In this paper, we emphasize the need for data cleansing when clustering large-scale transaction databases and propose a new data cleansing method that improves clustering quality and performance. We evaluate our data cleansing method through a series of experiments. As a result, the clustering quality and performance were significantly improved by up to 165% and 330%, respectively.
@article{arxiv.1004.4718,
title = {A Data Cleansing Method for Clustering Large-scale Transaction Databases},
author = {Woong-Kee Loh and Yang-Sae Moon and Jun-Gyu Kang},
journal= {arXiv preprint arXiv:1004.4718},
year = {2015}
}