An approach for the extensional integration of data sources heterogeneous representation formats

In this paper we propose an approach for the extensional integration of data sources with heterogeneous representation formats. The proposed approach is based on the exploitation of a new model, called E-SDR-Network, for representing and handling, at the extensional level, heterogeneous data sources, ranging from databases to XML documents, OEM graphs and other semi-structured data. Due to the specific features of E-SDR-Network, the proposed extensional integration methodology is capable of: (i) easily handling null or unknown values, (ii) producing consistent query answers from possibly inconsistent data and (iii) reconstructing, at the extensional level, the content of each data source involved in the integration task. Finally, we show that E-SDR-Network and the proposed extensional integration algorithm are the counterpart, at the extensional level, of the SDR-Network conceptual model and the associated intensional integration algorithm, already proposed in the literature. Therefore, in the whole, we obtain a complete approach consisting of two components performing synergically both the intensional and the extensional integration of data sources having heterogeneous data representation formats.