Multiple Data Owners Research Articles

Highly Secure Privacy-Preserving Outsourced k-Means Clustering under Multiple Keys in Cloud Computing

Data clustering is the unsupervised classification of data records into groups. As one of the steps in data analysis, it has been widely researched and applied in practical life, such as pattern recognition, image processing, information retrieval, geography, and marketing. In addition, the rapid increase of data volume in recent years poses a huge challenge for resource-constrained data owners to perform computation on their data. This leads to a trend that users authorize the cloud to perform computation on stored data, such as keyword search, equality test, and outsourced data clustering. In outsourced data clustering, the cloud classifies users’ data into groups according to their similarities. Considering the sensitive information in outsourced data and multiple data owners in practical application, it is necessary to develop a privacy-preserving outsourced clustering scheme under multiple keys. Recently, Rong et al. proposed a privacy-preserving outsourced k-means clustering scheme under multiple keys. However, in their scheme, the assistant server (AS) is able to extract the ratio of two underlying data records, and key management server (KMS) can decrypt the ciphertexts of owners’ data records, which break the privacy security. AS can even reduce all data records if it knows one of the data records. To solve the aforementioned problem, we propose a highly secure privacy-preserving outsourced k-means clustering scheme under multiple keys in cloud computing. In this paper, noncolluded cloud computing service (CCS) and KMS jointly perform clustering over the encrypted data records without exposing data privacy. Specifically, we use BCP encryption which has additive homomorphic property and AES encryption to double encrypt data records, where the former cryptosystem prevents CCS from obtaining any useful information from received ciphertexts and the latter one protects data records from being decrypted by KMS. We first define five protocols to realize different functions and then present our scheme based on these protocols. Finally, we give the security and performance analyses which show that our scheme is comparable with the existing schemes on functionality and security.

Secure Cloud Data Audit using Protocol and Digital Signature Techniques

Cloud computing is one of the important business models in the modern Information Technology. It provides various services (hardware, software) to the users with minimal interaction and low-cost. Storage service is one of the most useful services in cloud computing, which move data owners data from local computing system to the cloud. In this paradigm, once the data moves from the local computing system to the cloud, the data owner lost the physical control of the outsourced data on the cloud. So that, storage service creates data security challenges. Therefore, the integrity of the outsourced data has to be verified frequently using public or private verification method. In this paper we focus on two data security concern such as data confidentiality and remote data integrity on cloud storage system. In order to ensure the data integrity and reduce the data owners computational resources, in this work we have proposed a remote data integrity auditing methods such as Remote Data Audit using Protocol(RDAP) and Remote Data Audit using Digital Signature (RDADS) methods. To analyze the performance of the system, first, we define the single data owner on multiple servers and then multiple data owners on a single server for public data verification. Besides, these methods not only verify the integrity of data, but also detect the invalid data block during the verification process.

Privacy-Preserving Deep Learning via Weight Transmission

This paper considers the scenario that multiple data owners wish to apply a machine learning method over the combined dataset of all owners to obtain the best possible learning output but do not want to share the local datasets owing to privacy concerns. We design systems for the scenario that the stochastic gradient descent (SGD) algorithm is used as the machine learning method because SGD (or its variants) is at the heart of recent deep learning techniques over neural networks. Our systems differ from existing systems in the following features: {\bf (1)} any activation function can be used, meaning that no privacy-preserving-friendly approximation is required; {\bf (2)} gradients computed by SGD are not shared but the weight parameters are shared instead; and {\bf (3)} robustness against colluding parties even in the extreme case that only one honest party exists. We prove that our systems, while privacy-preserving, achieve the same learning accuracy as SGD and hence retain the merit of deep learning with respect to accuracy. Finally, we conduct several experiments using benchmark datasets, and show that our systems outperform previous system in terms of learning accuracies.

Efficient Ranked Multi-Keyword Retrieval With Privacy Protection for Multiple Data Owners in Cloud Computing

With the significant rise in the demand for infinite storage resource and high-quality retrieval service, cloud computing is deemed to be a highly remarkable technology. For privacy concerns, several research works on ranked multi-keyword search over the encrypted cloud data under multiple data owner model have been developed. However, most of these schemes are vulnerable to keyword guessing attacks and equivalence test attack. Besides, they cannot accurately return the top - $k$ ranked search results from the encrypted data of different data owners to the data users. Obviously, these drawbacks may incur leakage of privacy-sensitive keyword information and the inaccuracy of returned search results. Therefore, in this article, we first propose a novel efficient ranked multi-keyword retrieval scheme with keyword privacy for multiple data owners, which empowers the cloud server to perform multi-keyword search over the cloud data and then, return the top - $k$ ranked search results to data users without leaking any keyword and trapdoor information. Additionally, we show through rigorous security analysis that our scheme is secure against the attacks launched by inside attackers and outside attackers. Finally, the performance evaluation indicates that our scheme has more satisfactory features than the existing ranked multi-keyword search schemes.

