%0 Conference Paper %1 Dhillon:2003:IC:956750.956764 %A Dhillon, Inderjit S. %A Mallela, Subramanyam %A Modha, Dharmendra S. %B Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining %C New York, NY, USA %D 2003 %I ACM %K co-clustering information main thema thema:co-clustering theoretic %P 89--98 %R 10.1145/956750.956764 %T Information-theoretic co-clustering %U http://doi.acm.org/10.1145/956750.956764 %X Two-dimensional contingency or co-occurrence tables arise frequently in important applications such as text, web-log and market-basket data analysis. A basic problem in contingency table analysis is <i>co-clustering: simultaneous clustering</i> of the rows and columns. A novel theoretical formulation views the contingency table as an empirical joint probability distribution of two discrete random variables and poses the co-clustering problem as an optimization problem in <i>information theory</i>---the optimal co-clustering maximizes the mutual information between the clustered random variables subject to constraints on the number of row and column clusters. We present an innovative co-clustering algorithm that monotonically increases the preserved mutual information by intertwining both the row and column clusterings at all stages. Using the practical example of simultaneous word-document clustering, we demonstrate that our algorithm works well in practice, especially in the presence of sparsity and high-dimensionality. %@ 1-58113-737-0

@inproceedings{Dhillon:2003:IC:956750.956764, abstract = {Two-dimensional contingency or co-occurrence tables arise frequently in important applications such as text, web-log and market-basket data analysis. A basic problem in contingency table analysis is <i>co-clustering: simultaneous clustering</i> of the rows and columns. A novel theoretical formulation views the contingency table as an empirical joint probability distribution of two discrete random variables and poses the co-clustering problem as an optimization problem in <i>information theory</i>---the optimal co-clustering maximizes the mutual information between the clustered random variables subject to constraints on the number of row and column clusters. We present an innovative co-clustering algorithm that monotonically increases the preserved mutual information by intertwining both the row and column clusterings at all stages. Using the practical example of simultaneous word-document clustering, we demonstrate that our algorithm works well in practice, especially in the presence of sparsity and high-dimensionality.}, acmid = {956764}, added-at = {2013-04-08T11:03:49.000+0200}, address = {New York, NY, USA}, author = {Dhillon, Inderjit S. and Mallela, Subramanyam and Modha, Dharmendra S.}, biburl = {https://www.bibsonomy.org/bibtex/209da4110fbee857edabb7ecc64f68e02/becker}, booktitle = {Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining}, description = {Information-theoretic co-clustering}, doi = {10.1145/956750.956764}, interhash = {30fe4c22011ee3c5565d35709d9ce1f1}, intrahash = {09da4110fbee857edabb7ecc64f68e02}, isbn = {1-58113-737-0}, keywords = {co-clustering information main thema thema:co-clustering theoretic}, location = {Washington, D.C.}, numpages = {10}, pages = {89--98}, publisher = {ACM}, series = {KDD '03}, timestamp = {2013-04-19T09:34:55.000+0200}, title = {Information-theoretic co-clustering}, url = {http://doi.acm.org/10.1145/956750.956764}, year = 2003 }