How Can News Break through “Algorithmic Fragmentation”? An ABT Algorithmic Auditing Study from the Perspective of News Public Visibilit

Key Points

• This research aims to investigate how news can break through algorithmic fragmentation and the characteristics of such news.
• Conducted virtual agent testing to analyze algorithmic distribution logic on a major short video platform.
• Examined public visibility distribution of news under varying conditions.
• Investigated the relationship between public visibility and news value features.
• Explored the influence of selective exposure on the dissemination of publicly visible news.
• News articles that are proximate and authoritative have higher public visibility in algorithmic recommendations.
• Personalized news exposure and public visibility can coexist and even complement each other under algorithmic adjustments.
• Algorithms can connect community relationships by impacting news public visibility.

Abstract