about the id mapping

[xiangwang1223]

about the id mapping

[xiangwang1223]

Thanks for your interest! Sorry for causing the confusion.

1. Each triplet in kg_final.txt is composed of (head-entity-id, relation-id, tail-entity-id). Here entities include items and other KG entities. That is, the item entity id starts from 0 to (#items-1) (say, 99); the other KG entity id starts from #items (say, 100) to #all entities (say, 1000).
2. As for the question from @yuanyuansiyuan, if you consider features or profiles of users (i.e., organizing users as triples in kg_final.txt), you need to change the data loader. Our code so far only takes item knowledge into consideration.
   Thanks.

[ZhichaoOuyang]

I also want to ask the same question.
I hope the author has free time to answer.
Thanks!

[ZhichaoOuyang]

Thank you for your patience, I understand now.