Effectiveness of multimodal theatre mapping from four Uncle Vanya versions in self-reconstruction of trauma narratives

Against the backdrop of increasing integration between digital humanities and psychological narrative research, theatre, as a performative art combining language, movement and emotion, is gradually being recognized as a potential intervention tool for trauma memory reconstruction and emotional regulation. However, the mechanism by which multimodal theatrical information activates emotional processing and influences trauma narrative in audiences remains underexplored. This study focuses on four stylistically distinct versions of Chekhovs Uncle Vanya, constructing a structured multimodal dataset comprising text, facial expressions, speech, and movement. A multimodal Transformer and Graph Attention Network were employed to automatically map performance information to emotional structures, and a behavioral experiment was conducted to evaluate its effectiveness in trauma narrative intervention. The results show that the model significantly enhances participants abilities in emotional recognition, narrative coherence, and emotional integration, with more pronounced effects observed in trauma-experienced individuals. Moreover, stylistic differences among theatrical versions exhibited significant moderating effects on emotional activation intensity and propagation pathways, underscoring the critical value of multimodal synergy in emotional reconstruction. This research offers a technical pathway for emotion modeling and intervention via theatre mapping and expands the practical boundaries of performative arts in psychological reconstruction and educational applications.