Relevant-Based Feature Ranking (RBFR) Method for Text Classification Based on Machine Learning Algorithm

Algorithm 1

RBFR.

Input: F= set of features in the text corpus

Output: S – top N rich features

Begin:

1 For each f in F

2  TPR score =

3  FPR score =

4 L = {top k₁ features with high TPR-FPR score}

5 For each f in L

6  If FPR(f)<TH then

7   Remove f from L

8  F1 = top N features from IG =

9  F2 = top N features from CHI =

10  F3 = top N features from Pearson Correlation

11  Common Features =

Return