WELCOME TO COLOURING LONDON

DATA ETHICS POLICY

INTRODUCTION

Colouring London is designed for everyone. It looks to create a safe, positive, constructive space for users of all genders, ages, cultural backgrounds and abilities to enjoy and benefit from, where the sharing of knowledge is encouraged and supported. 

 

User trust in Colouring London and adherence to high data ethics standards is essential to the platform's success and longevity. This includes: trust that the stated aims and objectives are actually as described; that user contributions will be treated with respect; that GDPR principles of lawfulness, fairness and transparency; purpose limitation; data minimisation; accuracy; storage limitation; integrity and confidentiality (security); and accountability will be met (we in fact actively discourage users from giving us personal data wherever possible); that   building occupiers' privacy is prioritised in data collection and that this is also carefully balanced against the urgent need to collect information on the physical form of all London's buildings to help aid emissions reduction and increase the sustainability of the city as a whole; that ongoing efforts will be made to make data as accurate and accessible as possible, and available to the widest possible audience; that the project is being responsibly and efficiently managed for the public good,

cost effectively run, and that it is designed to last. Breakdown of trust in any of these areas is considered to pose a significant risk to the sustainability of the project. 

Colouring London specifically requests permission from its public sector partners to display partner logos. These are considered helpful to users when assessing the trustworthiness of the project, and, in effect, act as trust marks. The Open Data Institute's Data Ethics Canvas is also used to identify and address potential ethical concerns and is discussed in detail below.

 

THE DATA ETHICS CANVAS

Data ethics are described by the Open Data Institute (ODI) as a 
" A branch of ethics that evaluates data practices with the potential to adversely impact on people and society-in data collection, sharing and use.
Ethical use of data brings about trust and helps allow data to work for everyone.
Colouring London is using the ODI Data Ethics Canvas to help identify and
manage ethical issues throughout the lifecycle of the project. 
 
As part of the process of development, existing, and new features within the platform are checked against the questions posed by the Ethics Canvas. First stage responses to core questions are given below. 

WHERE ARE DATA FROM? ARE PERSONAL OR SENSITIVE DATA INVOLVED?

Colouring London collects over fifty subcategories of data to support research into the sustainability of London's building stock. These relate to building location, use, type, age and history, size, materials and construction, sustainability (including energy rating and estimated lifespan), design/construction team, planning/ designation/demolition status, streetscape/green context, whether the building is community owned, and whether the user thinks it contributes to the city.
 
Our data are accessible at no cost and will also be available on the GLA's London Datastore. Most data we are collating, collecting and/or generating relate to physical characteristics of building, already able to be seen from street or  satellite images. Much of this information is also already held within government or commercial databases, but is restricted to the public and academia. Some datasets, such as building designation and energy rating are already publicly available.  Detailed information on London's buildings is also provided by the commercial sector with images of the interiors of homes, their price and sale history now commonly available on property websites.  
We do not collect personal data, other than optional, emails needed to enable users to reset site passwords, and actively discourage users on our 'sign up' page from contributing even their real names.  Though it is important  for us to understand what sectors and disciplines, groups our users are coming from to help us reach and ensure relevance to as wide  an audience as possible, we believe requests for such information for this purpose should, if introduced,  be optional only, with minimal information asked for. In the meantime we continue to explore how this issue can best be addressed by working and consulting across sectors, disciplines and community groups.  
 
Our sign-up agreement also tries to be as transparent as possible. It emphasises that when users make a contribution to Colouring London they are creating a permanent, public record of all data they add, remove, or change; that the database will record the username and ID of the editor, along with the time and date of the change and that all of this information will be made publicly available through the website and through bulk downloads of the edit history.
 
 Data are gathered in three ways.  Firstly by identifying and collating existing datasets held by central and regional government bodies and other organisations. Secondly by harnessing knowledge held within the community, whether this be from, for example, building specialists, building users, civic societies or schools. Thirdly, through large-scale computational data generation programmes run with results encouraged to be checked by the community as well.
 
Our job is to bring together and visualise fragmented information, make this more accessible and increase data accuracy through verification.  Some data may perhaps derive from observation of the building itself, some from historical texts or ready-to-go spatial statistics.  Our 'Community' section differs slightly in that it also asks users' whether they think buildings contribute to the city and/or local area. Users are informed on sign up that data cannot be accepted on the site where any restrictions to its open release may apply. 
The platform is also being designed as a collaborative data maintenance project as described in the ODI's handbook at https://collaborative-data.theodi.org/.

