S²NMF: Information Self-Enhancement Self-Supervised Nonnegative Matrix Factorization for Recommendation

Algorithm 1

Optimization algorithm of S²NMF objective Equation (16).

 Input: Observed user-item interaction matrix R  B;

 Output: Observed user-item interaction matrix ;

 Initialize:N_iter =1, maxIter =10, S = R;

 while N_iter < maxIter do

   Update according to algorithm 2: ;

   Update according to equation (3): S;

   if satisfy stop condition then

     Break;

   N_iter = N_iter+1;

 ;

 Return ;