中文摘要：

描述了基于统计的蒙古文自动词性标注系统的功能和总体结构,并对系统的性能进行测试.以规模为95万词的语料库作为训练语料,对5万词的测试文本进行一级词性标注,结果表明封闭测试和开放测试的准确率分别达到96.96%和96.79%

英文摘要：

The functions and overall structure of a Mongolian part-of-speech tagging system based on statistics method is presented,and its performance test is completed with 950000 words training corpus and 50000 words test text.The close and open test result of the pos tagging on level one shows that tagging accuracies are about 96.96% and 96.79% respectively.