IEEE 2017: SociRank: Identifying and Ranking Prevalent NewsTopics Using Social Media Factors

Abstract: Mass media sources, specifically the news media, have traditionally informed us of daily events. In modern times, social media services such as Twitter provide an enormous amount of user-generated data, which have great potential to contain informative news-related content. For these resources to be useful, we must find a way to filter noise and only capture the content that, based on its similarity to the news media, is considered valuable. However, even after noise is removed, information overload may still exist in the remaining data—hence, it is convenient to prioritize it for consumption. To achieve prioritization, information must be ranked in order of estimated importance considering three factors. First, the temporal prevalence of a particular topic in the news media is a factor of importance, and can be considered the media focus (MF) of a topic. Second, the temporal prevalence of the topic in social media indicates its user attention (UA). Last, the interaction between the social media users who mention this topic indicates the strength of the community discussing it, and can be regarded as the user interaction (UI) toward the topic. We propose an unsupervised framework—SociRank—which identifies news topics prevalent in both social media and the news media, and then ranks them by relevance using their degrees of MF, UA, and UI. Our experiments

Click for more details

IEEE 2017: RAPARE: A Generic Strategy for Cold-Start Rating Prediction Problem

Abstract:I n recent years, recommender system is one of indispensable components in many e-commerce websites. One of the major challenges that largely remains open is the cold-start problem, which can be viewed as a barrier that keeps the cold-start users/items away from the existing ones. In this paper, we aim to break through this barrier for cold-start users/items by the assistance of existing ones. In particular, inspired by the classic Elo Rating System, which has been widely adopted in chess tournaments; we propose a novel rating comparison strategy (RAPARE) to learn the latent profiles of cold-start users/items. The center-piece of our RAPARE is to provide a fine-grained calibration on the latent profiles of cold-start users/items by exploring the differences between cold-start and existing users/items. As a generic strategy, our proposed strategy can be instantiated into existing methods in recommender systems. To reveal the capability of RAPARE strategy, we instantiate our strategy on two prevalent methods in recommender systems, i.e., the matrix factorization based and neighborhood based collaborative filtering.

Click for more details

IEEE 2017: l-Injection: Toward Effective Collaborative Filtering Using Uninteresting Items

Abstract: We develop a novel framework, named as l-injection, to address the sparsity problem of recommender systems. By carefully injecting low values to a selected set of unrated user-item pairs in a user-item matrix, we demonstrate that top-N recommendation accuracies of various collaborative filtering (CF) techniques can be significantly and consistently improved. We first adopt the notion of pre-use preferences of users toward a vast amount of unrated items. Using this notion, we identify uninteresting items that have not been rated yet but are likely to receive low ratings from users, and selectively impute them as low values. As our proposed approach is method-agnostic, it can be easily applied to a variety of CF algorithms. Through comprehensive experiments with three real-life datasets (e.g., Movielens, Ciao, and Watcha), we demonstrate that our solution consistently and universally enhances the accuracies of existing CF algorithms (e.g., item-based CF, SVD-based CF, and SVD++) by 2.5 to 5 times on average. Furthermore, our solution improves the running time of those CF methods by 1.2 to 2.3 times when its setting produces the best accuracy.

Click for more details

IEEE 2017: Vehicular Cloud Data Collection for Intelligent Transportation Systems

Abstract:  The Internet of Things (IoT) envisions connecting billions of sensors to the Internet, in order to provide new applications and services for smart cities. IoT will allow the evolution of the Internet of Vehicles (IoV) from existing Vehicular Ad hoc Networks (VANETs), in which the delivery of various services will be offered to drivers by integrating vehicles, sensors, and mobile devices into a global network. To serve VANET with computational resources, Vehicular Cloud Computing (VCC) is recently envisioned with the objective of providing traffic solutions to improve our daily driving. These solutions involve applications and services for the benefit of Intelligent Transportation Systems (ITS), which represent an important part of IoV. Data collection is an important aspect in ITS, which can effectively serve online travel systems with the aid of Vehicular Cloud (VC). In this paper, we involve the new paradigm of VCC to propose a data collection model for the benefit of ITS. We show via simulation results that the participation of low percentage of vehicles in a dynamic VC is sufficient to provide meaningful data collection.
Click for more details

