Institutional Repository Deposit: SWORD v2.0: Deposit Lifecycle

JISC has released SWORD v2.0: Deposit Lifecycle.

Here's an excerpt:

SWORD is a hugely successful JISC project which has kindled repository interoperability and built a community around the software and the problem space. It explicitly deals only with creating new repository resources by package deposit a simple case which is at the root of its success but also its key limitation. This next version of SWORD will push the standard towards supporting full repository deposit lifecycles by using update, retrieve and delete extensions to the specification. This will enable the repository to be integrated into a broader range of systems in the scholarly environment, by supporting an increased range of behaviours and use cases.

"Asking for Permission: A Survey of Copyright Workflows for Institutional Repositories"

Ann Hanlon and Marisa Ramirez have self-archived their presentation "Asking for Permission: A Survey of Copyright Workflows for Institutional Repositories" in DigitalCommons@CalPoly.

Here's an excerpt:

Most survey respondents reported providing mediated deposit (material is deposited on behalf of the author by a third party, usually someone associated with the IR), whether it is completely mediated by the library or whether the author, in partnership with the library, deposits their work. The only respondents to report author self-deposit as the primary method of IR deposit were in Australia and Europe.

DuraSpace Announces Registered Service Provider Program

DuraSpace has announced its new Registered Service Provider Program.

Here's an excerpt from the announcement:

Today, the DuraSpace not-for-profit organization announced its new Registered Service Provider Program to establish partnerships with companies that provide support services to institutions using the DSpace and Fedora digital repository software.

The new program will establish an network of service providers offering a range of services including customer support, technical consulting, software development, and systems integration. The program will benefit universities, libraries, museums, research institutions, and others that require support in building or maintaining repository-based systems built with DSpace, Fedora, and related open source and commercial technologies.

Registered Service Providers will be easily identified. They will be featured on the DuraSpace web site (duraspace.org) and relevant project websites, with contact information and a profile of their service offerings. Providers will also exhibit their affiliation with DuraSpace by displaying the special DuraSpace Service Provider Logo on their marketing materials and websites.

Registered Service Providers share the DuraSpace commitment to ensuring that current and future generations have access to our collective digital heritage. Service Providers are active participants in open source software communities and are committed to providing expertise and technical consulting to enable customers to achieve their goals with open technologies.

Institutional Repositories: BibApp 1.0

The BibApp development team has released BibApp 1.0.

Here's an excerpt from the announcement:

BibApp is a campus research gateway and expert finder. It matches researchers on your campus or research center with their publication data and mines that data to see collaborations, create visualizations of areas of research, and find experts in research areas. With BibApp, it is easy to see what publications can be placed on the Web for greater access and impact. BibApp can push those publications directly into an institutional repository.

BibApp allows researchers and research groups to promote research, find collaborators on campus, and make research more accessible. It also allows libraries to better understand research happening in local departments, facilitate conversations about author rights with researchers, and ease the population of the institutional repository. Finally, BibApp allows campus administrators to achieve a clearer picture of collaboration and scholarly publishing trends on campus.

BibApp is the result of a collaboration between the University of Wisconsin-Madison and the University of Illinois at Urbana-Champaign. The Illinois Informatics Institute at the University of Illinois (https://www.informatics.illinois.edu/icubed/) provided generous funding for the development of the 1.0 release of BibApp.

BibApp is a Ruby on Rails application, coupled with the Solr/Lucene search engine, and either MySQL or PostgreSQL as its datastore. It uses open standards and protocols such as OpenURL and SWORD and automatically pulls in data from third party sources such as Google Books and the Sherpa/Romeo publisher policy database. BibApp imports publication data in RIS, MEDLINE and Refworks XML bibliography formats and exports data in several citation formats (APA, Chicago, IEEE, MLA, more) via CiteProc. BibApp also provides a web services API for delivering data as XML, YML, JSON, and RDF. BibApp is released under a University of Illinois/NCSA Open Source License (http://www.opensource.org/licenses/UoI-NCSA.php).

RIAN, Ireland's National Portal for Open Access, Launched

RIAN, Ireland’s national open access portal, has been launched by the Irish Universities Association Librarians' Group.

Here's an excerpt from the press release:

Ireland’s new national portal for Open Access to Irish published research goes live today. RIAN – http://www.rian.ie will act as a single point of access to national research output, and contains content harvested from the institutional repositories of the seven Irish Universities and Dublin Institute of Technology. RIAN will significantly increase the visibility and impact of Irish research and will expand to harvest content from other Irish Open Access providers as the service develops.

