Year: 2025
Type: article
Abstract: This paper proposes an Intuitionistic Fuzzy Rubric (IFR) for grading multimodal student work that explicitly models supportive evidence, counterevidence, and uncertainty via membership (), non-membership (), and hesitation (), with . Linguistic terms are calibrated to intuitionistic fuzzy numbers using triangular prototypes with a soft-complement for doubt; indicator-level evidence is fused by intuitionistic weighted operators (IFWA/IFWG). Across criteria, we aggregate scores with either an Ordered Weighted Average (OWA) when interaction is negligible or a Choquet integral under a 2-additive capacity when synergy/redundancy matters; Shapley values provide criterion importances for feedback. The pipeline outputs a two-dimensional decision -net support and certainty-mapped to policy bands via transparent thresholds. A case study on 24 projects spanning four criteria and three indicators each demonstrates strong alignment with instructor references: Kendall's and grade accuracy 0.833, outperforming a crisp weighted rubric () and a type-1 fuzzy baseline (). Capacity analysis highlights positive Textual-Visual and Textual-Interaction synergy, while Shapley importances () align with learning objectives. Sensitivity checks show grade changes for only of projects under parameter perturbations. Overall, IFR yields reliable, interpretable, and pedagogically actionable assessment for multimodal work, with clear pathways for calibration, robustness analysis, and semester-over-semester governance.
Author: Suleiman Shelash Mohammad
Language: English
Field: Decision Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: diamond
Funder: Zarqa University
OpenAlex ID: W4414474144