The main contributions of this work are as follows: 

 - Proposing a new hybrid framework for the neighbors' ratings modeling and the prediction process within the belief function theory.  
- Building a trustworthy recommender that improves access and proactively provides relevant items to users by considering their preferences. It improves users' confidence and enhances the chance to succeed for such systems.  
- Helping the users making a better decision by allowing them to get a closer picture about their potentially future ratings. 
- Experiments on real world data sets to evaluate the effectiveness of our proposal.