knn最近邻算法在给定新文本后，考虑在训练文本集中与该新文本距离最近（最相似）的 K 篇文本，根据这 K 篇文本所属的类别判定新文本所属的类别，具体的算法步骤如下：



一、:根据特征项集合重新描述训练文本向量

二、:在新文本到达后，根据特征词分词新文本，确定新文本的向量表示

三、:在训练文本集中选出与新文本最相似的 K 个文本-knn nearest neighbor algorithm in the given text, to consider in the training text set with the new text from the nearest (most similar), K-chapter text, according to the category of K chapter text to determine the category of new text, the specific steps of the algorithm as follows: a: re-descr iption based on the feature set of training text vector to determine the vector of the new text: new text arrives, according to the new text of the characteristic word word said three: in the training text set to elect the most similar to the new text the K-text