基于HowNet的用户兴趣挖掘研究及应用  

User's Interest Mining Based on HowNet and its Application

作　　者：朱宝华[1] 

机构地区：[1]江苏商贸职业学院,南通226011

出　　处：《计算机与数字工程》2014年第10期1949-1951,1961,共4页Computer & Digital Engineering

摘　　要：搜索引擎中用户查询和网页资源之间的相似度研究一直是页面排序的研究核心。利用HowNet对词语的语义层次架构模型,对用户的检索词进行兴趣挖掘,同时对检索词和挖掘出的兴趣关键词的语义相似度计算方法进行改进,用户的检索请求与分块后的网页资源进行相似度迭代计算。实验结果表明,改进的算法使得页面排序的准确率和首页命中率有了较大提高。Research on similarity between user＇s query and web resource in search engine has been a very important is-sue on page ranking algorithm .The semantic architecture among terms are used to establish by HowNet ,mining user＇s re-trieval interest and a novel kind of term semantic similarity calculation is proposed ,the similarity calculation between user＇s query and web resource could be acquired through web segmentation iteratively .The experimental results show that the im-proved algorithm could acquire a high precision rate and first-page hitting rate compared with traditional algorithm .

关 键 词：HOWNET 用户兴趣挖掘 网页分块 页面排序 搜索引擎 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]