The EVIA Digital Archive project is a joint effort of Indiana University and the University of Michigan, supported by the Andrew W. Mellon Foundation, to build a digital preservation and access system for ethnographic field video. Since its inception in 2001, the project has involved a collaborative team of ethnomusicology scholars, librarians, legal experts, and university administrators to design and develop policies, workflows, tools, and infrastructure. In addition to preserving and providing access to important source content contributed by scholars, the project brings these scholars together to create detailed, peer-reviewed annotations of the content, which is then used as metadata to enable scene-level search and access. Software developed by the project, including a video annotation tool and controlled vocabulary management tool, will be made available to the larger community as open source.

In this presentation, we will provide an overview of the project and its current status, and then discuss issues of collaboration, sustainability, and the project's significance for ethnomusicology and allied disciplines such as anthropology and folklore. We will also discuss EVIA's role in the new Institute for Digital Arts and Humanities at Indiana University, a collaborative support unit for innovative digitally-based arts and humanities research and creative activity.

What is web archiving? To collection managers (librarians, archivists and faculty) it means collecting, preserving, and providing access to web content in the same way that they have always done for analog material. To lawyers it means entering a world of risk to be mitigated. To technologists (architects, programmers, graphic designers, preservationists) it means working with systems and content that is much more complicated then usual.

Harvard's Web Archive Collection Service (WAX) brought all of these players together in the last two years to design and implement a pilot web archiving service that takes into consideration all of these diverse viewpoints. Challenges include defining collections with amorphous boundaries, managing a multitude of complex IP and other legal issues, addressing QC for material too vast to review comprehensively, navigating crawler traps, an explosion of formats, and handling of duplicate content. Come hear the challenges encountered from multiple perspectives and see a demonstration of the WAX system.

DLF has long supported community interest in improving access to digital library collections and has sponsored projects and initiatives to support this goal. Sharing digital collections through metadata aggregation is one path to providing a higher level of exposure. Both the Distributed Library: OAI for Digital Library Aggregation project (http://www.diglib.org/architectures/oai/imls2004/index.htm) and the Aquifer initiative have been focused on making digital content easier to find and use. This presentation will describe a data analysis project that bridges the Distributed Library and Aquifer projects and will describe assessment activities for Aquifer's American Social History Online Web site.

Colin Koteles will report the results of a data analysis project that examines metadata harvested for the DLF MODS portal (http://quod.lib.umich.edu/m/mods/)created through the Distributed Library project. The analysis examines conformance to the DLF/Aquifer Implementation Guidelines for Sharable MODS Records. Colin completed this study for a Certificate of Advanced Study in Digital Libraries at the Graduate School of Library and Information Science at UIUC.

Colin used existing OAI/PMH metadata from a sampling of DLF partners to test conformance to the DLF/Aquifer Implementation Guidelines for Shareable MODS Records. The results of this analysis are used to predict both the nature and extent of future normalization processes required by OAI service providers and the nature and extent of training and education required by OAI data providers hoping to expose MODS records that are useful in a variety of shared contexts. Results emphasize the need for application profile authors to leverage the automatic computing processes available to service providers and the unique, subjective knowledge data providers have of their collections.

Deb Holmes-Wong and Susan Harum will offer a brief status report on the American Social History Online Web site, including ongoing assessment activities. The past year has seen the implementation of a faceted browsing feature, a federated search feature, a doubling of collection items, and the optimization of the American Social History Online site for Google and Zotero. This presentation will: 1. Report assessment results for the implementation of the federated search feature 2. Describe preliminary results of assessment of the Web site itself 3. Describe what assessment efforts using Google analytics and web log analysis reveal about use of the website and 4. Outline assessment plans for the remainder of the project including a closer look at Zotero and Google and a connection to American Social History Online from within the Sakai course management system.

OCLC's WorldCat Grid offers a wide array of bibliographic data and services via a suite of application program interfaces (APIs), including the ability to search and retrieve records from the massive WorldCat database and the ability to interact with several controlled vocabularies.

WorldCat Grid Services and the supporting OCLC Developer's Network will be introduced, and specific Grid services such as the WorldCat Search API and Terminology Services will be highlighted. Attendees will learn how they can take advantage of these services to enhance their web sites, services, or software applications.

Staff from Indiana University Digital Library Program (IU DLP) will demonstrate a prototype application that queries a subject vocabulary for graphic materials available from OCLC. IU DLP staff will discuss their experiences using OCLC's Terminology Services and outline future directions for tools and services that use networked vocabularies.

Current methods of access to Electronic Resources are far from ideal. Availability of library-licensed electronic resources addresses the need for convenience and the internet-acclimated expectation of ubiquitous availability of information. However, the problems that users encounter when accessing resources from off campus, as well as the machinations needed by library staff to manage access control and troubleshoot problems, raises the question of how to improve usability and manageability.

