Data Warehouse Design Methods Review: Trends, Challenges and Future Directions for the Healthcare Domain

Abstract

Abstract. In secondary data use context, traditional data warehouse design methods don't address many of today's challenges; particularly in the healthcare domain were semantics plays an essential role to achieve an effective and implementable heterogeneous data integration while satisfying core requirements. Forty papers were selected based on seven core requirements: data integrity, sound temporal schema design, query expressiveness, heterogeneous data integration, knowledge/source evolution integration, traceability and guided automation. Proposed methods were compared based on twenty-two comparison criteria. Analysis of the results shows important trends and challenges, among them (1) a growing number of methods unify knowledge with source structure to obtain a well-defined data warehouse schema built on semantic integration; (2) none of the published methods cover all the core requirements as a whole and (3) their potential in real world is not demonstrated yet.

Generating Relational Database using Ontology Review

“…This implies better management of missing information, and participation of individuals according to the property. In addition, the normalization is even more important in the context of physical data warehousing (especially for temporal database and big data) or virtual data warehousing (mediation) where the data extracted from multiple sources are heterogeneous, highly fragmented and context dependent [17], [32]. A -high‖ normal form (like 5NF and 6NF) reduce uncontrolled redundancy and facilitate schema extension as every part of the schema represents one predicate [33].…”

“…This implies changes to ontologies with ensuing repercussions on the related relational schema. The schema must, therefore, cope with it while maintaining earlier knowledge interpretations and preserving coherent data [17]. Moreover, with the opportunity to easily access data, new needs will emerge, and existing needs may change.…”

“…Yet, since ontologies offer excellent data integration support for disparate systems, a relational database derived directly from an ontology can be hypothesized to be the best way to ensure data integrity and a unique data access point for a large volume of data. First, modeling a relational database using ontologies (rather than entity-relationship) allows the reuse of ontology in many tasks facilitating queries expression as well as ensuring semantic and structural uniformity [16], [17]. Second, with the axiomatic model underlying the ontology, data storage and verification can be automated, increasing data integrity.…”

Clinical Data Warehouse a Review

“…However, if the location of the CDW is outside of hospitals, integration with non-clinical data may be easier. Several factors are required to deal with barriers for implementing a CDW, such as data integrity, sound temporal schema design, query expressiveness, heterogeneous data integration, knowledge evolution integration, source evolution integration, traceability and guided automation [11]. Figure 2 shows a simplified architecture of the CDW, how the data go through ETL into the CDW and how the analysis tools are used by clinical stakeholders for decision making, research and management purposes.…”

“…CDW construction is a difficult task from planning to implementation. Different clinical procedures from intensive care to treatment contain a variety of data and produce heterogeneous data [11]. The implementation process of CDW if full of obstacles start from analyzing data sources and ending with implementing access tools (OLAP, KPI, and reports console).…”