Anchoring the Consistency Dimension of Data Quality Using Ontology in Data Integration

Data quality is crucial for data integration and the consistency dimension is an important issue in data quality. Traditional methods of data consistency focus on the conflict or inconsistency that occurs in the same concept. However, it is sometimes insufficient to ensure the data consistency only using these methods. In this paper, we divide the conflicts among different data sources into the traditional intra-concept conflict and the neglected inter-concept conflict based on ontology, and then we propose a detection model for these conflicts. Ontology mapping, including concept mapping and restriction verification, is the key issue in our model. We analyze the consistency dimension of data quality using the model. Both the classification and the model help us ensure the data consistency in data integration efficiently. Data from the third party and business processes of the applications can be used to resolve the inconsistency when conflicts are detected.