IEEE 2017: Optimizing Green Energy, Cost, and Availability in Distributed Data Centers

Abstract:  Integrating renewable energy and ensuring high availability are among two major requirements for geodistributed data centers. Availability is ensured by provisioning spare capacity across the data centers to mask data center failures (either partial or complete). We propose a mixed integer linear programming formulation for capacity planning while minimizing the total cost of ownership (TCO) for highly available, green, distributed data centers. We minimize the cost due to power consumption and server deployment, while targeting a minimum usage of green energy. Solving our model shows that capacity provisioning considering green energy integration, not only lowers carbon footprint but also reduces the TCO. Results show that up to 40% green energy usage is feasible with marginal increase in the TCO compared to the other cost-aware models.

Click for more details

IEEE 2016 : SPORE : A Sequential Personalized Spatial Item Recommender System

Abstract: With the rapid development of location-based social networks (LBSNs), spatial item recommendation has become an important way of helping users discover interesting locations to increase their engagement with location-based services. Although human movement exhibits sequential patterns in LBSNs, most current studies on spatial item recommendations do not consider the sequential influence of locations. Leveraging sequential patterns in spatial item recommendation is, however, very challenging, considering 1) users’ check-in data in LBSNs has a low sampling rate in both space and time, which renders existing prediction techniques on GPS trajectories ineffective; 2) the prediction space is extremely large, with millions of distinct locations as the next prediction target, which impedes the application of classical Markov chain models; and 3) there is no existing framework that unifies users’ personal interests and the sequential influence in a principled manner.In light of the above challenges, we propose a sequential personalized spatial item recommendation framework (SPORE) which introduces a novel latent variable topic-region to model and fuse sequential influence with personal interests in the latent and exponential space. The advantages of modeling the sequential effect at the topic-region level include a significantly reduced prediction space, an effective alleviation of data sparsity and a direct expression of the semantic meaning of users’ spatial activities. Furthermore, we design an asymmetric Locality Sensitive Hashing (ALSH) technique to speed up the online top-k recommendation process by extending the traditional LSH. We evaluate the performance of SPORE on two real datasets and one large-scale synthetic dataset. The results demonstrate a significant improvement in SPORE’s ability to recommend spatial items, in terms of both effectiveness and efficiency, compared with the state-of-the-art methods.

Click for more details

IEEE 2016 : Truth Discovery in Crowdsourced Detection of Spatial Events

Abstract:The ubiquity of smartphones has led to the emergence of mobile crowdsourcing tasks such as the detection of spatial events when smartphone users move around in their daily lives. However, the credibility of those detected events can be negatively impacted by unreliable participants with low-quality data. Consequently, a major challenge in quality control is to discover true events from diverse and noisy participants’ reports. This truth discovery problem is uniquely distinct from its online counterpart in that it involves uncertainties in both participants’ mobility and reliability. Decoupling these two types of uncertainties through location tracking will raise severe privacy and energy issues, whereas simply ignoring missing reports or treating them as negative reports will significantly degrade the accuracy of the discovered truth. In this paper, we propose a new method to tackle this truth discovery problem through principled probabilistic modeling. In particular, we integrate the modeling of location popularity, location visit indicators, truth of events and three-way participant reliability in a unified framework. The proposed model is thus capable of efficiently handling various types of uncertainties and automatically discovering truth without any supervision or the need of location tracking. Experimental results demonstrate that our proposed method outperforms existing state-of-the-art truth discovery approaches in the mobile crowdsourcing environment.

