Chapter 8, Problem 7PE

Program Plan Intro

Relevance of a set of terms

• Frequency is defined as the number of times a term occurs in a document.
• If the number of terms in a document is n(d) and the number of occurrences of term t in document d is n (d, t) then the total frequency(TF) is

  $\text{TF}\left(\text{d,t}\right)\text{ = log}\left(\text{1+n}\left(\text{d,t}\right)\text{/ n}\left(\text{d}\right)\right)$

• Inverse document frequency (IDF) is used to assign weights to a term.

  $\text{IDF}\left(\text{t}\right)\text{ = 1/n}\left(\text{t}\right)$

• The relevance of a set of terms is

  $r\left(d,Q\right)=\sum TF\left(d,t\right)\times IDF\left(t\right)$