Ontology Construction and Sentiment Computation Analysis in Animated Series: A Case Study of Mobile Suit Gundam

This research involves the construction of structured event templates and the definition of metadata within the animated narrative. By integrating methods such as normalization, authority control, and content analysis, subjects appearing in events are extracted and their relationships defined to form the topological information structure of the story ontology for the animated work Mobile Suit Gundam 0079. This data structure, combined with screenshots from the work, is used as the event background and provided to ChatGPT for character sentiment analysis. The study explores whether providing event backgrounds influences ChatGPT’s analysis of character emotions in the animation. In the case study, the emotional states of the characters Amuro and Char in episode 41 of Mobile Suit Gundam 0079 are analyzed under different contextual information.