Title :
Text mining with information-theoretic clustering
Author :
Kogan, Jacob ; Nicholas, Charles ; Volkovich, Vladimir
Author_Institution :
Dept. of Math. & Stat., Maryland Univ., Baltimore, MD, USA
Abstract :
Motivated by the success of hybrid information-retrieval algorithms, the authors report on the development of their hybrid clustering scheme. Scheme experiments on data in a reduced vector space model indicate a higher performance level over several existing clustering algorithms.
Keywords :
data mining; information retrieval; pattern clustering; text analysis; hybrid clustering; hybrid information-retrieval algorithms; information-theoretic clustering; text mining; Bioinformatics; Clustering algorithms; Data mining; Information retrieval; Mutual information; Partitioning algorithms; Probability distribution; Random variables; Singular value decomposition; Text mining;
Journal_Title :
Computing in Science & Engineering
DOI :
10.1109/MCISE.2003.1238704