DATA INTEGRATION MODEL FOR FASTER DATA EXTRACTION AND RETRIEVAL FROM SEMI-STRUCTURED DATA FORMAT

Abstract

Collections of data is crucial across a wide variety of field because of increasing data rapidly year by year. These collections are important for many organizations to make a correct decision using business intelligent applications. A business intelligent application must have capability to collect and integrate all data from different data sources. One of the challenges in development of business intelligent application is data integration. This challenge is happened because of data structure are different. This research is looking for suitable data integration model in order to allows data integration from different data sources. Native XML (NXD) is one the model has been used in data integration. In this model, elements and attributes for each data are extract and store into Relational Database Management System (RDBMS). Meanwhile, the value for each element and attribute are stored in XML format. Based on the experiments has been done by previous researchers, NXD can produce a better performance during data insertion response time and query processing response time using SigmodRecord and DBLP datasets. However, the efficiency of NXD still has room for improvement. In the initial experiment, list of special character can be removed to improve the performance of NXD has been idenfitied in the SigmodRecord and DBLP datasets.