对北大语料进行地名前后次的抽取，通过设置一个阈值，来控制选择。-Names of the Beijing University corpus to carry out before and after the meeting of the extraction, by setting a threshold, to control the choice.

文本分类文本的选择似乎是根据内容来的， 而非一般所采用的语体分类-Text classification is based on the text of the choice seems to be content, rather than the general language used in Classification

基于隐马尔可夫模型的语音单字识别研究：本文针对线性模型在语音识别中的不足, 进行了隐马尔可夫模型(HMM)在 语音单字识别中的研究,主要对观察输出概率求解、 最佳状态序列寻找、 参数估计和 模型参数的选择进行了探讨.-Based on hidden Markov model speech word recognition: the lack of the linear model in speech recognition, hidden Markov model (HMM) speec