Source Themes

Fundamental statistics of higher-order networks: a survey(高阶网络统计指标综述)

*物理学报*. 本文综述了两种最常见的高阶网络——超图和单纯形网络——常用的统计指标及其物理意义. 本文有助于加深对高阶网络的理解, 促进对高阶网络结构特征的定量化研究, 也有助于研究者在此基础上开发更多适用于高阶网络的统计指标.

Identifying vital nodes through random walks on higher-order networks

*Information Sciences.* Developed a Higher-order Augmented Random Walk (HoRW) model to identify influencers, enabling multi-scale analysis according to the strength of higher-order effects.

Hyper-Null Models and Their Applications

*Entropy.* This study presents a novel hyperedge swapping method to construct hyper-null models for hypergraphs, which preserves certain network properties while altering others, and demonstrates their applicability in assessing network randomness, statistical properties, and dynamics across multiple datasets.

Cooperative Network Learning for Large-Scale and Decentralized Graphs

Developed a cooperative network learning (CNL) framework using technologies like homomorphic encryption, enabling decentralized, multi-party trusted, and privacy-preserving graph learning.

Higher-order Graph Convolutional Network with Flower-Petals Laplacians on Simplicial Complexes

*AAAI24*. Introduced a novel higher-order representation, the flower-petals (FP) model, and higher-order graph convolutional network (HiGCN), which achieves SOTA in various tasks and quantifies higher-order strength.