Application research of rough set and fuzzy evaluation in data credibility assessment

In the domain of education there are many investigations, such as teaching evaluation for teachers, learning evaluation for students, etc. Different student have different attitude towards these investigations. Thus some students evaluate at will and their data are not creditable. In this paper an approach for assessing data credibility is proposed. Firstly, the approach uses rough set to produce reductions and then calculates significance degree of attributes in a reduction, which is regarded as attribute weight. Secondly, data credibility is assessed by fuzzy evaluation, in which fuzzy relationship matrix is obtained by option reliability. At last, the approach is tested by an investigation data. The result shows it may assess data credibility and it is effective and feasible to many extents.