摘要
基于网络信息检索,从理论上和实证上探讨单个关键词出现概率与信息量的关系。分析不同概率的检索词在需求表达信息量上的差异,在信息需求的多维描述基础上研究高频关键词在需求信息量上对低频关键词的排挤效应。针对这种排挤效应,结合叙词表词间关系提出了关键词归类去重的检索相关性测量方案。
This paper study the relationship between frequency and information quantity of single key-words both theoretically and empirically. We analysis the difference on demand expression of retrievalwords with different probabilistic. Research the crowding out effect of high frequency keywords to low fre-quency of keywords based on multidimensional description of information demand. Aiming at this effect,we provide a new relevant measurement program of retrieve in which before the calculate keywords will beclassified according to its relationships in the thesaurus.
出处
《情报科学》
CSSCI
北大核心
2015年第12期62-65,82,共5页
Information Science
基金
国家科技支撑项目(2012BAH90F03)
关键词
出现概率
信息量
信息需求维度
检索相关性
frequency of retrieval words
information quantity
dimension of Information needs
retrieval relevance