Export Reviews, Discussions, Author Feedback and Meta-Reviews

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. The paper addresses the problem of estimation of eigenvalues and eigenvectors of large sparse graph matrices. It uses a nice divide-and-conquer approach to obtain better estimates of top-k eigenspace. Such an estimate can be used in several classic tasks, such as link prediction and recommender systems. The paper is build upon a divide-and-conquer method, taking advantage of the particular characteristics of some graphs (e.g.