You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ctakes.apache.org by Andrew Phillips <ap...@luc.edu> on 2017/06/28 17:13:54 UTC
Building a Custom cTAKES Dictionary
Hello,
I am new to cTAKES and I'm trying to create a custom dictionary of
additional terms related to alcohol. How can I go about adding semantic
types such as food (T168), etc that aren't available by default to the
dictionary? The cTAKES install is located on a headless Linux server that I
access via SSH.
Thank you,
Andrew
*Andrew Phillips*
GitHub: github.com/skeledrew
LinkedIn: www.linkedin.com/in/aphillipstech
RE: Building a Custom cTAKES Dictionary [EXTERNAL]
Posted by "Finan, Sean" <Se...@childrens.harvard.edu>.
Hi Andrew,
Can you xWin (or other) to the server to launch gui applications?
If so, try the dictionary creator gui: https://cwiki.apache.org/confluence/display/CTAKES/Dictionary+Creator+GUI
Sean
-----Original Message-----
From: Andrew Phillips [mailto:aphillips5@luc.edu]
Sent: Wednesday, June 28, 2017 1:14 PM
To: dev@ctakes.apache.org
Subject: Building a Custom cTAKES Dictionary [EXTERNAL]
Hello,
I am new to cTAKES and I'm trying to create a custom dictionary of additional terms related to alcohol. How can I go about adding semantic types such as food (T168), etc that aren't available by default to the dictionary? The cTAKES install is located on a headless Linux server that I access via SSH.
Thank you,
Andrew
*Andrew Phillips*
GitHub: github.com/skeledrew
LinkedIn: www.linkedin.com/in/aphillipstech
RE: Building a Custom cTAKES Dictionary [EXTERNAL]
Posted by "Finan, Sean" <Se...@childrens.harvard.edu>.
Hi Andrew,
You will need to download the umls data from the nlm. Go to their website
https://www.nlm.nih.gov/research/umls/
and use the "Downloads" button in the "Access" panel.
I will put some more instructions on the wiki page when I get a chance.
After you download and build a local copy of the umls, run the gui. In your ctakes resources/ directory, go to org/apache/ctakes/dictionary/lookup/fast/ and you will see an xml file and a directory named after your custom dictionary. Copy those to the headless server in that ctakes' resources/org/apache/../fast/ directory.
See the (info) panel at the bottom of https://cwiki.apache.org/confluence/display/CTAKES/Dictionary+Creator+GUI
To see how to point ctakes to your custom gui.
By the way, if you have a small number of terms and don't need the umls you can manually create a bar separated value (bsv) file. I am guessing that you have more than a few terms.
Sean
-----Original Message-----
From: Andrew Phillips [mailto:skeledrew@gmail.com]
Sent: Wednesday, June 28, 2017 6:28 PM
To: dev@ctakes.apache.org
Subject: Re: Building a Custom cTAKES Dictionary [EXTERNAL]
Hi Sean,
I haven't found a good way to launch GUIs remotely (BTW I'm also using Linux on my machine). However I also have a local cTAKES install and I'm trying to do the modification there. However I'm not sure what to enter into the UMLS installation field, as I cannot find a META directory or RRF files. Also what file(s) woud I transfer to the server once I have built the new dictionary?
Thanks,
Andrew
*Andrew Phillips*
Computer Technician / Programmer and Mobile Experience Consultant
Phone: (678) 753-5313
Email: skeledrew@gmail.com
LinkedIn: www.linkedin.com/in/aphillipstech
"A man may imagine things that are false, but he can only understand things that are true, for if the things be false, the apprehension of them is not understanding." - Isaac Newton
On 28 June 2017 at 12:35, Finan, Sean <Se...@childrens.harvard.edu>
wrote:
> Hi Andrew,
>
> Can you xWin (or other) to the server to launch gui applications?
