Strategy for Selecting A Widely-Applicable Algorithm for Finding Semantic Relationships Between Products in The Field of E-Commerce

Semantic relationship is the logical connection between the parameters, application, and significance of two or more objects. In this paper, we analyze and select appropriate algorithms for determining semantic relationships in e-commerce, where effective product recommendations are critical for boosting customer trust and increasing sales. The study evaluates several key algorithms, including associative matching, contextual matching, and graph-based methods, in terms of scalability, suitability for dynamic data, resource consumption, and other relevant characteristics. The findings suggest that content-based filtering, the Hungarian algorithm, and the Simplex method are the most promising candidates. Also included are potential improvements through hybrid approaches and iterative refinement. The results provide a structured methodology for selecting optimal algorithms in varying scenarios.