@esther this appears to be a bug of some sort. The words and their size relative to each other when displayed are correct, but what I’m seeing that is misleading is that we’re displaying one of the actual dimension values (i.e. AppDynamics_GFTS…) for each tokenized word with the Record Count (or whatever metric you’re using) grouped by that specific dimension value inside the tooltip. For example, that “AppDynamics_GFTS…” value likely appears 17 times in your data if you were to run another query.
I think ideally we would want to display the tokenized word (i.e. NSROOT) and the number of times that word appeared within that column. Is that what you were expecting?