WHY DOES THE X MEAN CLUSTERING CHANGES THE NUMBERS OF ITEMS IN THE DESCRIPTION ?

Hi,

I was working on a data set and used the K means and then later tried X means too. Though, its quite interesting to notice that the Number of Items in the Description after running the process is completely doubled on using X means clustering !!! ( whereas it remained the same while using K means)

       This really fascinates me. I assume no matter what ever the type of clustering methodology we are using, rapidminer  should rather not manipulate the number of items in the description  . 

*does it have something to do with Overlapping lapping Clustering 

Any help would be greatly appreciated !:)

Thanks

Regards,
Preet