An Algorithmic Solution for the “Hair Ball” Problem in Data Visualization
报告人：刘云凯（Gannon University计算机系副教授，主要学术方向是图论算法， 图数据库， 和图视觉显示的研究，还包括各种大数据的应用和解决方案， 包括 Smart City, Urban Computation, Social Media Mining and Real-time Database. ） 报告摘要：The investigation and analysis of large and complex graphs is an important aspect of data visualization research, yet there is a need for entirely new, scalable approaches and methodologies for graph visualization. This can ultimately provide more insight into the structure and function of this complex graph. To explain more, a methodology is needed as a solution to present a “tidy” graph with the minimal crossover between edges in the “Hair Balls.” In spite of the expanding significance of investigating and extensively analyzing and understanding very large graphs of data, the traditional way of visualizing graphs has difficulties scaling up, and typically ends up depicting these large graphs as “Hair Balls”. This traditional approach does indeed have a deeply intuitive foundation: nodes are depicted with a shape such as a circle, triangle or square, which are then connected by lines or curves that represent the edges. In any case, although there are many different ways to apply this basic underlying idea, it needs to be revisited in light of current and emerging needs for understanding increasingly complex crossover between edges in the graphs. The complex “Hair Ball,” which appears as an indecipherable graph, came from the crossover between edges. In our preliminary research, a feasible computational solution is developed based on the discovery of K2,3 subgraphs. Some large graphs with “hair balls” are tested. The computational performance is identical with the Big-O results.