Antonym - Processes for Antonym Generation

This page describes the processes of antonym generation in the annual Lexicon release.

I. directory

• base directory: ${ANTONYM_DIR}
• binary scripts: ./bin
• data: ./data
  • 0.Antonym
  • 1.Lexicon
  • 2.SuffixD
  • 3.PrefixD
  • 4.TtSet
  • 5.Medline
  • 6.WordNet

Some aPairs might come from different sources, such as

absolute|E0006593|relative|E0052609|adj|Y|B|O|quality|CC
absolute|E0006593|relative|E0052609|adj|Y|B|O|quality|SN

adjusted|E0007411|unadjusted|E0312692|adj|Y|B|O|quality|PD
adjusted|E0007411|unadjusted|E0312692|adj|Y|B|O|quality|SN

Thus, a heuristics rule is made to use the priority of source as follows: LEX > SD > PD > CC > SN to uniquely assign one source model.
Accordingly, the order of running process on all models must be followed.
Also, directories and input/output files requires manual modifications.

II. Processing commands

shell>cd ${ANTONYM_DIR}/bin
shell>GetAntonyms ${YEAR}