Currently Korean tokenizer parses adjectives and verbs in a superficial level. Add in-depth parsing recovering the original forms.

Example:
Adjective parsing (바쁜 -> 바쁘다 + ㄴ)
Verb parsing (할지도 -> 하다 + 지 + 도)