Derived data and derived data exemptions

This section is about what we mean by derived data.

It also provides guidance on the arrangements available to PSMA / OSMA members to obtain an exemption from the Member Licence for their derived dataset(s).

Once we have been notified and an exemption granted for the dataset(s), this means that a member can make the data available using the Open Government Licence.

Please note: this guidance only applies to our intellectual property rights (IPR) and interests and does not change or supersede any other third party IPR rights (for example, Royal Mail IPR).

It is your responsibility to ensure that you have the necessary permission from any other third party IPR owners.

There are four parts to this guidance about derived data and exemptions from the Member Licence: 

Tip: 'right click' a hyperlink to open it in a new browser tab.

Derived data - general

What is meant by derived data?

Derived data is a term used to describe new datasets or information that have been created using an existing dataset or information asset as its source.

Where, during the creation of new data, an existing dataset or information asset, including data attribution or textual content, has been copied, replicated, reproduced, and or generalised; that data could be considered to be derived if it encapsulates a significant proportion of the original, or source, data.

As the derived data will be protected by copyright and/or database rights, each user in the information chain would need to be appropriately licensed.

For example:

  • if you make a complete copy (i.e. a duplicate) of a dataset, that is not a derived dataset.
  • if you select certain features (e.g. points, lines or polygons) and or their attributions, that is a derived dataset.  
  • if you derive data from a dataset that is already compiled from multiple sources, the IP licensing arrangements from all those sources needs to continue.
What is OS derived data?

By this we mean digital data and or hard copy information that you have created using OS data products, see our Intellectual Property (IP) policy

Are there occasions when OS doesn't claim IP in our data?

Yes - when you have created data independently of any OS data. This is defined in the Member Licence as 'independently sourced data' and is also known as 'Licensees data' 

In practice this means that we do not claim intellectual property rights (IPR) in that data, although it could still be used in conjunction with (or overlaid) OS data. For example:

  • If you create data (including descriptive information or attribution) by means of an independent survey which can be used independently, we can't claim any ownership in your data. It is not, therefore, subject to any of our licensing terms.
  • If you then use OS data as the backdrop to your data, our licencing terms will apply to the use of that backdrop.
  • If you create data as described above, which is then overlaid onto OS licensed data for the purpose of verification, OS doesn't claim any ownership of that data - unless your features are repositioned as a result. 
  • If you reposition a feature, then our IPR will be present in it as you have now created derived data. In these circumstances the data will be subject to the licensing terms applicable to the OS source data used.
  • If additional independently-sourced descriptive information or attribution is attached to OS licensed data, we don't claim any ownership of that additional information. The additional attribution is not, therefore, subject to any OS licensing terms.
How can we release our derived data to third parties?

There are a number of options available to you.

As a member, do we have to make our derived data available?

The Member Licence lets you make your derived data available to others in a number of ways in support of your non-commercial public task.

To clarify the position and to reassure members, whilst you have these options, it is still entirely down to your own business decision-making and governance procedures to determine which is appropriate. It can also depend on your organisation’s preferences for releasing your derived data under the options we are providing.

Should you decide to release data that contains OS Licensed Data, in addition to the data sharing options available under the member Licence, we also have the exemption process that you could use to publish more of your OS derived data on Open Government Licence (OGL) terms. 

Where can I get additional advice before releasing my derived dataset?

We are happy to discuss your ideas with you if you are not sure whether the terms of your Member Licence or our open data arrangements for making your derived data available meet your requirements.

Before you make contact, please have a look at these points as they are likely to come into that conversation: 

You have established that your organisation is the owner of the data (or you have the owner's written permission to publish their data). 

If you do contact the PSMA Helpdesk or OSMA Helpdesk, it may be that we will need more detail from you. If so, this is the sort of information we will need:

  • which of our datasets you have used to create the derived data
  • the extent of the coverage, the data content and what you have done to make your data different from ours
  • a representative sample of the derived dataset, in a recognised industry standard format
  • a detailed description of the purpose(s) that the derived data has been produced for (and any other expected uses)
  • how you are intending to make the data available, including your approach to licensing and charging arrangements.

Derived data exemptions

What does a derived data exemption mean?

A derived data exemption is where you ask for your derived dataset, which you have created using our IPR, to be used beyond the terms of the Member Licence.

Where we grant an exemption, you can make your derived data available under Open Government Licence (OGL) terms. 

It does not mean that any other third-party IPR rights in your derived data have been superseded. You will still need the permission of those other IPR owners.

