A Review on Text Sanitization

Abstract

Information is essential for all purpose of activities such as research, business decision making, etc. In this internet technology age there is no scarcity of information also. But if the information reveals the identity of a person or if it discloses confidential matters, then such information is a serious threat to privacy. So before publishing or sharing documents, the sensitive information should be removed or masked. This is the major goal of Text sanitization. Several semi-automatic and automatic methods are used for identifying sensitive information and thereby sanitizing the document by removing such terms. This broadens the users using the document due to their lowered classification level and also privacy is preserved.

Conceptual Model of National Intellectucal System for Children Safety in Internet Environment

“…The gist of this method is the prevention of audience's access to vulnerable, personal, age-restricted, confidential, not recommended etc. contents, in other words, the application of specific limitations on contents [22]. DS is usually used during the editing of published materials, daily online media, cloud technologies and in other issues [23].…”

Verilənlərin sanitarizasiyasının informasiya təhlükəsizliyinin təmin olunmasında rolu

“…Məlumat şəxsin kimliyini aşkar edikdə və ya məxfi hesab olunan məsələləri açıqladıqda, konfidensionallıq üçün ciddi bir təhlükə yaradır. Araşdırmalar göstərir ki, fərdlərə və ya təşkilatlara məxsus olan məlumatların məxfiliyinin pozulması riski getdikcə artır [1]. Hazırda mütəxəssisləri ən çox narahat edən məsələlərdən biri verilənlərin onlayn mühitdə təhlükəsizliyinin təmin edilməsidir.…”

ER-AE: Differentially Private Text Generation for Authorship Anonymization

“…However, the studies on privacy protection for textual data are still preliminary. Most related works only focus on replacing the sensitive key phrases in the text (Vasudevan and John, 2014) without considering the author's writing style, which is indeed a strong indicator of a person's identity. Even though some textual data, such as double-blind academic reviews, is released anonymously, the adversaries may recover the author's identity using the personal traits in writing.…”