An evolutionary approach to query-sampling for heterogeneous systems

In distributed environment, resource descriptions about information sources (e.g., databases and repositories) are significantly important to determine which information source queries should be sequentially sent to. The goal of this work is to find out semantic relationships between the sources for efficiently integrating the resource descriptions. Thereby, in this paper, we propose an evolutionary approach to maximize the discrimination performance among the sources. We combine a query-sampling method and evolutionary method. These two methods are used for extracting the resource descriptions, and optimally integrating the descriptions, respectively. For evaluating the performance of the proposed query-sampling method, we have set up 53 information sources. We have compared the ranking lists recommended by the proposed method with user feedbacks (i.e., ideal ranks), and also computed the precision and recall of discovered semantic relationships between the sources.