Information Gain and Clustering

Hi everyone, I noticed a node called Information Gain Calculator which (I guess) has more applications for text classification (supervised), however I was thinking if it can be used in clustering (unsupervised) to determine which variable “helped the most to split the data into groups” or in other words a kind of variable importance.?
I am attaching an example.

Thank you

Information_Gain.knwf (31.0 KB)