In 2007, the Library Shibboleth Pilot Project was started to investigate these issues, and explore how federated authentication could improve the user and library experiences. This session will address use cases, recommended technologies and proposed future directions, as well as identifying opportunities for international collaboration and real-world testing.

The University of Illinois at Urbana-Champaign Library introduced a gateway portal in September 2007 that serves as a front-end to numerous information resources. Many academic libraries have deployed these "single-entry search box" type gateways. The Illinois gateway is built around a locally developed metasearch recommender and discovery system.

While there has been a great deal of research on user behavior within web search engine environments, there has been little work, other than OPAC use studies, of user behaviors within a multi-function library portal. This paper will report on user search and clickthrough behaviors within the Illinois portal.

Our metasearch software offers a variety of assisted search techniques to improve user search strategy formulation and modification, and aid in directing users to the appropriate information resources. This includes providing spelling suggestions, limiting of results to phrase and title searches, author search prompts, direct resource links, and re-ordering of search results to match perceived user needs. The software suite is heavily instrumented to record user-entered search arguments, software changes in their search arguments (such as stopword removal), system suggested actions, and user actions in response to the assisted search functions. Many of the assisted search techniques were suggested by a detailed transaction log analysis of user search arguments. We have found that user information seeking behavior in our academic environment--with regard to types of queries, user modification behaviors, the number of terms per query, search session dynamics, and other factors--differs from the behaviors reported in the web search engine literature.

The Digital Archive Services (DASe) project at UT Austin is a joint effort of the UT Libraries, the College of Liberal Arts, and the College of Fine Arts. DASe is a digital repository of images, audio, video, and pdf documents, and is used for managing, sharing, repurposing, and presenting digital media. Given the heterogeneous nature of both the assets themselves and the user populations, we have relied on the Atom Syndication Format and Atom Publishing Protocol as the primary data format for import, export, and storage.

The DASe application draws design principles from the Architecture of the World Wide Web (http://www.w3.org/TR/webarch/) and the principles of Representational State Transfer (REST) that underly it. DASe provides a uniquely open approach to metadata schemas ("Data First" as opposed to "Structure First", cf. http://www.betaversion.org/~stefano/linotype/news/93/), which allows minimal obstacles to adding new materials and the ability to import from and export to a wide range of metadata schemas, e.g, Atom, VRA Core, MODS, etc.

Our presentation will highlight the challenges, benefits, and outcomes of this approach. We will also discuss our investigations of OAI-ORE and the possible benefits that such a standard might offer our project. Special attention will be given to DASe not as an all-in-one solution, but rather DASe as an important player in the digital repository ecosystem that that includes DSpace, Fedora, EPrints, Google Base, Flickr, Metaweb, and others.

With the advent of information technology, it is imperative that we have a scheme to preserve digital born objects that were the products of office automation, geospatial images, and multimedia artifacts. As the first step of any preservation process, it is critical to be able to identify and validate the format of a digital object, and generate technical metadata embedded in that object. Some institutions have developed tools that can discover the format of a digital object beyond using file extensions.

In this paper, we studied and evaluated two tools, which are well-known in the domain of digital archiving and preservation: o JHOVE (JSTOR/Harvard Object Validation Environment), o DROID (Digital Record Object Identification).

The extent of capabilities against a same corpus of sample documents will be examined. We also will look at their respective performance, which is very important, due to the massive amount of digital objects, and practical deadlines of the ingest phase.

For types that are known to JHOVE, we found that there was no statistically significant difference of execution times between the two tools. But, with a mix of file types known and unknown to JHOVE, DROID seemed to fare better than JHOVE in terms of execution time. But JHOVE's processing generated richer metadata, which can be used as information in the asset catalog. Due to this duality, we would like to propose a software architecture which could leverage the benefits of JHOVE and DROID in terms of performance and metadata generation.

OpenMIC is an open source, web-based cataloging tool that can be used as a standalone application or integrated with other repository architectures by a range of organizations. It provides a complete metadata creation system for analog and digital materials, with services to export these metadata in standard formats.

• Full METS support
• Low overhead and infrastructure requirements
• Events-based data model for management and rights documentation
• Customization capabilities
• Mapping from both standard and local in-house schema
• Import and export utilities
• Unicode and CJK vernacular character support

The OpenMIC cataloging utility is one component of the MIC portal for moving images. MIC (Moving Image Collections) is a preservation, access, and education initiative co-sponsored by the Library of Congress and the Association of Moving Image Archivists (AMIA), in partnership with Rutgers University Libraries, the technology lead. MIC (pronounced 'Mike') integrates a union catalog, directories, and the cataloging utility with informational resources in a portal structure delivering customized information on moving images, their preservation, and the images themselves to archivists, researchers, educators, and the general public. MIC's mandate is both to document the national moving image collections environment and to serve as a coordinator and catalyst for the preservation of the nation's moving image heritage. MIC's mission is to immerse moving images into the education mainstream, recognizing that what society uses, it values, and what it values, it preserves.