Data sets for NLG blog - Revision history

Raojinfeng at 20:09, 6 August 2020

2020-08-06T20:09:33Z

← Older revision		Revision as of 20:09, 6 August 2020
Line 18:		Line 18:

	The WebNLG dataset was used in the [http://webnlg.loria.fr/pages/results.html WebNLG challenge].		The WebNLG dataset was used in the [http://webnlg.loria.fr/pages/results.html WebNLG challenge].

			=== Weather ===
			The weather dataset leverages tree-structured meaning representations for better discourse-level structuring, and collects ~30K human annotated utterances.

			This is available on [https://github.com/facebookresearch/TreeNLG Github].

	=== Weathergov ===		=== Weathergov ===
	The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.		The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://github.com/facebookresearch/TreeNLG Weather corpus] and [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.

	=== WikiBio ===		=== WikiBio ===
	No manual verification or filtering [https://ehudreiter.com/2019/09/26/generated-texts-must-be-accurate/#comment-15983]		No manual verification or filtering [https://ehudreiter.com/2019/09/26/generated-texts-must-be-accurate/#comment-15983]

Ereiter: update webnlg info

2019-10-15T11:43:49Z

update webnlg info

← Older revision		Revision as of 11:43, 15 October 2019
Line 15:		Line 15:

	=== WebNLG ===		=== WebNLG ===
	Thiago Castro Ferreira and Diego Moussallem spent six months ~~cleaning up~~ WebNLG~~, so~~ quality ~~should be good~~		Thiago Castro Ferreira and Diego Moussallem spent six months producing an enriched version of WebNLG with high-quality annotations. This is available on [https://github.com/ThiagoCF05/webnlg GitHub]

			The WebNLG dataset was used in the [http://webnlg.loria.fr/pages/results.html WebNLG challenge].

	=== Weathergov ===		=== Weathergov ===

Ereiter at 13:00, 27 September 2019

2019-09-27T13:00:34Z

← Older revision		Revision as of 13:00, 27 September 2019
Line 20:		Line 20:
	The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.		The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.

	=== ~~Wikibio~~ ===		=== WikiBio ===
	No manual verification or filtering [https://ehudreiter.com/2019/09/26/generated-texts-must-be-accurate/#comment-15983]		No manual verification or filtering [https://ehudreiter.com/2019/09/26/generated-texts-must-be-accurate/#comment-15983]

Ereiter at 12:59, 27 September 2019

2019-09-27T12:59:18Z

← Older revision		Revision as of 12:59, 27 September 2019
Line 14:		Line 14:
	[http://www.lrec-conf.org/proceedings/lrec2010/pdf/251_Paper.pdf Dutch] and [https://www.aclweb.org/anthology/W17-3532 Mandarin] versions of Tuna have been developed.		[http://www.lrec-conf.org/proceedings/lrec2010/pdf/251_Paper.pdf Dutch] and [https://www.aclweb.org/anthology/W17-3532 Mandarin] versions of Tuna have been developed.

	=== ~~Web NLG~~ ===		=== WebNLG ===
	Thiago Castro Ferreira and Diego Moussallem spent six months cleaning up WebNLG, so quality should be good		Thiago Castro Ferreira and Diego Moussallem spent six months cleaning up WebNLG, so quality should be good

Ereiter: webnlg

2019-09-27T12:59:00Z

webnlg

← Older revision		Revision as of 12:59, 27 September 2019
Line 13:		Line 13:
	=== Tuna ===		=== Tuna ===
	[http://www.lrec-conf.org/proceedings/lrec2010/pdf/251_Paper.pdf Dutch] and [https://www.aclweb.org/anthology/W17-3532 Mandarin] versions of Tuna have been developed.		[http://www.lrec-conf.org/proceedings/lrec2010/pdf/251_Paper.pdf Dutch] and [https://www.aclweb.org/anthology/W17-3532 Mandarin] versions of Tuna have been developed.

			=== Web NLG ===
			Thiago Castro Ferreira and Diego Moussallem spent six months cleaning up WebNLG, so quality should be good

	=== Weathergov ===		=== Weathergov ===

Ereiter: add wikibio note

2019-09-26T14:39:50Z

add wikibio note

← Older revision		Revision as of 14:39, 26 September 2019
Line 10:		Line 10:

	A good example of the use of Sumtime is [https://doi.org/10.1017/S1351324907004664 Automatic generation of weather forecast texts using comprehensive probabilistic generation-space models].		A good example of the use of Sumtime is [https://doi.org/10.1017/S1351324907004664 Automatic generation of weather forecast texts using comprehensive probabilistic generation-space models].

