Text Analytics

42.Given the Pivoted Length Normalization VSM and the BM25/Okapi equations. What are the following terms?

: c(w,q)

: avdl

: M

: df(w)

: b

16. For the following values, explain whether we can efficiently get the value from a default (standard) inverted index postings file. By efficiently, we mean with one lookup not scanning the entire index.

|d| - total number of terms in a given document

|d|u - number of unique terms in a given document

Df - number of documents a given term appears in

c(w, C) number of times w occurs in the corpus

c(w,d) count of w in d

p(w,d) probability of w occurring in d

Pivoted Length Normalization VSM 1 log df (w) NEgnd awdl b E 0.11 BM25 /Okapi M +1 (w, d) c(w, q) log df (w) we and (w,d) k (1 b b avedl