Comparative study of the QoS criteria weighting methods and their effects on ranking of web services

This research investigates the impact of QoS criteria weights on Web Service (WS) ranking. The study divides the WS ranking process into domain-wise WS selection for ranking, QoS criteria weight calculation, and final ranking of WS. To assess the weight calculation a noval criteria weighting method “Geometric Progression Based User Preference Weighting Method” (GP-UPW) is proposed. Additionally, the existing MCDM methods: MEREC and CRITIC are implemented, and detailed experimental and procedural descriptions are provided for all methods. Experimental results demonstrate the limitations of CRITIC and MEREC, which GP-UPW effectively addresses. For ranking the WS established MCDM method: TOPSIS and PROMETHEE-II are implemented and results are examined to evaluate the effect of weights on WS ranking. Furthermore, the study reveals that PROMETHEE-II is more sensitive to weight variations than TOPSIS. The correlation analysis between PROMETHEE-II and TOPSIS rankings using identical weighting methods highlights significant discrepancies, suggesting distinct ranking behaviors. These findings contribute to a deeper understanding of WS ranking and provide insights for selecting appropriate weighting and ranking methods.