home edit page issue tracker

flat: flat multiword expression

The flat relation is one of three relations for multiword expressions (MWEs) in UD (the other two being fixed and compound). It is used for exocentric (headless) structures like names and dates. The head of the relation is always the first word of the expression, while all the other words in the expression are its dependents. In the Slovene treebank three different varieties of the flat relation are used.

The more general flat relation is used for individual tokens that make up dates and multi-word numerals.

edit

  

    
input
    
brat
    
info
  
Rojena sem bila 5. 10. 1957 \n Born I-aux-PAST was 5. 10. 1957
flat(5.-4,10.-5)
flat(5.-4,1957-6)
flat(5.-11,10.-12)
flat(5.-11,1957-13)
{
    "text": "Rojena sem bila 5. 10. 1957 \n Born I-aux-PAST was 5. 10. 1957 ",
    "entities": [
        [ "embedded-1-T1", "token", [ [ 0, 6 ] ] ],
        [ "embedded-1-T2", "token", [ [ 7, 10 ] ] ],
        [ "embedded-1-T3", "token", [ [ 11, 15 ] ] ],
        [ "embedded-1-T4", "token", [ [ 16, 18 ] ] ],
        [ "embedded-1-T5", "token", [ [ 19, 22 ] ] ],
        [ "embedded-1-T6", "token", [ [ 23, 27 ] ] ],
        [ "embedded-1-T8", "token", [ [ 30, 34 ] ] ],
        [ "embedded-1-T9", "token", [ [ 35, 45 ] ] ],
        [ "embedded-1-T10", "token", [ [ 46, 49 ] ] ],
        [ "embedded-1-T11", "token", [ [ 50, 52 ] ] ],
        [ "embedded-1-T12", "token", [ [ 53, 56 ] ] ],
        [ "embedded-1-T13", "token", [ [ 57, 61 ] ] ]
    ],
    "relations": [
        [ "embedded-1-R1", "flat", [ [ "arg1", "embedded-1-T4" ], [ "arg2", "embedded-1-T5" ] ] ],
        [ "embedded-1-R2", "flat", [ [ "arg1", "embedded-1-T4" ], [ "arg2", "embedded-1-T6" ] ] ],
        [ "embedded-1-R3", "flat", [ [ "arg1", "embedded-1-T11" ], [ "arg2", "embedded-1-T12" ] ] ],
        [ "embedded-1-R4", "flat", [ [ "arg1", "embedded-1-T11" ], [ "arg2", "embedded-1-T13" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "1" ]
}

The flat:name subtype is used for personal names.

edit

  

    
input
    
brat
    
info
  
Vladimir  Putin
flat:name(Vladimir,Putin)
{
    "text": "Vladimir Putin ",
    "entities": [
        [ "embedded-2-T1", "token", [ [ 0, 8 ] ] ],
        [ "embedded-2-T2", "token", [ [ 9, 14 ] ] ]
    ],
    "relations": [
        [ "embedded-2-R1", "flat:name", [ [ "arg1", "embedded-2-T1" ], [ "arg2", "embedded-2-T2" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "2" ]
}

The flat:foreign subtype is used for foreign language expressions.

edit

  

    
input
    
brat
    
info
  
V svoji uspešnici You Can Heal Your Life 
flat:foreign(You,Can)
flat:foreign(You,Heal)
flat:foreign(You,Your)
flat:foreign(You,Life)
{
    "text": "V svoji uspešnici You Can Heal Your Life ",
    "entities": [
        [ "embedded-3-T1", "token", [ [ 0, 1 ] ] ],
        [ "embedded-3-T2", "token", [ [ 2, 7 ] ] ],
        [ "embedded-3-T3", "token", [ [ 8, 17 ] ] ],
        [ "embedded-3-T4", "token", [ [ 18, 21 ] ] ],
        [ "embedded-3-T5", "token", [ [ 22, 25 ] ] ],
        [ "embedded-3-T6", "token", [ [ 26, 30 ] ] ],
        [ "embedded-3-T7", "token", [ [ 31, 35 ] ] ],
        [ "embedded-3-T8", "token", [ [ 36, 40 ] ] ]
    ],
    "relations": [
        [ "embedded-3-R1", "flat:foreign", [ [ "arg1", "embedded-3-T4" ], [ "arg2", "embedded-3-T5" ] ] ],
        [ "embedded-3-R2", "flat:foreign", [ [ "arg1", "embedded-3-T4" ], [ "arg2", "embedded-3-T6" ] ] ],
        [ "embedded-3-R3", "flat:foreign", [ [ "arg1", "embedded-3-T4" ], [ "arg2", "embedded-3-T7" ] ] ],
        [ "embedded-3-R4", "flat:foreign", [ [ "arg1", "embedded-3-T4" ], [ "arg2", "embedded-3-T8" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "3" ]
}

Finally, the ‘flat’ relation is also used for some verbatim transcriptions in speech, such as spelled-out numbers, acronyms, or URL addresses, which would typically be written with different symbols in writing.

edit

  

    
input
    
brat
    
info
  
dva cela pet odstotka \n two point five percent
flat(dva,cela)
flat(dva,pet)
nummod(odstotka,dva)
flat(two,point)
flat(two,five)
nummod(percent,two)
{
    "text": "dva cela pet odstotka \n two point five percent ",
    "entities": [
        [ "embedded-4-T1", "token", [ [ 0, 3 ] ] ],
        [ "embedded-4-T2", "token", [ [ 4, 8 ] ] ],
        [ "embedded-4-T3", "token", [ [ 9, 12 ] ] ],
        [ "embedded-4-T4", "token", [ [ 13, 21 ] ] ],
        [ "embedded-4-T6", "token", [ [ 24, 27 ] ] ],
        [ "embedded-4-T7", "token", [ [ 28, 33 ] ] ],
        [ "embedded-4-T8", "token", [ [ 34, 38 ] ] ],
        [ "embedded-4-T9", "token", [ [ 39, 46 ] ] ]
    ],
    "relations": [
        [ "embedded-4-R1", "flat", [ [ "arg1", "embedded-4-T1" ], [ "arg2", "embedded-4-T2" ] ] ],
        [ "embedded-4-R2", "flat", [ [ "arg1", "embedded-4-T1" ], [ "arg2", "embedded-4-T3" ] ] ],
        [ "embedded-4-R3", "nummod", [ [ "arg1", "embedded-4-T4" ], [ "arg2", "embedded-4-T1" ] ] ],
        [ "embedded-4-R4", "flat", [ [ "arg1", "embedded-4-T6" ], [ "arg2", "embedded-4-T7" ] ] ],
        [ "embedded-4-R5", "flat", [ [ "arg1", "embedded-4-T6" ], [ "arg2", "embedded-4-T8" ] ] ],
        [ "embedded-4-R6", "nummod", [ [ "arg1", "embedded-4-T9" ], [ "arg2", "embedded-4-T6" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "4" ]
}

---

flat in other languages: [bg] [bm] [cop] [cs] [de] [el] [en] [et] [eu] [fi] [fr] [ga] [gd] [hy] [it] [ka] [kk] [ky] [pcm] [pt] [ru] [sl] [swl] [tr] [u] [vi] [xcl] [yue] [zh]

© 2014–2024 Universal Dependencies contributors. Site powered by Annodoc and brat

.