home edit page issue tracker

reparandum: overridden disfluency

The reparandum relation is used to annotate self-repairs in speech, i.e. instances where a speaker replaces a previously uttered content with a new one. The repaired unit can be syntactically complete or incomplete (i.e. unfinished words, phrases or clauses). In case of shared dependents between the reparandum and its repair, such as modifiers applicable to both, the dependent is attached to the repair, as illustrated in the second example below.

edit

  

    
input
    
brat
    
info
  
da so te eee ti stroški čim manjši \n that are these-Fem eee these-Masc expenses very small
mark(manjši,da)
cop(manjši,so)
nsubj(manjši,stroški)
det(stroški,ti)
reparandum(ti,te)
mark(small,that)
cop(small,are)
nsubj(small,expenses)
det(expenses,these-Masc)
reparandum(these-Masc,these-Fem)
{
    "text": "da so te eee ti stroški čim manjši \n that are these-Fem eee these-Masc expenses very small ",
    "entities": [
        [ "embedded-1-T1", "token", [ [ 0, 2 ] ] ],
        [ "embedded-1-T2", "token", [ [ 3, 5 ] ] ],
        [ "embedded-1-T3", "token", [ [ 6, 8 ] ] ],
        [ "embedded-1-T4", "token", [ [ 9, 12 ] ] ],
        [ "embedded-1-T5", "token", [ [ 13, 15 ] ] ],
        [ "embedded-1-T6", "token", [ [ 16, 23 ] ] ],
        [ "embedded-1-T7", "token", [ [ 24, 27 ] ] ],
        [ "embedded-1-T8", "token", [ [ 28, 34 ] ] ],
        [ "embedded-1-T10", "token", [ [ 37, 41 ] ] ],
        [ "embedded-1-T11", "token", [ [ 42, 45 ] ] ],
        [ "embedded-1-T12", "token", [ [ 46, 55 ] ] ],
        [ "embedded-1-T13", "token", [ [ 56, 59 ] ] ],
        [ "embedded-1-T14", "token", [ [ 60, 70 ] ] ],
        [ "embedded-1-T15", "token", [ [ 71, 79 ] ] ],
        [ "embedded-1-T16", "token", [ [ 80, 84 ] ] ],
        [ "embedded-1-T17", "token", [ [ 85, 90 ] ] ]
    ],
    "relations": [
        [ "embedded-1-R1", "mark", [ [ "arg1", "embedded-1-T8" ], [ "arg2", "embedded-1-T1" ] ] ],
        [ "embedded-1-R2", "cop", [ [ "arg1", "embedded-1-T8" ], [ "arg2", "embedded-1-T2" ] ] ],
        [ "embedded-1-R3", "nsubj", [ [ "arg1", "embedded-1-T8" ], [ "arg2", "embedded-1-T6" ] ] ],
        [ "embedded-1-R4", "det", [ [ "arg1", "embedded-1-T6" ], [ "arg2", "embedded-1-T5" ] ] ],
        [ "embedded-1-R5", "reparandum", [ [ "arg1", "embedded-1-T5" ], [ "arg2", "embedded-1-T3" ] ] ],
        [ "embedded-1-R6", "mark", [ [ "arg1", "embedded-1-T17" ], [ "arg2", "embedded-1-T10" ] ] ],
        [ "embedded-1-R7", "cop", [ [ "arg1", "embedded-1-T17" ], [ "arg2", "embedded-1-T11" ] ] ],
        [ "embedded-1-R8", "nsubj", [ [ "arg1", "embedded-1-T17" ], [ "arg2", "embedded-1-T15" ] ] ],
        [ "embedded-1-R9", "det", [ [ "arg1", "embedded-1-T15" ], [ "arg2", "embedded-1-T14" ] ] ],
        [ "embedded-1-R10", "reparandum", [ [ "arg1", "embedded-1-T14" ], [ "arg2", "embedded-1-T12" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "1" ]
}

edit

  

