Ontology Development & Information Extraction - ODIE

ODIE stands for “Ontology Development and Information Extraction”. It is a software toolkit that uses ontologies to perform information extraction tasks from clinical documents and uses clinical documents to enhance existing ontologies. ODIE codes document sets with ontologies or enriches existing ontologies with new concepts from the document set. It contains algorithms for Named Entity Recognition, Co-reference resolution, concept discovery, discourse reasoning and attribute value extraction.

To see Seminar Series Presentations on ODIE, please see the talks: ODIE Toolkit (12/2007) and ODIE 0.5 (2/2009).


The ODIE Project aims to do the following:

1. Develop and evaluate methods for information extraction (IE) tasks using existing BioPortal ontologies, including

  • Named Entity Recognition
  • Coreference Resolution
  • Discourse Reasoning
  • Attribute Value Extraction

2. Develop and evaluate general methods for clinical-text mining to assist in ontology development, including

  • Preprocessing
  • Concept discovery and clustering
  • Taxonomic positioning and automated completion of role relationships

3. Develop reusable software for performing information extraction and ontology development leveraging existing NCBO tools and compatible with NCBO architecture.

4. Enhance National Cancer Institute Thesaurus Ontology using the ODIE toolkit.

5. Test the ability of the resulting software and ontologies to address important translational research questions in hematologic cancers.

The end-product of the project will be a software toolkit that will be released under an open source license.

The demonstration version is now available for download below. It is released more as a proof of concept, and it serves to introduce what the ODIE project has set out to accomplish. In that regard it is limited in functionality. ODIE v1.0 is scheduled for release in Fall 2009 and is anticipated to contain many more features.

Please use the user forums and bug tracker at the project homepage to report bugs or to post questions or comments.
The user manual can be downloaded at: https://bmir-gforge.stanford.edu/gf/download/docmanfileversion/408/640/demomanual.doc
Project homepage: https://bmir-gforge.stanford.edu/gf/project/odie

For help, feedback of support, please access the User Forums: https://bmir-gforge.stanford.edu/gf/project/odie/forum/


Download ODIE 0.5 Now - ODIE 1.0 is scheduled for release in Fall 2009.