The image is a collage of the government's departmental organisation structure charts - labelled "Cameron's biggest nightmare" - which Cameron requested from the departments towards the end of 2010 (along with salaries and other information) - in an effort to increase government transparency.
The data.gov.uk team (and others including Talis, Epimorphics and TSO) have been working hard on modelling, converting and producing all of this data in machine-readable formats. To cut a long story short about why machine-readable formats are being used - it's because a) relationships between data can be expressed - letting machines infer and conclude things about the data using higher levels of semantic information and b) because standardised formats increase interoperability between machines and applications allowing the data to be used - and linked to - more easily and more often.
Since July, I have been designing and implementing an interactive governmental organogram as a web-application that's powered by linked data (RDF/Turtle) - which, once every department has published their organogram data, will be a generic application that's able to visualise every department's structure. Each department will also be able to embed the web-application onto their website and customise the look and feel as they wish.
Each department roughly publishes three types of data - their organisation structure, junior posts & grades and salary information within the structure. As linked data is being used, connections and relationships can be established across each of these three datasets and the web-application is able to automatically display the connecting information in real-time - allowing the public to ask questions such as "who is earning more than their boss?". Without the use of linked data, asking this question would have required searching for the people and their salaries across three datasets (which may be in spreadsheet format for example). The benefits are tenfold once you start linking together more datasets each containing hundreds of thousands of data entries (i.e. the COINS dataset).
Work is still on-going for the web-application as I've just got my hands on the real-data which is slightly different to what I've been working with throughout the prototyping stages, but everything's working as planned and looking great!
* Update: Version 1 of the visualisation can be found at http:/labs.data.gov.uk/gov-structure
2 comments:
of course one the more useful thing with government staff is noting predecessors?... but how
As long as the data has been captured so different periods of time can be established, this could be done using a time slider and some animation/visual notifications, so that when the slider is changed to a different date, it updates the information within the visualisation. I have plans to implement this at a later stage.
Post a Comment