| 摘要: |
| 建立海洋数据仓库以实现海洋多源信息的汇集融合已成为趋势,数据清洗是数据仓库建立过程中的重要环节。文章介绍基于滤波和离差法的数据清洗规则,并以厦门市海洋水温数据为例,采用该规则方案对异常值进行批量探测和修正,得到较好效果,可为数据挖掘技术的研究和应用提供支持。 |
| 关键词: 滤波 离差法 数据仓库 数据清洗 数据挖掘 |
| DOI: |
|
| 基金项目:国家海洋局东海分局青年科技基金项目(201525). |
|
| Application of Data Cleaning Rules of Marine Data Warehouse |
| LI Zhiming,YU Rongjun,YE Furong |
| Xiamen Marine Forecast Station,SOA;Ningde Marine Environmental Monitoring Center,SOA |
| Abstract: |
| Building marine data warehouse to realize multisource marine information fusion has become the developing trend.Data cleaning is an important issue for the whole marine data warehouse.In this paper,a processing method based on filter and interpolation to clean data was introduced.With the benchmark of water temperature in Xiamen,the processing method was applied to detect outliers.The results proved the efficiency of the proposed method,which would provide support for the research and application of marine data in the future. |
| Key words: Filter,Interpolation,Data warehouse,Data cleaning,Data mining |