A Method for Data Quality Validation Based on Shapes Constraint Language

Abstract

The quality of data is a critical factor for all kinds of decision-making and transaction processing. Data quality validation is a key step in improving data quality, but the traditional data quality validation methods have received little attention to the semantic technology. In this paper, we propose a SHACL-based data validation method. With this method, the users can build ontologies to establish constraints in terms of classes, which is more intuitive. In addition, the constructed ontologies can be reused in describing the same thing, which avoid users from constructing duplicate constraint rules and reduces the labor cost. With semantic inference used in this method, it has stronger logical reasoning ability to find logical abnormal data that is difficult to find by traditional data quality validation methods, which can improve the completeness and correctness of validation. In order to verify the feasibility of the method, we construct the nucleic acid testing ontology and constraints, and complete the data quality validation of Chongqing nucleic acid testing system, which proves that the method is of great practical value.