Rationale Aware Contrastive Learning Based Approach to Classify and Summarize Crisis-Related Microblogs

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Autoren

  • Thi Huyen Nguyen
  • Koustav Rudra

Organisationseinheiten

Externe Organisationen

  • Indian School of Mines University
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Titel des SammelwerksCIKM 2022
UntertitelProceedings of the 31st ACM International Conference on Information and Knowledge Management
Herausgeber (Verlag)Association for Computing Machinery (ACM)
Seiten1552-1562
Seitenumfang11
ISBN (elektronisch)9781450392365
PublikationsstatusVeröffentlicht - 17 Okt. 2022
Veranstaltung31st ACM International Conference on Information and Knowledge Management, CIKM 2022 - Atlanta, USA / Vereinigte Staaten
Dauer: 17 Okt. 202221 Okt. 2022

Publikationsreihe

NameInternational Conference on Information and Knowledge Management, Proceedings

Abstract

Recent fashion of information propagation on Twitter makes the platform a crucial conduit for tactical data and emergency responses during disasters. However, the real-time information about crises is immersed in a large volume of emotional and irrelevant posts. It brings the necessity to develop an automatic tool to identify disaster-related messages and summarize the information for data consumption and situation planning. Besides, explainability of the methods is crucial in determining their applicability in real-life scenarios. Recent studies also highlight the importance of learning a good latent representation of tweets for several downstream tasks. In this paper, we take advantage of state-of-the-art methods, such as transformers and contrastive learning to build an interpretable classifier. Our proposed model classifies Twitter messages into different humanitarian categories and also extracts rationale snippets as supporting evidence for output decisions. The contrastive learning framework helps to learn better representations of tweets by bringing the related tweets closer in the embedding space. Furthermore, we employ classification labels and rationales to efficiently generate summaries of crisis events. Extensive experiments over different crisis datasets show that (i). our classifier obtains the best performance-interpretability trade-off, (ii). the proposed summarizer shows superior performance (1.4%-22% improvement) with significantly less computation cost than baseline models.

ASJC Scopus Sachgebiete

Zitieren

Rationale Aware Contrastive Learning Based Approach to Classify and Summarize Crisis-Related Microblogs. / Nguyen, Thi Huyen; Rudra, Koustav.
CIKM 2022 : Proceedings of the 31st ACM International Conference on Information and Knowledge Management. Association for Computing Machinery (ACM), 2022. S. 1552-1562 (International Conference on Information and Knowledge Management, Proceedings).

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Nguyen, TH & Rudra, K 2022, Rationale Aware Contrastive Learning Based Approach to Classify and Summarize Crisis-Related Microblogs. in CIKM 2022 : Proceedings of the 31st ACM International Conference on Information and Knowledge Management. International Conference on Information and Knowledge Management, Proceedings, Association for Computing Machinery (ACM), S. 1552-1562, 31st ACM International Conference on Information and Knowledge Management, CIKM 2022, Atlanta, USA / Vereinigte Staaten, 17 Okt. 2022. https://doi.org/10.1145/3511808.3557426
Nguyen, T. H., & Rudra, K. (2022). Rationale Aware Contrastive Learning Based Approach to Classify and Summarize Crisis-Related Microblogs. In CIKM 2022 : Proceedings of the 31st ACM International Conference on Information and Knowledge Management (S. 1552-1562). (International Conference on Information and Knowledge Management, Proceedings). Association for Computing Machinery (ACM). https://doi.org/10.1145/3511808.3557426
Nguyen TH, Rudra K. Rationale Aware Contrastive Learning Based Approach to Classify and Summarize Crisis-Related Microblogs. in CIKM 2022 : Proceedings of the 31st ACM International Conference on Information and Knowledge Management. Association for Computing Machinery (ACM). 2022. S. 1552-1562. (International Conference on Information and Knowledge Management, Proceedings). doi: 10.1145/3511808.3557426
Nguyen, Thi Huyen ; Rudra, Koustav. / Rationale Aware Contrastive Learning Based Approach to Classify and Summarize Crisis-Related Microblogs. CIKM 2022 : Proceedings of the 31st ACM International Conference on Information and Knowledge Management. Association for Computing Machinery (ACM), 2022. S. 1552-1562 (International Conference on Information and Knowledge Management, Proceedings).
Download
@inproceedings{fe34d6f93cac47febe3d6341f71482c7,
title = "Rationale Aware Contrastive Learning Based Approach to Classify and Summarize Crisis-Related Microblogs",
abstract = "Recent fashion of information propagation on Twitter makes the platform a crucial conduit for tactical data and emergency responses during disasters. However, the real-time information about crises is immersed in a large volume of emotional and irrelevant posts. It brings the necessity to develop an automatic tool to identify disaster-related messages and summarize the information for data consumption and situation planning. Besides, explainability of the methods is crucial in determining their applicability in real-life scenarios. Recent studies also highlight the importance of learning a good latent representation of tweets for several downstream tasks. In this paper, we take advantage of state-of-the-art methods, such as transformers and contrastive learning to build an interpretable classifier. Our proposed model classifies Twitter messages into different humanitarian categories and also extracts rationale snippets as supporting evidence for output decisions. The contrastive learning framework helps to learn better representations of tweets by bringing the related tweets closer in the embedding space. Furthermore, we employ classification labels and rationales to efficiently generate summaries of crisis events. Extensive experiments over different crisis datasets show that (i). our classifier obtains the best performance-interpretability trade-off, (ii). the proposed summarizer shows superior performance (1.4%-22% improvement) with significantly less computation cost than baseline models.",
keywords = "classification, contrastive learning, crisis events, interpretability, summarization",
author = "Nguyen, {Thi Huyen} and Koustav Rudra",
note = "Funding Information: This work was partially funded by the DFG Grant NI-1760/1-1, and the European Union{\textquoteright}s Horizon 2020 research and innovation programme under grant agreement No. 101021866.; 31st ACM International Conference on Information and Knowledge Management, CIKM 2022 ; Conference date: 17-10-2022 Through 21-10-2022",
year = "2022",
month = oct,
day = "17",
doi = "10.1145/3511808.3557426",
language = "English",
series = "International Conference on Information and Knowledge Management, Proceedings",
publisher = "Association for Computing Machinery (ACM)",
pages = "1552--1562",
booktitle = "CIKM 2022",
address = "United States",

