Core2Vec: A core-preserving feature learning framework for networks

TL;DR

Core2Vec introduces a novel network embedding method that leverages node coreness to preserve core-based roles, outperforming existing techniques like node2vec and DeepWalk in word similarity tasks.

Contribution

It proposes a new core-preserving neighborhood concept and a corresponding embedding framework, Core2Vec, which captures node roles more effectively than traditional neighborhood-based methods.

Findings

Core2Vec outperforms node2vec, DeepWalk, and LINE in word similarity tasks.

Nodes with similar core numbers are closer in the learned vector space.

The method effectively captures core-based roles in networks.

Abstract

Recent advances in the field of network representation learning are mostly attributed to the application of the skip-gram model in the context of graphs. State-of-the-art analogues of skip-gram model in graphs define a notion of neighbourhood and aim to find the vector representation for a node, which maximizes the likelihood of preserving this neighborhood. In this paper, we take a drastic departure from the existing notion of neighbourhood of a node by utilizing the idea of coreness. More specifically, we utilize the well-established idea that nodes with similar core numbers play equivalent roles in the network and hence induce a novel and an organic notion of neighbourhood. Based on this idea, we propose core2vec, a new algorithmic framework for learning low dimensional continuous feature mapping for a node. Consequently, the nodes having similar core numbers are relatively closer in the vector space that we learn. We further demonstrate the effectiveness of core2vec by comparing word similarity scores obtained by our method where the node representations are drawn from standard word association graphs against scores computed by other state-of-the-art network representation techniques like node2vec, DeepWalk and LINE. Our results always outperform these existing methods