Automatic Document Classification of Biological Literature
Author Information
Author(s): Chen David, Müller Hans-Michael, Sternberg Paul W
Primary Institution: California Institute of Technology
Hypothesis
Can a two-step text categorization algorithm effectively classify biological literature?
Conclusion
The study demonstrates an improved method for classifying biological documents, particularly for Caenorhabditis elegans literature.
Supporting Evidence
- The classification engine improved the F-value from 0.49 to 0.55 compared to previous methods.
- The system allows researchers to navigate through a hierarchy of biological concepts.
- Support Vector Machines were used to classify papers into nine primary categories.
Takeaway
This study created a smart system that helps scientists find and organize biology papers better, especially about a tiny worm called C. elegans.
Methodology
The study used a support vector machine and a phrase-based clustering algorithm to classify a corpus of C. elegans papers.
Limitations
The classification results are optimized for C. elegans papers and may require human-created rules for other types of documents.
Statistical Information
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website