Assured information sharing (AIS) using the cloud

Assured Information Sharing (AIS) refers to the ability of multiple organizations to securely share information. Given the precarious times we live in today, a lack of information sharing, particularly between government agencies, could lead to incidents that cause irreparable damage to life and property. A neglected area in the context of AIS is technological infrastructure, which refers to the software ecosystem that allows organizations to efficiently, economically and securely share information. Although researchers have developed policy-based information sharing systems, none of these are cloud-based; therefore, they lack the scalability/efficiency needed for supporting a large user-base that utilizes vast quantities of data. In this dissertation, we try to remedy this situation by developing AIS implementations that operate under varying conditions.;The first part of this dissertation presents details of two prototypes that provide data sharing and analysis capabilities for a significant number of users. The first prototype, CAISS-X, uses a cloud framework for managing relational data and a non-cloud policy engine to enforce XACML policies. Although CAISS-X represents a discernible enhancement over prior AIS implementations, it still suffers from constraints related with usage of relational model and XACML-based policies. The second prototype, CAISS, overcomes these limitations by using a cloud framework for data storage/retrieval and policy enforcement.;The next part of this dissertation presents Hybridizer, a framework that allows organizations to automatically partition their data/processing tasks over hybrid clouds, while taking into account performance, security and financial requirements. Hybridizer allows organizations to fully exploit the benefits of hybrid clouds towards developing AIS solutions, while achieving the right mix of performance, security and financial costs.;The final part of this dissertation describes StormRider, a system that allows organizations to securely manage large-scale, evolving networks in real-time. StormRider stores/queries network data using RDF/SPARQL due to their expressivity and ability to capture evolving domain requirements. Since there did not exist a comprehensive, distributed RDF storage framework, the last part also describes Jena-HBase, a framework that uses existing cloud technologies to construct a distributed, scalable and efficient RDF storage framework.