METAPATH LEARNING IN GRAPH BASED RECOMMENDATION SYSTEMS
Gennady V. Ovechkin, D.I. Uspenskiy · 2025
This paper examines the application of knowledge graphs to solve the problem of generating product recommendations. Such approaches can be justified when the source data does not fit into a classical tabular model and involve a large number of diverse objects and relationships. A graph traversal algorithm based on pattern paths is proposed for information extraction. An experimental study of the method was conducted on some test datasets and compared with the classic recommendation algorithms.