Pei,
MedEx-UIMA is rule-based, instead of ML. :) What are the work involved in using
the cTAKES common type system? We could consider it when Min has time. Thanks.
Hua
From: , Pei
mailto:pei.c...@childrens.harvard.edu>>
Date: Monday, July 22, 2013 3:30 PM
To: Min Jiang mailto:min.ji...@uth.tmc
Dr. Xu,
That was a wonderful presentation on MedEx (ML based medication extraction)
during the iDASH webinar.
Would there be any interest in contributing/integrating MedEx-UIMA as an Apache
cTAKES [1] component/module? I had a quick look at the code and it looks like
it's already using Java/UIM
It's fixed and building correctly.
https://builds.apache.org/job/ctakes-trunk-compiletest/356/?auto_refresh=false
I had made some changes to update dependencies and then reverted back to the
old version (all on 7/18), but in the mean time, a directory that was still
needed was deleted. I'm ver
Here's that error again:
On 22 Jul2013, at 12:05 , Apache Jenkins Server
wrote:
> [ERROR] 'dependencies.dependency.version' for net.sf.mastif:mastif-zoner:jar
> is missing. @ line 36, column 15
Could whoever is taking care of "ctakes-assertion-zoner" fix this? We haven't
had a successful buil
Hi Sandeep,
I'll try and review this today.
Cheers,
Chris
++
Chris Mattmann, Ph.D.
Senior Computer Scientist
NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
Office: 171-266B, Mailstop: 171-246
Email: chris.a.mattm...@nasa.gov
Hi Sandeep,
I just took a peek at the JavaOcr code, and it looks like they perform image
filtering in the PixelImage class. This would probably cause a problem with
dot matrix images as every corner of every dot would be removed as "noise", so
dots that participate in curves on characters such
sir,
i have gone through some of the medical record such as bills,patient
details etc. most of them are printed using dot matrix printer,which is
very hard to extract such type text from scanned images.i have done testing
with some professional software such as abbyy fine reader which also given
a