Watermarking word documents using synonym replacement

Abstract

Current security operations are unable to predict a data leakage, while the number of such incidents are continuously increasing. "2020 Ponemon Institute Cost of Insider Threats" report states the mean cost of an insider threat is $8.7 million, the survey tallied the maximum cost at nearly $26.5 million, while the minimum cost, is still significant at $489,100. Current security controls and systems are able to prevent various incidents, but at the same time produces many false-positives. Such systems are unfortunately also prone to bypassing. Agent assigned to confirm if an user has caused a data leakage, usually cannot check the content of sent files. Insider can perform various actions to make sent file seem as a personal one, or temporary file used by various applications. When file becomes public, it is hard to deduce who is responsible for the leakage. Currently there is a lack of systems able to identify the owner of a leaked document. The incident can be detected, but the person responsible for it may still remain unknown. Present study presents an implementation of a word document’s owner identification, allowing to find a person responsible for a leakage.