Secure string pattern query for open data initiative

The open data initiative allows the organizations from different countries to share and integrate their data to produce value added information products and services. However, the trust relationship between any two organizations can hardly be established due to the data sensitivity and their specific responsibilities. In this paper, we propose a secure string pattern query processing framework for multiple data owners. With this framework, we design an efficient and secure indexing structure called Secure String PAttern Search tree (S2PAStree). S2PAStree allows the execution of the secure string pattern query with logarithmic time complexity. We further propose a set of secure index construction protocols under the scenario that multiple data owners share and integrate their sensitive data. Finally, we conduct comprehensive experiments on three public datasets to evaluate the efficiency and the scalability of our solutions.

Secure conjunctive multi-keyword ranked search over encrypted cloud data for multiple data owners

Recently, secure search over encrypted cloud data has become a hot research spot and challenging task. A number of secure search schemes have been proposed to try to meet this challenge. However, most of them only consider the single data owner model. In this paper, we propose a conjunctive multi-keyword ranked secure search scheme for multiple data owners. To guarantee data security and system flexibility in the multiple data owners environment, we design an ingenious secure query scheme that allows each data owner to adopt randomly chosen temporary keys to build secure indexes for different data files. An authorized data user does not need to know these temporary keys of constructing indexes and can instead randomly choose another temporary query keys to encrypt query keywords, while the cloud server can correctly perform keywords matching over encrypted data files. To rank the query results of a conjunctive multi-keyword query, the cloud server computes the similarity scores between the query and its query results according to encrypted relevance scores of keywords without obtaining any sensitive information. Extensive experiments demonstrate the correctness and practicality of the proposed scheme.

Efficient, dynamic and identity-based Remote Data Integrity Checking for multiple replicas

Nowadays, cloud storage plays an increasingly important role in our daily life. However, the cloud users do not have the physical possession of their own data anymore. To confirm whether the outsourced files are maintained intact without downloading them entirely, a mechanism namely Remote Data Integrity Checking (RDIC) is invented. Currently, some RDIC schemes allow the data owners with limited computation or communication power to delegate the checking task to a third-party verifier. However, most of these schemes rely on the complicated and resource consuming public key infrastructure (PKI). In this paper, we propose a novel identity-based RDIC scheme, namely Efficient, Dynamic and Identity-based Multiple Replication Provable Data Possession (EDID-MRPDP) without the burden of PKI. We introduce a new construction of Homomorphic Verifiable Tag (HVT) and a novel data structure namely Compressed Authentication Array (CAA), which allow EDID-MRPDP to perform batch verification for multiple data owners and cloud servers simultaneously and efficiently, both from computation and communication aspects. To the best of our knowledge, EDID-MRPDP is the first ID-based RDIC scheme with full dynamic updates and multi-replica batch checking. We provide comprehensive correctness and soundness proofs of EDID-MRPDP. Meanwhile, the detailed performance analyses and simulations show that EDID-MRPDP is practical for large-scale cloud applications.

Optimized Verifiable Fine-Grained Keyword Search in Dynamic Multi-owner Settings

Ciphertext-Policy Attribute-Based Keyword Search (CP-ABKS) schemes support both fine-grained access control and keyword-based ciphertext retrieval, which make these schemes attractive for resource-constrained users (i.e., mobile or wearable devices, sensor nodes, etc.) to store, share and search encrypted data in the public cloud. However, ciphertext length and decryption overhead in the existing CP-ABKS schemes grow with the complexity of access policies or the number of data users’ attributes. Moreover, such schemes generally do not consider the practical multi-owner setting (e.g., each file needs to be signed by multiple data owners before being uploaded to the cloud server) or prevent malicious cloud servers from returning incorrect search results. To overcome these limitations, in this paper we first design an optimized Verifiable Fine-grained Keyword Search scheme in the static Multi-owner setting (termed as basic VFKSM), which achieves short ciphertext length, fast ciphertext transformation, accelerated search process, and authentic search result verification. Then, we extend the basic VFKSM to support multi-keyword search and multi-owner update (also called as extended VFKSM). Finally, we prove that the basic (or extended) VFKSM resists the Chosen-Keyword Attack (CKA) and external Keyword-Guessing Attack (KGA). We also evaluate the performance of these schemes using various public datasets.

