Integration in Digital Libraries : Approaches a nd Challenges Ismail

Data integration is a central issue in digital libr aries to facilitate the access and manipulation of the highl y distributed, heterogeneous, and dynamic collection of information sources. In this paper, we survey the d ata integration approaches that meet the digital librar y requirements and identify the specific issues that contribute to the complexity of data integration in digital li braries. Information is power and digital libraries are buil t to provide a unified infrastructure for supporting the creation of information sources, facilitating the movement o f information across global networks and allowing eff ective and efficient interaction among knowledge producers , librarians, and information seekers [2]. A digital library is a vast collection of entities stored and maintained by multiple information sources includin g databases, image banks, file systems, email systems , the Web, and applications providing structured or semi-structured data. The dramatic growth of digital libraries in recent years has not only simplified the access to existing info rmation sources but also initiated the creation of numerous new sources. Paradoxically, this growth has made the ta sk of finding, extracting and aggregating relevant inform ation not easier. This is because most of the information systems underlying digital libraries are physically distributed, heterogeneous in the way how information is stored, organized and managed, and comprise heterogeneous software and hardware platforms on which they resid e. Additionally, they are autonomous in the sense that the content and format of data are determined by the organization owning the data not by the user [13]. Furthermore, the data in these information sources is mainlyofcompositemultimedianaturecomprising different media types such as text, video, images, and audio, and dynamic in the sense that the sources up date both the content and the form of the data on their own, possibly on unknown schedule. The goal of data integration system is to provide u sers with a uniform interface to access, relate, and com bine data stored in multiple, autonomous, and possibly hetero geneous information sources. The most important advantage of a data integration system is that it enables users to focus on specify ing what they want, rather than thinking about how to obtain the answers. Consequently, this frees users from the te dious tasksoffindingtherelevantinformationsources, interacting with each source in isolation using a p articular interface and combining data from multiple sources. Dataintegrationrequiresthatthedifferencesin modeling, semantics and capabilities of the sources together with the possible …