摘要
信息过滤技术的目的是滤除信息垃圾、建立主动服务机制。论述了信息过滤技术中的原理、关键技术及其实现方法。采用话题为基本单位,分析了信息的特征,建立的用户兴趣模型;根据用户反馈来动态地了解用户的兴趣变化,提出了潜在语义标引的过滤算法,并依据用户设定的不良信息库更有效地过滤出用户所需要的信息。利用VisualJ++(VJ)实现了信息过滤软件基本功能,实验证明上述过滤算法的有效性。最后给出了在信息过滤技术中正在研究的几个问题。
The purpose of information filtering technology is to filter the information garbage and provide the active services. The subject discusses some concepts, principles and implementation of important technology in the information filtering system. The topic of conversation is the elementary unit. And the information feature is analyzed; simultaneously the interesting profile of client users is established. At the same time, the feedback ntework is performed to learn user interests and latent semantic indexing is applied in filtering the information. So the interesting diversity can be understood dynamically, and the information of the user interest can be filtered efficiently. Utlizing Visual J++, the basic function of the information filtering software is brought into effect.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2004年第3期382-385,共4页
Systems Engineering and Electronics
基金
国家自然科学基金(60272014)
国家"836"高技术计划(2002AA121063)基金资助课题
关键词
信息过滤
特征提取
兴趣模型
潜在语义标引
information filtering
feature extraction
interesting profile
latent semantic indexing