Generally, data integration is performed through schema mapping representing high level correlation between heterogeneous data sources. Such mappings are generated using direct correspondences between data elements of source and target schemas, while other semantic relations are neglected. In this paper, we first use hierarchical relationships among properties (property precedence) as fundamental semantic relations within source and target schemas to semantically enhance schema mappings. Then, we use global property precedence relations between source and target elements to achieve Configurable Data Integration (CDI). This configurable setting allows trade-off between accuracy and completeness in query answering. Experiments using a working prototype of CDI show the potential of using this approach in various data integration scenarios.