Shayona@SMM4H23: COVID-19 Self diagnosis classification using BERT and LightGBM models
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913185732231168 |
|---|---|
| author | Chavda, Rushi Makwana, Darshan Patel, Vraj Shukla, Anupam |
| author_facet | Chavda, Rushi Makwana, Darshan Patel, Vraj Shukla, Anupam |
| contents | This paper describes approaches and results for shared Task 1 and 4 of SMMH4-23 by Team Shayona. Shared Task-1 was binary classification of english tweets self-reporting a COVID-19 diagnosis, and Shared Task-4 was Binary classification of English Reddit posts self-reporting a social anxiety disorder diagnosis. Our team has achieved the highest f1-score 0.94 in Task-1 among all participants. We have leveraged the Transformer model (BERT) in combination with the LightGBM model for both tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_02158 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Shayona@SMM4H23: COVID-19 Self diagnosis classification using BERT and LightGBM models Chavda, Rushi Makwana, Darshan Patel, Vraj Shukla, Anupam Computation and Language Artificial Intelligence This paper describes approaches and results for shared Task 1 and 4 of SMMH4-23 by Team Shayona. Shared Task-1 was binary classification of english tweets self-reporting a COVID-19 diagnosis, and Shared Task-4 was Binary classification of English Reddit posts self-reporting a social anxiety disorder diagnosis. Our team has achieved the highest f1-score 0.94 in Task-1 among all participants. We have leveraged the Transformer model (BERT) in combination with the LightGBM model for both tasks. |
| title | Shayona@SMM4H23: COVID-19 Self diagnosis classification using BERT and LightGBM models |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2401.02158 |