Click for more details

IEEE 2016 : Sentiment Analysis of Top Colleges in India Using Twitter Data

Abstract: Ttoday’s world, opinions and reviews accessible to us are one of the most critical factors in formulating our views and influencing the success of a brand, product or service. With the advent and growth of social media in the world, stakeholders often take to expressing their opinions on popular social media, namely Twitter. While Twitter data is extremely informative, it presents a challenge for analysis because of its humongous and disorganized nature. This paper is a thorough effort to dive into the novel domain of performing sentiment analysis of people’s opinions regarding top colleges in India. Besides taking additional preprocessing measures like the expansion of net lingo and removal of duplicate tweets, a probabilistic model based on Bayes’ theorem was used for spelling correction, which is overlooked in other research studies. This paper also highlights a comparison between the results obtained by exploiting the following machine learning algorithms: Naïve Bayes and Support Vector Machine and an Artificial Neural Network model: Multilayer Perceptron. Furthermore, a contrast has been presented between four different kernels of SVM: RBF, linear, polynomial and sigmoid.

Click for more details

IEEE 2016 : FRAppE: Detecting Malicious Facebook Applications

Abstract:With 20 million installs a day [1], third-party apps are a major reason for the popularity and addictiveness of Facebook. Unfortunately, hackers have realized the potential of using apps for spreading malware and spam. The problem is already significant, as we find that at least 13% of apps in our dataset are malicious. So far,the research community has focused on detecting malicious posts and campaigns. In this paper, we ask the question: given a Facebook application, can we determine if it is malicious? Our key contribution is in developing FRAppE—Facebook’s Rigorous Application Evaluator— arguably the first tool focused on detecting malicious apps on Facebook. To develop FRAppE, we use information gathered by observing the posting behavior of 111K Facebook apps seen across 2.2 million users on Facebook. First, we identify a set of features that help us distinguish malicious apps from benign ones. For example, we find that malicious apps often share names with other apps, and they typically request fewer permissions than benign apps. Second, leveraging these distinguishing features, we show that FRAppE can detect malicious apps with 99.5% accuracy, with no false positives and a low false negative rate (4.1%). Finally, we explore the ecosystem of malicious Facebook apps and identify mechanisms that these apps use to propagate. Interestingly, we find that many apps collude and support each other; in our dataset, we find 1,584 apps enabling the viral propagation of 3,723 other apps through their posts. Long-term, we see FRAppE as a step towards creating an independent watchdog for app assessment and ranking,so as to warn Facebook users before installing apps.

Click for more details

IEEE 2016: Secure Optimization Computation Outsourcing in Cloud Computing: A Case Study of Linear Programming

Abstract:Cloud computing enables an economically promising paradigm of computation outsourcing. However, how to protect customers confidential data processed and generated during the computation is becoming the major security concern. Focusing on engineering computing and optimization tasks, this paper investigates secure outsourcing of widely applicable linear programming (LP) computations. Our mechanism design explicitly decomposes LP computation outsourcing into public LP solvers running on the cloud and private LP parameters owned by the customer. The resulting flexibility allows us to explore appropriate security/efficiency tradeoff via higher-level abstraction of LP computation than the general circuit representation. Specifically, by formulating private LP problem as a set of matrices/vectors, we develop efficient privacy-preserving problem transformation techniques, which allow customers to transform the original LP into some random one while protecting sensitive input/output information. To validate the computation result, we further explore the fundamental duality theorem of LP and derive the necessary and sufficient conditions that correct results must satisfy. Such result verification mechanism is very efficient and incurs close-to-zero additional cost on both cloud server and customers. Extensive security analysis and experiment results show the immediate practicability of our mechanism design.

Click for more details

IEEE 2016: On Traffic-Aware Partition and Aggregation in MapReduce for Big Data Applications

Abstract:The MapReduce programming model simplifies large-scale data processing on commodity cluster by exploiting parallel map tasks and reduce tasks. Although many efforts have been made to improve the performance of MapReduce jobs, they ignore the network traffic generated in the shuffle phase, which plays a critical role in performance enhancement. Traditionally, a hash function is used to partition intermediate data among reduce tasks, which, however, is not traffic-efficient because network topology and data size associated with each key are not taken into consideration. In this paper, we study to reduce network traffic cost for a MapReduce job by designing a novel intermediate data partition scheme. Furthermore, we jointly consider the aggregator placement problem, where each aggregator can reduce merged traffic from multiple map tasks. A decomposition-based distributed algorithm is proposed to deal with the large-scale optimization problem for big data application and an online algorithm is also designed to adjust data partition and aggregation in a dynamic manner. Finally, extensive simulation results demonstrate that our proposals can significantly reduce network traffic cost under both offline and online cases.

