James Heilman wrote:
... Anyway who would base an expert system for medicine on Wikipedia would be a fool. I am sure the people at IBM are not.
The problem that concerns me isn't the quality of Wikipedia's medical content, which I am sure is proofread with a nanoscale-thin comb before it gets near Watson, but the quality of tangentially related concepts.
The examples I'd like to try to work with Watsoners at Rutgers etc. on are: saturation of medical-related concepts into articles with less quality, (e.g. the use of medical terms in articles about fiction, religion, disputes, disasters, economics, etc.) inference paths which involve controversial assumptions not subject to the usual rigor (e.g., cost savings from preventative care, proportionality of effort to harm, harm reduction, etc.) idioms, figures of speech, sarcastic quotations, etc.
http://www.nj.com/news/index.ssf/2013/01/ibms_jeopardy-winning_supercom.html
http://www.thenewnewinternet.com/2013/01/31/ibm-to-provide-rensselaer-polyte...
Who is our Campus Ambassador(s) to Rutgers, and do they want to try to enroll Watson as an editor?
Sincerely, James Salsman