System Report for CCL24-Eval Task 7: Multi-Error Modeling and Fluency-Targeted Pre-training for Chinese Essay Evaluation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914866374115328 |
|---|---|
| author | Zhang, Jingshen Yang, Xiangyu Su, Xinkai Chen, Xinglu Huang, Tianyou Qiu, Xinying |
| author_facet | Zhang, Jingshen Yang, Xiangyu Su, Xinkai Chen, Xinglu Huang, Tianyou Qiu, Xinying |
| contents | This system report presents our approaches and results for the Chinese Essay Fluency Evaluation (CEFE) task at CCL-2024. For Track 1, we optimized predictions for challenging fine-grained error types using binary classification models and trained coarse-grained models on the Chinese Learner 4W corpus. In Track 2, we enhanced performance by constructing a pseudo-dataset with multiple error types per sentence. For Track 3, where we achieved first place, we generated fluency-rated pseudo-data via back-translation for pre-training and used an NSP-based strategy with Symmetric Cross Entropy loss to capture context and mitigate long dependencies. Our methods effectively address key challenges in Chinese Essay Fluency Evaluation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_08206 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | System Report for CCL24-Eval Task 7: Multi-Error Modeling and Fluency-Targeted Pre-training for Chinese Essay Evaluation Zhang, Jingshen Yang, Xiangyu Su, Xinkai Chen, Xinglu Huang, Tianyou Qiu, Xinying Computation and Language This system report presents our approaches and results for the Chinese Essay Fluency Evaluation (CEFE) task at CCL-2024. For Track 1, we optimized predictions for challenging fine-grained error types using binary classification models and trained coarse-grained models on the Chinese Learner 4W corpus. In Track 2, we enhanced performance by constructing a pseudo-dataset with multiple error types per sentence. For Track 3, where we achieved first place, we generated fluency-rated pseudo-data via back-translation for pre-training and used an NSP-based strategy with Symmetric Cross Entropy loss to capture context and mitigate long dependencies. Our methods effectively address key challenges in Chinese Essay Fluency Evaluation. |
| title | System Report for CCL24-Eval Task 7: Multi-Error Modeling and Fluency-Targeted Pre-training for Chinese Essay Evaluation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2407.08206 |