基于频繁链接的Web权威资源挖掘

发布时间：2018-11-25 21:55

【摘要】：如何有效地利用Web这个巨大的信息库 ?传统的基于关键字的搜索引擎取得了一定的成绩 ,但是存在着查准率不高的问题 Web页面间链接结构事实上隐含地表达着权威的信息 ,这已被许多研究者用来试图改善Web信息检索(包括搜索引擎 )的性能 ,取得了较好的效果 ,但依然存在很大的改善空间为此 ,提出了FARMING(基于频繁度的Web图的权威资源挖掘 )算法诠释了新的权威页面定义 ,提出了带阶的频繁子图和权威社团等概念 ,并用实验证明了FARMING算法的有效性
[Abstract]:How to effectively utilize the huge information base of Web? The traditional keyword-based search engine has made some achievements, but there is a problem that the precision rate is not high. In fact, the link structure between Web pages implicitly expresses authoritative information. This has been used by many researchers to improve the performance of Web information retrieval (including search engines), but there is still much room for improvement. In this paper, FARMING (authoritative Resource Mining based on frequent Web Graph) algorithm is proposed to interpret the new definition of authoritative page, and the concepts of frequent subgraph with order and authoritative community are put forward, and the validity of FARMING algorithm is proved by experiments.
【作者单位】：复旦大学计算机与信息技术系复旦大学计算机与信息技术系复旦大学计算机与信息技术系复旦大学计算机与信息技术系复旦大学计算机与信息技术系
【基金】：国家自然科学基金 ( 6993 3 0 10 ) 国家“八六三”高技术研究发展计划基金 ( 2 0 0 2AA4Z3 43 0 )
【分类号】：TP393.09

【相似文献】