Semi-supervised Relation Extraction via Incremental Meta Self-Training
arXiv:2010.16410
Abstract
To alleviate human efforts from obtaining large-scale annotations, Semi-Supervised Relation Extraction methods aim to leverage unlabeled data in addition to learning from limited samples. Existing self-training methods suffer from the gradual drift problem, where noisy pseudo labels on unlabeled data are incorporated during training. To alleviate the noise in pseudo labels, we propose a method called MetaSRE, where a Relation Label Generation Network generates quality assessment on pseudo labels by (meta) learning from the successful and failed attempts on Relation Classification Network as an additional meta-objective. To reduce the influence of noisy pseudo labels, MetaSRE adopts a pseudo label selection and exploitation scheme which assesses pseudo label quality on unlabeled samples and only exploits high-quality pseudo labels in a self-training fashion to incrementally augment labeled samples for both robustness and accuracy. Experimental results on two public datasets demonstrate the effectiveness of the proposed approach.
arXiv commentIn Findings of EMNLP 2021 as a long paper. Code and data available at https://github.com/THU-BPM/MetaSRE
References in corpus (6)
- Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
- Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results
- Relation Classification via Recurrent Neural Network
- Learning to Self-Train for Semi-Supervised Few-Shot Classification
- Self-Supervised Generalisation with Meta Auxiliary Learning
- Exploit Multiple Reference Graphs for Semi-supervised Relation Extraction