Please note the Member Licence:

applies to:

  • all the products listed in Appendix 2 of the Member Licence
  • any data created by you (and or another member) that includes any elements of those products that has not been granted an exemption

does not apply to:

Why do I have to inform OS about releasing our OS derived data as open data?

It is because we have a responsibility to manage and protect the IP in our data on behalf of the Crown; see our Intellectual Property (IP) policy. In doing so, it also:

  • enables us to monitor the number and type of openly released OS derived datasets, and
  • means that we will be aware of how successful the exemptions process is, which will also assist us in ensuring that the process is being used appropriately.
Is a derived data exemption necessary?

No! 

An exemption is only needed where you want others to be able to download your licensed derived data so that the recipient can use it locally. 

Note: It also means that, by making your derived data available on Open Government Licence (OGL) terms, recipients can use your derived data for any purpose they choose, including commercial exploitation.

You don’t need an exemption from us if:

  • you are not going beyond one of the licensed activities allowed in the Member Licence (for example, you are making your information accessible through viewing on your website)
  • you are just using OS OpenData products, or 
  • we don’t have an IPR interest in the data you are using (it's independently sourced). 

Before considering applying for an exemption for your derived data, please look at these sections about the arrangements available for:

Is there a decision tree available to help me decide whether to apply?

Yes - we have put together this derived data exemptions decision diagram (Mar 2017)

How do I apply for a derived data exemption?

We have two processes available:

How do I make a 'Presumption to Publish' notification?

We have an online Presumption to Publish notification form (you need to log in to the members area).

It is important to read all the qualifying criteria, as you have to make a declaration that your derived dataset is meeting it. 

You should establish that your organisation is the owner of the data (or you have the owner's written permission to publish their data).

What is the core qualification criteria for Presumption to Publish?

To qualify under these arrangements:

a)  you must have advised us through the Presumption to Publish notification form  (you need to log in to the members area), before releasing the derived data

and your derived data:

b)  must have been created by you as part of the delivery of your core business activity

c)  by being released, it does not constitute a Competing Activity or a Commercial Activity;

d)  whilst it may incorporate our IPR in the Licensed Data by copying Features and / or Feature Attribution in part or whole, it must not consist of a substantial quantity of Features  and / or Feature Attribution from the Licensed Data used to create it (for more information on how a ‘substantial quantity’ is determined).

e) must have taken into consideration any additional qualifying criteria stated in this guidance

f)  must be released independently of the source Licensed Data used to create it and / or with any other Licensed Data (for example, the derived data must not be released with Licensed Data as a contextual backdrop) 

g)  includes your attribution for each point, line or polygon feature

h)  must be made 'available on' one or more of these websites as appropriate under Open Government Licence (v3) terms, including the open data acknowledgement we use.  

Notes:

  • by 'available on' we mean that at least the metadata (which signposts where the data is available from) for your derived dataset is listed on the appropriate website(s). It does not mean that the dataset itself has to be on that site.
  • you can submit multiple derived datasets on the same notification form (just separate the dataset names with a semi-colon), but all of those derived datasets need to meet the qualifying criteria.
Additional qualifying criteria for Presumption to Publish

There are a number of derived datasets where we have already agreed some additional qualifying criteria.

As a result, we have now withdrawn the old 'streamlined' derived data notification processes.

These are the applicable additional qualifying critera:

Brownfield Sites

The sites contained in your register must comply with the National Planning Policy Framework criteria for previously developed land (Brownfield sites) and the specification schema  as published by DCLG to meet the requirements of The Housing and Planning Act (May 2016).

Please note: we are aware that the current Schema includes scope to use a Royal Mail Postcode. If you intend to include that within a dataset you provide under OGL terms, you will need to confirm with Royal Mail that that is ok with them.

Public Sector assets

In addition to the Presumption to Publish criteria, we have agreed that data which meets the criteria for being a public sector asset dataset can be published on the same terms and can be notified using the same PtP form. 

This means that, along with the use of the AddressBase UPRN , we also allow you to release the associated OS x,y co-ordinates – as they specifically relate to your public sector asset datasets.

For the avoidance of doubt, this includes allowing local government members in England to release their address data with associated co-ordinates and AddressBase UPRNs see here for more information, for any of their public asset derived datasets that are created to meet the requirements of the Local Government Transparency Code (England)

Please note: the above is also subject to you having received permission from Royal Mail in relation to the release of any data derived from PAF.

