Service Quality Assessment of Cryptocurrency Exchanges Using the Fermatean Fuzzy TOPSIS Method: A Robust Multi-Criteria Framework with AI-Based Comparative Analysis (GAZI UNIVERSITY JOURNAL OF SCIENCE)

Summary

• The study evaluates cryptocurrency exchange service quality across 20 sub-criteria derived from 4 main dimensions of the ES-QUAL scale.p.1
• Market data from CoinMarketCap as of February 2026 lists 8786 cryptocurrencies and 252 cryptocurrency exchanges operating globally.p.2
• Criteria C2.1, C2.2, C2.3, C2.4, C3.4, C4.1, C4.3, and C4.4 achieved the highest weight value of 0.056, while C1.4 had the lowest weight of 0.038.p.11
• Under expert-based FF-TOPSIS rankings, Binance (A1) secured first place, while Bitstamp (A2) ranked last among the 5 analyzed exchanges.p.12
• Comparative tests demonstrate that ChatGPT 4o produced criteria weights and exchange rankings closest to human expert evaluations.p.17

Key insights

1. System Availability and Privacy dimensions exert the greatest influence when weighting service quality criteria for digital asset platforms.p.1
2. Binance ranks first in overall service quality performance, while Bitstamp ranks lowest among evaluated global cryptocurrency exchanges.p.1
3. The Es-Qual framework requires adaptation for cryptocurrency exchanges by revising fulfillment and privacy sub-criteria to fit digital asset trading.p.9
4. Selected industry experts possessed at least 5 years of active crypto market experience and trading background across all evaluated platforms.p.10
5. ChatGPT 4o exhibited higher analytical alignment with human experts than ChatGPT 3.5 in criteria weighting and exchange ranking tasks.p.17
6. Scenario analysis shows that under Scenario 3, where sub-criteria under efficiency sum to 1, KuCoin (A5) rises to second place in the rankings.p.18
7. Because large language models produce probabilistic outputs, AI evaluations are recommended as decision-support instruments rather than replacement evaluators.p.20