Privacy Preserving k-Nearest Neighbor for Medical Diagnosis in e-Health Cloud.

Cloud computing is highly suitable for medical diagnosis in e-health services where strong computing ability is required. However, in spite of the huge benefits of adopting the cloud computing, the medical diagnosis field is not yet ready to adopt the cloud computing because it contains sensitive data and hence using the cloud computing might cause a great concern in privacy infringement. For instance, a compromised e-health cloud server might expose the medical dataset outsourced from multiple medical data owners or infringe on the privacy of a patient inquirer by leaking his/her symptom or diagnosis result. In this paper, we propose a medical diagnosis system using e-health cloud servers in a privacy preserving manner when medical datasets are owned by multiple data owners. The proposed system is the first one that achieves the privacy of medical dataset, symptoms, and diagnosis results and hides the data access pattern even from e-health cloud servers performing computations using the data while it is still robust against collusion of the entities. As a building block of the proposed diagnosis system, we design a novel privacy preserving protocol for finding the k data with the highest similarity (PE-FTK) to a given symptom. The protocol reduces the average running time by 35% compared to that of a previous work in the literature. Moreover, the result of the previous work is probabilistic, i.e., the result can contain some error, while the result of our PE-FTK is deterministic, i.e., the result is correct without any error probability.

Privacy Preserving Ranked Multi-Keyword Search for Multiple Data Owners in SaaS Cloud Computing

The innovation in cloud computing has encouraged the data owners to outsource their data managing system from local sites to profitable public cloud to increase affordability of services. But people can like full benefit of cloud computing, if we are able to report very real secrecy and security concerns that come with loading sensitive personal information then we have to increase security, privacy. In view of the huge number of data users, documents in the cloud, it is important for the search facility to agree multi keywords query and arrange for result comparison ranking to meet the actual need of data recovery search and not regularly distinguish the search results. Related mechanisms on searchable encryption emphasis on single keyword search or Boolean keyword search, and often sort the search outcomes. In our system, we explain and solve the interesting problem of privacy preserving multi keywords ranked search over encrypted cloud data, and create a set of strict privacy necessities for such a safe cloud data application system to be effected in real. We can add Document owner accounts, Cloud Server account, Admin account and grant permission to various privileges to different user.

Cryptanalysis of a Symmetric Fully Homomorphic Encryption Scheme

Fully homomorphic encryption supports meaningful computations on encrypted data, and hence, is widely used in cloud computing and big data environments. Recently, Li et al . constructed an efficient symmetric fully homomorphic encryption scheme and utilized it to design a privacy-preserving-outsourced association rule mining scheme. Their proposal allows multiple data owners to jointly mine some association rules without sacrificing the data privacy. The security of the homomorphic encryption scheme against the known-plaintext attacks was established by examining the hardness of solving nonlinear systems. However, in this paper, we illustrate that the security of Li et al .’s homomorphic encryption is overvalued. First, we show that we can recover the first part of the secret key from several known plaintext/ciphertext pairs with the continued fraction algorithm. Second, we find that we can retrieve the second part of the secret key through the Euclidean algorithm for the greatest common divisor problem. Experiments on the suggested parameters demonstrate that in case of more than two homomorphic multiplications, all the secret keys of the randomly instantiated Li et al .’s encryption schemes can be very efficiently recovered, and the success probability is at least 98% for one homomorphic multiplication.

Expressive query over outsourced encrypted data

Data security and privacy concerns in cloud storage services compel data owners to encrypt their sensitive data before outsourcing. Standard encryption systems, however, hinder users from issuing search queries on encrypted data. Though various systems for search over encrypted data have been proposed in the literature, existing systems use different encrypted index structures to conduct search on different search query patterns and hence are not compatible with each other. In this paper, we propose a query over encrypted data system which supports expressive search query patterns, such as single/conjunctive keyword query, range query, boolean query and mixed boolean query, all using a single encrypted index structure. To the best of our knowledge, the proposed system enables the most expressive query pattern search among all the existing solutions. In addition, the system allows data users to simultaneously query over encrypted documents from multiple data owners using one query trapdoor and supports flexible user authorization and revocation. We show that our system is secure and resists keyword guessing attack. We also conduct extensive experiments and demonstrate that the system is more efficient than other public key searchable encryption systems.

