基于上下文语义的朴素贝叶斯文本分类算法

朴素贝叶斯分类器基于样本各属性相互条件独立的假设前提，它作为一种简单的词袋模型，忽略了上下文语境下同义词对分类的影响。本文提出相似词概念，使用相似词词簇代替传统的特征词典参与训练。首先训练word2vec得到词向量。然后，将特征词典用词向量表示后层次聚类，构建相似词词簇，并对其扩展。实验结果表明，改进后算法有效提高了文本分类的准确度，避免了因分类训练语料的差异导致分类效果的不稳定。

ContextSemantic-basedNaiveBayesianAlgorithmforTextClassification

TheNaiveBayesclassifierisbasedontheassumptionthatthesamples’attributesareindependentoneanother.Asasimplebag-of-wordsmodel,itignorestheinfluenceofsynonymsincontexttoclassification.Thispaperproposestheconceptofsimilarwordandusesclustersofsimilarwordsinsteadofkeyworddictionaryintraining.First,word2vecistrainedtogetwordembedding.Second,thekeyworddictionaryisrepresentedbywordembeddingwhichisthenclusteredhierarchically,theclustersofsimiliarwordsarebuiltandexpanded.Theexperimentalresultsshowthattheabovemethodcanimprovetheaccuracyoftextclassification,andavoidtheinstabilityofclassificationeffectduetothedifferencesintrainingcorpus.