Term distribution-based initialization of fuzzy text clustering

Authors:
Krzysztof Ciesielski;Mieczysław A. Kłopotek;Sławomir T. Wierzchoń
Affiliations:
Institute of Computer Science, Polish Academy of Sciences, Warszawa, Poland;Institute of Computer Science, Polish Academy of Sciences, Warszawa, Poland and Institute of Informatics, Univ. of Podlasie in Siedlce;Institute of Computer Science, Polish Academy of Sciences, Warszawa, Poland and Institute of Informatics, Univ. of Gdansk, Gdansk
Venue:
ISMIS'08 Proceedings of the 17th international conference on Foundations of intelligent systems
Year:
2008

Citing 4
Cited 0

Introduction to Modern Information Retrieval

Introduction to Modern Information Retrieval
Combining multiple clustering systems

PKDD '04 Proceedings of the 8th European Conference on Principles and Practice of Knowledge Discovery in Databases
Towards adaptive web mining: histograms and contexts in text data clustering

IDA'07 Proceedings of the 7th international conference on Intelligent data analysis
Text data clustering by contextual graphs

DS'06 Proceedings of the 9th international conference on Discovery Science

Quantified Score

Hi-index	0.00

Visualization

Abstract

We investigate the impact of an initialization strategy on the quality of fuzzy-based clustering, applied to creation of maps of text document collection. In particular, we study the effectiveness of bootstrapping as compared to traditional "randomized" initialization. We show that the idea is effective both for traditional Fuzzy K-Means algorithm and for a new one, applying histogram-based cluster description.