UD_Portuguese-Bosque icon indicating copy to clipboard operation
UD_Portuguese-Bosque copied to clipboard

copula

Open livyreal opened this issue 8 years ago • 5 comments

Some dependency relations with the verb "estar" are annotated as cop"(copula), but we agree it should be aux.

Example from [1] "estão a dormir"

1   Em  em  ADP PRP_@ADVL>  _   3   case        
2   a   o   DET ART_F_S_@>N PronType=Art|Number=Sing|Gender=Fem 3   det     
3   localidade  localidade  NOUN    N_F_S_@P<   Number=Sing|Gender=Fem  19  nmod        
4   de  de  ADP PRP_@N< _   5   case        
5   Flamengos   Flamengo    PROPN   PROP_M_P_@P<    Number=Plur|Gender=Masc 3   nmod        
6   ,   ,   PUNCT   PU_@PU  _   3   punct       
7   onde    onde    ADV ADV_@SA>    _   14  advmod      
8   ainda   ainda   ADV ADV_@ADVL>  _   14  advmod      
9   estão  estar   AUX V_PR_3P_IND_@FS-N<PRED  Mood=Ind|Tense=Pres|Person=3|Number=Plur    14  cop     
10  mais_de mais_de ADV ADV_@>A _   11  advmod      
11  200 200 NUM NUM_F_P_@>N _   12  nummod      
12  pessoas pessoa  NOUN    N_F_P_@<SUBJ    Number=Plur|Gender=Fem  14  nsubj       
13  a   a   ADP PRP_@PRT-AUX<   _   14  nmod        
14  dormir  dormir  VERB    V_INF_@ICL-AUX< VerbForm=Inf    3   acl:relcl       
15  em  em  ADP PRP_@<ADVL  _   16  case        
16  tendas  tenda   NOUN    N_F_P_@P<   Number=Plur|Gender=Fem  14  nmod        
17  ,   ,   PUNCT   PU_@PU  _   14  punct       
18  Sampaio Sampaio PROPN   PROP_M_S_@SUBJ> Number=Sing|Number=Sing|Gender=Masc 19  nsubj       
19  apelou  apelar  VERB    V_PS_3S_IND_@FS-STA Mood=Ind|Tense=Pret|Person=3|Number=Sing    0   root        
20  a   a   ADP PRP_@<PIV   _   23  case        
21  o   o   DET ART_M_S_@>N PronType=Art|Number=Sing|Gender=Masc    23  det     
22  «  «  PUNCT   PU_@PU  _   23  punct       
23  bom_senso   bom_senso   NOUN    N_M_S_@P<   Number=Sing|Gender=Masc 19  nmod        
24  e   e   CONJ    KC_@CO  _   23  cc      
25  solidariedade   solidariedade   NOUN    N_F_S_@P<   Number=Sing|Gender=Fem  23  conj        
26  típica típico ADJ ADJ_F_S_@N< _   25  amod        
27  de  de  ADP PRP_@A< _   29  case        
28  o   o   DET ART_M_S_@>N PronType=Art|Number=Sing|Gender=Masc    29  det     
29  povo    povo    NOUN    N_M_S_@P<   Number=Sing|Gender=Masc 26  nmod        
30  açoreano   açoreano   ADJ ADJ_M_S_@N< _   29  amod        
31  »  »  PUNCT   PU_@PU  _   23  punct       
32  para    para    ADP PRP_@<ADVL  _   34  mark        
33  «  «  PUNCT   PU_@PU  _   34  punct       
34  compreender compreender VERB    V_INF_@ICL-P<   VerbForm=Inf    19  advcl       
35  que que SCONJ   KS_@SUB _   39  mark        
36  não    não    ADV ADV_@ADVL>  _   39  neg     
37  se  se  PRON    PERS_F_3P_ACC_@ACC>-PASS    PronType=Prs|Case=Acc|Person=3|Number=Plur|Gender=Fem   39  dobj        
38  podem   poder   AUX V_PR_3P_IND_@FS-<ACC    Mood=Ind|Tense=Pres|Person=3|Number=Plur    39  aux     
39  construir   construir   VERB    V_INF_@ICL-AUX< VerbForm=Inf    34  ccomp       
40  todas_as    todo_o  DET DET_F_P_@>N _   41  det     
41  casas   casa    NOUN    N_F_P_@<SUBJ    Number=Plur|Gender=Fem  39  nsubj       
42  ao_mesmo_tempo  ao_mesmo_tempo  PP  PP_@<ADVL   _   39  advmod      
43  »  »  PUNCT   PU_@PU  _   34  punct       
44  .   .   PUNCT   PU_@PU  _   19  punct       

As we can see bellow, in previous versions of Bosque (Claudia gave the example, I don't know which version is it, but it is not [4]), the dependency relation used was "aux".

As    [o] <artd> ART F P @>N #1->2 
relações    [relação]  N F P @SUBJ> #2->8 
de    [de]  PRP  @N< #3->2 
Hong=Kong    [Hong_Kong]  PROP M S @P< #4->3 
com    [com]  PRP  @N< #5->2 
a    [o] <artd> ART F S @>N #6->7 
China    [China]  PROP F S @P< #7->5 
estão    [estar] <aux> V PR 3P IND @FS-STA #8->0 
a    [a]  PRP  @PRT-AUX< #9->8 
condicionar    [condicionar] <mv> V INF  @ICL-AUX< #10->8 

We believe, it is a conversion problem.

For reference:

[1] Bosque 7.5 Universal dependencies, file bosque_CP.udep.conll.gz,
[2] Bosque 7.5 Universal dependencies, file bosque_CF.udep.conll.gz
[3] Bosque version 7.3, converted by Dan Zeman available in 
http://github.com/UniversalDependencies/UD_Portuguese
[4] Linguateca Version of Bosque CoNLL (7.3), http://www.linguateca.pt/floresta/CoNLL-X/

livyreal avatar Oct 06 '16 18:10 livyreal