Graph Kernels exploiting Weisfeiler-Lehman Graph Isomorphism Test Extensions

TL;DR

This paper introduces a new graph kernel framework based on the Weisfeiler-Lehman isomorphism test, offering fast computation and state-of-the-art performance on multiple datasets.

Contribution

The paper proposes a novel relabelling scheme and derives two new kernels inspired by WL tests, improving efficiency and accuracy in graph analysis.

Findings

Achieves state-of-the-art results on five datasets.

Kernel computation is very fast.

Framework is based on WL isomorphism tests.

Abstract

In this paper we present a novel graph kernel framework inspired the by the Weisfeiler-Lehman (WL) isomorphism tests. Any WL test comprises a relabelling phase of the nodes based on test-specific information extracted from the graph, for example the set of neighbours of a node. We defined a novel relabelling and derived two kernels of the framework from it. The novel kernels are very fast to compute and achieve state-of-the-art results on five real-world datasets.