(Binary Independence Model) Consider the following document-term matrix, where a 1 entry indicates that the term occurs in a document, and 0 means it does not: Assume that the number of non-relevant documents is approximated by the size of the collection and that the probability of occurrence in relevant documents is constant over all the terms in the query (specifically, p_i = 0.9). For each of the following queries, rank the documents in decreasing order of relevance. q1 = {t1, t2} q2 = {t3} q3 = {t2, t4}