基于藏文Web的网络模型与免疫策略研究

发布时间：2018-06-21 16:12

本文选题：网络模型 + 藏文Web页面　；参考：《西北民族大学》2014年硕士论文

【摘要】：随着网络技术与民族语言文字处理技术的发展,以网页形式出现的藏文信息与日俱增,藏文网站的数量正以惊人的速度不断增长,藏文网页成为了互联网一个重要的组成部分。如何快速有效地保证这些数据庞大的藏文Web页面与链接的安全成为了一个亟需解决的问题,也是相关学者研究的热点。随机免疫策略、目标免疫策略和熟人免疫策略是三大免疫策略,分析三个免疫策略,对于随机免疫策略而言,若在无标度网络中实施随机免疫策略需要免疫全网节点才能保证最终遏止病毒的传播；目标免疫策略需要花费的计算代价太大；熟人免疫策略虽然避免了前两者的不足,但其免疫效果没有目标免疫策略好。首先对藏文Web网络拓扑结构进行分析,根据藏文Web网络的结构特性,提出合理的免疫算法。在藏文Web网络模型的基础上,设计了三个免疫算法： (1)基于藏文Web的最大度熟人免疫策略考虑到熟人免疫策略免疫目标不准确问题与目标免疫耗费代价问题本文提出对随机选出节点的最大度邻居进行免疫。 (2)基于藏文Web的双免疫策略本文进一步对熟人免疫策略进行改进,每一次免疫是对随机选出的节点以及其最大度邻居进行免疫,这个算法对具有无标度特性的网络和小世界特性的网络尤其适用。 (3)藏文Web中基于社团划分的免疫策略根据藏文Web网络的拓扑结构特性,依据社团划分的思想,在进行免疫前,首先对藏文Web网络进行社区划分,找到具有弱连接性质的节点。免疫开始时,首先免疫处于弱连接的节点,接着根据社团结构特性进行免疫。仿真结果表明,在藏文Web中,本文提出的三种免疫策略的免疫效果要优于熟人免疫策略,藏文Web中基于社团划分的免疫策略的免疫效果最好。
[Abstract]:With the development of network technology and national language word processing technology, Tibetan language information in the form of web pages is increasing day by day, the number of Tibetan language websites is increasing at an alarming speed, Tibetan web pages have become an important part of the Internet. How to quickly and effectively guarantee the security of these huge Tibetan Web pages and links has become a problem that needs to be solved, and it is also a hot research topic of relevant scholars. Random immunization strategy, target immunization strategy and acquaintance immunization strategy are three major immune strategies. If the implementation of random immunization strategy in scale-free networks requires immunization of the entire network nodes to ensure that the transmission of the virus will be stopped; the target immunization strategy will cost too much to calculate; although the acquaintance immunization strategy avoids the shortcomings of the first two, But its immune effect is not as good as the target immunization strategy. Firstly, the topological structure of Tibetan Web network is analyzed, and a reasonable immune algorithm is proposed according to the structural characteristics of Tibetan Web network. On the basis of Tibetan Web network model, Three immune algorithms are designed: (1) the maximum acquaintance immunization strategy based on Tibetan Web takes into account the inaccuracy of immune target and the cost of target immunization. Double immune Strategy based on Tibetan Web this paper further improves the immune strategy of acquaintances. Each immunization is to immunize randomly selected nodes and their maximum neighbors. This algorithm is especially suitable for scale-free networks and small-world networks. According to the idea of community division, before immunity, the community of Tibetan Web network is divided to find the node with weak connection. At the beginning of immunization, the weakly connected nodes are first immunized, and then immunity is carried out according to the characteristics of the community structure. The simulation results show that the immune effects of the three immune strategies proposed in this paper are better than those of acquaintances in Tibetan Web, and the immune effects of the immune strategies based on community division in Tibetan Web are the best.
【学位授予单位】：西北民族大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.092

【相似文献】