An Efficient Ranked Multi-Keyword Search for Multiple Data Owners Over Encrypted Cloud Data

With the development of cloud storage, more data owners are inclined to outsource their data to cloud services. For privacy concerns, sensitive data should be encrypted before outsourcing. There are various searchable encryption schemes to ensure data availability. However, the existing search schemes pay little attention to the efficiency of data users' queries, especially for the multi-owner scenario. In this paper, we proposed a tree-based ranked multi-keyword search scheme for multiple data owners. Specifically, by considering a large amount of data in the cloud, we utilize the TF × IDF model to develop a multikeyword search and return the top-k ranked search results. To enable the cloud servers to perform a secure search without knowing any sensitive data (e.g., keywords and trapdoors), we construct a novel privacypreserving search protocol based on the bilinear mapping. To achieve an efficient search, for each data owner, a tree-based index encrypted with an additive order and privacy-preserving function family is constructed. The cloud server can then merge these indexes effectively, using the depth-first search algorithm to find the corresponding files. Finally, the rigorous security analysis proves that our scheme is secure, and the performance analysis demonstrates its efficacy and efficiency.

Secure multi-keyword ranked search over encrypted cloud data for multiple data owners

Secure multi-keyword ranked search over outsourced cloud data has become a hot research field. Most existing works follow the model of “Single Owner”, which just supports searching on the outsourced data belong to only one data owner. But the more realistic scenario is “Multiple Owners”, users could search on all datasets outsourced by different data owners. However, directly extending “Single Owner” schemes into “Multiple Owners” scenario still face the major two challenges: (1) the inconvenient key management and the resulting high communication cost; (2) due to the different authorities of owners, the qualities of documents are also different even if they are about the similar topic, but current rank functions in this area cannot rank documents based on their qualities. In this paper, we propose a secure multi-keyword ranked search scheme for multiple data owners. A trusted third party is imported to solve the problem of key management. We exploit the vector space model for generating index and query, and our new-designed KDO algorithm is utilized for providing keyword weight, so that the rank function not only considers about the relevance between query and document, but also takes into account the document quality. In order to protect privacy for both owners and users, the Asymmetric Scalar-product Preserving Encryption approach is utilized for encrypting weighted index and query. Besides, we construct the Grouped Balanced Binary tree index, which could further improve the search efficiency by Greedy Depth-first search algorithm. Extensive experiments demonstrate that our proposed scheme is secure, accurate and efficient.

Secure Outsourced Association Rule Mining using Homomorphic Encryption

Several techniques are used in data analysis, where frequent itemset mining and association rule mining are very popular among them. The motivation for 'Data Mining as a Service' (DMaaS) paradigm is that when the data owners are not capable of doing mining tasks internally they have to outsource the mining work to a trusted third party. Multiple data owners can also collaboratively mine by combining their databases. In such cases the privacy of outsourced data is a major issue. Here the context includes necessity of 'corporate privacy' which means other than the data, the result of mining should also preserve privacy requirements. The system proposed uses Advanced Encryption Standard (AES) to encrypt the data items before outsourcing in order to prevent the vulnerability of 'Known Plaintext' attack in the existing system. Fictitious transactions are inserted to the databases using k-anonymity method to counter the frequency analysis attack. A symmetric homomorphic encryption scheme is applied in the databases for performing the mining securely. Based on the experiments and findings, though the running time of proposed solution is slightly greater than the existing system, it provides better security to the data items. Since the computations tasks are performed by the third party server, consumption of resources at the data owners' side is very less.

Supervised machine learning using encrypted training data

Preservation of privacy in data mining and machine learning has emerged as an absolute prerequisite in many practical scenarios, especially when the processing of sensitive data is outsourced to an external third party. Currently, privacy preservation methods are mainly based on randomization and/or perturbation, secure multiparty computations and cryptographic methods. In this paper, we take advantage of the partial homomorphic property of some cryptosystems to train simple machine learning models with encrypted data. Our basic scenario has three parties: multiple Data Owners, which provide encrypted training examples; the Algorithm Owner (or Application), which processes them to adjust the parameters of its models; and a semi-trusted third party, which provides privacy and secure computation services to the Application in some operations not supported by the homomorphic cryptosystem. In particular, we focus on two issues: the use of multiple-key cryptosystems, and the impact of the quantization of real-valued input data required before encryption. In addition, we develop primitives based on the outsourcing of a reduced set of operations that allows to implement general machine learning algorithms using efficient dedicated hardware. As applications, we consider the training of classifiers using privacy-protected data and the tracking of a moving target using encrypted distance measurements.

