Big data real - time security analytics

Διπλωματική εργασία--Πανεπιστήμιο Μακεδονίας, Θεσσαλονίκη, 2017The magic triangle of IoT, Big Data and Cloud is currently ubiquitous, permeating the digital air around us and pervading into our daily physical and cyber lives. Awareness however is the critical factor when contemplating deploying novelties. Primary goal of this paper is to lay down inescapable security issues and challenges in the new era. Daunting grim thoughts are rendered impotent if we change mindset and utilize the double sword of technological advances in favor of security. What if big data instruments & advanced analytics are deployed selectively to fortify our critical assets from the constant fear of possibly well funded and acutely organized premium attacks? Whether or not the market of security analytics is evolving, the necessity to apprehend advanced security features is a commonplace in the contemporary cyber confrontation. Enterprise editions may have arisen but open source solutions are indeed indispensable. Spanning from Cyber Threat Intelligence and Analytics to recent rapidly developing User Entity Behavioral Analytics, predictive and prescriptive analytics do gain momentum, promising enormous power and numerous security benefits for their users. The already entrenched Hadoop premise has gradually paved the way for advanced distributed confrontation of computationally intensive tasks, however nowadays the trend moves forward to fully capture and demystify the supernatural velocity of generated data in Real – Time, giving birth to real–time optimized decision making. Our Apache Strom deployment was an endeavor to prove that real-time stream processing accompanied with open Security Intelligence feeds can be utilized to enhance our Security countermeasures. Numerous applications of our approach are possible in order to complement a wider defense-at-depth security model