Finally, agent owners should closely interact with
the evolutionary process of their agents (Zhu & Guan, 2001).
Agents are both self-interested and social. They have their own goals, but they also
seek for collaboration. The interaction with other agents should in some way or
another help each individual agent to fulfill one or more of its goals. The motivation
for collaboration can arise for the purpose of one temporary task, for example,
information retrieval, or for a long-term objective??”co-evolution.
We exploit agent evolution in the knowledge level in this chapter, as we deem that
knowledge base is the cornerstone of all emergent behaviors of agents. Although
agents may gain new knowledge from self-reasoning, we concentrate on knowledge
exchange among agents as we argue that this is the most applicable and efficient
way to acquire knowledge in a trustworthy multi-agent environment. We use product
ontology as a typical instance of knowledge to test our design on evolutionary
agents and employ product-brokering agents with ontological process.
A Semantic Web is therefore developed where data is shared under a common framework.
Pages:
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369