Search

The vast majority of artificial intelligence solutions are founded on game theory, and differential privacy is emerging as perhaps the most rigorous and widely adopted privacy paradigm in the field. However, alongside all the advancements made in both these fields, there is not a single application that is not still vulnerable to privacy violations, security breaches, or manipulation by adversaries. Our understanding of the interactions between differential privacy and game theoretic solutions is limited. Hence, we undertook a comprehensive review of literature in the field, finding that differential privacy has several advantageous properties that can make more of a contribution to game theory than just privacy protection. It can also be used to build heuristic models for game-theoretic solutions, to avert strategic manipulations, and to quantify the cost of privacy protection. With a focus on mechanism design, the aim of this article is to provide a new perspective on the currently held impossibilities in game theory, potential avenues to circumvent those impossibilities, and opportunities to improve the performance of game-theoretic solutions with differentially private techniques.

IEEE In the context of Information Societies, a tremendous amount of information is daily exchanged or released. Among various information-release cases, medical document release has gained significant attention for its potential in improving healthcare service quality and efficacy. However, integrity and origin authentication of released medical documents is the priority in subsequent applications. Moreover, sensitive nature of much of this information also gives rise to a serious privacy threat when medical documents are uncontrollably made available to untrusted third parties. Redactable signatures allow any party to delete pieces of an authenticated document while guaranteeing the origin and integrity authentication of the resulting (released) subdocument. Nevertheless, most of existing redactable signature schemes (RSSs) are vulnerable to dishonest redactors or illegal redaction detection. To address the above issues, we propose two distinct RSSs with flexible release control (RSSs-FRC). We also analyse the performance of our constructions in terms of security, efficiency and functionality. The analysis results show that the performance of our construction has significant advantages over others, from the aspects of security and efficiency.

Privacy leakage becomes increasingly serious because massive volumes of data are constantly shared in diverse booming cyber-physical social networks (CPSN). Differential privacy is one of the dominating privacy-preserving methods, but most of its extensions assume all data users share the same privacy requirement, which fails to satisfy various privacy expectations in practice. To address this issue, customizable privacy preservation based on differential privacy is a potentially promising countermeasure. However, we found that customizable protection will trigger the composition mechanism of differential privacy and leads to unexpected correlations among injected noises that weakens privacy protection and reveal more sensitive inforamtion. As a result, customizable privacy protection is vulnerable to two primary attacks: background knowledge attack and collusion attack. To optimize the tradeoff between customizable privacy preservation and data utility, we propose a customizable reliable differential privacy model (CRDP), which provides customizable protection on each individual while being attack-proof. We define social distance as the shortest path between two nodes, which is used as an index to customize the privacy protection levels, followed by quantitatively modeling the attacks under the framework of differential privacy. We develop a modified Laplacian mechanism in which the noise generation complies with a Markov stochastic process.Consequently, the correlations of noises are properly de-coupled so that CRDP can simultaneously minimize background knowledge attacks and eliminate collusion attacks in this particular scenario. The evaluation results from real-world datasets show the feasibility and superiority of CRDP in terms of customizable privacy preservation and reliable attack resistance.

The cyber-physical social system (CPSS) enables human social interaction from cyberspace to the physical world by sharing an increasing volume of spatial-temporal data. The sensitive information in the shared data is appealing to the adversaries and various attacks. However, most existing research assumes that the privacy protection level is identical regardless of various requirements, which is not practical. Subsequently, this results in either over-protection or failing to resist the leading attacks like collusion attacks. Motivated by this, a personalized model is proposed by using generative adversarial nets (GAN) to achieve differential privacy and thereby enhancing spatial-temporal private data sharing. A Differential Privacy Identifier is added to the classic GAN with a Generator and a Discriminator . The deployment of the differentially Private GAN (P-GAN) enables the generation of the sanitized data that can perfectly approximate the spatial-temporal trajectory while providing high-level privacy protection. P-GAN optimizes the trade-off between personalized privacy protection and improved data utility while maintaining fast convergence. Simultaneously, the random noise generation guaranteed by GAN breaks the correlation of injected noises and make P-GAN attack-proof against the collusion attacks. Extensive evaluation results on real-world datasets show the superiority of P-GAN from the aspects of optimized trade-off and efficiency.

Privacy preserving in machine learning is a crucial issue in industry informatics since data used for training in industries usually contain sensitive information. Existing differentially private machine learning algorithms have not considered the impact of data correlation, which may lead to more privacy leakage than expected in industrial applications. For example, data collected for traffic monitoring may contain some correlated records due to temporal correlation or user correlation. To fill this gap, we propose a correlation reduction scheme with differentially private feature selection considering the issue of privacy loss when data have correlation in machine learning tasks. %The key to the proposed scheme is to describe the data correlation and select features which leads to less data correlation across the whole dataset. The proposed scheme involves five steps with the goal of managing the extent of data correlation, preserving the privacy, and supporting accuracy in the prediction results. In this way, the impact of data correlation is relieved with the proposed feature selection scheme, and moreover, the privacy issue of data correlation in learning is guaranteed. The proposed method can be widely used in machine learning algorithms which provide services in industrial areas. Experiments show that the proposed scheme can produce better prediction results with machine learning tasks and fewer mean square errors for data queries compared to existing schemes., Comment: This paper has been published in IEEE Transactions on Industrial Informatics

