作者: Tao-Wei Chiou , Shi-Chun Tsai , Yi-Bing Lin
DOI: 10.1007/S00500-013-1218-0
关键词: Spectral clustering 、 Domain (software engineering) 、 Big data 、 Cluster analysis 、 Denial-of-service attack 、 Network security management 、 Campus network 、 Airfield traffic pattern 、 Computer science 、 Data mining 、 Network security 、 Artificial intelligence 、 Machine learning
摘要: Profiling network traffic pattern is an important approach for tackling security problem. Based on campus infrastructure, we propose a new method to identify randomly generated domain names and pinpoint the potential victim groups. We characterize normal with so called popular 2gram (2 consecutive characters in word) distinguish between active nonexistent names. also track destination IPs of sources analyze their similarity connection uncover anomalous group behaviors. apply Hadoop technique deal big data classify clients as victims or not spectral clustering method.