Evaluating Graph Neural Networks for Link Prediction: Current Pitfalls and New Benchmarking

TL;DR

This paper critically assesses current graph neural network methods for link prediction, identifies evaluation pitfalls, and proposes a new, more realistic benchmarking approach using hard negative sampling to improve future research.

Contribution

It provides a fair comparison of existing methods and introduces HeaRT, a heuristic-based negative sampling technique for more practical link prediction evaluation.

Findings

Current benchmarks often underestimate performance due to evaluation flaws.

HeaRT improves the difficulty and realism of negative sampling.

The new evaluation setting reveals more accurate model capabilities.

Abstract

Link prediction attempts to predict whether an unseen edge exists based on only a portion of edges of a graph. A flurry of methods have been introduced in recent years that attempt to make use of graph neural networks (GNNs) for this task. Furthermore, new and diverse datasets have also been created to better evaluate the effectiveness of these new models. However, multiple pitfalls currently exist that hinder our ability to properly evaluate these new methods. These pitfalls mainly include: (1) Lower than actual performance on multiple baselines, (2) A lack of a unified data split and evaluation metric on some datasets, and (3) An unrealistic evaluation setting that uses easy negative samples. To overcome these challenges, we first conduct a fair comparison across prominent methods and datasets, utilizing the same dataset and hyperparameter search settings. We then create a more practical evaluation setting based on a Heuristic Related Sampling Technique (HeaRT), which samples hard negative samples via multiple heuristics. The new evaluation setting helps promote new challenges and opportunities in link prediction by aligning the evaluation with real-world situations. Our implementation and data are available at https://github.com/Juanhui28/HeaRT