Using Statistical Similarity to Identify Corresponding Attributes between Heterogeneous Spatial Databases

One step in integrating heterogeneous database systems is matching equivalent attributes: determine which fields in two databases refer to same real world entity character. In this paper, we present a methodology based on instance statistic similarity for identifying corresponding attributes in heterogeoneous spatial databases which has three basic steps. In first step, the corresponding of the same entities in heterogeneous spatial datasets is built by means of geometrical matching. In second step, attribute similarity, which is expressed by a value between 0 and 1, is computed according to attributes values of the same entity. We design different similarity compute models for different kind of attributes. At last, corresponding attributes are identified by means of comparing attribute similarity value with the predefined threshold. An experimental result is presented which shows the effectiveness and precision of this approach