Semantics is a central issue un Agreement Technologies.
This working group explores novel ways of semantic alignment of ontologies taking into account the current interaction state that agents are involved in, as well as the semantics of distributed knowledge based systems and argumentation framework. It studies the logical relations between ontologies, knowledge-bases, sets of non-monotonic rules (or any integrated combination of these) that are relevant for aligning and merging theories in meaning-preserving ways. In particular, the working group looks into how policies, measures of trust and argumentation can be used to decide which alignments and mappings between heterogeneous formats to apply, assuming that data, alignments, as well as trust-relevant meta-data is published and reusable by agents on the Web. Such published alignments shall allow combinations of knowledge bases based on static mapping rules, dynamic service calls and allow partial revocations of data published by other agents. Moreover, the group is interested in how multiple ontologies, knowledge bases or datasets from different sources can be exploited together, via various forms of mediation including the aforementioned alignments. The working group examines mathematical theories of information and information flow to describe the alignment of agent terminologies, as well as reasoning techniques to reuse and adapt previous alignments to new interactions. The working group also explore protocols and algorithms interweaving alignment and negotiation in cooperation with WG 4 and 5 so as to analyse and design basic resource management mechanisms for locating adequate services in open, large-scale, decentralized systems. In the spirit of the Semantic Web and Semantic Web Services initiatives, the WG will to this end further investigate how existing standardization efforts can be enhanced to cater for agreement relevant information. Particularly, this WG will address the following challenges around semantic alignment for agreement technologies:
- Integration of Ontologies and nonmonotonic Rules: Nonmonotonicity seems inevitable for ontological agreement between agents in order to deal with consistent, but closed subsets of ontologies in open environments which would otherwise involve logically inconsistent superfluous ontological information. This may include various forms of non-standard reasoning framework, including belief revision, defeasable logic, distributed logics and so on.
- Querying over distributed ontologies involving mapping rules and alignments: Queries for data, grounded in distributedly published ontologies and mapping/alignment rules, impose new challenges. The WG will carefully explore the right tradeoffs between expressivity and efficiency for ontology, mapping and query languages.
- Alignment with existing Semantic Web standards:The WG will investigate how existing standardization efforts in the area such as RDF, RIF, SPARQL and SA-WSDL can be enhanced to cater for agreement relevant information such as trust, provenance and policies and how this additions can be exploited for open, distributed access to heterogeneous data and service. For instance, the enhancement of current standards for Digital Rights Management (DRM) and Software Licensing by semantic annotations which enable automatic agreement on licence terms falls in this category.
Working group co-chairs:
- George Vouros
- Antoine Zimmermann