基于TREC目标优化的过滤阈值调整算法  

Filtering Threshold Adjusting Algorithm Based on TREC Target Optimization

作　　者：樊康新[1] 

机构地区：[1]南通大学计算机科学与技术学,院南通226019

出　　处：《图书情报工作》2009年第23期107-110,127,共5页Library and Information Service

基　　金：南通大学自然科学基金资助项目“基于VSM的自适应信息过滤研究”(项目编号:08Z030)研究成果之一

摘　　要：检出阈值的优化调整是自适应信息过滤的重点和难点之一。分析现有的阈值调整方法中普遍存在的问题,以TREC效用指标为目标函数,对阈值调整方法中的极大似然估计法和局部优化法进行比较分析,提出基于TREC目标优化的全局极大似然估计法与局部效用指标优化相结合的自适应过滤阈值调整算法。实验结果表明该方法能有效地提高信息过滤系统的性能。Threshold optimization is one of the most important and difficult problems in adaptive information filtering. This paper points out general drawbacks of current threshold optimization methods. It takes TREC evaluation measure as target function, and compares maximum likelihood estimation with local target optimization in the threshold tuning. As a result, it presents an adaptive information fitering threshold adjusting algorithm combining maximum likelihood estimation and local utility target optimization based on TREC target optimization. The experiments result shows that this approach can improve the quality of information filtering effectively.

关 键 词：TREC目标优化 自适应 信息过滤 阈值调整 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]