A national network of institutional repositories will increase the exposure of national research output, and allows services, such as enhanced searching, and statistics generation, to be developed using economies of scale. RIAN will demonstrate the impact of research to potential funders, who recognise the value of wider research dissemination.

The Irish Government has identified growth in research as critical to its future as a knowledge economy. Raising the research profile is a key strategy in the Universities’ strategic plans, and the ability to showcase research output and identify institutional research strengths is extremely important in attracting new funding and high quality staff.

The development of RIAN was managed by the Irish Universities Association Librarians' Group and is supported by the Association. This three year project was equally funded by the Universities and the Irish Government’s Strategic Innovation Fund which is administered by the Higher Education Authority.

Digital Repositories: Multiple-Deposit Function Added to EasyDeposit SWORD Client

Stuart Lewis has announced that a multiple-deposit function has been added to the EasyDeposit SWORD client.

Here's an excerpt from the announcement:

For those of you unfamiliar with EasyDeposit, it is an online tool that allows you to configure your own SWORD client. It is intended that you run multiple copies of EasyDeposit and configure each for a specific tailored use, such as thesis deposit, journal deposit, multiple deposit etc. The deposit process is made up of a set of 'steps' which you can configure and change into a preferred order to make your chosen client. . . .

The new multiple deposit functionality allows the administrator to 'hard code' the details of a set of repositories, and upon completion of the deposit process the item is deposited into each of those repositories. EasyDeposit has been designed with extensibility in mind, so if you wish to write your own 'steps,' for example to allow the depositor to select which repositories from a given list they would like to deposit into, this is easy and straightforward to write in PHP.

Digital Repository Software: DSpace 1.6.1 Released

DuraSpace has released DSpace 1.6.1.

Here's an excerpt from the announcement:

DSpace 1.6.1 is primarily a bug-fix release, which means it does not introduce any new features but improves existing features and fixes bugs discovered in earlier versions of DSpace. This is also the first [minor] version of DSpace to use a “time-driven” approach to release, rather than a “feature-driven” approach: we set a deadline and worked towards releasing on that date, instead of releasing when a certain number of issues were resolved. DSpace 1.7 will be the first major time-driven release.

Open Harvester Systems 2.3.0 Released

The Public Knowledge Project has released Open Harvester Systems 2.3.0.

Here's an excerpt from the announcement:

This is a major rewrite of numerous parts of the Harvester code, including metadata storage and indexing. It increases indexing flexibility to support plugin-based indexing, including Lucene/SOLR support. It also adds OAI Data Provider support, including the potential to convert between metadata formats (currently from various formats into Dublin Core).

Networked Digital Library of Theses and Dissertations Now Has More Than 1 Million ETD Records

The Networked Digital Library of Theses and Dissertations now contains records for over one million electronic theses and dissertations (ETDs).

Here's an excerpt from the press release:

The NDLTD, OCLC (Online Computer Library Center), VTLS and Scirus maintain and provide access related to the NDLTD Union Catalog of ETDs available in institutional repositories around the globe.

The NDLTD is an international non-profit organization dedicated to promoting the creation, dissemination, use, adoption, and preservation of digital theses and dissertations. The NDLTD assists students and universities in using electronic publishing and digital libraries to more effectively share knowledge in order to unlock potential benefits worldwide. The NDLTD also promotes student efforts to transform the genre of the print dissertation through the use of innovative software to create cutting edge hypertext/multimedia ETDs.

The NDLTD is comprised of many individual member institutions and consortia, each of which has or plans to put in place a process for archiving and distributing ETDs; others are welcome to join if they have similar interest. The Union Catalog Project is an attempt to make these individual collections appear as one seamless digital library of ETDs to students and researchers seeking out theses and dissertations.

In 1997 the first ETD program requirement was instituted at Virginia Tech. Over the course of thirteen years ETD programs have now been implemented in thousands of colleges and universities around the world. The one millionth ETD milestone indicates that ETD implementation is beginning to reach a critical mass. Indeed, in January, the count exceeded 800K records, while as of April 19, the record count reached 1.6 million, though there may be some duplicate records.

