ctakes-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Yi-Wen Liu <yiwen...@usc.edu>
Subject Re: CR descriptor
Date Mon, 09 Nov 2015 19:54:33 GMT
Hi,

Thanks! That helps a lot!

My data format is text file, can I assume other descriptors work like that?
Because there are also multiple AE xmls in
desc\ctakes-core\desc\analysis_engine, and CC xmls
in desc\ctakes-core\desc\cas_consumer, and I only need one of each.
It seems like they are AggregateAE and FilesInDirectoryCasConsumer.example
in my case?

Thanks,
Yi-Wen

On Sat, Nov 7, 2015 at 3:46 AM, Miller, Timothy <
Timothy.Miller@childrens.harvard.edu> wrote:

> Hi Yi-Wen,
> There are different collection readers for different data sources, and we
> usually try to give them descriptive names.
> FilesInDirectoryCollectionReader is one of the most useful ones -- it will
> look for a list of text files in a directory and put one file in each cas.
> If your data is in that format or is easy to convert to that format that's
> probably a good starting point.
> Tim
>
> ________________________________________
> From: Yi-Wen Liu <yiwenliu@usc.edu>
> Sent: Saturday, November 7, 2015 12:59 AM
> To: dev@ctakes.apache.org
> Subject: CR descriptor
>
> Hi,
>
> I am looking for the main collection reader(CR) in cTAKES in order to do
> scale out on UIMA DUCC. And in des/ctakes-core/des/collection_reader/,
> there are multiple CR xml files. I am not sure which is the one that should
> be specified in DUCC's job file...are they all necessary in cTAKES job or
> some of them are offered for other reference?
>
> I am not familiar with cTAKES structure so hope somebody can help me out,
> thanks!
>
> Thanks,
> Yi-Wen
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message