一种基于包含关系的空间面对象条件离群检测算法

山东大学学报(工学版) ›› 2011, Vol. 41 ›› Issue (2): 91-95.

一种基于包含关系的空间面对象条件离群检测算法

顾珊1,吉根林1, 2*

1. 南京师范大学计算机学院, 江苏南京 210097;
2. 江苏省信息安全保密技术工程研究中心, 江苏南京 210097

收稿日期:2011-02-14 出版日期:2011-04-16 发布日期:2011-02-14
通讯作者: 吉根林（1964- ），男，江苏海安人，教授，博士，博士生导师，主要研究方向为数据挖掘技术及应用. Email: glji@njnu.edu.cn E-mail:glji@njnu.edu.cn
作者简介:顾珊（1987- ），女，江苏无锡人，硕士研究生，主要研究方向为数据挖掘技术. Email: gushan.871019@163.com
基金资助:
国家自然科学基金资助项目（40871176）

An algorithm for detecting conditional outlier
polygons based on inclusion relations

GU Shan1, JI Genlin1, 2*

1. School of Computer, Nanjing Normal University, Nanjing 210097, China;
2. Jiangsu Research Center of Information Security ＆ Privacy Technology, Nanjing 210097, China

Received:2011-02-14 Online:2011-04-16 Published:2011-02-14

摘要/Abstract

摘要：

提出了一种基于包含关系的空间面对象条件离群检测算法DCOP-IR(detecting conditional outlier polygons based on inclusion relations),该算法针对空间面对象的包含关系和非空间属性,定义了一种相似度度量方法,利用基于密度的离群查找算法，先检测出整体的空间离群对象,然后在一定的条件属性下检测空间条件离群对象。实验结果表明,算法DCOP-IR能准确地检测出满足一定条件属性的空间离群对象,并具有较高的效率。

关键词: 包含关系, 空间离群, 条件离群

Abstract:

A new algorithm DCOP- IR was proposed for detecting conditional outlier polygons based on inclusion relations. In this algorithm, inclusion relations and nonspatial attributes were used as similarity measurement criteria, and a densitybased outlier detecting algorithm was first used for detecting spatial outliers for all database, then it was used for detecting spatial outliers which satisfy some certain conditions. The experimental results showed that DCOP-IR was effective and efficient.

Key words: inclusion relations, spatial outliers, conditional outliers

顾珊1,吉根林1, 2*. 一种基于包含关系的空间面对象条件离群检测算法[J]. 山东大学学报(工学版), 2011, 41(2): 91-95.

GU Shan1, JI Genlin1, 2*. An algorithm for detecting conditional outlier
polygons based on inclusion relations[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2011, 41(2): 91-95.

参考文献

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed