Classifying News Corpus by self-organizing maps
Taqlow Yanagida, T. Miura, Isamu Shioya · 2004
In this paper, we introduce extended self organization map (SOM), called k-propagated SOM (K-SOM, or SOM(k)), and discuss how to classify text documents. Also we discuss how we evaluate classification capabilities of points on SOM (K-SOM) maps. We discuss some experiments to Reuters News Corpus datasets and show the usefulness of K-SOM.