Related Experiment Videos
A graphical tool for ad hoc query generation
K Stoffel1, J D Davis, G Rottman
1Computer Science Department, University of Maryland, College Park 20742, USA.
Proceedings. AMIA Symposium
|February 3, 1999
Summary
Clinicians and researchers can now easily query complex medical databases using a new graphical tool. This tool simplifies ad hoc question formulation into SQL queries, improving data retrieval speed by up to 100 times.
Area of Science:
- Medical Informatics
- Database Management
- Clinical Research Tools
Background:
- Medical data presents challenges due to complex taxonomies and evolving terminology.
- Formulating complex questions for medical databases as standard SQL queries is difficult for users.
Purpose of the Study:
- To introduce a graphical tool for simplifying the creation of ad hoc SQL queries from medical databases.
- To enhance the speed and efficiency of data retrieval for clinical and research purposes.
Main Methods:
- Development of an interactive graphical tool to manage multiple attribute hierarchies.
- Navigation through hierarchies to generate SQL query strings.
- Optimization of the tool using indexing to accelerate query building and data retrieval.
Main Results:
- Indexed queries demonstrated a significant speed improvement, performing 5 to 100 times faster than non-indexed query strings.
- Query string generation time is dependent on taxonomy size; index generation time depends on data warehouse size.
Conclusions:
- The graphical tool effectively simplifies the process of formulating complex questions for medical databases.
- Indexing significantly enhances the performance of data retrieval, making medical data more accessible for analysis.