Closed rogerhome11 closed 4 years ago
@rogerhome11 Thanks for your interests in our work.
The weight in the file "uid_rid_pos_aid_weight.txt" represents the probability of the aspect belonging to this review. And the probability is generated by topic model (LDA), which is trained by all the reviews. We have give a brief instruction in the paper (Sec 4.4)
Merry Christmas Dr. zhao, sorry to disturb you. I have unzipped the amazon/yelp datasets and found the uid_rid_pos_aid_weight.txt in them. Are them the origin files in each datasets? If not , how did you calculate the weight for each user in each datasets?