Séminaires RALI-OLST

Extracting, representing and sharing environmental knowledge

Pilar León Araúz (pleon <at> ugr (point) es)

Lexicon Research Group, Université de Grenade

Le mercredi 10 juin 2015 à 11 h 30

Salle C-9019, Pavillon Lionel-Groulx


In this presentation we will focus on the principles that guided knowledge modelling in EcoLexicon as well as the future that lies ahead for the tool.

EcoLexicon is a multilingual terminological knowledge base (TKB) on the environment. The wide scope of the domain poses significant challenges for terminology work, namely a high degree of multidimensionality and contextual variation. In this sense, an important feature in the TKB is the representation of how conceptual propositions are more or less salient according to dynamic conceptualizations.

Propositions and their salience are extracted from a large classified corpora especifically compiled for EcoLexicon. So far, this has been done manually, based on iterative searches of keywords and knnowledge patterns. Given the scope of the domain, this task can be very tedious and time-consuming. Consequently, knowledge patterns are now being formalized in local grammars in order to identify and annotate the different structures that lexicalize semantic relations (i.e. hyponymy, meronymy and causality). There is still much work to be done, but the grammars have already been proven effective at: (1) extracting knowledge rich contexts; (2) accessing multidimensionality in a more straightforward way; (3) comparing the salience of conceptual propositions across different contexts in a more reliable way; (4) disambiguating polysemic KPs.

EcoLexicon was first conceived as a resource for knowledge acquisition, for users with both cognitive and linguistic needs (translators, tehnical writers, or even experts). However, in the future, we intend to take EcoLexicon a step further and fully exploit its contents for purposes of translation and natural language processing. This will entail sharing it across different platforms and linking it to others in the linguistic linked open data cloud.



Joignez-vous à nous avec Zoom à l'aide de cette URL.
ID de réunion : 916 9097 5818, Code : 343273.
Numéros de téléphone : https://umontreal.zoom.us/u/abitNZzLg.
One tap mobile : +14388097799,,91690975818#,,,,,,0#,,343273#


Pour recevoir les annonces hebdomadaires par courriel, envoyez un message à l'adresse majordomo@iro.umontreal.ca. Il suffit d'envoyer un message ne contenant que la ligne 'subscribe ralli' (sans les apostrophes).

Liste de tous les séminaires pour l'année :

1991 1992 1993 1994 1995 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020