SUMSUM@FNS-2020 Shared Task

Siyan Zheng, Anneliese Lu, Claire Cardie


Abstract
This paper describes the SUMSUM systems submitted to the Financial Narrative Summarization Shared Task (FNS-2020). We explore a section-based extractive summarization method tailored to the structure of financial reports: our best system parses the report Table of Contents (ToC), splits the report into narrative sections based on the ToC, and applies a BERT-based classifier to each section to determine whether it should be included in the summary. Our best system ranks 4<sup>th</sup>, 1<sup>st</sup>, 2<sup>nd</sup> and 17<sup>th</sup> on the Rouge-1, Rouge-2, Rouge-SU4, and Rouge-L official metrics, respectively. We also report results on the validation set using an alternative set of Rouge-based metrics that measure performance with respect to the best-matching of the available gold summaries.
Anthology ID:
2020.fnp-1.25
Volume:
Proceedings of the 1st Joint Workshop on Financial Narrative Processing and MultiLing Financial Summarisation
Month:
December
Year:
2020
Address:
Barcelona, Spain (Online)
Venue:
FNP
SIG:
Publisher:
COLING
Note:
Pages:
148–152
Language:
URL:
https://aclanthology.org/2020.fnp-1.25
DOI:
Bibkey:
Cite (ACL):
Siyan Zheng, Anneliese Lu, and Claire Cardie. 2020. SUMSUM@FNS-2020 Shared Task. In Proceedings of the 1st Joint Workshop on Financial Narrative Processing and MultiLing Financial Summarisation, pages 148–152, Barcelona, Spain (Online). COLING.
Cite (Informal):
SUMSUM@FNS-2020 Shared Task (Zheng et al., FNP 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/auto-file-uploads/2020.fnp-1.25.pdf