Loading [MathJax]/jax/output/CommonHTML/jax.js
World Scientific
Skip main navigation

Cookies Notification

We use cookies on this site to enhance your user experience. By continuing to browse the site, you consent to the use of our cookies. Learn More
×

System Upgrade on Tue, May 28th, 2024 at 2am (EDT)

Existing users will be able to log into the site and access content. However, E-commerce and registration of new users may not be available for up to 12 hours.
For online purchase, please visit us again. Contact us at customercare@wspc.com for any enquiries.
Special Issue — Security and Privacy in Big Data: Challenges and Formal MethodsNo Access

Practical Secure Naïve Bayesian Classification Over Encrypted Big Data in Cloud

    https://doi.org/10.1142/S0129054117400135Cited by:7 (Source: Crossref)

    Cloud can provide much convenience for big data storage and analysis. To enjoy the advantage of cloud service with privacy preservation, huge data is increasingly outsourced to cloud in encrypted form. Unfortunately, encryption may impede the analysis and computation over the outsourced dataset. Naïve Bayesian classification is an effective algorithm to predict the class label of unlabeled samples. In this paper, we investigate naïve Bayesian classification on encrypted large-scale dataset in cloud, and propose a practical and secure scheme for the challenging problem. In our scheme, all the computation task of naïve Bayesian classification are completed by the cloud, which can dramatically reduce the burden of data owner and users. We give a formal security proof for our scheme. Based on the theoretical proof, we can strictly guarantee the privacy of both input dataset and output classification results, i.e., the cloud can learn nothing useful about the training data of data owner and the test samples of users throughout the computation. Additionally, we not only theoretically analyze our computation complexity and communication overheads, but also evaluate our implementation cost by leveraging extensive experiments over real dataset, which shows our scheme can achieve practical efficiency.

    A preliminary version [14] of this paper was presented at the Tenth International Conference on Provable Security (ProvSec’ 16).

    Communicated by Jinguang Han, Yogachandran Rahulamathavan and Willy Susilo