As a developer, I want to be able to calculate the inverse document frequency, so that I can select words with better discrimination power.