OD-SIF: An Ontology-Driven Schema Integration Framework for e-Commerce Platform
Su-Cheng Haw, Kok-Why Ng, J. Jayapradha, Palanichamy Naveen · 2024
Automated data mapping is crucial in modern e-commerce ensuring seamless integration of diverse and heterogeneous datasets when migration from legacy systems to advanced platforms without any data loss or corruption. OD-SIF addresses these challenges by leveraging ontologies to unify data semantics, ensuring accurate and consistent schema integration. This framework excellently enhances the e-commerce operation by harmonizing product attributes such as brand, category, and specifications with a high precision of 93% and recall of 88%, completely surpassing baseline methods by 15-20% in the Fl-score. Due to OD-SIF's semantic matching capabilities, it can easily manage noisy and incomplete data with less dependency on manual intervention. On the other hand, special ontologies may further refine the limits in handling niche domains. While improving accuracy, OD-SIF allows real-time data enrichment and ensures cross-platform interoperability that supports core functionalities like inventory management, customer data processing, and transaction across diverse systems. All these advantages make OD-SIF a key enabler for digital transformation in e-commerce bridging various platforms with payment processors, logistics providers, and marketing tools into a unified efficient ecosystem.