University performance data was scraped from the Times Higher Education World University Rankings website for 6 consecutive years (2021–2026) using Python & BeautifulSoup. The dataset covers ~2,450 universities across 100+ countries with five core criteria scores per record.

Criteria weights are computed by combining two complementary methods: Shannon Entropy derives objective weights from the inherent variability of each score, while Analytic Hierarchy Process (AHP) captures expert-defined subjective priorities. The combined weight balances both perspectives for a robust MCDM model.

Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) computes a closeness coefficient for each university — measuring how close its profile is to the ideal best and how far from the ideal worst. This score (0–1) is used to produce the final MCDM ranking.

Model accuracy is validated by comparing MCDM rankings to official THE rankings using Spearman's rank correlation (ρ) and Kendall's Tau (τ). Sensitivity analysis tests how rankings shift when individual criteria weights are perturbed, confirming model robustness across all 6 years.