Proof for Multi-agent Communication of Semantic Web Information

In semantic Web, an application should not just giving a correct answer but also providing a proof or some kinds of explanations for that answer. A proof contains information related to the process that creates an answer like knowledge sources, reasoning systems, etc. Provided a proof or explanations, a user can trust and use a result with confidence. Such a means of answer provenance is even more important for distributed systems where an answer may be derived from many knowledge sources and by several systems. In this paper, basing on our existing logical framework for multi-agent communication of semantic Web information, we extend the framework so that the agents are capable of tracing and providing proofs of their answers. A proof contains information about sources of knowledge, agents participated in the reasoning process as well as facts, rules, and axioms that are employed for inferring an answer. Such proofs are essential for various tasks like checking and validating answers, computing trust information, or generating explanations of answers.