Hi @radhkan
The row-level security dataset can be setup on an individual user basis or on a user group basis. Depending on how you want to implement you can either have UserName, GroupName or Both in your row-level-security dataset. The names have to be named exactly as documented here as they are key to how QuickSight does the match with User and Group names. Additionally you will include field(s) from your dataset that you want to use to restrict data access. Again the field name should match with the field name in your dataset.
As you see in the below documentation Region, Segment etc are fields in the dataset and you are trying to setup to which of those the user/group have access to.
When you are creating this dataset you should now use the New Rules Dataset option in the dropdown for New Dataset.
Hope this helps.
Regards,
Giri