text2term: A Tool for Mapping Biomedical Terms to Ontologies
Author Information
Author(s): Gonçalves Rafael S, Payne Jason, Tan Amelia, Benitez Carmen, Haddock Jamie, Gentleman Robert
Primary Institution: Stanford Center for Biomedical Informatics Research, Stanford University
Hypothesis
The text2term tool can effectively standardize biomedical metadata by mapping free-text descriptions to controlled terms in ontologies.
Conclusion
The text2term tool accurately maps biomedical terms to ontologies, significantly reducing the need for human curation.
Supporting Evidence
- The text2term tool achieved an accuracy of 79%, 81%, and 73% in mapping compared to benchmark mappings.
- It is available as a Python package and can be used via a web application.
- The tool supports both programmatic and interactive mapping of biomedical terms.
- Users can deploy the tool locally using Docker for bulk mapping tasks.
- The tool has been downloaded over 22,000 times since its release.
Takeaway
The text2term tool helps researchers find the right names for diseases and other biomedical terms by matching them to a list of accepted terms, making it easier to share and use data.
Methodology
The tool uses various string-matching techniques, including TF-IDF and edit distance metrics, to map free-text terms to ontology terms.
Limitations
The tool may struggle with terms that include negation or are not well-represented in the ontologies.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website