COBE: A Conjunctive Ontology Browser and Explorer for Visualizing SNOMED CT Fragments

Area of Science:

• Medical Informatics
• Information Retrieval
• Ontology Engineering

Background:

• Ontology search interfaces require advanced information retrieval techniques.
• SNOMED CT is a comprehensive clinical terminology requiring efficient search capabilities.

Purpose of the Study:

• To introduce the Conjunctive Ontology Browser and Explorer (COBE) for searching and exploring SNOMED CT concepts.
• To combine navigational exploration (NE) and direct lookup (DL) for improved SNOMED CT concept retrieval.
• To visualize SNOMED CT fragments effectively.

Main Methods:

• COBE integrates NE and DL search modes.
• NE mode uses incremental addition of word stems as attribute constraints for narrowing the search space.
• DL mode utilizes keywords and concept identifiers for direct searching.

Main Results:

• The DL mode achieved a precision of 0.958, recall of 0.917, and F1 measure of 0.875.
• The NE mode, using a fraction of the SNOMED CT stem cloud, covered 98.97% of the entire concept collection.
• COBE effectively navigates and retrieves SNOMED CT concepts.

Conclusions:

• COBE offers a dual-mode approach for efficient SNOMED CT concept searching and exploration.
• The NE mode provides comprehensive coverage with minimal stem usage.
• COBE advances ontology search interfaces through integrated retrieval and exploration strategies.