Long Input Dialogue Summarization with Sketch Supervision for Summarization of Primetime Television Transcripts

Nataliia Kees, Thien Nguyen, Tobias Eder, Georg Groh


Abstract
This paper presents our entry to the CreativeSumm 2022 shared task. Specifically tackling the problem of prime-time television screenplay summarization based on the SummScreen Forever Dreaming dataset. Our approach utilizes extended Longformers combined with sketch supervision including categories specifically for scene descriptions. Our system was able to produce the shortest summaries out of all submissions. While some problems with factual consistency still remain, the system was scoring highest among competitors in the ROUGE and BERTScore evaluation categories.
Anthology ID:
2022.creativesumm-1.5
Volume:
Proceedings of The Workshop on Automatic Summarization for Creative Writing
Month:
October
Year:
2022
Address:
Gyeongju, Republic of Korea
Editor:
Kathleen Mckeown
Venue:
CreativeSumm
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
29–35
Language:
URL:
https://aclanthology.org/2022.creativesumm-1.5
DOI:
Bibkey:
Cite (ACL):
Nataliia Kees, Thien Nguyen, Tobias Eder, and Georg Groh. 2022. Long Input Dialogue Summarization with Sketch Supervision for Summarization of Primetime Television Transcripts. In Proceedings of The Workshop on Automatic Summarization for Creative Writing, pages 29–35, Gyeongju, Republic of Korea. Association for Computational Linguistics.
Cite (Informal):
Long Input Dialogue Summarization with Sketch Supervision for Summarization of Primetime Television Transcripts (Kees et al., CreativeSumm 2022)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-4/2022.creativesumm-1.5.pdf
Data
SummScreen