arXiv 2402.104682024-02-16cs.LGcs.AI

基于成对增强的对抗性课程图对比学习

Adversarial Curriculum Graph Contrastive Learning with Pair-wise Augmentation

City University of Hong Kong（香港城市大学）；Shenzhen Research Institute of Big Data（深圳大数据研究院）；Duke Kunshan University（昆山杜克大学）；ByteDance Research（字节跳动研究院）

Xinjian Zhao, Liang Zhang, Yang Liu, Ruocheng Guo, Xiangyu Zhao

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对图对比学习中样本相似度难以控制的问题，提出ACGCL框架，利用成对增强生成可控相似度的样本并结合子图对比学习，通过对抗性课程训练克服稀疏性问题，在六个基准数据集上显著超越基线。

摘要：

图对比学习（GCL）已成为图表示学习领域的一项关键技术。有效GCL的一个关键方面是生成的正负样本的质量，这本质上取决于它们与原始数据的相似度。然而，在样本生成过程中精确控制相似度是一个巨大的挑战，常常阻碍了对具有代表性的图模式的有效发现。为了应对这一挑战，我们提出了一个创新框架：Adversarial Curriculum Graph Contrastive Learning (ACGCL，对抗性课程图对比学习)，该框架利用成对增强的优势，生成具有可控相似度的图级正负样本，并结合子图对比学习来识别其中有效的图模式。在ACGCL框架内，我们设计了一种新颖的对抗性课程训练方法，通过依次增加区分生成样本的难度来实现渐进式学习。值得注意的是，该方法通过自适应地集中于更具挑战性的训练数据，克服了传统课程学习策略中普遍存在的稀疏性问题。最后，在六个知名基准数据集上进行了广泛的综合评估，ACGCL显著超越了一系列最先进的基线方法。

英文摘要：

Graph contrastive learning (GCL) has emerged as a pivotal technique in the domain of graph representation learning. A crucial aspect of effective GCL is the caliber of generated positive and negative samples, which is intrinsically dictated by their resemblance to the original data. Nevertheless, precise control over similarity during sample generation presents a formidable challenge, often impeding the effective discovery of representative graph patterns. To address this challenge, we propose an innovative framework: Adversarial Curriculum Graph Contrastive Learning (ACGCL), which capitalizes on the merits of pair-wise augmentation to engender graph-level positive and negative samples with controllable similarity, alongside subgraph contrastive learning to discern effective graph patterns therein. Within the ACGCL framework, we have devised a novel adversarial curriculum training methodology that facilitates progressive learning by sequentially increasing the difficulty of distinguishing the generated samples. Notably, this approach transcends the prevalent sparsity issue inherent in conventional curriculum learning strategies by adaptively concentrating on more challenging training data. Finally, a comprehensive assessment of ACGCL is conducted through extensive experiments on six well-known benchmark datasets, wherein ACGCL conspicuously surpasses a set of state-of-the-art baselines.

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。