Related Experiment Videos
Knowledge acquisition to qualify Unified Medical Language System interconceptual relationships
F Le Duff1, A Burgun, M Cleret
1Laboratoire d'Informatique Medicale, Faculte de Medecine-Rennes I 2 Rue Henri Le Guilloux 35033 Rennes, France.
Proceedings. AMIA Symposium
|November 18, 2000
Summary
This study enhances the Unified Medical Language System (UMLS) by automatically transferring medical relationship qualifiers from French databases. This improves UMLS medical relevance and consistency for diverse applications.
Area of Science:
- Medical Informatics
- Knowledge Representation
- Computational Linguistics
Background:
- Knowledge bases like the Unified Medical Language System (UMLS) benefit from automated relation extraction.
- Transferring qualified relationships enhances medical pertinence and consistency.
- Existing methods often lack the transfer of relationship qualifiers.
Purpose of the Study:
- To describe the automatic inheritance of medical inter-conceptual relationship qualifiers.
- To integrate these qualifiers into the UMLS knowledge base.
- To focus on knowledge transmission from French to English medical databases.
Main Methods:
- Developing methods for automatic extraction of relationship qualifiers from disease descriptions.
- Implementing algorithms for integrating extracted qualifiers into the UMLS.
- Utilizing a French medical database as the source and targeting the English UMLS.
Main Results:
- Demonstrated the feasibility of automatically inheriting medical relationship qualifiers.
- Successfully integrated qualifiers, enhancing UMLS compliance and medical pertinence.
- Showcased cross-lingual knowledge transfer from French to English medical data.
Conclusions:
- Automated transfer of qualified medical relationships significantly improves knowledge base utility.
- This approach enhances the UMLS for various medical applications.
- Cross-lingual knowledge transfer is a viable method for enriching medical knowledge bases.