Mailing-List: contact dev-help@ctakes.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@ctakes.apache.org
Received-SPF: pass (nike.apache.org: domain of
 Timothy.Miller@childrens.harvard.edu designates 134.174.20.73 as permitted
 sender)
From: "Miller, Timothy" <Timothy.Miller@childrens.harvard.edu>
To: "dev@ctakes.apache.org" <dev@ctakes.apache.org>
Subject: Re: suggestion for default pipelines
Thread-Topic: suggestion for default pipelines
Thread-Index: Ac9Ysf3ILp3kN1O0QBidCDkO3hyXvA==
Date: Thu, 24 Apr 2014 21:41:28 +0000
Message-ID: <E084D8EFE2B03A408B324458C5212E94244874FF@CHEXMBX3A.CHBOSTON.ORG>
References: <E084D8EFE2B03A408B324458C5212E9424346E3B@CHEXMBX3A.CHBOSTON.ORG>
 <CANuf76MZwZhOE3cOYc5Ey3ov=0Yo3Y5Sv+PSodS6MYBJDadnuw@mail.gmail.com>
 <400F3A89-0AB6-4596-86D0-8A37144C1292@apache.org>
Accept-Language: en-US
Content-Language: en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

Any preference for separate factory classes:=0A=
=0A=
class SentenceDetectorAnnotatorFactory:=0A=
=0A=
static AnalysisEngineDescription getSentenceDetectorAnnotator()=0A=
=0A=
VS=0A=
=0A=
static methods added to primitive annotators:=0A=
=0A=
class SentenceDetector (existing)=0A=
=0A=
static AnalysisEngineDescription getSentenceDetectorAnnotator()=0A=
=0A=
?=0A=
=0A=
The former can clutter up the class space while the latter extends the=0A=
length of classes, especially if there are multiple versions=0A=
(getUMLSDictionaryAnnotator(), getICD9DictionaryAnnotator(),=0A=
getMeshDictionaryAnnotator(), etc.)=0A=
=0A=
Tim=0A=
=0A=
On 04/16/2014 04:48 AM, Richard Eckart de Castilho wrote:=0A=
> It would be nice if uimaFIT provided a Maven plugin to automatically=0A=
> generate descriptors for aggregates. Maybe if we come up with a =0A=
> convention for factories, e.g. a "class with static methods that do=0A=
> not take any parameters and that return descriptors", or "methods=0A=
> that bear a specific Java annotation, e.g. @AutoGenerateDescriptor)"=0A=
> it should be possible to implement such a Maven plugin.=0A=
>=0A=
> Cheers,=0A=
>=0A=
> -- Richard=0A=
>=0A=
> On 16.04.2014, at 05:21, Steven Bethard <steven.bethard@gmail.com> wrote:=
=0A=
>=0A=
>> +1. And note that once you have a descriptor, you can generate the=0A=
>> XML, so we should arrange to replace the current XML descriptors with=0A=
>> ones generated automatically from the uimaFIT code. That should reduce=
=0A=
>> some synchronization problems when the Java code was changed but the=0A=
>> XML descriptor was not.=0A=
>>=0A=
>> Steve=0A=
>>=0A=
>> On Tue, Apr 15, 2014 at 8:52 AM, Miller, Timothy=0A=
>> <Timothy.Miller@childrens.harvard.edu> wrote:=0A=
>>> The discussion in the other thread with Abraham Tom gave me an idea I=
=0A=
>>> wanted to float to the list. We have been using some UIMAFit pipeline=
=0A=
>>> builders in the temporal project that maybe could be moved into=0A=
>>> clinical-pipeline. For example, look to this file:=0A=
>>>=0A=
>>> http://svn.apache.org/viewvc/ctakes/trunk/ctakes-temporal/src/main/java=
/org/apache/ctakes/temporal/pipelines/TemporalExtractionPipeline_ImplBase.j=
ava?view=3Dmarkup=0A=
>>>=0A=
>>> with the static methods getPreprocessorAggregateBuilder() and=0A=
>>> getLightweightPreprocessorAggregateBuilder()   [no umls].=0A=
>>>=0A=
>>> So my idea would be to create a class in clinical-pipeline=0A=
>>> (CTakesPipelines) with static methods for some standard pipelines (to=
=0A=
>>> return AnalysisEngineDescriptions instead of AggregateBuilders?):=0A=
>>>=0A=
>>> getStandardUMLSPipeline()  -- builds pipeline currently in=0A=
>>> AggregatePlaintextUMLSProcessor.xml=0A=
>>> getFullPipeline() -- same as above but with SRL, constituency parsing,=
=0A=
>>> etc., every component in ctakes=0A=
>>>=0A=
>>> We could then potentially merge our entry points -- I think Abraham's=
=0A=
>>> experience points out that this is currently confusing, as well as=0A=
>>> probably not implemented optimally. For example, either=0A=
>>> ClinicalPipelineWithUmls or BagOfCUIsGenerator would use that static=0A=
>>> method to run a uimafit-style pipeline. Maybe we can slowly deprecate=
=0A=
>>> our xml descriptors too unless people feel strongly about keeping those=
=0A=
>>> around.=0A=
>>>=0A=
>>> Another benefit is that the cTAKES API is then trivial -- if you import=
=0A=
>>> ctakes into your pom file getting a UIMA pipeline is one UimaFit call:=
=0A=
>>>=0A=
>>> builder.add(CTAKESPipelines.getStandardUMLSPipeline());=0A=
>>>=0A=
>>>=0A=
>>> I think this would actually be pretty easy to implement, but hoping to=
=0A=
>>> get some feedback on whether this is a good direction.=0A=
>>>=0A=
>>> Tim=0A=
>=0A=
=0A=
-- =0A=
Tim Miller=0A=
Instructor=0A=
Boston Children's Hospital and Harvard Medical School=0A=
timothy.miller@childrens.harvard.edu=0A=
617-919-1223=0A=
=0A=