学习写作--polyGCL.md

POLYGCL: GRAPH CONTRASTIVE LEARNING VIA LEARNABLE SPECTRAL POLYNOMIAL filters
这篇工作的摘要和引言写的特别好（不愧是ICLR spotlight）

摘要

第一步，设定背景

Recently, Graph Contrastive Learning (GCL) has achieved significantly superior performance in self-supervised graph representation learning.
这句话确立了索要解决问题的背景和重要性。他指明了对比学习在无监督图表示学习的重要性

第二步，提出问题：

However, the existing GCL technique has inherent smooth characteristics because of its low-pass GNN encoder and objective based on homophily assumption, which poses a challenge when applied to heterophilic graphs.
在这里，作者详细的介绍了传统方法的挑战或者缺点，其标明传统GCN使用的低频GNN的假设在异配图上不成立

第三步，引入有监督学习的方法：

In supervised learning tasks, spectral GNNs with polynomial approximation excel in both homophilic and heterophilic settings by adaptively fitting graph filters of arbitrary shapes.
在监督领域，自适应匹配任意形状滤波器的谱GNN表现情况很出色

第四步，很自然的，他们在无监督领域匮乏

Based on the above analysis, a natural question arises: Can we incorporate the excellent properties of spectral polynomial filters into graph contrastive learning?
那么，有监督领域该如何做呢？

第五步，引入解决方案：

从高通信息来实现解决异配性问题。

第六步，解决方案以及细节

We propose POLYGCL, a GCL pipeline that utilizes polynomial filters to achieve contrastive learning between the low-pass and highpass views. Specifically, POLYGCL utilizes polynomials with learnable filters to generate different spectral views and an objective that incorporates high-pass information through a linear combination.We theoretically prove that POLYGCL outperforms previous GCL paradigms when applied to graphs with varying levels of homophily.

最后一步，套话，我们在很多数据集上都取得了不错的结果

引言

第一步:确定话题的重要性

Self-supervised representation learning, which aims to learn informative representations without the demand of costly handcrafted labels, has achieved a wide range of applications in areas such as computer vision, natural language processing, and multimodal (Chen et al., 2020; Grill et al., 2020; Devlin et al., 2019; Radford et al., 2021; Gao et al., 2023). On non-Euclidean graph data, Graph Contrastive Learning (GCL), has become a mainstream research direction in self-supervised scenarios, namely learning representations by capturing consistency across different views and optimizing the objective function based on mutual information maximization to distinguish positive and negative examples (Velicˇkovic ́ et al., 2019; Hassani & Khasahmadi, 2020; Peng et al., 2020; Zhu et al., 2020b).
首先，第一句设置了问题的背景：在CV，NLP等领域，对比学习取得了巨大的进展。其次，第二句限定了当前的研究范围：图对比学习成为自监督场景的主流研究方向，它通过捕获不同视图的一致性并基于互信息最大化的优化目标函数来学习以区分积极和消极示例。
第三部分是过去的方法常用了什么模块。
现有的大部分GCL都依赖于同配性假设。其往往采用低通滤波器来是的编码器和平滑临近节点的表示目标。因此，GCL在同配图上表现出色（承上启下句子）