中文摘要：

隐喻现象是思维及语言的中心问题，而隐喻的机器理解若不能得到有效解决，将会成为制约自然语言理解和机器翻译技术发展的瓶颈，因此，隐喻的计算化研究成为自然语言处理的一个重要问题。相对于英语隐喻计算研究，汉语隐喻计算研究才刚刚起步。根据本体、喻体和相似性关系三者的认知结构，可以把汉语隐喻分为九种类别；在此基础上，隐喻网络模型的建构为汉语隐喻句的机器识别和分类算法提供了形式化方法；此外，从解决逻辑全知问题和隐喻的语义真值角度提出的汉语隐喻逻辑系统，则为汉语隐喻句的释义提供了很好的支持。在今后的研究工作中，还需加强隐喻分类识别研究，建立隐喻属性知识库，构建汉语隐喻的认知类比描述与转绎系统，改造汉英机器翻译系统并对面向隐喻的搜索引擎进行升级。

英文摘要：

Metaphor, an ordinary everyday phenomenon of language use, has become the focus of mind and language mechanism. The comprehension of metaphor by machine will be a bottleneck problem in natural language understanding and machine translation. Therefore, great attention has been paid to computational mechanisms of metaphor within the last few decades, with respect to cognitive mechanism, machine understanding, machine recognition, application, etc. Although metaphor computation may encounter many problems for the rich expressive power of natural language and cognitive nature of human being, scientists still have achieved a lot on constructing computational models of metaphor. In fact, most of the current achievements are limited to the English language, whereas research on Chinese metaphor computation has just got started. In this paper, the author attempts to make a summary of the research results concerning Chinese metaphor computation that have been achieved in recent years, and suggests reconsidering the existing problems and limitations revealed in the study, and makes proposals for further research on computational mechanisms of metaphor in general. Specifically, achievements in the study of Chinese metaphor computation have been scored from three main perspectives, i. e., categorization architecture, machine recognition and classification, logic representation and interpretation of metaphor. Firstly, based on the view of cognitive structures on the relations among tenor, vehicle and the similarity between them in a metaphor, we have put forward a system of Chinese metaphor categorization architecture, according to which there are altogether nine types of Chinese metaphor. Statistical verification and linguistic explanation are also proposed to support this categorization. Secondly, we have constructed a computational model of metaphor representation, called metaphorical semantic network （MSN）, according to the characteristics of metaphor. This model offers methods to formalize Chinese sentences, and