Biblio

More and more organizations are today using the cloud for their business as a quite convenient alternative to in-house solutions for storing, processing, and managing data. Cloud-based solutions are then permeating almost all aspects of business organizations, resulting appealing also for functions that, already in-house, may result sensitive or security critical, and whose enforcement in the cloud requires then particular care. In this paper, we provide an approach for securely relying on cloud-based services for the enforcement of Internal Controls and Audit (ICA) functions for corporate governance. Our approach is based on the use of selective encryption and of tags to provide a level of self-protection to data and for enabling only authorized parties to access data and perform operations on them, providing privacy and integrity guarantees, as well as accountability and non-repudiation.

We study the power of interactivity in local differential privacy. First, we focus on the difference between fully interactive and sequentially interactive protocols. Sequentially interactive protocols may query users adaptively in sequence, but they cannot return to previously queried users. The vast majority of existing lower bounds for local differential privacy apply only to sequentially interactive protocols, and before this paper it was not known whether fully interactive protocols were more powerful. We resolve this question. First, we classify locally private protocols by their compositionality, the multiplicative factor by which the sum of a protocol's single-round privacy parameters exceeds its overall privacy guarantee. We then show how to efficiently transform any fully interactive compositional protocol into an equivalent sequentially interactive protocol with a blowup in sample complexity linear in this compositionality. Next, we show that our reduction is tight by exhibiting a family of problems such that any sequentially interactive protocol requires this blowup in sample complexity over a fully interactive compositional protocol. We then turn our attention to hypothesis testing problems. We show that for a large class of compound hypothesis testing problems - which include all simple hypothesis testing problems as a special case - a simple noninteractive test is optimal among the class of all (possibly fully interactive) tests.

k-anonymity is a popular model in privacy preserving data publishing. It provides privacy guarantee when a microdata table is released. In microdata, sensitive attributes contain high-sensitive and low sensitive values. Unfortunately, study in anonymity for distributing sensitive value is still rare. This study aims to distribute evenly high-sensitive value to quasi identifier group. We proposed an approach called Simple Distribution of Sensitive Value. We compared our method with systematic clustering which is considered as very effective method to group quasi identifier. Information entropy is used to measure the diversity in each quasi identifier group and in a microdata table. Experiment result show our method outperformed systematic clustering when high-sensitive value is distributed.

Differential privacy, which is due to its rigorous mathematical proof and strong privacy guarantee, has become a standard for the release of statistics with privacy protection. Recently, a lot of dynamic data publishing algorithms based on differential privacy have been proposed, but most of the algorithms use a native method to allocate the privacy budget. That is, the limited privacy budget is allocated to each time point uniformly, which may result in the privacy budget being unreasonably utilized and reducing the utility of data. In order to make full use of the limited privacy budget in the dynamic data publishing and improve the utility of data publishing, we propose a dynamic data publishing algorithm based on reinforcement learning in this paper. The algorithm consists of two parts: privacy budget allocation and data release. In the privacy budget allocation phase, we combine the idea of reinforcement learning and the changing characteristics of dynamic data, and establish a reinforcement learning model for the allocation of privacy budget. Finally, the algorithm finds a reasonable privacy budget allocation scheme to publish dynamic data. In the data release phase, we also propose a new dynamic data publishing strategy to publish data after the privacy budget is exhausted. Extensive experiments on real datasets demonstrate that our algorithm can allocate the privacy budget reasonably and improve the utility of dynamic data publishing.

Personalized medicine performs diagnoses and treatments according to the DNA information of the patients. The new paradigm will change the health care model in the future. A doctor will perform the DNA sequence matching instead of the regular clinical laboratory tests to diagnose and medicate the diseases. Additionally, with the help of the affordable personal genomics services such as 23andMe, personalized medicine will be applied to a great population. Cloud computing will be the perfect computing model as the volume of the DNA data and the computation over it are often immense. However, due to the sensitivity, the DNA data should be encrypted before being outsourced into the cloud. In this paper, we start from a practical system model of the personalize medicine and present a solution for the secure DNA sequence matching problem in cloud computing. Comparing with the existing solutions, our scheme protects the DNA data privacy as well as the search pattern to provide a better privacy guarantee. We have proved that our scheme is secure under the well-defined cryptographic assumption, i.e., the sub-group decision assumption over a bilinear group. Unlike the existing interactive schemes, our scheme requires only one round of communication, which is critical in practical application scenarios. We also carry out a simulation study using the real-world DNA data to evaluate the performance of our scheme. The simulation results show that the computation overhead for real world problems is practical, and the communication cost is small. Furthermore, our scheme is not limited to the genome matching problem but it applies to general privacy preserving pattern matching problems which is widely used in real world.