Feature: Sub and Super Scripts for similarity

**Description**
The similarity calculation is a bit weird, particularly in cases where one script is exactly and sub/super script of another. Why don't we just have that as its own category in the similarity search?