search

dengyang17 / unicorn

The implementation of Unified Conversational Recommendation Policy Learning via Graph-based Reinforcement Learning (SIGIR 2021).
45 stars 8 forks source link

A error in env_enumerated_question.py #7

Closed Snnzhao closed 2 years ago

Snnzhao commented 2 years ago

in updata_reachable_feature(): "self.reachable_feature = list(set(next_reachable_feature) - set(self.user_acc_feature)) " should be " self.reachable_feature = list(set(next_reachable_feature) - set(self.user_acc_feature)- set(self.user_rej_feature))"?

dengyang17 commented 2 years ago

This is not an error. It is different between binary question setting and enumerate question setting. In enumerate question setting, we have features and small features, where only small features will be deleted during the update, i.e., "self.reachable_small_feature = list(set(next_reachable_small_feature) - filter_small_feature - self.invalid_small_feature)". You may check line 333-335.

Snnzhao commented 2 years ago

I get it. Thanks for your answer.