Mes: mayo 2009

La W3C está trabajando en una nueva especificación de OWL que está a punto de aprobarse como estándar.

Todavía no la he mirado con detalle. Dejo para una anotación posterior comentar las diferencias entre las dos, ver qué aporta realmente y si realmente es interesante para lo que estamos haciendo.

Por el momento, puedes acceder a la visión general, a las nuevas características y su sintaxis entre otras cosas.

Y aunque aún esté por aprobar el estándar, ya hay herramientas que permiten razonar con las nuevas especificaciones, como la nueva versión de Pellet.

This session begins explainig what some participants are doing related with trust because they haven’t finished yet :-(

About semantics… we’ve talk about semantics of datra, services, portocols, policies, preferences, context…

Trust issues related with semantics

• what is the data I can trust?
• what is the right ontologyu for me?
• what is the right alignment for me?

all those reasons are trust issues!!

Some idea: a big black swarn is flying above us. Trust can’t be totally predicted from past interractions (turkey example in Thaksgiving day). How uncertainy can be deal in such scenarios?

Discussion: data is related with trust, or we might talk about reliability? Trust seems to be someting more ‘deliberate’. Is based on the concept on commitment (in an agreement viewpoint)

For EUMAS meeting, two topics

• Trust in groups
• Trust in data

Coord. Axel Polleres

First a short review of the purpose of this WG. The main challenges are

• novel ways of semantic alignment
• take into account agent interaction
• ontologies+non-monotonic rules for meaning preserving and mergind theories

Main goals

• integration of ontologies and nonmon rules
• quering over distributed ontologies (maping rules and lalignmnets)
• alginment with web standards (RDF, RIF, OWL, SPARQL, SA-WSDL)

Some activities:

SPOT2009 and COIN@MALLOW

Participants (Axel’s remembered to join to the list and subscribe as ‘active member of the WG

Our points

• services
• open MAS: organisations
• content language (magentix) RDF and envelope (FIPA)
• agreement spaces -> contracts

other intersetng thigs

• coherence -> rewards, sanctions
• small world networks (navigable) to distribute

All people explaining thath are they doinjg…. I¡ll take that from the slides

URJC

Service matchmaking. Elena is going to work with them.

AI-LAB

mode ont. matching as CSP from multiple agents.

Evolution mof ontologies (RS-COE): agreed conceptualisation -> TELEX

NOMADIS Lab

Policy specification and enforcement. Not MAS, because it’s applied to sensor network (low compt. power) -> centralised policy. Using PDL (Policy description language). See workshop about contex.

DERI

Querying distributed semantic web data. Experiments: Lucene API p SoIR with >1000000 clases (data from Billion Triple Challenge)

Interesting, with a lot of work to do and a lot of things to be learned.

(NOTE: galen ontology -owl- for medical concepts)