Comparing biased random walks in graph embedding and link prediction

TL;DR

This paper compares various biased random walk strategies in graph embedding and link prediction, finding that different biases have minimal impact on link prediction performance, indicating robustness in network structure recovery.

Contribution

It provides a comprehensive comparison of multiple biased random walk strategies and their effects on embedding quality and link prediction, highlighting their similar effectiveness.

Findings

Minimal performance differences across walk biases in link prediction

Embeddings reliably recover network structure regardless of walk heuristic

Sequence data from unknown mechanisms can be reconstructed successfully

Abstract

Random walks find extensive application across various complex network domains, including embedding generation and link prediction. Despite the widespread utilization of random walks, the precise impact of distinct biases on embedding generation from sequence data and their subsequent effects on link prediction remain elusive. In this study, we conduct a comparative analysis of several random walk strategies, each rooted in different biases: true self-avoidance, unbiased randomness, bias towards node degree, and inverse node degree bias. Furthermore, we explore diverse adaptations of the node2vec algorithm to induce distinct exploratory behaviors. Our empirical findings demonstrate that despite the varied behaviors inherent in these embeddings, only slight performance differences manifest in the context of link prediction. This implies the resilient recovery of network structure, regardless of the specific walk heuristic employed to traverse the network. Consequently, the results suggest that data generated from sequences governed by unknown mechanisms can be successfully reconstructed.