一种基于粗集理论的文本特征约简新方法

A New Method to Select Eigenvectors from Documents Based on Rough Set Theory

信息过滤技术是解决“信息过载”和“信息迷向”问题的有效手段。为了高效地确立用户的信息需求模型,在粗集理论属性约简技术的基础上,提出主题特征选择的新方法RSAR。RSAR方法有效地克服了传统粗集方法不能直接处理连续值属性的缺陷,依据相对核和属性重要度从示例文本中抽取特征词,从而建立用户模型。实验验证了RSAR方法的有效性。

Information filtering（IF）is one of the methods that are rapidly evolving to manage large information flows.The aim of IF is to expose users to only information that is relevant to them.To build user model effectively ,this paper puts forward a novel algorithm-RSAR,which is based on reduction and core theory of Rough Set.By considering the continuous value attributes directly,RSAR can obtain the eigenvectors from the example set of documents selected by user effectively.Then,these eigenvectors will be used to construct user model.Simulation result illustrates the efficiency of RSAR.