基于语义Web挖掘的宽范围扫描环境弱信号识别  

Identification of Weak Signal in Wide Range Scanning Environment Based on Semantic Web Mining

作　　者：顾伟[1,2] 傅德胜[1,2] 蔡玮[3] 

机构地区：[1]南京信息工程大学江苏省网络监控中心,南京210044 [2]南京信息工程大学计算机与软件学院,南京210044 [3]南京工程学院计算机工程学院,南京211167

出　　处：《科学技术与工程》2013年第29期8791-8797,共7页Science Technology and Engineering

基　　金：江苏高校优势学科建设工程项目(PAPD)资助

摘　　要：针对宽范围扫描环境中传统方法在弱信号探测评估时通常需要过高的人工干预而导致结果质量较低的问题,提出了基于语义Web挖掘的弱信号识别方法。首先借助于网络信息来表示组织环境;然后利用隐含语义索引在不同的语境中识别出相似的文本模式;最后采用一种新的弱信号最大化方法计算出相关弱信号的最大数量;并由奇异值分解的维数表示,从而实现弱信号的高效识别。实验结果表明,相比其他最先进的弱信号识别方法,所提方法取得了更好的识别性能。For the problem that the quality of results is low due to traditional methods need too much human intervention in weak signal detection and assessment of wide range scanning environment,a method for detection of weak signal based on semantic Web mining is proposed.Internet information is used to represent organization＇ s environment at first.Then,latent semantic indexing （LSI） is used to identify similar textual patterns in different contexts.Finally,a new weak signal maximization approach is applied to calculate the largest number of relevant weak signals represented by singular value decomposition （SVD） dimensions,by which weak signal can be realized efficiently.Experimental results indicate that proposed method can more efficiently identify weak signal from the network comparing with the latest weak signal detection approaches.

关 键 词：弱信号 扫描环境 隐含语义索引 组织环境 奇异值分解 WEB挖掘 

分 类 号：TP311.11[自动化与计算机技术—计算机软件与理论]