Constrained Partitional Clustering of Text Data: An Overview - Text Mining: Classification, Clustering, and Applications - page 184

7.8 Conclusions
In this topic chapter, we gave an overview of different types of constrained
partitional clustering algorithms and how they can be used for improved text
clustering. We mainly focused on pairwise constraints and partitional clus-
tering algorithms that use these constraints in different ways (e.g., constraint
enforcement during inference, distance metric learning) for different distance
measures (e.g., cosine distance, Euclidean distance). There are other types of
constraints (e.g., size constraints on clusters (9)) and other categories of con-
strained clustering algorithms (e.g., hierarchical clustering (15), graph clus-
tering (29; 33)), which we could not cover in this chapter. Experiment results
on text datasets demonstrate that using constraints during clustering can sig-
nificantly improve the quality of the results, and also indicate that using the
cosine distance function is recommended for constrained clustering in the text
domain.
References
[1] R. Baeza-Yates and B. Ribeiro-Neto. Modern Information Retrieval .
ACM Press, New York, 1999.
[2] A. Banerjee, I. Dhillon, J. Ghosh, and S. Sra. Clustering on the unit
hypersphere using von Mises-Fisher distributions. Journal of Machine
Learning Research , 6:1345-1382, 2005.
[3] A. Banerjee, S. Merugu, I. Dhillon, and J. Ghosh. Clustering with Breg-
man divergences. Journal of Machine Learning Research , 6:1705-1749,
2005.
[4] A. Bar-Hillel, T. Hertz, N. Shental, and D. Weinshall. Learning distance
functions using equivalence relations. In Proceedings of ICML , pages 11-
18, Washington, DC, 2003.
[5] S. Basu, A. Banerjee, and R. J. Mooney. Semi-supervised clustering by
seeding. In Proceedings of ICML , pages 19-26, 2002.
[6] S. Basu, M. Bilenko, and R. J. Mooney. A probabilistic framework for
semi-supervised clustering. In Proceedings of ACM SIGKDD , pages 59-
68, Seattle, WA, 2004.
[7] T. De Bie, M. Momma, and N. Cristianini. Eciently learning the metric
using side-information. In Proc. of the 14th International Conference on