Parallel Computation of Random Walks on Distributed Graphs

A random walk is a process in which a random walker takes consecutive steps in space at equal intervals of time, with the length and direction of each step determined independently. Models related to random walks have been applied in diverse areas such as prediction, recommendation algorithms, and more recently, supervised learning and network analysis. Notably, there is a shortage of parallel computing models for random walks on distributed graphs across various scenarios. Therefore, the objective of this work is to offer a versatile model for random walks on distributed graphs, covering network principles and algorithmic models, followed by their parallel implementations. Evaluation results demonstrate the remarkable efficiency of our network model and highlight the future prospects of parallel random walks.