In location-sensitive applications, dishonest users may submit fake location claims to illegally access a service or obtain benefit. To address this issue, a number of location proof mechanisms have been proposed in literature. However, they confront various security and privacy challenges, including Prover–Prover collusions (Terrorist Frauds), Prover–Witness collusions, and location privacy threats. In this paper, we utilize the unique features of the blockchain technology to design a decentralized scheme for location proof generation and verification. In the proposed scheme, a user who needs a location proof (called a prover) broadcasts a request to the neighbor devices through a short-range communication interface, e.g. Bluetooth. Those neighbor devices that decide to respond (called witnesses) start to authenticate the requesting user. We integrate an incentive mechanism into the proposed scheme to reward such witnesses. Upon successful authentication, a transaction is generated as a location proof and is broadcast onto a peer-to-peer network where it can be picked up by verifiers for final verification. Our security analysis shows that the proposed scheme achieves a reliable performance against Prover–Prover and Prover–Witness collusions. Moreover, our prototype implementation on the Android platform shows that the proposed scheme outperforms other currently deployed location proof schemes.

Labeling malware or malware clustering is important for identifying new security threats, triaging and building reference datasets. The state-of-the-art Android malware clustering approaches rely heavily on the raw labels from commercial AntiVirus (AV) vendors, which causes misclustering for a substantial number of weakly-labeled malware due to the inconsistent, incomplete and overly generic labels reported by these closed-source AV engines, whose capabilities vary greatly and whose internal mechanisms are opaque (i.e., intermediate detection results are unavailable for clustering). The raw labels are thus often used as the only important source of information for clustering. To address the limitations of the existing approaches, this paper presents Andre, a new ANDroid Hybrid REpresentation Learning approach to clustering weakly-labeled Android malware by preserving heterogeneous information from multiple sources (including the results of static code analysis, the meta-information of an app, and the raw-labels of the AV vendors) to jointly learn a hybrid representation for accurate clustering. The learned representation is then fed into our outlier-aware clustering to partition the weakly-labeled malware into known and unknown families. The malware whose malicious behaviours are close to those of the existing families on the network, are further classified using a three-layer Deep Neural Network (DNN). The unknown malware are clustered using a standard density-based clustering algorithm. We have evaluated our approach using 5,416 ground-truth malware from Drebin and 9,000 malware from VirusShare (uploaded between Mar. 2017 and Feb. 2018), consisting of 3324 weakly-labeled malware. The evaluation shows that Andre effectively clusters weakly-labeled malware which cannot be clustered by the state-of-the-art approaches, while achieving comparable accuracy with those approaches for clustering ground-truth samples.

Agent advising is one of the main approaches to improve agent learning performance by enabling agents to share advice. Existing advising methods have a common limitation that an adviser agent can offer advice to an advisee agent only if the advice is created in the same state as the advisee's concerned state. However, in complex environments, it is a very strong requirement that two states are the same, because a state may consist of multiple dimensions and two states being the same means that all these dimensions in the two states are correspondingly identical. Therefore, this requirement may limit the applicability of existing advising methods to complex environments. In this article, inspired by the differential privacy scheme, we propose a differential advising method that relaxes this requirement by enabling agents to use advice in a state even if the advice is created in a slightly different state. Compared with the existing methods, agents using the proposed method have more opportunity to take advice from others. This article is the first to adopt the concept of differential privacy on advising to improve agent learning performance instead of addressing security issues. The experimental results demonstrate that the proposed method is more efficient in complex environments than the existing methods.

With the increasing development of cloud computing, a number of users choose to outsource their data to a remote database service provider and enjoy flexible data sharing with multiple parties. However, the integrity and confidentiality of outsourced data in a remote cloud server are the main threats users are concerned about. The tree structure is one of the most widely used data organization structures. It is crucial to guarantee the integrity and privacy not only for the content but also for the structure if sensitive information is organized in this structure. At present, despite some solutions have been put forward, none of these considers the additional redaction attack on tree-structured data from attackers while sharing data with others. In this article, we provide a construction of secure redactable signature scheme for tree-structured data which features a multi-party control on the redaction of vertexes in a tree while the authenticity and privacy are assured. Then, we prove that our scheme is unforgeable, private, and transparent. Furthermore, extensive theoretical and experimental analyses are conducted to assess the efficiency of our scheme. The results demonstrate our scheme achieves multi-party redaction control without sacrificing sensible resources especially when a tree has a larger number of vertexes and siblings per vertex.

With the support of portable electronic devices and crowdsensing, a new class of mobile applications based on the Internet of Things (IoT) application is emerging. Crowdsensing enables workers with mobile devices to travel to specified locations and collect data, then send it back to the requester for rewards. However, the majority of the existing crowdsensing systems are based on centralized servers, which are prone to a high chance of attack, intrusion, and manipulation. Further, during the process of transmitting information to and from the service server, the worker’s location is usually exposed. This raises the potential risk of a privacy infringement. In this paper, we first identify three ways locations can be disclosed in traditional crowdsensing systems. Then, we propose a novel solution, dubbed a blockchain privacy-preservation crowdsensing system, to address these privacy problems. The proposed system not only protects the privacy of worker locations but also increases the success rate of completing the assigned task. Specifically, the system entails a rewards-based task assignment process that, essentially, markets the given assignment and uses the anonymized characteristics of blockchain technology to hide the identity information of users. To prevent attacks through re-identification, we have introduced a private blockchain to distribute the worker’s transaction records.

