论文部分内容阅读
“语用意义”指话语在对话中传达出来的说话人意图。语用意义自动识别模式大体上可分为三类:一是以言语行为理论为基础的BDI(Belief/ Desire/Intention,信念/愿望/意图)模式,二是以会话发展和动态模型理论为出发点的DMS(Dialogue Management System,会话处理系统)模式;三是以概率理论为基础的Cue-based模式(线索模式)。本文拟从理论基础、研究方法、模型体系、算法等方面对上述三种模式予以评述。为面向中文信息处理的语用研究提供可借鉴的方法和理论。
“Pragmatic meaning ” refers to the speaker’s intention conveyed in dialogue. The patterns of automatic recognition of pragmatic meaning can be roughly divided into three categories: one is the Belief / Desire / Intention (BDI) model based on speech act theory; the other is from the perspective of conversation development and dynamic model theory (Dialogue Management System, Session Processing System) mode; the third is the Cue-based mode (clue mode) based on probability theory. This article intends to review the above three modes from the theoretical basis, research methods, model systems and algorithms. It provides a reference method and theory for pragmatic research oriented to Chinese information processing.