As I have read in my Database Course that normalization is efficient and but is it really efficient?
I think I am moving rather to a no-normalization database for my GSoC project GreenSMW and why is that?
As per our needs we have to answer queries for properties or subjects and very rarely for values. So, I am considering to store the data duplicated -once with sharding (horizontal partitioning) based on properties and secondly based on subjects. I am having other considerations also and will update this post if I come up with some other idea