Research On Important Hyperedge Recognition Method And Its Application In Hypernetwork

The field of network science,as an interdisciplinary discipline emerging in recent years due to the swift growth of complex systems and networks,has drawn considerable attention from various scholars in a variety of disciplines.Quantitatively assessing the significance of nodes or groups in complex networks is one of the most pressing issues to be addressed in this field.Compared with the fact that one edge in a complex network represents the direct adjacent relationship between two nodes,the unique "hyperedge" in a hypernetwork is more suitable for representing the group,team and community structure of multiple nodes.Identifying the importance of the team and community in the network is more conducive to accurately control the information transmission in the group,suppress the community-type epidemic outbreak,select important research achievements in the field,find important drug targets,and so on.The existing algorithms focus on identifying the important nodes in the hypernetwork,and the mining of important hyperedges is very few.This thesis discusses the important hyperedge recognition methods and applications of hypernetworks.The content of this thesis mainly includes the following two parts:(1)Based on the hypergraph theory and the minimum eigenvalue property of the grounded Laplacian matrix of the hypernetwork,this thesis proposes a new index MEGL to identify the important hyperedges in the hypergraph-based hypernetwork.It can be used in drug target hypernetwork to identify both important targets and important drugs.This method has important guiding significance for us to carry out drug development and target prediction,and also has certain reference significance for identifying important teams and communities in the hypernetwork.(2)This thesis proposes an algorithm to identify important hyperedges based on weighted hypernetworks,combining the local index of hyperedges with the weight of hyperedges,and proposes a comprehensive index,which combines the local and global importance of weighted hypernetworks.The graph theory theses published in the past six years have been applied to the scientific collaboration hypernetwork for empirical analysis.The evaluation results,when compared to the traditional index and the single index,demonstrate that this index is capable of accurately recognizing the significant theses in the journal,thereby confirming its efficacy.The weighted hypernetwork’s importance in recognizing teams and communities is highlighted by this method,which also has a significant reference value for recognizing theses of significance in a certain field.The purpose of this thesis is to discuss the recognition algorithm of important hyperedges in the hypernetwork,and propose two effective algorithms: one is based on the minimum eigenvalue of the Laplacian deleted matrix,and the other is based on the weighted hypernetwork.The two algorithms in this thesis are applied to the drug target hypernetwork and the scientific collaboration hypernetwork respectively.The experimental results show that these two algorithms can effectively identify the important hyperedges in the hypernetwork,and provide a certain reference basis and research ideas for the future research of the important hyperedge recognition and the importance ranking of hyperedges.