摘要
基于数据挖掘中决策树方法提出了一种新的W eb搜索策略。在该策略中,通过对预先采集的W eb页面进行学习得到一棵决策树,然后对其进行剪枝,得到简化布尔表达式形式的主题内在规律性信息,在其基础上进行查询修改,把修改后的查询转发到通用搜索引擎上,最终得到查询结果。实验结果表明,提出的查询策略对于特定主题的W eb搜索,查询结果的质量有明显的改善和提升。
A new Web search strategy based on decision tree method in data mining was proposed. In this strategy, a decision tree was obtained by studying the Web pages sampled in advance. Then the tree was pruned to get the inherent regularity of the subject in the form of simplified boolean expression, on the basis of which the query was modified and forwarded to general search engines, and finally the searching results were got. The experiment performed indicated that for subject-specific Web search, the quality of query results improved obviously.
出处
《计算机应用》
CSCD
北大核心
2006年第1期223-226,共4页
journal of Computer Applications
关键词
查询修改
决策树
信息检索
数据挖掘
机器学习
query modification
decision tree
information retrieval
data mining
machine learning
