The ShelfLife Collaborative

We are a loose coalition of people and institutions working toward a vision of the Digital Public Library of America (DPLA) embodied in software, as part of the DPLA "beta sprint."

Our vision

1. DPLA is not just a way to find online library materials, but a place.
2. In this place teeming with activity, users can engage with the available works and each other, thus adding to the cumulative understanding of our cultural heritage.
3. All users are able to easily locate materials if they know what they want, or browse easily and intelligently if they don't...and always can share what they learn.
4. DPLA uses every scrap of information, and information about information (metadata) to help users discover and learn, of course while respecting privacy.
5. The DPLA enhances the ability of local libraries to achieve their missions.
6. The DPLA not only supports open standards, but is a leading provider of linked open data, thus becoming an integral part of the cultural life of the Web.

Our DPLA proposal

The DPLA asked for 400-word proposals by June 15, 2011. Here is what we submitted:

ShelfLife is intended to provide users with a rich environment for exploring the combined content of the DPLA, discovering new works and engaging more deeply with them via social interactions. It displays the available content via a scrollable virtual representation of library shelves, providing users with their most familiar experience of physical libraries. But any work can instantly be shown in a variety of different shelves (facets), clustering it with other works to facilitate serendipity and contextual understanding. ShelfLife will support a "click and play" interface for reading/viewing available items. Users will be guided to new works by a recommendation system based on aggregated anonymous usage data and, importantly, via rich social interactions with friends and fellow fans of works. The information generated by users explicitly (by writing reviews, favoriting works, creating collections, etc.) and implicitly (by doing the digital equivalent of "checking works out", browsing interactively, etc.) will be fed back into the system's recommendation engine, so that it learns. ShelfLife will integrate data from local public libraries and will make available to its users the real-time expertise of librarians.

ShelfLife is designed to be an open application that can incorporate extensions and accommodate new features and types of data. For example, we are partnering with another DPLA beta sprinter, extraMUROS, by providing access within ShelfLife to the images, videos and audio in its distributed Web-wide collection, and to extraMUROs' creation tools.

LibraryCloud is the backend metadata server that supports ShelfLife. It will make the multi-library metadata ShelfLife uses available via open APIs, in order to spur and enable the development of innovative new applications, such as library browsers, library analytic tools, and recommendation engines. LibraryCloud will also explore the use of various tools to manage record matching and de-duplication across multiple data sources to aid in the aggregation of heterogeneous usage statistics.

The members of Shelflife Sprint Collaborative are collectively making available their library catalogs, anonymized usage data, search query logs, and other metadata, as well as assisting with (meta)data management and processing.

ShelfLife, and LibraryCloud are open source projects. Both will support linked open data.

Members

The members of the ShelfLife Sprint Collaborative come from a wide range of institutions, about evenly divided between public libraries and university libraries.