Found @ Your Repository: Increasing Visibility of IR Content

Eric Van de Velde

The number of institutional repositories worldwide has grown exponentially in the last three years. Much progress has been made on setting up and managing a repository but less on increasing the visibility of an institute’s research output. Today, being found is as important, if not more important, as being online. The more visible the content, the more it will be used and shared. Library Connect talked to Eric Van de Velde, Director of Library Information Technology at Caltech, to find out more about their institutional repository, Caltech CODA (  http://coda.library.caltech.edu).

LC: Can you tell us about Caltech’s institutional repository?

Eric Van de Velde: Our earliest repository goes back to 1998 when we put a collection of computer science technical reports online. The real kick-start was the Santa Fe meeting in October 1999, which established the Open Archives Initiative (www.openarchives.org). At the meeting, I was introduced to the Networked Digital Library of Theses and Dissertations (www.ndltd.org).

Upon returning from Santa Fe, we started experimenting with the NDLTD software, and the library started working with Professor Arden Albee, then our Dean of Graduate Studies, on a pilot project for electronic theses. In 2002, we introduced the requirement for Caltech theses to be submitted electronically. Currently our repository houses well over 2,000 theses and is the largest repository collection we have. Caltech produces about 200 theses per year. Our collection includes older theses we scanned. When starting this project, we didn’t intend to scan our archives, but after a flood damaged about 200 circulating theses, we scanned archival copies as a preventative measure. This project went so well that we started scanning older material as time permits.

After the Santa Fe meeting, Stevan Harnad started the EPrints initiative (www.eprints.org), and when the EPrints software became available about a year later, we started using it almost immediately. First, we moved the computer science technical reports. Then, we built a historical archive from the earthquake engineering research laboratory. As the service became more well-known, some faculty became interested in making their research papers, technical reports, books and even conference proceedings available online through the Institutional Repository. There’s also an interesting collection of Caltech oral histories, consisting of the archive department’s interviews with retiring professors and senior administrators.

LC: What’s the take-up of repository services at Caltech?

Van de Velde: We have to get to a situation where faculty consistently submit content to the repository so the library doesn’t have to do all the work. That’s a social change that still needs to happen.

LC: How are you increasing awareness of the repository?

Van de Velde: Lots of informal conversations! The librarians do more formal recruiting, but it’s at a low pitch at the moment. We’re making the service available, mentioning it wherever possible but not hitting people over the head with it. We’re always looking for ways to make the process as easy as possible. My guess is, students graduating now who have put their theses online and seen the response will be faculty in a couple of years, here and across the country. They will want the same interactivity for their research papers. I see this as an evolutionary process, not a revolution. It’s simple, but these changes take a while.

LC: What level of usage are you seeing on materials in the repository?

Van de Velde: Quite a bit actually. Via the library website, we provide usage statistics for all repository collections. I did some studies on thesis usage and discovered an average of six to seven unique accesses per thesis per year. It doesn’t sound like a lot, but compared to use in print or even microfilm, it’s a tremendous increase in visibility.

LC: Can you tell us a bit about the work you’ve been doing with Scirus to increase visibility of your IR’s content?

Van de Velde: The Open Archives Initiative is a three-layered system. First, you have the content providers, such as institutional and disciplinary repositories, but these could also include journals (open-access journals as well as conventional journals). Second is an intermediary layer, provided by the OAI initiaitve, giving a standard interface to the disparate repositories available. The vision of the Santa Fe meeting was that service providers would come along and use this standard intermediate layer to access data underneath. That’s what is starting to happen now with service providers like Scirus.

So, whenever a service provider comes along and wants to access our data we are happy to collaborate to make our collection visible. We want our content out there: the more service providers (i.e. the more entry points to our content), the better.

Our collaboration with Scirus has also provided us with one search interface for the disparate Caltech repositories. Until now you had to search each respository individually.

LC: Would you recommend other libraries explore similar collaborations?

Van de Velde: Yes, my big recommendation for academic libraries is to make their unique materials available through the Open Archives Initiative. Service providers can then make the material visible in a variety of contexts.

LC: What do you see coming down the line for institutional repositories?

Van de Velde: Well, this is certainly only the beginning. Right now the vast majority of theses are text files. The NDLTD recently announced awards for innovative multi-media theses. Once we are truly in a multi-media age, that will open up the realm of what a scientific publication really could be and what institutional repositories will really have to offer.

Scirus and Institutional Repositories

Institutional Repositories require serious and relevant search functionality to make their content more accessible. Scirus has developed a unique indexing process matching a repository’s metadata with full text and allowing them to be indexed together. This is achieved by considering the overall structure of the IR and determining the best method of indexing for optimal searching. By working with an institute to index its repository, Scirus can also power the search of the repository’s site to ensure more complete search results, relevant ranking of results based on in-depth classification and an interface optimized for the repository’s content.

“Search plays a critical role in increasing the visibility of a repository. Being online doesn’t guarantee visibility. Ensuring Institutional Repositories are made visible within Web search engines and implementing a good search capability on a repository’s site can be difficult to implement. Good search capability in particular requires expertise in search technology coupled with knowledge of the structure of scholarly material. Partnerships such as those Scirus is exploring will help to make this a successful effort.” Sharon Mombrú, Senior Product Manager, Scirus, Elsevier, Amsterdam, The Netherlands.   www.scirus.com