Searching on encrypted data has become a very important technique in cloud computing. Such searches enable the data owner to search on the encrypted data stored on the cloud without leaking any information. To obtain a better search experience, researchers have proposed many schemes which mainly focus on conjunctive and disjunctive keyword searches. However, conjunction of all the keywords may result in very few results, whereas a disjunction will return too many results. With the current schemes, customizing the relevancy of the keywords to obtain the desired results is difficult. To solve these problems, we propose a novel scheme that supports the search with the user-specified number of keywords contained in the search result. This number $n$ can be used to customize the keyword relevancy. As a result, the data owner could obtain the desired search results containing any $n$ keywords of a keyword set. The proposed scheme also supports the traditional disjunctive and conjunctive keyword searches when $n$ equals 1 or the size of the keyword set, respectively. The keyword could be positive or negative. We first formally define its security and then prove that the proposed scheme is secure against the adaptively chosen keyword attack in the standard model and can defend against the offline keyword guessing attack to some extent. Furthermore, we present a theoretical performance comparison with other schemes as well as the experimental performance evaluations on our implemented scheme.

IEEE Advanced persistent threat (APT) for cyber espionage poses a great threat to modern organizations. In order to mitigate the impact of APT on an organization, all the compromised systems in the organization must be quarantined and recovered in a timely and effective way. This article focuses on the problem of customizing a dynamic quarantine and recovery (QAR) scheme for an organization so that the APT impact is minimized. Based on a novel node-level epidemic model characterizing the effect of the QAR scheme on the expected state of the underlying network, we estimate the expected impact of APT under a QAR scheme. On this basis, we model the original problem as an optimal control problem. By use of optimal control theory, we derive the optimality system for the optimal control problem and thereby introduce the concept of normal potential optimal (NPO) control. Next, through comparative experiments, we find that the NPO control outperforms a set of heuristic controls. Hence, the QAR scheme associated with the NPO control is satisfactory in terms of the effectiveness of defending against APT. Finally, we examine the effect of some factors on the expected APT impact under the NPO control. This article would be helpful to the defense against APT for cyber espionage.

Differential privacy is a rigorous mathematical framework for evaluating and protecting data privacy. In most existing studies, there is a vulnerable assumption that records in a dataset are independent when differential privacy is applied. However, in real-world datasets, records are likely to be correlated, which may lead to unexpected data leakage. In this survey, we investigate the issue of privacy loss due to data correlation under differential privacy models. Roughly, we classify existing literature into three lines: 1) using parameters to describe data correlation in differential privacy, 2) using models to describe data correlation in differential privacy, and 3) describing data correlation based on the framework of Pufferfish. Firstly, a detailed example is given to illustrate the issue of privacy leakage on correlated data in real scenes. Then our main work is to analyze and compare these methods, and evaluate situations that these diverse studies are applied. Finally, we propose some future challenges on correlated differential privacy., Comment: This paper is accepted in Concurrency and Computation: Practice and Experience

Federated learning is an improved version of distributed machine learning that further offloads operations which would usually be performed by a central server. The server becomes more like an assistant coordinating clients to work together rather than micro-managing the workforce as in traditional DML. One of the greatest advantages of federated learning is the additional privacy and security guarantees it affords. Federated learning architecture relies on smart devices, such as smartphones and IoT sensors, that collect and process their own data, so sensitive information never has to leave the client device. Rather, clients train a sub-model locally and send an encrypted update to the central server for aggregation into the global model. These strong privacy guarantees make federated learning an attractive choice in a world where data breaches and information theft are common and serious threats. This survey outlines the landscape and latest developments in data privacy and security for federated learning. We identify the different mechanisms used to provide privacy and security, such as differential privacy, secure multi-party computation and secure aggregation. We also survey the current attack models, identifying the areas of vulnerability and the strategies adversaries use to penetrate federated systems. The survey concludes with a discussion on the open challenges and potential directions of future work in this increasingly popular learning paradigm., 21 pages, 4 figures, 4 tables, 76 references

© 2020, Springer-Verlag London Ltd., part of Springer Nature. Cooperation is an essential behavior in multi-agent systems. Existing mechanisms have two common drawbacks. The first drawback is that malicious agents are not taken into account. Due to the diverse roles in the evolution of cooperation, malicious agents can exist in multi-agent systems, and they can easily degrade the level of cooperation by interfering with agent’s actions. The second drawback is that most existing mechanisms have a limited ability to fit in different environments, such as different types of social networks. The performance of existing mechanisms heavily depends on some factors, such as network structures and the initial proportion of cooperators. To solve these two drawbacks, we propose a novel mechanism which adopts differential privacy mechanisms and reinforcement learning. Differential privacy mechanisms can be used to relieve the impact of malicious agents by exploiting the property of randomization. Reinforcement learning enables agents to learn how to make decisions in various social networks. In this way, the proposed mechanism can promote the evolution of cooperation in malicious social networks.

