SIENA: Semi-automatic semantic enhancement of datasets using concept recognition.

Area of Science:

• Biomedical Informatics
• Bioinformatics
• Computational Biology

Background:

• The increasing volume and diverse formats of scientific data present challenges for integrating multiple datasets.
• Effective integration is crucial for answering complex scientific research questions.

Purpose of the Study:

• To develop and evaluate a semi-automated framework for semantic enhancement of biomedical data, specifically gene datasets.
• To improve the integration and utility of disparate gene datasets for research.

Main Methods:

• A semi-automated framework was developed, incorporating concept recognition using machine learning.
• The framework utilized the BioPortal annotator for semantic enhancement.
• Performance was compared against methods relying solely on the BioPortal annotator.

Main Results:

• The proposed framework achieved superior results compared to methods using only the BioPortal annotator.
• Semantic enhancement of gene datasets was successfully demonstrated.

Conclusions:

• Combining concept recognition, machine learning, and biomedical ontology annotation enhances gene datasets.
• The framework enables datasets to realize their full potential for providing meaningful information to answer research questions.