Dynamic Graph Embedding Through Hub-aware Random Walks

TL;DR

This paper introduces DeepHub, a dynamic graph embedding method that incorporates hub sensitivity into random walk sampling, improving the representation of less-connected nodes and overall embedding stability in evolving networks.

Contribution

The paper proposes a novel hub-aware random walk strategy for dynamic graph embedding, systematically analyzing its impact across multiple real-world networks.

Findings

Hub-aware walks balance exploration of nodes with different degrees.

Standard walks overrepresent hubs, leading to less accurate embeddings.

Hub sensitivity improves downstream task performance.

Abstract

The role of high-degree nodes, or hubs, in shaping graph dynamics and structure is well-recognized in network science, yet their influence remains underexplored in the context of dynamic graph embedding. Recent advances in representation learning for graphs have shown that random walk-based methods can capture both structural and temporal patterns, but often overlook the impact of hubs on walk trajectories and embedding stability. In this paper, we introduce DeepHub, a method for dynamic graph embedding that explicitly integrates hub sensitivity into random walk sampling strategies. Focusing on dynnode2vec as a representative dynamic embedding method, we systematically analyze the effect of hub-biased walks across nine real-world temporal networks. Our findings reveal that standard random walks tend to overrepresent hub nodes, leading to embeddings that underfit the evolving local context of less-connected nodes. By contrast, hub-aware walks can balance exploration, resulting in embeddings that better preserve temporal neighborhood structure and improve downstream task performance. These results suggest that hub-awareness is an important yet overlooked factor in dynamic graph embedding, and our work provides a foundation for more robust, structure-sensitive representation learning in evolving networks.