Rubbish in, rubbish out
The Audit Commission’s recent report “Nothing but the Trueth” is important given the current emphasis on access to public data. While Tim Berners-Lee and Nigel Shadbolt rightly highlight the enormous opportunities associated with making public sector data readily available, accuracy and consistency are still a challenge.
There is a lot going on at the moment around public data and a number of themes overlap.
Both Government and Opposition see the freeing of public data as part of broader effort to make government accountable. Public scrutiny, enabled by accessible data, will drive out costs and inefficiencies.
Linked to this, but not the same, is the argument that more accessible public data can create economic and social value. If you can get hold of public data easily and at low cost, you can re-purpose it, integrate it with it with other sources of data, and create products and services lumbering public sector bodies can only dream about.
The Ordnance Survey sits neatly at the cross-hairs of both themes. Managing large parts of core geographic referencing data, its terms of trade critically influence the ability of other providers to create new products which include an element of this core data. Because of importance of geography in performance data, its terms of trade can also provide a block to the scrutiny developments Government and Opposition are both looking for.
The IDeA and local government have been a “victim” of Ordnance Survey’s terms of trade which have provided a fundamental block to the exploitation of the National Land and Property Gazetteer now generally recognised as the best national source of address data. But although the IDeA is no fan of the Ordnance Survey it does recognise the core points being made by the Audit Commission. Reliable data is important and although the processes associated with collecting and distributing data can be transformed, processes to define and maintain data standards still have to be in place. As the experience of the National Land and Property Gazetteer proves, quality involves effort, and if we want this quality it will have to be paid for.
I am the Director of Services at the 
I agree on the relevance of the quality data argument - particularly relating to OS.
The best practice in gathering information which is likely to be stored and re-used is to capture it at source, from the operational data that creates it. This is exactly what happens in NLPG for the descriptive elements of an address, where the property name or number, street name and other details are created and stored as part of the local authority’s statutory street naming and numbering processes.
Gathering geographic information does not always offer this route of collection at source, so OS have to organise real world checking services to record natural features, and also building positions etc. This must be collected in a different way, but still requires quality processes and standards.
For easy-to-use information, providers need to combine or “mash up” operational and geographic data. This was obvious from the competition sponsored by the Power of Information Taskforce ( http://www.showusabetterway.co.uk/ ).
Let’s hope that the current restrictions on derived data that OS have been looking at for so long are loosened in the current rethink of OS data provision. There is revenue to be earned from mash up data that could still support the OS data quality we need.
DD