Enterprise users at different geographic locations generate large-volume data and store their data at different geographic datacenters. These users may also issue ad hoc queries of big data analytics on the  stored data to identify valuable information in order to help them make strategic decisions. However, it is well known that querying such large-volume big data usually is time-consuming and costly. Sometimes, users are only interested in timely approximate rather than exact query results. When this approximation is the case, applications must sacrifice either timeliness or accuracy by allowing either the latency of delivering more accurate results or the accuracy error of delivered results based on the samples of the data, rather than the entire set of data itself. In this paper, we study the QoSaware data replications and placements for approximate query evaluation of big data analytics in a distributed cloud, where the original (source) data of a query is distributed at different geo-distributed datacenters. We focus on placing the samples of the source data with certain error bounds at some strategic datacenters to meet users’ stringent query response time. We propose an efficient algorithm for evaluating a set of big data analytic queries with the aim to minimize the evaluation cost of the queries while meeting their response time requirements. We demonstrate the effectiveness of the proposed algorithm through experimental simulations. Experimental results show that the proposed algorithm is promising.

Given a set of files that show a certain degree of similarity, we consider a novel problem of performing data redundancy elimination across a set of distributed worker nodes in a shared-nothing in-memory big data analytic system. The redundancy elimination scheme is designed in a manner that is: (i) space-efficient: the total space needed to store the files is minimized and, (ii) access-isolation: data shuffling among server is  also minimized. In this paper, we first show that finding an access-efficient and space optimal solution is an NP-Hard problem. Following this, we present the file partitioning algorithms that locate access-efficient solutions in an incremental manner with minimal algorithm time complexity (polynomial time). Our experimental verification on multiple data sets confirms that the proposed file partitioning solution is able to achieve compression ratio close to the optimal compression performance achieved by a centralized solution.

Healthcare has increased its overall value by  adopting big data methods to analyze and understand its data from various sources. This article presents big data from the perspective of improving healthcare services and, also, offers a holistic view of system security and factors determining security breaches.

In big data era, data are usually stored in databases for easy access and utilization, which are now woven into every aspect of our lives. However, traditional relational databases cannot address users’ demands for quick data access and calculating, since they cannot process data in a distributed way. To tackle this problem, non-relational databases such as MongoDB have emerged up and been applied in various Scenarios. Nevertheless, it should be noted that most MongoDB products fail to consider user’s data privacy. In this paper, we propose a practical encrypted MongoDB ( i.e., CryptMDB ). Specifically, we utilize an additive homomorphic asymmetric cryptosystem to encrypt user’s data and achieve strong privacy protection. Security analysis indicates that the CryptMDB can achieve confidentiality of user’s data and prevent adversaries from illegally gaining access to the database. Furthermore, extensive experiments demonstrate that the CryptMDB achieves better efficiency than existing relational database in terms of data access and calculating.

The increased density of base stations (BSs) may significantly add complexity to network management mechanisms and hamper them from efficiently managing the network. In this paper, we propose a big-data-driven network partitioning and optimization framework to reduce the complexity of the networking mechanisms. The proposed framework divides the entire radio access network (RAN) into multiple sub-RANs and each sub-RAN can be managed independently. Therefore, the complexity of the network management can be reduced.

Quantifying the relationships among BSs is challenging in the network partitioning. We propose to extract three networking features from mobile traffic data to discover the relationships. Based on these features, we engineer the network partitioning solution in three steps. First, we design a hierarchical clustering analysis (HCA) algorithm to divide the entire RAN into sub- RANs. Second, we implement a traffic load balancing algorithm to characterize the performance of the network partitioning. Third, we adapt the weights of networking features in the HCA algorithm to optimize the network partitioning. We validate the proposed solution through simulations designed based on real mobile network traffic data. The simulation results reveal the  impacts of the RAN partitioning on the networking performance and the computational complexity of the networking mechanism.

