BOAT: automatic alignment of biomedical ontologies using term informativeness and candidate selection.

Area of Science:

• Biomedical Informatics
• Computational Biology
• Knowledge Representation

Background:

• Ontologies are crucial for information retrieval and knowledge sharing in biomedical sciences.
• Interoperability challenges arise from using disparate ontologies without explicit concept mapping.
• Ontology alignment is essential for enabling knowledge sharing across different biomedical knowledge bases.

Purpose of the Study:

• To address the challenges of concept-pair equivalence determination and high runtime in biomedical ontology alignment.
• To introduce a novel approach, the Biomedical Ontologies Alignment Technique (BOAT), for efficient and accurate biomedical ontology alignment.
• To improve the F-measure, precision, and speed of ontology alignment in the biomedical domain.

Main Methods:

• Developed the Biomedical Ontologies Alignment Technique (BOAT), a novel approach for ontology alignment.
• Incorporated word informativeness into concept label analysis to enhance matching accuracy.
• Utilized annotation similarity to pre-select concept pairs with high equivalence likelihoods, optimizing runtime.

Main Results:

• Achieved state-of-the-art performance in F-measure, precision, and speed.
• Consideration of word informativeness increased F-measure by 12.2%.
• Achieved an F-measure of 0.88 in aligning mouse and human anatomy ontologies, comparable to AgreementMaker but with reduced runtime.

Conclusions:

• BOAT effectively addresses key challenges in biomedical ontology alignment, including concept equivalence and runtime.
• The technique demonstrates superior performance in accuracy and efficiency for large-scale biomedical ontologies.
• BOAT facilitates enhanced knowledge sharing and data integration within the biomedical domain.