Smartphone is equipped with a variety of sensors as well as a communication device. These devices can monitor their target and environments more accurately. Previous closed system prevented valuable sensing data from sharing with other systems and applications. It is also difficult to share sensing data without the infrastructure for large-scale data processing and metadata management. In this paper, we proposed a hierarchical back-end architecture for smartphone sensing. The hierarchical structure was designed for supporting the large scale and distributed data processing. We also designed the metadata management scheme based on RDF/S framework for sharing sensing data with their meaning. The evaluation results showed that our architecture is more efficient to process the large-scale metadata.