Crowdsourcing is a new distributed computing paradigm that leverages the wisdom of crowd and the voluntary human effort to solve problems or collect data. In this paradigm of soliciting user contributions, the trustworthiness of contributions becomes a matter of crucial importance to the viability of crowdsourcing. Prior mechanisms either do not consider the trustworthiness of contributions or assess the quality of contributions only after the event, resulting in irreversible effort exertion and distorted player utilities. In this paper, we propose a reputation system to not only assess but also predict the trustworthiness of user contributions. In particular, we explore an inter-worker relationship called endorsement to improve trustworthiness prediction using machine learning methods, while taking into account the heterogeneity of both workers and tasks.