Building Digital Collections Using Open Source Digital Repository Software: A Comparative Study

Introduction

The last decade a great number of Digital Library (DL) and Digital Repository (DR) systems have been developed and published as open-source software. The variety of available software systems becomes a headache when an organization plans to build a repository infrastructure to host its collections. Fortunately, there are many articles and surveys that evaluate or compare open-source DR and DL software. One of the first guides for selecting open-source repository software, based on the features and benefits of 9 different repositories, is provided by the Open Society Institute (2004). An extensive checklist for evaluating DL software is drafted by Goh et al. (2006). Also, two recent papers compare and evaluate some current open-source DR and DL software (Masrek & Hakimjavadi, 2012; Tramboo et al., 2012). The main scope of these papers is the comparison of the software systems based on some quantitive and quality characteristics, in order for interested organizations to select the proper system for their digital collections. In our study we try to go a step further and in addition to the comparison of DR software systems and their characteristics' description, we propose the most suitable systems for different collection types. Using five collection paradigms that represent case studies of different content and functionality, an organization can be directed to select a repository software matching its criteria.

We used the following three restrictions in order to select the repository software systems that participate in the comparative study. The repository systems:

• 1.

  are publicly available using an open-source license,

• 2.

  are compliant with the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) (Lagoze & Sompel, 2001),

• 3.

  have a large number of installations worldwide.

Using these restrictions we selected the five (5) widely used repository systems shown in Table 1. These systems are between the 10 most used open-source software participating in the Directory of Open Access Repositories – OpenDOAR (n.d.). Each of these systems has been thoroughly studied based on the core repository characteristics and supported features. We used the latest software versions which are (at August 2013): DSpace 3.2, Fedora 3.6, Greenstone 3, EPrints 3.3 and Invenio 1.1. In the following section, the characteristics needed by a current repository software are listed and described. In the third section, the five repository systems are compared based on each of the characteristics and the results are summarized in a score table. In the fourth section, one or two repository systems are proposed for the hosting of each of five different collection types.