W3C home > Mailing lists > Public > public-ontolex@w3.org > April 2012

can't participate to the audioconference: Introduction of my work and suggestions for requirements

From: Maxime Lefrançois <maxime.lefrancois@inria.fr>
Date: Fri, 13 Apr 2012 14:59:59 +0200 (CEST)
To: Philipp Cimiano <cimiano@cit-ec.uni-bielefeld.de>, public-ontolex@w3.org
Message-ID: <2048998993.54687.1334321999535.JavaMail.root@zmbs3.inria.fr>

Dear all, 

I won't be able to participate to today's discussion, so I'd like to write down the few facts that I would have liked to mention. 

First of all, I sent the open call for use cases to : < web.semantique@lists-sop.inria.fr > and "atala < ln@cines.fr > ". They were broadcasted on March the 22nd and I guess Philip should be the one to receive replies if any. 

Also, as I did not already introduce myself, I'll do that first. 



I'm in my 2nd year of Ph.D. in the WIMMICS team of INRIA Sophia-Antipolis, FR, under the supervision of Fabien Gandon (Semantic Web) and Christian Boitet (NLP, lig-lab Grenoble, GETALP team). 

I focus on porting and enhancing the lexicon model of a leading-edge linguistic theory to the semantic web formalisms. 

The theory is the Meaning-Text Theory - MTT - http://en.wikipedia.org/wiki/Meaning%E2%80%93text_theory 

The (not-formalized-) lexicon model is the Explanatory Combinatorial Dictionary - ECD - http://en.wikipedia.org/wiki/Explanatory_Combinatorial_Dictionary 

As I'm in the field of the Semantic Web and the Lexicology, I do want to get involved in the MSW-community group and the MLW-LT working group as I'd like my work to be complementary / compatible / integratable / ... with the futur web standards. 

For now my main contribution has been to show how simple Lexical Definitions in the ECD could be formalised in the SW formalisms, and to plan a roadmap to conceive a pivot-based ECD using SW formalisms - Maxime Lefrançois and Fabien Gandon, ILexicOn: toward an ECD-compliant interlingual lexical ontology described with semantic web formalisms , Proc. 5th International Conference on Meaning-Text Theory, MTT, September 2011, Barcelona, Spain , pp.155-164 

I'm currently working on more complex linguistic definitions, and on the formalization of Lexical Functions - http://en.wikipedia.org/wiki/Lexical_function 

I am also to port the growing database of a recently begun French ECD Lexicographers project to my formalisms (RELIEF project - http://www.atilf.fr/spip.php?article908 ) , and to design a platform to manage the SW-formalized ECD. 



I thus fall into the [LDOT] use case, i.e., rich Linguistic Descriptions. 

To formalize the ECD, I need to use the semantics of OWL 2 (SROIQ) with the following modeling choices (rough description): 
- S emantic relations are owl object properties that link lexical unit instances to form the semantic representation of a utterance, and the semantic part of the lexicon thus consists of the description of lexical unit classes (as owl classes), with restrictions over the use of semantic relations. To parallel with the lemon model, please tell me if I mistake but I feel like using the MTT linguistic theory makes me refine tremendously the description of Lexical Senses and arguments. 
- Lexical Functions are annotation properties or sparql construct rules. 

- Lexical unit classes are linked to any rdf resource through annotation properties (lemon's "reference" property) 
- Lexical unit classes are linked to multiple lexical unit lexicalization classes through annotation properties (lemon's "lexeme" property) 

What I suggest as a requirement for the ontolex is that model refinements such as mine don't necessarily fall into an unnecessary owl flavour (I don't want to fall into OWL-full for instance). 

Thus, if the model is an OWL ontology (requirement R1), it must be carefully designed so as not to limit its openness (requirement R4). 
Requirement R5 and the thread about SKOS and SKOS-XL makes me vote for integrating the SKOS model, 

About other requirements, I guess we'll need to classify them into (core model / extension model) 

General Issues: 
How to represent meaning of a lexical sense, and do we need lexical relations ? 
It could first be done simply by basic property in the core model, but I really deal with this precise point in my phd, (formal representation of lexical senses, and lexical functions (a generalization of lexical relations) ), I suggest we'll deal with this in an extension to the core model (this would be a great result for me). 

About the presentation of requirements, Fabien Gandon just told me that w3c working groups usually classify requirements into three categories: Main requirements, Optional requirements (i.e., if we have time), and Requirements not in the charter (no value judgment, this only means we don't think we'll have time to deal with them) 

Kind regards, 
Maxime Lefrançois 
Ph.D. Student, INRIA - WIMMICS Team 

p.s., anyone goes to the WWW2012 conference ? 

----- Mail original -----

De: "Philipp Cimiano" < cimiano@cit-ec.uni-bielefeld.de > 
À: public-ontolex@w3.org 
Envoyé: Jeudi 12 Avril 2012 15:07:53 
Objet: Reminder: 5th Ontolex telco tomorrow, 3-4 CET 

Dear all, 

this is a remainder that we will have our 5th Ontolex telco tomorrow 
April 13th, 3pm-4pm CET. 

Details about how to access the telco can be found here: 

Tomorrow, I would like to start a brainstorming on the requirements of 
the model. I have compiled a first list of requirements on the basis of 
our use case descriptions: 


Talk to you all tomorrow. 


Prof. Dr. Philipp Cimiano 
Semantic Computing Group 
Excellence Cluster - Cognitive Interaction Technology (CITEC) 
University of Bielefeld 

Phone: +49 521 106 1224 9 
Fax: +49 521 106 1241 2 
Mail: cimiano@cit-ec.uni-bielefeld.de 

Room H-127 
Morgenbreede 39 
33615 Bielefeld 

Received on Friday, 13 April 2012 13:00:35 UTC

This archive was generated by hypermail 2.3.1 : Monday, 23 October 2017 10:57:25 UTC