Closed kajad closed 3 years ago
Can we first resolve what actualy needs to be done for (I assume) ssj550k 2.3?
I did a diff between the last commit on UD_Slovenian-SSJ and on running the scripts here, and there are some differences, but not many. Below the diff between sl_ssj-ud-dev.conllu here and the one on UD_Slovenian-SSJ:
1178c1178
< 39 , , PUNCT Z _ 35 punct _ Dep=0|Rel=Root
---
> 39 , , PUNCT Z _ 42 punct _ Dep=0|Rel=Root
1783c1783
< 21 nadzoru nadzor NOUN Ncmsl Case=Loc|Gender=Masc|Number=Sing 18 conj _ Dep=18\
|Rel=Coord
---
> 21 nadzoru nadzor NOUN Ncmsl Case=Loc|Gender=Masc|Number=Sing 18 nmod _ Dep=18\
|Rel=Coord
1785c1785
< 23 , , PUNCT Z _ 2 punct _ Dep=0|Rel=Root
---
> 23 , , PUNCT Z _ 18 punct _ Dep=0|Rel=Root
4367c4367
< 29 . . PUNCT Z _ 16 punct _ SpaceAfter=No|Dep=0|Rel=Root
---
> 29 . . PUNCT Z _ 16 punct _ Dep=0|Rel=Root
5019c5019
< 7 , , PUNCT Z _ 30 punct _ Dep=0|Rel=Root
---
> 7 , , PUNCT Z _ 15 punct _ Dep=0|Rel=Root
5946c5946
< 26 še še PART Q _ 11 discourse _ SpaceAfter=No|Dep=0|Rel=Root
---
> 26 še še PART Q _ 28 advmod _ SpaceAfter=No|Dep=0|Rel=Root
6212c6212
< 7 goloto golota NOUN Ncfsa Case=Acc|Gender=Fem|Number=Sing 6 obj _ SpaceAfter=No|\
Dep=6|Rel=Obj
---
> 7 goloto golota NOUN Ncfsa Case=Acc|Gender=Fem|Number=Sing 6 obj _ Dep=6|Rel=Obj
6625c6625
< 21 celo celo PART Q _ 28 advmod _ Dep=0|Rel=Root
---
> 21 celo celo PART Q _ 27 advmod _ Dep=0|Rel=Root
7427c7427
< 27 primerni primeren ADJ Agpmpn Case=Nom|Degree=Pos|Gender=Masc|Number=Plur 19 \
conj _ Dep=26|Rel=Atr
---
> 27 primerni primeren ADJ Agpmpn Case=Nom|Degree=Pos|Gender=Masc|Number=Plur 18 \
conj _ Dep=26|Rel=Atr
8065c8065
< 15 . . PUNCT Z _ 4 punct _ SpaceAfter=No|Dep=0|Rel=Root
---
> 15 . . PUNCT Z _ 4 punct _ Dep=0|Rel=Root
10239c10239
< 9 , , PUNCT Z _ 6 punct _ Dep=0|Rel=Root
---
> 9 , , PUNCT Z _ 15 punct _ Dep=0|Rel=Root
11024c11024
< 10 , , PUNCT Z _ 41 punct _ Dep=0|Rel=Root
---
> 10 , , PUNCT Z _ 17 punct _ Dep=0|Rel=Root
11537c11537
< 25 , , PUNCT Z _ 22 punct _ Dep=0|Rel=Root
---
> 25 , , PUNCT Z _ 27 punct _ Dep=0|Rel=Root
12366c12366
< 19 razrašča razraščati VERB Vmpr3s Aspect=Imp|Mood=Ind|Number=Sing|Person=3|Tense=Pres|Ve\
rbForm=Fin 7 parataxis _ SpaceAfter=No|Dep=0|Rel=Root
---
> 19 razrašča razraščati VERB Vmpr3s Aspect=Imp|Mood=Ind|Number=Sing|Person=3|Tense=Pres|Ve\
rbForm=Fin 21 advcl _ SpaceAfter=No|Dep=0|Rel=Root
12804c12804
< 21 reče reči VERB Vmer3s Aspect=Perf|Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin \
4 parataxis _ SpaceAfter=No|Dep=0|Rel=Root
---
> 21 reče reči VERB Vmer3s Aspect=Perf|Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin \
14 advcl _ SpaceAfter=No|Dep=0|Rel=Root
12991c12991
< 3 ! ! PUNCT Z _ 6 punct _ SpaceAfter=No|Dep=0|Rel=Root
---
> 3 ! ! PUNCT Z _ 2 punct _ SpaceAfter=No|Dep=0|Rel=Root
12999c12999
< 11 Meta Meta PROPN Npfsn Case=Nom|Gender=Fem|Number=Sing 6 parataxis _ Dep=10\
|Rel=Atr
---
> 11 Meta Meta PROPN Npfsn Case=Nom|Gender=Fem|Number=Sing 6 advcl _ Dep=10|Rel=Atr
13006c13006
< 18 opotekla opoteči VERB Vmep-sf Aspect=Perf|Gender=Fem|Number=Sing|VerbForm=Part 11 \
conj _ Dep=0|Rel=Root
---
> 18 opotekla opoteči VERB Vmep-sf Aspect=Perf|Gender=Fem|Number=Sing|VerbForm=Part 6 \
conj _ Dep=0|Rel=Root
14730c14730
< 14 ! ! PUNCT Z _ 20 punct _ Dep=0|Rel=Root
---
> 14 ! ! PUNCT Z _ 10 punct _ Dep=0|Rel=Root
14902c14902
< 8 - - PUNCT Z _ 17 punct _ Dep=0|Rel=Root
---
> 8 - - PUNCT Z _ 10 punct _ Dep=0|Rel=Root
14906c14906
< 12 športno športno ADV Rgp Degree=Pos 10 advmod _ SpaceAfter=No|Dep=10|Rel=AdvM
---
> 12 športno športno ADV Rgp Degree=Pos 10 conj _ SpaceAfter=No|Dep=10|Rel=AdvM
15248c15248
< 18 , , PUNCT Z _ 13 punct _ Dep=0|Rel=Root
---
> 18 , , PUNCT Z _ 23 punct _ Dep=0|Rel=Root
We agreed today that the official SSJ-UD release (2.8) will be mapped to the new ssj500k version (2.3) in the next few months by @TomazErjavec.
OK, this is mostly done, except for the errors in SRL and possilbly MWE (for that, cf. Redmine). @simonkrek, can you pls. start the next version of ssj500k (so, http://hdl.handle.net/11356/1210) and share with me the token.
Token shared.
Token received, tnx. Given that the UD annotations have been added to the corpus (even though it has not been published yet) I think we can close this issue.
ssj500k v2.2 (http://hdl.handle.net/11356/1210) currently includes SSJ-UD data v2.4. As notified in an email on 19/11/2019, the SSJ-UD annotations have been changed in release 2.5 (Nov 2019), so the ssj500k should be updated accordingly. @TomazErjavec, can we decide on the optimal timeline for this, so it does not get forgotten?
(Ofc, https://github.com/UniversalDependencies/UD_Slovenian-SSJ/issues/1 needs to be resolved beforehand.)