Characterizing attitudinal network graphs through frustration cloud

Abstract Attitudinal network graphs are signed where edges capture an expressed opinion; two vertices connected by edge can be agreeable (positive) or antagonistic (negative). A graph is called balanced if each of its cycles includes even number negative edges. Balance often characterized the frustration index finding a single convergent state consensus. In this paper, we propose to expand measures consensus from associated with set nearest states. We introduce cloud as all states and use graph-balancing algorithm find in deterministic way. Computational concerns addressed measuring probabilistically, new vertex metrics quantify status , agreement influence . also global measure controversy for given show that zero-sum game network. efficient scalable calculating cloud-based social survey data up 80,000 half-a-million demonstrate power proposed approach provide discriminant features community discovery when compared spectral clustering automatically identify dominant anomalous decisions