Autogeneration of Database Applications from XML Metadata for Web-Based Data Entry

In a biomedical research environment, there is frequently a need for dynamically generated Web interfaces for patient-related data that can easily accommodate changes in the underlying data model by altering metadata rather than changing the interface code. In this paper, we present AutoGen, a tool developed at the Heart and Vascular Institute (HVI) of Cleveland Clinic for autogenerating the database schema and data-entry-specific Web front end from XML metadata. To specify metadata, we established a tested and proven methodology for systematically capturing data and data entry requirements and encapsulating them into standard XML schema. During the development process, we created a successful approach to incorporating model-view-controller and template technologies for implementing and optimizing AutoGen. Additionally, we conducted a cost-benefit analysis of AutoGen during its first 20 months of use within HVI. Results from both the development perspective and practical usage show that AutoGen and its associated metadata specification method provide a unique approach that can efficiently and effectively address the data entry needs of similar biomedical research environments.