Using Commercial Big Data for Economic & Social Development

Impact Area: Public Policy
Institution: N-LAB - Nottingham University Business School
Leading Academic: Professor Andrew Smith, Assistant Professor James Goulding, Assistant Professor Gavin Smith

Project background

Many developing economies lack the data infrastructures that are taken for granted in countries like the UK. Rapid population change and a paucity of data capture mean that understanding the population is a considerable challenge for businesses and policymakers alike. Everything from retail locations to transport and disaster resilience planning rely on credible insights. This case study gives an example of how proprietary data can be used by policymakers in the developing world in the absence of data from the ‘traditional’ sources. The technology described has been used in projects undertaken in Africa and Asia with the involvement of the World Bank and numerous multinational commercial partners.

Solution

The N/LAB team at the Nottingham University Business School used machine learning driven data science and analytics to provide invaluable insights into a range of problems and applications in the commercial and policy realm to encourage economic and social development. An excellent example of this was the Projected Augmented Relief Model (PARM) of cellular activity in Tanzania. The PARM is a new display system that provides a physical, 3D approach to data visualisation. Using digital projection onto physical models the PARM provides an engaging and informative display, offering an intuitive frame of reference for placing objects, activities or events into their spatial context.

Benefits and impact

In Tanzania the PARM was generated using anonymised mobile phone records and the display was used to develop socio-economic maps, inform infrastructure and transport development among many other applications. A mobile PARM installation was used to examine how projection models might assist with urban planning and disaster resilience in East Africa. Using the mobile phone records, human mobility was visualised and the results layered over satellite imagery of the region. The PARM system was then used as a tool to aid discussions with Dar es Salaam City Council and World Bank members concerning the biannual flooding of the region.

The temporal nature of the data visualisation demonstrated how projection mapping technologies may be successfully combined with ‘Big Data’ to help invigorate decision-making processes around flooding, especially in areas requiring public consultation.

Experimental research involving direct comparisons between 2D maps and PARM showed that PARM enabled users to make more accurate judgements on questions such as the relative height of two points or the intervisibility between them (this refers to the ability to see in a direct line of sight from one position on the earth’s surface to another). Informed by these and other experiences the PARM technique is now being developed for use as an aid to spatial decision support and for consultation and education, using flood mapping as a primary case study.

Sources of research funding: This project was supported by funding received under two grants from the Engineering and Physical Sciences Research Council: From Human Data to Personal Experience (£4.1m) and Neo-demographics: Opening Developing World Markets by Using Personal Data and Collaboration (£612,744).