Abstract
A serious problem on the WWW is finding reliable information. Statements that are published on the Semantic Web should be seen as claims rather than as facts, and is a need for a mechanism to decide which among many sources is most reliable. Trust management refers to mechanisms for verification that is the source of information really who the source claims to be and whether the information provided by source can be trusted or not? Trust management is currently handled by two different approaches: a "strong and crisp" approach called policy based trust management, where decisions are founded on logical rules and digital credentials, and a "soft and social" approach called reputation based approach, based on reputation gathered and shared in a distributed environment. We analyze the differences between the two models of trust and give an integrated model that would improve significantly trust management systems. We support our claim with Buyer Seller scenario and illustrate how the two approaches are integrated.