TY - GEN
T1 - SAFE
T2 - 24th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2020
AU - Zhou, Xinyi
AU - Wu, Jindi
AU - Zafarani, Reza
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2020.
PY - 2020
Y1 - 2020
N2 - Effective detection of fake news has recently attracted significant attention. Current studies have made significant contributions to predicting fake news with less focus on exploiting the relationship (similarity) between the textual and visual information in news articles. Attaching importance to such similarity helps identify fake news stories that, for example, attempt to use irrelevant images to attract readers’ attention. In this work, we propose a Similarity-Aware FakE news detection method (SAFE) which investigates multi-modal (textual and visual) information of news articles. First, neural networks are adopted to separately extract textual and visual features for news representation. We further investigate the relationship between the extracted features across modalities. Such representations of news textual and visual information along with their relationship are jointly learned and used to predict fake news. The proposed method facilitates recognizing the falsity of news articles based on their text, images, or their “mismatches.” We conduct extensive experiments on large-scale real-world data, which demonstrate the effectiveness of the proposed method.
AB - Effective detection of fake news has recently attracted significant attention. Current studies have made significant contributions to predicting fake news with less focus on exploiting the relationship (similarity) between the textual and visual information in news articles. Attaching importance to such similarity helps identify fake news stories that, for example, attempt to use irrelevant images to attract readers’ attention. In this work, we propose a Similarity-Aware FakE news detection method (SAFE) which investigates multi-modal (textual and visual) information of news articles. First, neural networks are adopted to separately extract textual and visual features for news representation. We further investigate the relationship between the extracted features across modalities. Such representations of news textual and visual information along with their relationship are jointly learned and used to predict fake news. The proposed method facilitates recognizing the falsity of news articles based on their text, images, or their “mismatches.” We conduct extensive experiments on large-scale real-world data, which demonstrate the effectiveness of the proposed method.
KW - Fake news
KW - Multi-modal analysis
KW - Neural networks
KW - Representation learning
UR - http://www.scopus.com/inward/record.url?scp=85085729056&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85085729056&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-47436-2_27
DO - 10.1007/978-3-030-47436-2_27
M3 - Conference contribution
AN - SCOPUS:85085729056
SN - 9783030474355
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 354
EP - 367
BT - Advances in Knowledge Discovery and Data Mining - 24th Pacific-Asia Conference, PAKDD 2020, Proceedings
A2 - Lauw, Hady W.
A2 - Lim, Ee-Peng
A2 - Wong, Raymond Chi-Wing
A2 - Ntoulas, Alexandros
A2 - Ng, See-Kiong
A2 - Pan, Sinno Jialin
PB - Springer
Y2 - 11 May 2020 through 14 May 2020
ER -