Modern methods of filtering and searching for goods in electronic commerce systems

Abstract

The rapid expansion of electronic commerce has resulted in product catalogs of unprecedented scale, making efficient information retrieval one of the central technical and design challenges of modern digital platforms. As the number of available products on a single platform can reach hundreds of thousands or even millions of items, the ability to locate relevant goods quickly and accurately becomes a critical factor in determining user satisfaction and overall platform performance. This article examines modern methods of filtering and searching for goods in e-commerce systems, with a focus on their role in improving usability, search accuracy, and commercial outcomes.The study analyzes the main approaches to organizing product search in large-scale digital catalogs: multi-parameter (faceted) filtering and recommendation systems. Multi-parameter filtering is considered as a mechanism for constraining result sets through simultaneously applied criteria such as price range, product category, brand, and technical specifications. Its implementation relies on optimized indexing strategies, including inverted indexes and precomputed facet counts, which ensure low query latency even under high filter complexity. The usability dimension of filtering is also addressed, with attention given to interface design challenges such as decision fatigue and empty result states.Recommendation systems are examined as tools of proactive personalization, covering content-based filtering, collaborative filtering, and hybrid approaches. Each paradigm is analyzed in terms of its data requirements, strengths, and known limitations, including the cold-start problem for new users and items, recommendation diversity constraints, and filter bubble effects that may narrow user exposure to the broader catalog. The role of machine learning in continuously improving search relevance and adapting to evolving user preferences is investigated throughout.A generalized structural model of the product search system in modern online stores is proposed, illustrating the interaction between the user interface layer, the core search and personalization layer, and the underlying data infrastructure. The analysis shows that the combined application of these methods enables e-commerce platforms to significantly reduce irrelevant results, accelerate product discovery, and increase user engagement. The results obtained can be applied in the development or improvement of e-commerce information systems and serve as a basis for further research in intelligent search and personalization technologies.