Click for more details

IEEE 2016: DeyPoS: Deduplicatable Dynamic Proof of Storage for Multi-User Environments

Abstract: Dynamic Proof of Storage (PoS) is a useful cryptographic primitive that enables a user to check the integrity of out sourced files and to efficiently update the files in a cloud server. Although researchers have proposed many dynamic PoS schemes in single user environments, the problem in multi-user environments has not been investigated sufficiently. A practical multi-user cloud storage system needs the secure client-side cross-user deduplication technique, which allows a user to skip the uploading process and obtain the ownership of the files immediately, when other owners of the same files have uploaded them to the cloud server. To the best of our knowledge, none of the existing dynamic PoSs can support this technique. In this paper, we introduce the concept of deduplicatable dynamic proof of storage and propose an efficient construction called DeyPoS, to achieve dynamic PoS and secure cross-user deduplication, simultaneously. Considering the challenges of structure diversity and private tag generation, we exploit a novel tool called Homomorphic Authenticated Tree (HAT). We prove the security of our construction, and the theoretical analysis and experimental results show that our construction is efficient in practice.

Click for more details

IEEE 2016: Fine-Grained Two-Factor Access Control for Web-Based Cloud Computing Services

Abstract:In this paper, we introduce a new fine-grained two-factor authentication (2FA) access control system for web-based cloud computing services. Specifically, in our proposed 2FA access control system, an attribute-based access control  mechanism is implemented with the necessity of both a user secret key and a lightweight security device. As a user cannot access the system if they do not hold both, the mechanism can enhance the security of the system, especially in those scenarios where many users share the same computer for web-based cloud services. In addition, attribute-based control in the system also enables the cloud server to restrict the access to those users with the same set of attributes while preserving user privacy, i.e., the cloud server only knows that the user fulfills the required predicate, but has no idea on the exact identity of the user. Finally, we also carry out a simulation to demonstrate the practicability of our proposed 2FA system.

Click for more details

IEEE 2015 : Privacy-Preserving Detection of Sensitive Data Exposure

Abstract:Statistics from security firms, research institutions and government organizations show that the number of data-leak instances have grown rapidly in recent years. Among various data-leak cases, human mistakes are one of the main causes of data loss. There exist solutions detecting inadvertent sensitive data leaks caused by human mistakes and to provide alerts for organizations. A common approach is to screen content in storage and transmission for exposed sensitive information. Such an approach usually requires the detection operation to be conducted in secrecy. However, this secrecy requirement is challenging to satisfy in practice, as detection servers may be compromised or outsourced. In this paper, we present a privacypreserving data-leak detection (DLD) solution to solve the issue where a special set of sensitive data digests is used in detection. The advantage of our method is that it enables the data owner to safely delegate the detection operation to a semihonest provider without revealing the sensitive data to the provider. We describe how Internet service providers can offer their customers DLD as an add-on service with strong privacy guarantees. The evaluation results show that our method can support accurate detection with very small number of false alarms under various data-leak scenarios.

