Analýza vícevariačních sítí

Abstract

Multivariate networks are made up of nodes and relationships in a form of links, but also data about these nodes and links as attributes. Real-world networks are often associated with several attributes a many analysis tasks depend on analyzing both, relationships and attributes. Most analysis methods are focused only on network topology contructed from multivariate data. In addition, these methods are often domain specific and require expert knowledge. The aim of this thesis is to create an web explorative tool which combines well-known methods for multivariate network exploration and analysis to enable structural and multivariate analysis. The manual or automatic creation of selections and attribute-based filtration of nodes enable the user to gain insights and understand the network better. Finally, examples of dataset analysis are presented to demonstrate the methods for multivariate network exploration.