}

Download

TY - GEN

T1 - Rationale Aware Contrastive Learning Based Approach to Classify and Summarize Crisis-Related Microblogs

AU - Nguyen, Thi Huyen

AU - Rudra, Koustav

N1 - Funding Information: This work was partially funded by the DFG Grant NI-1760/1-1, and the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 101021866.

PY - 2022/10/17

Y1 - 2022/10/17

N2 - Recent fashion of information propagation on Twitter makes the platform a crucial conduit for tactical data and emergency responses during disasters. However, the real-time information about crises is immersed in a large volume of emotional and irrelevant posts. It brings the necessity to develop an automatic tool to identify disaster-related messages and summarize the information for data consumption and situation planning. Besides, explainability of the methods is crucial in determining their applicability in real-life scenarios. Recent studies also highlight the importance of learning a good latent representation of tweets for several downstream tasks. In this paper, we take advantage of state-of-the-art methods, such as transformers and contrastive learning to build an interpretable classifier. Our proposed model classifies Twitter messages into different humanitarian categories and also extracts rationale snippets as supporting evidence for output decisions. The contrastive learning framework helps to learn better representations of tweets by bringing the related tweets closer in the embedding space. Furthermore, we employ classification labels and rationales to efficiently generate summaries of crisis events. Extensive experiments over different crisis datasets show that (i). our classifier obtains the best performance-interpretability trade-off, (ii). the proposed summarizer shows superior performance (1.4%-22% improvement) with significantly less computation cost than baseline models.

AB - Recent fashion of information propagation on Twitter makes the platform a crucial conduit for tactical data and emergency responses during disasters. However, the real-time information about crises is immersed in a large volume of emotional and irrelevant posts. It brings the necessity to develop an automatic tool to identify disaster-related messages and summarize the information for data consumption and situation planning. Besides, explainability of the methods is crucial in determining their applicability in real-life scenarios. Recent studies also highlight the importance of learning a good latent representation of tweets for several downstream tasks. In this paper, we take advantage of state-of-the-art methods, such as transformers and contrastive learning to build an interpretable classifier. Our proposed model classifies Twitter messages into different humanitarian categories and also extracts rationale snippets as supporting evidence for output decisions. The contrastive learning framework helps to learn better representations of tweets by bringing the related tweets closer in the embedding space. Furthermore, we employ classification labels and rationales to efficiently generate summaries of crisis events. Extensive experiments over different crisis datasets show that (i). our classifier obtains the best performance-interpretability trade-off, (ii). the proposed summarizer shows superior performance (1.4%-22% improvement) with significantly less computation cost than baseline models.

KW - classification

KW - contrastive learning

KW - crisis events

KW - interpretability

KW - summarization

UR - http://www.scopus.com/inward/record.url?scp=85140823314&partnerID=8YFLogxK

U2 - 10.1145/3511808.3557426

DO - 10.1145/3511808.3557426

M3 - Conference contribution

AN - SCOPUS:85140823314

T3 - International Conference on Information and Knowledge Management, Proceedings

SP - 1552

EP - 1562

BT - CIKM 2022

PB - Association for Computing Machinery (ACM)

T2 - 31st ACM International Conference on Information and Knowledge Management, CIKM 2022

Y2 - 17 October 2022 through 21 October 2022

ER -