系统功能语言学的知识表示方法研究

系统功能语言学应用于语篇生成系统中需要进行知识表示。通过考察PENMAM，发现其一种主要的表示方法是系统网络。然而，这种方法不能充分表示语言的交际性以及三元功能，在进行计算机实现时也存在一些缺陷。基于此，文章进一步提出改进系统网络表示法的两种基本途径:基于复杂性科学的理论途径和基于人工智能的技术途径。此研究能够促进系统功能语言学与其他学科之间的交叉、融合，从而更好地开展面向自然语言处理用途的语言研究。 

Research on the Knowledge Representation Method of Systemic Functional Linguistics

When systemic functional linguistics is used in natural language generation systems, it needs knowledge representa-tion. Focusing on Penman, this paper finds that a major representation method is the system network. However, this knowledge representation method cannot fully represent the communication goal of language and the organization principle of three meta-functions. Besides, there also exist some technical flaws when it is implemented in computer. Based on this, this paper puts forward two basic ways to improve the system network: the theoretical approach which is based on complex science and the technical approach which is based on artificial intelligence. Through this research, this paper is to develop the communication between systemic functional linguistics and other subjects, thus promoting the linguistic research oriented in natural language processing.