The United Nations Population Fund (UNFPA), recently convened a group of staff interested in big data. UNFPA is traditionally one of the more data-driven entities of the UN, and in his statement (pdf) at the recent First Regular Session of the Executive Board, their Executive Director, Dr. Babatunde Osotimehin, laid out the establishment of UNFPA "as a center of excellence in data for development" and highlighted that data for development is a corporate priority, and specifically mentioned the promise of big data for their work:
Data is another priority area for UNFPA and a powerful tool for development. Data is essential to make evidence-based decisions and drive accountability. It can indicate whether our investments are making meaningful, measurable change in the lives of the most vulnerable. It can strengthen communities’ understanding of their own needs and assist people in holding decision makers accountable.
This year UNFPA will focus on strengthening its strategic direction, internal capacity and key partnerships for data for development. Currently, the world knows the least about people who have the least – in other words, the very people we need most to reach. The data revolution and the promise of big data have particular resonance for our work to promote equity and human rights and ensure that all women, adolescent girls and young people everywhere are able to achieve their full potential. […] Measuring progress towards the new Sustainable Development Goals will require a steady flow of high-quality, timely, authoritative, and accessible data. Dr. Babatunde Osotimehin, UNFPA Executive Director
Against this supportive backdrop, 30 employees from all regions of the world and from every HQ division gathered in Barcelona, Spain where the Bootcamp took place. Also invited were 20 big data experts, including Global Pulse (see full list of participant biographies).
The Big Data Bootcamp
The bootcamp was hosted by event and co-working space Betahaus in Barcelona with technical support from Data-Pop Alliance. With great success, Betahaus introduced a "design thinking" methodology that sought to combine information sessions with brainstorming and creative solution-generation.
Day 1: Discovering Big Data
On the first day, the agenda focused on how UNFPA use and plan to use data, introductions to the concept of Big Data, and case studies of how it has previously been used for development. Global Pulse's Chief of Research, Miguel Luengo-Oroz, presented "Applied Use of Big Data at the United Nations", and we also presented 3 case studies:
- Understanding the Post-2015 Development Agenda through Social Media Main Post-2015 project website / Uganda Post-2015 Dashboard
- Analyzing Social Media to Understand Attitudes towards Family Planning (with UNFPA Uganda) PPT Presentation / Live Dashboard
- Big Data for Climate Action Data for Climate Action Challenge / 2014 Climate Change Social Media Dashboard
Day 2: Designing solutions with Big Data
The second day was all about brainstorming about big data potentials for UNFPA around the world. Some groups based their discussions on a particular data source while others based their ideation on one of the thematic areas of UNFPA's work:
Data Sources
- Call Detail Records (mobile phone data)
- Citizen reporting / crowd-sourcing
- Financial Transaction Data
- Physical sensors
- Online information
- Web Search
Thematic Areas
- Adolescents and Youth
- Gender, Human Rights, and Culture
- Humanitarian
- Population and Development
- Sexual and Reproductive Health
I can't speak to the other groups, but the work done and the lessons learned through this group engagement was quite phenomenal in the groups I led on: Online Information and Web Search Data. After this day's sessions we all understood better what it takes to do a big data project based on online data, when it can be expected to work and when not, and how it can best be of value to UNFPA country offices specifically. All groups came up with a potential UNFPA big data project and it really got creativity flowing. UNDP's Benjamin Kumpf tweeted this as they were looking at Financial Transaction Data:
Building a prototype for mobile payments for girls' university attendance at #UNFPABigData w. @juka_ma @trevmon28 pic.twitter.com/C0XQG6iyMt
— Benjamin Kumpf (@bkumpf) February 4, 2015
Later, we performed our ideas which will surely feed into the UNFPA folklore:
The @UNFPA gets 'betahaused' #UNFPABigData pic.twitter.com/WcmToEovKc
— betahaus barcelona (@betahausbcn) February 4, 2015
Day 3: Defining The Future
To round off the bootcamp, the last day looked towards the future of big data analysis in UNFPA. The morning was spent looking at every main aspect of the big data lifecycle, and how UNFPA could go about answering questions like "Who should we partner with?", "How do we get access to data", "Do we hire or train internal data science capacities?" , "How do we best disseminate analytical results", and "How do we make sure results are used for policy change?" The last hours were also spent in groups. One group analyzed and visualized some bootcamp participant data, while another group talked about how some of the proposed Sustainable Development Goals could be monitored using Big Data.
The group I was in tried to hone in on UNFPA's vision for big data and feed into a UNFPA big data strategy. Our group settled on three areas to be described in the strategy:
- Partnerships: Partners with experience in (analyzing) big data and with interesting data are paramount to any success.
- Capacity Development: Without any current substantial in-house big data expertise, it is necessary to make sure to build on the initial learnings from the bootcamp, and make sure that a group of UNFPA employees at least know where big data can be used for what so that the above mentioned partnerships can be as strategically founded as possible, and leveraged fully.
- Do!: To get some early institutional learning, UNFPA can embark on pilot projects with current partners to try working through the main parts of a big data process.
Big Data: Learning, Exploring, Testing
The format of the bootcamp turned out to be a great way of learning about a new topic in general, and big data in particular. The 3-step setup of "Information" –> "Ideation" –> "Doing" corresponds very well with the project initiation cycle we use at Global Pulse.
As a result of the bootcamp, UNFPA is producing its 'Big Data Vision' with an outline underway, developing its internal culture and capacity and is taking forward plans for several big data prototype projects with partners. Finally, a data visualization app developed during the Bootcamp is now available for use by UNFPA in future meetings.
The UNFPA video below is a great recap of how the event unfolded: