On task effects in NLG corpus elicitation: A replication study using mixed effects modeling

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

9 Downloads (Pure)

Abstract

Task effects in NLG corpus elicitation recently started to receive more attention, but are usu- ally not modeled statistically. We present a controlled replication of the study by Van Mil- tenburg et al. (2018b), contrasting spoken with written descriptions. We collected additional written Dutch descriptions to supplement the spoken data from the DIDEC corpus, and an- alyzed the descriptions using mixed effects modeling to account for variation between par- ticipants and items. Our results show that the effects of modality largely disappear in a con- trolled setting.
Original languageEnglish
Title of host publicationProceedings of the 12th International Conference on Natural Language Generation (INLG 2019)
EditorsKees van Deemter, Chenghua Lin, Hiroya Takamura
Publication statusPublished - 2019
Event12th International conference on Natural Language Generation (INLG 2019) - Tokyo, Japan
Duration: 29 Oct 20191 Nov 2019
https://www.inlg2019.com

Conference

Conference12th International conference on Natural Language Generation (INLG 2019)
CountryJapan
CityTokyo
Period29/10/191/11/19
Internet address

    Fingerprint

Keywords

  • Natural Language Generation
  • Image Description
  • Modality

Cite this

van Miltenburg, E., van de Kerkhof, M., Koolen, R., Goudbeek, M., & Krahmer, E. (2019). On task effects in NLG corpus elicitation: A replication study using mixed effects modeling. In K. van Deemter, C. Lin, & H. Takamura (Eds.), Proceedings of the 12th International Conference on Natural Language Generation (INLG 2019)