基于HMM的维吾尔语词性标注研究

Research on Uyghur Part-of-Speech Tagging Model Based on Hidden Markov Model

在维吾尔语与汉语的机器翻译的研究中,词性标注起到很大的作用,词性标注也是自然语言处理的基础性工作。介绍基于隐马尔可夫模型的词性标注算法和词性标注器Citar,并且将Citar标注器应用到维吾尔语上进行词性标注。为了能对维吾尔语进行词性标注,在在布朗词性标注集的基础上,定义一套适用于维吾尔语的词性标注集,采用基于隐马尔可夫模型的方法,对部分维吾尔语进行词性标注实验,经过实验表明,Citar标注器能准确对维吾尔语进行词性标注,从而表明此标注器适用于维吾尔语。

The part-of-speech tagging plays a very important role in the research on machine translation in Uyghur and Chinese. The part-of- speech tagging is the groundwork for natural language processing. Introduces the part-of-speech tagging algorithm based on HMM and the part-of-speech tools named Citar, improves Citar in order to make the part-of-speech tagging tools apply to the Uyghur. On the basis of brown part-of-speech tagging sets, defines part-of-speech tagging sets used in the Uyghur for the part-of-speech tagging of Uyghur. Uses the method based on hidden Markov model, carried out the part of speech tagging experiment. The experiment result show that Citar has a good result on the part-for-speech tagging of Uyghur and the label machine is suitable for the Uyghur.