基于属性拓展的数据挖掘预处理技术研究

目前的数据挖掘技术基本上依据的是原始数据库中的信息，数据预处理技术要维护原始数据库的信息基本不变，只是进行简单的数据标准化、数据平整、异常点发现、缺失数据修补、数据离散等基本预处理工作，不能从根本上拓展原始数据库中的信息。同时，为保密起见，兴起的隐私保护数据挖掘技术对原始数据库中的敏感数据进行处理，隐藏了一些基本信息，进一步弱化了原始数据库中的信息含量。基于属性拓展的数据挖掘预处理技术，从原始数据库出发，通过属性拓展，拓展基础数据库所蕴含的信息，使数据挖掘能产生更深的隐藏关联规则。

Research on Data Mining Preprocessing Technology Based on Attribute Extension

Nowadays the data mining technology is basically based on the information of the original database. Data preprocessing technol-ogy must maintain the information of the original database unchanged,just for simple data standardization,data level,abnormal points found,missing data repair work,data discrete and other basic data pretreatment,not radically expanded the information in the original da-tabase. Meanwhile,for the sake of secrecy,the privacy preserving of data mining technology deals with the sensitive data in the original database,and hides some basic information,further weakening the information content in the original database. The data mining prepro-cessing technology based on the attribute extension starts from the original database,expands the information implicated in basic database by extending the attribute from the original database perspective,so that data mining can produce more deeply hidden rules.