Data Trading With Multiple Owners, Collectors, and Users: An Iterative Auction Mechanism

In the big data era, it is vital to allocate the vast amount of data to heterogeneous users with different interests. To clinch this goal, various agents including data owners, collectors, and users should cooperate to trade data efficiently. However, the data agents (data owners, collectors, and users) are selfish and seek to maximize their own utilities instead of the overall system efficiency. As such, a sophisticated mechanism is imperative to guide the agents to distribute data efficiently. In this paper, the data trading problem of a data market with multiple data owners, collectors, and users is formulated and an iterative auction mechanism is proposed to coordinate the trading. The proposed mechanism guides the selfish data agents to trade data efficiently in terms of social welfare and avoids direct access of the agents’ private information. We theoretically prove that the proposed mechanism can achieve the socially optimal operation point. Moreover, we demonstrate that the mechanism satisfies appealing economic properties such as individual rationality and weakly balanced budget. Then, we expand the mechanism to nonexclusive data trading, in which the same data can be dispensed to multiple collectors and users. Simulations as well as real data experiments validate the theoretical properties of the mechanism.

VKSE-MO: verifiable keyword search over encrypted data in multi-owner settings

Searchable encryption (SE) techniques allow cloud clients to easily store data and search encrypted data in a privacy-preserving manner, where most of SE schemes treat the cloud server as honest-but-curious. However, in practice, the cloud server is a semi-honest-but-curious third-party, which only executes a fraction of search operations and returns a fraction of false search results to save its computational and bandwidth resources. Thus, it is important to provide a results verification method to guarantee the correctness of the search results. Existing SE schemes allow multiple data owners to upload different records to the cloud server, but these schemes have very high computational and storage overheads when applied in a different but more practical setting where each record is co-owned by multiple data owners. To address this problem, we develop a verifiable keyword search over encrypted data in multi-owner settings (VKSE-MO) scheme by exploiting the multisignatures technique. Thus, our scheme only requires a single index for each record and data users are assured of the correctness of the search results in challenging settings. Our formal security analysis proved that the VKSE-MO scheme is secure against a chosen-keyword attack under a random oracle model. In addition, our empirical study using a real-world dataset demonstrated the efficiency and feasibility of the proposed scheme in practice.

Multi-key privacy-preserving deep learning in cloud computing

Deep learning has attracted a lot of attention and has been applied successfully in many areas such as bioinformatics, imaging processing, game playing and computer security etc. On the other hand, deep learning usually requires a lot of training data which may not be provided by a sole owner. As the volume of data gets huge, it is common for users to store their data in a third-party cloud. Due to the confidentiality of the data, data are usually stored in encrypted form. To apply deep learning to these datasets owned by multiple data owners on cloud, we need to tackle two challenges: (i) the data are encrypted with different keys, all operations including intermediate results must be secure; and (ii) the computational cost and the communication cost of the data owner(s) should be kept minimal. In our work, we propose two schemes to solve the above problems. We first present a basic scheme based on multi-key fully homomorphic encryption (MK-FHE), then we propose an advanced scheme based on a hybrid structure by combining the double decryption mechanism and fully homomorphic encryption (FHE). We also prove that these two multi-key privacy-preserving deep learning schemes over encrypted data are secure.

Dynamic Privacy Pricing: A Multi-Armed Bandit Approach With Time-Variant Rewards

Recently, the conflict between exploiting the value of personal data and protecting individuals’ privacy has attracted much attention. Personal data market provides a promising solution to this conflict, while determining the price of privacy is a tough issue. In this paper, we study the pricing problem in a setting where a data collector sequentially buys data from multiple data owners whose valuations of privacy are randomly drawn from an unknown distribution. To maximize the total payoff, the collector needs to dynamically adjust the prices offered to owners. We model the sequential decision-making problem of the collector as a multi-armed bandit problem with each arm representing a candidate price. Specifically, the privacy protection technique adopted by the collector is taken into account. Protecting privacy generally causes a negative effect on the value of data, and this effect is embodied by the time-variant distributions of the rewards associated with arms. Based on the classic upper confidence bound policy, we propose two learning policies for the bandit problem. The first policy estimates the expected reward of a price by counting how many times the price has been accepted by data owners. The second policy treats the time-variant data value as a context and uses ridge regression to estimate the rewards in different contexts. Simulation results on real-world data demonstrate that by applying the proposed policies, the collector can get a payoff which is close to that he can get by setting a fixed price, which is the best in hindsight, for all data owners.