Top-k dominance range-based uncertain queries

Abstract

Most of the existing efforts for probabilistic skyline queries have used data modeling where the appearance of the object is uncertain while the attribute values of objects are certain. In many real-life applications, the values of an uncertain object can be in a continuous range that a probability density function is employed to describe the distribution of the values. In addition, the “interest-ingness” of the objects as a single criterion for measuring skyline probability may result in missing some desirable data objects. In this paper, we introduce a new operator, namely, the Top-k Dominating Range (TkDR) query, to identify the subset of truly interesting objects by considering objects’ dominance scores. We devise the ranking criterion to formalize the TkDR query and propose three algorithms for processing the TkDR query. Performance evaluations are conducted on both real-life and synthetic datasets to demonstrate the efficiency, effectiveness and scalability of our proposed approach.