The TIES team is excited to preview the upcoming release of NOBLE Suite with a seismic change in NOBLE’s capabilities. The new release is currently scheduled for June 1st. Congratulations to team members Eugene Tseytlin and Melissa Castine, whose extraordinary collaboration brought us to this important next step in the development of the software.
What’s Next for NOBLE?
The new NOBLE Mentions release (NOBLE v 1.1) provides the ability to input a domain ontology specifying entities to be annotated along with their relations within specific text documents and sections. For example, create an annotator that marks all mentions of hypertension. Even create specific quantitative definitions for observations such as high blood pressure (BP systolic > 140, diastolic > 90). And use these to reason from observation to disease.
NOBLE uses the concept of an anchor and links relevant modifiers (semantic attributes, linguistic attributes, qualifiers and quantifiers) to annotate more complex statements. Use these fully articulated ontology-based features in downstream processing such as predictive analytic tools or population based analyses. Extraction can even be conditioned on axioms in the ontology.
The major advance of NOBLE Mentions is that it gives you the ability to define your phenotypes and configure the system without changing a line of code.
More about the NOBLE Suite of NLP tools
Noble Coder is a generic tool and Application Programming Interface (API) that automatically encodes free text with concepts from a controlled terminology using natural language processing. NOBLE tools can be used to create terminologies from all the leading terminology and ontology formats, even by using the Bioportal API. The tool comes bundled with Apache UIMA AEs and GATE processing resources to further help with processing needs.
NOBLE has been evaluated against all the other major systems for biomedical NLP and achieves excellent performance, while providing unique advantages in terms of terminology building and scaling to cloud environments. Learn more about NOBLE from our BMC Bioinformatics paper “NOBLE – Flexible concept recognition for large-scale biomedical natural language processing”).
Ever wonder how NOBLE got its name? A previous system we worked on included an implementation of the IndexFinder algorithm, but Eugene Tseytlin was already working on a better method. After another team member joked that we should call the new algorithm “EugeneFinder,” we settled on using the translation of Eugene, which means ‘noble’ in Greek. With our goal to make the system ‘concept coding royalty’, NOBLE seemed just right.
Use NOBLE in your research
We hope you’ll continue to make NOBLE a top choice for anyone looking for customization and usability in an NLP tool. The team welcomes bug reports, suggestions, and extensions through the Noble Coder GitHub page.
Want more information about Noble Coder? Visit our documentation page here: http://ties.dbmi.pitt.edu/noble-coder/