HOW ARE WE ADDRESSING ACCURACY, BIAS AND INCOMPLETENESS?

To help address issues of accuracy and bias a number of features are being included. Each subcategory will have an accessible edit history, a source box, a verification button, and a query button to enable problems that cannot be addressed within the editing system to be raised. Moderated dropdown options for sources range from 'expert assessment viewed at first hand', and link option allow references to sources and routes to further information. Easy to access edit histories also allow users to assess the accuracy of data. Phrasing of specific subcategory questions is also employed in certain cases to address uncertainty. Detailed data are collected, usually at building level.
Our landing page contains a clear statement that data are derived from multiple sources and that accuracy of the data must be determined by the user. 
As with the Wikipedia model the project is being planned as a low cost model, overseen by expert contributors. Our stewarding structure is currently being developed. 

WHO ARE WE SHARING DATA WITH AND UNDER WHAT CONDITIONS? 

 Colouring London has been designed  by University College London as a free knowledge exchange platform that collates, collects and generates open data on London's building stock, able to be used by everyone. We do not sell data and we will not share user’s personal data (e.g email address) with any other organisation.
 
The site is explicit in the user agreement, required to be accepted on our sign up page, on the way that contributed data can be used. Colouring London contributions are licensed under the Open Data Commons Open Database License (ODbL) by Colouring London contributors.

Users are free to copy, distribute, transmit and adapt our data, as long as they credit Colouring London and its contributors. If users alter or build on Colouring London data, they may distribute the result only under the same licence.

The sign up agreement emphasises that when you make a contribution to Colouring London, you are creating a permanent, public record of all data added, removed, or changed by you as noted above. It is also stated that Colouring London is unable to accept any data derived from copyright or restricted sources, other than as covered by fair use. Data sources are encouraged to be recorded wherever possible

 

Our platform code are also open and we encourage its use by other cities and towns. Code are available on our GitHub site 

https://github.com/colouring-london/colouring-london  under the following licensing terms: 
'Colouring London Copyright (C) 2018 Tom Russell and Colouring London contributors'. This program is free software: you can redistribute it and/or

modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

WHAT LEGISLATION & POLICY ARE SHAPING OUR DATA?

We closely monitor changes to UK policy in relation to the opening up of mapping data at building/property level particularly OS Mastermap footprints,  Valuation Office Agency property tax data (which contain comprehensive building attribute data relating to many of our categories) and OS historical map data. Our work also involves pressing government for the open release of these datasets - with footprints and property tax data now freely available in many other countries. Once OSMM footprint geometry has been released, all our data will be able to be mapped as on our site, rather than point data only.   
 
Our data are also specifically designed to support the UK's net-zero greenhouse gas targets for 2050, the United Nation's 17 Sustainable Development Goals , and its New Urban AgendaChanges to The UK National Planning Policy Framework and to energy legislation are also relevant to us

WHAT RIGHTS WILL THE SOURCE HAVE? 

Open Data Commons Open Database License (ODbL) Colouring London contributors are free to copy, distribute, transmit and adapt our data, as long as they credit Colouring London and its contributors. If users alter or build on Colouring London data, they may distribute the result only under the same licence

ARE WE SURE WE ARE NOT CONTRAVENING ETHICAL FRAMEWORKS

 Data privacy and data ethics are of the highest priority to the Colour Cities programme.  We work to the best of our ability to ensure we do not contravene any existing ethical frameworks. Each data category is widely consulted on and rigorously checked for potential issues prior to release. Issues on  contravention can be raised by users on our Githib site which is constantly monitored. We also actively discourage the contribution of personal data, avoid the collection of data within the building fabric,  incorporate controlled dropdown menus (with an internal moderation system for sources), and moderate all bulk uploads. We also  state that all data uploaded must be from an open source or generated by the user themselves. Our working partnerships also allow additional feedback routes and we are actively seeking to learn from, and collaborate with organisations advancing the data ethics agenda.  

WHY ARE WE COLLECTING DATA? ARE WE REPLACING A SERVICE? ARE WE MAKING THINGS BETTER AND FOR WHOM?

We are collating and collecting open data on London's building stock to provide  essential information for citizens, researchers, education providers and policy makers, to support sustainable development. We also want to assist those designing, constructing, caring for, managing and studying London's buildings to help solve urban problems, make the city more efficient and promote cross disciplinary/collaborative work. 
 
