Skip to main content
QUICK REVIEW

[论文解读] The Hyperspherical Geometry of Community Detection: Modularity as a Distance

Martijn Gösgens, Remco van der Hofstad|TU/e Research Portal|Jul 6, 2021
Complex Network Analysis Techniques参考文献 25被引用 6
一句话总结

本文提出了一种用于社区检测的超球面几何方法,将模块度最大化问题重新表述为在超球面上最小化与模块度向量的夹角距离。该方法将社区检测重新定义为两步投影过程——将网络映射到超球面上的点,再投影到聚类向量——为模块度提供了几何解释,解决了其分辨率限制问题,并实现了不依赖于模块度的新方法,这些方法在真实网络上的表现优于传统方法。

ABSTRACT

We introduce a metric space of clusterings, where clusterings are described by a binary vector indexed by the vertex-pairs. We extend this geometry to a hypersphere and prove that maximizing modularity is equivalent to minimizing the angular distance to some modularity vector over the set of clustering vectors. In that sense, modularity-based community detection methods can be seen as a subclass of a more general class of projection methods, which we define as the community detection methods that adhere to the following two-step procedure: first, mapping the network to a point on the hypersphere; second, projecting this point to the set of clustering vectors. We show that this class of projection methods contains many interesting community detection methods. Many of these new methods cannot be described in terms of null models and resolution parameters, as is customary for modularity-based methods. We provide a new characterization of such methods in terms of meridians and latitudes of the hypersphere. In addition, by relating the modularity resolution parameter to the latitude of the corresponding modularity vector, we obtain a new interpretation of the resolution limit that modularity maximization is known to suffer from.

研究动机与目标

  • 开发一种基于聚类超球面嵌入的社区检测几何框架。
  • 将模块度最大化重新解释为在超球面上最小化与模块度向量的夹角距离。
  • 将基于模块度的方法推广为更广泛的基于投影的社区检测技术类别。
  • 为模块度优化中的分辨率限制提供新的几何解释。
  • 设计并评估不依赖于零模型或分辨率参数的新型社区检测方法。

提出的方法

  • 将每个聚类表示为以顶点对为索引的二值向量,并将其嵌入度量空间。
  • 将聚类空间扩展至超球面,以支持夹角距离计算等几何操作。
  • 将模块度向量定义为超球面上的点,其纬度编码分辨率参数。
  • 将模块度最大化重新表述为最小化网络查询向量与模块度向量之间的夹角距离。
  • 引入一类通用的“投影方法”,即先将网络映射到超球面上的点,再投影到最近的聚类向量。
  • 采用相关距离(通过皮尔逊相关系数的反余弦函数获得)作为主要评估指标,实现几何不变性并避免偏差。
Figure 3 : Heatmap of the correlation between the ground truth and the candidate clusterings for the Karate network. We take query vectors from the meridians that $\left(\bm{q}_{M}^{\textrm{CM}}(G;\gamma)\right)_{\gamma\in[-1.5,2]}$ runs through and vary the query latitude between $\tfrac{1}{3}\pi$
Figure 3 : Heatmap of the correlation between the ground truth and the candidate clusterings for the Karate network. We take query vectors from the meridians that $\left(\bm{q}_{M}^{\textrm{CM}}(G;\gamma)\right)_{\gamma\in[-1.5,2]}$ runs through and vary the query latitude between $\tfrac{1}{3}\pi$

实验结果

研究问题

  • RQ1如何通过超球面嵌入的几何视角重新诠释社区检测?
  • RQ2模块度最大化能否等价地表述为在超球面上最小化与模块度向量的夹角距离?
  • RQ3当模块度被推广至超越零模型和分辨率参数时,会涌现出哪些新型社区检测方法?
  • RQ4该几何框架如何解释基于模块度方法中的分辨率限制?
  • RQ5不依赖于模块度的基于投影的方法是否仍能在真实世界网络中优于传统方法?

主要发现

  • 模块度最大化在几何上等价于最小化网络查询向量与超球面上模块度向量之间的夹角距离。
  • 模块度中的分辨率限制被重新解释为与超球面上模块度向量纬度相关的几何约束。
  • 所提出的投影方法类别推广了基于模块度的检测方法,并包含无法通过零模型或分辨率参数描述的方法。
  • 基于加权向量 w(G) 的新型投影方法在 Karate、Dolphins 和 Political Blogs 等真实网络中,优于标准模块度和 Louvain 算法。
  • 通过皮尔逊相关系数的反余弦函数获得的相关距离,是一种稳健且无偏差的评估指标,对聚类粒度具有不变性。
  • Louvain 算法可被解释为在超球面上的近似最近邻投影,提示可通过先进近似最近邻算法实现性能提升。
Figure 4 : Heatmap of the correlation between the ground truth and the candidate clusterings for the Football network. We take query vectors from the meridians that $\left(\bm{q}_{M}^{\textrm{CM}}(G;\gamma)\right)_{\gamma\in[-1,14]}$ runs through. The horizontal coordinates are given by $\textrm{sgn
Figure 4 : Heatmap of the correlation between the ground truth and the candidate clusterings for the Football network. We take query vectors from the meridians that $\left(\bm{q}_{M}^{\textrm{CM}}(G;\gamma)\right)_{\gamma\in[-1,14]}$ runs through. The horizontal coordinates are given by $\textrm{sgn

更好的研究,从现在开始

从阅读论文到最终审阅,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。