IJCAI填坑3 Recommender Systems with Social Regularization
recommendation systems with social regularization
主要解决问题:
(1) social RS 和 trust-aware RS的区别:This process of trust generation is a unilateral action that does not require user ut to confirm the relationship. This also indicates that user ui does not need to even know user ut in the real life.
(2) Secondly, trust-aware recommender systems are based on the assumption that users have similar tastes with other users they trust.
(3) More time on social
Social Regularization
- model 1: average-based regularization
- model 2: individual-based regularization
- similarity function
conclusion
In this paper, we only constrain user feature vectors while ignoring the item side. Actually, if we can find some cor- relations between the items, we can also incorporate item regularization terms into our framework to further improve the prediction accuracy. We can achieve this if we have so- cial tagging data in the recommender systems.