In the higher-GDP countries, institutions are rapidly adopting ETDs on a per-institutional or state-wide basis. Many lower-GDP countries are adopting ETDs at a national level as one means of jump-starting and disseminating research and development activities. The NDLTD anticipates that the number of ETDs worldwide will increase rapidly as more schools in every region around the globe implement ETD programs.. . .

Many institutions around the world are represented in the NDLTD Union Catalog. Universities can participate by implementing the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) to contribute metadata records to the union catalog. The NDLTD provides free resources to implement OAI-PMH in an institutional repository.

Digital Curation and Preservation Policies in Scottish HEIs: Survey Results

The Digital Curation Centre has released Digital Curation and Preservation Policies in Scottish HEIs: Survey Results.

Here's an excerpt:

The survey showed that repositories are still relatively young, and that preservation is not yet the highest priority for them. The situation with preservation policies also reflects the early stage of repository development, where the need to apply explicit curation policies is only beginning to be acknowledged.

The survey did not identify any institution level preservation policies, but given the heterogeneity of digital information across any higher education institution, it is not surprising that institution-wide preservation policies have yet to be formulated. Repository level policies were found to be in place at four institutions that also reported to be offering preservation services.

The survey reported a very low level of awareness of both existing preservation policies and digital preservation issues in general, especially amongst administrative and research staff. Enforcing preservation policies and making them effective are challenges that all HEIs face; at the same time, this formative period could be considered an opportunity for the ERIS project to develop supporting tools and guidance, especially since the prospect of additional guidance and the possibility of centralised services for preservation were welcomed by the respondents.

"Digital Repositories at a Crossroads: Achieving Sustainable Success through Campus-wide Engagement"

Jean-Gabriel Bankier and Courtney Smith, both of Berkeley Electronic Press, have self-archived "Digital Repositories at a Crossroads: Achieving Sustainable Success through Campus-wide Engagement" in SelectedWorks.

Here's an excerpt:

Repository initiatives were, at the outset, driven by two noble desires: to remove barriers to access; and, to begin to address the scholarly communications crisis. For universities across the globe, this specifically meant a focus on collecting peer reviewed journal articles. As we discovered together, neither faculty nor other campus constituents were impelled to invest or take ownership in the endeavor and the failure rate among digital repositories was very high.

Over the past few years a new model for the institutional repository has begun to emerge. To guarantee the long-term viability of the institutional repository (IR), the IR must be made integral to units on campus beyond the library. By working closely with Senior Administrators (like Provosts, Deans, and Department Heads), as well as faculty and students, librarians are offering valuable, targeted services that meet constituents’ needs and fulfill the goals of the repository. With this approach, the scope and value of the IR transcend a limited administrative or library function to fundamentally change the role of the library on campus.

DSpace 1.6.0 Demonstration Repositories

DSpace has released DSpace 1.6.0 demonstration repositories.

Here's an excerpt from the announcement:

The DSpace team announced today that an updated DSpace Demonstration Repository running DSpace 1.6.0 is now available for the community to use. The DSpace Demonstration Repository is a good place to run demonstrations, or to use as a sandbox for testing DSpace software before installing it. . . .

This demonstration site provides a sample repository with new DSpace 1.6.0 features enabled. This demonstration site also includes all DSpace interfaces (JSPUI, XMLUI, SWORD, OAI-PMH, LNI), connected to the same underlying database (so items created via XMLUI will also appear under JSPUI).

Also of interest: "screencast showing DSpace 1.6 authority control for author names and publishers from @mire.

Digital Audio: What's New in Fedora 3.3 and DSpace 1.6

DuraSpace has released What's New in Fedora 3.3 and DSpace 1.6.

Here's an excerpt from the announcement:

On March 17, 2010 Chris Wilper and Stuart Lewis offered a one-hour overview of new features in Fedora 3.3 and DSpace 1.6. The session concluded with a lively question and answer period with some of the 55 participants from around the globe.

Preservation and Curation in Institutional Repositories

The Digital Curation Centre has released Preservation and Curation in Institutional Repositories.

