A novel cryptographic protocol for privacy preserving classification over distributed encrypted databases

A novel cryptographic protocol for privacy preserving classification over distributed encrypted databases

연구 분야: Databases

학회: Journal of Banking and Financial Technology

Recent progression in Information Technology facilitated the collection and storage of large amounts of data to be accessed by multiple parties in a distributed manner. Privacy is an important concern while mining sensitive data. In a distributed data scenario, when the data is available in encrypted form, mining it without sharing original data among the involved parties is a challenging task. One of the activities in privacy preserving data mining is privacy preserving data classification. In this work, we propose a privacy preserving -NN data classification technique for distributed encrypted databases. Our classification approach uses a private Jaccard similarity measure, which is based on privacy equality testing protocol. We also discuss the security analysis of the proposed protocol with respect to various cryptographic attacks.