基于集束搜索算法改进的机器翻译系统研究

Research on machine translation system based on improved cluster search algorithm

机器翻译在当前社会中具有广泛的应用,对促进人们交流、提升工作效率等具有重要意义。研究对当前英语机器翻译系统进行改进,将集束搜索算法与语言模型相结合,并在其中引入长句惩罚措施。为防止机器翻译出现信息缺失、重复翻译等问题,研究通过对比实验对提出的模型进行性能验证,结果显示研究提出的模型在运行时间中可达到0.32 s;模型翻译精度达到0.93。通过训练集与测试集反应,模型的损失率可降低至0.08;模型PLEU评分可达到0.937分。结果表明,该模型无论在中译英还是英译中的情况下,均具有较好性能。研究为进一步提升机器翻译性能提供了新的思路。

Machine translation has a wide range of applications in the current society.It is of great significance to promote people’s communication and improve work efficiency.The research is to improve the current machine translation system,combine the cluster search algorithm with the language model,and introduce the penalty measures for long sentences.In order to prevent the problems of missing information and repetitive translation in machine translation,the research verifies the performance of the proposed model through comparative experiments,and the results show that the proposed model can reach 0.32 s in running time;The model translation accuracy reaches 0.93.According to the training set and test set,the loss rate of the model can be reduced to 0.08;The PLEU score of the model can reach 0.937.The results show that the model has good performance in both Chinese-English and Chinese-English translation.The research provides a new way to further improve the performance of machine translation.