Here's an excerpt from the announcement:

The DCC has produced a report that provides a snapshot of the state of the art of preservation and curation in an institutional repository context in early 2010, noting areas of recent and current research and development. It should be of interest principally to institutional repository managers and others concerned with the strategic planning for these services. The report begins with a brief introduction to preservation and curation, followed in chapter 3 by a summary of the current provision for these activities in EPrints, DSpace and Fedora. Some repository models and architectures relevant to preservation and curation are presented in chapter 4 and chapter 5 respectively, while a selection of preservation planning tools of possible use in a repository context are described in chapter 6. Pertinent developments in metadata are reviewed in chapter 7, while tools for working with such metadata are presented in chapter 8. Technologies that assist in performing emulation, reverse engineering and migration are described in chapter 9. The issue of identifiers for repository materials is tackled in chapter 10. A selection of guidelines and tools for auditing curatorial aspects of institutional repositories is presented in chapter 11, and a selection of tools for calculating the costs and benefits of curation is presented in chapter 12. Finally, some conclusions are drawn in chapter 13.

Papers from Open Repositories 2009 Published in Journal of Digital Information

Selected papers from the Open Repositories 2009 conference have been published in a special issue of the Journal of Digital Information.

DSpace 1.6 Released

DuraSpace has released DSpace 1.6.

Here's an excerpt from the announcement:

Community-requested features in the new release include an enhanced statistics package which provides more information about how your repository is being used, an embargo facility so items can be kept dark for a period of time, and a batch metadata editing tool which can be used to change, add, find/replace metadata as well as facilitate mass moves, re-order values or add new items in bulk. And there’s more such as authority control which contains an integration with the Sherpa Romeo Service for publisher names, as well as the Library of Congress Nameservice. Other new features include:

  • Delegated administration
  • OpenSearch
  • Command launcher
  • OAI-PMH harvesting of items from remote repositories
  • Configurable OAI-PMH dublin core output
  • Move item functionality in XMLUI
  • If-Modified-Since / Last-Modified header support in XMLUI
  • Change to logging behaviour to ensure better log retention and management
  • Update to the latest handle server library
  • Ability to perform batch imports and exports from zip files of items
  • New test scripts to test database and email settings
  • Ability to set legal jurisdiction in creative commons licensing

University of Rochester's IR+ Institutional Repository Software

The University of Rochester Libraries' IR+ institutional repository software has been overviewed in a recent Inside Higher Ed article ("Encouraging Open Access") and a University of Rochester press release.

Here's an excerpt from the press release:

Created by a team at the University's River Campus Libraries, UR Research (https://urresearch.rochester.edu/home.action) is a one-stop Web site for managing the academic workflow. The suite of online tools meets an array of research needs, from authoring manuscripts to showcasing work to storing digital materials securely and when needed, permanently.

The system is designed to give researchers the incentive they need to upload their work to Web-based archives. "It's a win-win relationship," explains Suzanne Bell, the librarian charged with introducing the system to the University community. "Researchers get the tailor-made functions and online storage they need, Internet users get free and open access to academic research and priceless collections." For libraries, the program was designed as open-source software available for download to other institutions at no cost.

"It's a cyber work space that's collaborative," says Natalie Klein, a doctoral candidate in brain and cognitive sciences who quickly became enamored with the site after its recent launch. Klein particularly likes the ability to create and share documents in virtually any format. As a psycholinguist, her research data is quantitative and requires specialized text formatting, coding that is typically stripped when using popular online text sharing programs like Google Documents.

Klein also loves the user control built into the program, allowing her to create a customized researcher page and to easily post and update her CV and research. "I can upload documents myself instead of harassing a Web developer," she points out. "And the support has been great. I can't get that with Google Documents or even my department's Web site."

Doug Guiffrida, associate professor of counseling and human development in the Warner School of Education, agrees. The library's support team has always been "fantastic," he says. "They have made using this super easy."

Guiffrida uses the online archiving system to keep his work organized in one place and available for search engines and the public, who are free to download documents from the system. (Users also may have private work areas that cannot be searched or viewed by the public.) When a colleague asks for a paper, Guiffrida emails the UR Research link; when he gives a presentation, he posts the PowerPoint file on the system and shares the URL, a more earth friendly practice than printing out stacks of handouts, he points out. Even in his own office, Guiffrida says, "I actually use it instead of keeping hard copies around, if I need to look at something that I've written. I'm kind of a minimalist." And, unlike most departmental Web sites, the system tracks how often files are downloaded. "It's nice to see what people are reading and what they are ignoring," says Guiffrida.

An early adopter of UR Research, Guiffrida first began using the system when it was originally introduced in 2003 as simply a digital repository, a place for scholars to preserve and make available online preprints, dissertations, working papers, photographs, research data, music scores, and other work. But five years after its launch, only 6,500 documents had been entered into UR Research, says Nathan Sarr, the library's senior software engineer who designed the expanded system.

