Towards a Common Framework for Computational Models of Emotion: Unifying Cognitive and Emotional Processing
Description
Towards a Common Framework for Computational Models of Emotion: Unifying Cognitive and Emotional Processing This repository contains the reference implementation of the framework proposed in the research article: "Towards a Common Framework for Computational Models of Emotion: Unifying Cognitive and Emotional Processing". Overview Current Computational Models of Emotion (CMEs) often operate as isolated systems characterized by architectural rigidity and fixed execution cycles. This framework serves as a mediating core designed to integrate heterogeneous affective components into unified, theoretically coherent processes. By decoupling domain-specific implementations, the system employs an ontology-based semantic controller to align disparate terminologies and a dependency coordinator to enable configurable execution sequences. Key Features Ontology-Based Semantic Alignment: Resolves terminological inconsistencies using a formal affective ontology comprising 139 elements. Dynamic Execution Orchestration: Employs a dependency-aware execution planner that identifies independent component groups for parallel execution while managing sequential data flow. Machine-Readable Traceability: Enriches every affective output with formal metadata and ontological provenance, yielding a semantic trace suitable for explainability analysis. Architectural Extensibility: Supports the modular integration of diverse third-party components (e.g., appraisal evaluation, mood dynamics, behavioral generators) without requiring structural redesigns. Repository Contents FRAMEWORKV7_(English).py: The main Python implementation containing the core modules (Central Executive, Semantic Controller, and Execution Coordinator). CMEs_ontology.owl: The formal OWL ontology file containing the 139 affective elements used for semantic alignment. Supplementary_Material_Tables.pdf: This document includes: Table I: Comprehensive listings of affective labels and theories extracted per model (ALMA, EEGS, EMIA, FLAME, EMA). Table II: Unified list of variability scenarios used for the semantic matching validation in Case Study 2.