Architecture for Accessing Heterogeneous Databases

Abstract

This paper presents the architecture for accessing heterogeneous databases. Two major processes in this architecture which are extracting SQL statement and ontology. The algorithms for extracting SQL statement was created and tested in order to improve time performance during searching and retrieving process. Ontology approach was implemented and combined with these algorithms. In ontology approach, web semantic was implemented in order to retrieve only relevant data from database. A prototype based on this architecture was developed using JAVA technology. JAVA technology was chosen because this technology have Jena library. This library is provide API and support SPARQL. Several experiments have been executed and tested. The result indicates this architecture able to improve web query processing in term of time. The result also indicates this architecture able to retrieve and displayed more relevant data to web users. (http://www.malaysia.gov.my) to access information about education, carrier, tourism, etc. The web users can surffing internet using laptop, notebook, PDA, etc anywhere and anytime to access the information. These information can help web users to make any decision in research, education, etc. Most of web applications stored data into a single database. A single database stored lot of data. Increasing number of data will reduce the space size of storage database. In this case, performance of web query processing become slow to search and retreive lot of data from a single database. In web based applications, three main component invloved; interaction among web user, web application and web server. User interface is provided to web user to enter information (a query) such as education, toursim, etc. The user interface also will display the information/results to web user. Web application server will get a query from web user. This query will manipulate before searching and retrieving process. Then, the web application server will communicate with database server for searching and retrieving process. After data is found, the information/results will display to web user through web user interface. Figure 1 shows the interaction among web user, web application server and database server.