How are these three files OpSub_train.json, OpSub_dev.json and OpSub_test.json generated?

[gongshaojie12]

Hi,I would like to ask a question, how are these three files OpSub_train.json, OpSub_dev.json and OpSub_test.json generated? Thank you！

[wxjiao]

Hi,I would like to ask a question, how are these three files OpSub_train.json, OpSub_dev.json and OpSub_test.json generated? Thank you！

Sorry for the late reply. We first crawl the raw conversation data from OpenSubtile and conduct preprocessing (see our paper Appendix A.2 Dataset Creation). Then we save the pre-processed data into JSON files following the format of EmotionLine datasets (including Friends and EmotionPush).

[gongshaojie12]

OK,Thanks a lot!