A Human-Centered Evaluation of AI-Augmented Educational Intelligence Ecosystems Using a T-Spherical Fuzzy SWARA–COPRAS Decision Framework

Abstract

The rapid adoption of artificial intelligence (AI) within the higher education sector has transformed the traditional learning system into a multifaceted AI-based educational intelligence system, as required by solid, transparent, and human-centered evaluation systems. Such ecosystems involve the multidimensional elements of pedagogical effectiveness, ethical responsibility, technological reliability, organizational readiness, and learner support, all of which are always uncertain, subjective, and professionally tentative. Conventional evaluation techniques are normally incapable of assessing this type of complexity, and thus, give wrongful evaluations and unreliable evaluation outcomes. To overcome this challenge, the proposed T-spherical fuzzy SWARA-COPRAS (TSF-SWARA-COPRAS) decision framework (the human-based) has been proposed: the criterion importance is determined by Step-wise Weight Assessment Ratio Analysis (SWARA), and it ranks the alternatives by using Complex Proportional Assessment (COPRAS). In the model of uncertain and imprecise expert judgments in membership and non-membership levels as well as hesitation levels, the proposed framework presupposes the application of T-spherical fuzzy sets (TSFSs) in the model, which would strengthen the capacity of the model to maximize the modeling flexibility and the realism of decisions. The integrated framework enables the flexible prioritization of the key dimensions of assessment and open assessment of the options of educational intelligence systems. To prove the effectiveness, strength, and usability of the proposed solution, it is represented with the help of a detailed numerical case study and a comparative analysis. The results demonstrate that the framework is effective in managing uncertainty, enhancing ranking stability, and enabling rational and transparent decision-making, which can be of significant value as a decision support tool to policy-makers, educators, and AI system developers as a way of ensuring responsible AI integration and quality improvement in the long-term in higher education.