From b3d1ccbc40034c103eebc301cf1a42d6f777b92a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?bc=C2=B2?= Date: Thu, 7 Apr 2022 08:37:04 +0000 Subject: [PATCH] (gf2ud) label updates (#97) * (gf2ud) add ConsNP "," rule - use CCONJ instead of CONJ * (gf2ud) dobj -> obj (ud 2 -> ud 1) - clean up whitespace * (gf2ud) fix eng labels - comma exists in SSubjS, not SubjS - add local rule for comma in VocNP - auxpass (UD1) -> aux:pass (UD2) --- src/Lang.labels | 268 ++++++++++++++++++------------------- src/english/LangEng.labels | 10 +- 2 files changed, 134 insertions(+), 144 deletions(-) diff --git a/src/Lang.labels b/src/Lang.labels index 7b574b34..fb55ff6d 100644 --- a/src/Lang.labels +++ b/src/Lang.labels @@ -1,157 +1,157 @@ -PredVP nsubj head -PredVPS nsubj head -DirectComplVQ empty nsubj head ccomp -- Temp argument can only be empty in English -DirectComplVS empty nsubj head ccomp -- Temp argument can only be empty in English -SlashVP nsubj head -SlashVS nsubj head ccomp -FocusObjS nsubj head -- again might be wrong; more correct to call it dobj -QuestIAdv advmod head -- "where does John sleep" cf. AdvVP -QuestIComp head nsubj -- "where is John": John is clearly nsubj, but is where the head? -QuestQVP nsubj head -QuestSlash obj head -QuestVP nsubj head +PredVP nsubj head +PredVPS nsubj head +DirectComplVQ empty nsubj head ccomp -- Temp argument can only be empty in English +DirectComplVS empty nsubj head ccomp -- Temp argument can only be empty in English +SlashVP nsubj head +SlashVS nsubj head ccomp +FocusObjS nsubj head -- again might be wrong; more correct to call it obj +QuestIAdv advmod head -- "where does John sleep" cf. AdvVP +QuestIComp head nsubj -- "where is John": John is clearly nsubj, but is where the head? +QuestQVP nsubj head +QuestSlash obj head +QuestVP nsubj head IdetCN det head -PredSCVP csubj head +PredSCVP csubj head -ComplSlash head obj -ComplSlashIP head obj -Slash3V3 head obj -SlashV2VNP head obj xcomp -PastPartAgentAP head obj -AdvQVP head obj +ComplSlash head obj +ComplSlashIP head obj +Slash3V3 head obj +SlashV2VNP head obj xcomp +PastPartAgentAP head obj +AdvQVP head obj -Slash2V3 head iobj +Slash2V3 head iobj -ComplVS head ccomp -ComplVQ head ccomp -SlashV2S head ccomp -SlashV2Q head ccomp -CleftAdv head advcl -- not sure +ComplVS head ccomp +ComplVQ head ccomp +SlashV2S head ccomp +SlashV2Q head ccomp +CleftAdv head advcl -- not sure -ComplVA head xcomp -ComplVV head xcomp -SlashVV head acl -SlashV2A head xcomp -SlashV2V head xcomp +ComplVA head xcomp +ComplVV head xcomp +SlashVV head acl +SlashV2A head xcomp +SlashV2V head xcomp -AdvNP head nmod -- not sure -ExtAdvNP head nmod -- not sure -PassAgentVPSlash head nmod -- not sure -CleftNP head acl -- this cannot be nmod, not with type RS +AdvNP head nmod -- not sure +ExtAdvNP head nmod -- not sure +PassAgentVPSlash head nmod -- not sure +CleftNP head acl -- this cannot be nmod, not with type RS -SSubjS head mark advcl -RelS head advcl +SSubjS head mark advcl +RelS head advcl -AdNum advmod head -AdAP advmod head -AdvAP head advmod -ComparAdvAdj advmod amod head -ComparAdvAdjS advmod amod head -AdvIAdv advmod head -AdVVP advmod head -AdvVP head advmod -ExtAdvVP head advmod -AddAdvQVP head advmod -AdVVPSlash advmod head -AdvVPSlash head advmod -AdvSlash head advmod -ExistIPAdv head advmod -AdvS advmod head -ExtAdvS advmod head +AdNum advmod head +AdAP advmod head +AdvAP head advmod +ComparAdvAdj advmod amod head +ComparAdvAdjS advmod amod head +AdvIAdv advmod head +AdVVP advmod head +AdvVP head advmod +ExtAdvVP head advmod +AddAdvQVP head advmod +AdVVPSlash advmod head +AdvVPSlash head advmod +AdvSlash head advmod +ExistIPAdv head advmod +AdvS advmod head +ExtAdvS advmod head -SubjS mark head -RelVP mark head -RelSlash mark head +SubjS mark head +RelVP mark head +RelSlash mark head -ComplN2 head nmod -ComplN3 nmod head -AdvCN head nmod -PossNP head nmod -PartNP head nmod -ExistNPAdv head nmod +ComplN2 head nmod +ComplN3 nmod head +AdvCN head nmod +PossNP head nmod +PartNP head nmod +ExistNPAdv head nmod -OrdNumeralSuperl nummod head -- quite wrong; cannot find relevant details in the documentation; but most probably head lies outside the score of this function -DetQuant head nummod -- quite wrong; I think the head lies outside the scope of the function -CNNumNP head nummod +OrdNumeralSuperl nummod head -- quite wrong; cannot find relevant details in the documentation; but most probably head lies outside the score of this function +DetQuant head nummod -- quite wrong; I think the head lies outside the scope of the function +CNNumNP head nummod -ApposCN head appos -ApposNP head appos +ApposCN head appos +ApposNP head appos -RelNP head acl -RelCN head acl -SentCN head acl -SentAP head acl +RelNP head acl +RelCN head acl +SentCN head acl +SentAP head acl -DetQuantOrd head nummod amod -- quite wrong; again for same reason as DetQuant function -DetCN det head -IdetCN det head +DetQuantOrd head nummod amod -- quite wrong; again for same reason as DetQuant function +DetCN det head +IdetCN det head IdetQuant head nummod -CountNP det head -PredetNP det head +CountNP det head +PredetNP det head PPartNP head amod -- only in core RGL -AdjCN amod head -AdjDAP head amod -CAdvAP case amod head -ComparA amod head -ComplA2 amod head -CompoundAP head amod -AdAdV amod head -AdAdv amod head +AdjCN amod head +AdjDAP head amod +CAdvAP case amod head +ComparA amod head +ComplA2 amod head +CompoundAP head amod +AdAdV amod head +AdAdv amod head -CompoundN compound head +CompoundN compound head -IIDig head goeswith +IIDig head goeswith -PrepNP case head -PrepIP case head -SlashPrep head case -VPSlashPrep head case +PrepNP case head +PrepIP case head +SlashPrep head case +VPSlashPrep head case -UttImpPl empty head -UttImpPol empty head -UttImpSg empty head +UttImpPl empty head +UttImpPol empty head +UttImpSg empty head -MkVPS empty empty head +MkVPS empty empty head -UseCl empty empty head -UseQCl empty empty head -UseRCl empty empty head -UseSlash empty empty head -UseVC empty empty head +UseCl empty empty head +UseQCl empty empty head +UseRCl empty empty head +UseSlash empty empty head +UseVC empty empty head -BaseAP head conj -BaseAdV