Towards Adaptively Approximated Search in Distributed Architectures

摘要: Innovative applications over distributed architectures, like the Web, often require analysis of strongly related, highly heterogeneous data, stored in remote and autonomous data sources, that can be either totally available at query processing time (stored data) or become a continuous stream (data stream). In these contexts, search efficiency is key issue. However, classical techniques, according to which queries are executed exactly, both for what concerns request technique, set beginning execution, may not ensure adequate performance quality (in terms completeness accuracy) returned result. To overcome such problem, approximate adaptive techniques have been proposed. Adaptive aim ensuring an efficient whenever priori information, needed statically select once most available. Approximation, by contrast, has proposed higher result presence heterogeneity limited knowledge. dynamic environments, two approaches usually considered as orthogonal. we claim exist could benefit from combined approach. An example Web allowing specify on (streams), retrieved through mash-up different sites. Since dynamically acquired, they cannot reconciled, before queries. Moreover, adopting single strategy, fixed priori, penalize system and/or result, only characterizes subsets input data. The this chapter make one step towards integration introducing Approximate Search with Processing (ASAP short) systems. ASAP, decisions concerning when, how, how much taken dynamically, goal optimizing processing.