Financial data visualization based on power-law degree distribution

Problems in the field of data visualization are "how to define the layout algorithm to discover the hidden relationships in the financial data" and "how to design and develop a reliable system to implement the layout algorithm". Although there were many solutions to deal with the problems, the methods mainly focused on visualizing network topologies or social networks. This thesis develops a novel tree layout algorithm based on a power-law degree distribution and a robust flash web application system to investigate relationships in financial time-series data. Compared to previous tree layout algorithms, this novel layout algorithm is flexible, scalable, and easy to implement. Additionally, this thesis presents the design structure and important features in the novel graph visualization system. This system provides a number of unique characteristics. For example, if a new focus node is selected, the feature of animating zooming with degree separations will track an animating transition from one to another layout for users. At last, based on many experiments, the thesis demonstrates the profound applications and practices of this system. When utilizing the visualization system to display the data in world crude oil price, we discover that the price fluctuations of crude oil within Organization of the Petroleum Exporting Countries (OPEC) or non-OPEC group are correlated, but price fluctuations of crude oil between difference groups are uncorrelated.