A Systematic Review of Combinatorial Testing Strategies for Cross-Border E-Commerce Platforms: Methods, Architectures, and Future Research Directions

Abstract

Cross-border e-commerce platforms have significantly transformed global trade by enabling seamless transactions between buyers and sellers across diverse geographical regions. These platforms operate in complex environments characterized by varied user preferences, multiple payment gateways, diverse regulatory requirements, and multilingual interfaces, making reliability, scalability, and correctness critical challenges. Combinatorial testing has emerged as an effective approach to systematically evaluate interactions among multiple input parameters, reducing testing effort while maintaining high fault detection capability. This review examines combinatorial testing strategies applied to cross-border e-commerce systems, focusing on testing methods, architectural considerations, and emerging research trends. Techniques such as pairwise testing, t-way testing, constraint-based testing, and adaptive combinatorial testing are analyzed alongside modern architectures including microservices, cloud-based infrastructures, and distributed systems. The study highlights trends such as the integration of artificial intelligence, automated test generation, and optimization methods for large-scale applications. Despite its advantages, challenges related to handling complex constraints, scalability, and real-time testing persist. The findings underscore the need for hybrid testing strategies, enhanced tool support, and the incorporation of machine learning techniques to improve testing efficiency, with future directions emphasizing intelligent and adaptive testing frameworks.