Prototyping Digital Libraries Handling Heterogeneous Data Sources – An ETANA-DL Case Study

Information systems used in archaeological research have several needs that can be summarized as follows: interoperability among diverse, heterogeneous systems, making information available without significant delay, providing a sustainable approach to longterm preservation of data, and providing a suite of services to users of the system. In this thesis, we describe how digital library techniques can be employed to provide solutions to these problems and describe our experiences in creating a prototype for ETANA-DL. ETANA-DL is a model-based, componentized, extensible, archaeological Digital Library that manages complex information sources using the client-server paradigm of the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH). We have designed and developed the prototype system with the following main goals: 1) to achieve information sharing between different heterogeneous archaeological systems, 2) to make primary archaeological data rapidly available to users, 3) to provide useful services to users of the DL, 4) to elicit requirements that users of the system will have beyond the services that it supports, and 5) to provide a sustainable solution to long-term preservation of valuable archaeological data. Consequently, we describe our approach to handling heterogeneous archaeological information from disparate sources; suggest an architecture for ETANADL, to be validated through prototyping; and show that given a pool of components that implement common DL services, a prototype DL can be rapidly created that supports several useful services over integrated data. Further, and most fundamentally, we note that understanding complex information systems is a difficult task. Finally, therefore, we describe our efforts to model complex archaeological information systems using the 5S framework, and show how we have used the resulting partial models to implement ETANA-DL with cross-collection searching and browsing capabilities. This work is supported by the National Science Foundation’s Information Technology Research program (ITR-0325579). Any opinions, findings, and conclusions or recommendations expressed in this thesis are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.