摘要: 为了解网络话题内容组成和演化情况,提出基于有向图的在线分类(OCBDG)方法,并设计一个基于网络搜索引擎的话题分析框架。通过搜索引擎查询话题内容,OCBDG将查询结果分成若干子话题,分析子话题间的关系和演变。结果证明,该方法能够以大约70%的正确率分析出子话题,并能准确、及时地反映话题在网络上任意时间跨度的变化情况。
关键词:
有向图,
分类,
网络搜索引擎,
网页摘要,
快照
Abstract: This paper presents a method of Online Classifying Based On Directed Graph(OCBDG), and designs a framework based on Web search engines in order to know about the content composition and evolvement of Web topics. It gets information about some topics from Web search engines, classifies the results into subtopics, and analyzes the relations between the subtopics and the evolvements of the subtopics. Experimental results prove that the framework can extract the subtopics in an about 70% precision and can show the evolvements of topics on Web in any time span truly and timely.
Key words:
directed graph,
classification,
Web search engine,
Web snippet,
snapshot
中图分类号:
王 巍;曾剑平;吴承荣;张世永.
基于网络搜索引擎的网络话题分析框架
[J]. 计算机工程, 2009, 35(3): 257-259,.
WANG Wei; ZENG Jian-ping; WU Cheng-rong; ZHANG Shi-yong. Framework of Web Topics Analysis Based on Web Search Engines[J]. Computer Engineering, 2009, 35(3): 257-259,.