Privacy Preserving Text Document Summarization

Abstract

: Data Anonymization provides privacy preservation of the data such that input data containing sensitive information is converted into anonymized data. Hence, nobody can identify the information either directly or indirectly. During the analysis of each text document, the unique attributes reveal the identity of an entity and its private data. The proposed system preserves the sensitive data related to an entity available in text documents by anonymizing the sensitive documents either entirely or partially based on the sensitivity context which is very specific to a domain. The documents are categorized based on sensitivity context as sensitive and not-sensitive documents and further, these documents are subjected to Summarization. The proposed Privacy Preserving Text Document Summarization generates crisp privacy preserved summary of the input text document which consists of the most relevant domain-specific information related to the text document without defying an entity privacy constraints with the compression rate of 11%, the precision of 86.32%, and the recall of 84.28%.