基于最大熵模型的汉语隐喻现象识别

Recognition of the Chinese Metaphor Phenomena Based on the Maximum Entropy Model

隐喻是我们日程生活中常见的语言现象,利用计算机识别隐喻已经成为自然语言处理、人工智能乃至应用语言学领域中的一个具有重要价值的研究课题。本文根据隐喻特点,基于最大熵原理建立了一个隐喻识别模型,并论证了利用统计手段建立该模型的合理性。实验结果表明,该模型具有较高的准确度和召回率,以及较为理想的f值,是非常有前途的。

Metaphor is a usual language phenomenon in our daily life,and recognizing them by the use of computers becomes a valuable research task in the fields of natural language processing, artificial intelligence, and even applied linguistics. This paper proposes a way to recognize metaphors based on the maximum entropy model after analyzing the features of metaphor, and reasons the rationality to build a recognition model using statistical methods. The results of the experiment show that the model performs well at a high precision and recall rate, as well as the f value, thus we come to the conclusion that such a metaphor recognization model based on the maximum entropy principle is promising.