📄 Abstract
Adult media platforms remain underexamined in recommender-systems and algorithmic fairness research despite their scale, intimate data environment, and reliance on personalization. This paper develops a socio-technical framework for studying algorithmic amplification in adult-platform recommendation systems, focusing on whether sensitive content is surfaced at rates above a user’s baseline exposure. Building on warning-amplification logic from sensitivity-aware recommendation research, the paper formalizes amplification through prevalence-based and normalized measures that compare recommendation outputs with user histories. It combines these measures with metadata analysis, semantic clustering of tags, comparative benchmarking against general-media datasets, and persona-based black-box auditing. The paper argues that adult-platform recommenders should be treated as a distinct high-risk domain because sensitive categories are central to content organization rather than peripheral. It concludes with an ethical protocol and a roadmap for future empirical audits under GDPR- and DSA-relevant governance concerns.
🏷️ Keywords
📚 How to Cite:
Abhijit Joshi , Soumya Ranjan Bhoi, Nikhil Agrawal , ALGORITHMIC AMPLIFICATION IN ADULT MEDIA PLATFORMS: A STUDY ON RECOMMENDATION OF SENSITIVE CONTENT , Volume 12 , Issue 6, June 2026, EPRA International Journal of Multidisciplinary Research (IJMR) , Pages: 208 - 216 , DOI: https://doi.org/10.36713/epra28279