基于RNN编码器的交互式平台控制技术开发
2020-12-08吕娜
吕娜



摘 要:自然语言是人类智慧的结晶,自然语言处理是人工智能中最为困难的环节之一。智能化时代的来临给人们带来了不一样的体验,人机对话、人脸识别、在线翻译技术等为现代交互式平台控制技术的发展奠定了基础。利用智能技术获取数据、处理语言、实现与智能机器间的交流成为可能。文章对RNN编码器交互式翻译平台技术开发进行研究,结果显示该交互式平台在RNN编码器嵌入后稳定性和控制能力有一定程度的提升,减少了交互机在工作时被外界干扰的信息,以日汉互译为例对基于RNN编码器的交互式平台控制技术应用进行了论证。
关键词:RNN编码器;交互式平台;控制技术
中图分类号:TP391.2 文献标识码:A 文章编号:1001-5922(2020)10-0162-03
Abstract:Natural language is the crystallization of human intelligence. Natural language processing is one of the most difficult problems in artificial intelligence. The coming of intelligent era brings people different experiences. Human-computer conversation, face recognition, online translation technology and so on lay the foundation for the development of modern interactive platform control technology. It is possible to acquire data, process language and communicate with intelligent machine by using intelligent technology. This paper studies the development of RNN encoder interactive translation platform technology. The results show that the stability and control ability of the interactive platform are improved to a certain extent after the RNN encoder is embedded, which reduces the information that the interactive machine is interfered by the outside world when it is working. Taking Japanese Chinese translation as an example, this paper demonstrates the application of RNN encoder based interactive platform control technology.
Key words:RNN encoder; interactive platform; control technology
0 引言
目前,國内外对于交互式翻译平台控制开发领域研究的比较彻底,技术也越来越成熟[1]。随着国家经济飞速发展,我国与日本等国之间的沟通愈来愈多,针对沟通中的语言障碍,可以将交互式平台装置安装在计算机上解决语言问题。为保证交互式平台工作的准确性,当语言信息发生变化时,该平台也会随之改变[2]。虽然这种交互式翻译控制平台在发展过程中得以应用,但在还是有一些难以攻克的问题,比如工作时容易受到外界的干扰、控制的精准度不达标等。文章主要是以RNN编码器作背景,以研发理论技术为前提,对交互式翻译平台的结构和稳定控制技术做了深切研究,以日语翻译为例对平台的实际应用情况进行了考量。……
