A Framework for Mining Heterogeneous Dataset

13 Pages Posted: 9 Mar 2018

See all articles by J. Dhayanithi

J. Dhayanithi

Sona College of Technology

J. Akilandeswari

Sona College of Technology

Date Written: November 15, 2017

Abstract

Many have explored various clustering strategies to partition heterogeneous datasets with numeric, binary, nominal and ordinal attributes. The clustering algorithms seek to identify similar groups of objects based on the values of their attributes. These algorithms either assume the attributes to be of homogeneous types or are transformed to homogeneous types. In the real world, the dataset are often with heterogeneous nature. If those data are converted it leads to information loss. This paper proposes a hybrid methodology with unified similarity measure to cluster dataset of heterogeneous nature. The proposed hybrid clustering algorithm identifies the similar set of objects with heterogeneous data types without changing its characteristics. The empirical results show that the proposed clustering algorithm yields better results.

Keywords: Distance Measure, Hybrid Clustering, Heterogeneous Data, Fusing Technique

Suggested Citation

Dhayanithi, J. and Akilandeswari, J., A Framework for Mining Heterogeneous Dataset (November 15, 2017). Proceedings of the International Conference on Intelligent Computing Systems (ICICS 2017 – Dec 15th - 16th 2017) organized by Sona College of Technology, Salem, Tamilnadu, India, Available at SSRN: https://ssrn.com/abstract=3134291 or http://dx.doi.org/10.2139/ssrn.3134291

J. Dhayanithi (Contact Author)

Sona College of Technology ( email )

Junction Main Road
Suramangalam
Salem, Tamil Nadu 636005
India

J. Akilandeswari

Sona College of Technology ( email )

Junction Main Road
Suramangalam
Salem, Tamil Nadu 636005
India

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
38
Abstract Views
378
PlumX Metrics