Abstract: Based on the statistical characteristics of Chinese maximal noun phrases (MNPs) in a Chinese corpus with 5 573 sentences,two efficient identifying algorithms for Chinese MNPs:(1) To identify MNPs by using boundary distribution probabilities; (2)
Abstract: Based on the statistical characteristics of Chinese maximal noun phrases (MNPs) in a Chinese corpus with 5 573 sentences,two efficient identifying algorithms for Chinese MNPs:(1) To identify MNPs by using boundary distribution probabilities; (2)
文摘:通过对包含5573个汉语句子的语料文本中的最长名词短语的分布特点的统计分析,提出了两种有效的汉语最长名词短语自动识别算法：基于边界分布概率的识别算法和基于内部结构组合的识别算法.实验结果显示,后者的识别正确率和召回率分别达到了85.4%和82.3%,取得了较好的自动识别效果.