TY - CONF T1 - Satisfying Complex Data Needs using Pull-Based Online Monitoring of Volatile Data Sources T2 - Data Engineering, International Conference on Y1 - 2008 A1 - Roitman,Haggai A1 - Gal,Avigdor A1 - Raschid, Louiqa AB - Emerging applications on the Web require better management of volatile data in pull-based environments. In a pull based setting, data may be periodically removed from the server. Data may also become obsolete, no longer serving client needs. In both cases, we consider such data to be volatile. To model such constraints on data usability, and support complex user needs we define profiles to specify which data sources are to be monitored and when. Using a novel abstraction of execution intervals we model complex profiles that access simultaneously several servers to gain from the used data. Given some budgetary constraints (e.g., bandwidth), the paper formalizes the problem of maximizing completeness. JA - Data Engineering, International Conference on PB - IEEE Computer Society CY - Los Alamitos, CA, USA SN - 978-1-4244-1836-7 M3 - http://doi.ieeecomputersociety.org/10.1109/ICDE.2008.4497591 ER -