Thanks to recent advance in massive social data and increasingly mature big data mining technologies, information diffusion and its control strategies have attracted much attention, which play pivotal roles in public opinion control, virus marketing as well as other social applications. In this paper, relying on social big data, we focus on the analysis and control of information diffusion. Specifically, we commence with analyzing the topological role of the social strengths, i.e., tie strength, partial strength, value strength, and their corresponding symmetric as well as asymmetric forms. Then, we define two critical points for the cascade information diffusion model, i.e., the information coverage critical point (CCP) and the information heat critical point (HCP). Furthermore, based on the two real-world datasets, the proposed two critical points are verified and analyzed. Our work may be beneficial in terms of analyzing and designing the information diffusion algorithms and relevant control strategies.

In the age of big data, a traditional scanning search pattern is gradually becoming unfit for a satisfying user experience due to its lengthy computing process. In this paper, we propose a sampling-based approximate search framework called Hermes, to meet user’s query demand for both accurate and efficient results. A novel metric, (ε, δ)-approximation, is presented to uniformly measure accuracy and efficiency for a big data search service, which enables Hermes to work out a feasible searching job. Based on this, we employ the bootstrapping technique to further speed up the search process. Moreover, an incremental sampling strategy is investigated to process homogeneous queries; in addition, the reuse theory of historical results is also studied for the scenario of appending data. Theoretical analyses and experiments on a real-world dataset demonstrate that Hermes is capable of producing approximate results meeting the preset query requirements with both high accuracy and efficiency.

As more than 2.5 quintillion bytes of data are generated every day, the era of big data is undoubtedly upon us. Running analysis on extensive datasets is a challenge. Fortunately, a significant percentage of the data accrued can be omitted while maintaining a certain quality of statistical inference in many cases. Censoring provides us a natural option for data reduction. However, the data chosen by censoring occur nonuniformly, which may not relieve the computational resource requirement. In this paper, we propose a dynamic, queuing method to smooth out the data processing without sacrificing the convergence performance of censoring. The proposed method entails simple, closed-form updates, and has no loss in terms of accuracy comparing to the original adaptive censoring method.Simulation results validate its effectiveness.

Bringing Big data technologies into agriculture presents a significant challenge; at the same time, this technology contributes effectively in many countries’ economic and social development. In this work, we will study environmental data provided by precision agriculture information technologies, which represents a crucial source of data in need of being wisely managed and analyzed with appropriate methods and tools in order to extract the meaningful information. Our main purpose through this paper is to propose an effective Big data architecture based on profiling system which can assist (among others) producers, consulting companies, public bodies and research laboratories to make better decisions by providing them real time data processing, and a dynamic big data service composition method, to enhance and monitor the agricultural productivity. Thus, improve their traditional decision making process, and allow better management of the natural resources.

Internet of Things adoption across industries has proven to be beneficial in providing business value by transforming the way data is utilized in decision making and visualization. Power industry has for long struggled with traditional ways of operating and has suffered from issues like instability, blackouts,etc. The move towards smart grid has thus received lot of acceptance. This paper presents the Internet of Things deployment in grid, namely WAMS, and the challenges it present in terms of the Big Data it aggregates. Better insight into the problem is provided with the help of Indian Grid case studies.

Big sensor data is prevalent in both industry and scientific research applications where the data is generated with high volume and velocity it is difficult to process using on-hand database management tools or traditional data processing applications. Cloud computing provides a promising platform to support the addressing of this challenge as it provides a flexible stack of massive computing, storage, and software services in a scalable manner at low cost. Some techniques have been developed in recent years for processing sensor data on cloud, such as sensor-cloud. However, these techniques do not provide efficient support on fast detection and locating of errors in big sensor data sets. For fast data error detection in big sensor data sets, in this paper, we develop a novel data error detection approach which exploits the full computation potential of cloud platform and the network feature of WSN. Firstly, a set of sensor data error types are classified and defined. Based on that classification, the network feature of a clustered WSN is introduced and analyzed to support fast error detection and location. Specifically, in our proposed approach, the error detection is based on the scale-free network topology and most of detection operations can be conducted in limited temporal or spatial data blocks instead of a whole big data set. Hence the detection and location process can be dramatically accelerated. Furthermore, the detection and location tasks can be distributed to cloud platform to fully exploit the computation power and massive storage. Through the experiment on our cloud computing platform of U-Cloud, it is demonstrated that our proposed approach can significantly reduce the time for error detection and location in big data sets generated by large scale sensor network systems with acceptable error detecting accuracy.