> If so, try the dictionary creator gui:
> https://urldefense.proofpoint.com/v2/url?u=https-3A__cwiki.apache.org_
> &d=DwIBaQ&c=qS4goWBT7poplM69zy_3xhKwEW14JZMSdioCoppxeFU&r=fs67GvlGZstT
> pyIisCYNYmQCP6r0bcpKGd4f7d4gTao&m=375tbajCm8PjEi_45ahKX7bDghJ_w3fOBIh9
> vAMpylU&s=0JwVENy0eLeCkfQVSfkUBwhF_xAzBzXGYCWVX9_e65U&e=
> confluence/display/CTAKES/Dictionary+Creator+GUI
>
>
> Sean
>
> -----Original Message-----
> From: Andrew Phillips [mailto:aphillips5@luc.edu]
> Sent: Wednesday, June 28, 2017 1:14 PM
> To: dev@ctakes.apache.org
> Subject: Building a Custom cTAKES Dictionary [EXTERNAL]
>
> Hello,
>
> I am new to cTAKES and I'm trying to create a custom dictionary of
> additional terms related to alcohol. How can I go about adding
> semantic types such as food (T168), etc that aren't available by
> default to the dictionary? The cTAKES install is located on a headless
> Linux server that I access via SSH.
>
> Thank you,
> Andrew
>
> *Andrew Phillips*
> GitHub: github.com/skeledrew
> LinkedIn: www.linkedin.com/in/aphillipstech
>
RE: Building a Custom cTAKES Dictionary [EXTERNAL]
Posted by "Finan, Sean" <Se...@childrens.harvard.edu>.
Hi Andrew,
I am glad that it worked for you. I agree that the umls/metamorphosis installation could use a few hints here and there, but I don't want to be the pot calling the kettle ...
Thanks,
Sean
-----Original Message-----
From: Andrew Phillips [mailto:aphillips5@luc.edu]
Sent: Friday, June 30, 2017 10:29 PM
To: dev@ctakes.apache.org
Subject: Re: Building a Custom cTAKES Dictionary [EXTERNAL]
Hi Sean,
It took a while to figure out how to setup everything and run tests, but I have now successfully generated CUI files containing the terms that were missing before.
Thanks for your help.
Andrew
*Andrew Phillips*
GitHub: github.com/skeledrew
LinkedIn: www.linkedin.com/in/aphillipstech
On 28 June 2017 at 19:31, Finan, Sean <Se...@childrens.harvard.edu>
wrote:
> Hi Andrew,
>
> You will need to download the umls data from the nlm. Go to their
> website
> https://urldefense.proofpoint.com/v2/url?u=https-3A__www.nlm.nih.gov_r
> esearch_umls_&d=DwIBaQ&c=qS4goWBT7poplM69zy_3xhKwEW14JZMSdioCoppxeFU&r
> =fs67GvlGZstTpyIisCYNYmQCP6r0bcpKGd4f7d4gTao&m=F5WVZckx_3sdlvoaIdIevDj
> eldXHgjUG7qjyUqYK9qQ&s=S-FPNAvXFdKt7MWWJUkn5X2s1PiDkbuhZCFUphCH3t4&e=
> and use the "Downloads" button in the "Access" panel.
>
> I will put some more instructions on the wiki page when I get a chance.
>
> After you download and build a local copy of the umls, run the gui.
> In your ctakes resources/ directory, go to
> org/apache/ctakes/dictionary/lookup/fast/
> and you will see an xml file and a directory named after your custom
> dictionary. Copy those to the headless server in that ctakes'
> resources/org/apache/../fast/ directory.
>
> See the (info) panel at the bottom of
> https://urldefense.proofpoint.com/v2/url?u=https-3A__cwiki.apache.org_
> &d=DwIBaQ&c=qS4goWBT7poplM69zy_3xhKwEW14JZMSdioCoppxeFU&r=fs67GvlGZstT
> pyIisCYNYmQCP6r0bcpKGd4f7d4gTao&m=F5WVZckx_3sdlvoaIdIevDjeldXHgjUG7qjy
> UqYK9qQ&s=XQZYxmJ5KLB6jcx4ZAPQ-EtQt05nUbUim-hiI5snN_M&e=
> confluence/display/CTAKES/Dictionary+Creator+GUI
> To see how to point ctakes to your custom gui.
>
> By the way, if you have a small number of terms and don't need the
> umls you can manually create a bar separated value (bsv) file. I am
> guessing that you have more than a few terms.
>
> Sean
>
> -----Original Message-----
> From: Andrew Phillips [mailto:skeledrew@gmail.com]
> Sent: Wednesday, June 28, 2017 6:28 PM
> To: dev@ctakes.apache.org
> Subject: Re: Building a Custom cTAKES Dictionary [EXTERNAL]
>
> Hi Sean,
>
> I haven't found a good way to launch GUIs remotely (BTW I'm also using
> Linux on my machine). However I also have a local cTAKES install and
> I'm trying to do the modification there. However I'm not sure what to
> enter into the UMLS installation field, as I cannot find a META
> directory or RRF files. Also what file(s) woud I transfer to the
> server once I have built the new dictionary?
