2. The experimental design

A : Training users
B : Active users
C : Training items
D : Active items
E : Data for calculating the similarities between A and B in the improved CF approach
F : Test data set
Figure. Division of the experimental data set for the CF approach.
3. R Shiny GUI instruction
- Upload a csv file comprising zeros and ones
- Only the first 20 users are selected as training users(A), while the first 80% of the items is selected as training items(C).
- The performance measure is based on the precision, which is generally used in information retrieval research and defined by
Precision = Hitting number/Top-N
where Top-N is the number of first N items that are recommended by a CF scheme and ‘Hitting number’ is the actual Top-N obtained from the section F.
- The precision for Top-N (N=1,...,10) is calculated.