• Botnet Campaign Detection on Twitter

      Fields, Jeremy; Sengupta, Saumendra; Adviser; White, Joshura; Reviewer; Spetka, Scott; Reviewer (2016-08)
      The goal of this thesis is to investigate and analyze botnet activity on social media networks. We first start by creating an algorithm and scoring method for “likely bots,” and analyze them in conjunction with their neighboring messages to determine whether there is a likely group of bots, or botnet. Chapters 1 & 2 cover the overview of the work, and the previous research done by others. Multiple datasets were collected from Twitter, over different time frames, including random samples, and targeted topics. Chapters 3 & 4 cover the methodology and results of the approach using these datasets. The method is shown to have high accuracy.
    • Swarm Intelligence in the Multi Agent System Environment

      Bollam, Sravani; Sengupta, Saumendra; Adviser; Chiang, Chen-Fu; Reviewer; Rezk, Mohamed; Reviewer (2016-12-20)
      This thesis is focused on the use of naturally occurring concept of “swarm intelligence” to multiagentcsystems, namely the relatively new system-theoretic framework known as Swarm Intelligence Systems (SIS). In this work, outlined is the general framework of an interacting agents behaving as a swarm group on a mission to deliver a service such as transporting a set of goods from a given starting point to a destination. As an example, one could think of a team of drones, for instance, providing a delivery service in a milieu where normal transportation mode may be prohibitively expensive. The entire logistic, in detail as a commercial project, is too involved for our thesis focus; instead, some of the structural interaction issues with a group are outlined and discussed here to articulate the behavior of multi-agent based swarm as a group necessary to function as a distributed entity somewhat different from the standard swarm models found in literature.