摘要 |
One embodiment of the present invention provides a system for tracing information leaks. The system introduces linguistic and syntactic changes to a document, and associates these changes with a user identifier, which facilitates identification of a user that may have leaked the document. During operation, the system receives a document. The system then determines a most similar original document based on the received document. The system determines difference between the most similar original document and the received document, and determines a user identifier based on the determined difference. |