https://data.blog.gov.uk/2015/06/05/small-step-for-local-giant-leap-for-open-data/

Small step for local... giant leap for open data

Almost a year ago, I blogged "Every Forward Step is Progress" just prior to the launch of the Local Authority (LA) Incentive Scheme. The scheme was allocated funding (£721,360) from the Cabinet Office’s Release of Data Fund (RoDF) to support the first ever bulk release of Demand-led Open Data, focusing on the priorities of the UK Open Data community.

The Open Data Request Roadmap developed by the Open Data User Group (ODUG) is the evidence base for Demand-led Open Data. This roadmap has been, and continues to be, used by the Open Data community with almost 15,000 viewers and over 1,000 data requests raised to request that new data be released as Open Data, with local government data a key focus for release in many cases.

Launching the LA Incentive Scheme in the midst of world cup fever (which admittedly was short-lived in the UK), and at the start of the holiday season, led to some concerns. We need not have worried. The local government sector signalled within the first six weeks of the scheme launch that they were intending to participate, and we had to put plans in place to deal with a possible over-subscription for the scheme.  A good challenge to have, and a reflection of the agility of local government to collaborate with ODUG under the stewardship of the Local Government Association (LGA).

The LA Incentive Scheme closed at the end of March 2015 with 209 datasets committed from 90 authorities. This data is across all three agreed themes: Public Conveniences, Planning and Licensed Premises. This is the first bulk release of Demand-led Open Data in the UK and the data also carries a quality mark - Open Data Institute (ODI) Open Data certificates at the RAW level. On behalf of ODUG, I would like to thank all the participants in this incentive scheme for championing the Open Data cause across their own organisations. We know from the feedback we have received - this has been a significant undertaking.

One of the key feedback points we received from the scheme participants was related to data quality. The use of standards, schemas and certificates of data quality from the ODI allowed organisations to release their data at an agreed level of quality, in a way which is repeatable. Since the launch of the scheme, the LGA have reused some of the same mechanisms and technology built for the LA Incentive Scheme to deliver the Open Data mandated under the Transparency Code. The intention is to reuse technology from the RoDF, where possible, on an ongoing basis. We want to encourage the release of more local data in a consistent way so that it can be compared and combined at regional or national level.

As part of the LA Incentive Scheme, free Open Data training (provided by a separate RoDF project) was supplied to participants, alongside a small grant towards the cost of releasing data to the scheme’s standards. These incentives were well received and have helped several Local Authorities to build open data capability. However, meeting the basic data quality standards was a struggle for most Local Authorities and the internal resources to be able to complete this work were not always available. Whilst this Scheme was always intended to be a pilot, it is clear that further investment is needed to build the capacity and capability to make local Open Data more usable. As with all RoDF projects, the delivery of the promised outcomes was required by 31/03/2015. For the LA incentive scheme, some LA's were unable to complete the delivery of the datasets within this project timetable and this resulted in £88,000 being returned to the RoDF.

ODUG was very keen to understand whether releasing data on a thematic basis, with some but not all of local government involved, would be enough to create applications or services based on the newly released data. We now have 50% of the scheme data loaded onto data.gov.uk and we already have positive examples to point to: for public conveniences this data has been used to help create “The Great British Toilet map” - not just useful for tourists but being used by business travellers when they visit new cities in the UK. I am in no doubt that many more use cases will emerge once all of the data is made available.

Ordnance Survey (OS) lifted their derived data licensing restrictions (a closely associated ODUG benefits case) on the use of location data in public conveniences datasets after the scheme had been launched. We have not been as lucky (so far) for either the planning data or the licensed premises data. Local government has data (with location coordinates) that is ready to be released once these licensing restrictions are lifted. Despite this, Surrey County Council have been able to utilise the planning data that has been released and launched. Ordnance Survey’s continued licensing restrictions will, sadly, inhibit its usefulness from a data entrepreneurs’ perspective. Derived data licensing restrictions continue to be one of the main barriers to beneficial Open Data use in the UK.

>Local Government teams will continue to load and release their data as it becomes available, via data.gov.uk, and we have sent a survey out to all participants to gather evidence on the lessons learned. As a project team, this pilot scheme has prepared us well to extend this approach across the local government sector. All that is needed is some relatively small investment funding in order to support this work in local authorities. Do let me know if you can help with this!

Jacqui Taylor 

Open Data User Group

June 2015

1 comment

  1. exstat

    I note various references in the blog entry to data quality. 

    Is this intended to reflect the intrinsic quality of the data, or simply the extent to which it adheres to Open Data principles?  A dataset may meet the most stringent Open Data requirements and still be unreliable.  Results from a statistical survey may be extremely unreliable at the lowest leval at which they are released, because the sample is so small or the response low and biased.  While an administrative dataset may be unreliable because of flaws in the data recording or because reporting is very incomplete.

    I doubt that this comes as news to anyone seriously involved in Open Data initiatives but it often does not come over that way in what people write.  Could I suggest that anyone referring to qauality should make clear exactly what they are referring to? (If they don't, perhaps the contribution should be deemed to be low quality"!...)

    That's not meant to belittle the great work that is clearly being done.

    Link to this comment