Cloud computing offers a new way of service provision by re-arranging various resources over the Internet. The most important and popular cloud service is data storage. In order to preserve the privacy of data holders, data are often stored in cloud in an encrypted form. However, encrypted data introduce new challenges for cloud data deduplication, which becomes crucial for big data storage and processing in cloud. Traditional deduplication schemes cannot work on encrypted data. Existing solutions of encrypted data deduplication suffer from security weakness. They cannot flexibly support data access control and revocation. Therefore, few of them can be readily deployed in practice. In this paper, we propose a scheme to deduplicate encrypted data stored in cloud based on ownership challenge and proxy re-encryption. It integrates cloud data deduplication with access control. We evaluate its performance based on extensive analysis and computer simulations. The results show the superior efficiency and effectiveness of the scheme for potential practical deployment, especially for big data deduplication in cloud storage.

CLOUD computing [1] assembles large networks of virtualized ICT services such as hardware resources (such as CPU, storage, and network), software resources (such as databases, application servers, and web servers) and applications.In industry these services are referred to as infrastructure as a service (IaaS), platform as a service (PaaS), and software as a service (SaaS). Mainstream ICT powerhouses such as Amazon, HP, and IBM are heavily investing in the provision and support of public cloud infrastructure. Cloud computing is rapidly becoming a popular infrastructure of choice among all types of organisations. Despite some initial security concerns and technical issues, an increasing number of organisations have moved their applications and services in to “The Cloud”. These applications range from generic word processing software to online healthcare. The cloud system taps into the processing power of virtualized computers on the back end, thus significantly speeding up the application for the user, which just pays for the used services.

Nowadays, cloud-based storage services are rapidly growing and becoming an emerging trend in data storage field. There are many problems when designing an efficient storage engine for cloud-based systems with some requirements such as big-file processing, lightweight meta-data, low latency, parallel I/O, deduplication, distributed, high scalability. Key-value stores

played an important role and showed many advantages when solving those problems. This paper presents about Big File Cloud (BFC) with its algorithms and architecture to handle most of problems in a big-file cloud storage system based on keyvalue store. It is done by proposing low-complicated, fixed-size meta-data design, which supports fast and highly-concurrent, distributed file I/O, several algorithms for resumable upload, download and simple data deduplication method for static data. This research applied the advantages of ZDB - an in-house keyvalue store which was optimized with auto-increment integer keys for solving big-file storage problems efficiently. The results can be used for building scalable distributed data cloud storage that support big-file with size up to several terabytes

Frequent itemset mining is an exploratory data mining technique that has fruitfully been exploited to extract recurrent co-occurrences between data items. Since in many application contexts items are enriched with weights denoting their relative importance in the analyzed data, pushing item weights into the itemset mining process, i.e., mining weighted itemsets rather than traditional itemsets, is an appealing research direction. Although many efficient in-memory weighted itemset mining algorithms are available in literature, there is a lack of parallel and distributed solutions which are able to scale towards Big Weighted Data. This paper presents a scalable frequent weighted itemset mining algorithm based on the MapReduce paradigm. To demonstrate its actionability and scalability, the proposed algorithm was tested on a real Big dataset collecting approximately 34 millions of reviews of Amazon items. Weights indicate the ratings given by users to the purchased items. The mined itemsets represent combinations of items that were frequently bought together with an overall rating above average.

