ganjinzero/icd-msmn

Code Synonyms Do Matter: Multiple Synonyms Matching Network for Automatic ICD Coding [ACL 2022]

icd-msmn's Issues

Missing method "clean" in load_umls.py ?

I can't find the method called clean in class UMLS and it reports an error when initializing the class UMLS.
In load_umls.py, clean_string = self.clean(string, clean_bracket=False)
Besides, in match.py, icd_dict[icd] = [desc_dict[icd]] + umls.icd2str(icd)
It calls the method icd2str, which is also missing in UMLS.

Where does preprocessing use match.py

I studied the code in preprocess folder. I could understand how description of the code and synonyms are combined. But I didn't find the result is used in the real training process. Because icd_dict{} generated in generate_data_new.ipynb isn't the one in match.py. Where is the result of match.py used? Or which file import match.py? Do I need to preprocess by myself according to the code in preprocess and then generate new data?

Thanks,
Best regards.

Pre-trained model

Is there any plan to release pre-trained model ?

Lack of 'word2vec_sg0_100.model.wv.vectors.npy’ in Word Embedding

Dear Author: @GanjinZero Lacking of 'word2vec_sg0_100.model.wv.vectors.npy’ in Word Embedding，while word2vec_sg0_100.model is ready.

A mistake: line 118 in main.py in run

I met a mistake: line 118 in main.py, should i rewrite it like best_train_metric, _, threshold_ = eval_func() ?

PS: line 83 in main.py, the same mistake

.

Can you please upload the full code?

Not able to convert yhat or yhat_raw outputs into icd codes

Hi. Thanks for the great work. Im trying to write a inference script for this repo. But not able to understand how to convert yhat or yhat_raw predictions into ICD codes. If possible can you please suggest how to go ahead.
Thanks.

Question about R-Drop

If one does not want to use R-Drop but just binary cross entorpy, then one could remove rdrop_alpha and set it to 0?

label_feat cannot be reshaped to [12, 4, 512]

Code Synonyms Generator

Where can we find the code generated the code synonyms from UMLS?