SMAT: An Attention-Based Deep Learning Solution to the Automation of Schema Matching

Schema matching aims to identify the correspondences among attributes of database schemas. It is frequently considered as most challenging and decisive stage existing in many contemporary web semantics systems. Low-quality algorithmic matchers fail provide improvement while manually annotation consumes extensive human efforts. Further complications arise from data privacy certain domains such healthcare, where only schema-level should be used prevent leakage. For this problem, we propose SMAT, a new deep learning model based on state-of-the-art natural language processing techniques obtain semantic mappings between source target schemas using attribute name description. SMAT avoids directly encoding domain knowledge about systems, which allows it more easily deployed across different sites. We also introduce benchmark dataset, OMAP, real-world healthcare domain. Our evaluation various datasets demonstrates potential help automate tasks.