Related Experiment Videos
Conceptual search in electronic patient record.
1Medical Informatics Division, University Hospital of Geneva, Switzerland. Robert.Baud@dim.hcuge.ch
Studies in Health Technology and Informatics
|October 18, 2001
Summary
This study enhances medical text searching by adding a conceptual model to string matching, improving results beyond traditional GREP methods. The new system accepts natural language queries, offering better discoverability of domain-specific knowledge.
Area of Science:
- Medical Informatics
- Natural Language Processing
- Information Retrieval
Background:
- Current content search in large medical text corpora is limited.
- Traditional GREP (Get Regular Expression and Print) methods struggle with domain-specific knowledge and require Boolean queries.
- Poor search results occur when Boolean query constraints are not met.
Purpose of the Study:
- To enhance content search in medical texts.
- To overcome limitations of traditional string matching and GREP approaches.
- To improve the quality and accessibility of search results.
Main Methods:
- Implemented an enhancement to string matching search.
- Integrated a light conceptual model with the word lexicon.
- Utilized advanced indexing algorithms during a pre-processing phase for efficiency.
Main Results:
- The new system accepts natural language sentences as queries.
- Radically improved the quality of search results.
- Achieved efficiency in execution time through advanced indexing.
Conclusions:
- The enhanced search system offers a significant improvement over existing methods.
- The integration of a conceptual model broadens the applicability of text search in the medical domain.
- The system provides a more effective way to retrieve domain-specific knowledge from medical texts.