ABSTRACT
We protect provenance with policies. These policies determine who can access a document and under what conditions access is to be granted, and when access is allowed or disallowed. We need to develop policies that scale with the large provenance graphs and also take into account any implicit relationships among various components in the provenance graph. In other words, we need to build large data stores for provenance. Some major challenges are how to successfully manage the huge datasets involved and how to effectively query large provenance graphs efficiently while still maintaining system responsiveness and enforcing security constraints. We believe that cloud computing offers a viable solution to this problem.
