基于关联矩阵的短信自动分类

SMS Automatic Classification Based on Relational Matrix

短信自动分类是短文本研究的热点问题。针对此问题,提出了关联强度和关联矩阵特征提取方法,并设计了基于关联矩阵的全监督学习算法。为了实现系统的自我学习,探讨了基于关联矩阵的半监督学习算法,其结合了人工矫正的主动学习算法。最后通过实例验证说明了算法的有效性。

SMS automatic classification is a hot issue of short text study.In this problem,this paper put forward to the feature extraction method of relational strength and the relational matrix,and designed a fully supervised learning algorithm based on relational matrix.In order to implement the system of self learning,this paper also discussed a semi-supervised learning algorithm based on relational matrix,which combines with active learning algorithm of the artificial modification.Finally the experiment results illustrate the effectiveness and efficiency of this algorithm.