摘要
基于Transformer的神经机器翻译在高资源语言对上取得了巨大的成功,但是在资源稀缺的情况下模型翻译效果较差.针对旅游口语领域的俄汉低资源机器翻译,本文通过扩充领域内数据和多种微调方法改善目标领域翻译质量.首先利用多种语料过滤方法得到通用领域俄汉语料,然后在不同参数配置的Transformer模型上进行预训练,探索多种领域适应性微调方法,得到最优模型,最后在译码阶段采用模型平均和后处理生成翻译结果.实验结果表明,该方法可以有效提升低资源场景下的俄汉机器翻译效果.
Objective:To observe the effect of health education combined with psychological nursing intervention on mental state and sleep quality of cesarean section parturient.Methods:A total of 64 cases of cesarean section women admitted to the Third Hospital of Heilongjiang Province from July 2019 to July 2020 were selected as the research objects,and were divided into observation group and control group according to the random number table method,with 32 cases in each group.The control group was given routine nursing,the observation group was given health education combined with psychological nursing intervention.Self-rating Depression Scale(SDS)and self-rating Anxiety Scale(SAS)were used to compare the improvement of anxiety and depression before and after intervention.Pittsburgh Sleep Quality Index(PSQI)was used to compare the improvement of sleep quality before and after intervention between 2 groups,and the incidence of adverse reactions in 2 groups were compared at the same time.Results:After intervention,SAS score and SDS score of the observation group were significantly lower than the control group,PSQI score was significantly lower than control group,and the incidence of adverse reactions was significantly lower than control group,the difference between the two groups was statistically significant(Ps<0.05).Conclusion:Health education combined with psychological nursing intervention can effectively improve maternal adverse mood,improve sleep quality,reduce the incidence of adverse reactions,worthy of clinical application.
作者
刘欢
刘俊鹏
黄锴宇
黄德根
LIU Huan;LIU Junpeng;HUANG Kaiyu;HUANG Degen(College of Computer Science and Technology,Dalian University of Technology,Dalian 116024,China)
出处
《厦门大学学报（自然科学版）》
CAS
CSCD
北大核心
2022年第4期654-659,共6页
Journal of Xiamen University：Natural Science
基金
国家科技创新2030“新一代人工智能”重大项目(2020AAA0108000)。
关键词
神经机器翻译
低资源语言
领域适应
mRASP
Cesarean section
Health education
Psychological intervention
Nursing
Anxiety
Depression
Sleep
作者简介
通信作者:黄德根,huangdg@dlut.edu.cn。
