Corpus Refactoring: a Feasibility Study
Author Information
Author(s): Helen L Johnson, William A Baumgartner Jr, Martin Krallinger, K Bretonnel Cohen, Lawrence Hunter
Primary Institution: Center for Computational Pharmacology, University of Colorado School of Medicine
Hypothesis
Is corpus refactoring a feasible goal that can be accomplished with a semi-automatable process and in a time-efficient way?
Conclusion
Refactoring of publicly available corpora is a technically and economically feasible method for increasing the usage of data already available for evaluating biomedical language processing systems.
Supporting Evidence
- The refactored corpus is available for download at the BioNLP SourceForge website.
- The total time expended was just over three person-weeks, consisting of about 102 hours of programming time and 20 hours of manual validation.
- Corpus refactoring can increase the usage of existing biomedical data without the cost of repeating their annotation.
Takeaway
This study shows that changing the format of biomedical data can help more people use it, and it can be done without spending too much time or money.
Methodology
The study involved refactoring the Protein Design Group corpus into two new formats using simple text processing methods and limited human validation.
Limitations
The original corpus had idiosyncratic formatting and lacked standard annotations, which complicated the refactoring process.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website