Bounded link prediction for very large networks

TL;DR

This paper introduces a new framework and algorithm for efficient and accurate evaluation of link prediction methods in very large networks, focusing on CN-based indices and introducing a new performance measure.

Contribution

It proposes a fast parallel algorithm and a novel metric called self-predictability for assessing link prediction performance in large heterogeneous networks.

Findings

Efficiently identifies node pairs with high common neighbors.

Introduces the self-predictability measure for network link predictability.

Demonstrates improved evaluation accuracy in large-scale networks.

Abstract

Evaluation of link prediction methods is a hard task in very large complex networks because of the inhibitive computational cost. By setting a lower bound of the number of common neighbors (CN), we propose a new framework to efficiently and precisely evaluate the performances of CN-based similarity indices in link prediction for very large heterogeneous networks. Specifically, we propose a fast algorithm based on the parallel computing scheme to obtain all the node pairs with CN values larger than the lower bound. Furthermore, we propose a new measurement, called self-predictability, to quantify the performance of the CN-based similarity indices in link prediction, which on the other side can indicate the link predictability of a network.