Abstract:
The use of mobile phones continues to rise with no indication of slowing down. The increase in digitization of services to mobile platforms has resulted in more users becoming reliant on their mobile phones and in most cases trusting organizations offering these digital services. Fraudsters continue taking advantage of this trust and finding ways to defraud users. More than 25% of the 43 million mobile phone subscribers in Kenya have been victims of fraud. The purpose of this study was to design and implement a threat intelligence platform that would enable the reporting and sharing of fraudulent phone numbers as a way of combating fraud. The study had three specific objectives, which required the review of existing classification frameworks for fraudulent phone numbers, design and development of a web-based threat intelligence system with a classification framework, and evaluation of the developed system in terms of efficacy, validity and generality. The study adopted the Design Science Research (DSR) approach and a quasi-experimental methodology to test the developed system. Experiments with seven test cases were performed to establish the effectiveness of the developed platform named Coues. Coues had three modules namely: Crius, a control phone and a malware information-sharing platform. The research found that existing classification frameworks for fraud were not adequate in comprehensively classifying mobile phone fraud. By reviewing multiple frameworks, the study was able to establish a classification baseline that was built into Crius. To support public users and organizations in need of threat intelligence, a web-based system with a form and API was used. The web form allowed users to search the threat database, while the API allowed organizations to easily integrate Crius into their processes. Cyber threat intelligence sharing was achieved via the malware information sharing platform (MISP) and Twitter integration. As a recommendation, Coues is a repository of fraudulent phone numbers data which should be used as dataset for machine learning systems aimed at fighting mobile phone fraud. Additional work should be done to progressively mature the detection and cyber threat intelligence capabilities of Coues to help it scale.
