Well, not exactly. I do have in mind the views of data that has to be pulled from my database. For example: JobOpenings, CompanyProfiles, CompanyOpenings.
Search can be on all columns to get a view of Job Openings regarding the column as a key search.
My question is this:
Should I start with ONE BIG table (about twenty columns), and work my way down to 3NF/4NF, or do I start with 3 tables from my View?
Right now I have columns:
Company | JobOpening | Province | City | Description | Skills | Branche | etc. Just all columns after each other...
If I think about Job requirements: Skills | HourPw | etc
Trouble with skills is that the choice is minimum 1 maximum 5, and each have 5 choices in it. So a separate table would be good, but against what key?
If I think about Job location: Province | City | JobOpening | Company
While I watched video's, read books and articles about normalisation, they all indicate one should already have separate tables. If that is true, I would at least need 3: JobOpenings, CompanyProfiles, CompanyOpenings.
But in the end, I thought, it looks like many to many relationship is almost a table for every column...
Any suggestions should be welcome