1Introduction The existing privacy protection methods of trajectory data publication have two problems.First,existing methods[1-3]cannot resist attribute link attacks effectively.In practical scenarios,it is necessary...1Introduction The existing privacy protection methods of trajectory data publication have two problems.First,existing methods[1-3]cannot resist attribute link attacks effectively.In practical scenarios,it is necessary to integrate data from a variety of fields for data analysis,such as cross-domain recommendations.We usually need to obtain the user attributes outside of the trajectory data as background knowledge.展开更多
基金supported in part by the Guangxi"Bagui Sechola"Teams and the Guangxi Natural Science Foundation(2020JA170023).
文摘1Introduction The existing privacy protection methods of trajectory data publication have two problems.First,existing methods[1-3]cannot resist attribute link attacks effectively.In practical scenarios,it is necessary to integrate data from a variety of fields for data analysis,such as cross-domain recommendations.We usually need to obtain the user attributes outside of the trajectory data as background knowledge.
