The field of recommender systems (RecSys) has been extensively studied to
enhance accuracy by leveraging users' historical interactions. Nonetheless,
this persistent pursuit of accuracy frequently engenders diminished diversity,
culminating in the well-recognized "echo chamber" phenome