Computing ranking scores in a search engine with the Inc.lt weighting scheme. Let the query be "good student" and the document be "good bad student good bad instructor. Fill out the empty columns in the following table and then compute the cosine similarity between the query vector and the document vector. In the table, df denotes document frequency, idf denotes inverse document frequency (i.e., idft = log10N/dft), tf denotes term frequency, log tf denotes the tf weight based on log-frequency weighting as shown in slides (i.e., 1+logiotft,d for tft.d> 0 and 0 otherwise), q is the query vector, q' is the length-normalized q, d is the document vector, and d' is the length-normalized d. Assume N = 10,000,000. query dfidftf log tf q 1000 document log tf d q ' tf terms bad good instructor student 10000 10 50000 The cosine similarity between q and d is the dot product of q' and d', which is: Computing ranking scores in a search engine with the Inc.lt weighting scheme. Let the query be "good student" and the document be "good bad student good bad instructor. Fill out the empty columns in the following table and then compute the cosine similarity between the query vector and the document vector. In the table, df denotes document frequency, idf denotes inverse document frequency (i.e., idft = log10N/dft), tf denotes term frequency, log tf denotes the tf weight based on log-frequency weighting as shown in slides (i.e., 1+logiotft,d for tft.d> 0 and 0 otherwise), q is the query vector, q' is the length-normalized q, d is the document vector, and d' is the length-normalized d. Assume N = 10,000,000. query dfidftf log tf q 1000 document log tf d q ' tf terms bad good instructor student 10000 10 50000 The cosine similarity between q and d is the dot product of q' and d', which is