| Option | Description | input | Output | Notes | Option
|
|---|
| 30 |
- get antonym candidates from prefixD
- derivation.getantcandfromprefixd.java
|
- ${PD_DIR}/input/derivation.data
- ${LEX_DIR}/input/inflvars.data
- ${ANT_DIR}/input/antcand.data.tag.${year}
- ${ANT_DIR}/input/domain.data
|
- ./output/Cand/antCandPrefixD.data
=> main output file, include tagged and not tagged.
- ./output/Cand/antCandPrefixD.data.tag
=> aPairs already tagged
- ./output/Cand/antCandPrefixD.data.tbd
=> aPairs to be tagged, should be 0 when is completed (|CANON_TBD|)
- ./output/candTagged/antCandPrefixD.data.tag.tagged
=> should be the same as ./output/Cand/antCandPrefixD.data.tag (beacuse all tagged aPairs are PD) at this step. It is used s the base (plus new tagged) to check tagged aPair form PD.
|
- if the first time:
- mkdir ./${year}/output/Cand
- mkdir ./${year}/output/candTagged
- use updated derivation.data and inflvars.data
- send antCandPrefixD.data.tbd.${YEAR}.${VERSION} to linguist to complete the tags
| 30
|
| 31 |
- Validate and fix tags of antonym candidates (PD)
- Antonym.ValidateTaggedCand.java
|
- ./output/candTagged/antCandPrefixD.data.tag.tagged
- ${ANT_DIR}/input/domain.data
|
- ./output/candTagged/antCandPrefixD.data.tag.fixed
|
- Append linguist's tag to ${PD_DIR}/output/candTagged/antCandPrefixD.data.tag.tagged
- Run this step until the tag and fixed file are the same.
- Fixed file is the auto-fixes on [TYPE_TBD] and [DOMAIN_TBD] to [NA] and [DOMAIN_NONE].
- shell> sort -u antCandPrefixD.data.tag.fixed > antCandPrefixD.data.tag.fixed.uSort
- Manually make a backup copy the sorted-fixed file to tagged file antCandPrefixD.data.tag.tagged.${YEAR}.${NO}
- Use the fixed file as input tag file to re-run this program until input and output is the same.
- Manually make a copy the tagged file as release file antCandPrefixD.data.tag.tagged.${YEAR}
| 31
|
| 32 |
- Update release antonyms tagged file form PD
- Antonym.UpdateAllTaggedFile
|
- ./output/candTagged/antCandPrefixD.data.tag.tagged.${YEAR}
- ${ANT_DIR}/input/antCand.data.tag.${YEAR}
- ${ANT_DIR}/input/domain.data
|
- ${ANT_DIR}/input/antCand.data.tag.updated
- ${ANT_DIR}/input/antCand.data.tag.updated.srcConflict
- ${ANT_DIR}/input/antCand.data.tag.updated.tarConflict
|
- This step auto-update all antonym candidate tag file.
- Add new tags to the tag file.
- If tags exist, update the source in the order of LEX, SD, PD, CC, SN.
- print out conflicts (duplicates) of src. This is OK! For example:
- activate|E0007090|deactivate|E0417566|verb|Y|UB|BN2|quality|SN
- activate|E0007090|deactivate|E0417566|verb|Y|UB|BN2|quality|PD
- print out conflicts of tags. This myst be fixed (manually).
- conflicted tags could be type, negation and domain.
- send antCand.data.tag.updated.tagConflict to linguist to fix.
- Then, manually fix one by one on both input files:
- antCandPrefixD.data.tag.tagged.${YEAR}
- antCand.data.tag.$[YEAR}
- cd ${ANT_DIR}/input
- Manully copy antCand.data.tag.updated to antCand.data.tag.updated.3.PD
- Manully copy/link antCand.data.tag.updated to antCand.data.tag.${YEAR}
- The output file is used to generate antonym and negation files for the release.
- Re-run steps 30-32 until it passes all steps.
=> update ${ANT_DIR}/input/antCand.data.tag.${YEAR} in step 32
=> update ${PREFIXD_DIR}/output/candTagged/antCandPrefixD.data.tag.tagged.${YEAR} in step 31
--- antCand.data.tag.$[YEAR} ---
- Total tag conflict no = 0
- Total source conflict no = 0
- Total duplicate tag = 0
--- antCandPrefixD.data.tag.tagged.${YEAR} ---
- tag conflict no = 0
| 32
|
