Research On Methods Of Large-scale Hierarchical Information Visualization

Information Visualization, as a precise mathematical model of visualization technology, is a fast growing and far-reaching cross-cutting research areas and has been widely used in information retrieval (IR) with the World Wide Web as an information carrier, and digital libraries (DL) and the design of human-computer interaction interface (HCI) model. It is a computer-aided technology of interactive mining the potential links within the complex data structures based on the human cognitive abilities.Hierarchical structure, as a data organization, is more common organization of data in the real world. In the study of hierarchy theory of information visualization technology, abstract information as a node in the tree, while the link between information within the node will be abstracted an edge within the tree, it will eventually form a tree structure corresponds to the abstract information, using visualization techniques to display the tree structure within a valid physical reality region, and ultimately achieve the purpose of enhancing the user cognitive abilities. There are many important research results in the field of theory and method of the hierarchical information visualization of tree structure. However, the existing theory and technology of visualization for hierarchical information visualization within a limited area, with increasing the complexity of the data structure and scale, will have significant limitations.Firstly, we will discuss and make improvements to the classical random tree generation algorithms, make the final random generated tree has more good performance, use the tree structure data set generated by the improved algorithm to test efficiency of tree layout algorithm, visualize the tree structure on the physical screen. At the same time, we will discuss the problem arising in the visualization and make the improved algorithm can be applied to large-scale tree visualization and navigation.The main contents of this study include the following aspects: (1) Discussing the research status theory of the current typical tree visualization technology, comparing the advantages and disadvantages of each kind of algorithms, at the same time, we point out the drawback of existing algorithms for large-scale visualization of hierarchical information.(2) The existing Random Tree generation algorithms are mainly designed for artificial intelligence, it does not need to generate large-scale tree structure, and therefore, there are few researches of random tree generation algorithm for testing the large-scale tree layout algorithm. In this paper, we propose the algorithm with limited scale, which produce a balance tree by joining the controllable factors, the result shows that, the data sets generated by the improved algorithm can be better used to test the performance of the tree layout algorithm.(3) Circular drawing algorithm has good effect on small-scale and uniform tree structure visualization, but the scalability is poor. There will be many nodes overlap when the scale increases, and the poor dynamic interaction. User can’t select the focus content to observe the interesting data. In this paper, we will join the dynamic interaction properties and fix the calculation formula of the radius on the basis of the original algorithm, realizing the drawing with different spacing. It will make the focus circular more clearly by using the dynamic interaction with different spacing, the improved algorithm can be better used for large-scale tree structure visualization and navigation.(4) The experimental procedure involved, in this paper, is designed based on the eclipse platform; the experimental results indicate that our proposed algorithm can be used to solve the difficulty of information visualization techniques.