基于RNN汉语语言模型自适应算法研究

深度学习在自然语言处理中的应用越来越广泛。相比于传统的n-gram统计语言模型,循环神经网络(Recurrent Neural Network,RNN)建模技术在语言模型建模方面表现出了极大的优越性,逐渐在语音识别、机器翻译等领域中得到应用。然而,目前RNN语言模型的训练大多是离线的,对于不同的语音识别任务,训练语料与识别任务之间存在着语言差异,使语音识别系统的识别率受到影响。在采用RNN建模技术训练汉语语言模型的同时,提出一种在线RNN模型自适应(self-adaption)算法,将语音信号初步识别结果作为语料继续训练模型,使自适应后的RNN模型与识别任务之间获得最大程度的匹配。实验结果表明:自适应模型有效地减少了语言模型与识别任务之间的语言差异,对汉语词混淆网络进行重打分后,系统识别率得到进一步提升,并在实际汉语语音识别系统中得到了验证。

Research on a Self-Adaption Algorithm of Recurrent Neural Network Based Chinese Language Model

Deep learning is used more and more widely in natural language processing. Compared with the conventional n-gram statistical language model,recurrent neural network(RNN)modeling technology shows great superiority in the aspect of language model modeling,which is gradually applied in speech recognition,machine translation and other fields. However,most RNN language models are trained off-line at present,for different speech recognition tasks,there exist many language differences between training corpus and recognition tasks that affects the recognition rate of speech recognition system deeply. The authors adopt RNN model technical in training the Chinese language model and put forward a online self-adaption model training algorithm at the same time,with this algorithm,we treat the voice signal preliminary recognition results as the additional training corpus to retrain the model to ensure that the adaptive RNN model can match with different tasks mostly. The experiment results show that the self-adaption model can reduce the differences against recognition tasks effectively and further improve the system recognition rate after rescoring the lattice,it is proved in the actual Chinese speech recognition system at the same time.