基于本体的关联规则挖掘算法研究  

Method of mining association rules based on ontology

作　　者：庞淑侠[1] 袁占亭[1] 张秋余[1] 李明[1] 

机构地区：[1]兰州理工大学计算机与通信学院,甘肃兰州730050

出　　处：《计算机工程与设计》2006年第10期1772-1774,共3页Computer Engineering and Design

基　　金：国家科技攻关基金项目(2001BA201A32);甘肃省自然科学基金项目(ZS021-A25-018-G)

摘　　要：在图像数据库中,针对图像的检索方法很多,但以传统的基于内容的图像检索方法效率低,而且误差较大。提出一种基于OBFM(ontology_basedfeedbackmodel)方法首先得到基本相似图像库,从中提取相应的特征元素,并在此基础上进行关联规则挖掘。该算法充分考虑了图像具有空间可重复性的特点,从而提高了图像检索的准确率。With the increasing of multimedia information, many papers are focus on image retrieval in several years. But there are always many errors from those. The paper puts forwards to a method based on OBFM, from which approximate image database are rcquired. The features got, association rules are extracted in approximate image database. The character of space reiteration is taken into account by the method. Good results are acquired well and truly.

关 键 词：图像检索 本体 关联规则 挖掘算法 图像数据库 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]