Being at the forefront of cutting-edge research, COSMOS publishes its research studies in leading international scientific forums.

In the paper Examining Content and Emotion Bias in YouTube’s Recommendation Algorithm published at the Ninth International Conference on Human and Social Analytics (HUSO 2023) in Barcelona, Spain, the authors, Ivory Okeke, Mert Can Cakmak, Billy Spann, and Nitin Agarwal look at bias in YouTube recommendations. 

The study’s goal is to “explore YouTube’s video recommendation bias to determine if an inherent bias has an unintended impact of ‘hiding’ content related to vulnerable communities and minority groups,” said Ivory. The research found that “YouTube’s algorithm tends to recommend more videos which evoke positive emotions and encourage higher user engagement.” In terms of future research, authors plan to  “explore biases that may be present in other recommender systems.” 

In the paper Commenter Behavior Characterization on YouTube Channels published at the Fifteenth International Conference on Information, Process, and Knowledge Management (eKNOW 2023) in Venice, Italy, the authors, Shadi Shajari, Mustafa Allasad, and Nitin Agarwal discover and analyze mob behavior in YouTube comment sections. 

Shadi outlines how the paper uses “a social network analysis-based methodology for detecting commenter mobs on YouTube.” She says, “With this methodology, we analyzed thousands of videos and discovered coordination among channels, more specifically, coordinated misinformation amplification on the U.S. Military.” She notes the paper’s importance on how citizens and researchers should become more aware of online mob activities and misinformation propagation.