TY - GEN
T1 - Defend
T2 - 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2019
AU - Shu, Kai
AU - Cui, Limeng
AU - Wang, Suhang
AU - Lee, Dongwon
AU - Liu, Huan
N1 - Funding Information: This material is in part supported by the NSF awards #1614576, #1742702, #1820609, and #1915801, ONR grant N00014-17-1-2605 and N000141812108, and ORAU-directed R&D program in 2018. Publisher Copyright: © 2019 Association for Computing Machinery.
PY - 2019/7/25
Y1 - 2019/7/25
N2 - In recent years, to mitigate the problem of fake news, computational detection of fake news has been studied, producing some promising early results. While important, however, we argue that a critical missing piece of the study be the explainability of such detection, i.e., why a particular piece of news is detected as fake. In this paper, therefore, we study the explainable detection of fake news. We develop a sentence-comment co-attention sub-network to exploit both news contents and user comments to jointly capture explainable top-k check-worthy sentences and user comments for fake news detection. We conduct extensive experiments on real-world datasets and demonstrate that the proposed method not only significantly outperforms 7 state-of-the-art fake news detection methods by at least 5.33% in F1-score, but also (concurrently) identifies top-k user comments that explain why a news piece is fake, better than baselines by 28.2% in NDCG and 30.7% in Precision.
AB - In recent years, to mitigate the problem of fake news, computational detection of fake news has been studied, producing some promising early results. While important, however, we argue that a critical missing piece of the study be the explainability of such detection, i.e., why a particular piece of news is detected as fake. In this paper, therefore, we study the explainable detection of fake news. We develop a sentence-comment co-attention sub-network to exploit both news contents and user comments to jointly capture explainable top-k check-worthy sentences and user comments for fake news detection. We conduct extensive experiments on real-world datasets and demonstrate that the proposed method not only significantly outperforms 7 state-of-the-art fake news detection methods by at least 5.33% in F1-score, but also (concurrently) identifies top-k user comments that explain why a news piece is fake, better than baselines by 28.2% in NDCG and 30.7% in Precision.
KW - Explainable machine learning
KW - Fake news
KW - Social network
UR - http://www.scopus.com/inward/record.url?scp=85071180295&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85071180295&partnerID=8YFLogxK
U2 - 10.1145/3292500.3330935
DO - 10.1145/3292500.3330935
M3 - Conference contribution
T3 - Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
SP - 395
EP - 405
BT - KDD 2019 - Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
PB - Association for Computing Machinery
Y2 - 4 August 2019 through 8 August 2019
ER -