Collaborative prediction of web service quality based on user preferences and services

Area of Science:

• Internet Services
• Web Service Quality Prediction

Background:

• Traditional collaborative filtering methods overlook user personalization and preferences.
• Accurate prediction of web service quality is crucial for enhancing user experience.

Purpose of the Study:

• To propose an improved web service quality prediction method.
• To address the limitations of existing methods in handling user individuality and QoS inconsistencies.

Main Methods:

• A novel prediction method for web service quality based on diverse Quality of Service (QoS) attributes.
• Extraction of user preference matrices using distinct rules from web data.
• Integration of a negative value filtering-based top-K method for collaborative prediction.

Main Results:

• The proposed method effectively exploits individualized differences in user preferences.
• It successfully resolves issues related to inconsistent QoS values.
• Experimental results validate the method's effectiveness and superior performance.

Conclusions:

• The developed method enhances web service quality prediction by considering user-specific factors.
• It offers more accurate predictions compared to existing approaches.
• This research contributes to more personalized and reliable internet services.