Controlled Personalization in Legacy Media Online Services: A Case Study in News Recommendation
A new paper explores "controlled personalization" in news recommendation systems, a strategy combining editorial curation with algorithmic selection. An A/B test on a Norwegian news website showed that this approach significantly increased user engagement, measured by click-through rates and reduced navigation effort. The study also found that controlled personalization enhanced content diversity, broadened catalog coverage, and mitigated popularity bias, suggesting a balanced method for traditional media to adopt personalization technologies while maintaining journalistic integrity. AI
IMPACT Suggests a balanced approach for legacy media to leverage AI-driven personalization while upholding editorial standards.