Top-K Query Answering for Probabilistic Data Integration Systems in Pervasive Computing Environment

In pervasive computing environment, a challenge is how to exchange and share information on heterogeneous devices. The mediated-base approach of data integration provide a uniform view by creating the semantic relationship between sources and mediated schema. Since the manual match is unfeasible, many automatic approaches are developed. The probability of mappings should be considered inevitably during a query proposed on the global schema of mediated process. Since each tuple of the results has different probability, we may adapt Top-K algorithm to get the most approximate k answers. In this paper, based on a given pervasive computing scenario, we formulate a theoretical model for data integration with probability, demonstrate the probabilities for schema mappings and query answering. We also describe a distributed top-k algorithm cooperating with a local top-k algorithm.