Dialogue Response Evaluation Model with Conversational Feature Sensitive Negative Sampling

Dongjun Kang, Jin Yeong Bak

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

1 Scopus citations

Abstract

Evaluating the conversational responses is a challenging task. This is because there are so many possible responses in open-domain conversation. Recent work finds that appropriate negative samples are effective in practice, but there was problem that labeled conversational negative sample was insufficient. It is important to create an appropriate negative sample automatically because it is too costly to create all possible responses by person and annotate the response's coherence score. To address this problem, we propose a method for generating and labeling feature sensitive negative responses for conversation automatically. Besides, we show that the model learned with the generated negative sample performs well with high correlation between human score and model score.

Original languageEnglish
Title of host publicationProceedings - 2023 IEEE International Conference on Big Data and Smart Computing, BigComp 2023
EditorsHyeran Byun, Beng Chin Ooi, Katsumi Tanaka, Sang-Won Lee, Zhixu Li, Akiyo Nadamoto, Giltae Song, Young-guk Ha, Kazutoshi Sumiya, Wu Yuncheng, Hyuk-Yoon Kwon, Takehiro Yamamoto
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages183-186
Number of pages4
ISBN (Electronic)9781665475785
DOIs
StatePublished - 2023
Event2023 IEEE International Conference on Big Data and Smart Computing, BigComp 2023 - Jeju, Korea, Republic of
Duration: 13 Feb 202316 Feb 2023

Publication series

NameProceedings - 2023 IEEE International Conference on Big Data and Smart Computing, BigComp 2023

Conference

Conference2023 IEEE International Conference on Big Data and Smart Computing, BigComp 2023
Country/TerritoryKorea, Republic of
CityJeju
Period13/02/2316/02/23

Keywords

  • Conversation
  • Dialogue
  • Evaluation
  • Language Model
  • Negative Sampling

Fingerprint

Dive into the research topics of 'Dialogue Response Evaluation Model with Conversational Feature Sensitive Negative Sampling'. Together they form a unique fingerprint.

Cite this