Effectiveness of Pre-training for Few-shot Intent Classification
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2021
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913499432615936 |
|---|---|
| author | Zhang, Haode Zhang, Yuwei Zhan, Li-Ming Chen, Jiaxin Shi, Guangyuan Lam, Albert Y. S. Wu, Xiao-Ming |
| author_facet | Zhang, Haode Zhang, Yuwei Zhan, Li-Ming Chen, Jiaxin Shi, Guangyuan Lam, Albert Y. S. Wu, Xiao-Ming |
| contents | This paper investigates the effectiveness of pre-training for few-shot intent classification. While existing paradigms commonly further pre-train language models such as BERT on a vast amount of unlabeled corpus, we find it highly effective and efficient to simply fine-tune BERT with a small set of labeled utterances from public datasets. Specifically, fine-tuning BERT with roughly 1,000 labeled data yields a pre-trained model -- IntentBERT, which can easily surpass the performance of existing pre-trained models for few-shot intent classification on novel domains with very different semantics. The high effectiveness of IntentBERT confirms the feasibility and practicality of few-shot intent detection, and its high generalization ability across different domains suggests that intent classification tasks may share a similar underlying structure, which can be efficiently learned from a small set of labeled data. The source code can be found at https://github.com/hdzhang-code/IntentBERT. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2109_05782 |
| institution | arXiv |
| publishDate | 2021 |
| record_format | arxiv |
| spellingShingle | Effectiveness of Pre-training for Few-shot Intent Classification Zhang, Haode Zhang, Yuwei Zhan, Li-Ming Chen, Jiaxin Shi, Guangyuan Lam, Albert Y. S. Wu, Xiao-Ming Computation and Language This paper investigates the effectiveness of pre-training for few-shot intent classification. While existing paradigms commonly further pre-train language models such as BERT on a vast amount of unlabeled corpus, we find it highly effective and efficient to simply fine-tune BERT with a small set of labeled utterances from public datasets. Specifically, fine-tuning BERT with roughly 1,000 labeled data yields a pre-trained model -- IntentBERT, which can easily surpass the performance of existing pre-trained models for few-shot intent classification on novel domains with very different semantics. The high effectiveness of IntentBERT confirms the feasibility and practicality of few-shot intent detection, and its high generalization ability across different domains suggests that intent classification tasks may share a similar underlying structure, which can be efficiently learned from a small set of labeled data. The source code can be found at https://github.com/hdzhang-code/IntentBERT. |
| title | Effectiveness of Pre-training for Few-shot Intent Classification |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2109.05782 |