Generalized SOMs with Splitting-Merging Tree-Like Structures for WWW-Document Clustering
Marian B. Gorzałczany, Filip Rudziński, Jakub Piekoszewski · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2015
This paper presents our clustering technique based on generalized SOMs with evolving splittingmerging tree-like structures and its application to complex clustering problems including some benchmark data sets and, first of all, WWW-document clustering.Our approach that works in a fully unsupervised way (i.e., without the pre-defined cluster number and using unlabelled data), automatically detects the number of clusters and generates multiprototypes for them.The collection of 548 abstracts of technical reports as well as its 476-element subset, both available at WWW server of the Department of Computer Science, University of Rochester, USA (www.cs.rochester.edu/trs) are the subjects of clustering.A comparative analysis with five alternative clustering techniques is also carried out.The reported results prove that our approach is a powerful tool (that outperforms several alternative approaches) for complex cluster-analysis tasks including the problems of WWW-document clustering.