Training and Learning

[borobudur <bo...@gmx.ch>]

Hi, I had a look at the UIMA architecture and I was asking myself if there is a
konzept for a training and learning phase in UIMA. 
What I learned is that UIMA chains Analysis Engines together. There is no two
phase concept like training and extraction in UIMA, isn't it?
Thanks
borobudur

Re: Training and Learning

[Tommaso Teofili <to...@gmail.com>]

Hi,
also CFE [1] has been used for machine learning purpose as you can read in
[2].
Cheers,
Tommaso

[1] :
http://uima.apache.org/sandbox.html#configurable.feature.extractor.annotator
[2] :
http://domino.research.ibm.com/comm/research_projects.nsf/pages/medicalinformatics.pubs.html/$FILE/CFE_sominsky-A4.pdf

2010/11/12 Philip Ogren <ph...@ogren.info>

> Borobudur,
>
> Do you mean "training" in the machine learning sense?  If so, UIMA does not
> directly support any notion of training and using statistical classifiers.
>  You might check out ClearTK <http://cleartk.googlecode.com> which is a
> UIMA-based project that provides support for a number of machine learning
> classifiers and supports feature extraction, training, and classification.
>
> Philip
>
>
> On 11/12/2010 7:39 AM, borobudur wrote:
>
>> Hi, I had a look at the UIMA architecture and I was asking myself if there
>> is a
>> konzept for a training and learning phase in UIMA.
>> What I learned is that UIMA chains Analysis Engines together. There is no
>> two
>> phase concept like training and extraction in UIMA, isn't it?
>> Thanks
>> borobudur
>>
>>
>>
>> -----
>> No virus found in this message.
>> Checked by AVG - www.avg.com
>> Version: 10.0.1153 / Virus Database: 424/3251 - Release Date: 11/11/10
>>
>>
>>

Re: Training and Learning

[Philip Ogren <ph...@ogren.info>]

Borobudur,

Do you mean "training" in the machine learning sense?  If so, UIMA does 
not directly support any notion of training and using statistical 
classifiers.  You might check out ClearTK 
<http://cleartk.googlecode.com> which is a UIMA-based project that 
provides support for a number of machine learning classifiers and 
supports feature extraction, training, and classification.

Philip

On 11/12/2010 7:39 AM, borobudur wrote:
> Hi, I had a look at the UIMA architecture and I was asking myself if there is a
> konzept for a training and learning phase in UIMA.
> What I learned is that UIMA chains Analysis Engines together. There is no two
> phase concept like training and extraction in UIMA, isn't it?
> Thanks
> borobudur
>
>
>
> -----
> No virus found in this message.
> Checked by AVG - www.avg.com
> Version: 10.0.1153 / Virus Database: 424/3251 - Release Date: 11/11/10
>
>

Re: Training and Learning

[Thilo Götz <tw...@gmx.de>]

On 11/12/2010 15:39, borobudur wrote:
> Hi, I had a look at the UIMA architecture and I was asking myself if there is a
> konzept for a training and learning phase in UIMA. 
> What I learned is that UIMA chains Analysis Engines together. There is no two
> phase concept like training and extraction in UIMA, isn't it?
> Thanks
> borobudur
> 

Not per se.  People use UIMA both for training and extraction,
but those are not notions that are built in.

--Thilo