摘要：
Trust plays important roles on effective interaction and cooperation
for multi-agent systems(MAS). This study aims at finding out the
current situation and future trends of computational trustfor multi-agent
systems. Through defining seven common compositional elements for
the computational trust models, the study points out significant
weaknesses in the current design. Finally, the paper figures out
the future research trends through discussion and analysis around
the strengths and weaknesses identified. Also the paper proposes
an idea of using ontology and XML technologies such as RDF that
allow systems to provide both human and machine readable annotations
for trust models.
