摘要
人们学习说话的教育过程就像是在语言的超级棋盘暨形式化字符组成的单词矩阵中做各种选择。提出一种必然涵盖所有言语(话语或术语)的新方法:构建语言的集合暨字符棋盘或词表;通过人机交互和协作,生成大量的言语(话语或术语)大数据,涵盖代表知识本体的各种话语或术语;通过机器学习和人机交互过程,比较、查询或重复使用这些话语或术语。结果表明,话语或术语的选取都可通过双语或多语转换以多种方式自动查询。该方法不仅可用于创建大数据与人机对话的环境平台,还可用于智能化文本分析和知识模块精加工,从而搭起大数据与知识大生产的桥梁。
The educational process of learning to speak is like making choices in a word matrix of formal characters or a language s super chessboard. This paper introduced a new approach that must cover all speech(discourse or terminology). We constructed a single set of languages, that is, a character or vocabulary matrix. Through human-computer interaction and collaboration, a large number of big data of speech(discourse or terminology) were generated, covering all kinds of discourse or terminology representing knowledge ontology. We compared, queried or reused these words or terms through machine learning and human-computer interaction. The results show the choice of discourse or terminology can be automatically queried in a variety of ways through bilingual or multilingual conversion. This method and can be used not only to create a platform for big data and human-machine dialogue environment, but also for intelligent text analysis and knowledge module finishing, so as to bridge the gap between big data and knowledge big production.
作者
邹晓辉
王肖群
邹顺鹏
Zou Xiaohui;Wang Xiaoqun;Zou Shunpeng(Subject Group,Sino-American Searle Research Center, Beijing 100871, China;Peking University Teacher Teaching Development Center, Beijing 100871, China;Higher Education Group, Sino-American Searle Research Center, Beijing 100083, China)
出处
《计算机应用与软件》
北大核心
2019年第9期186-191,共6页
Computer Applications and Software
基金
中美合作创新基金项目(SSRC2018-2019-01c)
关键词
大数据
软件
人机对话
机器学习
Big data
Software
Man-machine dialogue
Machine learning