Exciting developments with use of our Solr server. As a resource, it supplies a readily available source of metadata for NLP and AI implementations. Using the Solr repository schema, we can access key elements of the guideline document. For example, we can point to the keywords field to retrieve MESH and SNOMED codes from the UMLS or screen for guidelines that consider Risks and Harms or Recommendation Grading. We have utilized a Java interface to convert the Solr JSON response to objects which can then be stored or used to query other repositories such as the UMLS. Combinations of fields should prove useful in identifying and screening for relevant recommendations. View some preliminary results here.
GEM Guideline Transformations
We have introduced a site where users can transform there GEM markup to nicely displayed reports for summary review or decision logic. Click here to access the site.