Data integration is a challenging task due to the large numbers of autonomous data sources. This necessitates the development of techniques to reason about the benefits and costs of acquiring and integrating data. Recently the problem of source selection (i.e., identifying the subset of sources that maximizes the profit from integration) was introduced as a preprocessing step before the actual integration. The problem was studied for static sources and used the accuracy of data fusion to quantify the integration profit.

In this paper, we study the problem of source selection considering dynamic data sources whose content changes over time. We define a set of time-dependent metrics, including coverage, freshness and accuracy, to characterize the quality of integrated data. We show how statistical models for the evolution of sources can be used to estimate these metrics. While source selection is NP-complete, we show that for a large class of practical cases, near-optimal solutions can be found, propose an algorithmic framework with theoretical guarantees for our problem and show its effectiveness with an extensive experimental evaluation on both real-world and synthetic data.