Users store vast amounts of sensitive data on a big data platform. Sharing sensitive data will help enterprises reduce the cost of providing users with personalized services and provide value-added data services. However, secure data sharing is problematic. This paper proposes a framework for secure sensitive data sharing on a big data platform, including secure data delivery, storage, usage, and destruction on a semi-trusted big data sharing platform. We present a proxy re-encryption algorithm based on heterogeneous ciphertext transformation and a user process protection method based on a virtual machine monitor, which provides support for the realization of system functions. The framework protects the security of users’ sensitive data effectively and shares these data safely. At the same time, data owners retain complete control of their own data in a sound environment for modern Internet information security.

Due to the high volume and velocity of big data, it is an effective option to store big data in the cloud, because the cloud has capabilities of storing big data and processing high volume of user access requests. Attribute-Based Encryption (ABE) is a promising technique to ensure the end-to-end security of big data in the cloud. However, the policy updating has always been a challenging issue when ABE is used to construct access control schemes. A trivial implementation is to let data owners retrieve the data and re-encrypt it under the new access policy, and then send it back to the cloud. This method incurs a high communication overhead and heavy computation burden on data owners. In this paper, we propose a novel scheme that enabling efficient access control with dynamic policy updating for big data in the cloud. We focus on developing an outsourced policy updating method for ABE systems. Our method can avoid the transmission of encrypted data and minimize the computation work of data owners, by making use of the previously encrypted data with old access policies. Moreover, we also design policy updating algorithms for different types of access policies. The analysis show that our scheme is correct, complete, secure and efficient.

The explosive growth of demands on big data processing imposes a heavy burden on computation, storage, and communication in data centers, which hence incurs considerable operational expenditure to data center providers. Therefore, cost minimization has become an emergent issue for the upcoming big data era. Different from conventional cloud services, one of the main features of big data services is the tight coupling between data and computation as computation tasks can be conducted only when the corresponding data is available. As a result, three factors, i.e., task assignment, data placement and data movement, deeply influence the operational expenditure of data centers. In this paper, we are motivated to study the cost minimization problem via a joint
optimization of these three factors for big data services in geo-distributed data centers. To describe the task completion time with the consideration of both data transmission and computation, we propose a two-dimensional Markov chain and derive the average task completion time in closed-form. Furthermore, we model the problem as a mixed-integer non-linear programming (MINLP) and propose an efficient solution to linearize it. The high efficiency of our proposal is validated by extensive simulation based studies.

Spurred by service computing and cloud computing, an increasing number of services are emerging on the Internet. As a result, service-relevant data become too big to be effectively processed by traditional approaches. In view of this challenge, a Clustering-based Collaborative Filtering approach (ClubCF) is proposed in this paper, which aims at recruiting similar services in the same clusters to recommend services collaboratively. Technically, this approach is enacted around two stages. In the first stage, the available services are divided into small-scale clusters, in logic, for further processing. At the second stage, a collaborative filtering algorithm is imposed on one of the clusters. Since the number of the services in a cluster is much less than the total number of the services available on the web, it is expected to reduce the online execution time of collaborative filtering. At last, several experiments are conducted to verify the availability of the approach, on a real dataset of 6,225 mashup services collected from ProgrammableWeb.

A framework for maintaining security & preserving privacy for analysis of sensor data from smart homes, without compromising on data utility is presented. Storing the personally identifiable data as hashed values withholds identifiable information from any computing nodes. However the very nature of smart home data analytics is establishing preventive care. Data processing results should be identifiable to certain users responsible for direct care. Through a separate encrypted identifier dictionary with hashed and actual values of all unique sets of identifiers, we suggest re-identification of any data processing results. However the level of re-identification needs to be controlled, depending on the type of user accessing the results. Generalization and suppression on identifiers from the identifier dictionary before re-introduction could achieve different levels of privacy preservation. In this paper we propose an approach to achieve data security & privacy through out the complete data lifecycle:data generation/collection, transfer, storage, processing and sharing.