A scoping review on ontology-driven mapping processes in healthcare and beyond: requirements and remaining challenges

The increasing variety of data produced in various domains, along with the evolution of data management and analysis methods, has led to the development of many data integration solutions. Most of these solutions use a common data model to harmonize the semantics and the structure of the data across multiple organizations. However, mapping data to a common data model still poses major challenges due to the multiple forms of heterogeneity. In this study, we examine papers that provide a detailed description of a mapping process between an ontology and a relational database. We conducted a scoping review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. In addition, we defined a list of requirements based on the concrete needs of the healthcare domain to analyze and compare the selected papers. Our findings provide a comprehensive overview of 15 papers. The investigation uncovered that none of the selected papers fully meets all the specified requirements. This involves maintaining consistent identification of entities, providing tools for documenting mapping decisions, and incorporating validation functions for testing the mapping process and its results. Future research should focus on designing a guided mapping process that guarantees the reproducibility and reliability of the generated data. Although motivated by the needs in the healthcare domain, the results are relevant to any domain where integrating distributed data is essential.