Categorizing counselors' intentions on Counseling Conversation Dataset (test)
0.9025AccuracyChinese RoBERTa-large classifier
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Chinese RoBERTa-large classifierBackbone=Chinese RoBERTa-large, Auxiliary task=Masked Language Modeling (MLM), Context=Conversation history, Aggregation=Averages across five random seeds2023.06 | 0.9025 | 0.8821 | 0.8446 | 0.8612 |