DataStorm An Ontology-Driven Framework for Cloud-Based Data Analytic Systems

Cloud-based systems have proven to be a powerful technology for building data-intensive applications. However, the process of designing and deploying such applications is still primarily a manual one. There is a need for mechanisms and tools to help automate the required development steps. Using the Semantic Web ontology language OWL and the Hadoop platform we have developed a number of models and associated software tools that provide an end-to-end solution for designing and deploying cloud-based systems. This solution supports the construction of detailed models of data dependencies and their validation. It also enables generation and deployment of cloud-based data flows from those models. We illustrate its use for detecting alarm scenarios using data from vast underwater sensor-network.