Labeled Corpus of Icelandic Homographs (24.04)

  Authors

Nikulásdóttir, Anna Björk

  Item identifier

http://hdl.handle.net/20.500.12537/327

 Date issued

2024-04-16

 Type

corpus, text

 Size

9 mb

 Language(s)

Icelandic

 Description

This labeled corpus of Icelandic homographs contains a) a list of all labeled homographs (ice_homographs.txt) and b) a corpus of sentences extracted from a selection of the Icelandic Gigaword Corpus (IGC-2022 - annotated version, http://hdl.handle.net/20.500.12537/254 ), containing these homographs, labeled by pronunciation. The folder ice_homographs_labeled contains one tsv-file for each homograph. The files contain a selection of sentences where the homograph in question is marked: [[homograph]] and tab-separated a label, 0 or 1. The homographs all follow the pattern V-ll-(V|$), i.e. vowel-ll-vowel_or_end. Examples: 'villa', 'holl'. There are two pronunciation possibilities for the homographs: /tl/ and /l/ and the labels are set accordingly, /tl/ = 1, /l/ = 0. The corpus can be used e.g. for the training of a homograph classifier, and for linguistic research. Þetta gagnasett inniheldur a) lista af samstafa orðum (homographs) í skjalinu ice_homographs.txt og b) gagnasett af setningum úr Risamálheildinni (IGC-2022 - annotated version, http://hdl.handle.net/20.500.12537/254 ) sem innihalda þessi orð, merkt eftir framburði. Mappan ice_homographs_labeled inniheldur eitt tsv-skjal fyrir hvert samstafa orð í gagnasettinu. Skjölin innihalda safn setninga þar sem viðkomandi samstafa orð er merkt með hornklofum: [[samstafa]] og merkingin 0 eða 1 er í enda línu á eftir "tab". Öll samstafa orðin fylgja mynstrinu V-ll-(V|$), þ.e. sérhljóð-ll-sérhljóð_eða_lok_orðs. Dæmi: 'villa', 'holl'. Þessi orð eru borin fram á mismunandi hátt eftir merkingu þeirra, annars vegar sem /tl/ og hins vegar sem /l/. Merkingarnar í gagnasettinu eru: /tl/ = 1 og /l/ = 0. Gangasettið má m.a. nota til þess að þjálfa líkan sem flokkar samstafa orð eftir framburði og til málvísindarannsókna.

 Publisher

Grammatek ehf.

 Acknowledgement

Ministry of Culture and Business Affairs

Project code: Prosody and Intonation Analysis (T11)

Project name: Language Technology for Icelandic 2019-2023

 Subject(s)

homographs tts

 Collection(s)

Clarin IS

Show full item record