Evolving Ontological Knowledge Bases through Agent Collaboration
Evolving Ontological Knowledge Bases through Agent Collaboration
This paper presents initial work that will enable an agent to augment its ontology to incorporate required knowledge from other agents, in order to let it answer domain related queries. Specifically, our agents are heterogeneous, whereby an agent has its own interest domain and represents this with an ontology that contains relevant conceptualisations. These agents have intersecting domain interests and their ontologies represent a set of overlapping concepts with alternative symbolic representations. In this setting, our proposed approach focuses on reducing the costs associated with acquiring knowledge through collaboration, and augmenting axioms into an agent’s ontology. In order to achieve this, we consider incorporating knowledge to reduce the number of messages required to answer repetitive domain related queries that require mediation, and select a shared set of axioms that represent conceptual knowledge. We present results from our approach and identify the number of messages and axioms required for a repeated transaction. These preliminary results show that augmenting an agent’s ontology can indeed reduce the number of messages and axioms required.
agent, ontology, augmentation, collaboration, fragment
Packer, Heather S.
0e86c31f-6460-4bbd-b6ac-c717ee2cbd96
Payne, Terry
0bb13d45-2735-45a3-b72c-472fddbd0bb4
Gibbins, Nicholas
98efd447-4aa7-411c-86d1-955a612eceac
Jennings, Nicholas R.
ab3d94cc-247c-4545-9d1e-65873d6cdb30
2008
Packer, Heather S.
0e86c31f-6460-4bbd-b6ac-c717ee2cbd96
Payne, Terry
0bb13d45-2735-45a3-b72c-472fddbd0bb4
Gibbins, Nicholas
98efd447-4aa7-411c-86d1-955a612eceac
Jennings, Nicholas R.
ab3d94cc-247c-4545-9d1e-65873d6cdb30
Packer, Heather S., Payne, Terry, Gibbins, Nicholas and Jennings, Nicholas R.
(2008)
Evolving Ontological Knowledge Bases through Agent Collaboration.
Sixth European Workshop on Multi-Agent Systems, Bath.
Record type:
Conference or Workshop Item
(Paper)
Abstract
This paper presents initial work that will enable an agent to augment its ontology to incorporate required knowledge from other agents, in order to let it answer domain related queries. Specifically, our agents are heterogeneous, whereby an agent has its own interest domain and represents this with an ontology that contains relevant conceptualisations. These agents have intersecting domain interests and their ontologies represent a set of overlapping concepts with alternative symbolic representations. In this setting, our proposed approach focuses on reducing the costs associated with acquiring knowledge through collaboration, and augmenting axioms into an agent’s ontology. In order to achieve this, we consider incorporating knowledge to reduce the number of messages required to answer repetitive domain related queries that require mediation, and select a shared set of axioms that represent conceptual knowledge. We present results from our approach and identify the number of messages and axioms required for a repeated transaction. These preliminary results show that augmenting an agent’s ontology can indeed reduce the number of messages and axioms required.
Text
eumas.pdf
- Accepted Manuscript
More information
Submitted date: December 2008
Published date: 2008
Additional Information:
Event Dates: November 2008
Venue - Dates:
Sixth European Workshop on Multi-Agent Systems, Bath, 2008-11-01
Keywords:
agent, ontology, augmentation, collaboration, fragment
Organisations:
Web & Internet Science, Agents, Interactions & Complexity
Identifiers
Local EPrints ID: 266941
URI: http://eprints.soton.ac.uk/id/eprint/266941
PURE UUID: e163d69e-662d-4063-869c-76049bf0745e
Catalogue record
Date deposited: 25 Nov 2008 19:46
Last modified: 15 Mar 2024 02:59
Export record
Contributors
Author:
Heather S. Packer
Author:
Terry Payne
Author:
Nicholas Gibbins
Author:
Nicholas R. Jennings
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics