IPAB Workshop - 31/07/2025 Speaker: Ruyi Wang Title: From Narratives to Minds: Computational Affective-Cognitive Theory of Mind via caseKG Abstract: Affective and cognitive components of Theory of Mind (ToM) have historically developed as separate research strands, largely due to disparities in data availability, annotation practices, and modelling paradigms. Whereas affective ToM has centred on emotional state inference from perceptual signals, cognitive ToM has focused on belief and intention modelling from linguistic or task-based data. The absence of unified frameworks has constrained progress towards integrative computational models of mental state reasoning.The caseKG framework addresses this challenge through a theory-informed approach to structured modelling of belief, emotion, cognition, and behaviour from psychological case narratives. Grounded in the Belief–Emotion–Situation–Cognition (BSEC) schema and formalised via OWL-based ontologies, caseKG extracts structured triplets aligned with clinical theory to construct interpretable, compositionally valid mental state graphs. By operationalising both affective and cognitive dimensions within a unified representational scaffold, the framework establishes a foundation for transparent, theory-aligned, and scalable computational ToM.Implications include applications in hypothesis generation, psychological case formulation, and human-centred AI, with ongoing work focused on cross-cultural generalisation, temporally dynamic modelling, and interactive clinical deployment. Jul 31 2025 13.00 - 14.00 IPAB Workshop - 31/07/2025 Ruyi Wang G.03
IPAB Workshop - 31/07/2025 Speaker: Ruyi Wang Title: From Narratives to Minds: Computational Affective-Cognitive Theory of Mind via caseKG Abstract: Affective and cognitive components of Theory of Mind (ToM) have historically developed as separate research strands, largely due to disparities in data availability, annotation practices, and modelling paradigms. Whereas affective ToM has centred on emotional state inference from perceptual signals, cognitive ToM has focused on belief and intention modelling from linguistic or task-based data. The absence of unified frameworks has constrained progress towards integrative computational models of mental state reasoning.The caseKG framework addresses this challenge through a theory-informed approach to structured modelling of belief, emotion, cognition, and behaviour from psychological case narratives. Grounded in the Belief–Emotion–Situation–Cognition (BSEC) schema and formalised via OWL-based ontologies, caseKG extracts structured triplets aligned with clinical theory to construct interpretable, compositionally valid mental state graphs. By operationalising both affective and cognitive dimensions within a unified representational scaffold, the framework establishes a foundation for transparent, theory-aligned, and scalable computational ToM.Implications include applications in hypothesis generation, psychological case formulation, and human-centred AI, with ongoing work focused on cross-cultural generalisation, temporally dynamic modelling, and interactive clinical deployment. Jul 31 2025 13.00 - 14.00 IPAB Workshop - 31/07/2025 Ruyi Wang G.03
