数字服务贸易中共同语言促进效应式微的实证  

An Empirical Analysis of Decreasing Positive Effect of Common Language on Digital Services Trade

作　　者：白树强[1] 肖雯 BAI Shuqiang;XIA Wen(School of International Trade and Economics,University of Intemnational Business and Economics,100029)

机构地区：[1]对外经济贸易大学,北京

出　　处：《制度经济学研究》2022年第4期180-201,共22页Research on Institutional Economics

摘　　要：基于深度学习算法和大数据驱动的人工智能技术演化推动世界产业数字化转型升级,全球数字服务贸易逆势高速增长。神经机器翻译智能应用的全球共享显著提高了多语言转换效率,基于贸易成本理论和随机前沿引力模型实证研究发现,2005~2019年共同语言对50国双边数字服务贸易的促进作用自2009年起逐年弱化,神经机器翻译技术的普及即时缓解了语言的工具性贸易障碍,但其对语言的文化性贸易障碍的缓解存在一定时滞性。数字鸿沟增加了南北国家对语言便利性技术获取的非均衡性和其数字服务贸易的国际市场份额差距。及时推广人工智能技术的产业规模化应用,加快培养人工智能复合专业型人才,有效依托人工智能提高双边文化交流效率将是推动中国数字服务贸易更高质量发展的政策参考。On the foundation of deep learning and big data,artificial intelligence has promoted the digitalized transformation of global industry.Especially in digital trade,the appearance and spread of neural machine translation has improved multilingual conversion efficiency.Based on trade cost theory and stochastic frontier gravity model,it is found that the positive trade effect of common language has been decreasing since 2009,which implies that the development of AI technology has instantly alleviated functional negative trade effect of language.However,there is some time delay in the alleviation of cultural negative trade effect of language by AI.“Digital divide”between developed and developing countries will exacerbate the imbalance of access to neural machine translation as well as gap of international market share.In order to improve competitiveness and quality of China's digital services trade,we propose three practical ways which include timely promotion large-scale industrial application of AI technology,accelerating cultivation of inter disciplinary talents of mastering AI and translation,and exploring effective combination of AI and current bilateral cultural exchange platform.

关 键 词：人工智能 数字贸易 机器翻译 服务贸易 

分 类 号：F746.18[经济管理—国际贸易] H085[经济管理—产业经济]