The task of this project is to mark the knowledge points of the Chinese college entrance exam , which belongs to the multi-label text classification task. This project is a learning-based project, mainly to deepen the understanding of the different classifiers through code practice. The dataset contains four main subjects in high school (Geography,History,Politics,Biology), and each subject has many different themes.
- My defined Fasttest: Micro_f1 0.54
- My defined Naive_bayes: Micro_f1 0.78
- TextCNN: Micro_f1 0.83
- Bert: Micro_f1 0.92
- ERNIE: Micro_f1 0.94