Enhancing E-commerce Recommender Systems through Multi-Objective Immune Algorithm
Fatima Ezzahra Zaizi, Sara Qassimi, Said Rakrak · 2024
In the dynamic world of e-commerce, recommendation systems are essential for adapting the shopping experience to individual preferences. Traditional systems often struggle to balance accuracy and diversity in recommendations. This paper introduces a novel approach, SVD-MOIA, combining the Multi-Objective Immune Algorithm with Singular Value Decomposition to optimize both accuracy and diversity. Inspired by the adaptability of the immune system, SVD-MOIA fine-tunes recommendations, achieving precision and variety. Experimental evaluations utilizing the Amazon Product Review Dataset underscore the superior performance of the SVD-MOIA approach compared to traditional algorithms, without relying on specific performance metrics. The results highlight the potential of bio-inspired, multi-objective optimization for revolutionizing e-commerce recommender systems, ensuring enhanced user experiences and improved platform performance.