Middleware and open data

Just a quick one.  Last week the project multitouch group met to discuss moving forward with the development of the project website and multitouch interface.  The upshot of the conversation was that Russell has commited to building a ‘middleware’.  Essentially we have our project data stored in a database for use in a program called ‘NVivo’ which is used for analysing qualitative data.  Russell’s team are going to build some code to allow this database to be searched from outside the program and then call up the data tagged with different themes.  While the middleware itself won’t be pretty to look at (little more than a search box), we’ll then be able to layer on top of that some more attractive and intuitive interfaces suitable for web and touch devices.

This is very exciting and links to a meeting I was at last week in Southampton as part of the HESTIA2 project, which was considering issues of linked and open data.  One of the conditions of research council funding these days is that you make your data open to all.  Traditionally this has meant putting things like interview transcripts in an online repository.  This is fine, but it means that people accessing your data don’t get any sense of the analysis that you have done – tagging different pieces of text and images against different themes.  The middleware that Russell’s team are building will actually make both the data and, crucially, our analysis, both open and linkable, which is a very exciting prospect.