Dimensional modeling of medical data warehouse based on ontology

The two driving forces for designing dimensional models of data warehouse are data sources and business requirements. Medical data sources are characterized by too many data types, large amount of data, lack of associations among data, mixed dictionary tables, etc., which result in lacking uniform representation and effective organization among a large number of concepts. These features may lead to serious semantic heterogeneity among different data sources. Medical data warehouse design involves a wide range of business requirements, while the users' perspectives are different and the expression of semantic heterogeneity is substantial. So how to obtain and integrate these business requirements is a challenge. Aiming at the characteristics of medical data sources and business requirements, this paper proposes an ontology-based medical data warehouse dimensional modeling method. Experiment shows that this method can not only optimize the data warehouse requirements analysis process, but also effectively eliminate the semantic heterogeneity in data sources and business requirements, thereby greatly improving the efficiency of medical data warehouse dimensional modeling.