KDD中的数据清理技术研究

Research on data cleaning technique of KDD

数据清理是KDD的首要步骤;没有好的数据环境,就不会有理想的挖掘结果。介绍了数据的一般特征,讨论了KDD中数据清理技术的清除空缺、噪声处理及不一致数据等问题,指出通用性和自适应性差是目前数据清理工具存在的主要问题。

Data cleaning is a principal step of KDD; there is no ideal result without appropriate data environment. The general data character was presented. Some questions were discussed including the missing value cleaning, noise processing and unconformable data. And it was pointed that the improving of universal and adaptive properties was key of data cleaning technique.