Title

Mining Distance-Based Outliers From Categorical Data

Keywords

Categorical data; Data mining; Distance-based outliers; Outlier detection; Similarity measure

Abstract

Distance-based outlier detection is an important data mining technique that finds abnormal data objects according to some distance function. However, when applying this technique to categorical data, a traditional simple matching dissimilarity measure is not an adequate model for high dimensional categorical data. In this article, we employ a new commonneighbor- based distance function to measure the proximity between a pair of data points. Experiments show that better outlier mining results can be achieved when the new distance function is utilized instead of a conventional simple matching dissimilarity measure. © CGU 2007.

Publication Date

12-1-2007

Publication Title

DESRIST 2007 Conference Proceedings - 2nd International Conference on Design Science Research in Information Systems and Technology

Number of Pages

75-88

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

Socpus ID

84880166024 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/84880166024

This document is currently not available here.

Share

COinS