arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2204.01376cs.LGcs.SI
收藏

用于基准图学习的合成图生成

Synthetic Graph Generation to Benchmark Graph Learning

University of Bonn(波恩大学);The University of Edinburgh(爱丁堡大学);Google Research(谷歌研究院)

Anton Tsitsulin, Benedek Rozemberczki, John Palowitch, Bryan Perozzi

摘要:

图学习算法在许多图分析任务上已经达到了最先进的性能,例如节点分类、链接预测和聚类。然而,追踪该领域的迅速发展已经变得很难。一个原因是由于在实践中用于对图学习算法的性能进行基准测试的数据集数量非常少。这个令人震惊的小样本量(~10)只允许对问题进行有限的科学洞察。在这项工作中,我们的目标是解决这一缺陷。我们建议生成合成图,并研究在受控情况下的图学习算法的行为。我们开发了一个功能齐全的合成图生成器,可以对不同的模型进行深入检查。我们认为,合成图生成允许算法的彻底调查,并提供了更多的见解比过拟合三个引文数据集。在案例研究中,我们展示了我们的框架如何提供对无监督和监督图神经网络模型的洞察。

英文摘要:

Graph learning algorithms have attained state-of-the-art performance on many graph analysis tasks such as node classification, link prediction, and clustering. It has, however, become hard to track the field's burgeoning progress. One reason is due to the very small number of datasets used in practice to benchmark the performance of graph learning algorithms. This shockingly small sample size (~10) allows for only limited scientific insight into the problem. In this work, we aim to address this deficiency. We propose to generate synthetic graphs, and study the behaviour of graph learning algorithms in a controlled scenario. We develop a fully-featured synthetic graph generator that allows deep inspection of different models. We argue that synthetic graph generations allows for thorough investigation of algorithms and provides more insights than overfitting on three citation datasets. In the case study, we show how our framework provides insight into unsupervised and supervised graph neural network models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