Quantifying narratives: a comparative social network analysis of Upin & Ipin across series and film

Abstract Social network analysis offers a quantitative approach for examining narrative structures in animated media through character interactions. This study focuses on Upin & Ipin, a Malaysian animated franchise, by constructing directed and weighted conversational networks from two seasons of the series and the feature film, Geng: The Adventure Begins , based on speaker-listener identification. A Boolean-based weighting approach encodes interactions over fixed narrative units, and the resulting networks are analysed using community detection, network measures, centrality metrics and null configuration models. The findings reveal decentralised connectivity with small-world characteristics, where certain characters occupy structurally important bridging positions while antagonists remain peripheral. Comparative analysis indicates that the series and film differ in how their interactions are organised and distributed, reflecting variations in narrative structure between formats. Role classification based on directed and weighted degree centrality is explored as a heuristic for character prominence, with its internal consistency evaluated relative to other structural measures. Null model analysis further suggests that some centrality patterns may not be entirely explained by degree sequence alone. Overall, this study presents a reproducible framework for modelling conversational character networks in narrative media, demonstrated through its application to comparing network structure across two formats within a single franchise.