6. Need Of MLT-PPDM
Need of MLT PPDM Service.
Diversity Attack
Two Solution On Preventing Diversity Attack
Multi-party Computation Security Approach
Perturbation Approach
7. Multi-party Computation Security
Approach
provides a great level of security to the privacy by only providing them the
information which they have mined rather than any other preferences to data.
This algorithms are very effective but the are very expensive and very rarely used.
To avoid the high cost various other solutions have got designed like constructing
decision trees over horizontal data while k-means clustering and association rule
mining are designed for the vertical data mining.
8. Perturbation Approach
1)Single Level Trust PPDM
Before Data is Published to third parties for Mining Purposes
A Data Owner Generates only one perturbed copy of its Data with affixed
Amount of Uncertainty.
2)Multi-Level Trust PPDM
This Introduces the Multi-level Trust on Data Miner.
In this Different Perturbed copies are available to Data miner at different trust Levels.
IT Allows a data owner to generate perturbed copies of its data for arbitrary trust levels
on demand. This feature offers data owners maximum flexibility.
9. 2) Multi-Level Trust PPDM
Perturbation of Data
Gaussian noise
Without noise With Gaussian noise
11. Continues…
Modules of Architecture
1) User
2) Security System
3) Fraud User
4) Admin
5)Database
12. DFD of MLT PPDM
Start
Extract The DataBatch Generation
Miner Request
Load The Data
Check
Trust
level
High low Medium
Generates Perturbed copies
Database
13. Conclusion :
the Expanded PPDM to multilevel trust (MLT) is introduced
and we increases the scope of PPDM, where in existing
system single level trust is available .
Multilevel Trust Privacy Preserving Data Mining allows to
generate multilevel trust fragmented copies of data for
developed by data owner.
14. References
1) S. Zhang, C. Zhu, J. K. O. Sin, and P. K. T. Mok, A novel ultrathin elevated channel low-
temperature poly-Si TFT, IEEE Electron Device Lett., vol. 20, pp. 569571, Nov. 1999.
2) Yaping Li, Minghua Chen, Qiwei Li, and Wei Zhang IEEE paper on Enabling Multilevel Trust in
Privacy Preserving Data Mining
3) https://www.youtube.com/watch?v=Ohdm2sQF-as
4) https://www.youtube.com/watch?v=o67IDJt4_IM
5) https://www.youtube.com/watch?v=-nFJ6EODe5U
6)https://www.youtube.com/watch?v=ygwkdmgoy3E
7) K. Liu, H. Kargupta, and J. Ryan, Random Projection-Based Multiplica-tive Data Perturbation for
Privacy Preserving Distributed Data Mining, IEEE Trans. Knowl-edge and Data Eng., vol. 18, no. 1,
pp. 92-106, Jan. 2006.
Yaping Li, Minghua Chen, Qiwei Li and Wei Zhang IEEE MANUSCRIPT ACCEPTED FOR
PUBLICATION IN IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, 2011