Homepage of Aaron Michael Cohen, M.D. M.S.

Current contact information:

	Aaron M. Cohen Associate Professor School of Medicine, Department of Medical Informatics and Clinical Epidemiology (DMICE) Oregon Health & Science University 3181 S.W. Sam Jackson Park Road Mail Code: BICC Portland, Oregon, USA 97239-3098 email: cohenaa(at-sign-goes-here)ohsu.edu
Future Informatics Associate Professor?

Professional Information

Publications:

1. Cohen AM, Hersh W. A Survey of Current Work in Biomedical Text Mining. Briefings in Bioinformatics 2005;6(1):57-71. [pdf]
2. Cohen AM, Hersh WR, Dubay C, Spackman K. Using co-occurrence network structure to extract synonymous gene and protein names from MEDLINE abstracts. BMC Bioinformatics 2005;6(103). [pdf]
3.  Cohen AM, Hersh WR, Bhupatiraju RT. Feature generation, feature selection, classifiers, and conceptual drift for biomedical document triage. In: Proceedings of the Text Retrieval Conference (TREC) 2004; Gaithersburg, MD. [pdf]
4. Cohen AM. Unsupervised gene/protein entity normalization using automatically extracted dictionaries. In: Linking Biological Literature, Ontologies and Databases: Mining Biological Semantics, Proceedings of the BioLINK2005 Workshop; Detroit, MI: Association for Computational Linguistics; 2005. p. 17-24. [pdf]
5. Cohen AM, Yang J, Hersh WR. A Comparison of Techniques for Classification and Ad Hoc Retrieval of Biomedical Documents. In: Proceedings of the Fourteenth Annual Text REtrieval Conference, TREC 2005, Gaithersburg, MD. [pdf]
6. Cohen AM, Hersh WR, Peterson K, Yen PY. Reducing Workload in Systematic Review Preparation Using Automated Citation Classification. JAMIA 2006;13(2):206-219. [pdf]
7. Cohen AM. An Effective General Purpose Approach for Automated Biomedical Document Classification. In: Proceedings of the American Medical Informatics Association (AMIA) 2006 Annual Symposium; 2006. [pdf]
8. Cohen AM.Five-way Smoking Status Classification using Text Hot-spot Identification and Error-Correcting Output Codes. J Am Med Inform Assoc, 2007. [pdf] [pdf of full paper and data supplement]
9. Cohen AM. Optimizing feature representation for automated systematic review work prioritization. AMIA Annu Symp Proc 2008:121-5. [pdf]
10. Cohen AM, Ambert K, McDonagh MS. Cross-topic Learning for Work Prioritization in Systematic Review Creation and Update. J Am Med Inform Assoc 2009. [pdf]
11. Ambert KH, Cohen AM. k-Information Gain Scaled Nearest Neighbors: A Novel Approach to Classifying Protein-Protein Interaction-Related Documents. IEEE IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2011. [pdf]
12. Cohen AM, Ambert K, McDonagh M. Studying the potential impact of automated document classification on scheduling a systematic review update. BMC Medical Informatics and Decision Making. 2012 Apr 19;12(1):33. [pdf]
    

Curriculum Vitae:

• View my Curriculum Vitae here.

Data Sets:

• Gold standard files for work described in my masters thesis and in the paper Using co-occurrence network structure to extract synonymous gene and protein names from MEDLINE abstracts. You can also view the whole directory at once.
• Systematic drug class review data extracted from Endnote files created by the OHSU Evidence-based Practice Center.

Software:

• RMEQ: A tool for Computing Equivalence Groups in Repeated Measures Studies

Personal Information

I like (in no particular order) programming the iPhone, experimenting with programming languages, building and playing ukuleles and guitars, folding origami, listening to podcasts, reading (to myself and my kids), playing board games, taking photographs, and watching Doctor Who (when I have the time).

Web Resources

Here are a few of my favorite programming and biomedical text mining resources on the Web:

• The ORBIT Project. An Online Registery of Biomedical Informatics Tools.
  
• Alex Morgan's page on BioNLP resources, although it has been a while since it has been updated.
• BLIMP (Biomedical Literature and Text Mining Publications), a forum for collection, compilation, and exchange of publications on biomedical text mining. Sadly, this is no longer kept up and has gone out of date.
  
• TREC Genomics Track home page.
• TREC Medical Records Track home page.
• BioPython project home page. Not geared specifically to text mining BioPython contains lots of tools for working with resources useful in biomedical text mining.
• Python.org, the home of the Python language. You do know about Python, right? After spending years getting paid to program in Assembly, C, C++, and Java, I now do almost all of my text mining research using Python. It's my favorite language. But that GIL is going prevent Python being a great multicore language.
  
• Clojure.org, the home of the Clojure programming language. LISP on the JVM, really? But it's got immutable data and software transaction memory. It may be the language best-prepared to take advantage of 21st century CPUs (read: multicore).