D-Lib Magazine

May/June 2015
Volume 21, Number 5/6
Table of Contents

An Assessment of Institutional Repositories in the Arab World

Scott Carlson
Rice University
scarlson@rice.edu
orcid.org/0000-0002-6764-1238

DOI: 10.1045/may2015-carlson

Printer-friendly Version

Abstract

Compared to those of the Western world, institutional repositories from the Arabic-speaking countries of the Middle East have been described by scholars of the region as occupying an "infancy stage". In this article, repositories from countries in the Arab world were selected and assessed in terms of accessibility and transparency from the viewpoint of an external user. A set of assessment criteria was formed by analyzing trends and similarities in established repositories from the rest of the world, in hopes of analyzing the "infancy stage" appraisal. The results provide not only a current view of digital scholarship and institutional memory in the Middle East, but may also provide a helpful set of criteria for developing repositories for the rest of the world.

Keywords: Open Repositories, Institutional Repositories, Middle East, Repository Assessment, Open Access, Arab World, Arabian Gulf

1 Introduction and Background

Since the release of EPrints, DSpace, and other software packages in the early 2000s, the history and development of the digital repository in Western archives and libraries have been well documented. This is less true for other regions of the world, especially for Arabic-speaking countries and territories. A 2012 article by Syed Sajjad Ahmed and Saleh Al-Baridi of the King Fahd University of Petroleum and Minerals Library in Dhahran, Saudi Arabia, as well as a 2010 conference presentation by Mohamed Boufarss of the Petroleum Institute in Abu Dhabi, note several factors that have contributed to institutional repositories in the Arabian Gulf region remaining at an "infancy stage." These factors include a lack of published literature on the topic originating from the region [1], the majority of academics in the Arab world having little knowledge of or experience with institutional repositories [2], and the lack of a coherent long-term repository preservation strategy among regional institutions [3]. In fact, Ahmed and Al-Baridi noted that much of their research on the state of repositories was gathered from colleagues and acquaintances from other institutions in the region.

I was no stranger to this "infancy stage." My interest in institutional repositories in the Arab world began in late 2013 in a project that combined professional interest with class work for a certificate in Digital Stewardship. Between 2012 and 2014, I worked in the technical services department of the library of the American University of Sharjah (AUS), an accredited, multicultural institution in the United Arab Emirates. Not long after joining AUS, my role in working with the library's institutional repository (formally launched in March, 2012, the same month I joined) increased, as did faculty interest in the repository. In 2013, a series of working papers created by faculty from the university's School of Business Administration was implemented; later that year, a test community was set up to deposit student work produced by the university's English for Engineering class.

Despite this interest, feedback from the faculty community indicated that the school's repository was difficult to locate online, as was information about the repository and its policies on the library's web site. Thus, a plan was initiated to visit institutional repositories in the Arab world, as if they were being visited by anonymous external users, and compare trends or similarities found to see if this "infancy stage" appraisal was apparent through repository assessment. (This comparison plan was inspired by a 2008 Association of Research Libraries Conference presentation, wherein Robert H. McDonald and Charles Thomas argued that institutional repositories do not exist as "stand-alone phenomena." [4])

Originally, the selection of repositories was to begin by visiting the Directory of Open Access Repositories (OpenDOAR); however, both the Ahmed/Al-Baridi and Boufarss pieces pointed out that OpenDOAR lacks information on a significant number of existing repositories in the region. Currently, OpenDOAR lists only 21 registered repositories between Egypt, Iraq, Saudi Arabia, Lebanon, the Sudan and Qatar. Meanwhile, the Ranking Web of Repositories —an initiative of the Consejo Superior de Investigaciones Científicas (CSIC), a public research body in Spain — lists only 19 repositories in their "Arab World" category, with some overlap to OpenDOAR. All together, OpenDOAR, the Ranking Web, and the Registry of Open Access Repositories (ROAR) contained information on the repositories of 29 individual institutions in 10 countries of the Arab world. Ultimately, this pool of candidate repositories was supplemented by an informal survey of the major universities and institutions within the Arabian Gulf region, which was found to be under-represented in the 29 registered and ranked repositories. The final pool of candidates grew to 36 repositories in 11 countries, from which 25 were selected for assessment, representing each one of the 11 countries from the pool.

