Relaxed Queries over Data Streams

Relaxation skyline queries have been proposed, in the relational context, as a solution to the so-called empty answer problem. Given a query composed of selection and join operations, a relaxation skyline query relies on the usage of a relaxation function (usually, a numeric function) to quantify the distance of each tuple (pair of tuples in case of join) from the specified conditions and uses a skyline-based semantics to compute the answer. Though the empty answer problem is extremely relevant also in a streaming context, where users may not be acquainted with the actual data arriving on the stream, it has been largely neglected. Specifically, no solutions have been proposed so far for skyline-based relaxation over data streams. In this paper, we define relaxation skyline queries for window-based join over data streams, propose one processing algorithm and present a preliminary experimental evaluation of the designed technique.