Analyzing dynamical systems is challenging due to their high dimensionality and chaotic, nonlinear behavior. Motivated by this challenge, we develop graph representations of these systems that track both the structure of states and their temporal evolution. More precisely, we introduce two reconstruction methods: a direct binning approach that discretizes phase space into grid-based nodes, and a k-nearest neighbors (k-NN) approach in which nodes are defined by local neighborhoods. In both cases, directed edges encode temporal transitions between nodes. These graph constructions capture local geometric organization, recurrent behavior, and transition structure in the discretized state space. Consequently, graph-theoretic techniques such as community detection and loop detection can be used to identify structural signatures of the underlying dynamics.