Advances in proteomics technologies have enabled novel protein interactions detected at high speed, but they come at the expense of relatively low quality. Therefore, a crucial step in utilizing the high-throughput protein interaction data is evaluating their confidence and then separating the subsets of reliable interactions from the background noise for further analyses. Using Bayesian network approaches, we combine multiple heterogeneous biological evidences, including model organism protein-protein interaction, interaction domain, functional annotation, gene expression, genome context, and network topology structure, to assign reliability to the human protein-protein interactions identified by high-throughput experiments. This method shows high sensitivity and specificity to predict true interactions from the human high-throughput protein-protein interaction datasets. This method has been developed into an online confidence scoring system specifically for the human high-throughput protein-protein interactions. Users may submit their protein-protein interaction data online, and will be returned the detailed information about the supporting evidence for query interactions together with the confidence scores. The web interface of PRINCESS is available at http://www.hupo.org.cn/princess/.