Click for more details

IEEE 2015: An Energy-Efficient and Delay-Aware Wireless Computing System for Industrial Wireless Sensor Networks

Abstract: Industrial wireless sensor networks have attracted much attention as a cornerstone to making the smart factories real. Utilizing industrial wireless sensor networks as a base for smart factories makes it possible to optimize the production line without human resources since it provides industrial Internet of Things (IoT) service, where various types of data are collected from sensors and mined to control the machines based on the analysis result. On the other hand, a fog computing node, which executes such real-time feedback control, should be capable of real-time data collection, management, and processing. To achieve these requirements, in this paper, we introduce Wireless Computing System (WCS) as a fog computing node. Since there are a lot of servers and each server has 60 GHz antennas to connect to other servers and sensors, WCS has high collecting and processing capabilities. However, in order to fulfill a demand for real-time feedback control, WCS needs to satisfy an acceptable delay for data collection. Additionally, lower power consumption is required in order to reduce the cost for factory operation. Therefore, we propose an Energy-Efficient and Delay-Aware Wireless Computing System (E2DA-WCS). Since there is a tradeoff relationship between the power consumption and the delay for data collection, our proposed system controls the sleep schedule and the number of links to minimize the power consumption while satisfying an acceptable delay constraint. Furthermore, the effectiveness of our proposed system is evaluated through extensive computer simulations.

Click for more details

IEEE 2015: Cost-Effective Authentic and Anonymous Data Sharing with Forward Security

Abstract: Data sharing has never been easier with the advances of cloud computing, and an accurate analysis on the shared data provides an array of benefits to both the society and individuals. Data sharing with a large number of participants must take into account several issues, including efficiency, data integrity and privacy of data owner. Ring signature is a promising candidate to construct an anonymous and authentic data sharing system. It allows a data owner to anonymously authenticate his data which can be put into the cloud for storage or analysis purpose. Yet the costly certificate verification in the traditional public key infrastructure (PKI) setting becomes a bottleneck for this solution to be scalable. Identity-based (ID-based) ring signature, which eliminates the process of certificate verification, can be used instead. In this paper, we further enhance the security of ID-based ring signature by providing forward security: If a secret key of any user has been compromised, all previous generated signatures that include this user still remain valid. This property is especially important to any large scale data sharing system, as it is impossible to ask all data owners to reauthenticate their data even if a secret key of one single user has been compromised. We provide a concrete and efficient instantiation of our scheme, prove its security and provide an implementation to show its practicality.

Click for more details

IEEE 2015 : k Nearest Neighbor Search for Location-Dependent Sensor Data in MANETs

Abstract:K nearest neighbor (kNN) queries, which retrieve the k nearest sensor data items associated with a location (location-dependent sensor data) from the location of the query issuer, are useful for location-based services (LBSs) in mobile environments. Here, we focus on kNN query processing in mobile ad hoc networks (MANETs). Key challenges in designing system protocols for MANETs include low-overhead adaptability to network topology changes due to node mobility, and query processing that achieves high accuracy of the query result without a centralized server. In this paper, we propose the Filling Area (FA) method to efficiently process kNN queries in MANETs. The FA method achieves low overhead in query processing by reducing a search area. In the FA method, data items remain at nodes near the locations with which the items are associated, and nodes cache data items whose locations are near their own so that the query issuer retrieves kNNs from nearby nodes. Through extensive simulations, we verify that our proposed approach achieves low overhead and high accuracy of the query result.

Click for more details

IEEE 2015 : The Mason Test: A Defense Against Sybil Attacks in Wireless Networks Without Trusted Authorities

IEEE 2015 Transactions on Parallel and Distributed Systems 

 Abstract : Wireless networks are vulnerable to Sybil attacks, in which a malicious node poses as many identities in order to gain disproportionate influence. Many defenses based on spatial variability of wireless channels exist, but depend either on detailed, multi-tap channel estimation—something not exposed on commodity 802.11 devices—or valid RSSI observations from multiple trusted sources, e.g., corporate access points—something not directly available in ad hoc and delay-tolerant networks with potentially malicious neighbors. We extend these techniques to be practical for wireless ad hoc networks of commodity 802.11 devices. Specifically, we propose two efficient methods for separating the valid RSSI observations of behaving nodes from those falsified by malicious participants. Further, we note that prior signalprint methods are easily defeated by mobile attackers and develop an appropriate challenge-response defense. Finally, we present the Mason test, the first implementation of these techniques for ad hoc and delay-tolerant networks of commodity 802.11 devices. We illustrate its performance in several real-world scenarios.

