基于多次权重鉴别的信息雷达及应用  

The Inforadar based on Multi-Ply Weight Discrimination and Its Application

作　　者：曹龄兮[1] 李建华[1] 娄悦[1] 

机构地区：[1]中南大学信息科学与工程学院,长沙410075

出　　处：《现代计算机》2006年第2期12-15,共4页Modern Computer

摘　　要：信息雷达不仅能有效发挥搜索引擎的数据挖掘、知识发现等功能,同时能较大地加强数字图书馆个性化服务的主动性、智能性优势。在信息过滤和预处理过程中,为了提高信息雷达区分相似主题的精度,提出了多次权重鉴别的方法,即将特征项权值评价函数分层次考虑,将相似主题作为下一层特征集,依此类推。本文分析了多次权重鉴别的一个特例——二次权重的基本思路和方法,介绍了基于多次权重鉴别的信息雷达在数字图书馆个性化服务体系中的应用。Inforadars are not only able to effectively unleash the function of search engines, including data mining knowledge discovery, but also improve the go-aheadism and intelligence of the individual service of digital library. In the process of information filtrating and pretreatment, in order to improve the precision of information classification, we put forward a multi-ply weight discrimination method, which considers the evaluating function of feature term in distinguishing the similar classes and makes the similar topics as lower feature term. The rest may be deduced by analogy. This paper introduces the fundamental thinking and method of an especial example of multi-ply discrimination--the second weight discrimination, and introduees the application of the tnforadar which is based on multi-ply discrimination in individual service of digital library from the practice of the procedure of developing projects.

关 键 词：信息雷达 个性化服务 多次权重鉴别 信息过滤 权重 雷达 鉴别 应用 数字图书馆 预处理过程 

分 类 号：TN956[电子电信—信号与信息处理]