Please use this identifier to cite or link to this item:
Title: An Enhanced Anonymization Approach for Privacy Preservation in Social Networks
Researcher: Kaur, Inderjit
Guide(s): Bhardwaj, Vijay
Keywords: Computer Science
Computer Science Information Systems
Engineering and Technology
University: Guru Kashi University
Completed Date: 2021
Abstract: Data analysis over social media has become an interesting area of research since social media has become hype over users. Sixty percent of the users of this world are using different social media platforms like Facebook, Twitter, Linked-in, etc. Social media applications do not put data in the normalized form as the user is un-aware of technicalities of background data. Social media can become dangerous while handling user data from researchers. This research work aims to solve the anonymities of the research data of real-life used for futuristic researches. Anonymization is a technique which normalizes the user data when the data is used for the purpose of the research by the researchers. In general, a lot of scientists have developed and optimized anonymization algorithms for privacy preservation. K anonymity is one of the most advanced and secures algorithms for the anonymization of public data for research. Node editing and labeling is a common practice in this algorithm. The issue rises when the data set is humongous. The data label which should not be disclosed is termed as a sensitive label. The K anonymity algorithms fail to prevent the sensitive label information and hence the enhancement of K anonymity was presented by the modern frame researcher named as Yuan. Both these algorithm fails to reduce the ratio of the noisy nodes to the original set of nodes, which results in high information loss and lost integrity. This research work aims to resolve these issues by applying Machine Learning based Feed Forward back propagation neural network. The evaluation parameters are Average Path Length (APL), Average Change in Path Length(ACPL) and Ratio of Top Influential Users(RRTI) and Information Loss(IL) The proposed algorithm is compared with Zeng et al. and Wei et al. work The proposed algorithms architecture demonstrates an average precision improvement of 1.29% against Wei et al. work and 2.55% against Zeng et al. work newlineKeywords: Social Media Sites, Cuckoo Search, Privacy Preservation, Support Vector Machine
Pagination: 115
Appears in Departments:Department of Computer Applications

Files in This Item:
File Description SizeFormat 
80_recommendation.pdfAttached File104.02 kBAdobe PDFView/Open
chapter1 introduction.pdf1.14 MBAdobe PDFView/Open
chapter 2 literature survey.pdf517.33 kBAdobe PDFView/Open
chapter3 aims and objectives.pdf328.98 kBAdobe PDFView/Open
chapter4 methodology.pdf1.02 MBAdobe PDFView/Open
chapter5 result and discussions.pdf1.18 MBAdobe PDFView/Open
chapter6 conclusion and future scope.pdf259.81 kBAdobe PDFView/Open
dec.pdf141.16 kBAdobe PDFView/Open
preliminary section.pdf122.33 kBAdobe PDFView/Open
refrences.pdf564.55 kBAdobe PDFView/Open
title page.pdf425.67 kBAdobe PDFView/Open
Show full item record

Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).

Altmetric Badge: