Paper Type
Complete
Paper Number
PACIS2026-1295
Description
Editorial screening on content platforms operates under a structural mismatch: submission volume exceeds the capacity for careful human review. This study examines whether AI-based quality evaluation surfaces such overlooked content by providing information distinct from the market signals editors rely on. Analyzing 405 webtoon titles with archival promotion data (2013-2024) and multidimensional AI scoring, we find that AI evaluation and pre-promotion reader ratings capture largely non-overlapping aspects of content value, and their combination accounts for editorial decisions more effectively than either alone. We identify 137 Hidden Gems, works rated highly by AI but not promoted, whose quality profiles are comparable to those of promoted titles. Genre saturation reduces promotion chances, yet AI quality scores carry greater weight in more competitive genres. These findings reflect documented cognitive mechanisms: editors under cognitive load default to salient market signals, while AI detects quality dimensions that heuristic screening tends to underweight.
Recommended Citation
Kang, Minah and Zo, Hangjung, "Discovering Hidden Gems: AI-Human Complementarity in Content Selection on Webtoon Platforms" (2026). PACIS 2026 Proceedings. 4.
https://aisel.aisnet.org/pacis2026/ai_fow/ai_fow/4
Discovering Hidden Gems: AI-Human Complementarity in Content Selection on Webtoon Platforms
Editorial screening on content platforms operates under a structural mismatch: submission volume exceeds the capacity for careful human review. This study examines whether AI-based quality evaluation surfaces such overlooked content by providing information distinct from the market signals editors rely on. Analyzing 405 webtoon titles with archival promotion data (2013-2024) and multidimensional AI scoring, we find that AI evaluation and pre-promotion reader ratings capture largely non-overlapping aspects of content value, and their combination accounts for editorial decisions more effectively than either alone. We identify 137 Hidden Gems, works rated highly by AI but not promoted, whose quality profiles are comparable to those of promoted titles. Genre saturation reduces promotion chances, yet AI quality scores carry greater weight in more competitive genres. These findings reflect documented cognitive mechanisms: editors under cognitive load default to salient market signals, while AI detects quality dimensions that heuristic screening tends to underweight.
Comments
02-FutureofWork