search

ECMLPKDD / CapsRater-FeatureCapture

Feature Enhanced Capsule Networks for Robust Automatic Essay Scoring
3 stars 2 forks source link

5-fold cross validation #1

Open Elenore1997 opened 3 years ago

Elenore1997 commented 3 years ago

Hi, thanks for sharing the codes. But I have a question on the dataset you used. I saw that in your codes, you evaluate your method on extra valid and test dataset, have you ever tried to use 5-fold cross validation like previous works do? (eg: https://github.com/nusnlp/nea or skipflow, etc.)

ahaliu1 commented 2 years ago

Hi, I have the same question. In your code file data.py line 47 and line 49, you use the domain1_score of the validation set and test set. But in the data file, this data column is none. As far as I know. the validation and test golden label hasn't been released. So I wonder If you have the golden label of test set and validation set. Could you please share the data file?