Contents

Annual Report

Looking back on 2020

Introduction

2020 was a year of transition for Open Knowledge Foundation after we renewed our mission in 2019, unveiling a new vision of a fair, free and open future.

By developing open data tools through our Frictionless Data or CKAN work, delivering world-class training and services for our partners as well as planning fantastic community events including csv,conf and Open Data Day, we pursued our mission to create a more open world – a world where all non-personal information is open, free for everyone to use, build on and share; and creators and innovators are fairly recognised and rewarded.

The Covid-19 pandemic brought much of the world to a halt this year and offered stark examples of why an open future has never been more important. Our remote working organisation continued operating as normal delivering projects and work for our partners throughout the year as well as launching our new Open Knowledge Justice Programme to ensure public impact algorithms cause no harm.

We also stepped up our campaigning efforts launching the Our Open Future campaign aimed at spreading the word about our vision, supporting open data efforts around the globe being pursued by our Network chapters, joining the Open Covid Pledge and commissioning an opinion poll on the importance of openness in government responses to the crisis.

In July we sadly bid farewell to our CEO Catherine Stihler who left to take up an amazing opportunity to lead the Creative Commons team. A new, interim leadership team assumed the CEO role collectively whilst the search for a new CEO takes place.

Highlights from our projects

Frictionless Data

Over the past year, we have collaborated directly with scientists in a variety of ways, from working with early career researchers to enact cultural change, to working with established researcher groups on large-scale projects. Our three main methods of collaboration are our pilots, Tool Fund, and Fellows programme.

Our pilots are hands-on, intensive collaborations between the technical Frictionless team and a researcher group to solve an existing data workflow issue. Last year, we finished a pilot with the Catalyst Cooperative to package open data about US utilities to make the data more accessible and understandable for climate researchers. We also finished the second round of a pilot with the Biological and Chemical Oceanography Data Management Office to incorporate Frictionless into their data processing pipeline to help their data managers more reproducibly process open oceanography data. We started another pilot with Dr. Philippe Rocca-Serra to create Frictionless templates for reporting scientific results. Finally, we are currently starting a new pilot with the data repository Dryad, to improve their data submission system for researchers. The pilots have gone beyond proof of concept and are showing the true value of the Frictionless approach to researchers.

The Fellows programme provides early career researchers with a grounding in the Frictionless approach and the skills needed to begin integrating it into their work and building in adoption in their respective fields. We are half-way through the second round of Frictionless fellows, bringing the total number of fellows to 11. You can read about their learnings on the blog - such as their blogpost about Open Access Week 2020 - and you can watch their workshop on how to use Frictionless tooling here.

The Tool Fund gives small grants to developers to build novel tooling for open science based on the open Frictionless software. This fund has provided support for developers to add to the growing suite of Frictionless libraries and tools, while also solving practical problems for researchers. Last year, we supported five grantees during the second round of the Tool Fund, bringing our total number of Tool Fund grantees to nine.

We are seeing month-on-month improvements both in terms of engagement with the project though community channels and in terms of adoption. Part of this was due to our improved Frictionless codebase, where we’ve been hard at work to release a new Frictionless Framework in Python.

2020 was also a banner year for virtual presentations. You can watch videos of Frictionless presentations from CarpentryCon at Home, FOSDEM, Research Reproducibility Conference, SciPy 2020, and All Things Open.

Training and data literacy

International Republican Institute

The Data for Democracy programme intends to increase the capacity of democracy activists to collect, process and interpret data. To do this, we built a data literacy curriculum customised to the work of up to 20 democracy activists. This work is going well and is due to be completed by the end of Q1 2021.

CoAct

CoAct is a European Union-funded consortium of research institutions, NGOs and global networks of open science and open data activism. The project’s goal is to identify and develop a general model for citizen social science that engages citizen bodies concerned with specific social issues in co-research. In the first year of this work, we participated in the forming and planning of the project, and advised participants on data management best practices.

Transparency International

We worked to deliver data literacy training, focused on the Global Corruption Barometer survey. As part of this work, we are also creating a written guide to the Global Corruption Barometer data that is planned to be released alongside future corruption surveys by Transparency International.

Open Spending

OpenSpending is one of the longest running projects both at Open Knowledge Foundation and within the open data ecosystem in its entirety. In July, we announced that Datopian would take on the stewardship of the project going forward with Rufus Pollock, the original creator of OpenSpending, in the lead.

Open Knowledge Justice Programme

In 2020, Open Knowledge Foundation made a commitment to apply our skills and network to the increasingly important topics of artificial intelligence (AI) and algorithms.

As a result, we launched the Open Knowledge Justice Programme in April 2020, with the objective of ensuring public impact algorithms cause no harm. Public impact algorithms are those that involve automated decision-making using AI and/or algorithms, are deployed by governments and corporations and have the potential to cause serious harm to individuals and communities. We are pursuing our mission through three strands of activities - training, advocacy and strategic litigation.

Training - We hosted a webinar for lawyers and legal campaigners called ‘AI and Algorithms for Lawyers in the Covid context and beyond’. Over 70 participants attended and we received great feedback. We also held a webinar for police officers, which covered an introduction to how AI and algorithms work and a detailed exploration of these systems in the context of law enforcement: Automatic Number Plate Recognition (ANPR), facial recognition and automated parole decisions, for example. We focused on the risks and harms of deploying AI and algorithms in policing, identifying the serious problems of bias, especially as directed towards people of colour.

