Skip navigation.
Home
Semantic Software Lab
Concordia University
Montréal, Canada

Accessing MultiPaX features through JAPE

Printer-friendly versionPrinter-friendly versionPDF versionPDF version

Hi,

How can I extract the features of MultiPaX annotations? This JAPE code for extracting the subject feature doesn't seem to work.

Rule: Extract_MultiPaX_Features
(
{MultiPaX.sub =~ ".*"}
):label
-->
:label.MultPaX_Annotation = {rule = "Extract_MultiPaX_Features"}

Also, what JAPE code can I use to put the sentence which MultiPaX has parsed, as a feature inside an annotation? The annotation "Sentence" always has a value: {}

Also, I have been able to use the "antecedent" example code, to place the pronominal referent of a Person or Organization as a feature inside these annotations, but is there an easy way to transfer them to the MultiPaX annotation?

As you can see, I'm new to JAPE and java, I would appreciate any help.. thanks.

Edan

rene's picture

same as for other GATE annotations

There's nothing special about MultiPAX annotations with respect to JAPE. I suggest you have a look at the JAPE slides from the latest GATE Training Course (Track 1, Module 3): http://gate.ac.uk/wiki/TrainingCourseAug2010/ and the GATE Wiki JAPE Tips: http://gate.ac.uk/wiki/jape-repository/

In your code, it's not really clear what you try to match on the LHS of the rule: Is your goal to match only specific subjects? Otherwise, why not simply match all MultiPAX annotations? And what do you want to do with the extracted subjects -- create new annotations?