Interactive Visualization of the Market Graph

Financial markets are a fruitful area for data exploration, but the overwhelming size and dimension of the datasets usually prohibit meaningful analysis, especially on a large scale. Thus, there is a need for effective visualization tools to assist in efficiently exploring the data space. In this paper, we present a novel visualization tool that empowers a user with an interactive tool for finding meaningful relationships in historical or real-time financial market data. To reduce the size of data to be visualized, we summarize the areas of interest within the market graph by displaying only the pre-computed clusters, and aggregated interand intra-cluster edges. Target graph structures and their associated attributes are encoded using several visually intuitive schemes, and a modified force-directed model is used to layout the graph with minimal visual clutter while retaining important spatial properties. We also provide a brief overview of an underlying parallel data-mining pipeline which enables us to apply this visualization tool to real-time stock market data.