1 Introduction
The digital information collected by enterprises, public administrations, and governments has created enormous opportunities for knowledge-based applications. Driven by these benefits, there exists a high demand for the publication and exchange of collected data among numerous parties. However, sensitive information about users is typically contained in the original documents, and the privacy would be violated if such data is released without being processed. Document redaction, a straightforward method for privacy-preserving, is to remove sensitive information from the document. For example, document redaction is a critical approach for companies to prevent inadvertent or even malicious disclosure of proprietary formation while sharing data with outsourced operations.