- Network Security and Intrusion Detection
- Spam and Phishing Detection
- Internet Traffic Analysis and Secure E-voting
- Complex Network Analysis Techniques
- Software System Performance and Reliability
- Anomaly Detection Techniques and Applications
- Advanced Malware Detection Techniques
- Sentiment Analysis and Opinion Mining
- Data Management and Algorithms
- Nonlinear Partial Differential Equations
- Advanced Mathematical Modeling in Engineering
- Advanced Computational Techniques and Applications
- Geographic Information Systems Studies
- Robotic Path Planning Algorithms
- Text and Document Classification Technologies
- Network Traffic and Congestion Control
- Differential Equations and Boundary Problems
- Advanced Graph Neural Networks
- Technology and Security Systems
- Network Packet Processing and Optimization
- Caching and Content Delivery
- Differential Equations and Numerical Methods
- Safety and Risk Management
- Modular Robots and Swarm Intelligence
- Public Relations and Crisis Communication
Xi'an University of Technology
2019-2022
Beijing University of Posts and Telecommunications
2021
Xi'an Jiaotong University
2014-2019
Jiangsu Police Officer College
2015
Chinese Academy of Sciences
2011
Northeast Institute of Geography and Agroecology
2003-2011
Capital Normal University
2004
Shandong University
2002
Abstract We present various results on multiple solutions for superlinear elliptic equations in a bounded domain or the whole space. With Nehari type condition assumed, we show that standard Ambrosetti-Rabinowitz can be replaced by more natural super-quadratic condition.
Logs are generated by systems to record the detailed runtime information about system operations, and log analysis plays an important role in anomaly detection at host or network level. Most existing methods require a priori knowledge, which cannot be used detect new unknown anomalies. Moreover, growing volume of logs poses challenges detection. In this paper, we propose integrated method using K-prototype clustering k-NN classification algorithms, uses novel clustering-filtering-refinement...
Mastering user's behavior character is important for efficient network management and security monitoring. In this paper, we develop a novel framework named as multilevel user cluster mining (MUCM) to measure similarity under different prefix levels. Focusing on aggregated traffic prefixes cannot only reduce the number of flows but also reveal detailed patterns group users sharing similar behaviors. First, employ bidirectional flow bipartite graphs model characteristics in large-scale...
Anomaly detection is an important technique used to identify patterns of unusual network behavior and keep the under control. Today, attacks are increasing in terms both their number sophistication. To avoid causing significant traffic being detected by existing techniques, many new tend involve gradual adjustment behaviors, which always generate incomplete sessions due running mechanisms. Accordingly, this work, we employ symmetry degree profile anomalies further behaviors. We first...
Log analysis is an efficiency way to detect threats by scrutinizing the events recorded operating systems and devices. However, it more difficult discover accurately due massive amount of logs their various formats. Focusing on this problem, authors propose a method for potential mining based correlation multi-type logs. Firstly, they extract 12 features, including behavior-related, attribute-related measurable from characteristics known attacks. They also normalization deal with these...
With the number of Internet users and applications continues to grow, it becomes increasingly important understand users' behavior character for efficient network management security monitoring. Web is one most popular applications, which can help obtain anything they want. In this paper, we propose a new framework measure monitor web access character. Firstly, services are divided into 12 types according content provide. As user like different at time instants, proposed spectrum describe in...
The widespread use of social media, cloud computing, and Internet Things generates massive behavior data recorded by system logs, how to utilize these improve the stability security systems becomes more difficult due increasing number users amount data. In this paper, we propose a novel model named rhythm (BR) characterize visualize user's behaviors from logs apply it management. Based on BR model, conduct clustering analysis mine user clusters. Different management access control policies...
Recommender system, as a data-driven way to help customers locate products that match their interests, is increasingly critical for providing competitive customer suggestions in many web services. However, recommender systems are highly vulnerable malicious injection attacks due fundamental vulnerabilities and openness. With the endless emergence of new attacks, how provide feasible defending different threats against online recommendations still an under-explored issue. In this paper, we...
Device identification is of great importance in system management and network security. Especially, it the priority industrial internet things (IIoT) scenario. Since there are massive devices producing various kinds information manufacturing process, robustness, reliability, security real-time control whole based on IIoT devices. Previous device solutions mostly a centralized architecture, which brings lot problems scalability In addition, most traditional systems can only identify inherent...
Abstract Behavior security management refers to monitoring and guiding the user's opinions in online social networks reduce their harmful influence public security. Pushing designed tweets with specific contains them is one of promising ways solve this problem. In paper, we developed a new method for high‐quality supervision tweet generation, which not only considers semantics but also includes requirement different aspects. Firstly, collect millions six typical events. Following, construct...
The quickly development of many online applications benefit our daily life. But on the other hand, user usually holds several aliases in different applications. across multi-online detection are becoming more and important for E-marketing user's behavior monitoring. In this paper, we propose a method detecting Firstly, employ active positive methods to collect alias information from famous applications, including Email, RenRen etc. Then analyzed characteristics specific some interesting...
Users can participate in variety topics and express their opinions using different kinds of online applications, the IDs (nickname) they used are usually virtual difficult for finding physical person. Which pose great challenges network security management user's behavior supervision. Focus on this problem, we proposed methods nickname detection based connection structure similarity propagation model. Firstly, collected profile information from two popular sina microblog RenRen network. Then...
Distributed denial of service (DDoS) attacks have become a major security threat in data center scenarios, and the rapid growth traffic volume sophistication DDoS further increases difficulty detection. Most existing works are hard to distinguish from legitimate flash crowds, suffer high computational cost detecting attacks. In this paper, we propose an efficient detection method based on Renyi cross entropy attention-based BiGRU, which uses prescreening-and-detection framework detect...
Road information is a kind of significant data sources in navigable geo-spatial database, the model and storage structure which are directly related to application efficiency navigation system. This paper starts with analysis several common spatio-temporal models, then we put forward an improved based state amendments named as reversely meet requirements fast updating road center. The basic principle that database center on large server stores current time changing each historical moment....