IEEE 2015 : Secure and Distributed Data Discovery and Dissemination in Wireless Sensor Networks

IEEE 2015 Transactions on Parallel and Distributed Systems 

Abstract : A data discovery and dissemination protocol for wireless sensor networks (WSNs) is responsible for updating configuration parameters of, and distributing management commands to, the sensor nodes. All existing data discovery and dissemination protocols suffer from two drawbacks. First, they are based on the centralized approach; only the base station can distribute data item. Such an approach is not suitable for emergent multi-owner-multi-user WSNs. Second, those protocols were not designed with security in mind and hence adversaries can easily launch attacks to harm the network. This paper proposes the first secure and distributed data discovery and dissemination protocol named DiDrip. It allows the network owners to authorize multiple network users with different privileges to simultaneously and directly disseminate data items to the sensor nodes. Moreover, as demonstrated by our theoretical analysis, it addresses a number of possible security vulnerabilities that we have identified. Extensive security analysis show DiDrip is provably secure. We also implement DiDrip in an experimental network of resource-limited sensor nodes to show its high efficiency in practice.

IEEE 2015 : Automatic Face Naming by Learning Discriminative Affinity Matrices From Weakly Labeled Images

IEEE 2015 TRANSACTIONS ON IMAGE PROCESSING

Abstract : Given a collection of images, where each image contains several faces and is associated with a few names in the corresponding caption, the goal of face naming is to infer the correct name for each face. In this paper, we propose two new methods to effectively solve this problem by learning two discriminative affinity matrices from these weakly labeled images. We first propose a new method called regularized low-rank representation by effectively utilizing weakly supervised information to learn a low-rank reconstruction coefficient matrix while exploring multiple subspace structures of the data. Specifically, by introducing a specially designed regularizer to the low-rank representation method, we penalize the corresponding reconstruction coefficients related to the situations where a face is reconstructed by using face images from other subjects or by using itself. With the inferred reconstruction coefficient matrix, a discriminative affinity matrix can be obtained. Moreover, we also develop a new distance metric learning method called ambiguously supervised structural metric learning by using weakly supervised information to seek a discriminative distance metric. Hence, another discriminative affinity matrix can be obtained using the similarity matrix (i.e., the kernel matrix) based on the Mahalanobis distances of the data. Observing that these two affinity matrices contain complementary information, we further combine them to obtain a fused affinity matrix, based on which we develop a new iterative scheme to infer the name of each face. Comprehensive experiments demonstrate the effectiveness of our approach.

IEEE 2015 : Query Aware Determinization of Uncertain Objects

IEEE 2015 TRANSACTIONS ON IMAGE PROCESSING

Abstract :The determinizing probabilistic data to enable such data to be stored in legacy systems that accept only deterministic input. Probabilistic data may be generated by automated data analysis/enrichment techniques such as entity resolution, information extraction, and speech processing. The legacy system may correspond to pre-existing web applications such as Flickr, Picasa, etc. The goal is to generate a deterministic representation of probabilistic data that optimizes the quality of the end-application built on deterministic data. We explore such a determinization problem in the context of two different data processing tasks—triggers and selection queries. We show that approaches such as thresholding or top-1 selection traditionally used for determinization lead to suboptimal performance for such applications. Instead, we develop a query-aware strategy and show its advantages over existing solutions through a comprehensive empirical evaluation over real and synthetic datasets.

IEEE 2015 : RRW—A Robust and Reversible Watermarking Technique for Relational Data

