Jer Thorp at IBM Cambridge

Image by johnnybelmont via Flickr

Yesterday I had the pleasure of hearing (and seeing) Jer Thorp, a kind of computer Renaissance man, who presented his work at the IBM Cambridge offices.  He is a scientist, computer programmer, and artist rolled into one.  His art, as he describes it, are the software that he creates to work with visualizations of data. 

He showed a range of examples of his work.  He has been exploring the NY Times archives using their API (Application Programming Interface...it allows one piece of software to speak to another).  He explained that where Open Source was once the center of computer interest, that interest has now shifted to Open Data.  

With hundreds of thousands of data items from the NY Times--the articles--he demonstrated how small, simple programs can illustrate different aspects of the information.  One that caught my eye, was a comparison of the word strength and connections within two articles.  This is like IBM's "Many Eyes", but that tool is still related to a single piece of text.  It is not comparative in the way that Thorp's was. 

Of particular interest to me were the "facets".  The NY Times has been tagging their articles since the 1800's (using real live human indexers).  Facets are "parent nodes" in NVivo speak that relate to categories like description, geography, person, etc.  Within the facets there are "child nodes".

He demonstrated another interesting piece of work tracking people who say "Good Morning" on Twitter.  In this piece he turned to Magna Carta, a Google tool, that gives the Latitude and Longitude for a text.  As the globe turned and colors popped out at us, reflecting the time zones coming on as risers tweeted "Good Morning" to the world, he made the point that "huge amounts of data are trailing behind us".

His blog blprnt@blprnt.com provides more information on his work. 


Oh--he also introduced me to an unusual off-beat artist Mark Lombardi, whose art consisted of graphs or node networks related to conspiracies he tracked.  These are truly like what one develops in Qualitative Data Analysis Software.  He, and his friends, considered them art.  Here's what Wikipedia said about it:

 Lombardi called his diagrams Narrative Structures ^[2] and they are structurally similar to sociograms – a type of graph drawing used in the field of social network analysis, and to a lesser degree to earlier artists like Hans Haacke – but in Lombardi's historical diagrams, each node or connection was drawn from news stories from reputable media organizations. The aesthetic impact is unique – the schematics are elaborate and delicate, yet precise and factual spiderwebs of illustrations depicting alleged networks of criminal conspiracies.

This talk gave me hope in regard to the ideas I've been playing around with regarding artful computing.

The talk was sponsored by the Center for Social Software at IBM.  Their new application, SAND, is looking a lot like QDAS.  I'm looking forward to learning more about it.