Idea is to have a unique id for each job which is a hash of the sourceid and the source. Then have common features. Then have features unique to a certain source as {source}{feature} columns.
As we are starting with just UN jobs, I'm parking this issue for later.
Idea is to have a unique id for each job which is a hash of the sourceid and the source. Then have common features. Then have features unique to a certain source as {source}{feature} columns.
As we are starting with just UN jobs, I'm parking this issue for later.