Collusion attacks are considered to be challenging attacks in audio copyright protection. The traditional watermarking algorithms cannot identify the traitors when other attacks, such as desynchronization attacks, are applied with a collusion attack. Instead of tracing the traitors, in this paper we aim to tackle collusion attacks by removing the commercial value from the colluded copy, which will demotivate the attackers from launching collusion attacks. Since the commercial value of an audio signal is directly reflected by its perceptual quality, we propose a novel non-linear-echo generation (NLEG) based algorithm to significantly degrade the perceptual quality of the colluded copy by embedding a time delay sequence into the host signal. The proposed NLEG is also designed to be resilient to common signal processing attacks and desynchronization attacks. Furthermore, the proposed NLEG can be combined with other digital watermarking techniques to enhance its performance on protecting the copyright information. Experimental results show the validity of the proposed NLEG.

A growing specter in the rise of machine learning is whether the decisions made by machine learning models are fair. While research is already underway to formalize a machine-learning concept of fairness and to design frameworks for building fair models with sacrifice in accuracy, most are geared toward either supervised or unsupervised learning. Yet two observations inspired us to wonder whether semi-supervised learning might be useful to solve discrimination problems. First, previous study showed that increasing the size of the training set may lead to a better trade-off between fairness and accuracy. Second, the most powerful models today require an enormous of data to train which, in practical terms, is likely possible from a combination of labeled and unlabeled data. Hence, in this paper, we present a framework of fair semi-supervised learning in the pre-processing phase, including pseudo labeling to predict labels for unlabeled data, a re-sampling method to obtain multiple fair datasets and lastly, ensemble learning to improve accuracy and decrease discrimination. A theoretical decomposition analysis of bias, variance and noise highlights the different sources of discrimination and the impact they have on fairness in semi-supervised learning. A set of experiments on real-world and synthetic datasets show that our method is able to use unlabeled data to achieve a better trade-off between accuracy and discrimination., Comment: This paper has been published in IEEE Transactions on Knowledge and Data Engineering

© 2018, Springer Nature B.V. A big challenge in educational resources construction is the intelligent and personalized resource recommendation for learners. This paper proposes a semantic recommendation framework of educational resources based on semantic web and pedagogics. In this framework, a domain ontology is constructed to describe the knowledge structure of the domain. All the resources and user portfolio are described with ontology technology and resource description framework to support semantic inference. Based on the semantic resource organization, we made a set of reasoning rules based on pedagogics. These rules are made from the synthesis of the type of the knowledge, the internal structure of knowledge and learner’s learning performance. A case study was implemented on the course “theory and practice of database”. In this case, learners are recommended different learning materials according to the different knowledge structure and different learning performance. Three typical learning modes are proposed to describe the personalized learning experience. This framework can be used as a guide for teachers and resource designers.

© 2004-2012 IEEE. Accurate malware propagation modeling in wireless ad hoc networks (WANETs) represents a fundamental and open research issue which shows distinguished challenges due to complicated access competition, severe channel interference, and dynamic connectivity. As an effort towards the issue, in this paper, we investigate the malware propagation under two spread schemes including Unicast and Broadcast, in Spread Mode and Communication Mode, respectively. We highlight our contributions in three-fold in the light of previous literature works. First, a bound of malware infection rate for each scheme is provided by applying the wireless network capacity theories. Second, the impact of mobility on malware propagations has been studied. Third, discussion of the relationship between different schemes and practical applications is provided. Numerical simulations and detailed performance analysis show that the Broadcast Scheme with Spread Mode is most dangerous in the sense of malware propagation speed in WANETs, and mobility will greatly increase the risk further. The results achieved in this paper not only provide insights on the malware propagation characteristics in WANETs, but also serve as fundamental guidelines on designing defense schemes.

Privacy preserving data release is a hot topic that attracts a lot of attentions in data mining, machine learning, and social network communities. Most studies on privacy preserving focus on static data releases; however, data are usually updated periodically. As a potential solution, differential privacy addresses continual data release by simplifying it into an event stream release problem. This approach overlooks the relationship between events, which is defined as coupled information in this paper. We argue that datasets cannot be simplified as an event stream due to the coupled information. In addition, the coupled information may reveal more private information than expected. This work proposes a privacy-preserving mechanism that explicitly identify the coupled information in continually released datasets. In stead of simplifying datasets to event streams, this mechanism considers the continual released datasets as coupled datasets based on the relationship between the same individual in different datasets, and the relationship between different individuals in the same dataset. We also propose the notion of coupled sensitivity for answering differentially private queries and develop an iterative based coupled continual release algorithm, called CCR, that answers these queries with a large set of differentially private results. Theoretical analysis proves the privacy of this method, and an extensive performance study shows that CCR outperforms traditional differential privacy mechanisms when answering a large set of queries.

The multiple relationships among objects in complex systems can be described well by multiplex networks, which contain rich information of the connections between objects. The null model of networks, which can be used to quantify the specific nature of a network, is a powerful tool for analysing the structural characteristics of complex systems. However, the null model for multiplex networks remains largely unexplored. In this paper, we propose a null model for multiplex networks based on the node redundancy degree, which is a natural measure for describing the multiple relationships in multiplex networks. Based on this model, we define the modularity of multiplex networks to study the community structures in multiplex networks and demonstrate our theory in practice through community detection in four real-world networks. The results show that our model can reveal the community structures in multiplex networks and indicate that our null model is a useful approach for providing new insights into the specific nature of multiplex networks, which are difficult to quantify.

