Answers are grounded in this paper and cite their sources.
Mobile money fraud has become so prevalent with users losing their hard earn income and Service Providers finding it difficult to cope with this new reality.Several authors in literature have proposed different approaches ranging from expert systems, machine learning and hybrid approaches.However, these approaches have not been able to handle the problem of efficiency and rapidity in mobile money fraud detection, while taking into consideration the imbalance nature of financial datasets and the dimensionality explosion problem associated with these datasets.This article proposes a novel dynamic Locality Sensitive Hashing approach to efficient and rapid detection and prediction of mobile money fraud related to SMS and Transaction data, based on the application of Search Based Software Engineering in machine learning while tackling the limitation brought about as a result of imbalance data inherent in financial datasets and the dimensionality explosion problem.A comparative analysis of our approach with the state-of-the-art techniques demonstrates promising results on both efficiency and rapidity.The results showed a 98.6% efficiency for SMS fraud data detection and prediction and 99% for transaction data while yielding an O(n) running time for both dataset where n is the dataset size.These results are very encouraging and will go a long way to reducing the amount of fraud on mobile money networks and hence an increase in customer trust and adoption of mobile money services.
No takes yet. Share an insight, caveat, or question.
Youh et al. (2025) studied this question.
Synapse has enriched 4 related papers. Consider them for comparative context: