期刊名称：International Journal of Engineering and Computer Science
印刷版ISSN：2319-7242
出版年度：2015
卷号：4
期号：3
页码：11066-11071
出版社：IJECS
摘要：The availability of open source traffic classification[1][2][3][4][5] systems designed for both experimentaland operational use can facilitate cooperation, union on standard definitions and events and trustedevaluation of methods. In this paper, we describe Traffic Identification Engine (TIE), an open source tool fornetwork transfer classification, Investigating the optimal combination strategy and set of classifiers togenerate reliable ground truth while preserving privacy Extending the support for sharing labeled traffic withanonym zed traces Investigating strategies for multi-threaded classification, exploiting: off-load techniquesfocuses by recent traffic capturing engines such as multi queue adapters. Comparing the accuracy ofdifferent classifiers and classification performance. Investigating multi-classification and combinationstrategies.
关键词：patterns; accuracy of classifier;garbage collector; plug-in