© 2018 IEEE. Data aggregation plays an important role in the Internet of Things, and its study and analysis has resulted in a range of innovative services and benefits for people. However, the privacy issues associated with raw sensory data raise significant concerns due to the sensitive nature of the user information it often contains. Thus, numerous schemes have been proposed over the last few decades to preserve the privacy of users' data. Most methods are based on encryption technology, which is computationally and communicationally expensive. In addition, most methods can only handle a single aggregation function. Therefore, in this paper, we propose a multifunctional data aggregation method with differential privacy. The method is based on machine learning and can support a wide range of statistical aggregation functions, including additive and non-additive aggregation. It operates within a fog computing architecture, which extends cloud computing to the edge of the network, alleviating much of the computational burden on the cloud server. And, by only reporting the results of the aggregation to the server, communication efficiency is improved. Extensive experimental results show that the proposed method not only answers flexible aggregation queries that meet diversified aggregation goals, but also produces aggregation results with high accuracy.

The location privacy issue has become a critical research topic recently. The existing solutions do not solve one typical problem in practice: people may only want to protect certain privacy-sensitive locations among a group of temporal and spatial correlated points in a trajectory. As an effort towards this issue, we analyze the impact of space-time relationship on location privacy preservation. In addition, we propose new privacy definitions to better evaluate the privacy level and prove that the target location's privacy can be enhanced by randomizing its time and space related points. Moreover, under the constraint of the total noise power, the problem of obfuscating a location in a temporal and spatial correlated trajectory is formulated as finding the best noise allocation vector which can achieve the highest privacy level. This problem is solved by our proposed location privacy preserving method which applies differential privacy scheme on a series of points with noise budget allocation. Lastly, the performance of the proposed scheme is evaluated by simulations.

© 2019 Association for Computing Machinery. Human activity recognition plays a key role in the application areas such as fitness tracking, healthcare and aged care support. However, inaccurate recognition results may cause an adverse effect on users or even an unpredictable accident. In order to improve the accuracy of human activity recognition, multi-device and deep learning based approaches have been proposed. However, they are not practical on a daily basis due to the limitations that devices are difficult to wear, and deep learning requires large training dataset and incurs expensive computational costs. To address this problem, we propose a novel approach, multiscale feature engineering for human activity recognition (MFE-HAR), which exploits the properties of arm movement from global and local scales using the accelerometer and gyroscope sensors on a single wearable device. Our method takes advantage of having important features at multiple scales over previous single-scale methods. We evaluated the performance of the proposed method on two public datasets and achieved the mean classification accuracy of 93% and 98% respectively. Our proposed system performs better than the state of the art multi-device based approaches, and is more practical for real-world applications.

Agent advising is one of the key approaches to improve agent learning performance by enabling agents to ask for advice between each other. Existing agent advising approaches have two limitations. The first limitation is that all the agents in a system are assumed to be friendly and cooperative. However, in the real world, malicious agents may exist and provide false advice to hinder the learning performance of other agents. The second limitation is that the analysis of communication overhead in these approaches is either overlooked or simplified. However, in communication-constrained environments, communication overhead has to be carefully considered. To overcome the two limitations, this paper proposes a novel differentially private agent advising approach. Our approach employs the Laplace mechanism to add noise on the rewards used by student agents to select teacher agents. By using the differential privacy technique, the proposed approach can reduce the impact of malicious agents without identifying them. Also, by adopting the privacy budget concept, the proposed approach can naturally control communication overhead. The experimental results demonstrate the effectiveness of the proposed approach.

© Springer Nature Switzerland AG 2019. This chapter provides some primary knowledge about identifying propagation sources of malicious attacks. We first introduce different types of observations about the propagation of malicious attacks. Then, we present the maximum-likelihood estimation method adopted by many approaches in this research area. We finally introduce the evaluation metrics for source identification.

Biomedical data are widely accepted in developing prediction models for identifying a specific tumor, drug discovery and classification of human cancers. However, previous studies usually focused on different classifiers, and overlook the class imbalance problem in real-world biomedical datasets. There are a lack of studies on evaluation of data pre-processing techniques, such as resampling and feature selection, on imbalanced biomedical data learning. The relationship between data pre-processing techniques and the data distributions has never been analysed in previous studies. This article mainly focuses on reviewing and evaluating some popular and recently developed resampling and feature selection methods for class imbalance learning. We analyse the effectiveness of each technique from data distribution perspective. Extensive experiments have been done based on five classifiers, four performance measures, eight learning techniques across twenty real-world datasets. Experimental results show that: (1) resampling and feature selection techniques exhibit better performance using support vector machine (SVM) classifier. However, resampling and Feature Selection techniques perform poorly when using C4.5 decision tree and Linear discriminant analysis classifiers; (2) for datasets with different distributions, techniques such as Random undersampling and Feature Selection perform better than other data pre-processing methods with T Location-Scale distribution when using SVM and KNN (K-nearest neighbours) classifiers. Random oversampling outperforms other methods on Negative Binomial distribution using Random Forest classifier with lower level of imbalance ratio; (3) Feature Selection outperforms other data pre-processing methods in most cases, thus, Feature Selection with SVM classifier is the best choice for imbalanced biomedical data learning., Comment: This paper is scheduled for inclusion in V16 N3 2020, International Journal of Bioinformatics Research and Applications (IJBRA)

