Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/578891| Title: | Review of Context-Based Similarity Measure for Categorical Data |
| Authors: | Nurul Adzlyana M. S (UITM) Rosma M. D (UITM) Nurazzah A. R (UITM) |
| Keywords: | Categorical data Context-based Data mining Similarity measure |
| Issue Date: | Apr-2017 |
| Description: | Data mining processes such as clustering, classification, regression and outlier detection are developed based on similarity between two objects. Data mining processes of categorical data is found to be most challenging. Earlier similarity measures are context-free. In recent years, researchers have come up with context-sensitive similarity measure based on the relationships of objects. This paper provides an in-depth review of context-based similarity measures. Descriptions of algorithm for four context-based similarity measure, namely Association-based similarity measure, DILCA, CBDL and the hybrid context-based similarity measure, are described. Advantages and limitations of each context-based similarity measure are identified and explained. Context-based similarity measure is highly recommended for data-mining tasks for categorical data. The findings of this paper will help data miners in choosing appropriate similarity measures to achieve more accurate classification or clustering results. |
| News Source: | Pertanika Journals |
| ISSN: | 0128-7680 |
| Volume: | 25 |
| Pages: | 619-630 |
| Publisher: | Universiti Putra Malaysia Press |
| Appears in Collections: | Journal Content Pages/ Kandungan Halaman Jurnal |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_116307+Source01+Source010.PDF | 1.52 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.