Context development for internet security in the big data analytics world

Abstract

For the first time computer launched foreign assaults on U.S. infrastructure were ranked higher in the U.S. intelligence community's annual review of worldwide threats than worries about terrorism (Dilanian, 2013). That alone points to the importance and criticality of internet security for national as well as international security. As the world heads towards massive penetration of mobile devices, as the era of IOT (Internet of Things) dawns and as the world ultimately transitions from early-Pervasive to Ubiquitous computing, the internet and therefore internet security are bound to be massively and inextricably enmeshed with human existence, while multiplying by many orders on complexity, sensitivity and criticality. This paper traverses the vast landscape of internet security quickly to construct and provide a first ever big picture of the internet security context. It finds that most existing paradigms, already insufficient to provide sustainable internet security, are going to fall woefully short either on protection or coverage, or even be rendered obsolete. The paper then discusses the necessity for Big Data analytics and machine learning. Taking cues from the shortcomings pointed out by early Big Data practitioners, the paper recommends metadata and context driven Big Data practice for security analytics. The importance of machine learning for delivering context and context-awareness, is emphasized, to emerge with a set of potential winning approaches centering on malbehavior prevention and resilience. A contextual framework for malbehavior that may serve as the core of context-aware security systems, and a wider contextual framework for sustainable internet security are also proposed.