摘要
文章根据心理语言学关于语言理解的主要理论框架,介绍并讨论了关于语音听辨、句法处理和语篇理解等三个方面计算机模拟的主要模型。在语音听辨阶段,研究者针对听辨过程中需要解决的切分和恒定问题提出了模糊逻辑模型和跟踪模型。句法分析重点关注句子歧义的问题,主要有移进——归约句法分析和联合空间两种模型。语篇理解的代表性模型包括命题模型和关联模型两种。
In this paper, the writer introduces and discusses the computational simulation of the three main steps of language comprehension: sound recognition, grammatical processing and text comprehension. In order to solve the problems of segmentation and invariance, researchers put forward fuzzy logical model of perception and trace model. Researches on syntactic processing are mainly concerned with the ambiguity of sentences and two representative models including shift-reduce parsing and unification space were developed. In the discourse comprehension, there are two models, which are proposition model and connectionist model of text comprehension.
出处
《清华大学学报(哲学社会科学版)》
CSSCI
北大核心
2004年第2期10-16,共7页
Journal of Tsinghua University(Philosophy and Social Sciences)
关键词
计算心理语言学
语言理解
计算机模拟
computational psycholinguistics
computational simulation
language comprehension