Streaks and flashes of lightening intermittently lit the sky in Chapel Hill, North Carolina as JCDL conference-goers picked their way through puddles to attend the 2006 Joint Conference on Digital Libraries (JCDL)–Opening Information Horizons, held June 11-15.

Highlights of the conference included a plenary, workshops, tutorials, informal meetings, and a North Carolina “Pig Pull” where the winners of the ACM Vannevar Bush Best Paper Award were announced. Carl Lagoze, Tim Cornwell, Naomi Dushay, Dean Eckstrom, Dean Krafft and John Saylor won the $1,000 award sponsored by the Association for Computing Machinery (ACM) for their paper “Metadata Aggregation and “Automated Digital Libraries”: A Retrospective on the NSDL Experience.

In presenting the paper Lagoze commented that it reported on “A lot of hard work by many smart people.” His talk reviewed successes and benefits that the NSDL Project has achieved for the digital library production and research community in the context of analyzing organizational practices. He offered his remarks in the spirit of gaining an understanding of future efforts for supporting community-based deployment of well-established digital library technology. His talk did not offer solutions for distributed metadata generation, collection, and management. Rather, he suggested that the results reported in the paper should lead to collaborative discussion in the digital library community to understand the impact of these results and what they indicate about future work on metadata generation and aggregation.

He reviewed NSDL 1.0 architecture and workflow and emphasized that NSDL had found “low barriers” to participation were not low enough for many NSDL metadata providers. He also presented some surprising statistics: “A cross section of email (primary method of interaction with all providers) archives of 8 representative providers revealed over 2,700 messages, or around 170 messages per provider per year.” The paper concluded that while the NSDL 1.0 architecture led to the successful deployment of a digital library, the human effort costs were inordinately high and the resource discovery benefits gained from the structured metadata were compromised due to quality and incompleteness.

Lagoze commented, “I’ve said it elsewhere, but it may be time that we (digital) library types understand the real costs and benefits of human-generated descriptive metadata and the role in can practically play in large-scale digital libraries. It may be time to put aside notions that descriptive metadata can play a dominant role in resource discovery, and put our efforts into the mechanisms for knowledge enhancement and context building that can distinguish our efforts from the search engine vendors. Our reaction in NSDL is to focus our efforts on the deployment of a FEDORA-powered NSDL 2.0 architecture that focuses on resource context rather than structured description.”