Re: [gate-users] How to extract the words that annotated

[Vuong D. N. <vuo...@ya...>]

Dear Pradhan,

Thanks alot. In the waiting time, I also tried  and find out a method:

This part I wrote
           AnnotationSet defaultAnnots = doc.getAnnotations();
//AnnotationSet  annSet = ...;
String type = "Person";
//Get all "type"  annotations
AnnotationSet persSet = defaultAnnots.get(type);


I  believe equal to this in your part:
AnnotationSet  personSet = (AnnotationSet)  bindings.get("person");

 but, when you write:
if  (personSet.size() > 3)
what does this mean ?

For  example personSet annotations contain:

A, B, C, B, B, D, E, B, O ,  X, Y , Z, C, 
can the above "if" condition return me value "B" which  appears more than 3 
times?
If yes, its really smart method, because I  have to write longer code to get it:

List  persList = new ArrayList(persSet);
Collections.sort(persList, new  gate.util.OffsetComparator());



//Iterate
Iterator  persIter = persList.iterator();
while(persIter.hasNext()){
      Annotation temp=(Annotation)persIter.next();
//                String  nameOfAnnotation=temp.getType();
//      annotationSets.add(doc.getAnnotations((String)persIter.next()));
        String content = stringFor(doc, temp);

Only From here, I  can get  content as A, B, C ... separately and I start to 
count the time  it appears. 


That's for discussion purpose only. :)

Thank  you all ! 

Thomas





________________________________
From: Shekhar Pradhan <She...@ma...>
To: Vuong Dao Nghe <vuo...@ya...>; gat...@li...
Sent: Wed, November 24, 2010 11:34:02 PM
Subject: Re: [gate-users] How to extract the words that annotated

Hi,
Try this (assumes that on the LHS of the JAPE rule) you use the label "person" 
for occurrences of Person:
AnnotationSet personSet = (AnnotationSet) bindings.get("person");
if (personSet.size() > 3) //more than 3 Person annotations
{String personStr;
for (Annotation personAnn: AnnotationSet)
{
try {
      personStr = doc.getContent().getContent(
                personAnn.getStartNode().getOffset(),
        personAnn.lgetEndNode().getOffset()).toString();
//Now you can do what you want with personStr
}
catch (InvalidOffsetException e)
    {
        throw new GateRuntimeException(e);
    }
}

Hope this helps.

Cheers,
Shekhar Pradhan

-----Vuong Dao Nghe <vuo...@ya...> wrote: -----


To: gat...@li...
>From: Vuong Dao Nghe <vuo...@ya...>
>Date: 11/22/2010 03:21AM
>Subject: [gate-users] How to extract the words that annotated
>
>
>Hi all,
>
>I am looking for the answer for this but long time couldn't find the correct 
>way.
>
>When you run a gate pipeline, then open the document, choose annotation Sets 
>then tick an Annotation, for example Person, you will see all words related 
>Person will be highlighted.
>How can I print out all these highlighted words ? Now I can write in java to 
>load the pipeline with document and all necessary PRs. But I dont know how is 
>the code to print out the words the be long to Person annotation for example ?
>
>Can I print the words annotated that appear for more than 3 times ?
>
>Thomas
>------------------------------------------------------------------------------
>Beautiful is writing same markup. Internet Explorer 9 supports
>standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
>Spend less time writing and  rewriting code and more time creating great
>experiences on the web. Be a part of the beta today
>http://p.sf.net/sfu/msIE9-sfdev2dev_______________________________________________
>
>GATE-users mailing list
>GAT...@li...
>https://lists.sourceforge.net/lists/listinfo/gate-users
>