Hierarchical Data Visualization And Visual Comparison

Recent years,data science is widely developed.Meanwhile,more and more people pay attention to data visualization,which is not only an art of presentation,but also a science combining information retrieving,data processing and result showing.Hierarchical data is very common in our daily life,such as file systems,company organization,relation of news topic,taxonomy.Visualization of hierarchical data is a main topic in data visualization and the data is mainly visualized as tree structure.In era of data driven,different data are produced in all fields,and the data volume handled with is getting larger and larger.However,computer is still the primary presenter in visualization,whose screen size doesn't change that much these years.Therefore,it is a big challenge to handle massive data with limited visual area in visualization.It is the same to hierarchical data.For such kind of data,the node size increases exponentially by depth,which possibly causes visual cluster in the visualization,and degrades the user experience.Hence,it is necessary to study the methods of hierarchical data visualization and to make it easier to retrieve information from data.Methods of hierarchical data visualization are divided into four categories by the way of how edges visualized: indent form,node-link form,adjacency form and space-filling form.Base on previous work,a prototype of hierarchical data visualization system is presented and implemented herein.All four categories of methods are included in the prototype and basic analysis needs are satisfied.Tree comparison methods are also supported in our prototype.Comparison is a common means for us to understand things,acquire knowledge and discover patterns.In hierarchical data visualization,comparison plays an important role too,but there isn't a systematic summary of visual comparison when it comes to hierarchical data.In this paper,we give a detailed discussion of all comparison methods in literatures and then a classification methods is proposed.According to the present style,tree comparison methods can be divided into four categories: juxtaposition,superposition,matrix and animation.We analyse all methods for their advantages and disadvantages and then build the basic principles of the choice of methods.As we know,it is also the first try to conclude and study the methods of hierarchical data comparison in visualization systematically.