Abstract
Unprecedented technological advancements in information technology have ushered in a data science revolution, allowing scholars, companies and policy makers to conduct analyses at a scale, speed and granularity previously unimaginable 1 , 2 . However as Elshawi et al . 3 note, “in practice, big data science lives and dies by the data. It mainly rests on the availability of massive datasets, of that there can be no doubt.” From fields as varied as socio-economics 2 , 4 , 5 , 6 to ecology 7 and the ‘Internet of Things’ 8 , data scientists report the lack of big data itself is a major bottleneck in using big data tools. Increasingly, data scientists must first sort through heterogeneous, incongruent, and fragmented datasets before any analyses can be conducted 9 , 10 , 11 , 12 , 13 , 14 , 15 . Such problems with data availability are often exacerbated in emergency situations where real time analyses are often stymied by unevenly documented or unclean data 16 [pg. 358].