Photos of 2016 AIM-WEST Workshop in Grenoble

img_7217 img_7220 img_7223  img_7183   img_7186 img_7187  img_7189 img_7190   img_7193 img_7194  img_7196  img_7242 img_7244 img_7246 img_7247 img_7248 img_7250 img_7254 img_7255 img_7259 img_7265 img_7270 img_7275dsc_0043dsc_0044

Posted in Uncategorized

Final AIM-WEST Workshop will be held in Grenoble (3-4 Oct)

Final AIM-WEST Workshop will be held in Grenoble (3-4 Oct)

See details here

(The workshop is joint with PARSEME-FR WS)

Posted in Uncategorized

AIM-WEST 2015 short report for CNRS

The AIM-WEST 2015 short report for CNRS can be downloaded from HERE.

Posted in Uncategorized

Presentation of AIM-WEST during LICIA Workshop on Oct 23d 2015

The AIM-WEST project was presented during LICIA Workshop on Oct 23d 2015.

pres-licia

Posted in Uncategorized

Photos of 2015 AIM-WEST Workhop in Porto Alegre

IMG_4601

IMG_4583IMG_4596

IMG_4607IMG_4608

IMG_4589IMG_4587

Posted in Uncategorized

Partners’ Plans for next year in AIM-WEST

LIG
-MWE annotation: release of the corpus soon (after verifications, etc)
-MWE translation: smarter approaches ; collab with UFRGS ; add compositionality features into the SMT model… apply this to one or two language pairs
-Deep Multilingual Parsing: with Pablo & Helena (LIG-UFScar)
LIF
-Monolingual identification of MWEs
-Could be integrated into SMT (maybe try to work on the shared task data)
-SOme collegues at LIF work on lexical simplification,  this could be a future activity
LATTICE
-Consolidate the internship student’s work
-Parsing MWEs
UFSCAR
-Deep multilingual parsing
-Automatically extracting lists of bilingual MWEs
-(need to talk more with Helena)
UFRGS
-Distributionnal models for MWE extraction
-Applications: simplification,
-Interest in resources that have been developed in the project
WE SHOULD THINK ABOUT THE NEXT PROJECT (renew this kind of project? cafes-cofecub?)
Posted in Uncategorized

Program of 2d AIM-WEST Workshop in Porto-Alegre

The program of 2d AIM-WEST Workshop in Porto-Alegre is now online !!

http://aim-west.imag.fr/project-meetings/2nd-aim-west-workshop-porto-alegre/

Posted in Uncategorized

LIG work on MWE annotation now online

An annotation guide for MWE annotation, as well as a paper has been put online on the publication page – see the links below

  1. Agnès Tutin “Annotation guidelines for multi-word expressions“. Scientific Report – AIM-WEST. 2015. Annotation guidelines for MWE_20_april_2015
  2. Agnès Tutin, Emmanuelle Esperança-Rodier, Manolo Iborra, Justine Reverdy “Annotation of multiword expressions in French” Proceedings of Conference EUROPHRAS2015 –  Malaga, Spain, 1-2 July 2015. Europhras2015_Tutin_et_al_vfinale
Posted in Uncategorized

Next AIM-WEST Meeting on Oct 20-21 in Porto Alegre (Brasil)

Draft Planning

20 and 21 October – AIM-West workshop – Possible presentations
FRANCE
*Laurent Besacier –
1 or 2 presentations  (work done at LIG in the framework of AIMWEST)
TUTORIAL – Intro course about ASR for Master (and under-gradutate students ?) :  2 hours course with a break – show some resources and tools) – afternoon of the 21st
*Carlos Ramisch – work on MWEs
*Isabelle Tellier – present the work of the intership student

UFSCAR
*Helena Caseli – cannot come but would give her slides to Laurent (they co-supervised a student this summer)
*Someone from UFSCAR – see if we can find funding for a student ?

UFRGS
*Viviane Moreira – focused crawling
*Aline
*Marco Idiart – Distributional thesaurus
*Rodrigo Wilkens – Simplification architecture

22 and 23 October
LICIA workshop – 1 day dedicated to talk about projects related to LICIA
*1 joint general talk to present AIM-WEST – Laurent and Aline
*1 talk about smart cities
*1 talk by Viviane Moreira (multilingual IR?)

*some talks about HPC (other project between LIG and UFRGS)
Posted in Uncategorized

MWE Translation Shared Task is Now Launched !!

All the training/testing material (with baseline LIG system outputs) has been made available on (thanks to Zied Elloumi) :

(this is a non permanent link – so please download this before April 15th)
READMEs are now both in French and English – Basically we provide train/dev/test data for Fr-En SMT ; including corpora with higher density of MWEs and some of them with MWE annotation on the source side ; the evaluation script is also made available so that we all use the same one…
We’ll have internship student that will analyze the SMT errors on these corpora and they will start in May… So early May can be a (informal) deadline for us to submit systems output
Posted in Uncategorized