Deep SMAnE-Deep Similarity Matrix Adjustment and Evaluation to Improve Schema Matching

Schema matching is at the basis of integrating structured and semistructured data, serving as a handy tool in multiple contemporary business and commerce applications. Being investigated in the fields of databases, AI, semantic Web and data mining for many years, the core challenge still remains the ability to create quality matchers, automatic tools for identifying correspondences among data concepts (e.g., database attributes). In this work, we offer a novel post processing step to schema matching that improves the final matching outcome without human intervention. We present a new mechanism, similarity matrix adjustment, to calibrate a matching result and propose an algorithm (dubbed ADnEV) that manipulates, using deep neural networks, similarity matrices, created by state-ofthe-art matchers. ADnEV learns two models that iteratively adjust and evaluate the original similarity matrix. We show conditions for dominance and convergence of ADnEV and demonstrate empirically the effectiveness of the proposed algorithmic solution for improving matching results, using real-world benchmark ontology and schema sets. ACM Reference Format: Roee Shraga, Avigdor Gal, and Haggai Roitman. 2019. Deep SMAnE Deep Similarity Matrix Adjustment and Evaluation to Improve Schema Matching. In Proceedings of ACM Conference (Conference’17). ACM, New York, NY, USA, 11 pages. https://doi.org/10.1145/nnnnnnn.nnnnnnn