Skyline and Top-k Processing in Web Bargaining

Skyline and top-k queries are gaining increasing importance in many emerging applications. Current skyline and top-k query processing techniques work on deterministic object attributes and known scores. However, in many practical scenarios these settings are inapplicable. In this paper we focus on web interaction scenarios where each interaction is a data object with a set of possible outcomes (scores). Obtaining exact interaction scores is expensive as it involves complex arbitration between interacting parties. Moreover, the outcome of each interaction might redefine other interactions scores. We demonstrate that the search space for solving such problems is very large. Based on this we formulate and present skyline and top-k processing algorithms that can efficiently reduce the search space. We present the results of a thorough experimental evaluation quantifying the relative performance of the algorithms we propose herein with respect to costly exact solutions. Our results indicate that our techniques can efficiently reduce the space and identify precise solutions.