本文通过短语结构限定的方式,全面地考察汉语名词性隐喻的分布和构成特点,总结和发现名词性隐喻的表达规律,同时在大规模语料考察的基础上,尝试规则约束和机器学习方法,探讨了汉语名词性隐喻的理解和识别。实验证明,机器学习方法在汉语名词隐喻识别中取得了较为满意的效果。 This paper examines the distribution and compositional characteristics of Chinese nominal metaphors in a comprehensive way through the definition of phrase structure. It summarizes and finds the expression of nominal metaphors. At the same time, on the basis of large-scale corpus investigation, this paper attempts to explore the rules and machine learning methods Understanding and Recognition of Chinese Nominal Metaphor. Experiments show that the machine learning method has achieved satisfactory results in the Chinese noun metaphor recognition.