2 Assessment Criteria

The foundations of the assessment criteria for this investigation were found in the metrics developed by the Center for Research Libraries (CRL), which tend to focus on the transparency of repositories. The CRL's Ten Principles, developed along with the U.K.'s Digital Curation Center, DigitalPreservationEurope, and Germany's NESTOR in January of 2007, spell out a number of expectations for digital preservation repositories, including policy frameworks, stated ingestion criteria, and requisite metadata on digital objects, before and during preservation [5]. These principles would be further refined in the Trustworthy Repositories Audit & Certification: Criteria and Checklist (TRAC), developed and used by the Center for Research Libraries in auditing and certification of digital repositories, specifically sections A2 (Organizational structure & staffing), A3 (Procedural accountability & policy framework), B1 (Ingest), B3 and B4 (Preservation planning), and B5 (Information management) [6].

Additionally, the latest edition of Institutional Digital Repository Benchmarks was consulted for the web visibility metrics contained in its Marketing section. [7]. Specifically, the Benchmarks survey focused on whether repository web sites provide URL links back to either the institution's library website or main website. The survey questions, however, effectively discount the ways that users might arrive at the repository. The arguable assumption is that libraries tend to either be given or willfully take on the responsibility of managing the institutional repository, and are thus often shuttled to repositories via the institution's library web site. This, too, would become an assessment metric.

However, because the assessment would also focus on the accessibility of repositories, the assessment would take place much in the same way a critic would visit a restaurant (i.e., without announcement and without interviews). Thus, it seemed appropriate to conduct more informal surveying — this time, repositories outside of the Arab world — for accessibility trends. Repositories that were already long-established (in relative terms, as the digital archiving community is itself still a blooming field) could provide insight in assessing the comparatively "younger" pool of the Arab world; 18 baseline repositories were chosen, each founded sometime between 2002 and 2007 and representing institutions of the United States, Canada, the United Kingdom, Australia, China, and the Netherlands.

The full assessment criteria were structured in order to answer the following questions:

1. What is the name of the repository's home institution, country of origin, and software platform?
2. Is the repository registered with OpenDOAR or ROAR? If so, what is the link?
3. Does the repository have a dedicated "landing page" with associated information?
4. Does the repository have a connecting URL link from the institution's library web page?
5. Does the repository or the library site offer contact information for the repository staff or administrators? If so, who is listed and how many?
6. Is there a stated submission policy for ingested objects?
7. Is there a stated preservation plan?
8. Is there a stated collection policy?
9. Is there a stated metadata policy?
10. Are there stated standards on what file types can be deposited?
11. Is self-deposit allowed for the repository user community?
12. Is the repository subject to some kind of institutional mandate? Is there a mandate registered with the Registry of Open Access Repositories Mandatory Archiving Policies (ROARMAP)?
13. Does the repository or library offer Open Access information? Is there an Open Access policy?

3 Assessment Results

The full results of the investigation of Arab world repositories were recorded in a Google Sheets document which is freely available for reference. The data collected from the baseline repositories, which informed the assessment criteria, is also freely available for reference.

3.1 Web Presence and Accessibility

As mentioned previously, one of the main focuses of this investigation was the web presence of each sampled repository and the process of reaching each of them, starting at the institution's library website. All of the baseline repositories were reachable from somewhere within the library's web site, mostly from the front page, or through minimal searching. Roughly half (10, or 55 percent) opted to utilize IR landing pages in their library website — pages where information about the repository "lives" on the site, as opposed to storing relevant information on the repository itself.

Of the 25 repositories from the Arab world, 15 (60 percent) were directly linked to from the university library's web page; only three of those (12 percent of the overall group) elected to utilize landing pages in their library website. A further four repositories were searchable via a portal on the library web page; this left six repositories which either were not linked to directly from the library site — instead, linked from somewhere else on the institutional web site — or in some cases, appeared not to be linked anywhere at all.

Also investigated was whether the repository (or its associated institutional web presence) provided concrete contact information for those responsible for the repository's administration or oversight. Section A3 of TRAC (Procedural accountability & policy framework) recommends that a trustworthy repository "ensure[s] that feedback from producers and users is sought and addressed over time" and "commits to transparency and accountability in all actions supporting the operation and management of the repository [8]; again, all of the baseline repositories included at least one specific point of contact. However, less than half of the sampled repositories (11, or 44 percent) provided a contact email for the responsible party (or parties). This number is slightly inflated to 18 (72 percent) if non-email feedback forms are included in the number; however, this still leaves five repositories that included no form of communication at all.

