Big data on a small scale
Mirjam van Reisen favours big data built up from local inputs in developing countries and suitable for local use. The new Professor of Computing for Society at Leiden's Faculty of Science connects data science with development sociology. Inaugural lecture 10 March.
In her lecture Van Reisen will talk about the Ebola outbreak that started in Liberia in 2014. The western world only began to take notice of the epidemic after the first infected person arrived in the US. Then, all kinds of help was possible. The satellite company SES used high-tech methods to chart the epidemic in Liberia, Sierra Leone and Guinea in order to contain it; the key focus was on limiting the epidemic geographically, making satellite observations and focusing on data integration. New techniques were introduced, such as combining data from mobile telephones, diagnostic apps and social media, and using advanced mathematical models to predict the spread of the disease. In May 2015, after thousands of deaths, including 500 medical staff, the epidemic was finally under control.
And then all the high tech was gone as rapidly as it had appeared, including the data collected; international collaboration stopped unilaterally. In particular the healthcare system in Liberia was left in a desperate state: the country had no digital support and had to cope with a shortage of staff. Fortunately, this situation did not go unnoticed. It was the impetus for the international debate on International Cooperation in the Digital Era, the title of Van Reisen's inaugural lecture. Researchers, ministries in developing countries and companies started to work together. The Ebola crisis had been halted, but the question now was how a new outbreak of the same or another disease could be prevented. The real issue was to build a real-life system of e-health aiming at help and prevention, with data collection and analysis playing a major role. Data was to be collected and analysed locally and nationally, for the country's own use, but the data also had to be available for researchers elsewhere in the world. The most important tool in this project was the mobile phone because, unlike PCs and internet, mobiles are very widely used in the region.
Zambia has set up a project in which health workers spread over the country record health and medical data in the Zambia SmartCare Card central database. The system is rooted in the actual situation in the local area, which means that crucial steps can be taken quickly if the health situation in the country warrants it. India has introduced a similar system, the India SmartCard for Health, that focuses on citizens who are living below the poverty threshold. Just as in Zambia, patient and health information is collected, but the system is also used to support the poorest people; their card gives them access to particular health facilities, for example. It is also possible to channel financial assistance directly to the right person and healthcare can be linked to a health insurance.
Mobile payment common in Africa
This brings Van Reisen to the role of digitisation in financial practices in developing countries: ‘In Sub-Sahara Africa, it is common practice to pay using a mobile phone; a large part of the economy relies on person-to-person financial traffic within mobile neworks of family and friends.' This offers opportunities for a data-driven, personal approach to connecting health facilities, e-health and mobile payments. Van Reisen will also discuss the rapidly increasing volume of transfers to developing countries by family members or friends who have made a better life for themselves elsewhere. They transfer money by phone directly to the recipient. And what's surprising is the donors want to see that their money is well used for the recipients: they demand adequate healthcare for their loved ones.
Open Science Cloud
In spite of the progress made, there are still some hurdles to be crossed. Both the data on the health condition of an individual and financial traffic need more protection. In the case of personal medical data, this is for privacy reasons, while money transfers run the risk of cyber-criminality. Blockchain is one possible solution, but there's still some way to go.
Research within a set of data on the health situation in different places, regions and countries is only possible if there is an Open Science Cloud that is accessible for scientists worldwide. And that needs a strict protocol. Work is being done on this by a European committee headed by Leiden Professor of Bio-Information Technology Barend Mons. The committee has developed the FAIR system, a protocol for the Open Science Cloud. This states that the data included in the database has to be Findable, Accessible, Interoperable and Reusable, so that a well-functioning data internet can be set up. According to Van Reisen, it is crucial for developing countries to be involved in this. The countries will not lose their data, and they can also take part in joint international scientific projects.
In Van Reisen's opinion, all this stands or falls with international collaboration. Big data researchers are dependent on the locally recorded data and what comes out of their research has to serve the local situation, for example in the area of disease prevention. A start has been made, and rapid progress is possible.
Photo: Department of Foreign Affairs and Trade/Flickr