Data Transformed not Transferred
IG Mitchell, Tma Harvest · eCite Digital Repository (University of Tasmania) · 2013
In the processing of research metadata there are typically many iterationsin order to quality check,maintain and report data in a meaningful way. The transition from raw data to meaningful reportabledata is also usually via many routes and via many systems. Data manipulation in a downstream data storeis potentially dangeroussince the changes cannot easily be propagated back up. This issue becomesexponentially more difficult to control the more the data transitions from one system to another. Weargue that is it more efficient to collect data and store it once. Data processing is replaced with functionaldata transformation. The transformation can either be in-place or temporary. Data transformed is aculmination of the functions applied to it. At any functional transformation an in-place update can bemade and the changes will automaticallypropagate though the functional transformations to the output.This poster highlights the benefit of functionally transforming data using an all-electronic HERDC for publications.