A new study, coordinated by Walter Quattrociocchi of the Centro per la Data Science e la complessità per la società at the Department of Computer Science of Sapienza University of Rome, published in the prestigious journal Nature, revealed a constant in the dynamics of online interaction between users on different platforms, including a comparison with platforms of the past. The analysis suggests the persistent nature of "toxic" interactions within digital communities, highlighting a human component that remains constant in spite of platform variations, changing social norms and the passage of decades.

"The study of digital communication and the dynamics revolving around the new media is a highly topical issue that requires rigorous analysis, given the many implications that arise from it" - says Rector Antonella Polimeni. "Sapienza can boast researchers of the highest profile who study the many aspects of communication. This publication, in a prestigious journal such as Nature, confirms and consolidates the quality of our University's research activities in this field as well: an important recognition for the team coordinated by Walter Quattrociocchi and for the entire University".

The research, focused on the dynamics of online conversations and conducted by Sapienza, identified recurring behavioural patterns within the various social media, demonstrating a remarkable consistency in interactions between users despite the evolution of platforms and social norms. In particular, the study used a comparative approach on various platforms - from Facebook, Reddit, Gab, and YouTube to the less recent USNET on more than 500 million comments - to explore crucial aspects related to the persistence of "toxic" interactions in digital communities.

Key elements identified by the researchers include the length of conversations, with prolonged discussions more prone to toxicity, and polarisation, i.e. when divergent viewpoints lead to an escalation of online disagreement.

Surprisingly, toxic interactions do not act as a deterrent to user engagement, with users continuing to actively participate in conversations. This indicates a complex interaction between harmful content and participation in online debates, suggesting the resilience of users to negativity in digital environments.

"This research represents a significant advance in our understanding of online social dynamics," says Walter Quattrociocchi, "and how these are influenced by algorithms, moving beyond a single platform focus. The results underline the broad implications of algorithmic influence on social interactions. The study"- concludes Quattrociocchi - "highlights the fundamental importance of data science in analysing and interpreting human behaviour online, confirming that toxic behaviour is a deeply rooted aspect of digital interactions".

References:

Persistent interaction patterns across social media platforms and over time - Michele Avalle, Niccolò Di Marco, Gabriele Etta, Emanuele Sangiorgio, Shayan Alipour, Anita Bonetti, Lorenzo Alvisi, Antonio Scala, Andrea Baronchelli, Matteo Cinelli & Walter Quattrociocchi - Nature (2024). https://doi.org/10.1038/s41586-024-07229-y

Further Information

Walter Quattrociocchi
Department of Computer Science
walter.quattrociocchi@uniroma1.it