Closed Edouard-Legoupil closed 5 years ago
The initial scope of this function is basic anonymisation (i.e. removal of direct identifiers...) so that the dataset can be manipulated by approved researcher...
In order to release the dataset more widely, there might be links to be created with more advanced tool like SDCmicro - http://www.ihsn.org/anonymization - and it's shiny app:
library(sdcMicro)
sdcApp()
Need to integrate some functions from sdcMicro in order to perform k-anonymity and l-senstivity or Special Uniques Detection Algorithm (SUDA) on the dataset - the idea would be to produce a data anonymisation analysis report that will provide some benchmark in terms of disclosure risk vs information loss balance. Users will be able to apply further statistical disclosure control method till they get an acceptable anonymisation level
Work in progress here: https://github.com/unhcr/koboloadeR/blob/master/R/kobo_anonymisation_report.R
See some brainstorming on the expected scope for the function: https://github.com/unhcr/koboloadeR/blob/master/R/kobo_anonymise.R