    
input
    
brat
    
info
  
kako orož- orožje pa to \n some weap- weapon or something
det(orožje,kako)
reparandum(orožje,orož-)
conj(orožje,to)
cc(to,pa)
det(weapon,some)
reparandum(weapon,weap-)
conj(weapon,something)
cc(something,or)
{
    "text": "kako orož- orožje pa to \n some weap- weapon or something ",
    "entities": [
        [ "embedded-2-T1", "token", [ [ 0, 4 ] ] ],
        [ "embedded-2-T2", "token", [ [ 5, 10 ] ] ],
        [ "embedded-2-T3", "token", [ [ 11, 17 ] ] ],
        [ "embedded-2-T4", "token", [ [ 18, 20 ] ] ],
        [ "embedded-2-T5", "token", [ [ 21, 23 ] ] ],
        [ "embedded-2-T7", "token", [ [ 26, 30 ] ] ],
        [ "embedded-2-T8", "token", [ [ 31, 36 ] ] ],
        [ "embedded-2-T9", "token", [ [ 37, 43 ] ] ],
        [ "embedded-2-T10", "token", [ [ 44, 46 ] ] ],
        [ "embedded-2-T11", "token", [ [ 47, 56 ] ] ]
    ],
    "relations": [
        [ "embedded-2-R1", "det", [ [ "arg1", "embedded-2-T3" ], [ "arg2", "embedded-2-T1" ] ] ],
        [ "embedded-2-R2", "reparandum", [ [ "arg1", "embedded-2-T3" ], [ "arg2", "embedded-2-T2" ] ] ],
        [ "embedded-2-R3", "conj", [ [ "arg1", "embedded-2-T3" ], [ "arg2", "embedded-2-T5" ] ] ],
        [ "embedded-2-R4", "cc", [ [ "arg1", "embedded-2-T5" ], [ "arg2", "embedded-2-T4" ] ] ],
        [ "embedded-2-R5", "det", [ [ "arg1", "embedded-2-T9" ], [ "arg2", "embedded-2-T7" ] ] ],
        [ "embedded-2-R6", "reparandum", [ [ "arg1", "embedded-2-T9" ], [ "arg2", "embedded-2-T8" ] ] ],
        [ "embedded-2-R7", "conj", [ [ "arg1", "embedded-2-T9" ], [ "arg2", "embedded-2-T11" ] ] ],
        [ "embedded-2-R8", "cc", [ [ "arg1", "embedded-2-T11" ], [ "arg2", "embedded-2-T10" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "2" ]
}

In case of sequence of self-repairs, for example when a speaker keeps restarting their intended verbalization, all reparandums attach to the same head, i.e. the head of the final repair.

edit

  

