In this paper we present a system architecture for trust management inspired by a human model for the evaluation of trust. We focus on large-scale marketplaces for diverse software resources and services. In such marketplaces, user agents (consumer, provider and broker) are largely autonomous with vastly diverse requirements, capabilities, and trust preferences. We hypothesize the need for personalized trust management system for the robustness and wide-scale adoption of such large-scale heterogeneous marketplaces. Our proposed system architecture for personalized trust management comprises five layers, namely Monitoring, Data, Analytics, Expectation, and Decision management. Additionally, it utilizes a trust model based on the four aspects of intent, integrity, capability and results. Our simulation results demonstrate the effectiveness, efficiency and resilience of our proposed system.