Research on Data Quality Evaluation Considering the Duality of Data Dimension and Quality Dimension
[Objective/Significance]This article constructs a dimensional evaluation model based on Shapley value to address the high cost issues caused by poor relevance and insufficient integrity in data development.The purpose is to evaluate the quality of each dimension of data and conduct value analysis.[Methods/Processes]The research first analyzes the literature related to data quality,and uses the modified Delphi method to construct a data quality evaluation index system.Then,the analytic hierarchy process is used to select indicators,and the Monte Carlo sampling method is used to calculate the Shapley value and obtain the dimensional evaluation results of data.In the case study of real estate data quality evaluation,the quality evaluation model is applied to evaluate 12 data dimensions.The results show that three dimensions have poor quality and are recommended for deletion.At the same time,suggestions for modification are proposed for other dimensions.[Limitations]This article does not address the difficulty of quantifying data quality dimensions,nor does it classify data,only proposing a general evaluation model.[Results/Conclusions]The research integrates data dimensions and data quality dimensions,and constructs a relatively complete data quality evaluation model based on the data quality evaluation index system and Shapley value.It can conduct more detailed quality evaluation of the database.
万方数据
维普
