The Research Of Chinese Nominal Metaphor Computation

Metaphor which is a kind of ubiquitous phenomenon in natural language, is not only a philological and rhetoric problem, but also a kind of cognitive nature of human being. Metaphor helps us to construct our mind and concept system.Metaphor computation has attracted more and more attention because metaphor, to some extent, is the focus of mind and language mechanism. However, it encounters problems not only due to the rich expressive power of natural language but also due to cognitive nature of human being. Therefore machine-understanding of metaphor is now becoming a bottle-neck in natural language processing and machine translation.Chinese is one of the oldest languages in the world, Chinese Metaphor has some characteristics which are different from English Metaphor. Compared with development in English language Chinese metaphor computation is just at its starting stage. To improve the computation of Chinese Metaphor, this paper first presents a survey of current computational approaches, then suggests how a metaphor is understood in two ways, such as logical method and corpus method. Because there are various kinds of Chinese Metaphor, this paper concentrates on the research of Chinese nominal Metaphor.First, based on the epistemic logic, we put forward a new logical system which is related to metaphor understanding, including cognitive similar logic,cognitive dependence logic and cognitive comprehension logic. Then we expand the metaphor inference which can deal with fuzzy similar inference.Second, this paper brings forward two kinds of novel Metaphor computation which is based on corpus method. We use the cooperate mechanism of tenor and vehicle to calculate simple Chinese Nominal Metaphor, use semantic sense to catch the latent meaning of metaphor from the metaphor context. The experiments show the methods are effient.