Season's Greetings from the GIZ Data Lab Team

intense – but happy – most of the time. This is how we would describe our year 2021. If you had told us, that December 2021 kind of feels like December 2020, we wouldn’t have believed a word. But here we are, still in the midst of a pandemic and also the most sophisticated data analysis has not led to policy interventions which have made a difference. The virus just didn’t care. So what else to do than to continue with all our creativity, innovation power and hope to look for solutions which do not only fight against this pandemic but also lead to more impact, more evidence, more ownership in development cooperation. And that’s what we did:

This year, we have made a tremendous step in our Data Powered Positive Deviance (DPPD) initiative. We have not only published a scientific paper on our DPPD work on Covid-19, we have also advanced in our five other pilots, in which we look for positive outliers through big data. We have published our handbook, which has so much in store for all data for development enthusiasts, like checklists, vivid examples, diagrams etc. which help you apply this unique method – or parts of it. You can find all about the pilots on the initiative’s  Medium Blog.

We are very proud of our team’s contribution in the experiment with Colombia’s National Statistics Authority DANE on SDG 11 (sustainable cities and communities) tracking utilizing alternative data sets and Machine Learning to e.g. calculate open and green areas for public use.

Citizen generated data and crowd sourcing have big potential for context monitoring even in fragile countries. Based on the results in our experiment with GIZ’s RePLECA (Restoration of Peace, Livelihoods, and Economic Cycles in Anbar, Iraq) project and Premise, the usage of such data will continue…and guess what, on more use cases and with larger reach. We are happy to make a difference by gathering citizen data for project planning and steering in the world’s most challenging regions. 

In our experiment with the Sustainable Rice Platform we explored the usage of OCR (Optical Character Recognition) technologies to automate the digitization of paper-based surveys in English and Thai languages. We concluded that current technologies are promising but not good enough to justify large scale implementation. Though not in favor of innovative technologies, such evidence-based decisions are of great value for our projects like the SRP. Nevertheless, there is no doubt that the time of scaled OCR usage will come… 

Our most recent experiment with ETH Zürich and Sustainable Development Solutions Network (SDSN) on text mining of policy documents to track the implementation of the Nationally Determined Contributions revealed the power of Natural Language Processing for unstructured information mining. Stay tuned for more details on our blog to be released in January 2022! 
  
Following the Paris Peace Forum in 2020, we have worked hard on putting Gender Data on the development agenda and where thrilled to learn that our commissioning parties are seeking ways to uptake it. We don’t want to state that it was our contribution which lead to a mention of Gender Data in the Coalition agreement of Germany’s new Government, but hey – you never know! By the way, data is mentioned in many different parts of the document – what a good basis to continue our work. Same goes for the strong mandate the Data Lab has received from our Management Board: after our first three pilot years, we can count on becoming an integral and long-term unit within GIZ. With entering this new phase, we had to say good-bye to some our founding members and incredible colleagues, but with a good mix of new colleagues and experienced labsters, we will rock this year 2022. Join us with your ideas and maybe even with your wo*menpower!

Andreas, Catherine, Christian, Erik, Lena and Valentin