Enhancing DSpace Digital Library with Topic Maps

Abstract

As information becomes more rich yet more complex, alternative ways of organizing the information are required to retrieve that information more effectively and efficiently from the specific source such as Web, local PC or a remote server. However, the problem faced by most digital libraries today is that there is a huge amount of information that has become difficult to be accessed speedily and accurately. Moreover, the actual strength of any digital repository for improving the education can only be achieved when users start using it vigorously and persistently. However, there is a serious lack of research work in this area. We need to manage the enormous information flow, data and knowledge, by improving information retrieval. For this purpose, a tool is required that caters for these issues besides navigating through the information. The Topic Map (TM) is one such concept used for organizing information and knowledge representation. Though various attempts have been made in the past and a number of solutions exist but most of them do not make use of Topic Map model at its length. Therefore, the objective of this study is to develop a Topic Maps-based ontology approach for DSpace Digital Repository software being used at NUST, to augment user’s ability to locate a particular thesis / project report effectively and efficiently, also be able to compare the two different topics without having to send a complex query to the server. This study also tends to explore how that approach affects users' ability to search and his searching preferences. The Topic Maps [ISO 13250] is an international standard for representing knowledge and the information stored as database schemas. While databases only captures relations between information objects, Topic Maps also allow these objects to connect to various places where they can occur. Since they are real-world representation of Topics, Topic Maps provide more flexibility, interoperability, and scalability to the system where they are incorporated. Our implementation of the Topic Maps in the DSpace incorporates features, such as, quick and easy information retrieval, comparison between various thesis reports and their metadata, graphical display of searched results, clickable nodes for viewing the thesis documents / reports etc. In order to evaluate our system, we conducted a survey and evaluated that the mean scores were consistently higher in preference to using topic maps based approach rather than a conventional tabular approach for information retrieval.