Different from simplifying trajectories using turning points to keep their coarse shapes, this paper summarizes trajectories using stop points.We define a new concept of stay stability (i.e., reciprocal of speed) between any two GPS points to detect stop points on individual trajectories.We discover repeat travel behaviors by finding common sequences of stop regions where a certain number of objects visit with similar stop duration.The experiments on 20 labeled trajectories in Geolife demonstrated the semantic effect, accuracy and near linear efficiency of our method. GPS trajectory dataset with high sampling-rates is usually in large volume that challenges the processing efficiency. Most of the data points on trajectories are useless. This paper summarizes trajectories using stop points. We define a new concept of stay stability (i.e., time dividing distance or reciprocal of speed) between any two GPS points to detect stop points on individual trajectories. We propose a novel Mining Repeat Travel Behaviors Using Stop Regions (MRTBUSR) method. In MRTBUSR, a stop region is a popular region containing a certain number of close stop points that can be grouped into a cluster. We then retrieve common sequences of stop regions to denote repeat route patterns and further analyze the stop durations on a stop region to find repeat travel behaviors. The experiments on 20 labeled trajectories selected from GeoLife demonstrated the semantic effect, accuracy and near linear efficiency of our proposed method.

© 2018 John Wiley & Sons, Ltd. Deep learning methods have become the basis of new AI-based services on the Internet in big data era because of their unprecedented accuracy. Meanwhile, it raises obvious privacy issues. The deep learning–assisted privacy attack can extract sensitive personal information not only from the text but also from unstructured data such as images and videos. In this paper, we proposed a framework to protect image privacy against deep learning tools, along with two new metrics that measure image privacy. Moreover, we propose two different image privacy protection schemes based on the two metrics, utilizing the adversarial example idea. The performance of our solution is validated by simulations on two different datasets. Our research shows that we can protect the image privacy by adding a small amount of noise that has a humanly imperceptible impact on the image quality, especially for images of complex structures and textures.

Digital signatures, with the properties of data integrity and authenticity authentication, protect a signed message from any alteration. However, appropriate alteration of signed message should be allowed for the purposes of privacy protection in some scenarios, such as medical data sharing, outsourced databases, etc. Redactable signatures, a branch of homomorphic signatures for editing, allow any party to delete some submessage blocks from a signed message and generate a valid signature on the remaining message without any help of the original signer. This chapter provides a basic introduction on the state-of-the-art redactable signature schemes. We mainly consider the redaction control problem of redactable signature schemes in different applications. We also present three integrated solutions, which hopefully offer more insights into this crucial problem.

© 2018 IEEE. Fast proliferation of mobile devices significantly promotes the development of mobile social networks. Users tend to interact with friends via multiple social networks. Multiple social networks identification is of great significance in terms of both attack and defense. Current methods either focus on the profile matching or network structure to re-identify a specific user. However, the accuracy are not satisfying with relative high error rate. In this paper, we propose a new Friendship learning-Based Identification (FBI) method to discriminate multiple pseudo identities of a real-world individual. We aim at providing potential attack mechanism to following privacy protection research. Firstly, we develop a new identification method based on friendship matching. Then, we implement a weighted mechanism which takes profile, network structure, and friendship into consideration. Furthermore, machine learning is leverage to further optimize the parameters and improve the accuracy. In addition, extensive experimental results show the superior of the FBI comparing to existing ones.

© 2002-2012 IEEE. The proliferation of wireless devices and appliances is facilitating the rapid development of the Internet of Things (IoT). Numerous state-of-the-art applications are being used in, for example, smart cities, autonomous vehicles, and biocomputing. With the popularization of IoT, new challenges are emerging with respect to privacy issues. In this article, we first summarize privacy constraints and primary attacks based on new features of IoT. Then we present three case studies to demonstrate principal vulnerabilities and classify existing protection schemes. Built on this analysis, we identify three key challenges: a lack of theoretical foundation, the trade-off optimization between privacy and data utility, and system isomerism over-complexity and high scalability. Finally, we illustrate possible promising future directions and potential solutions to the emerging challenges facing wireless IoT scenarios. We aim to assist interested readers in investigating the unexplored parts of this promising domain.

Networks portray a multitude of interactions through which people meet, ideas are spread, and infectious diseases and malicious rumors propagate within a society. Recently, researchers have found that unsolicited malicious attacks spread extremely fast through influential spreaders. For example, in April 23, 2013, the twitter account of Associated Press was hacked to spread the rumor that explosions at White House injured Obama. This led to both the DOW Jones industrial average and Standard & Poor’s 500 Index plunging about 1% before regaining their losses. Hence, identifying the most efficient ‘spreaders’ in a network becomes an important step towards restraining spread of malicious attacks. In this chapter, we investigate the methods of measuring influence of network nodes.

© Springer Nature Switzerland AG 2019. The global diffusion of epidemics, computer viruses and rumors causes great damage to our society. One critical issue to identify the multiple diffusion sources so as to timely quarantine them. However, most methods proposed so far are unsuitable for diffusion with multiple sources because of the high computational cost and the complex spatiotemporal diffusion processes. In this chapter, we introduce an effective method to identify multiple diffusion sources, which can address three main issues in this area: (1) How many sources are there? (2) Where did the diffusion emerge? (3) When did the diffusion break out? For simplicity, we use rumor source identification to present the approach.

