Multiview Ensemble Method for Detecting Shilling Attacks in Collaborative Recommender Systems

Algorithm 1

Feature extraction algorithm.

Input: D

Output: M

1  Initializing M

2  UR=conver(D)

3  US=converS(D)

4  if (PIS and NIS are not constructed) then

5  Calculate the novelty of items and construct the item novelty series

6  Construct the popular item set PIS and novel item set NIS

7  end if

8  if (TPI is not calculated ) then

9  Calculate the TPI of everyday

10  for each    do

11  decompose TPI by DWT

12  using wavelet reconstruction to obtain TPI of everyday

13   end for

14  end if

15  UT=converT(D)

16  for each    do

17  Calculate user features   according to formula ,-

18  end for

19  Return M