Open Context Upgrade Progress

Open Context Upgrade Progress
May 18, 2015 

Last summer, we launched a major programming effort to upgrade Open Context. The upgrade involves completely rewriting all of Open Context’s software so as to more efficiently scale Open Context and take advantage of technology standards that have emerged to prominence since our last major upgrade back in 2009-2011.

We’ve now deployed the new version of Open Context on a testing / development server generously provided by the German Archaeological Institute (DAI). The new testing / development version is hosted here:
http://opencontext.dainst.org/

The most important “back-end” aspects of the new version of Open Context are in place and functioning. We still need to add several user-interface features, site documentation, and some data visualization features, so this is still a work-in-progress. We are also still testing mapping interface features against different browsers, noting bugs, and addressing issues that cause confusion or broken functionality. If you notice any bugs or have suggestions for improvement, please make raise issues and comments at our source-code repository!

A major reason for our upgrade centers on the need to more fully implement Linked Open Data methods for interoperability. We’ll write more on these developments shortly, but for now, we’ll highlight how the new version of Open Context is starting to use linked data to situate content in more clearly defined temporal contexts. We are now testing use of PeriodO URIs for chronological metadata. PeriodO (a project led by Adam Rabinowitz, Ryan Shaw and Open Context’s Eric Kansa) provides a framework and data model for researches and other authorities (data repositories, museums, etc.) to publish assertions about time periods. Here’s an example of items in Open Context relating to the “Orientalizing” period in Italy, as modeled by PeriodO.

The new API (application program interface) for Open Context is now fully implemented (except for finishing touches on files that define certain namespaces). Already, the API is supporting original research, including this intriguing text-analysis and topic modeling project launched by Shawn Graham (described here and here and other places). In addition, Ben Marwick, Lincoln Mullen and Scott Chamberlain other colleagues with rOpenScience have started developing an rStats package built on top of Open Context‘s new API. This rStats package is particularly exciting, since it can open a whole new world of statistical analyses and data visualization.