<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Recommender Systems on C.CUI's Log</title><link>https://cuicaihao.github.io/tags/recommender-systems/</link><description>Recent content in Recommender Systems on C.CUI's Log</description><generator>Hugo</generator><language>en-AU</language><lastBuildDate>Sun, 30 Aug 2026 07:00:00 +1000</lastBuildDate><atom:link href="https://cuicaihao.github.io/tags/recommender-systems/index.xml" rel="self" type="application/rss+xml"/><item><title>When Social Conflict Becomes a Recommendation Signal: Why Recommendation Algorithms Cannot Stay Neutral</title><link>https://cuicaihao.github.io/posts/2026-08-30-when-social-conflict-becomes-a-recommendation-signal/</link><pubDate>Sun, 30 Aug 2026 07:00:00 +1000</pubDate><guid>https://cuicaihao.github.io/posts/2026-08-30-when-social-conflict-becomes-a-recommendation-signal/</guid><description>Social conflict is not created out of thin air by recommendation algorithms. The economic costs, reproductive risks, and domestic responsibilities in marriage are not evenly distributed in reality, and algorithmic distribution absorbs these grievances and anxieties into polarizing signals. When platforms interpret users primarily through dwell time, comments, and shares, anger can easily be mistaken for interest and fed into the next recommendation round. Yet the same technology can also improve discourse, identify common ground, and return control to users. Technology learns from society and in turn reshapes it, making the public interest an integral part of algorithm design.</description></item></channel></rss>