Abstract | In this paper, we present and study solutions for the efficient processing of queries over string attributes in a large P2P data network implemented with DHTs. The proposed solutions support queries with equality, prefix, suffix, and containment predicates over string attributes. Currently, no known solution to this problem exists.
We propose and study algorithms for processing such queries and their optimizations. As event-based, Publish/Subscribe information systems are a champion application class where string attribute (continuous) queries are very common, we pay particular attention to this type of data networks, formulating our solution in terms of this environment. A major design decision behind the proposed solution is our intention to provide a solution that is general (DHT-independent), capable of being implemented on top of any particular DHT. |