Local degree blocking model for link prediction in complex networks

TL;DR

This paper introduces a local degree blocking model for link prediction in complex networks, leveraging local structures to improve accuracy over traditional methods, and reveals the underlying principles of degree and short-path influence.

Contribution

The paper proposes a novel parameter-free local blocking predictor based on local link density, outperforming traditional similarity measures in real-world network tests.

Findings

LB index outperforms traditional methods on most networks

LB captures large degree and short path principles jointly

Scores of LB correlate with PA and short-path-based indices

Abstract

Recovering and reconstructing networks by accurately identifying missing and unreliable links is a vital task in the domain of network analysis and mining. In this article, by studying a specific local structure, namely a degree block having a node and its all immediate neighbors, we find it contains important statistical features of link formation for complex networks. We therefore propose a parameter-free local blocking (LB) predictor to quantitatively detect link formation in given networks via local link density calculations. The promising experimental results performed on six real-world networks suggest that the new index can outperform other traditional local similarity-based methods on most of tested networks. After further analyzing the scores' correlations between LB and two other methods, we find that the features of LB index are analogous to those of both PA index and short-path-based index, which empirically verify that large degree principle and short path principle simultaneously captured by the LB index are jointly driving link formation in complex networks.