A5101754102- Software System Performance and Reliability
- Cloud Computing and Resource Management
- Big Data and Business Intelligence
- Privacy-Preserving Technologies in Data
- Internet Traffic Analysis and Secure E-voting
- Network Traffic and Congestion Control
- Adversarial Robustness in Machine Learning
- Privacy, Security, and Data Protection
- Human Mobility and Location-Based Analysis
- Parallel Computing and Optimization Techniques
- Security and Verification in Computing
- Advanced Malware Detection Techniques
- Traffic Prediction and Management Techniques
- Distributed systems and fault tolerance
- Security in Wireless Sensor Networks
- Music and Audio Processing
- Digital Media Forensic Detection
- Advanced Steganography and Watermarking Techniques
- Network Security and Intrusion Detection
- Gaze Tracking and Assistive Technology
- Generative Adversarial Networks and Image Synthesis
- Vehicular Ad Hoc Networks (VANETs)
- Caching and Content Delivery
Guizhou University of Finance and Economics
2024
University of Illinois Urbana-Champaign
2020
University of Illinois System
2019
Shanghai Jiao Tong University
2017-2018
Though representing a promising approach for personalization, targeting, and recommendation, aggregation of user profiles from multiple social networks will inevitably incur serious privacy leakage issue. In this paper, we propose Novel Heterogeneous De-anonymization Scheme (NHDS) aiming at de-anonymizing heterogeneous networks. NHDS first leverages the network graph structure to significantly reduce size candidate set, then exploits profile information identify correct mapping users with...
A large percentage of real-world software configuration issues, such as misconfigurations, involve multiple interdependent parameters. However, existing techniques and tools either do not consider dependencies among parameters— termed dependencies—or rely on one or two dependency types code patterns input. Without rigorous understanding dependencies, it is hard to deal with many resulting issues.
Deep neural networks (DNNs) have recently been widely adopted in various applications, and such success is largely due to a combination of algorithmic breakthroughs, computation resource improvements, access large amount data. However, the large-scale data collections required for deep learning often contain sensitive information, therefore raising many privacy concerns. Prior research has shown several successful attacks inferring training as model inversion, membership inference,...
System call checking is extensively used to protect the operating system kernel from user attacks. However, existing solutions such as Seccomp execute lengthy rule-based programs against calls and their arguments, leading substantial execution overhead.To minimize overhead, this paper proposes Draco, a new architecture that caches IDs argument values after they have been checked validated. are first looked-up in special cache and, on hit, skip all checks. We present both software hardware...
Usage-based Insurance (UBI) is regarded as a promising way to offer more accurate insurance premium by profiling driving behaviors. Compared with traditional which considers drivers' history of accidents, traffic violations and etc, UBI focuses on data can give reasonable based the current Insurers use sensors in smartphone or vehicle collect (e.g. mileage, speed, hark braking) compute risk score these recalculate premium. Many programs, are advertised being privacy-preserving, do not...
As smart phones gradually become the dominant network traffic generators, app analysis methods have gained great interests for management and targeted advertisement. Specifically, previous works shown that scalability of inference via meta-data has edge over traditional payload based analysis. However, such mainly considered ideal scenario where only one is running on client's device, without any background noise interfered. In this paper, we extend research to a more practical scenario, by...
Last-mile geo-localization plays an essential role in many location-based services, such as fraud detection and targeted advertising. In this study, we point out that round trip time (RTT) latency shows extremely weak correlation with physical distance estimation China's Internet, since a path between vantage destination can often be circuitous inflated by queuing processing delays. To sidestep the measurement, perform three-tier hop count based IP mapping for on assumption each provincial...
To enjoy various utility and services, people are active in multiple social networks nowadays. With tons of data generated on platforms, accounts the same user different can be used to de-anonymize a large scale. The aggregation profiles poses threat privacy. concern privacy leakage, de-anonymization techniques, including graph based approaches profile approaches, widely studied recent years. However, few works throw light deanonymization between real-world heterogeneous networks. In this...