Knowledge Graph Construction and Applications in E-Retailing: A Review of Literature

A knowledge graph (KG) describes entities and their relations in the form of nodes and edges. The property of KG’s modeling topological data has brought possible solutions for many tasks, including web search and information retrieval. Recently, e-retailers start utilizing KGs to optimize their work processes, aiming to develop efficient approaches to handle massive, diverse, and interconnected data about suppliers, products, and customers. Various approaches have been proposed, but extant knowledge seems to be rather scarce. Thus, this study summarizes how KGs have recently been constructed and applied in the e-retailing domain. We perform a systematic literature review and categorize KG literature into multiple aspects based on their construction methodologies, application areas, and roles in supporting retailing workflow. We also describe few challenges (e.g., product variety and domain complexity), which can provide insights for future research.