自动化学报2018，Vol.44Issue(5)：891-900,10.DOI:10.16383/j.aas.2018.c170464
基于对抗训练策略的语言模型数据增强技术
Data Augmentation for Language Models via Adversarial Training
摘要
Abstract
The conventional approach to data augmentation for language models based on maximum likelihood estima-tion(MLE)causes the exposure bias problem,which leads to generated text lacking of long-term semantics. We propose a novel data augmentation approach via adversarial training,which uses a convolutional neural network as a discriminator to guide the training of a recurrent neural network based generative model. The matter of augmentation for language models can be regarded as discrete sequential data generation. When outputs of the generative model are discrete, backforward propagation algorithm fails to update the generative model via the gradient of discriminator errors. To deal with this problem, we treat the generative model as a stochastic policy in reinforcement learning and optimize it by rewards from the discriminator. Since the discriminator can only judge completed sequences,we evaluate intermediate states by Monte Carlo search. Experiments on rescoring the n-best lists of speech recognition outputs show that with the increase of train-ing corpus, the proposed approach achieves a lower character error rate (CER) and always outperforms the MLE-based approach. When training corpus reaches 6 million tokens,the proposed approach provides a relative 5.0 % CER reduction on THCHS 30 dataset and a relative 7.1 % CER reduction on AISHELL dataset compared with the baseline.关键词
数据增强/语言模型/生成对抗网络/强化学习/语音识别Key words
Data augmentation/language modeling/generative adversarial nets(GAN)/reinforcement learning/speech recognition引用本文复制引用
张一珂,张鹏远,颜永红..基于对抗训练策略的语言模型数据增强技术[J].自动化学报,2018,44(5):891-900,10.基金项目
国家自然科学基金(11590770-4,U1536117,11504406,11461141004),国家重点研发计划(2016YFB0801203,2016YFB0801200),新疆维吾尔自治区科技重大专项(2016A03007-1)资助 Supported by National Natural Science Foundation of China(11590770-4,U1536117,11504406,11461141004),National Key Research and Development Plan(2016YFB0801203,2016YFB0801200),and Key Science and Technology Project of Xinjiang Uygur Autonomous Region(2016A03007-1) （11590770-4,U1536117,11504406,11461141004）