Public Rights of Way

The Public Rights of Way derived dataset must have been created by you in your capacity as the Highway Authority (and Surveying Authority) with a legal obligation to maintain the Definitive Map and Statement of Public Rights of Way (PRoW) for the area covered by it.

Conservation Areas

The Conservation Areas derived dataset must have been created by you in your capacity as the authority responsible for defining Conservation Areas for the area in question under the terms of the Planning (Listed Buildings and Conservation Areas) Act 1990.

Areas for Development

The Areas for Development derived dataset must have been created by you, the authority responsible for granting planning permission for the area in question and only depicts the extent of areas for which planning permission has been requested and/or granted for large-scale housing, employment or mixed-use developments.

This means the areas identified should relate to:

  • The creation of 10 or more residential units; and/or
  • A residential development on a site of 0.5 hectares or more; and/or
  • A non-residential development on a site of at least one hectare; and/or
  • A change of use of 1,000 square metres or more of gross floor space (not including housing).

 

Making a full derived data exemption application

If your derived dataset  does not meet the Presumption to Publish criteria, you can still apply for your derived data it to be available on Open Government Licence (OGL) terms.

We will work with you to ensure you are clear about what this entails.

So that we can help you, we will ask you to provide the following information using our online Derived data exemption proposal form (you need to log in to members area):

  • which of our datasets you have used to create the derived data
  • the extent of the coverage,
  • a description of the dataset you are seeking an exemption for, including how you have used our data and what you have done to make your data different,
  • a detailed description of the purpose(s) that the derived data has been produced for (and any other expected uses)
  • how you intend to make the data available, including your approach to licensing and charging arrangements.

You must also include a suitably representative sample of your derived dataset, which you can upload at the same time (up to 5Mb).

Note: if we find that the sample you provide is insufficient for us to make the assessment, we will ask you for a larger area. 

To remind you, it may not be necessary to seek an exemption as many data sharing circumstances are already catered for under the Member Licence, so please familiarise yourself with the Sharing data with others section of this guidance.

Under what circumstances wouldn't our derived dataset meet OS' criteria?

If your derived dataset doesn’t meet our open data criteria, it may be because it:

In these circumstances, we will inform you in writing that you should not release the dataset, or if you have already made it available, to withdraw and stop that release.

You can still share your derived data with others, but it will have be as Licensed Data as one of the licensed activities allowed under the terms of your Member Licence.

If your Member Licence or Presumption to Publish does not cover your proposed use of the derived dataset, then you may still be able to release the data under our Partner commercial terms. We will work with you to ensure that you are clear about what this entails.

Why are OS concerned about 'substantial' copying?

It is because we are in the information business and our income depends on the licensing of our intellectual property (IP).

Generally speaking, making a substantial copy means that you are copying a considerable amount, or a significant part, of something.

So, if data which substantially copies our Licenced Data and is then made available on open terms, there is a risk that such release will negatively impact our commercial business, or that of our partners, see our Intellectual Property (IP) policy.

It is not possible to give you a precise definition of when your derived data starts to substantially copy our Licensed Data, as any test needs to consider both qualitative and quantitative aspects; but in considering the meaning, the following factors may be relevant.

  • We consider the number of features (including real world objects or networks) and or Feature Attribution represented in the Licensed Data, by any line, polygon, symbol or text (or combination of the same) in the Licensed Data being used by you.
  • Whilst your Derived Data may incorporate our IPR by copying Features and/or Feature Attribution in part or whole, the Derived Data must not include a substantial number of Features or Feature Attribution from the Licensed Data used to create it; if the dataset does include a substantial number, then it will be a substantial copy.
  • The meaning of ‘substantial number’ can be determined by:
    i)  the total of the relevant Features and or Feature Attribution in the underlying Licensed Data, by reference to Great Britain as a whole or any of England, Wales or Scotland, and
    ii)  whether the Derived Data is capable of forming part of a series of connected datasets, whether created by you or other PSMA or OSMA members.
Can you illustrate what does / does not make a 'substantial' copy?

This presentation contains examples of derived datasets to indicate what does and does not form substantial copying.

We would like to include other examples from members who have been granted an exemption for their derived data. Please contact us with details if you have any example that we could use.

How does OS support the open release of public sector asset datasets

In addition to the Presumption to Publish criteria, we have agreed that data which meets the criteria for being a public sector asset dataset can be published on the same terms and can be notified using the same PtP form. 

