Data publishing and data destruction are both actions with significant, irreversible consequences.
Big data has made increasingly clear that knowledge, or data, is directly correlated with power. But what about the destruction of knowledge?
Knowledge destruction destroys trust, and irreversibly destroys the ability to make decent forecasting models. When a nation destroys its data, it will no longer be able to forecast national trends to the accuracy that other nations can – leaving it in a dangerous position to fall into the 'dark ages' as the rest of the world moves on. Similarly, if a company destroys its data, it risks falling behind the competition.
As a result, knowledge destruction is an intentional abuse of power with unfortunate long-term consequences. And while there are situations in which data deletion may be justified (e.g. to give a person under death threat from a criminal group a new identity), in general, destroying data is not a good idea.
/via +M Sinclair Stevens
Records deleted, burned, tossed in Dumpsters. A Maclean’s investigation on the crisis in government data
+Jeffrey J Davis If you read the article, what is being discussed is, in my opinion, actually a case of knowledge destruction (knowledge being defined in my previous comment), which is why I find it particularly concerning, and also different from other applications such as the EU 'right to forget', which affects personal data.
Allow me to elaborate: A targeted destruction of physical and digital historical dataset records means that we can no longer construct new knowledge based on the destroyed data, nor can we differentiate existing claims from bogus claims on the basis of the destroyed data. And given the political context of the current Canadian administration, it seems that the latter – the inability to differentiate between legitimate and bogus claims on development – is what is currently at stake.
+Sophie Wrobel it's impossible to grow insightful knowledge without a robust dataset. Your initial premise is legit.
+Jeffrey J Davis I'd agree that data and knowledge are two different things:
Data refers to facts, which are immutable pieces of information.
Knowledge refers to contextualized analysis derived from data.
That said, data and knowledge are very closely intertwined in the contemporary information ecosystem!
I'd argue that data and knowledge are two totally separate things, but as data storage marginally approaches zero cost, I would agree that destruction of data is a very short sighted strategy.