			=== Tuna ===
			[http://www.lrec-conf.org/proceedings/lrec2010/pdf/251_Paper.pdf Dutch] and [https://www.aclweb.org/anthology/W17-3532 Mandarin] versions of Tuna have been developed.

	=== Weathergov ===		=== Weathergov ===
	The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.		The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.

	=== ~~Tuna~~ ===		=== Wikibio ===
	[~~http~~://~~www~~.~~lrec-conf.org~~/~~proceedings~~/~~lrec2010~~/~~pdf~~/~~251_Paper.pdf Dutch] and [https:~~/~~/www.aclweb.org/anthology/W17~~-~~3532 Mandarin~~] ~~versions of Tuna have been developed.~~		No manual verification or filtering [https://ehudreiter.com/2019/09/26/generated-texts-must-be-accurate/#comment-15983]

Ereiter at 12:12, 29 August 2019

2019-08-29T12:12:24Z

← Older revision		Revision as of 12:12, 29 August 2019
Line 1:		Line 1:
	This blog is a supplement to [[Data sets for NLG]], which lists comments about these data sets from users, authors and other interested parties. We are especially interested in comments about appropriate and inappropriate usage of a data set, "best practice" use of a data set, useful additional information about a data set (eg, scope, how it was constructed), and pointers to related data sets which may be more appropriate for some users. Links to relevant papers and other resources are welcome		This blog is a supplement to [[Data sets for NLG]], which lists comments about these data sets from users, authors and other interested parties. We are especially interested in comments about appropriate and inappropriate usage of a data set, "best practice" use of a data set, useful additional information about a data set (eg, scope, how it was constructed), and pointers to related data sets which may be more appropriate for some users. Links to relevant papers and other resources are welcome.

			We'd love to see more content here, please email Ehud Reiter (e.reiter@abdn.ac.uk) with contributions or other comments

	=== E2E ===		=== E2E ===

Ereiter: another SumTime comment

2019-08-21T13:56:21Z

another SumTime comment

← Older revision		Revision as of 13:56, 21 August 2019
Line 5:		Line 5:

	=== SumTime ===		=== SumTime ===
			The SumTime corpus is structured as a database, and presented in text (CSV) and MDB (Microsoft Access) formats.

	A good example of the use of Sumtime is [https://doi.org/10.1017/S1351324907004664 Automatic generation of weather forecast texts using comprehensive probabilistic generation-space models].		A good example of the use of Sumtime is [https://doi.org/10.1017/S1351324907004664 Automatic generation of weather forecast texts using comprehensive probabilistic generation-space models].

Ereiter: add Tuna blog

2019-08-21T13:46:45Z

add Tuna blog

← Older revision		Revision as of 13:46, 21 August 2019
Line 9:		Line 9:
	=== Weathergov ===		=== Weathergov ===
	The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.		The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.

			=== Tuna ===
			[http://www.lrec-conf.org/proceedings/lrec2010/pdf/251_Paper.pdf Dutch] and [https://www.aclweb.org/anthology/W17-3532 Mandarin] versions of Tuna have been developed.

Ereiter: add SumTime blog

2019-08-21T13:34:10Z

add SumTime blog

← Older revision		Revision as of 13:34, 21 August 2019
Line 3:		Line 3:
	=== E2E ===		=== E2E ===
	The E2E dataset was used in the [http://www.macs.hw.ac.uk/InteractionLab/E2E/ E2E challenge].		The E2E dataset was used in the [http://www.macs.hw.ac.uk/InteractionLab/E2E/ E2E challenge].

			=== SumTime ===
			A good example of the use of Sumtime is [https://doi.org/10.1017/S1351324907004664 Automatic generation of weather forecast texts using comprehensive probabilistic generation-space models].

	=== Weathergov ===		=== Weathergov ===
	The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.		The Weathergov corpus contains the output of a template-based weather forecast generator, not human-written forecasts ([https://ehudreiter.com/2017/05/09/weathergov/ blog post]). Hence ML on Weathergov is an exercise in reverse engineering a template-based NLG system, not in training an NLG system from human data. If you want to train on human-written weather forecasts, consider using the [https://ehudreiter.files.wordpress.com/2016/12/sumtime.zip SumTime corpus] instead.