Abstract
Knowledge-Level Integration for JaCaMo
The specification of agent systems comprises different dimensions normally defined using distinct formalisms. This lack of uniform representations makes it harder to express how each level affects the others. To address this problem, we propose a semantic representation that integrates the formalisms that originally cover a single multi-agent system dimension, such as agency, environment, or social organisation. In doing this, we align current trends in semantic technologies and in knowledge representation for agents, environments,
and organisations thus providing MAS designers with a unified approach at the
knowledge-level for the development of complex systems based on collaborative
agents. Our approach uses three ontologies to represent the multi-agent system dimensions, and we exemplify how separate platforms operating at each dimension can be integrated. We discuss the implications of such integrated view for designing agents, and highlight advantages of semantic representations for agent-based software development.
The specification of agent systems comprises different dimensions normally defined using distinct formalisms. This lack of uniform representations makes it harder to express how each level affects the others. To address this problem, we propose a semantic representation that integrates the formalisms that originally cover a single multi-agent system dimension, such as agency, environment, or social organisation. In doing this, we align current trends in semantic technologies and in knowledge representation for agents, environments,
and organisations thus providing MAS designers with a unified approach at the
knowledge-level for the development of complex systems based on collaborative
agents. Our approach uses three ontologies to represent the multi-agent system dimensions, and we exemplify how separate platforms operating at each dimension can be integrated. We discuss the implications of such integrated view for designing agents, and highlight advantages of semantic representations for agent-based software development.
| Original language | English |
|---|---|
| Title of host publication | 5th International Workshop on Collaborative Agents – Research & Development, CARE for Intelligent Mobile Services |
| Number of pages | 6 |
| Publication status | Published - 2014 |
| Externally published | Yes |