Visual Analysis of Multisource Heterogeneous Data Based on Improved DPCA Algorithm

This paper proposes a multiview collaborative visual analysis system of network security based on a DPCA (clustering by fast search and find of density peaks) clustering algorithm called DPCANETVis, with network security analysis requirements for multisource heterogeneous data. Firstly, the system p...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Mathematical problems in engineering Ročník 2022; s. 1 - 10
Hlavní autoři: Zhou, Yun, Mengfei, Wen, Youzhi, He, Yun, Cheng, Jinhui, Lv, Yi, Zuo
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York Hindawi 06.12.2022
John Wiley & Sons, Inc
Témata:
Algorithms
Behavior
Clustering
Data mining
Graphical representations
Mathematical problems
Network security
Security
Systems design
Visualization
ISSN:1024-123X, 1563-5147
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:This paper proposes a multiview collaborative visual analysis system of network security based on a DPCA (clustering by fast search and find of density peaks) clustering algorithm called DPCANETVis, with network security analysis requirements for multisource heterogeneous data. Firstly, the system proposes an improved DPCA clustering algorithm based on the hierarchical relationship of mail sending and receiving to achieve the purpose of accurate classification. Secondly, a three-layer visual layout is designed to display relevant information such as data hierarchy, node relationship, behavior model, and other relevant information, by mixing a variety of interactive visual analysis methods such as tree diagram, word cloud, line diagram, subject river, and parallel coordinate. Lastly, based on the exploration of events, all suspicious nodes and their abnormal behaviors can be displayed in the system. Finally, the prototype system is used to analyze the network security log data set provided by the ChinaVis 2018, and the feasibility of the multilevel interactive visual analysis method for network security is verified through many experiments and discussions.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1024-123X
1563-5147
DOI:10.1155/2022/7895544