Occurrences of catastrophes such as natural or man-made disasters trigger the spread of rumours over social media at a rapid pace. Presenting a trustworthy and summarized account of the unfolding event in near real-time to the consumers of such potentially unreliable information thus becomes an important task. In this work, we propose MTLTS, the first end-to-end solution for the task that jointly determines the credibility and summary-worthiness of tweets. Our credibility verifier is designed to recursively learn the structural properties of a Twitter conversation cascade, along with the stances of replies towards the source tweet. We then take a hierarchical multi-task learning approach, where the verifier is trained at a lower layer, and the summarizer is trained at a deeper layer where it utilizes the verifier predictions to determine the salience of a tweet. Different from existing disaster-specific summarizers, we model tweet summarization as a supervised task. Such an approach can automatically learn summary-worthy features, and can therefore generalize well across domains. When trained on the PHEME dataset 29, not only do we outperform the strongest baselines for the auxiliary task of verification/rumour detection, we also achieve 21 - 35% gains in the verified ratio of summary tweets, and 16 - 20% gains in ROUGE1-F1 scores over the existing state-of-the-art solutions for the primary task of trustworthy summarization.
%0 Conference Paper
%1 10.1145/3488560.3498536
%A Mukherjee, Rajdeep
%A Vishnu, Uppada
%A Peruri, Hari Chandana
%A Bhattacharya, Sourangshu
%A Rudra, Koustav
%A Goyal, Pawan
%A Ganguly, Niloy
%B Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining
%C New York, NY, USA
%D 2022
%I Association for Computing Machinery
%K leibnizailab myown
%P 755–763
%R 10.1145/3488560.3498536
%T MTLTS: A Multi-Task Framework To Obtain Trustworthy Summaries From Crisis-Related Microblogs
%U https://doi.org/10.1145/3488560.3498536
%X Occurrences of catastrophes such as natural or man-made disasters trigger the spread of rumours over social media at a rapid pace. Presenting a trustworthy and summarized account of the unfolding event in near real-time to the consumers of such potentially unreliable information thus becomes an important task. In this work, we propose MTLTS, the first end-to-end solution for the task that jointly determines the credibility and summary-worthiness of tweets. Our credibility verifier is designed to recursively learn the structural properties of a Twitter conversation cascade, along with the stances of replies towards the source tweet. We then take a hierarchical multi-task learning approach, where the verifier is trained at a lower layer, and the summarizer is trained at a deeper layer where it utilizes the verifier predictions to determine the salience of a tweet. Different from existing disaster-specific summarizers, we model tweet summarization as a supervised task. Such an approach can automatically learn summary-worthy features, and can therefore generalize well across domains. When trained on the PHEME dataset 29, not only do we outperform the strongest baselines for the auxiliary task of verification/rumour detection, we also achieve 21 - 35% gains in the verified ratio of summary tweets, and 16 - 20% gains in ROUGE1-F1 scores over the existing state-of-the-art solutions for the primary task of trustworthy summarization.
%@ 9781450391320
@inproceedings{10.1145/3488560.3498536,
abstract = {Occurrences of catastrophes such as natural or man-made disasters trigger the spread of rumours over social media at a rapid pace. Presenting a trustworthy and summarized account of the unfolding event in near real-time to the consumers of such potentially unreliable information thus becomes an important task. In this work, we propose MTLTS, the first end-to-end solution for the task that jointly determines the credibility and summary-worthiness of tweets. Our credibility verifier is designed to recursively learn the structural properties of a Twitter conversation cascade, along with the stances of replies towards the source tweet. We then take a hierarchical multi-task learning approach, where the verifier is trained at a lower layer, and the summarizer is trained at a deeper layer where it utilizes the verifier predictions to determine the salience of a tweet. Different from existing disaster-specific summarizers, we model tweet summarization as a supervised task. Such an approach can automatically learn summary-worthy features, and can therefore generalize well across domains. When trained on the PHEME dataset [29], not only do we outperform the strongest baselines for the auxiliary task of verification/rumour detection, we also achieve 21 - 35% gains in the verified ratio of summary tweets, and 16 - 20% gains in ROUGE1-F1 scores over the existing state-of-the-art solutions for the primary task of trustworthy summarization.},
added-at = {2022-02-19T20:15:34.000+0100},
address = {New York, NY, USA},
author = {Mukherjee, Rajdeep and Vishnu, Uppada and Peruri, Hari Chandana and Bhattacharya, Sourangshu and Rudra, Koustav and Goyal, Pawan and Ganguly, Niloy},
biburl = {https://www.bibsonomy.org/bibtex/202fecdd7dcfd34b4a45df207a96bf87b/niloy},
booktitle = {Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining},
doi = {10.1145/3488560.3498536},
interhash = {c19c46fbe7e74fbf20a8420b46856c7d},
intrahash = {02fecdd7dcfd34b4a45df207a96bf87b},
isbn = {9781450391320},
keywords = {leibnizailab myown},
location = {Virtual Event, AZ, USA},
numpages = {9},
pages = {755–763},
publisher = {Association for Computing Machinery},
series = {WSDM '22},
timestamp = {2022-02-25T09:02:52.000+0100},
title = {MTLTS: A Multi-Task Framework To Obtain Trustworthy Summaries From Crisis-Related Microblogs},
url = {https://doi.org/10.1145/3488560.3498536},
year = 2022
}