Correlation vs. Causality

I know I shouldn’t mistake correlation from causality but in a very large data set how do I tell the difference? The reason we went to the big data in the first place was that we couldn’t establish causality in the first place!

What approaches can I use to test this?