Statis&cal Properties of Text (a) Zipf's Law expresses the proportion of term-occurrences associated with a term as a function of the term's frequency-based rank. Zipf's law can be stated as follows: Pt - the proportion of term-occurrences associated with term t r = the frequency-based rank associated with term t c = a constant (in English c = 0.10) What proportion of term-occurrences would be removed from the collection if we ignored all occurrences of the five most frequent terms in the collection? [10 points]