摘要： 以自组织特征映射（Self-organizing map,SOM）算法作为理论基础，实现了对 文本文档的聚类，并且给出了相应的类别标注，同时对聚类的准确度进行了评价，由于准确 度较低而后提出了把tf·idf计算词权重的方法应用到SOM算法的输入文本向量中去，因此提 高了聚类的平均准确度。

关键词: 文本聚类, 自组织特征映射算法, 向量

Abstract: This paper reports the implementation of document clustering based o n Self-organizing map. The map can be labeled with automatically identified desc riptive words that convey properties of each area. The paper also evaluates the average accuracy of this method. Because of the low accuracy, it presents incorp orating the tf·idf term weighting method into the input document vectors to in crease the average accuracy.

中图分类号: 

• TP391.1