摘要
目前机器翻译无法进行高质量翻译,以提升机器翻译译文质量为目标,提出了用户反馈和模式识别相融合的机器翻译优化方法。首先提取不同用户置信度特征,建立用户置信度评价模型,通过评价模型区分不同用户反馈的翻译内容,将置信度评价结果较高的机器翻译结果利用模式识别算法进行优化,通过搜索算法判断关键词是否属于椭球公式,符合要求的关键词即为机器翻译最终优化结果。最后进行了仿真实验,结果表明,文中方法的机器翻译优化结果BLEU值、NIST值高,困惑度低,可以提高机器翻译质量。
At present,machine translation cannot perform high-quality translation.With the goal of improving the quality of machine translation,an optimization method of machine translation is proposed,which combines user feedback and pattern recognition.First of all,we extract the characteristics of different users’confidence,establish the evaluation model of users’confidence,distinguish the translation model fed back by different users through the evaluation model,optimize the machine translation results with higher confidence evaluation results by using pattern recognition algorithm,and judge whether the key words belong to ellipsoid formula through search algorithm.The key words that meet the requirements are the final optimization results of machine translation Finally,simulation experiments are carried out.The results show that the BLEUand NIST values of the machine translation optimization results of the method in the article are high and the degree of confusion is low,which can improve the quality of machine translation.
作者
岳佩
张浩
YUE Pei;ZHANG Hao(College of Humanities and Education,Shaanxi Energy Institute,Xianyang 712000,Shaanxi Province,China;No.202 Institute,Norinco Group,Xianyang 712000,Shaanxi Province,China)
出处
《信息技术》
2021年第1期126-130,共5页
Information Technology
关键词
用户反馈
模式识别
机器翻译
矢量量化
置信度
user feedback
pattern recognition
machine translation
vector quantization
confidence
