Tooling Up for Digital Humanities

Let’s look at the first category, or what some people call exploratory data analysis. In this case, you’re simply throwing your data set at a wall to see what’s in it. If it’s numeric data, what are the highs and lows? The median, mean, and standard deviation (a measure of the statistical distribution of your values)? Are there spatial patterns in geographic data, or trends over time? A number in isolation doesn’t tell you anything; visualization can help you understand the context  for your data. For data analysis, free online tools can be the very useful; while the final interface of the visualization might not be the most beautiful, and the interactivity might be limited, free tools can provide great visual sandboxes for you to play around with your data. Among the most accessible tools for quantitative information are Tableau Public, a free PC application that lets you apply quite powerful graphical analysis to your data, and ManyEyes, a web application that IBM developed to help people share data sets and look for patterns.

Tableau has quite good mapping abilities for data that contain spatial information, but there are other map-focused tools that can help you make more sophisticated data maps. One of these, called Geocommons, enables you to upload data sets that reference specific place names or geographic coordinates (in the form of latitude and longitude); it can recognize spatial references in your data set and plot it on a map or match it with administrative boundaries like counties, census tracts, states etc. You can then output a static map or an embeddable interactive map.

There are also tools that support plotting your information over time, including the free timeline creation tools Dipity and MIT’s Simile Timeline, and the more sophisticated free desktop application called TimeFlow, originally developed for investigative journalists.

For plotting the non-geographic relationships between people, organizations and ideas, there are also a number of network analysis tools available. (See “Network Analysis Software and Services” postscript for more information.)

Finally, for building something more interactive and application-like, Adobe Flash and Flex remain powerful and popular tools. For more than a decade, they have served as the standard for building online infographics and multimedia, particularly on news and information sites (Geocommons, shown above, is one of several visualization tools built with Flash). While the lack of support for Flash on the iPhone and iPad have hurt its cross-platform compatibility, Flash is still a far more full featured tool for authoring interactive data visualizations than Javascript and HTML, although the latter have made great advances in recent years.  Moreover, Adobe now offers a tool called “Flash Packager for iPhone” that can convert Flash documents into iOS applications, and they have a similar tool in development for translating Flash to HTML.

The key to using Flash lies in mastering Actionscript, its internal language for creating interactions and processing data for visualization. Syntactically, it is very similar to Javascript, but because it is “compiled” by the development tool into a self-contained application, it can run much faster.

In addition to the Spatial History Project and the Republic of Letters, probably the best showcase of the type of work possible with Flash is the online work done by The New York Times’ graphics desk.