数据清洗方法研究综述  

Review of Data Cleaning Methods

作　　者：张晨雪 肖祥慧 顾阳 ZHANG Chenxue;XIAO Xianghui;GU Yang(School of Mechanical and Electrical Engineering,Beijing Institution of Graphic Communication,Beijing 102600,China)

机构地区：[1]北京印刷学院机电工程学院,北京102600

出　　处：《北京印刷学院学报》2025年第3期49-55,共7页Journal of Beijing Institute of Graphic Communication

摘　　要：随着信息技术的迅速发展,数据总量呈指数级增长,涵盖了传感器网络、社交媒体、商业交易等多种来源。然而,噪声、缺失值、不一致性和冗余数据等问题对大数据分析的准确性构成了挑战。在此背景下,数据清洗成为提升数据质量的关键环节之一。本文阐述数据清洗的步骤和方法,针对脏数据的不同类型提出相应的数据清洗方法,并对数据清洗领域未来的发展方向与研究重点进行展望。With the rapid development of information technology,the total volume of data has grown exponentially,encompassing various sources such as sensor networks,social media,and business transactions.However,challenges to the accuracy of big data analysis arise from issues such as noise,missing values,inconsistencies,and redundant data.In this context,data cleaning has become a critical step in improving data quality.This paper outlines the steps and methods of data cleaning,and presents corresponding data cleaning approaches for different types of dirty data.It offers a prospect for future research directions and key areas of focus in the field of data cleaning.

关 键 词：数据预处理 数据清洗 缺失数据 脏数据 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]