Questions & Answers
-
What is DNF?
-
If DNF is not owned by anyone, who will manage it?
- Isn't joined up information just something that is too complicated and too difficult?
- I support the concept of the DNF, but doesn't it need support from ministers and key organisations to make it a reality?
- Doesn't the DNF oblige adopters to use OS MasterMap?
- There are costs required to change the way we do things, how can I be reassured that such an investment will be worthwhile?
- Is the DNF a blueprint or strategy for an SDI (spatial data infrastructure)?
- This sounds like a re-launch of something proposed 4 years ago. Since little real progress has been made since then why will it be different this time?
- Why is DNF so biased towards a topographic base?
- Why aren't addresses a component of Base data?
- Are other countries thinking along the same lines?
- The DNF is an inclusive model but promotes the TOID as an identifier. How does it relate and include other identifiers such as NLPG?
- The DNF needs all organisations to adopt in order for it to be a success. Will there really be tangible savings if I invest in the concept and others do not?
- As a government user of Ordnance Survey information I create my own geography overlays. My licence allows me to pass these derived data layers on to other organisations. That works well.
- Why should I change, as I have interoperability already?
- How does the DNF relate to ODPM's "Maps on Tap" vision, which promotes interoperability and sharing of data to improve the knowledge economy?
- When will the DNF be complete?
- How can I keep informed as the thinking on DNF develops?
- What happens next?
What is DNF?
DNF is a set of principles, concepts and methods evolved through best practice with the intention of establishing better integrity, promoting greater reuse and easier sharing of application information. It is neither a standard nor a database.
If DNF is not owned by anyone, who will manage it?
A strength of DNF is that real users and practitioners, and through them the wider industry are evolving methods and approaches using a few common-sense guidelines and principles. Through small formal and informal groups, DNF will be managed at an appropriate level by practitioners for practitioners.
Isn't joined up information just something that is too complicated and too difficult?
No one pretends it will be easy, but the banking and electronic point of sale industries evolved simple models through their pressing need to handle massive volumes of transactions in a reliable and regular way. These transactions build up into more complex processes. By promoting DNF principles in the creation and maintenance of different datasets, we are taking a similar approach to securing early benefits - but we are being careful to avoid being too ambitious in one step.
I support the concept of the DNF, but doesn't it need support from ministers and key organisations to make it a reality?
While Ministerial endorsement would be welcome, it is people and organisations that will take the concept forward and make things happen. There is growing evidence of both take-up and emerging benefits from the DNF concept - examples are given in the paper.
Doesn't the DNF oblige adopters to use OS MasterMap?
No, although undoubtedly many organisations will find there are very real advantages in incorporating some components of OS MasterMap in their applications. OS MasterMap (and its large-scale mapping predecessors) have been created and are maintained by Ordnance Survey to provide the geographic framework infrastructure for Great Britain. This function, successively adapted and evolved over the last 200 years to anticipate customer needs, aims to avoid massive duplication of survey effort and to support the widest possible integration of information. Many organisations already have easy access to OS MasterMap through some form of SLA, and a lot of work has also gone into making the base information affordable while maintaining it to very high standards.
That said the greater take up of DNF principles does not mean that OS MasterMap features should necessarily be directly involved in all applications; linkages might, for example, be made through DNF-compliant identifiers rather than references to a mapping base (for example in gazetteers).
There are costs required to change the way we do things, how can I be reassured that such an investment will be worthwhile?
Most organisations renew their IT infrastructure every 5-7 years or so to ensure they have the necessary business capability. Inevitably there are major choices to be made at these times. In terms of geographic referencing, an organisation could continue to use traditional methods or adopt the DNF approach, which would make its information more transportable and easier to integrate and analyse with other datasets. Whether or not to invest in the new approach will be an individual decision for each organisation, but we anticipate that more and more will adopt DNF principles.
Is the DNF a blueprint or strategy for an SDI (spatial data infrastructure)?
DNF is neither of these things but is an enabler to both and more. It could be described as a blueprint to support better integration of business information with geographic information. Equally, it could be described as a component of an SDI. Adopting the DNF creates significant potential to improve our existing SDI framework, making it stronger.
This sounds like a re-launch of something proposed 4 years ago. Since little real progress has been made since then why will it be different this time?
The White Paper is, in effect, a progress report on the DNF initiative first proposed four years ago. It includes updated definitions and highlights next steps. Progress over the last four years has actually been significant, with both the launch of OS MasterMap and several key initiatives adopting DNF principles. Some confusion over the scope and nature of OS MasterMap and DNF crept in during that time and hopefully the paper now makes the difference clearer.
Why is DNF so biased towards a topographic base?
It leans towards a topographic base because the relationship between different kinds of geographic information is important, and is often best understood by linkages to a topographic base All things happen somewhere and ultimately the topographic base provides a set of atomic units that can be configured, in most cases, to reflect the user's geography - although this does not mean topographic mapping has to be visible in every subsequent application.
Why aren't addresses a component of base data?