So the development team went back to the drawing board to find out what was wrong. Under the leadership of Susan Gibbons, Vice Provost and Andrew H. and Janet Dayton Neilly Dean of River Campus Libraries; David Lindahl, software development director; and Nancy Fried Foster, director of anthropological research, they conducted an anthropological study of faculty members and a second study of graduate students. What they discovered was that researchers needed software for working collaboratively with their colleagues, whether down the hall or across the globe. They needed online tools for sharing different versions of manuscripts in a safe and secure environment. And they needed a place to showcase their research. Graduate students, in particular, needed safe storage for their theses and the ever growing mounds of digital data from research. And all researchers, juggling demanding schedules, needed a system that saved time and was easy to use.

"We saw there was a disconnect between the repository and faculty needs," says Gibbons. "Then we set out to build a system that addressed all of the misalignments we found. Every significant feature of this system can be traced back to a finding of our faculty and/or grad student research."

"What's novel about this system is that it has been built around user needs," adds Mike Bell, assistant dean for information technology. "Most other repositories have been focused on systems architecture and public access, not on what is most important to contributors."

Presentations from Repositories and the Cloud Meeting

Presentations from the recent Repositories and the Cloud meeting, which was sponsored by Eduserv and JISC, and are now available.

Presentations included "Cloud-Based Projects at Belfast e-Science Centre," "Cloud Services for Repositories", "DuraCloud—Open Technologies and Services for Managing Durable Data in the Cloud," and "EPrints and the Cloud."

Read more about it at "Slides and Observations from “Repositories in the Cloud” London."

JISC Digital Repository infoKit

JISC has released the Digital Repository infoKit, which was created in association with the Repository Support Project.

Here's an excerpt from the press release:

The Digital Repository infoKit is a practical 'how to' guide to setting up and running digital repositories. The kit contains information on a broad range of topics running from the initial idea of a digital repository and the planning process, via detailed sections on repository set up and promotion, through to the maintenance and ongoing management of the repository. The main focus is on institutional repositories and the kit reflects current repository community best practice. This resource has been written for repository administrators. It assumes no prior knowledge of repository matters and, more importantly, assumes no prior technical knowledge. The kit can be used by anyone who needs an introduction to any of the topics covered.

"DSpace Manakin UI: Case Study of Value and Costs"

Eric Jansson has self-archived "DSpace Manakin UI: Case Study of Value and Costs" in the National Institute for Technology in Liberal Education's repository.

Here's an excerpt from:

Our experience in this project pointed towards three main ideas:

  • Many smaller institutions are staffed sufficiently to use the Manakin technology for branding and interface tweaking, as they can leverage web development skills, providing that system administration support for maintaining a development DSpace is available.
  • Development of advanced Manakin themes (defined as development that significantly alters, removes, or rearranges interface components, or that integrates new functionality into the interfaces) is comparable to web-application development in terms of complexity. As such, more advanced development is only likely to succeed for institutions whose staffing includes a dedicated expert in web application development, and one who is familiar with XML technologies (i.e. XML, XPath, XSL), and possibly AJAX, and Javascript.
  • Smaller institutions with more advanced goals should consider working together and using outside expertise in theme development. Outside expertise could be supplied through partnering with other institutions pursuing Manakin development projects or through vendors.

Read more about it at "Using Manakin to Expand the Capabilities of DSpace Repositories."

Hong Kong University Institutional Repository Uses Scopus API for Researcher Citation Data

Researcher pages in Hong Kong University's institutional repository will be updated with citation data generated by Elsevier's Scopus API.

Here's an excerpt from the press release:

The Scopus API offers users the opportunity to creatively interact with Scopus data by building mashups. It also allows access and usage of Scopus data inside and outside of the traditional library domain through applications based on the API. The API returns Scopus data in a format that easily integrates into an application or a web site. The majority of Scopus data is already available through the API, which can currently be used to request very specific information about article references, citations and affiliations.

HKU is the first institution to show Scopus h-index, and counts of citations, documents, and co-authors for each current HKU author across the institution, in its local institutional repository, The HKU Scholars Hub (The Hub). These details are shown on The Hub ResearcherPages, an expert profiling system which showcases the research of each current HKU author. HKU uses the Scopus API to build these pages, and update them in real time.

