Multi-Stage Pre-training for Low-Resource Domain Adaptation
Preprint
Abstract
Transfer learning techniques are particularly useful in NLP tasks where a sizable amount of high-quality annotated data is difficult to obtain. Current approaches directly adapt a pre-trained language model (LM) on in-domain text before fine-tuning to downstream tasks. We show that extending the vocabulary of the LM with domain-specific terms leads to further gains. To a bigger effect, we utilize structure in the unlabeled data to create auxiliary synthetic tasks, which helps the LM transfer to downstream tasks. We apply these approaches incrementally on a pre-trained Roberta-large LM and show considerable performance gain on three tasks in the IT domain: Extractive Reading Comprehension, Document Ranking and Duplicate Question Detection.Keywords
All Related Versions
This publication has 18 references indexed in Scilit:
- Multilingual Denoising Pre-training for Neural Machine TranslationTransactions of the Association for Computational Linguistics, 2020
- Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks2020
- BioBERT: a pre-trained biomedical language representation model for biomedical text miningBioinformatics, 2019
- Neural Duplicate Question Detection without Labeled Training Data2019
- SciBERT: A Pretrained Language Model for Scientific Text2019
- Simple, Scalable Adaptation for Neural Machine Translation2019
- Neural Domain Adaptation for Biomedical Question Answering2017
- Two-Stage Synthesis Networks for Transfer Learning in Machine Comprehension2017
- An overview of the BIOASQ large-scale biomedical semantic indexing and question answering competitionBMC Bioinformatics, 2015
- The Probabilistic Relevance Framework: BM25 and BeyondFoundations and Trends® in Information Retrieval, 2009