基于统计机器学习的互联网暗链检测方法
发布时间：2019-06-25 14:05
【摘要】：互联网搜索引擎排名算法中,外部链接是一个重要因素,而利用链接作弊现象普遍存在于互联网中。暗链是链接作弊其中的一种手段,难以检测和清除,被称为"网络牛皮癣"。为了维护公平的搜索引擎排名机制,保证搜索结果质量,针对暗链这种作弊手段,提出了一种基于机器学习的互联网暗链检测方法,该方法结合网页源码锚文本的特征检测暗链。给出了相关性能分析,在真实的网络环境下的实验验证表明了所提出的方法可行有效。该研究为搜索引擎打击链接隐藏的作弊行为提供了理论和实践支撑。
[Abstract]:In the Internet search engine ranking algorithm, the external link is an important factor, and the link cheating phenomenon is ubiquitous in the Internet. The hidden chain is a means of linking cheating, which is difficult to detect and clear and is called a "network psoriasis". In order to maintain a fair search engine ranking mechanism and to guarantee the quality of the search results, a method for detecting the hidden chain based on the machine learning is proposed in view of the cheating means of the hidden chain, and the method combines the characteristics of the web page source code anchor text to detect the hidden chain. The relevant performance analysis is given, and the experimental verification in the real network environment shows that the proposed method is feasible and effective. The research provides theoretical and practical support for search engine's anti-link hidden cheating behavior.
【作者单位】： 中国科学院计算机网络信息中心;中国互联网络信息中心;
【基金】：国家自然科学基金资助项目(61375039,61005029) 中国科学院计算机网络信息中心“一三五”规划重点培育方向专项基金资助项目(CNIC_PY_1402)
【分类号】：TP391.3;TP181
[Abstract]:In the Internet search engine ranking algorithm, the external link is an important factor, and the link cheating phenomenon is ubiquitous in the Internet. The hidden chain is a means of linking cheating, which is difficult to detect and clear and is called a "network psoriasis". In order to maintain a fair search engine ranking mechanism and to guarantee the quality of the search results, a method for detecting the hidden chain based on the machine learning is proposed in view of the cheating means of the hidden chain, and the method combines the characteristics of the web page source code anchor text to detect the hidden chain. The relevant performance analysis is given, and the experimental verification in the real network environment shows that the proposed method is feasible and effective. The research provides theoretical and practical support for search engine's anti-link hidden cheating behavior.
【作者单位】： 中国科学院计算机网络信息中心;中国互联网络信息中心;
【基金】：国家自然科学基金资助项目(61375039,61005029) 中国科学院计算机网络信息中心“一三五”规划重点培育方向专项基金资助项目(CNIC_PY_1402)
【分类号】：TP391.3;TP181
【共引文献】
相关期刊论文 前10条
1 杨绪兵,韩自存;ε不敏感的核Adaline算法及其在图像去噪中的应用[J];安徽工程科技学院学报(自然科学版);2003年04期
2 陶秀凤,唐诗忠,周鸣争;基于支持向量机的软测量模型及应用[J];安徽工程科技学院学报(自然科学版);2004年02期
3 许高程;张文君;王卫红;;支持向量机技术在遥感影像滑坡体提取中的应用[J];安徽农业科学;2009年06期
4 郭立萍;唐家奎;米素娟;张成雯;赵理君;;基于支持向量机遥感图像融合分类方法研究进展[J];安徽农业科学;2010年17期
5 冯学军;;最小二乘支持向量机的研究与应用[J];安庆师范学院学报(自然科学版);2009年01期
6 邹心遥;姚若河;;基于LSSVM的威布尔分布形状参数估计(英文)[J];半导体技术;2008年06期
7 邹心遥;姚若河;;基于LSSVM的小子样元器件寿命预测[J];半导体技术;2011年09期
8 李卓远,吴为民,王e，
本文编号：2505738
本文链接：https://www.wllwen.com/kejilunwen/sousuoyinqinglunwen/2505738.html