The Scopus search API draws on live data from Scopus, the world's largest abstract and citation database. By using the API, HKU is able to populate The Hub with real-time Scopus information, increasing accuracy and enriching data with valuable citation information. The API also enables HKU to highlight its overall performance and automate the process of keeping faculty publication lists up to date through continuous electronic tracking of individual researcher output. Research metrics cumulated by paper, and by author, are brought seamlessly into The Hub and displayed on appropriate records. This flexibility is a result of a recent enhancement to the Scopus API which allows for easier and more scalable ways of implementing citation counts to instantly enrich the content available on a given platform. . . .

HKU plans to further enhance ResearcherPages to display other output and citation details and expects there will be new opportunities to leverage the results in key activities such as the grant application process. HKU is also using the additional bibliometric data from Scopus and others to prepare for an impending Research Assessment Exercise.

EasyDeposit, Toolkit for Creating SWORD Deposit Interfaces, Released

Stuart Lewis has released EasyDeposit, a toolkit for creating SWORD deposit interfaces.

Here's an excerpt from the announcement:

EasyDeposit allows you to create customised SWORD deposit interfaces by configuring a set of 'steps'. A typical flow of steps may be: login, select a repository, enter some metadata, upload a file, verify the information is correct, perform the deposit, send a confirmation email. Alternatively a deposit flow may just require a file to be uploaded and a title entered. A configuration file is used to list the steps you require.

EasyDeposit makes use of the CodeIgniter MVC PHP framework. This means each 'step' is made up of two files: a 'controller' which looks after the validation and processing of any data entered, and a 'view' which controls the web page that a user sees. This separation of concerns makes it easy for web programmers to edit the controllers, and web designers to tinker with the look and feel of the interface in the views.

PEER Behavioural Research: Authors and Users vis-à-vis Journals and Repositories; Baseline Report

The Publishing and the Ecology of European Research (PEER) project has released PEER Behavioural Research: Authors and Users vis-à-vis Journals and Repositories; Baseline Report.

Here's an excerpt from the press release:

The PEER Behavioural Research Team from Loughborough University (Department of Information Science & LISU) has completed its behavioural baseline report, which is based on an electronic survey of authors (and authors as users) with more than 3000 European researchers and a series of focus groups covering the Medical sciences; Social sciences, humanities & arts; Life sciences; and Physical sciences & mathematics. The objectives of the Behavioural Research within PEER are to:

  • Track trends and explain patterns of author and user behaviour in the context of so called Green Open Access.
  • Understand the role repositories play for authors in the context of journal publishing.
  • Understand the role repositories play for users in context of accessing journal articles.

The baseline report outlines findings from the first phase of the research and identifies the key themes to emerge. It also identifies priorities for further analysis and future work. Some interesting points to emerge from the first phase of research that may be of interest to a number of stakeholders in the scholarly communication system include:

  • An individual's attitude towards open access repositories may change dependant on whether they are an author or a reader; readers being interested in the quality of the articles but authors also focused on the reputation of the repository itself
  • Reaching the target audience is the overwhelming motivation for scholars to disseminate their research results and this strongly influences their choice of journal and/or repository
  • Researchers in certain disciplines may lack confidence in making preprints available, and to some extent this is not only a matter of confidence in the quality of a text but also due to differences in work organisation across research cultures (e.g. strong internal peer review of manuscripts versus reliance on journals for peer review). Other factors are likely to include career stage and centrality of research to the parent discipline
  • Value-added services, such as download statistics and alert services, would contribute to the perceived usefulness of repositories and could help them gain popularity in what is an increasingly competitive information landscape
  • Readers often need to go through a variety of processes to access all the articles that they require and widespread open access may reduce the need for this time consuming practice.

OCLC Makes New OAIster Interfaces Available

OCLC has made basic and advanced OAIster search interfaces available. Access is free.

OAIster is a database of over 23 million records from OAI-PMH-compliant digital repositories, which was originally developed by the University of Michigan Library. Initially, OCLC made OAIster available only as part of WorldCat and as a FirstSearch database (these access points remain). (Thanks to ResourceShelf.)

Read more about it at "OCLC Makes OAIster Records Available through WorldCat.org," "OCLC makes OAIster Records Available through WorldCat.org to Ensure Long-Term Public Access to Digital Resources," and "University of Michigan and OCLC Form Partnership to Ensure Long-Term Access to OAIster Database."