transfers are so common, there are numerous opportunities for an individual to lose money in a single day. Smishing is prevalent in Sub-Saharan Africa, particularly in Tanzania and Kenya, where the virtual banking service M-Pesa is widely used. The program lets users send money with their cell phones, and as part of the transaction, both the sender and the recipient receive an SMS confirmation notice. The greater the prevalence of mobile money transfers, the greater the risk, and the greater the number of people who are susceptible to smishing. Although mobile money transactions in Sub-Saharan Africa are projected to exceed $3 billion by the end of 2022, few robust resources are available in the Swahili language to monitor for smishing activity. There are several models for preventing attacks in high-resource languages such as English and Chinese, but not in the most common language of the most active mobile banking region. Therefore, Carnegie Mellon University (CMU)-Africa Assistant Teaching Professor Jema Ndibwile and his fellow researchers, Iddi S. Mambina and Kisangiri F. Michael, both of the Nelson Mandela Institute of Science and Technology, developed a Machine Learning (ML) based hybrid model that classifies Swahili smishing text messages targeting mobile money users. The algorithm determines which messages aimed at mobile money users are legitimate and which are smishing. According to the researchers, it has an accuracy rating of more than 99 percent. Over 30,000 SMS samples provided by college students were used to test the model. This article continues to discuss the ML-based model that provides smishing protection in Swahili. 

CMU Africa reports "New Model Provides Smishing Protection in Swahili"