Alibaba-NLP / HiAGM

Hierarchy-Aware Global Model for Hierarchical Text Classification
MIT License
209 stars 44 forks source link

How to get the vocab files of wos dataset? #6

Open liug0324 opened 3 years ago

liug0324 commented 3 years ago

It seems that preprocess_wos.py doesn't generate the files, word.dict and label.dict. I generate them myself by collecting every word appearing in the training set, but the test result is significantly worse than the result published. Could you tell me how you generate the vocab files?

gunquxuexia commented 8 months ago

看来 preprocess_wos.py 不会生成文件 word.dict 和 label.dict。我通过收集训练集中出现的每个单词来自己生成它们,但测试结果明显比发布的结果差。你能告诉我你是如何生成词汇文件的吗?

同样的问题,请问您解决了吗?