CurryTang / LLMGNN

Label-free Node Classification on Graphs with Large Language Models (LLMS)
67 stars 7 forks source link

不好意思,请问一下cora、pubmed、citeceer中节点embedding的原始输入是什么 #3

Closed valleysprings closed 11 months ago

valleysprings commented 11 months ago

主要网上好像没有公开的论文引用图原始数据,torch geometric下下来的embedding都是Bag-of-words的,请问作者是怎么获取原始数据的,因为只有获得原始论文的summary/abstract/full之后才能放进language model之后生成embeddings。谢谢作者QwQ

valleysprings commented 11 months ago

不好意思,没看到文章最后的部分,谢谢作者!