Network-based computational communication

Network-based computational communication research has become a cornerstone in understanding the mechanisms governing opinion formation, dissemination, and polarization within modern digital ecosystems. This letter synthesizes recent advances across three levels: individual-, group-, and algorithmic-level communications. At the individual level, factors such as network structure, behavioral heterogeneity, and external influences are identified as critical in shaping opinion diffusion dynamics. Group-level interactions emphasize peer pressure, community structure, and high-order coupling as drivers of collective behaviors, while algorithmic mechanisms are shown to reshape opinion landscapes, often resulting in polarization and communities fragmentation. The evolution of large language models (LLMs) and their use in social simulations provide novel opportunities to examine intricate phenomena such as polarizing and misinformation. This paves the way for extensive future research, which has the potential to fundamentally deepen our understanding of communication processes within digital ecosystems and their societal impacts. By bridging methodological advancements with interdisciplinary perspectives, network-based computational communication stands poised to provide critical insights into the dynamics of modern public discourse.