A Cross-Domain Collaborative Filtering Algorithm Based on Feature Construction and Locally Weighted Linear Regression

Algorithm 2

The FCLWLR algorithm.

Input: the incomplete rating matrix corresponding to

Output: the complete rating matrix

Filter out the top rated items in each auxiliary domain ( may be different across different

auxiliary domains) to obtain denser sub-matrices;

Use N-CF-U algorithm to fill the missing ratings in the sub-matrices;

Feature construction in the target domain;

Feature construction in the auxiliary domain;

Convert the recommendation problem into a regression problem;

Train a regression model on the obtained training set based on Algorithm 1;

Predict the missing ratings in the target domain.