TY - GEN
T1 - Multimodal Emergent Fake News Detection via Meta Neural Process Networks
AU - Wang, Yaqing
AU - Ma, Fenglong
AU - Wang, Haoyu
AU - Jha, Kishlay
AU - Gao, Jing
N1 - Funding Information:
7 IMPACT STATEMENT Fake news can manipulate important public events and becomes a global concern. If the fake news detection algorithm can function as intended, it is beneficial to prevent the spread of fake news in the early stage and correspondingly many negative public events caused by fake news may be avoided. However, we are also aware that automatic detection may suppress the public discussion. The failure modes may lie in the negation cases: if someone tries to spot the fake news by citing false information contents, the automatic algorithm may not understand the logic behind the post and incorrectly identify it as fake news. The bias may be unavoidable included in the dataset especially when the events are controversial or lacking a clear standard for annotation. Our proposed model explicitly uses the labeled sample as reference information and thus it is possible to replace the incorrect annotated support set by correct ones to correct the bias. To reduce harm brought by the automatic algorithm, both technology and human review are needed and an effective user appeal system should be employed in case the incorrect detection happened. ACKNOWLEDGMENT The authors thank the anonymous referees for their valuable comments and helpful suggestions. This work is supported in part by the US National Science Foundation under grants NSF IIS-1553411 and IIS-1956017. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. REFERENCES
Publisher Copyright:
© 2021 Owner/Author.
PY - 2021/8/14
Y1 - 2021/8/14
N2 - Fake news travels at unprecedented speeds, reaches global audiences and puts users and communities at great risk via social media platforms. Deep learning based models show good performance when trained on large amounts of labeled data on events of interest, whereas the performance of models tends to degrade on other events due to domain shift. Therefore, significant challenges are posed for existing detection approaches to detect fake news on emergent events, where large-scale labeled datasets are difficult to obtain. Moreover, adding the knowledge from newly emergent events requires to build a new model from scratch or continue to fine-tune the model, which can be challenging, expensive, and unrealistic for real-world settings. In order to address those challenges, we propose an end-to-end fake news detection framework named MetaFEND, which is able to learn quickly to detect fake news on emergent events with a few verified posts. Specifically, the proposed model integrates meta-learning and neural process methods together to enjoy the benefits of these approaches. In particular, a label embedding module and a hard attention mechanism are proposed to enhance the effectiveness by handling categorical information and trimming irrelevant posts. Extensive experiments are conducted on multimedia datasets collected from Twitter and Weibo. The experimental results show our proposed MetaFEND model can detect fake news on never-seen events effectively and outperform the state-of-the-art methods.
AB - Fake news travels at unprecedented speeds, reaches global audiences and puts users and communities at great risk via social media platforms. Deep learning based models show good performance when trained on large amounts of labeled data on events of interest, whereas the performance of models tends to degrade on other events due to domain shift. Therefore, significant challenges are posed for existing detection approaches to detect fake news on emergent events, where large-scale labeled datasets are difficult to obtain. Moreover, adding the knowledge from newly emergent events requires to build a new model from scratch or continue to fine-tune the model, which can be challenging, expensive, and unrealistic for real-world settings. In order to address those challenges, we propose an end-to-end fake news detection framework named MetaFEND, which is able to learn quickly to detect fake news on emergent events with a few verified posts. Specifically, the proposed model integrates meta-learning and neural process methods together to enjoy the benefits of these approaches. In particular, a label embedding module and a hard attention mechanism are proposed to enhance the effectiveness by handling categorical information and trimming irrelevant posts. Extensive experiments are conducted on multimedia datasets collected from Twitter and Weibo. The experimental results show our proposed MetaFEND model can detect fake news on never-seen events effectively and outperform the state-of-the-art methods.
UR - http://www.scopus.com/inward/record.url?scp=85114172969&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85114172969&partnerID=8YFLogxK
U2 - 10.1145/3447548.3467153
DO - 10.1145/3447548.3467153
M3 - Conference contribution
AN - SCOPUS:85114172969
T3 - Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
SP - 3708
EP - 3716
BT - KDD 2021 - Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining
PB - Association for Computing Machinery
T2 - 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD 2021
Y2 - 14 August 2021 through 18 August 2021
ER -