| G. Drakopoulos, A. Krouska, Ph. Mylonas |
| F For Fake: A Survey On Graph Learning For News Veracity Classification In Social Media |
| 6th International Conference on Novel & Intelligent Digital Systems (NIDS 2026), September 23-25, 2026, Athens, Greece |
|
ABSTRACT
|
| Fake news are currently considered a major contributing factor to the overall online climate of social toxicity and political polarization. Moreover, they are the primary reason the trust in and within the digital ecosystem decays. A considerable part of their disruptive potential can be attributed to affective and cognitive mechanisms designed to attract readership while making them look plausible. Moreover, fake news are frequently propagated in a combination of mainstream and less obvious paths along social media in an effort to reach as many communities as possible. These traits along with the excessive complexity of processing social data because not only of their sheer volume but also because of their enormous generation rate and semistructured nature call for employing computational methodologies of comparable complexity if meaningful conclusions are to be reached. Additionally, higher order patterns may be necessary in order to distinguish between real and fake news. Graph deep learning techniques are explored and possible strategies based on them for discovering fake news are outlined. The latter may well lead to mining actions for effectively countering fake news.
|
| 23 September, 2026 |
| G. Drakopoulos, A. Krouska, Ph. Mylonas, "F For Fake: A Survey On Graph Learning For News Veracity Classification In Social Media", 6th International Conference on Novel & Intelligent Digital Systems (NIDS 2026), September 23-25, 2026, Athens, Greece |
[
BibTex] [
Print] [
 Back] |
