Internet applications have evolved from centralized computing platforms to distributed collaborative network computing systems. Recent work suggests that reputation-based trust systems are an effective way for collaborative computing systems and applications to identify and avoid malicious nodes, minimize the threat, and protect the system from possible misuses and abuses in decentralized network computing systems and applications. With the increasing importance of reputation-based trust systems for Internet-computing applications, we’re faced with trust management issues from several perspectives. First, we must establish attack-resilient reputation trust among nodes that don’t have prior knowledge about one another. Second, we must utilize the historical data in reputation management to deal with sparse and malicious manipulation of reputation trust. Finally, we must exploit reputation-based trust inference for building more reliable large-scale Internet applications.