IEEE 2015 TRANSACTIONS ON IMAGE PROCESSING

Abstract : Advancement in information technology is playing an increasing role in the use of information systems comprising relational databases. These databases are used effectively in collaborative environments for information extraction; consequently, they are vulnerable to security threats concerning ownership rights and data tampering. Watermarking is advocated to enforce ownership rights over shared relational data and for providing a means for tackling data tampering. When ownership rights are enforced using watermarking, the underlying data undergoes certain modifications; as a result of which, the data quality gets compromised. Reversible watermarking is employed to ensure data quality along-with data recovery. However, such techniques are usually not robust against malicious attacks and do not provide any mechanism to selectively watermark a particular attribute by taking into account its role in knowledge discovery. Therefore, reversible watermarking is required that ensures; (i) watermark encoding and decoding by accounting for the role of all the features in knowledge discovery; and, (ii) original data recovery in the presence of active malicious attacks. In this paper, a robust and semi-blind reversible watermarking (RRW) technique for numerical relational data has been proposed that addresses the above objectives. Experimental studies prove the effectiveness of RRW against malicious attacks and show that the proposed technique outperforms existing ones.

IEEE 2015 : Secure and Practical Outsourcing of Linear Programming in Cloud Computing

IEEE 2015 TRANSACTIONS ON SERVICE COMPUTING

Abstract :Cloud Computing has great potential of providing robust computational power to the society at reduced cost. It enables customers with limited computational resources to outsource their large computation workloads to the cloud, and economically enjoy the massive computational power, bandwidth, storage, and even appropriate software that can be shared in a pay-per-use manner. Despite the tremendous benefits, security is the primary obstacle that prevents the wide adoption of this promising computing model, especially for customers when their confidential data are consumed and produced during the computation. Treating the cloud as an intrinsically insecure computing platform from the viewpoint of the cloud customers, we must design mechanisms that not only protect sensitive information by enabling computations with encrypted data, but also protect customers from malicious behaviors by enabling the validation of the computation result. Such a mechanism of general secure computation outsourcing was recently shown to be feasible in theory, but to design mechanisms that are practically efficient remains a very challenging problem. Focusing on engineering computing and optimization tasks, this paper investigates secure outsourcing of widely applicable linear programming (LP) computations. In order to achieve practical efficiency, our mechanism design explicitly decomposes the LP computation outsourcing into public LP solvers running on the cloud and private LP parameters owned by the customer. The resulting flexibility allows us to explore appropriate security/ efficiency tradeoff via higher-level abstraction of LP computations than the general circuit representation. In particular, by formulating private data owned by the customer for LP problem as a set of matrices and vectors, we are able to develop a set of efficient privacy-preserving problem transformation techniques, which allow customers to transform original LP problem into some arbitrary one while protecting sensitive input/output information. To validate the computation result, we further explore the fundamental duality theorem of LP computation and derive the necessary and sufficient conditions that correct result must satisfy. Such result verification mechanism is extremely efficient and incurs close-to-zero additional cost on both cloud server and customers. Extensive security analysis and experiment results show the immediate practicability of our mechanism design.

IEEE 2015 : Face Recognition and Retrieval Using Cross-Age Reference Coding With Cross-Age Celebrity Dataset

IEEE 2015 TRANSACTIONS ON IMAGE PROCESSING

Abstract :  This paper introduces a method for face recognition across age and also a dataset containing variations of age in the wild. We use a data-driven method to address the cross-age face recognition problem, called cross-age reference coding (CARC). By leveraging a large-scale image dataset freely available on the Internet as a reference set, CARC can encode the low-level feature of a face image with an age-invariant reference space. In the retrieval phase, our method only requires a linear projection to encode the feature and thus it is highly scalable. To evaluate our method, we introduce a large-scale dataset called cross-age celebrity dataset (CACD). The dataset contains more than 160 000 images of 2,000 celebrities with age ranging from 16 to 62. Experimental results show that our method can achieve state-of-the-art performance on both CACD and the other widely used dataset for face recognition across age. To understand the difficulties of face recognition across age, we further construct a verification subset from the CACD called CACD-VS and conduct human evaluation using Amazon Mechanical Turk. CACD-VS contains 2,000 positive pairs and 2,000 negative pairs and is carefully annotated by checking both the associated image and web contents. Our experiments show that although state-of-the-art methods can achieve competitive performance compared to average human performance, majority votes of several humans can achieve much higher performance on this task. The gap between machine and human would imply possible directions for further improvement of cross-age face recognition in the future.