© Springer Nature Switzerland AG 2019. The global diffusion of epidemics, rumors and computer viruses causes great damage to our society. It is critical to identify the diffusion sources and promptly quarantine them. However, one critical issue of current methods is that they are far are unsuitable for large-scale networks due to the computational cost and the complex spatiotemporal diffusion processes. In this chapter, we introduce a community structure based approach to efficiently identify diffusion sources in large networks.

Restraining the propagation of malicious attacks in complex networks has long been an important but difficult problem to be addressed. In this chapter, we particularly use rumor propagation as an example to analyze the methods of restraining malicious attack propagation. There are mainly two types of methods: (1) blocking rumors at the most influential users or community bridges, and (2) spreading truths to clarify the rumors. We first compare all the measures of locating influential users. The results suggest that the degree and betweenness measures outperform all the others in real-world networks. Secondly, we analyze the method of the truth clarification method, and find that this method has a long-term performance while the degree measure performs well only in the early stage. Thirdly, in order to leverage these two methods, we further explore the strategy of different methods working together and their equivalence. Given a fixed budget in the real world, our analysis provides a potential solution to find out a better strategy by integrating both kinds of methods together.

This chapter provides an extensive literature review on identifying the propagation source of malicious attacks by tracing research trends and hierarchically reviewing the contributions along each research line regarding identifying the propagation source of malicious attacks. This chapter consists of three parts. We first review the existing approaches and analyze their pros and cons. Then, numerical studies are provided according to various experiment settings and diffusion scenarios. Finally, we summarize the remarks of existing approaches. Here, we particularly use rumor propagation as an example to analyze these approaches.

© 2018 IEEE. Road safety is a significant issue in any intelligent transportation system (ITS). Intersections are the most complex part of the road network, as they involve various participants, such as vehicles and pedestrians. Therefore, providing an accurate traffic flow prediction model will enhance traffic efficiency and reduce common problems, such as accidents, congestion and air pollution. However, there are two challenges in the traffic flow prediction problem: first, traffic is a dynamic nonlinear problem due to nonrecurrent events, such as accidents and roadworks, that occur near intersections as an unexpected event will impact the accuracy of the prediction method. The second challenge is that there is a large amount of data which needs a scalable model to efficiently handle big data. To overcome the first issue, in this study, accidents and roadworks data are used, in addition to sensor data that are updated in real time. The datasets are published by VicRoads for the state of Victoria, Australia. Moreover, ensemble decision trees for regression, namely the gradient boosting regression trees (GBRT) and random forest (RF), are proposed. To address the second challenge, the extreme gradient boosting Tree (XGBoost) algorithm, which is a scalable system, is examined to explore its ability to handle traffic data. Finally, a comparative analysis of the proposed methods in terms of time and accuracy is presented.

© 2014 IEEE. The rapid proliferation of smart mobile devices has significantly enhanced the popularization of the cyber-physical social network, where users actively publish data with sensitive information. Adversaries can easily obtain these data and launch continuous attacks to breach privacy. However, existing works only focus on either location privacy or identity privacy with a static adversary. This results in privacy leakage and possible further damage. Motivated by this, we propose a hybrid privacy-preserving scheme, which considers both location and identity privacy against a dynamic adversary. We study the privacy protection problem as the tradeoff between the users aiming at maximizing data utility with high-level privacy protection while adversaries possessing the opposite goal. We first establish a game-based Markov decision process model, in which the user and the adversary are regarded as two players in a dynamic multistage zero-sum game. To acquire the best strategy for users, we employ a modified state-action-reward-state-action reinforcement learning algorithm. Iteration times decrease because of cardinality reduction from n to 2, which accelerates the convergence process. Our extensive experiments on real-world data sets demonstrate the efficiency and feasibility of the propose method.

© 2018 IEEE. Due to dramatically increasing information published in social networks, privacy issues have given rise to public concerns. Although the presence of differential privacy provides privacy protection with theoretical foundations, the trade-off between privacy and data utility still demands further improvement. However, most existing works do not consider the impact of the adversary in the measurement of data utility. In this paper, we firstly propose a personalized differential privacy based on social distance. Then, we analyze the maximum data utility when users and adversaries are blind to the strategy sets of each other. We formulize all the payoff functions in the differential privacy sense, which is followed by the establishment of a Static Bayesian Game. The trade-off is calculated by deriving the Bayesian Nash Equilibrium. In addition, the in-place trade-off can maximize the user' data utility if the action sets of the user and the adversary are public while the strategy sets are unrevealed. Our extensive experiments on the real-world dataset prove the proposed model is effective and feasible.

© 2018 Association for Computing Machinery. Machine learning (ML) is commonly used in multiple disciplines and real-world applications, such as information retrieval, financial systems, health, biometrics and online social networks. However, their security profiles against deliberate attacks have not often been considered. Sophisticated adversaries can exploit specific vulnerabilities exposed by classical ML algorithms to deceive intelligent systems. It is emerging to perform a thorough security evaluation as well as potential attacks against the machine learning techniques before developing novel methods to guarantee that machine learning can be securely applied in adversarial setting. In this paper, an effective attack strategy for crafting foreign support vectors in order to attack a classic ML algorithm, the Support Vector Machine (SVM) has been proposed with mathematical proof. The new attack can minimize the margin around the decision boundary and maximize the hinge loss simultaneously. We evaluate the new attack in different real-world applications including social spam detection, Internet traffic classification and image recognition. Experimental results highlight that the security of classifiers can be worsened by poisoning a small group of support vectors.

