双语信息处理方法及原理

BILINGUAL INFORMATION PROCESSING METHOD AND ITS PRINCIPLE

旨在阐述双语信息处理方法及其原理。该方法的特征在于人机双方都可按各自的特点来彼此协作或响应,具体含三个基本步骤:首先,从汉英翻译的困难发现问题;接着从跨学科、跨领域和跨行业的角度分析问题;最后,从人机分工与合作的角度解决问题。其结果是,不仅发现了大、小字符串并行的双重形式化进路,而且还发现并验证了广义双语信息处理方法的优点及其所遵循的科学原理。其意义在于:既可从理论上开阔思路,又可在实践中发挥作用。例如可兼容形式信息论的收敛性或封闭性与语义信息论的发散性或开放性;再如可采用算术和语文这样的广义双语信息处理方法驾驭强、弱两派的人工智能观点。不仅可很好地解释信息术语的约束条件,而且还可较为满意地规范不同的信息观和人工智能观的应用约束情景即兼顾人际、人机、机际、机人一系列双语应用的实际语境。

This paper aims to elaborate bilingual information processing method and its principle. The method is characterised by mutual collaboration or response of both human and machine according to their own characteristics. Specifically speaking, it involves three steps： First, identifying problems in Chinese-English translation difficulties; Then, analysing problems from the interdisciplinary and cross-sectorial point of view; Finally, solving problems from the perspective of human-machine division and cooperation. The result is that not only dual formal approach, parallel processing of both large string and small string, was found, but also the advantages of broad bilingual information processing method were found and its scientific principles were verified. Its significance lies in not only broadening the mind theoretically but also playing a role in practice. For example, convergence or closeness of formal information theory is compatible with divergence or openness of semantic information theory. In addition, broad bilingual information processing method of arithmetic and language can be adopted respectively to balance and surpass views of both strong AI and weak AI. Furthermore, not only constraints of information terminology can be well explained, but also applications＇ constrained scenarios of distinct viewpoints on information and artificial intelligence can be regulated accordingly. In another word, a series of actual contexts of bilingual applications can be balanced including the context between humans, human and machine, machines as well as machine and human.