Home > Research > Publications & Outputs > Joint abstractive and extractive method for lon...

Links

View graph of relations

Joint abstractive and extractive method for long financial documentsummarization

Research output: Contribution to conference - Without ISBN/ISSN Conference paperpeer-review

Published

Standard

Joint abstractive and extractive method for long financial documentsummarization. / Zmandar, Nadhem; Singh, Abhishek ; El-Haj, Mahmoud et al.
2021. 99-105 Paper presented at 3rd Financial Narrative Processing Workshop (FNP 2021).

Research output: Contribution to conference - Without ISBN/ISSN Conference paperpeer-review

Harvard

Zmandar, N, Singh, A, El-Haj, M & Rayson, P 2021, 'Joint abstractive and extractive method for long financial documentsummarization', Paper presented at 3rd Financial Narrative Processing Workshop (FNP 2021), 15/09/21 - 16/09/21 pp. 99-105. <https://aclanthology.org/2021.fnp-1.19/>

APA

Zmandar, N., Singh, A., El-Haj, M., & Rayson, P. (2021). Joint abstractive and extractive method for long financial documentsummarization. 99-105. Paper presented at 3rd Financial Narrative Processing Workshop (FNP 2021). https://aclanthology.org/2021.fnp-1.19/

Vancouver

Zmandar N, Singh A, El-Haj M, Rayson P. Joint abstractive and extractive method for long financial documentsummarization. 2021. Paper presented at 3rd Financial Narrative Processing Workshop (FNP 2021).

Author

Zmandar, Nadhem ; Singh, Abhishek ; El-Haj, Mahmoud et al. / Joint abstractive and extractive method for long financial documentsummarization. Paper presented at 3rd Financial Narrative Processing Workshop (FNP 2021).7 p.

Bibtex

@conference{283116eb906249d0999d0fd701c01058,
title = "Joint abstractive and extractive method for long financial documentsummarization",
abstract = "In this paper we show the results of our participation in the FNS 2021 shared task. In our work we propose an end to end financial narrative summarization system that first selects salient sentences from the document and thenparaphrases extracted sentences. This method generates an overall concise summary that maximises the ROUGE metric with the gold standard summary. The end to end system is developed using a hybrid extractive and abstractive architecture followed by joint training using policy-based reinforcement learning to bridge together the two networks. Empirically, we achieve better scores than the proposed baselines and toplines of FNS 2021 (LexRank, TextRank, MUSE topline and POLY baseline) and we were ranked 2nd in the shared taskcompetition. Keywords: Summarization, Neural networks, Reinforcement learning, sequence to sequence learning; actor-critic methods; policy gradients.",
author = "Nadhem Zmandar and Abhishek Singh and Mahmoud El-Haj and Paul Rayson",
year = "2021",
month = oct,
day = "26",
language = "English",
pages = "99--105",
note = "3rd Financial Narrative Processing Workshop (FNP 2021) ; Conference date: 15-09-2021 Through 16-09-2021",
url = "http://wp.lancs.ac.uk/cfie/",

}

RIS

TY - CONF

T1 - Joint abstractive and extractive method for long financial documentsummarization

AU - Zmandar, Nadhem

AU - Singh, Abhishek

AU - El-Haj, Mahmoud

AU - Rayson, Paul

PY - 2021/10/26

Y1 - 2021/10/26

N2 - In this paper we show the results of our participation in the FNS 2021 shared task. In our work we propose an end to end financial narrative summarization system that first selects salient sentences from the document and thenparaphrases extracted sentences. This method generates an overall concise summary that maximises the ROUGE metric with the gold standard summary. The end to end system is developed using a hybrid extractive and abstractive architecture followed by joint training using policy-based reinforcement learning to bridge together the two networks. Empirically, we achieve better scores than the proposed baselines and toplines of FNS 2021 (LexRank, TextRank, MUSE topline and POLY baseline) and we were ranked 2nd in the shared taskcompetition. Keywords: Summarization, Neural networks, Reinforcement learning, sequence to sequence learning; actor-critic methods; policy gradients.

AB - In this paper we show the results of our participation in the FNS 2021 shared task. In our work we propose an end to end financial narrative summarization system that first selects salient sentences from the document and thenparaphrases extracted sentences. This method generates an overall concise summary that maximises the ROUGE metric with the gold standard summary. The end to end system is developed using a hybrid extractive and abstractive architecture followed by joint training using policy-based reinforcement learning to bridge together the two networks. Empirically, we achieve better scores than the proposed baselines and toplines of FNS 2021 (LexRank, TextRank, MUSE topline and POLY baseline) and we were ranked 2nd in the shared taskcompetition. Keywords: Summarization, Neural networks, Reinforcement learning, sequence to sequence learning; actor-critic methods; policy gradients.

M3 - Conference paper

SP - 99

EP - 105

T2 - 3rd Financial Narrative Processing Workshop (FNP 2021)

Y2 - 15 September 2021 through 16 September 2021

ER -