© 2018 IEEE. Network inference has been extensively studied to better understand the information diffusion in online social networks. In this field, state-of- art widely adopted a priori knowledge related to users' infection timestamps. Researchers also assume that the smaller the time difference between two nodes, the higher the likelihood of an edge between the pair of users. However, according to our technical analyses and empirical studies, existing methods have two critical problems 1) alternative spreading paths 2) users' delivery delay, which leads to the inaccuracy of previous methods. In this paper, we developed an innovative method to address the inference inaccuracy caused by the exposed two problems. This method determined the existence of an edge between a pair of users according to part of the users' features. The experiment results suggested that our method achieved around 70% accuracy in inferring network structures while existing methods failed in the same tasks.

© Springer Nature Switzerland AG 2018. Managing security in an RFID system is a complex activity considering that it is imperatively challenging to implement trust among tags and readers. There is always the chance that an unauthorized individual might assume the identity of a trusted tag and manage to gain confidential data in an RFID system. The situation becomes worse in systems that use a backend server and a private Internet connection. In such a system, there is no comprehensive mechanism for authenticating a tag into the system. It is thus essential to consider the implementation of a robust framework that improves the trust and the authentication levels in an RFID system. In this paper, a system known as Secure Cloud-Assisted RFID Authentication (SCARA) is proposed, which uses cloud-assisted RFID authentication to reap benefits of cloud-like scalability, availability and fault tolerance. It has three parties such as a cloud server, RFID reader and issuer involved. Issuer provides system parameters to other parties through a secure channel. Server and RFID reader are included in the authentication process with the help of information obtained from the issuer. The proposed system is secure even if the private keys associated with server and RFID tag are compromised. It does mean that it can prevent server-side insider attack in addition to external attacks. Amazon EC2 is used to have experiments. We built a prototype application to demonstrate proof of the concept. The empirical results revealed that the proposed system is able to withstand various kinds of attacks and provides a more efficient solution with less overhead.

Location privacy has drawn much attention among mobile social network users, as the geo-location information can be used by the adversaries to launch localization attacks which focus on finding people’s sensitive locations such as home and office place. In this chapter, we propose a community-based information sharing scheme to help the users to protect their home locations. First, we study the existing home location prediction algorithms and conclude that they are all mainly based on the spatial and temporal features of the check-in data. Then we design the community-based information sharing scheme which aggregates the check-ins of all community members, thus change the overall spatial and temporal features. Finally, our simulation results validate that our proposed scheme greatly reduces the home location prediction accuracy and therefore can protect the user’s privacy effectively.

© 2018, Springer Nature Switzerland AG. The advent of cloud computing has inspired an increasing number of users outsourcing their data to remote servers to enjoy flexible and affordable data management services. However, storing data in a remote cloud server raises data privacy and security concerns, i.e., the integrity and origin of the query results. Although some solutions have been proposed to address these issues, none of them consider the arbitrary dissemination control of authenticated tree-structured data while disseminating to other users. To address the above concerns, in this paper, we first propose a novel and efficient redactable signature scheme which features editable homomorphic operation and redaction control on tree-structured data. Subsequently, we prove the security properties of our scheme and conduct extensive theoretical and experimental analyses. The experimental results show that our scheme outperforms the existing solutions in disseminating of authenticated tree-structured data with privacy protection and dissemination control in outsourced database (ODB) model.

© Springer Nature Switzerland AG 2018. One of the latest and most important research topics in the field of information diffusion, which has attracted many social network analyst experts in recent years, is how information is disseminated on social networks. In this paper, a new method is proposed by integration of ant colony algorithm and node centrality to increase the prediction accuracy of information diffusion paths on social networks. In the first stage of our approach, centrality of all nodes in the network is calculated. Then, based on the distances of nodes in the network and also ant colony algorithm, the optimal path of propagation is detected. After implementation of the proposed method, 4 real social network data sets were used to evaluate its performance. The evaluation results of all methods showed a better outcome for our method.

This chapter reviews the existing location privacy-preserving mechanisms (LPPMs). There have been some survey papers discussing this topic. For example, Shokri et al. [1] discuss LPPMs in two groups: obfuscation mechanisms and anonymization mechanisms. Our review of the literature reveals two further mechanisms, cryptographic mechanisms and shared information reduction mechanisms, creating four categories for existing LPPMs.

The privacy issue strongly impedes the development of mobile crowd sensing (MCS) applications. Under the current MCS framework, processes including bidding, task assignment, and sensed data uploading are all potentially risky for participants. As an effort toward this issue, we propose a framework which enhances the location privacy of MCS applications by reducing the bidding and assignment steps in the MCS cycle. Meanwhile, to reduce the unnecessary privacy loss while maintaining the required quality of service, the economic theory is used to help both the service provider (SP) and participants to decide their strategies. We propose schemes based on both the Monopoly and Oligopoly models. In the former case, the participants cooperate to gain exclusive control of the supply of crowd sensing data, while the latter case is a state of limited competition. The parameters in different schemes are analyzed, and the strengths and weaknesses of both schemes are discussed. Additionally, the proposed schemes are evaluated by extensive simulations, and the results are discussed in detail.