Posted by Barry Graubart, Vice President, Product Management
Identifying and understanding relationships across large data sets is difficult. Traditional row & column approaches can't scale or support non-linear relationships. For a number of years, specialized data visualization
tools have allowed analysts to find relevant patterns in data.
IBM's Visual Communications Lab has launched a social visualization platform called Many Eyes, led by data visualization gurus Martin Wattenberg and Fernanda Viegas. The site offers what they refer to as "social data analysis". Many Eyes users can upload time series or other data sets, which can then be used by members of the community to build visual analyses.
Many Eyes is similar in some ways to the recently launched Swivel, which bills itself as the "YouTube for Data". While Swivel is more focused on data mining, Many Eyes has put their efforts into more advanced visualization. Both offer simple charts, graphs and cross-tabulation, but Many Eyes includes geographical maps, histograms, bubble diagrams, network diagrams and tree maps. The maps also support data mining techniques such as co-occurrence of terms (terms that appear together in a document or data record).
As with Swivel, they've seeded the site with publicly available data from the Census Bureau, Bureau of Labor Statistics and Federal Reserve. As more data gets added, the relevance of the visual maps will surely increase.
Data visualization allows users to identify patterns in vast sets of data. As users become accustomed to navigating and mining data visually, these tools will allow them to use data in creative new ways.