>
> Thanks,
> Andrew
>
> *Andrew Phillips*
> Computer Technician / Programmer and Mobile Experience Consultant
> Phone: (678) 753-5313
> Email: skeledrew@gmail.com
> LinkedIn: www.linkedin.com/in/aphillipstech
>
> "A man may imagine things that are false, but he can only understand
> things that are true, for if the things be false, the apprehension of
> them is not understanding." - Isaac Newton
>
> On 28 June 2017 at 12:35, Finan, Sean
> <Se...@childrens.harvard.edu>
> wrote:
>
> > Hi Andrew,
> >
> > Can you xWin (or other) to the server to launch gui applications?
> > If so, try the dictionary creator gui:
> > https://urldefense.proofpoint.com/v2/url?u=https-3A__cwiki.apache.or
> > g_
> > &d=DwIBaQ&c=qS4goWBT7poplM69zy_3xhKwEW14JZMSdioCoppxeFU&r=fs67GvlGZs
> > tT
> > pyIisCYNYmQCP6r0bcpKGd4f7d4gTao&m=375tbajCm8PjEi_45ahKX7bDghJ_w3fOBI
> > h9 vAMpylU&s=0JwVENy0eLeCkfQVSfkUBwhF_xAzBzXGYCWVX9_e65U&e=
> > confluence/display/CTAKES/Dictionary+Creator+GUI
> >
> >
> > Sean
> >
> > -----Original Message-----
> > From: Andrew Phillips [mailto:aphillips5@luc.edu]
> > Sent: Wednesday, June 28, 2017 1:14 PM
> > To: dev@ctakes.apache.org
> > Subject: Building a Custom cTAKES Dictionary [EXTERNAL]
> >
> > Hello,
> >
> > I am new to cTAKES and I'm trying to create a custom dictionary of
> > additional terms related to alcohol. How can I go about adding
> > semantic types such as food (T168), etc that aren't available by
> > default to the dictionary? The cTAKES install is located on a
> > headless Linux server that I access via SSH.
> >
> > Thank you,
> > Andrew
> >
> > *Andrew Phillips*
> > GitHub: github.com/skeledrew
> > LinkedIn: www.linkedin.com/in/aphillipstech
> >
>
Re: Building a Custom cTAKES Dictionary [EXTERNAL]
Posted by Andrew Phillips <ap...@luc.edu>.
Hi Sean,
It took a while to figure out how to setup everything and run tests, but I
have now successfully generated CUI files containing the terms that were
missing before.
Thanks for your help.
Andrew
*Andrew Phillips*
GitHub: github.com/skeledrew
LinkedIn: www.linkedin.com/in/aphillipstech
On 28 June 2017 at 19:31, Finan, Sean <Se...@childrens.harvard.edu>
wrote:
> Hi Andrew,
>
> You will need to download the umls data from the nlm. Go to their website
> https://www.nlm.nih.gov/research/umls/
> and use the "Downloads" button in the "Access" panel.
>
> I will put some more instructions on the wiki page when I get a chance.
>
> After you download and build a local copy of the umls, run the gui. In
> your ctakes resources/ directory, go to org/apache/ctakes/dictionary/lookup/fast/
> and you will see an xml file and a directory named after your custom
> dictionary. Copy those to the headless server in that ctakes'
> resources/org/apache/../fast/ directory.
>
> See the (info) panel at the bottom of https://cwiki.apache.org/
> confluence/display/CTAKES/Dictionary+Creator+GUI
> To see how to point ctakes to your custom gui.
>
> By the way, if you have a small number of terms and don't need the umls
> you can manually create a bar separated value (bsv) file. I am guessing
> that you have more than a few terms.
>
> Sean
>
> -----Original Message-----
> From: Andrew Phillips [mailto:skeledrew@gmail.com]
> Sent: Wednesday, June 28, 2017 6:28 PM
> To: dev@ctakes.apache.org
> Subject: Re: Building a Custom cTAKES Dictionary [EXTERNAL]
>
> Hi Sean,
>
> I haven't found a good way to launch GUIs remotely (BTW I'm also using
> Linux on my machine). However I also have a local cTAKES install and I'm
> trying to do the modification there. However I'm not sure what to enter
> into the UMLS installation field, as I cannot find a META directory or RRF
> files. Also what file(s) woud I transfer to the server once I have built
> the new dictionary?
