Abstract

Just like mobile telephony had an astounding boom in Africa because the wired telecommunication infrastructures were very inadequate, we are expecting innovative data science methods based on new data sources to flourish on the continent due to the weak state of the statistical systems and the resulting wide gap between the demand and the offer of data for evidence-based decision-making. We are expecting Africa to be the one of the epicentres of the data revolution in the coming years. From the use of satellite imagery to monitor the degradation of natural capital stock and measure land use and land cover to the use of mobile tracking data to improve knowledge of migrations, cross-border formal and informal trade, and the development of digital id solutions to tackle the scandal of invisibility that makes people be born, live and die without their government being aware of their existence, the opportunities for applying innovative data science techniques to solve practical problems on the African continent are immense.
Our main argument is that the translation barrier has hindered the integration of knowledge in Africa and the lack of data and knowledge integration has stalled the awareness raising about the value of evidence-based decision-making, creating the well know vicious circle of low demand, low funding, and low quality of statistical production. AI has lowered the barrier for the knowledge translation but did not completely eliminate it. Knowledge translation has become a low hanging fruit that we can easily pick by adding data stewards in all domains of official statistics. Thanks to big data, machine learning and in particular transfer learning, a little effort in data curation can go a long way in terms of value added and this is an opportunity Africa should not miss.
ECA is therefore advocating for the African statistical community to tap into this huge potential by giving training in data science and initiating innovative data science projects through among others, the UN Regional Data Hub on big data and data science and the African Research Centre on Artificial Intelligence that was established in Brazzaville, Congo.