This means that, along with the use of the AddressBase UPRN , we also allow you to release the associated OS x,y co-ordinates – as they specifically relate to your public sector asset datasets.

For the avoidance of doubt, this includes allowing local government members in England to release their address data with associated co-ordinates and AddressBase UPRNs, for any of their public asset derived datasets that are created to meet the requirements of the Local Government Transparency Code (England)

Please note: the above is also subject to you having received permission from Royal Mail in relation to the release of any data derived from PAF.

Can our point ‘planning applications’ dataset become an exempted dataset?

No - if you are wanting to include geocoding (either x,y or lat/long) or content from Royal Mail's PAF.

However, it can, provided that you leave those out elements in your derived data.

You can also include the appropriate AddressBase UPRNs in your derived data.

Please note: members can, and many already do, make their planning application data available to view on their websites and they can also share this data with other third parties under End User Licence terms. 

Can our polygon ‘planning applications’ dataset become an exempted dataset?

No.

Please note: members can, and many already do, make their planning application data available to view on their websites and they can also share this data with other third parties under End User Licence terms. 

How did OS come to this conclusion with planning applications datasets?

We came to this conclusion because, by granting them exemption status it is likely to mean that these dataset(s), either on their own or aggregated with others, they are potentially unfairly competing with or substituting addressing products supplied by OS and /or its Licensed Partners.

OS considers the risk still exists to the sustainability of our addressing product, despite this further analysis and is not something we are currently prepared to exempt. This is because the data contained in the planning dataset is premium address information and exempting that element of AddressBase Premium would significantly devalue that element of the product.

For example, those scenarios where members want to release their derived data under OGL terms in such a way that either they or a recipient could generate financial gain through making the data available for commercial exploitation by third parties. 

We are working with a number of members and undertaking further analysis of the derived datasets they have provided to us. The purpose is to learn more about what sort of addressing product could be created from this type of data, whether individually or collectively.

The challenge is that, while not necessarily being a whole dataset, over time recipients would be able to build up the history; where the total impact is then greater. Furthermore, OS need to consider the actual value of the total of all the planning data across the country, which even if it were 1%, it is still a ‘substantial’ copy.

It should be noted, it is only the extra geocoding (either x,y or lat/long) element where we have concerns, albeit Royal Mail still need to be consulted about the use of PAF.

 

Exemptions process governance

What is the process for assessing a full derived data exemption request?

Note: This is the process we use when assessing a ‘compliant’ application from a Licensee for a derived data exemption.

By ‘compliant’ we mean that the application includes all the necessary information to enable us to determine whether or not to approve the request. 

  • Where there is clear precedent, we will inform the Licensee in writing whether we have approved (either with or without conditions) or declined the request within 15 Working Days of receipt. If it is declined, that decision may be appealed by the applicant by writing to us (detailing the reasons) within 15 Working Days of receipt of the decision.
  • Where there is no clear precedent, or there is an appeal (see above), the request will be referred to our Pricing and Trading Group for a decision. Within 35 Working Days of receipt of the application, we will inform the applicant that we have either approved (either with or without conditions) the request, or declined the request.
  • If the application has been declined, the Licensee is entitled, within 20 Working Days of being informed to write to OS stating that they wish to appeal the decision, along with their reasoning. Within 5 Working Days of receipt of that notice, we will refer the Application to the Independent Advisory Group (IAG) for their consideration.
  • Within 20 Working Days of receiving a referral from OS or an appeal from the applicant, the IAG will make a written recommendation to OS’s Chief Executive and Director General as to whether the request should be approved (either with or without conditions) or declined, in each case containing details justifying such recommendation.
  • Within 20 Working Days of receiving a recommendation from IAG, OS’s Chief Executive and Director General shall, having given due consideration to such recommendation, decide in their absolute discretion whether to approve (either with or without conditions) or decline the request, and in each case inform the Licensee of the decision. 
    The Licensee acknowledges and agrees that the OS Chief Executive and Director General has sole discretion in relation to the decision, which shall not be subject to the Dispute Resolution Procedure (which is set out in the Member Licence).

Note: If, at any of the above stages, either OS or the IAG reasonably considers that it needs further information in relation to the initial request or an appeal, OS can request such information (whether on its own behalf or on behalf of the IAG) in writing from the Licensee and the relevant timescale will be extended by the period of time that it takes to receive that information from the Licensee.

Diagram of the full derived data assessment process (Mar 2017)

Diagram of the appeals process (Mar 2017).

