Web Materialization Formulation: Modelling Feasible Solutions

摘要

Performance of multi domain web search applications is typically hindered by the availability and accessibility of the web data sources. In this paper we consider web data materialization as a solution. The web data services are modelled via binding schema patterns - access patterns - thereby defining input and output dependencies between the participating data sources. Web materialization is formulated as a set of interdependent blocks, each being a deciding factor in formulating an obtainable materialization. In this work consideration is given to the feasibility of the proposed set of web sources for the given materialization task. The model for analysing the feasible materialization solution in terms of reachability and bound is established. To demonstrate the effectiveness of such a feasibility analysis model, an empirical study is performed on a set of materialization tasks ranging in their schema dependency complexity.