Graph modeling of software communities

Pervasive computing is an active field of research today and Pervasive Information Community Organization (PICO) is a novel approach towards building a framework for pervasive environments. PICO involves the concept of community computing based on a framework of intelligent software delegates namely the delegents, which work on behalf of hardware devices namely the camileuns. Delegents are autonomous software agents that possess reactive and proactive behavior and they work alongside other delegents to accomplish the goals of a pervasive community application.; This thesis work proposes a graph representation for delegents and their join and leave operations within communities. Simulation experiments were conducted to analyze the time and message complexities involved in merging and splitting of delegent communities. The graph-based approach was applied to model joining and leaving of delegents in a service provider community and experiments were conducted to observe the performance of service provisioning due to this approach.