Advocacy - The programme received good press via the Economist’s Babbage podcast and a full article on algorithms in Covid exams in The Guardian in July. In November, the director of the Justice Programme, Meg Foulkes, was interviewed by The Privacy Collective: ‘The future of tech should be fair, transparent, accountable’

Strategic litigation - In December, we began a brand new activity for OKF: strategic litigation. Supported by a generous grant from the Digital Freedom Fund, we are pursuing a stronger, interventionist approach in ensuring that public impact algorithms do no harm through legal action. This might mean we will go to court to make sure AI and algorithms are used fairly. But it will also include advocacy in the form of letters and negotiation. We think it’s wrong that governments can purchase software and not be transparent around key points of accountability such as its objectives, an assessment of the risk it will cause harm and its accuracy. As far as we are able, we’ll share our cases for others to use, re-use and modify for their own legal actions, wherever they are in the world.

Services

UNHCR

This year we continued our work with the United Nations Refugees Agency by developing their RIDL Platform further. RIDL is a centralised and secure data repository that ensures their team is able to use its valuable raw data to its full potential and make it available externally for operational partners, as well as preserving it for future use.

Highlights from our collaborations

Open Data Day 2020

Flyers from Open Data Day 2020 events across the world

On Saturday 7th March 2020, the tenth Open Data Day took place with people around the world organising over 300 events to celebrate, promote and spread the use of open data.

Thanks to the generous support of our funders – Datopian, the Foreign & Commonwealth Office, Hivos, the Latin American Open Data Initiative (ILDA), Mapbox, Open Contracting Partnership and Resource Watch – we were able to give out 67 mini-grants, the most ever.

Sadly several events had to be cancelled or delayed as the Covid-19 pandemic affected countries around the world but some of our grantees were able to swiftly adapt their plans in order to deliver engaging virtual Open Data Day celebrations.

The community registered a total of 307 events on the Open Data Day map with events taking place in every timezone and our team captured some of the great conversations across Asia/Oceania, Africa/Europe and the Americas by using Twitter Moments.

csv,conf,v5

csv,conf is a much-loved community conference bringing together diverse groups to discuss data topics, featuring stories about data sharing and data analysis from science, journalism, government, and open source. Over two days, attendees had the opportunity to hear about ongoing work, share skills, exchange ideas and kickstart collaborations. As in previous years, the Open Knowledge Foundation were members of the organising team.

csv,conf,v5 was planned to go ahead in May 2020 in person in Washington DC. Due to the Covid-19 situation this was not possible and the organising team made the decision to do the event online. It went ahead on May 13-14 2020. The event was generally accepted to be a huge success and one of the first in person conferences to move online. with over 1,000 registrations for the event. Most talks had well over 300 attendees. All the talks are available online. We have written more about our experience of organising an online conference with the hope that it will help others.

Global heatmap showing how people from across the world joining the csv,conf,v5 opening session

CKAN

2020 saw an important milestone in CKAN history with the release of the 2.9 version, which marked the culmination of a multi-year effort to modernise the source code and bring it up to speed with modern development practices. We led this effort from the beginning, collaborating on the development and planning and coordinating contributions from many different parties.

We remain involved in the day to day maintenance of the project and support the ever-growing community across different channels, and will continue to help steer the project in this new phase it is starting.

Campaigns

To push for a more open world, we pursued the following issues in 2020:

• We demanded answers to questions about the UK government’s plans to build an NHS COVID-19 datastore alongside dozens of leading digital rights organisations.

• We revealed that the British public would overwhelmingly value openness, transparency and respect for human rights in the government’s response to the COVID-19 pandemic.

• We appealed to governments around the world to embrace openness and transparency in their responses to the Covid-19 pandemic.

• We welcomed a call by the Centre for Data Ethics & Innovation to overhaul social media regulation in the UK.

New team members

  • James Hamilton, Development Manager
  • Chris Shaw, Senior Developer

Network highlights

The coronavirus pandemic turned the world upside down in 2020 but Open Knowledge chapters from around the world stepped up, lending their knowledge and expertise to efforts to tackle COVID-19.

Open Knowledge Brazil launched a successful legal challenge to get the Brazilian government to release more open data to help in the fight against COVID-19. The chapter also launched a COVID-19 data transparency site. Members of the Open Knowledge Switzerland team joined with other open data experts, public health officials and government workers to generate comprehensive and timely open data on the spread of COVID-19 in all the Swiss cantons that make up the country. Code for Germany – an Open Knowledge Germany project – was one of the partners which initiated the huge WirVsVirus hackathon aimed at solving COVID-19 issues with over 40,000 people taking part. Open Knowledge Belgium partnered for the Hack the Crisis Belgium initiative, bringing together the tech and start-up world to launch a public platform to create dedicated solutions.

To learn more about the activities of Network chapters, groups or affiliates in your country or region, visit okfn.org/network to connect with people where you live who also want to open information and use it to empower individuals and communities.

Multimedia

How to get involved

Join the Open Knowledge forum to connect with people across the world pursuing open knowledge and open data projects. To follow the work of the Open Knowledge Foundation, you can also connect with us via Twitter, Facebook, Github, Gitter, YouTube or subscribe to receive our newsletter every two months which features updates on our projects, network and events.