P1EDA options

[Tom Bosc]

Hello!

I am interested in using P1EDA. I understand that it is still experimental.

I know that there is java code to play with but I don't have so much tim to play with it.

I'm working with very noisy social media data so I'd need to tweak the preprocessing.

What are the possible options?

For example, can I add any preprocessing in the "activatedLAP" field? Will it be used by the EDA?

I send you again the config. file I'm using.

<?xml version="1.0" encoding="utf-8"?>

<configuration>
<section name="PlatformConfiguration">
            <!-- <property name="activatedEDA">eu.excitementproject.eop.alignmentedas.p1eda.instances.SimpleWordCoverageEN</property> -->
            <property name="activatedEDA">eu.excitementproject.eop.alignmentedas.p1eda.sandbox.FNR_EN</property>
            <property name="language">EN</property>
            <property name="activatedLAP">eu.excitementproject.eop.lap.dkpro.TreeTaggerEN</property>
</section>

<section name="eu.excitementproject.eop.alignmentedas.p1eda.sandbox.FNR_EN">
    <property name="modelFile">/tmp//P1EDA_Base_EN.model</property>
    <property name="trainDir">/tmp/EN/dev/</property>
    <property name="testDir">/tmp/EN/test/</property>
</section>

</configuration>

Thanks a lot,

Tom

[Tae-Gil Noh]

Hello Tom, Roberto asked me to support you if possible :-)

I am sorry but this configuration file and utilizing the configuration is not part of P1EDA itself. I am not trying to avoid the question --- but, literally I have no idea how they work at all.  (e.g. I never used them before). 

For example, the options like activatedEDA or activatedLAP are part of "runner" (top level application) code that initiates and utilizes EDAs as library --- I guess this "runner" has been designed as common command line wrappers. But I never use them and have no idea how they work.

Roberto, can you help me/him to understand those things? how activatedLAP and activatedEDA works? Who reads those configuration file? activatedLAP initiates and run those LAPs as designed in common procedure?

P1EDA itself follows all the same interfaces that other EDA follows. So basically, if would work the same way as long as the runner code supports it correctly. So my expectation would be they would work. But I have no good idea if they really will or not.

Sincerely,
Gil

[Tae-Gil Noh]

(Correct me if I am wrong, Roberto --- e.g. in case if runner wrapper is also generalized for P1EDA, and that parameter should be passed to some other runner module, etc. ) I guess this part of the configuration means / expects P1EDA to read those configuration values. Is this correct? If so, no, it won't work. --- Sorry about that.

<section name="eu.excitementproject.

eop.alignmentedas.p1eda.sandbox.FNR_EN">
    <property name="modelFile">/tmp//P1EDA_Base_EN.model</property>
    <property name="trainDir">/tmp/EN/dev/</property>
    <property name="testDir">/tmp/EN/test/</property>
</section>

Sincerely,
Gil

On Wednesday, January 27, 2016 at 4:03:05 PM UTC+1, tom.bosc wrote:

[Roberto Zanoli]

Basically the PlatformConfiguration section in the configuration file is used by EOPRunner for preprocessing data and running EDAs. You can see EOPRunner as the entry point of the EOP. It is useful when you want to use the EOP from the command line. The quick start tutorial shows you hot to use it. In this case for example the preprocessing is done by the TreeTagger (activatedLap) while the EDA to be used is P1EDA (activatedEDA). Given that different EDAs could require a different preprocessing of the data developers (for example an EDA could require dependency parsing while another one could require pos tagging only) generally developers put all the possible options that one can use in the configuration file  itself or in the wiki page of their EDA. This means that if you substitute the current lap with another one that is not thought to work with that EDA,

the EDA will not work correctly and probably it will generate an exception when running.

--
You received this message because you are subscribed to the Google Groups "eop-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to eop-users+...@googlegroups.com.
To post to this group, send email to eop-...@googlegroups.com.
Visit this group at https://groups.google.com/group/eop-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/eop-users/503722294.13447446.1453906970310.JavaMail.zimbra%40inria.fr.

--

Roberto Zanoli
Technologist
Human Language Technologies Research Unit
Fondazione Bruno Kessler
Via Sommarive 18, 38123 Povo - Trento (Italy)

T: +39 0461 314565

zan...@fbk.eu

[Tae-Gil Noh]

Thanks for update, Roberto. That makes sense. So the following question from Tom is already answered.

> For example, can I add any preprocessing in the "activatedLAP" field? Will it be used by the EDA?

Yes. It will be. P1EDA requires LAPs with "Lemma" and "Token". Any descent LAP that provides those will  with P1EDA.

===

Roberto --- I've checked the P1EDA code and realized that Vivi has extended the P1EDA to support EOPRunner. For example, config values such as "modelFile", "trainDir" are added by her effort.
If Tom wants to use EOPRunner to run P1EDA --- then I would suggest the best way would be asking Vivi for advice and/or any working configuration of EOPRunner on P1EDA.

Sincerely,
Gil