Skip to Main Content
 

Global Search Box

 
 
 
 

Files

ETD Abstract Container

Abstract Header

USING GRAPH MODELING IN SEVERAL VISUAL ANALYTIC TASKS

Abstract Details

2016, PHD, Kent State University, College of Arts and Sciences / Department of Computer Science.
Graph models can represent a variety of data types such as social media, cyber business and security, web, urban networks, and more. They are extensively studied and widely used in data management, mining, and analysis in many important application areas. On the other hand, graph visualization has been a major topic in information visualization to manifest graph structure and features for effective and intuitive data exploration. In this thesis, we present a set of visual analytics solutions for several important applications by integrating graph models with visualization tools, including the visualization systems of urban trajectory data, text stream data, and categorical data. Our approaches utilize graphs to abstract and manage various data, to discover hidden knowledge with graph algorithms, and to help users gain insights from graph-based visualizations and interaction. Our research widens the horizon and enhances the capability of visual analytics methodologies. First, we propose a new visual analytics method, TrajGraph, for studying urban mobility patterns. In particular, a graph model represents taxi trajectories traveling over road networks. Then graph computation is applied to identify graph centralities that find the time varying hubs and backbones of road networks from massive taxi trajectories. The graph is further visualized and interacted for users to explore the important roles of city streets and regions. Second, we employed a parallel-graph model to enhance visual analytics of the large-scale urban trajectory datasets. Specifically, we designed a novel, scalable parallel-graph model for trajectory data management. It supports fast computation over various information queries in distributed environments. A new visualization tool that allows users to get statistics information, and relationship of cars and roads in the big trajectory data by employing the functionalities of the parallel-graph model. Third, we develop a dynamic visualization system to tell city stories extracted from massive city news. A news stream is rendered by gradually evolving animated visualization aimed to help users observe and understand the dynamical topics, events and trends of cities. An incremental clustering scheme is developed over an evolving graph consisting of the streaming news. The clusters are visualized based on text summarization so that users can easily explore salient and changing focuses of the text stream. Finally, we optimize the parallel set visualization of large categorical datasets by applying the graph algorithm of the traveling salesman problem (TSP). The graph represents visual layouts with different orders of parallel coordinates. The graph edge weights are defined as the mutual information computed from categorical datasets. We modify the minimization optimization to a maximization solution of the sum of mutual information to create optimal ordering and visual results.
Ye Zhao (Advisor)
Ruoming Jin (Committee Member)
Chengchang Lu (Committee Member)
Xinyue Ye (Committee Member)
Donald White (Committee Member)
210 p.

Recommended Citations

Citations

  • Huang, X. (2016). USING GRAPH MODELING IN SEVERAL VISUAL ANALYTIC TASKS [Doctoral dissertation, Kent State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=kent1467738860

    APA Style (7th edition)

  • Huang, Xiaoke. USING GRAPH MODELING IN SEVERAL VISUAL ANALYTIC TASKS. 2016. Kent State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=kent1467738860.

    MLA Style (8th edition)

  • Huang, Xiaoke. "USING GRAPH MODELING IN SEVERAL VISUAL ANALYTIC TASKS." Doctoral dissertation, Kent State University, 2016. http://rave.ohiolink.edu/etdc/view?acc_num=kent1467738860

    Chicago Manual of Style (17th edition)