AI-Driven Personalized Recommendation Systems: Effects on Source Credibility and Information Quality
Paper Type
Complete
Paper Number
PACIS2026-1369
Description
Although personalization has long been a central goal of recommendation systems, earlier technological constraints limited the implementation and empirical examination of more sophisticated personalization strategies. Recent advances in AI have enabled researchers to investigate personalization at both the source level (who delivers the recommendation) and the content level (what information is delivered). Drawing on the self-reference effect and information relevance theory, this study examines how source and content personalization influence users’ perceptions of source credibility and information quality. A 2 × 2 between-subjects laboratory experiment was conducted in the AI-enabled video product recommendation context (N = 161). Results indicate that source personalization significantly enhances source credibility and information quality (except for usefulness), whereas content personalization significantly improves goodwill, information understandability, and usefulness, with an interaction effect observed for goodwill. These findings advance theoretical understanding of personalization mechanisms in recommendation systems and offer practical insights for designing personalized recommendation strategies.
Recommended Citation
Li, Chengyuan; Hamari, Juho; Holopainen, Jani; and Xi, Nannan, "AI-Driven Personalized Recommendation Systems: Effects on Source Credibility and Information Quality" (2026). PACIS 2026 Proceedings. 4.
https://aisel.aisnet.org/pacis2026/hci_robotic/hci_robotic/4
AI-Driven Personalized Recommendation Systems: Effects on Source Credibility and Information Quality
Although personalization has long been a central goal of recommendation systems, earlier technological constraints limited the implementation and empirical examination of more sophisticated personalization strategies. Recent advances in AI have enabled researchers to investigate personalization at both the source level (who delivers the recommendation) and the content level (what information is delivered). Drawing on the self-reference effect and information relevance theory, this study examines how source and content personalization influence users’ perceptions of source credibility and information quality. A 2 × 2 between-subjects laboratory experiment was conducted in the AI-enabled video product recommendation context (N = 161). Results indicate that source personalization significantly enhances source credibility and information quality (except for usefulness), whereas content personalization significantly improves goodwill, information understandability, and usefulness, with an interaction effect observed for goodwill. These findings advance theoretical understanding of personalization mechanisms in recommendation systems and offer practical insights for designing personalized recommendation strategies.
Comments
12-HCI