The rapid development of the Network makes the comprehensive analysis as well as the quantitative evaluation of its security become more and mere important. This paper illustrates the major realization process of a Ne...The rapid development of the Network makes the comprehensive analysis as well as the quantitative evaluation of its security become more and mere important. This paper illustrates the major realization process of a Network Security Quantitative Evaluation System,which,from an intruder's angle ,established a Hierarchy Intrusion Relationship Graph by analyzing the credit degree fusion and relevancy of the secure information of the target network and by combining with powerful database information. At last, by applying some relative mathematics model and arithmetic, the paper analyzes and evaluates the security of this Network Hierarchy Intrusion Relationship Graph comprehensively and quantitatively.展开更多
随着自媒体的蓬勃发展,任何人都可以在网上随意发布和转发信息,而这些信息可能是真实的,也可能是道听途说或被故意篡改的。互联网上数据的严重冗余和弱可信问题,导致现有数据的可用性很差。Bi-LSTM-CRF(Bi-Long Short Term Memory with ...随着自媒体的蓬勃发展,任何人都可以在网上随意发布和转发信息,而这些信息可能是真实的,也可能是道听途说或被故意篡改的。互联网上数据的严重冗余和弱可信问题,导致现有数据的可用性很差。Bi-LSTM-CRF(Bi-Long Short Term Memory with Conditional Random Field Layer)网络虽然能够解决数据中命名实体识别的准确率问题,但不能满足识别出的实体是可信的这一要求。文中提出一种基于编辑距离的多实体可信确认算法,并通过人物命名实体识别实例对该算法进行验证。首先通过分布式爬虫抓取同一个邮箱地址在多个搜索引擎上的Top N网页记录,然后使用经过双语语料训练后的Bi-LSTM-CRF模型抽取每个页面内的人物命名实体,最后通过实体多参数融合确定邮箱所对应的人物命名实体。实验结果表明,多实体可信确认算法能够将邮箱地址与邮箱真实主人的匹配准确率MRR(Mean Reciprocal Rank)提高到91.32%,相比只使用词频的算法其MRR提升了23.08%。实验数据充分说明,多实体可信确认算法能很好地从弱可信数据中获得强可信度的实体,降低海量数据中的低质特性,从而有效地增强实体数据源的可信度。展开更多
文摘The rapid development of the Network makes the comprehensive analysis as well as the quantitative evaluation of its security become more and mere important. This paper illustrates the major realization process of a Network Security Quantitative Evaluation System,which,from an intruder's angle ,established a Hierarchy Intrusion Relationship Graph by analyzing the credit degree fusion and relevancy of the secure information of the target network and by combining with powerful database information. At last, by applying some relative mathematics model and arithmetic, the paper analyzes and evaluates the security of this Network Hierarchy Intrusion Relationship Graph comprehensively and quantitatively.
文摘随着自媒体的蓬勃发展,任何人都可以在网上随意发布和转发信息,而这些信息可能是真实的,也可能是道听途说或被故意篡改的。互联网上数据的严重冗余和弱可信问题,导致现有数据的可用性很差。Bi-LSTM-CRF(Bi-Long Short Term Memory with Conditional Random Field Layer)网络虽然能够解决数据中命名实体识别的准确率问题,但不能满足识别出的实体是可信的这一要求。文中提出一种基于编辑距离的多实体可信确认算法,并通过人物命名实体识别实例对该算法进行验证。首先通过分布式爬虫抓取同一个邮箱地址在多个搜索引擎上的Top N网页记录,然后使用经过双语语料训练后的Bi-LSTM-CRF模型抽取每个页面内的人物命名实体,最后通过实体多参数融合确定邮箱所对应的人物命名实体。实验结果表明,多实体可信确认算法能够将邮箱地址与邮箱真实主人的匹配准确率MRR(Mean Reciprocal Rank)提高到91.32%,相比只使用词频的算法其MRR提升了23.08%。实验数据充分说明,多实体可信确认算法能很好地从弱可信数据中获得强可信度的实体,降低海量数据中的低质特性,从而有效地增强实体数据源的可信度。
