Case Study on Modeling Approaches and Framework of Scientific Data Cleaning

This paper is the case study on scientific data cleaning, it proposes some new ideas or applies some hot technologies for scientific data cleaning. There are three challenges of scientific data cleaning tool: domain knowledge representation and usage, customized cleaning flow and building dynamically. We adopt knowledge based rule modeling, workflow based flow modeling and pluggable components based cleaning framework to solve the problems. Proposed approaches have being used in a project which faces to oceanography data cleaning. Theories and practices prove that the proposed approaches and framework are contributed to build a flexible and extendable data cleaning tool.