Implementation of Tibetan-Chinese translation platform based on LSTM algorithm

摘要：

　　With the rapid economic development and increasingly frequent language exchanges in Tibet,the traditional statistical machine translation methods are faced with problems such as lack of data and over-fitting of training,resulting in poor translation quality.Combined with the current development of natural language processing(NLP),the LSTM algorithm based on Google's TensorFlow framework is proposed to realize the Tibetan and Chinese neural machine translation method.In the preprocessing stage of corpus,word segmentation module is constructed for tibetan-chinese bilingual parallel corpus by using the combination algorithm of gated circulatory neural network(GRU)and conditional random field(CRF).In the model construction stage,the LSTM method is used to construct the model.The experimental results show that the short and long time memory networks have good translation effects.

关键词： bilingual parallel corpus GRU-CRF algorithm LSTM networks

作者: XiaoFeng Chen Hao Wang Wei Xiang

作者单位: Southwest Minzu University,Chengdu,Sichuan,China

会议类型: 国际会议

会议名称: 2019国图灵大会(ACM Turing Celebration conference-China 2019 )

会议地点: 成都

会议语种:英文

页码: 869-873

在线出版日期: 2019-05-17（万方平台首次上网日期，不代表论文的发表时间）