Cyprien Lomas

Can Higher Education Repository Projects Learn from Flickr?

Cyprien Lomas, University of British Columbia
Wednesday, September 28, 2005, 1:15-2:00 pm

mp3 podcast Right click and 'Save Target As' for manual download

Repositories have been part of the academic landscape for many years. Stores of information collected by individuals or groups have been used for research and academic purposes. While the library has often been the location for repositories, they are also created by research groups, teaching groups or just about anybody wishing to store and organize information. Unfortunately, little communication by creators of repositories often results in ‘silos’ of information that evolve with little attention paid to standards, interoperability or thoughts of preservation. Repositories may be as simple as files stored on a computer or as complicated as custom applications residing on networked computers. Sharing of resources may or may not be part of the design of the repository, and consequently, intellectual property issues are seldom in dispute.

In recent years, renewed (continued) interest in repositories has resulted in new cadres of individuals (or groups) proposing, designing, creating and implementing data stores for use within their particular disciplines. While the creators of the new databases are building them with standards and reuse in mind, the challenge of promoting reuse by anyone other than the owners and creators of the resources still remains. Inadequate promotion of databases combined with general lack of knowledge of them results in their under use. While older databases are left to wither and die, they are duplicated and ultimately superseded by newer initiatives.

Meanwhile, outside of the institution, photo sharing applications like Flickr and BuzzNet are enjoying great popularity and success with the general public. While Flickr is primarily a website to share pictures, it has many features in common with repositories. Photos are stored online and are available in several resolutions. They can be categorized for easier retrieval. Photo creators can restrict access to their content by setting permissions; subsets can be shared with trusted friends and peers. Metadata and usage data is collected and stored along with the photos.

In addition to repository-like features, Flickr incorporates a number of other features that may contribute to its success. All pictures have an absolute URL and can be easily included in other webpages. In addition, tools for easy incorporation into weblogs have been built for most of the existing weblog packages. Publishing an open API has resulted in the creation of several plug-ins that extend the functionality of Flickr.

Friends and colleagues can collect and view photos without sharing them with the general public. Pictures can be ‘tagged’ to allow their discovery by parties with similar interests and groups can self discover and self organize using the ‘friend of a friend’, recommender/referrer and other social software features of Flickr. One additional significant feature of Flickr is its embrace of standards including RSS and Atom. All of these features go far to ensure that Flickr is versatile, works with almost every other system out there, can be ‘hacked’ by those who wish to do so and promotes a ‘social’, fun type of engagement.

The University of British Columbia created several image repositories as part of a campus wide Learning Objects pilot. These projects were part of a larger pilot project looking at institutional repositories and their utility as campus wide learning object repositories. Of the pilot projects undertaken, image repositories were amongst the most successful having the largest number of loyal users and greatest number of records. While these projects included some of the features seen in Flickr, an analysis of the successful features of Flickr and a comparison with UBC projects may help to enhance the success of the institutional projects.

A Biological image repository that was discipline specific, locally ‘owned’, and sensitive to the requirements of its users was created. Its intended purpose was to store research data for use by fellow scientists as well as instructors and students. The repository managed to fit the needs of researchers, instructors and students resulting in records serving as research objects, teaching objects and learning objects. In addition to serving as a repository, some of the additional features of this repository promote community, flexibility and sharing.

In a second UBC project, a community site related to a Botanical Garden serves as the place for amateur botanists and gardening enthusiasts to meet and discuss a wide variety of botanical topics. The site consists of a weblog, moderated discussion boards, and links to the scientific image repositories of plant scientists. Volunteers from the general public are enlisted to ensure that the site runs well. Enthusiasts have helped to make this site one of the more vibrant amateur botanical communities on the web and a potential outreach point to make scientific research accessible to the general public.

This paper attempts to explore those design features that help to make a repository useful to its users. By examining the goals of institutional projects and fitting them to some of the features that have been successful outside the academy, we hope to identify design principles that contribute to the success of open, useful research object /learning object repositories.