What assessment criteria is used by OS for an exemption application?

Where you notify us of a Presumption to Publish dataset, you will have confirmed that it meets the Presumption to Publish criteria at the point you submit your request.

If you submit, or have submitted, a full exemption application then, at that time, you will have provided answers to the questions we asked.

When deciding whether an exemption can be granted or not, we consider the following questions.

Would allowing the release of your derived dataset:

  • have a negative impact on OS’s commercial business and ability to operate as a sustainable government company, or create a precedent that may have such an impact?
  • damage OS’s reputation or brand?
  • release a dataset that would compete with (or act as a substitute for) an existing or planned product or service offered either by OS or by one of our business partners?
  • infringe (or potentially infringe) the IP of a third party?
  • be for the public good, or in the public interest?
  • support the effective delivery of government or citizen services?
  • enable greater efficiency in Government – reducing costs or supporting cross-Government initiatives?
  • support government policy; for example the Transparency agenda and Public Data Principles?
  • fulfil statutory or legislative obligations?

The answers are documented and may be provided to the applicant as part of a formal response. 

Please note: If you have already made the derived dataset available and an exemption is refused, it may mean that you have to withdraw it.

Is there an 'appeals' process?

Yes.

In the event that there is a disagreement between you and OS as to whether your dataset meets the requirements of either the 'presumption to publish' or the full derived data exemption criteria, either party may notify the other in writing that it intends to refer the matter to the Independent Advisory Group (IAG) and the parties shall then discuss how to resolve the matter.

Where a resolution cannot be agreed within 20 Working Days of notification or, where the dataset has already been released and you refuse to cease release immediately, either you or OS may then refer the matter to the IAG for review and recommendation.

For more information on the IAG and their Terms of Reference

For a diagram of the appeals process (Mar 2017).

 

Additional guidance

Do we need to keep customer records for openly released derived data?

No - you don’t need to keep records on whom you have supplied data to when making your derived data available on Open Data terms.

You just need to have notified us about the derived data you are making available in that manner, so you may wish to keep the confirmation you received from us.

Where can we find details of other derived datasets that have been exempted?

As these exemptions processes only relate to our interests, there is no guarantee that the data has been made available by the member, especially if there are other third party IPR interests involved.

We would suggest that the best place to start is either Data.gov, or Scottish Spatial Data Infrastructure Metadata Catalogue to see if they have made it available.

Otherwise, you need to look at the member's own website and or contact them to find out.

There are some circumstances where it is not possible for the derived data to be released on Open Data terms, for more information see the 'Sharing your data with others' section of this guidance.

I am updating a previously approved exemption, what should I do?

You may be thinking of using the Open Government Licence (v3) terms.

If so, you don’t need to re-apply if the derived dataset that you’re making available on Open Data terms has already been notified to us:

  • under one of the previously available streamlined options (i.e. Public Rights of Way, School Catchment Areas, Conservation Areas, Areas for Development, Public Conveniences or 'Generic'), or 
  • we had previously approved your derived data exemption application on terms equivalent to OS OpenData terms, or 
  • if there has been no change to your drived dataset

However, if you are refreshing the dataset with updated/changed content and or attribution, then you should check that your derived data still meets the Presumption to Publish criteria and you should notify us.

Notes:

  • If you are using hyperlinks to the OS OpenData licence, this already links to the latest Open Government Licence.
  • As we update our policy, we review those applications that were originally turned down and may have already been in contact with you. Even if we haven't, you are still welcome to consider whether your dataset now fits the new arrangements and, if it does, make your notification to us.
  • If your derived dataset doesn't meet the Presumption to Publish criteria and you still wish to make it available on Open Data terms, you will need to submit a full derived data exemption application.
Can we aggregate our derived datasets?

You can aggregate derived datasets for your use, but you need check what the licensing arrangements are before doing so.

This is because you should not assume that all the component datasets that you are interested in using have been (or can be) released on terms equivalent to Open Government Licence terms.

You also need to be mindful that the more of your derived data that becomes (or is capable of being) aggregated either by you or other recipients, the greater the potential of that aggregated dataset becoming a ‘substantial copy’ of our Licenced Data.

This may mean that we will have revisit the assessment criteria used and may mean that subsequent applications may be refused.

What has happened to the other 'streamlined' exemption notification forms?

They have all been withdrawn from use as the specifications for the datasets they applied to are now covered by the Presumption to Publish criteria.