New terminology services based on term comparison using semantic definitions and similarity computation

Area of Science:

• Medical Informatics
• Computational Linguistics
• Bioinformatics

Background:

• Medical information is stored in diverse terminological systems, hindering interoperability.
• Current systems primarily use semantic definitions and subsumption relations for term comparison.
• Existing terminology servers lack support for similarity-based term comparison.

Purpose of the Study:

• To develop novel terminology services to address the challenge of comparing medical terms across different systems.
• To enable semantic mediation between diverse medical applications.
• To enhance data analysis and communication in healthcare.

Main Methods:

• Implementation of new terminology services based on a semantic distance metric.
• Development of services for term comparison, k-nearest neighbors, and concept coding.
• Creation of tools for automatic similarity table generation, distance-based queries, and clustering.

Main Results:

• A suite of terminology services supporting semantic distance-based comparisons.
• Enabled functionalities including k-nearest neighbors search and concept coding.
• Facilitated automatic generation of similarity tables and distance-based queries for enhanced data mining.

Conclusions:

• The proposed terminology services effectively utilize semantic distance for term comparison.
• These services offer a valuable solution for semantic mediation in medical informatics.
• The developed tools can improve data integration and analysis in healthcare settings.