A survey of privacy preserving data integration

Today there is necessity to integrate and share data from variety of data sources for mutual benefits and analysis purpose. Moreover, data integration involves matching of schema metadata and data across databases. Also, the databases contain personally identifying information and other sensitive information of individuals. Most of the privacy preserving data integration techniques are vulnerable to linking/re-identification attacks and some of them compromise accuracy while maintaining privacy of integrated data. Thus, protecting privacy of schemas or ontologies and data is a crucial requirement during data integration. This survey provides the various challenges, review of existing work and research directions for privacy preserving data integration.