6,585
edits
L'en-droit de Laval: Difference between revisions
no edit summary
(Created page with "{{Subventions |import ref=Subventions 3939 |Code postal=H7G2T2 |Rue=111, boulevard des Laurentides, bureau 201 |Ville=Laval |Ministère=MESSF(SACA) }}") |
No edit summary |
||
| Line 1: | Line 1: | ||
{{ | {{MailMessage | ||
|import ref=class org.zooid.mediawiki.MailToSMW | |import ref=class org.zooid.mediawiki.MailToSMW | ||
|Subject=Discuss:National Library of Medicine working on search engine that will use natural language processing | |||
|From=[[email protected]] | |||
|To=[[email protected]] | |||
|CC=[[email protected]] | |||
|Date=10 Feb 2012 22:17:04 GMT | |||
}} | }} | ||
<code>http://t.co/ZejfUAeR |<br /> | |||
<br /> | |||
The National Library of Medicine <http://www.nlm.nih.gov/> is working on a<br /> | |||
search engine that will scan thousands of medical records to turn up<br /> | |||
documents related to patient queries. That sounds exactly like what Google<br /> | |||
and Bing can already do, but it���s different.<br /> | |||
<br /> | |||
It will employ Watson-like search and retrieval techniques to return a very<br /> | |||
specific set of one to three documents.<br /> | |||
Watson<http://www-03.ibm.com/innovation/us/watson/index.html>,<br /> | |||
IMB���s robot with artificial intelligence capabilities,<br /> | |||
smashed*Jeopardy!* records<br /> | |||
when it appeared on the gameshow last year. One trait that makes Watson<br /> | |||
intelligent is that the computer can understand natural language as opposed<br /> | |||
to a Google query (i.e. ���cure common cold��� instead of ���what���s the cure for<br /> | |||
the common cold?���), something that researchers at the National Library of<br /> | |||
Medicine (NLM) want to try to replicate with their own machine.<br /> | |||
<br /> | |||
Another characteristic of intelligent computers is that they learn. This<br /> | |||
occurs over time and through repetition.<br /> | |||
<br /> | |||
���In order to train the machine we had to have a large set of questions in<br /> | |||
the vernacular, not cleaned up to grammar, and we had to have a lot of<br /> | |||
associated, or reasonably associated, answers,��� Dr. Milton Corn, deputy<br /> | |||
director for research and education at NLM, said.<br /> | |||
<br /> | |||
Having a database of answers ��� not just questions ��� is key because when the<br /> | |||
computer turns up a wrong answer, researchers need to show the computer<br /> | |||
what the right answer looks like. Corn said NLM gets millions of queries<br /> | |||
from patients every year, but they���re answered through results generated by<br /> | |||
NLM���s databases PubMed <http://www.ncbi.nlm.nih.gov/pubmed/> and<br /> | |||
MedlinePlus<http://www.nlm.nih.gov/medlineplus/>,<br /> | |||
as opposed to health care professionals.<br /> | |||
<br /> | |||
In his search for a set of published medical questions and answers, Corn<br /> | |||
came across Ask the Doctor <http://www.askthedoctor.com/>, a Canadian<br /> | |||
website where doctors answer individual patient questions. Ask the Doctor<br /> | |||
agreed to share a set of about 8,000 medical questions and their answers<br /> | |||
with NLM for its artificial intelligence project.<br /> | |||
<br /> | |||
Right now the computer is in Rocky Balboa mode.<br /> | |||
<br /> | |||
Researchers are training it by feeding it questions, then feeding it<br /> | |||
answers and testing it. Then they repeat the process. The major hurdle for<br /> | |||
the computer is getting to a point where it can understand natural language.<br /> | |||
<br /> | |||
And if the machine ever becomes ready for public use, patients, too, will<br /> | |||
have to be trained to approach the Q&A process differently. Patients will<br /> | |||
need to phrase their questions like they���re talking to their doctor, not<br /> | |||
Google. This will lay the groundwork for the possible incorporation of<br /> | |||
voice technology later on.<br /> | |||
<br /> | |||
Is NLM also laying the groundwork for computer replacement of physicians?<br /> | |||
Not quite.<br /> | |||
<br /> | |||
���We are not contemplating ��� unless this were an unbelievably fantastic<br /> | |||
success ��� letting a machine practice medicine,��� Corn said.<br /> | |||
<br /> | |||
If this early stage does prove to be successful, the project might expand<br /> | |||
with a partnership with IBM, which has expressed an interest in working<br /> | |||
with NLM.<br /> | |||
<br /> | |||
���We would like to see whether Watson could be adapted to health questions,���<br /> | |||
Corn said.<br /> | |||
<br /> | |||
But first, Watson will be in for some medical education.<br /> | |||
<br /> | |||
���It doesn���t know much about health at the moment,��� Corn said.<br /> | |||
</code> | |||