arXiv reaDer
Multi-modal Misinformation Detection: Approaches, Challenges and Opportunities
As social media platforms are evolving from text-based forums into multi-modal environments, the nature of misinformation in social media is also transforming accordingly. Taking advantage of the fact that visual modalities such as images and videos are more favorable and attractive to the users and textual contents are sometimes skimmed carelessly, misinformation spreaders have recently targeted contextual connections between the modalities e.g., text and image. Hence many researchers have developed automatic techniques for detecting possible cross-modal discordance in web-based content. We analyze, categorize and identify existing approaches in addition to challenges and shortcomings they face in order to unearth new research opportunities in the field of multi-modal misinformation detection.
updated: Wed Mar 27 2024 23:27:58 GMT+0000 (UTC)
published: Fri Mar 25 2022 19:45:33 GMT+0000 (UTC)
参考文献 (このサイトで利用可能なもの) / References (only if available on this site)
被参照文献 (このサイトで利用可能なものを新しい順に) / Citations (only if available on this site, in order of most recent)アソシエイト