Mentor 1
Jerry Scripps
Abstract
For most African nations, the cost of an Internet subscription is more than the average yearly per capita income – which inherently makes Internet access a quixotic amenity for the majority of Africans. This disproportionate connectivity creates a disadvantage for Africa’s academic potential because most of its Internet traffic is routed through international fiber optic links – which is costlier than direct connections within Africa [1]. One of the causes for slower Internet traffic is the lack of cooperation between the Internet Service Providers (ISPs) across Africa. This study will explore whether we can increase the average closeness of a sample network representing African Internet traffic by designing and testing strategic link-prediction algorithms versus a random link-prediction algorithm.