Papers › Diacritics Restoration Using Neural Networks

Diacritics Restoration Using Neural Networks

1 May 2018LREC 2018 5archive 2025-07-28

Jakub N{\'a}plava, Milan Straka, Pavel Stra{\v{n}}{\'a}k, Jan Haji{\v{c}}

PaperPDFConference PDFCode

Code

arahusky/diacritics_restoration officialmentioned in papertf report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Croatian Text DiacritizationCzech Text DiacritizationFrench Text DiacritizationHungarian Text DiacritizationIrish Text DiacritizationLanguage ModelingLatvian Text DiacritizationPolish Text DiacritizationRomanian Text DiacritizationSlovak Text DiacritizationSpanish Text DiacritizationTurkish Text DiacritizationVietnamese Text Diacritization

Datasets

Introduced by this paper, per the archive.

Multilingual Dataset for Training and Evaluating Diacritics Restoration Systems

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Czech Text Diacritization Multilingual Dataset for Training and Evaluating Diacritics Restoration Systems RNN + LM Alpha-Word accuracy 99.06 #2 of 3 Archive leaderboard report
Czech Text Diacritization Multilingual Dataset for Training and Evaluating Diacritics Restoration Systems RNN Alpha-Word accuracy 98.71 #3 of 3 Archive leaderboard report
Vietnamese Text Diacritization Multilingual Dataset for Training and Evaluating Diacritics Restoration Systems RNN + LM Alpha-Word accuracy 97.73 #2 of 3 Archive leaderboard report
Vietnamese Text Diacritization Multilingual Dataset for Training and Evaluating Diacritics Restoration Systems RNN Alpha-Word accuracy 97.55 #3 of 3 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections