基于非负矩阵分解特征提取的垃圾邮件过滤

随着垃圾邮件的不断增多,它的危害性越来越严重,为了消除这种危害性,垃圾邮件的过滤技术就显得异常重要。由于垃圾邮件数据具有稀疏性、高特征维数和多重相关性等,所以直接对它进行分类的话会造成运算量很大和错误分类的问题。本文针对这些问题,先用非负矩阵分解的方法对实验数据进行特征提取,然后再用分类方法对它进行分类。在实验中,比较之后发现经过分解之后的数据比没有经过分解的数据有更高的分类准确率。

Spam Filtering Based on Non-negative Matrix Factorization

The growing spam is more and more harmful to people,so the technology of filtering spam becomes extremely important.With data sparseness,high dimensionalities and multiple correlation of spam,its direct classification will lead to large computation and false classification.Therefore,the paper firstly uses the non-negative matrix factorization method to extract the features of the experimental data,and then classifies it.In the experiment,comparing results show that data of decomposition has high accuracy than original data.