This was one of the best value for money events I’ve attended recently - £20 for a half day mini-conference with four outstanding speakers; the Information Commissioner, the Director, Information Policy and Services at TNA, Professor Charles Oppenheim and the CIO for a District Council in Somerset.
Over 100 delegates listened to four very different but interlinked talks, ranging from how to achieve a balance between protection of individual privacy (Data Protection) and transparency and accountability in the spending of public funds (Freedom of Information, to how public bodies can move from a protective gatekeeper stance over data towards a new default position of proactive release of information.
Charles Oppenheim gave a very entertaining and controversial look at what’s wrong with UK information law, and the CIO of Sedgemoor District Council had us all puzzling over the difference between linkable data and linked data, and how the latter means different data sets can be amalgamated to allow more sensible questions to be posed.
Many people’s favourite anecdote of the day was Charles’ revelation that a valid exemption to an FoI release in Norway was if the requester was ‘too inebriated’. He was puzzled as to how the FoI officers could tell...
The more senior questions he challenged us with about information law, included:
- Why is some information considered ‘sensitive’ under DP but not others? Eg – sexual preference or religious belief is ‘sensitive’ but DNA profiles, financial information or job performance information isn’t.
- While there is a right to sue for damages arising from DP breaches, why is it only possible to sue for distress caused if the media is involved? Eg – if a utility accuses you of not having paid a bill which you have, and threaten to bring in the bailiffs, that is likely to cause distress but you have no redress.
- How is the DP going to catch up with modern technology? Eg – with cloud computing, it is all too easy for personal data to end up ‘in the cloud’ which in practice is likely to mean it being sent to a USA based server – and despite the use of Safe Harbour companies, the Patriot Act is likely to over-ride those and thus potentially expose personal data to review.
Christopher Graham, the Information Commissioner, made a valid point about the call for his role to act as a ‘Privacy Czar’, in that this would simply push the point of decision (between right to privacy and right of access) elsewhere. Christopher also argued that FoI can be a driver of efficiency and that proactive disclosure promotes value for money, as it’s more cost effective to publish information in advance than it is to hunt for information later on in response to a FoI request.
Carol Tullo, TNA, talked about the ‘public data principles’ and the Open Government Licence, which has already been adopted by central government departments, over 180 local authorities, and many non-departmental government bodies. This is important if third party uses, volunteers and entrepreneurs are to feel comfortable to ‘mash up’ the data to produce innovative applications and products. Carol also reported that the Cabinet Office is considering the establishment of a Public Data Corporation, to collect, hold and manage public data centrally. It was clear that the main focus is now moving to proactive disclosure of as much information as possible, in a re-usable and linked data format.
Carol also briefly described the 5 stars of open linked data, from a 1 star of having data on a website, in any format, through to a 5 star of having data up which is already linked with other people’s data. This fed nicely into Paul Davidson, CIO of Segdemore District Council’s presentation on how to create meaningful linked local data. Paul made the distinction between ‘machine readable’ data, such as data in csv or rft format, open standards (such as rss feeds or GML), linkable data with identifiers, and true linked data, which has Uniform Resource Identifiers (URIs) embedded.
Paul gave several examples of the sorts of interesting and useful questions that can be answered if linked data is available:
Example 1:
Air pollution data (including locations)
+
Health incidence data (eg asthma)
+
demographic data
Could lead to very interesting questions about which local authority area has higher air pollution, do they show increased incidence of asthma, and which section of the populations is more affected?
Example 2:
Spend on road maintenance
+
Number of miles of road
+
Political control of councils over time
Could allow questions to be asked such as “which councils spend least on road maintenance per mile, while under control of a particular party?”
A lively Q&A panel debate session followed the speakers, which eventually ran out of time with several hands still up in the audience. Around half the audience then retired to the main UCL building for drinks and networking, and many were still there when I had to leave at 7.00pm.
- Nicola Franklin
No comments:
Post a Comment