Abstract
Recently, many researchers study various methods of agglomerative hierarchical clustering.
However, a method to handle asymmetric data matrices is not often discussed.
Then, we propose average linkage method that assume asymmetric data matrices
and text mining method with asymmetric measure using a multiset model.
Moreover, there is a
problem of reversals in the dendrogram in the output of agglomerative hierarchical clustering. If we have
a reversal, to obtain clusters from the dendrogram becomes difficult.
It is proved that the proposed linkage methods for asymmetric measures have no reversals in the dendrograms.
The effect of the proposed methods is shown by numerical examples.