>
> Thanks,
> Andrew
>
> *Andrew Phillips*
> Computer Technician / Programmer and Mobile Experience Consultant
> Phone: (678) 753-5313
> Email: skeledrew@gmail.com
> LinkedIn: www.linkedin.com/in/aphillipstech
>
> "A man may imagine things that are false, but he can only understand
> things that are true, for if the things be false, the apprehension of them
> is not understanding." - Isaac Newton
>
> On 28 June 2017 at 12:35, Finan, Sean <Se...@childrens.harvard.edu>
> wrote:
>
> > Hi Andrew,
> >
> > Can you xWin (or other) to the server to launch gui applications?
> > If so, try the dictionary creator gui:
> > https://urldefense.proofpoint.com/v2/url?u=https-3A__cwiki.apache.org_
> > &d=DwIBaQ&c=qS4goWBT7poplM69zy_3xhKwEW14JZMSdioCoppxeFU&r=fs67GvlGZstT
> > pyIisCYNYmQCP6r0bcpKGd4f7d4gTao&m=375tbajCm8PjEi_45ahKX7bDghJ_w3fOBIh9
> > vAMpylU&s=0JwVENy0eLeCkfQVSfkUBwhF_xAzBzXGYCWVX9_e65U&e=
> > confluence/display/CTAKES/Dictionary+Creator+GUI
> >
> >
> > Sean
> >
> > -----Original Message-----
> > From: Andrew Phillips [mailto:aphillips5@luc.edu]
> > Sent: Wednesday, June 28, 2017 1:14 PM
> > To: dev@ctakes.apache.org
> > Subject: Building a Custom cTAKES Dictionary [EXTERNAL]
> >
> > Hello,
> >
> > I am new to cTAKES and I'm trying to create a custom dictionary of
> > additional terms related to alcohol. How can I go about adding
> > semantic types such as food (T168), etc that aren't available by
> > default to the dictionary? The cTAKES install is located on a headless
> > Linux server that I access via SSH.
> >
> > Thank you,
> > Andrew
> >
> > *Andrew Phillips*
> > GitHub: github.com/skeledrew
> > LinkedIn: www.linkedin.com/in/aphillipstech
> >
>
Re: Building a Custom cTAKES Dictionary [EXTERNAL]
Posted by Andrew Phillips <sk...@gmail.com>.
Hi Sean,
I haven't found a good way to launch GUIs remotely (BTW I'm also using
Linux on my machine). However I also have a local cTAKES install and I'm
trying to do the modification there. However I'm not sure what to enter
into the UMLS installation field, as I cannot find a META directory or RRF
files. Also what file(s) woud I transfer to the server once I have built
the new dictionary?
Thanks,
Andrew
*Andrew Phillips*
Computer Technician / Programmer and Mobile Experience Consultant
Phone: (678) 753-5313
Email: skeledrew@gmail.com
LinkedIn: www.linkedin.com/in/aphillipstech
"A man may imagine things that are false, but he can only understand things
that are true, for if the things be false, the apprehension of them is not
understanding." - Isaac Newton
On 28 June 2017 at 12:35, Finan, Sean <Se...@childrens.harvard.edu>
wrote:
> Hi Andrew,
>
> Can you xWin (or other) to the server to launch gui applications?
> If so, try the dictionary creator gui: https://cwiki.apache.org/
> confluence/display/CTAKES/Dictionary+Creator+GUI
>
>
> Sean
>
> -----Original Message-----
> From: Andrew Phillips [mailto:aphillips5@luc.edu]
> Sent: Wednesday, June 28, 2017 1:14 PM
> To: dev@ctakes.apache.org
> Subject: Building a Custom cTAKES Dictionary [EXTERNAL]
>
> Hello,
>
> I am new to cTAKES and I'm trying to create a custom dictionary of
> additional terms related to alcohol. How can I go about adding semantic
> types such as food (T168), etc that aren't available by default to the
> dictionary? The cTAKES install is located on a headless Linux server that I
> access via SSH.
>
> Thank you,
> Andrew
>
> *Andrew Phillips*
> GitHub: github.com/skeledrew
> LinkedIn: www.linkedin.com/in/aphillipstech
>