PROJECT TITLE :
A Hybrid Approach to Private Record Matching
Real-world entities don't seem to be continuously represented by the same set of options in different data sets. So, matching records of the same real-world entity distributed across these knowledge sets may be a difficult task. If the info sets contain personal data, the matter becomes even more tough. Existing solutions to this downside typically follow two approaches: sanitization techniques and cryptographic techniques. We have a tendency to propose a hybrid technique that mixes these 2 approaches and enables users to trade off between privacy, accuracy, and price. Our main contribution is the use of a blocking phase that operates over sanitized information to filter out in a very privacy-preserving manner pairs of records that don't satisfy the matching condition. We have a tendency to additionally offer a proper definition of privacy and prove that the participants of our protocols learn nothing alternative than their share of the result and what can be inferred from their share of the result, their input and sanitized views of the input knowledge sets (that are thought-about public data). Our methodology incurs significantly lower costs than cryptographic techniques and yields significantly a lot of accurate matching results compared to sanitization techniques, even when privacy requirements are high.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here