Frank van Ham (IBM) – Data Visualization in Journalism

Van Ham was part of the IBM team who developed and created Many Eyes. 

Data visualization is the use of: computer-supported, interactive, visual representations of data (to “amplify cognition”). 

“Cognition amplificators”
Any representation technique that allows you to distort the truth is a medium. Data visualization as such is a medium and people have used it for propaganda, expression of self, art and communication of complex issues. 

Uses for data visualization in journalism are two-fold: explorative, transforming data into information. Communicative, transforming a story into a medium.

Data > information nuggets > story > representation of the story > consumption.

Transforming data into information, possible issues: messy data (government pdfs anyone?) data format and scale, which tools?, tiime constraints, structures (tables) vs unstructured formats (text).

Tag clouds: Analysis method maps directly to the presentation method.

Some insights for explorative data visualization
On data formats: from interview with data professional. Q: Do you need support for XML? a: What is XML?
Preferred: Tab delimited format. Copy and paste from Excel. The most important data interface in practice is that your tools allow easy copy and paste from and to Excel.

Transforming a story into an interactive medium. Print media already have a lot of experience in infographics. Possible issues: complexity of coding, exposing the data source, linking to other visualizations, to what extent do you want to take users by the hand when exploring, discussion and commentary.

On fostering discussion around visualization: allowing users to find and bookmark data points within the visualization is a great way to get them to interact. Combine structured view with free-range exploring. Integration with other technologies: sharing of visualizations is so much easier if you can just mail or Tweet a URL to a specific view.