中文摘要：

为了解决基于VSM方法在进行短文本分类时存在的严重数据稀疏问题,提出了基于语义与最大匹配度的短文本分类方法.以《知网》为知识源,设计了基于义原距离、义原深度与区域密度的义原相似度计算方法,实现基于词类的词语相似度计算;提出了基于语义与最大匹配度的方法计算短文本相似度,应用KNN算法进行短文本分类.实验结果表明,该方法与基于语义、基于AD_NB等方法相比,正确率、召回率和F值均得到了明显的提高.

英文摘要：

To deal with the serious data sparseness problem exists in the traditional VSM method of carrying out short text classification,a short text classification method based on the semantics and maximum matching degree is put forward.The primary similarity calculation method is designed based on the distance,the depth and area density.Word similarity calculation is carried out according to its part of speech and HowNet is utilized as a source of knowledge.The short text similarity calculation based on the method of combining the semantics and maximum degree is proposed.KNN algorithm is applied to the short text classification.Experimental results show that the precision,recall and F-measure are significantly improved in contrast with those of the method based on the semantics,AD_ NB and so on.