Home // eKNOW 2025, The Seventeenth International Conference on Information, Process, and Knowledge Management // View article

Identification and Characterization of Content Traps in YouTube Recommendation Network

Authors:
Monoarul Bhuiyan
Nitin Agarwal

Keywords: Content Traps, Characterization, YouTube Recommendation Network, Social Network Analysis.

Abstract:
YouTube’s recommendation algorithm accounts for a substantial portion of total video views, influencing what users see and engage with. This study investigates how the algorithm may contribute to the formation of content traps, which are clusters of videos that repeatedly expose users to topically similar content. We employ Focal Structure Analysis (FSA), a Social Network Analysis (SNA) approach, to identify structurally cohesive groups of videos within the recommendation network, focusing on the China–Uyghur dataset as a case study. Topic modeling and divergence metrics are used to evaluate the thematic composition of each focal structure, revealing reduced topical diversity in areas where content traps are present. Building on this, we characterize each focal structure by its topical dominance, clustering coefficient, and the relative size of the focal structures, which allows us to distinguish between structurally dense traps and large, loosely connected ones. Our results show that content traps often exhibit strong topical alignment through tightly interconnected nodes. This study contributes a framework for identifying and characterizing content traps and offers insights relevant to understanding algorithmic reinforcement in content recommendation systems.

Pages: 59 to 64

Copyright: Copyright (c) IARIA, 2025

Publication date: May 18, 2025

Published in: conference

ISSN: 2308-4375

ISBN: 978-1-68558-272-2

Location: Nice, France

Dates: from May 18, 2025 to May 22, 2025