Our aim is to create a one-stop-shop for open data on London's stock. The release of these data is also designed to stimulate the production of innovative and efficient products  within the academic, non-profit and commercial sectors which promote and support the UK's transition to a low carbon economy, the UN's Sustainable Development Goals and the UN New Urban Agenda.
 
We also believe that it is healthy for platforms collating data on the building stock to be curated by universities, and others whose stance is impartial and whose brief is to undertake research for the public good.

ARE WE CLEAR IN THE WAY THE DATA WILL BE USED?

We are already aware of areas of research, such as energy, where demand for accurate building level attribute data is very high. We also know, from extensive consultation, that these data are also important to the construction and property industry, housing suppliers, planning bodies and the education sector. We are therefore excited about the many ways in which the data might be used.
 
We are currently developing a data showcase facility to allow users to upload how data from Colouring London can be best applied to solve urban problems, and to in doing so to inspire and inform. However we still need to address issues of how this section will be moderated.

WHO WILL BE POSITIVELY IMPACTED AND HOW? HOW CAN WE MAXIMISE AND MEASURE THIS?

As noted above, access to building attribute data for London at building level will provide free information for all those involved in the design, research, construction, management and maintenance and analysis of London's buildings, and its sustainable development. Our project is also designed to encourage use by diverse audiences. This element is central to its design.
 
Our task at the moment is to begin to release data on all our c50 subcategories. The second stage of the project will involve the introduction of features relating to the monitoring of platform (using analytical software ) and the Showcase section which will illustrate how the data are being used. 

WHO COULD BE NEGATIVELY AFFECTED BY THE PROJECT? & HOW IS tHIS BEING ADDRESSED?

The ODI's 10th item on its Data Ethics Canvas addresses the issue of negative project impact. Could the manner in which this data is collected, shared and used cause harm? Or be used to target, profile or prejudice people, unfairly restrict access?  Could people perceive it as harmful?
All spatial data projects that collect information able to be linked to specific addresses need to be very careful with regard to the type of data collected and how it is held and accessed. A number of checks have had to be put in place to ensure the safety and privacy of building occupants, and platform users. 
 
The main ways in which we are working to minimise negative impacts are by 
a) discouraging the submission of personal data (e.g. email addressed, real names), b) not collecting data on the insides of homes, c) avoiding  freetext wherever possible and using preset dropdowns, to prevent cyberbullying and security risks for occupants, d) only allowing users one vote per user on 'like me?', e) having no negative option for 'Like me?' again to prevent cyberbulling, f) having a sign up page that provides clear guidelines for responsible and ethical use of the site and g) only allowing the copy and paste tool to be used on one building at a time to deter macilous behaviour, and moderating all bulk uploads. 
 
Owing to concerns raised during consultation with regard to privacy and ownership data, Colouring London also only collects data on buildings where the freehold is held by the state or 3rd sector owners.  Colouring London still however needs to set up a data ethics advisory group.

HOW CAN PEOPLE APPEAL OR AFFECT CHANGES TO THE SERVICE?

Comments can also be made on existing discussion threads.  We are also looking at site improvement forms.

HOW ARE WE BUILDING IN CONSIDERATIONS OF PEOPLE AFFECTED BY OUR PROJECT?

The ODI's 13th Ethics Canvas item addresses considerations of people affected by our project. Are we creating potential risks or issues? How are limitations being communicated to those the data is about, and those impacted by its use? And how are we doing this?
 
Colouring London has been designed from the outset in consultation with representatives from diverse sectors. These are listed on our 'Who's involved page'. Our aim is to work with our partners and their networks to allow possible risks, concerns and improvements to be raised at the earliest possible stage, as well as through our public discussions forum and feedback forms. For these issues to then be addressed and features added, adjusted and/or removed as appropriate.

HOW WILL ONGOING ISSUES RELATING TO DATA ETHICS BE MONITORED & DISCUSSED?

We will be discussing data ethics issues on an ongoing basis with our project partners, our proposed ethics committee and within our discussion threads. Relevant actions and updates to the canvas may be viewed on this page. Formal meetings for Colouring London's ethics committee, will take place three times a year.

WHAT DATA ETHICS ACTIONS ARE OUR CURRENT PRIORITES?

Our current data ethics priorities (Last updated July 2020) are setting up the Colouring London Data Ethics Committee and working on dropdowns, verification, potential alert features and on issues relating  to security and free text.

COLOURING

LONDON