Abstract:In the research of privacy preserving data publishing, the present method always removes the individual identification attributes and then anonymizes the quasi-identifier attributes. This paper analyzes the situation of multiple records one individual and proposes the principle of identity-reserved anonymity. This method reserves more information while maintaining the individual privacy. The generalization and loss-join approaches are developed to meet this requirement. The algorithms are evaluated in an experimental scenario, reserving more information and demonstrating practical applicability of the approaches.