As a user, I want to automatically select words that have a higher inverse document frequency, so that I can avoid words that have low page selection power.