Privacy Preserving Mining

This blog present latest IEEE projects for cse topics on privacy preserving mining. Here we list various topics with sample code for the topics. So various approaches adopt by the researcher for hiding the privacy of the people was done where techniques and related codes were listed on different pages as well. In this approach of privacy outsourced data was preserved by the latest researcher for the mining of sensitive information. 

So following are topics available over here:

Privacy-Preserving-Data-Mining-Using-Matlab-IEEE

Multiparty-Data-Privacy-Preserving-Mining-Using-Gaussian-Function

Text-Document-Retrieval-from-Bulk-Dataset

Suspicious-Tax-Evasion-User-Identification

Web-User-Prediction-by-Integrating-Markov-Model-with-Different-Features

Privacy-Based-Digital-Image-Retrieval-by-Using-Visual-and-Textual-Features

A-Robust-Fractal-Code-and-LSB-Based-Image-Image-Watermarking

Robust-Keywords-Based-Document-Retrieval-by-Utilizing-AES

Advanced-Encryption-algorithm-AES-code

Privacy-Preserving-Ranked-Keyword-Search-In-Hierarchical-Data-Structure

Cloud-based-secured-document-retrival-using-encryption-techniques

Encrypted-Document-File-Retrieval-Using-User-Encrypted-Query

Website-user-query-privacy-preserving-by-data-mining

Information-retrieval-under-data-mining-technique-for-text-document

In this work one can opt different approaches for hiding sensitive information which may be obtain by manual involvement or data mining approaches. As this is reverse process of mining so first it is required to find information which need to be hide.

1. K Annonymity

2. Super class Substitution

3. Perturbation

4. Subsitution

5. Removing

6. Rule suppression which may be obtain by markov model, aprior algorithm or decision tree, etc.

7. Column Removal

8. Gaussian Mixing by adding Noise

9. Robfrugal Mixing

10. Encryption (In worst case only as it loose privacy preserving property)

Bu using above methods one can easily apply privacy preserving on any kind of data which may contain only text or only numeric or both text and numeric.