Related Experiment Video
Updated: Dec 30, 2025
07:35
A Knowledge Graph Approach to Elucidate the Role of Organellar Pathways in Disease via Biomedical Reports
Published on: October 13, 2023
2.0K
SimConcept: a hybrid approach for simplifying composite named entities in biomedical text
IEEE Journal of Biomedical and Health Informatics
|April 17, 2015
Summary
This study introduces SimConcept, a novel hybrid method for identifying and resolving complex biomedical terms like gene and disease names. SimConcept significantly improves the accuracy of recognizing and normalizing these composite entities.
Area of Science:
- Biomedical Natural Language Processing
- Bioinformatics
- Computational Biology
Background:
- Biomedical Named Entity Recognition (NER) and normalization face challenges with composite entities (e.g., BRCA1/2) that represent multiple concepts.
- Existing methods inadequately handle composite mentions, necessitating a robust approach for multitype entities.
Purpose of the Study:
- To develop and evaluate SimConcept, a hybrid method for identifying and resolving composite biomedical named entities.
- To systematically address various types of composite mentions in biomedical text.
Main Methods:
- A hybrid approach integrating a machine-learning model with a pattern identification strategy.
- Development of the SimConcept method to identify individual components within composite mentions.
Main Results:
- SimConcept achieves high F-measure scores: 90.42% for genes, 86.47% for diseases, and 86.05% for chemicals.
- The method demonstrates effectiveness in identifying and resolving composite mentions across multiple biological entity types.
- Subsequent improvements in gene and disease concept recognition and normalization were observed using SimConcept.
Conclusions:
- SimConcept is the first method to systematically handle diverse composite biomedical entities.
- The proposed hybrid approach offers a robust solution for a significant challenge in biomedical NER and normalization.
- SimConcept enhances downstream tasks like concept recognition and normalization for genes and diseases.
