Comprehensive Study on Zero-Shot Text Classification Using Category Mapping
Existing zero-shot text classification methods based on large pre-trained models with added prompts exhibit strong representational capacity and scalability but have relatively poor commercial applicability. Approaches that fine-tune smaller models using label mappings and existing datasets for zero...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10870154/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823859628457328640 |
|---|---|
| author | Kai Zhang Qiuxia Zhang Chung-Che Wang Jyh-Shing Roger Jang |
| author_facet | Kai Zhang Qiuxia Zhang Chung-Che Wang Jyh-Shing Roger Jang |
| author_sort | Kai Zhang |
| collection | DOAJ |
| description | Existing zero-shot text classification methods based on large pre-trained models with added prompts exhibit strong representational capacity and scalability but have relatively poor commercial applicability. Approaches that fine-tune smaller models using label mappings and existing datasets for zero-shot classification are simpler but suffer from weaker generalization capabilities. This paper employs three strategies to improve the accuracy and generalization of pre-trained models in zero-shot text classification tasks: 1) Utilizing a pre-trained model that transforms inputs into a standardized multiple-choice format. 2) Constructing a text classification training set using Wikipedia text data to fine-tune the pre-trained model; 3) Proposing a zero-shot category mapping method based on GloVe text similarity, using Wikipedia categories as substitutes for text labels. Without fine-tuning on the target labels, this method achieves performance comparable to the best models fine-tuned with target labels. |
| format | Article |
| id | doaj-art-9876ddcdd6cf46d39f5087f41c52b46f |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-9876ddcdd6cf46d39f5087f41c52b46f2025-02-11T00:01:02ZengIEEEIEEE Access2169-35362025-01-0113235262354610.1109/ACCESS.2025.353810310870154Comprehensive Study on Zero-Shot Text Classification Using Category MappingKai Zhang0https://orcid.org/0009-0004-1688-1815Qiuxia Zhang1Chung-Che Wang2https://orcid.org/0000-0002-1414-9848Jyh-Shing Roger Jang3Yiwu Industrial and Commercial College, Yiwu, Zhejiang, ChinaDepartment of Computer Science and Information Engineering, National Taiwan University, Taipei, TaiwanDepartment of Computer Science and Information Engineering, National Taiwan University, Taipei, TaiwanDepartment of Computer Science and Information Engineering, National Taiwan University, Taipei, TaiwanExisting zero-shot text classification methods based on large pre-trained models with added prompts exhibit strong representational capacity and scalability but have relatively poor commercial applicability. Approaches that fine-tune smaller models using label mappings and existing datasets for zero-shot classification are simpler but suffer from weaker generalization capabilities. This paper employs three strategies to improve the accuracy and generalization of pre-trained models in zero-shot text classification tasks: 1) Utilizing a pre-trained model that transforms inputs into a standardized multiple-choice format. 2) Constructing a text classification training set using Wikipedia text data to fine-tune the pre-trained model; 3) Proposing a zero-shot category mapping method based on GloVe text similarity, using Wikipedia categories as substitutes for text labels. Without fine-tuning on the target labels, this method achieves performance comparable to the best models fine-tuned with target labels.https://ieeexplore.ieee.org/document/10870154/Natural language processingpre-trained language modelzero-shot text classificationclassificationGloVe |
| spellingShingle | Kai Zhang Qiuxia Zhang Chung-Che Wang Jyh-Shing Roger Jang Comprehensive Study on Zero-Shot Text Classification Using Category Mapping IEEE Access Natural language processing pre-trained language model zero-shot text classification classification GloVe |
| title | Comprehensive Study on Zero-Shot Text Classification Using Category Mapping |
| title_full | Comprehensive Study on Zero-Shot Text Classification Using Category Mapping |
| title_fullStr | Comprehensive Study on Zero-Shot Text Classification Using Category Mapping |
| title_full_unstemmed | Comprehensive Study on Zero-Shot Text Classification Using Category Mapping |
| title_short | Comprehensive Study on Zero-Shot Text Classification Using Category Mapping |
| title_sort | comprehensive study on zero shot text classification using category mapping |
| topic | Natural language processing pre-trained language model zero-shot text classification classification GloVe |
| url | https://ieeexplore.ieee.org/document/10870154/ |
| work_keys_str_mv | AT kaizhang comprehensivestudyonzeroshottextclassificationusingcategorymapping AT qiuxiazhang comprehensivestudyonzeroshottextclassificationusingcategorymapping AT chungchewang comprehensivestudyonzeroshottextclassificationusingcategorymapping AT jyhshingrogerjang comprehensivestudyonzeroshottextclassificationusingcategorymapping |