文章基本信息

标题：Enhanced Slicing Technique for Improving Accuracy in Crowdsourcing Database
本地全文：下载
作者：T.Malathi ; S. Nandagopal
期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2014
期号：ICETS
页码：278
出版社：S&S Publications
摘要：In recent years, privacy preserving hasseen rapid growth which leads to an increase in thecapability to store and retrieve personal datasetwithout revealing sensitive information about theindividuals. Different techniques have been proposedto improve accuracy in crowdsourcing database.Anonymization techniques such as, generalizationand bucketization, are designed for improvingaccuracy in privacy preserving method. But themalicious workers can hack the private informationof the user and misuse it. Recent work has beenshown that k-anonymity for generalization lossesconsiderable amount of information especially forhigher dimensionality data. l-diversity forbucketization does not able to prevent membershipdisclosure. In this paper we introduce a noveltechnique called overlapped slicing, which partitionsthe data in both horizontal and vertical manner.Slicing preserves better data utility thangeneralization and bucketization techniques. As anextension we proposed a technique called overlappedslicing, in which an attribute is divided into morethan one column. The release in each column consistsof more attribute correlations. Important advantageof this work is to handle high-dimensional data andalso preserves better privacy than the previoustechniques.
关键词：crowdsourcing; k-anonymity; ldiversity;generalization; bucketization