More troubling, though, was the connectivity issues affecting a handful of the sampled repositories. Over the course of research in 2014, KhartoumSpace — the repository of the University of Khartoum in the Sudan — seemed to suffer many server errors, resulting in unpredictable access to the repository and its contents. The repository of Khalifa University in Abu Dhabi has appeared to suffer from seemingly permanent connectivity errors for months, rendering the archive unavailable since at least the beginning of 2014; meanwhile, the URL for the British University in Egypt merely points to a welcome page for Internet Information Services (IIS) 7, a Microsoft Windows web server. The total inaccessibility of the latter two ultimately dropped the total of the sample pool down to 23 repositories.

3.2 Transparency of Policies

As noted in the assessment statements above, the repositories in this study and their institutions were analyzed for formal, publicly-available policies and procedures, specifically covering the submission of ingested objects, preservation, detailed scope of collection, metadata, recommended file types of objects [9], and the explicit allowance of users to self-deposit their work. An individual analysis of the specifics of each policy, however, was not undertaken; rather, the focus was placed on whether or not these policies exist via explicit references or availability on the library/repository websites.

Table 1 clearly demonstrates the low frequency of such policies and procedures in the Arab world repositories; only Egypt's American University in Cairo, Saudi Arabia's King AbdulAziz University and King Abdullah University, and Tunisia's Université Virtuelle de Tunis made their policies available, and even then, not all of the policies that were investigated. Only the King Abdullah University of Science and Technology in Saudi Arabia offered an application profile, or the available fields of metadata, for their digital archive; that being said, their FAQ page maintains that "the only compulsory field [for submitted objects] is title." Similarly, only the American University in Cairo offered recommendations on file types of submitted objects that moved beyond standard boilerplate repository language. None of the repositories offered written policy on the methods of preservation of their materials, apart from general statements detailing their overall goals. (It should be noted that these repositories are not assumed to be lacking in formal policies on their day-to-day operation and long-term goals; this is simply a report of the public availability of those institutional policies.)

Table 1: General Policies and Procedures of the Assessed Arab World Repositories

These results were compared to the availability of policies from the baseline repositories. Seventy-eight percent of the baseline group (14 repositories, which vary by actual policy) included collection and document submission policies, as well as self-deposit, while just over half (10 repositories, or 55 percent) included standards for file formats. Far less common in the baselines were explicit policies on metadata (6, or one-third) and preservation (4, or 22 percent); in the case of preservation, a further three repositories explicitly mentioned techniques such as migration from obsolete formats, maintaining regular backups, and checking bit integrity, but without any substantial policy or planning information accompanying these tidbits, they were judged to be no more helpful or informative than boilerplate statements about best practices.

3.3 Open Access and Mandate Information

Open Access information was also collected; repository, library and institution websites were directly investigated for both general information on the concept of Open Access and institution-specific policies, including any references to mandated deposits. The Registry of Open Access Repositories Mandatory Archiving Policies (ROARMAP) was then checked for any specific institutional entries found in the repository pool and cross-referenced with the collected data. While all of the baseline repositories included some basic information on open access (via either the repository, library or institutional website), over half (11 repositories, or 61 percent) carried an explicit open access policy and/or registered some kind of mandate with ROARMAP. (Interestingly, having one did not necessarily guarantee having the other.)

Table 2 shows the collected data for the Arab world repositories. Only one (King Abdullah University of Science and Technology in Saudi Arabia) explicitly references an institutional Open Access policy, along with a registered entry at ROARMAP. One might be tempted to conclude that similar mandates otherwise do not exist within the other repositories; the problem is that at least four of the other repositories — Bibliothèque Virtuelle de l'université d'Alger, the American University in Cairo, the Lebanese American University, and the American University of Sharjah — employ language, on either the library websites or the repositories themselves, that implies the existence of thesis mandates for graduating students. In the case of AUS, direct experience proves this to be the case, but for the other three, without clearly defined policies, we are left to read between the lines with statements such as "All theses published after 2003 are Open Access" [10], and "You will not receive your graduation invitations until you submit an electronic final approved version of your thesis to the repository" [11].

Table 2: Open Access Policies and Mandates of the Assessed Arab World Repositories

Ahmed and Al-Baridi identify an overall lack of Open Access discourse in the Arabian Gulf region [12], which may offer a possible explanation for the lack of Open Access and mandate policy information in the surveyed repositories. However, a healthy dose of skepticism may be in order for this particular assessment; the freely available metadata of the contents of the Directory of Open Access Journals lists more than 550 individual open access journals founded in countries in the Arab world (though almost 500 of them originate from Egypt) [13].

4 Conclusions and Further Applications

The impetus for this research was to assess the progress of the modern institutional repository of the Arab world, and to assess its appraisal as relatively "youthful" in comparison to repositories of the rest of the world. Many of the repositories in the Arab world have taken the initiative to build their own unique collections, make their presence felt within the community, and begin building a case for their necessity. However, these comparisons seem to bear witness to Ahmed and Al-Baridi's 2012 appraisal of a community in relative infancy. Obviously, not all procedural documentation requires full disclosure, but a sizable portion of the sampled repositories seem to occupy a plane of existence where material is deposited for safekeeping and public circulation, but without any public acknowledgment or information on how those goals will be met. And while the focus of investigation rested on the repositories from the Arab World, the sampled baseline repositories were notably lacking in explicit policies on file formats (55 percent), metadata (33 percent) and preservation (22 percent), revealing that even these repositories have areas in need of development. To once again refer to TRAC, "only a repository that exposes its design, specifications, practices, policies, and procedures for risk analysis can be trusted" [14].

A further indication of the need for serious commitment to the development and management of institutional repositories in the Arab world is that in the short time between when this article was written and its publication in D-Lib Magazine, the accessibility of a number of the assessed repositories has fluctuated greatly: Universite Kasdi Merbah Ouargla, Thi Qar University, Lebanese American University, Université Virtuelle de Tunis, and the American University of Sharjah.

The path to developing a secure, trustworthy repository — whether assessed through TRAC, DRAMBORA, or another auditing method — often requires a serious commitment of time and resources. Hopefully, the assessment criteria developed for this research project can be of use to repositories while preparing certification proper, especially relatively young repositories.

References & Notes

[1] Syed Sajjad Ahmed and Saleh Al-Baridi, "An overview of institutional repository developments in the Arabian Gulf Region." OCLC Systems & Services: 80. http://doi.org/10.1108/10650751211236613

[2] Mohamed Boufarss, "If we build it, will they come? A survey of attitudes toward institutional repositories among faculty at the Petroleum Institute." Special Libraries Association-Arabian Gulf Chapter 16th Annual Conference. http://www.ceser.in/ceserp/index.php/ijls/article/view/2915

[3] Ahmed and Al-Baridi, 87.

[4] Robert H. McDonald and Charles Thomas, "Cross-institutional Repository Assessment: a Standardized Model for Institutional Research Assessment." ARL Assessment Conference. Association of Research Libraries, Seattle, 4 August 2008.

[5] Center for Research Libraries, Digital Curation Centre, Digital Preservation Europe, and Competence Network for Digital Preservation, "Ten Principles." 2007.

[6] Center for Research Libraries, Trustworthy Repositories Audit & Certification: Criteria and Checklist. Version 1.0. Center for Research Libraries, Chicago, 2007.

[7] Institutional Digital Repository Benchmarks. 2014 ed. New York: Primary Research Group, Inc., 2014: 36-52. Print.

[8] Trustworthy Repositories Audit & Certification: 14.

[9] The DSpace platform includes a boilerplate list of supported file formats in its help guide. This study differentiated this standard list from a policy provided by the institution that specifies preferred or disallowed file formats available for submission to the repository.

[10] Available here.

[11] Available here.

[12] Ahmed and Al-Baridi: 80.

[13] Metadata can be acquired here.

[14] Trustworthy Repositories Audit & Certification: 6.

About the Author

Scott Carlson is the Metadata Coordinator at Rice University's Fondren Library. Between March 2012 and August 2014, he was the Cataloging and Metadata Librarian at the American University of Sharjah, an accredited, multicultural institution in the United Arab Emirates. He received his MLIS from Dominican University in River Forest, Illinois, and recently completed an Archives Certificate in Digital Stewardship from Simmons College.