The concept of reference information (which has been advanced in Europe in the last couple of years) can be viewed in two ways. 1) the physical features on the surface of the earth on the landscape and at sea and the height of that surface and 2) real world features that have been configured on that surface e.g. roads, property parcels, schools etc (either as simple point references or inclusive of their extents). Example 1 is very much a neutral view of the world, which does not reflect any particular user domain. Addresses are introduced in 2 which need referencing to the landscape via 1 (e.g. to a building which is in the base data). This does not indicate any significant "value" of any one dataset over another, but any relative values will normally be determined by users. The DNF model does introduce a logical set of links to connect physical and real world features to support data maintenance and the development of user models, which may or may not include addresses.
Are other countries thinking along the same lines?
Much of the early thinking around DNF was developed in conjunction with Ordnance Survey Ireland and Ordnance Survey of Northern Ireland, who are evolving similar approaches designed to suit their own political and commercial climate. The concept of unique geographic identifiers lies at the heart of all these developments. In Europe, several countries are contemplating database re-engineering and the most recent to announce plans with a similar structure are the Netherlands with TOP10NL. Elsewhere, work under the INSPIRE initiative over the last two years has been influenced by the DNF model and elements are reflected in the Architecture and Standards work. However, it is likely to be several years before such detailed information is adopted at a European wide level.
The DNF is an inclusive model but promotes the TOID as an identifier. How does it relate and include other identifiers such as NLPG?
The TOID family of identifiers represent the geographic location of a feature or object, whether this is an individual building or a collection of such features that collectively are described as, for example, "a school". Most businesses will already hold identifiers of clients, insured properties, assets etc as part of their business information. The BS7666 standard supports a gazetteer model (similar to that in an atlas) to create lists of business objects with a unique identifier known as a UPRN. Anyone can do this, but to date it has been used by local authorities to support their local land and property gazetteers. The DNF concept and TOIDs are entirely complementary to such a model by representing the geographic extents of an object. Where a user already has an existing identifier, these can normally be cross referenced directly on a 1:1 basis by referencing collections of individual features under a geographic [feature collection] identifier that can then be related to the user's business record identifier.
The DNF needs all organisations to adopt in order for it to be a success. Will there really be tangible savings if I invest in the concept and others do not?
Some of the benefits outlined in the white paper relate to internal business interests, and some to external. The benefits of wider take-up are related to reuse of information, data sharing and the evolution of "plug and play" information. Therefore the benefits are twofold - both internal and external - and the value of these benefits will naturally differ across organisations. To explore this further, contact one of the names in the paper.
As a government user of Ordnance Survey information I create my own geography overlays. My licence allows me to pass these derived data layers on to other organisations. That works well. Why should I change, as I have interoperability already?
The Digital National Framework is primarily a methodology to support information integration: licence issues will work independently. The main benefit of reusing source or derived datasets under a DNF model would be in promoting more reuse to wider audiences and in ensuring that all the various data components fit where they are meant to; thereby minimising expensive data cleansing programmes which are often encountered today.
How does the DNF relate to ODPM's "Maps on Tap" vision, which promotes interoperability and sharing of data to improve the knowledge economy?
Within Maps on Tap several examples of different datasets are loaded onto a server providing a quick way of visualising data relationships and also offering some GI analysis. However, the use of a common coordinate system (the National Grid) does not fully assure the internal integrity of the different information sources. The human eye can see where boundaries should fit, but in a true e-world, where automation and sound decision-making will become ever more prevalent, we need to encourage and adopt ways of georeferencing that promote internal consistency and information integrity. This will support the production of definitive and authoritative results from the integration and analysis of many different datasets. The DNF model would help bring this about if the datasets were to be built adopting the same referencing regime.
When will the DNF be complete?
DNF will continue to evolve to reflect and anticipate industry needs, so in a sense it will never be complete. New data and new applications can be expected to emerge and require future integration of some kind, supported by practicable models. However many emerging applications are likely to fall into a small number of generic models.
How can I keep informed as the thinking on DNF develops?
Over the next few months, this website will continue to be developed as a source of additional information on the DNF initiative. The current contact email addresses will be shown on the contacts page. You can also register interest with one of the contact names listed in the white paper and they will ensure that you are included in updates.
What happens next?
This depends on where you are at the moment. If DNF is new to you it is likely that you will like to find out more and perhaps speak to someone to explore the issues. If you are familiar and have developed some DNF-like ideas of your own, again you may wish to discuss this with other people who have done something similar e.g. the Expert Group. The paper contains a selection of names. please feel free to contact those most appropriate to your organisation's current position.
Other news
-
Linked Data Workshop - Agenda available - Registration open
Spaces are limited - first come first served ... read more
-
Getting Started with Linked Data
Workshop Announced ... read more
-
Coming up this week - DNF Expert Group Meeting
Plans and outline for EG#21 meeting, July 1st ... read more
-
BCS reports on recent geospatial event
"Time, Space and Meaning - a seminar on Geospatial semantics" ... read more
