Good question.
From what I can read about it online (
link 1, and
link 2) you could use existing database such as Unihan. I haven't tried it myself though.
Another way to do t would be to use the CC dictionary (
here) and break down the entries in monograms, as suggested in link 2.
You could also look into web converters (from Simplified to Traditional and vice-versa) such as
this one. This wouldn't tell you whether the string is in Trad or Simpl but would convert it if it was put in the wrong field.
I'm sure any of this could be done with a good script, to split your mixed Chinese database into Trad/Simpl.