    
input
    
brat
    
info
  
iskalnik je po krajih po imenih po po po ukrepih ne \n you can browse by places by names and by by by measures
reparandum(po-9,po-8)
reparandum(po-9,po-7)
case(ukrepih,po-9)
reparandum(by-23,by-21)
reparandum(by-23,by-22)
case(measures,by-23)
{
    "text": "iskalnik je po krajih po imenih po po po ukrepih ne \n you can browse by places by names and by by by measures ",
    "entities": [
        [ "embedded-3-T1", "token", [ [ 0, 8 ] ] ],
        [ "embedded-3-T2", "token", [ [ 9, 11 ] ] ],
        [ "embedded-3-T3", "token", [ [ 12, 14 ] ] ],
        [ "embedded-3-T4", "token", [ [ 15, 21 ] ] ],
        [ "embedded-3-T5", "token", [ [ 22, 24 ] ] ],
        [ "embedded-3-T6", "token", [ [ 25, 31 ] ] ],
        [ "embedded-3-T7", "token", [ [ 32, 34 ] ] ],
        [ "embedded-3-T8", "token", [ [ 35, 37 ] ] ],
        [ "embedded-3-T9", "token", [ [ 38, 40 ] ] ],
        [ "embedded-3-T10", "token", [ [ 41, 48 ] ] ],
        [ "embedded-3-T11", "token", [ [ 49, 51 ] ] ],
        [ "embedded-3-T13", "token", [ [ 54, 57 ] ] ],
        [ "embedded-3-T14", "token", [ [ 58, 61 ] ] ],
        [ "embedded-3-T15", "token", [ [ 62, 68 ] ] ],
        [ "embedded-3-T16", "token", [ [ 69, 71 ] ] ],
        [ "embedded-3-T17", "token", [ [ 72, 78 ] ] ],
        [ "embedded-3-T18", "token", [ [ 79, 81 ] ] ],
        [ "embedded-3-T19", "token", [ [ 82, 87 ] ] ],
        [ "embedded-3-T20", "token", [ [ 88, 91 ] ] ],
        [ "embedded-3-T21", "token", [ [ 92, 94 ] ] ],
        [ "embedded-3-T22", "token", [ [ 95, 97 ] ] ],
        [ "embedded-3-T23", "token", [ [ 98, 100 ] ] ],
        [ "embedded-3-T24", "token", [ [ 101, 109 ] ] ]
    ],
    "relations": [
        [ "embedded-3-R1", "reparandum", [ [ "arg1", "embedded-3-T9" ], [ "arg2", "embedded-3-T8" ] ] ],
        [ "embedded-3-R2", "reparandum", [ [ "arg1", "embedded-3-T9" ], [ "arg2", "embedded-3-T7" ] ] ],
        [ "embedded-3-R3", "case", [ [ "arg1", "embedded-3-T10" ], [ "arg2", "embedded-3-T9" ] ] ],
        [ "embedded-3-R4", "reparandum", [ [ "arg1", "embedded-3-T23" ], [ "arg2", "embedded-3-T21" ] ] ],
        [ "embedded-3-R5", "reparandum", [ [ "arg1", "embedded-3-T23" ], [ "arg2", "embedded-3-T22" ] ] ],
        [ "embedded-3-R6", "case", [ [ "arg1", "embedded-3-T24" ], [ "arg2", "embedded-3-T23" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "3" ]
}

edit

  

    
input
    
brat
    
info
  
kako k- kako hodiš \n how h- how do-you-walk
advmod(hodiš,kako-3)
reparandum(kako-3,kako-1)
reparandum(kako-3,k-)
advmod(do-you-walk,how-8)
reparandum(how-8,how-6)
reparandum(how-8,h-)
{
    "text": "kako k- kako hodiš \n how h- how do-you-walk ",
    "entities": [
        [ "embedded-4-T1", "token", [ [ 0, 4 ] ] ],
        [ "embedded-4-T2", "token", [ [ 5, 7 ] ] ],
        [ "embedded-4-T3", "token", [ [ 8, 12 ] ] ],
        [ "embedded-4-T4", "token", [ [ 13, 18 ] ] ],
        [ "embedded-4-T6", "token", [ [ 21, 24 ] ] ],
        [ "embedded-4-T7", "token", [ [ 25, 27 ] ] ],
        [ "embedded-4-T8", "token", [ [ 28, 31 ] ] ],
        [ "embedded-4-T9", "token", [ [ 32, 43 ] ] ]
    ],
    "relations": [
        [ "embedded-4-R1", "advmod", [ [ "arg1", "embedded-4-T4" ], [ "arg2", "embedded-4-T3" ] ] ],
        [ "embedded-4-R2", "reparandum", [ [ "arg1", "embedded-4-T3" ], [ "arg2", "embedded-4-T1" ] ] ],
        [ "embedded-4-R3", "reparandum", [ [ "arg1", "embedded-4-T3" ], [ "arg2", "embedded-4-T2" ] ] ],
        [ "embedded-4-R4", "advmod", [ [ "arg1", "embedded-4-T9" ], [ "arg2", "embedded-4-T8" ] ] ],
        [ "embedded-4-R5", "reparandum", [ [ "arg1", "embedded-4-T8" ], [ "arg2", "embedded-4-T6" ] ] ],
        [ "embedded-4-R6", "reparandum", [ [ "arg1", "embedded-4-T8" ], [ "arg2", "embedded-4-T7" ] ] ]
    ],
    "attributes": [],
    "error": false,
    "sentlabels": [ "4" ]
}

---

reparandum in other languages: [bej] [bm] [cs] [de] [el] [en] [es] [fr] [gsw] [hy] [pt] [qpm] [ru] [sl] [ssp] [sv] [swl] [tr] [u] [yue] [zh]

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

.