head conj -BaseAdv head conj -BaseCN head conj -BaseDAP head conj -BaseIAdv head conj -BaseNP head conj -BaseRS head conj -BaseS head conj -ConsAP head conj -ConsAdV head conj -ConsAdv head conj -ConsCN head conj -ConsDAP head conj -ConsIAdv head conj -ConsNP head conj -ConsRS head conj -ConsS head conj -ConjAP cc head -ConjAdV cc head -ConjAdv cc head -ConjCN cc head -ConjDet cc head -ConjIAdv cc head -ConjNP cc head -ConjRS cc head -ConjS cc head -PhrUtt cc head discourse +BaseAP head conj +BaseAdV head conj +BaseAdv head conj +BaseCN head conj +BaseDAP head conj +BaseIAdv head conj +BaseNP head conj +BaseRS head conj +BaseS head conj +ConsAP head conj +ConsAdV head conj +ConsAdv head conj +ConsCN head conj +ConsDAP head conj +ConsIAdv head conj +ConsNP head conj +ConsRS head conj +ConsS head conj +ConjAP cc head +ConjAdV cc head +ConjAdv cc head +ConjCN cc head +ConjDet cc head +ConjIAdv cc head +ConjNP cc head +ConjRS cc head +ConjS cc head +PhrUtt cc head discourse -PlusChunk head dep -TTAnt empty empty head +PlusChunk head dep +TTAnt empty empty head TExclMark head dep -- punctuation in the middle in these three TFullStop head dep @@ -159,13 +159,13 @@ TQuestMark head dep ComplV2 head obj -- shortcuts in App ComplV2V head obj xcomp -ComplV3 head iobj obj ---- could be dobj dobj +ComplV3 head iobj obj ---- could be obj obj PassAgentV2 head ncomp -- not sure RelV2 mark nsubj head QuestV2 obj nsubj head ModCN amod head -- in ResourceDemo -RSubjS mark advcl head +RSubjS mark advcl head SlashV2 nsubj head RAdvCN head case nmod SubjCl head mark advcl @@ -181,7 +181,7 @@ AdV ADV Adv ADV CAdv ADV Card NUM -Conj CONJ +Conj CCONJ Det DET Digits NUM IAdv ADV @@ -194,7 +194,7 @@ N2 NOUN N3 NOUN Numeral NUM Ord NUM -PConj CONJ +PConj CCONJ PN PROPN Predet DET Prep ADP @@ -214,7 +214,3 @@ VA VERB VQ VERB VS VERB VV VERB - - - - diff --git a/src/english/LangEng.labels b/src/english/LangEng.labels index e011bab4..625bd231 100644 --- a/src/english/LangEng.labels +++ b/src/english/LangEng.labels @@ -3,13 +3,13 @@ UseV,ComplVV,ComplVS,ComplVQ,ComplVA,SlashV2a,SlashV2V,SlashV2A,SlashV2S,Slash2V UseV,UseComp {"to"} PART mark head UseComp,CompAdv,CompAP,CompNP,CompCN,ProgrVP,QuestIComp {"is","are","am","was","were","been","be"} VERB cop head CompCN {"a","an"} DET det head -PassV2 {"is","are","am","was","were""been","be"} VERB auxpass head +PassV2 {"is","are","am","was","were""been","be"} VERB aux:pass head ComplVV {"to"} PART mark xcomp ComplVS {"that"} PART mark ccomp EmbedVP {"to"} PART mark head EmbedS {"that"} PART mark head SentCN {"that","to"} PART mark acl -ExtAdvS,SubjS {","} PUNCT punct head +ConsNP,ExtAdvS,SSubjS,VocNP {","} PUNCT punct head ExistNPAdv,ExistNP {"not","don't","doesn't","didn't","haven't","hasn't","hadn't","wouldn't","won't","isn't","aren't","wasn't","weren't"} PART neg head ExistNPAdv,ExistNP {"has","had","have","will","would","do","does","did"} AUX aux head ExistNPAdv,ExistNP {"is","are","am","was","were""been","be"} VERB cop head @@ -34,9 +34,3 @@ PossNP,PartNP {"of"} ADP case nmod @"am" PresSg1 @"is" PresSg3 @"are" PresPl - - - - - -