Improved Privacy Protecting in Distributed Grid Data Resource using Multiplicative Perturbation Based on Frequent Decision Classifier
Praveen Kumar. G1, S K Mohan Rao2, B.V. Swathi3
1Mr. Praveen Kumar. G, Research Scholar, Assistant Professor, JNTUH & CMR Engineering College, Hyderabad (Telangana), India.
2Dr. S K Mohan Rao, Professor & Principal, Gandhi Institute for Technology, Bhubaneswar (Odisha), India.
3Dr. B.V. Swathi, Professor, Geethanjali College of Engineering & Technology, Hyderabad (Telangana), India.
Manuscript received on 05 September 2019 | Revised Manuscript received on 14 September 2019 | Manuscript Published on 26 October 2019 | PP: 131-139 | Volume-8 Issue-11S2 September 2019 | Retrieval Number: K102109811S219/2019©BEIESP | DOI: 10.35940/ijitee.K1021.09811S219
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: The enormous amount of data process in distributed grid resource holds sensitive information on centralized access. The common privacy standard needs advancement to protect the sensitive data in various sectors like sharing, legal privacy and policies to access the data. The privacy standards depend the Distributed Data Mining (DDM) approaches like Association Rule Mining algorithms (ARM), clustering, and classification methods to preserve the data from unauthorized access. But the security standards have lacked privacy-preserving rules due to high dimensionality problems of data access leads more time complexity. To overcome the problem, to propose a Multiplicative Perturbation Swapping method based on Frequent Decision Classifier (MPS-FDC). This method is adaptive to data publishing secrecy to hold the privacy standards better than association rule prediction. This optimization resolves the forecasting leakages based on Persuasive Privacy Preserving Data Mining (P2PDM) to secure the data. Which this technique initially does the sanitization to reduce the dimensionality to remove un-variant the outliers. The data perturbation keeps the original data to modify using supportive noise delimiters with state matrix distortion (SMD). So the original data keep safe without effect from the outliers. The frequent rule prediction decides to classify the recurrent data from unauthorized access to disclosing crypto- privacy policy. The proposed system improves the privacy standard compared to the ARM specification rules.
Keywords: Privacy-Preserving, Data Mining, Perturbation, Swapping, Forecasting, Sanitization, Decision Classifiers.
Scope of the Article: Data Mining