Datasets:

benjamin
/

compoundpiece

ArXiv:

Tags:

License:

Dataset card Viewer Files Files and versions Community

Dataset Viewer

Auto-converted to Parquet

View in Dataset Viewer

Viewer

Subset (2)

web · 44M rows

Split (2)

train · 43.8M rows

word stringlengths 1 64 ⌀	segmentation stringlengths 1 64 ⌀	lang stringlengths 2 2
Alpakaland	Alpakaland	de
Schmidtrottmann	Schmidt-Rottmann	de
Семов	Семов	bg
Operatio	Operatio	en
bulmasa	bulmasa	tr
નોનએસીમાં	નોન-એસીમાં	gu
повішений	повішений	uk
PARANTAA	PARANTAA	fi
such	such	nl
tutoretzaorduen	tutoretza-orduen	eu
Fuentesaúco	Fuentesaúco	en
cazarile	cazarile	ro
HDD’s	HDD’s	de
Ρεμςγκινγκάμπ	Ρεμς-Γκινγκάμπ	el
GMYO	GMYO	tr
l’ENTPE	l’ENTPE	fr
Molsheimmutzig	Molsheim-Mutzig	fr
xulextubufox	xul-ext-ubufox	es
Guarapuava	Guarapuava	pt
голугычы	голу-гычы	ky
Auszahlungsprofilen	Auszahlungsprofilen	de
ORIGAN	ORIGAN	fr
Carpetpartial	Carpet-Partial	en
Astrologisk	Astrologisk	da
Magnetabzeichen	Magnet-Abzeichen	de
Rudsvedby	Ruds-vedby	da
brainplow	brain-plow	en
Akkulaufzeittests	Akkulaufzeit-Tests	de
dzieckowychowawca	dziecko-wychowawca	pl
De’Longhi	De’Longhi	pt
latrina	latrina	it
topmount	top-Mount	es
Knechtelsdorfer	Knechtelsdorfer	de
besparingsinitiatieven	besparings-initiatieven	nl
Aggitation	Aggitation	en
neolewaka	neo-lewaka	pl
Guideoracle	Guide-Oracle	en
<br>Pokój	<br>-pokój	pl
хакунаматата	хакуна-матата	mk
kilowatttimmar	kilowatt-timmar	sv
تهرانپاسدارانمیدان	تهران-پاسداران-میدان	fa
советскомонгольские	советско-монгольские	ru
Sederul	Seder-ul	ro
जूतोंमोज़ों	जूतों-मोज़ों	hi
Perjovschi	Perjovschi	de
Tses	Tses	de
Martínetxebarría	Martín-Etxebarría	es
Berlinkölnischen	Berlin-Kölnischen	de
ACRYLIC	ACRYLIC	id
Reißverschlussseitenzugriff	Reißverschluss-Seitenzugriff	de
d`échantillon	d`échantillon	fr
philippines,concasseur	philippines-,concasseur	fr
י'טחוברת	י'ט-חוברת	he
perineurium	perineurium	en
Cigniti	Cigniti	en
1%chance	1%-Chance	de
Polaroidhochzeitsideen	Polaroid-Hochzeitsideen	es
glucoseelectrolyte	glucose-electrolyte	en
тыптықыр	тып-тықыр	kk
htp	htp	en
cartoonsi	cartoon-si	sq
eganförvaltat	egan-förvaltat	sv
changed	changed	be
hozan	hozan	tr
автогенной	автогенной	ru
groziłaby	groziłaby	pl
Valmiustila	Valmiustila	fi
Pelagie	Pelagie	it
Cadoretrento	Cadore-Trento	eu
entreesysteem	entreesysteem	nl
mammals	mammals	fi
bateriajozailea	bateria-jozailea	eu
компаниейзаказчиком	компанией-заказчиком	ru
દઝાડતી	દઝાડતી	gu
d’aubergine	d’aubergine	fr
vatofehizoro	vato-fehizoro	mg
Grudziądzwłókniarz	Grudziądz-Włókniarz	pl
ausbildungsberechtigt	ausbildungsberechtigt	de
nationalisera	nationalisera	sv
Campingchecklist	Camping-Checklist	de
BFMTV	BFMTV	en
súvisiaci	súvisiaci	sk
Clevelandbeachwood	Cleveland-Beachwood	fr
pieczarkowopomidorowym	pieczarkowo-pomidorowym	pl
membrele	membrele	ro
azobisdiisobutyronitrile	azo-bisdiisobutyronitrile	en
Kleineberg	Kleineberg	de
рапирах	рапирах	ru
Stumpertenrod	Stumpertenrod	de
nit:It	nit:-it	mt
Unabhängigkeitskonflikt	Unabhängigkeitskonflikt	de
Kohlkartoffelmischung	Kohl-Kartoffelmischung	de
qualities	qualities	en
consultar	consultar	pt
ФШУ	ФШУ	ru
gəmiyə	gəmiyə	az
Sonuçlarını	Sonuçlarını	tr
Stockhausenfestival	Stockhausen-festival	da
colorfotoservice	colorfotoservice	de
Keesjan	Kees-Jan	nl

End of preview. Expand in Dataset Viewer.

CompoundPiece

Dataset of compound words for the paper CompoundPiece: Evaluating and Improving Decompounding Performance of Language Models.

Load the balanced dataset of hyphens and non-hyphenated words scraped from the web (used as pretraining data):

load_dataset("benjamin/compoundpiece", "web")

Load the dataset of compound and non-compound words (used for fine-tuning):

load_dataset("benjamin/compoundpiece", "wiktionary")

Citation

@article{minixhofer2023compoundpiece,
  title={CompoundPiece: Evaluating and Improving Decompounding Performance of Language Models},
  author={Minixhofer, Benjamin and Pfeiffer, Jonas and Vuli{\'c}, Ivan},
  journal={arXiv preprint arXiv:2305.14214},
  year={2023}
}

License

MIT

Downloads last month: 0

Edit dataset card

Size of downloaded dataset files:

1.44 GB

Size of the auto-converted Parquet files:

1.55 GB

Number of rows:

44,235,055

Models trained or fine-tuned on benjamin/compoundpiece

benjamin/compoundpiece

Text2Text Generation • Updated Jul 24, 2023 • 57

benjamin/compoundpiece-stage1

Text2Text Generation • Updated Jul 24, 2023 • 4