Executive Summary
The Task Force conducted a survey of visual resource collections in UC libraries to assess their scope, characteristics, usage, and digitization status The findings revealed an impressive total of over 11.4 million images, with approximately 236,000 already digitized These collections, primarily sourced from special, map, and arts collections, cater to diverse user needs and standards, although limited usage data suggests they support a blend of research and instructional purposes.
The Task Force conducted user studies to identify essential functions, services, and support needed for digital images The delivery of digital images for instructional purposes is intricate, necessitating enhanced development and evaluation of presentation systems Additionally, it is crucial to address policy matters, particularly concerning copyright and fair use, while also fostering stronger collaborations with campus professionals to improve user support.
The Task Force examined the potential for establishing a union catalog of digital visual resources, recognizing that UC libraries possess a robust collection and a solid foundation for developing effective discovery and delivery systems for digital images It is essential to clearly define the necessary access systems, determine an architectural model that allows for suitable services and flexibility, and establish the standards and management systems to support these initiatives Given the areas requiring further investigation and consensus, the Task Force advises delaying the decision to create a union catalog focused exclusively on digital visual resources.
A union catalog may not fully meet the diverse characteristics of the collection or the unique needs of various user groups To enhance decision-making, it is beneficial to explore interim approaches and evaluate ongoing studies.
The Task Force has put forth 15 recommendations aimed at enhancing the digital accessibility of visual resource collections The top three priorities include adopting an architectural model for future development, establishing minimum metadata and digital object standards for images, and creating comprehensive copyright and fair use policies Addressing these key areas promptly will lay a robust planning framework and establish essential infrastructure, allowing for incremental progress on other initiatives as opportunities emerge Additionally, the recommendations highlight how the CDL Image Demonstrator service can facilitate exploration and experimentation in this domain.
Survey of UC Libraries’ Visual Resources Collections
SOPAG charged the Task Force on Digital Visual Resources to conduct a survey of UC libraries to determine:
1) what visual resources collections, including both moving and still images, are now owned or managed by the UC libraries;
2) what user groups these collections now serve; and
3) what use statistics are currently available for these collections
The Task Force defines "image" as a visible representation that acts as a surrogate for an original work, encompassing photomechanical, photographic, and digital formats An image typically manifests as a slide, photograph, or digital graphic file, serving as a reproduction of the original work Additionally, "visual resource image collections" refer to entities that produce, conserve, classify, and provide access to both analog and digital images, including slide and photograph collections.
The Task Force surveyed libraries about the following areas in order to address the other aspects of its charge (see Appendix VIII):
• Nature and extent of the collections, including unique holdings, number of items by format
• Percentage of items in the collections that are licensed, commercially purchased, owned with or without digitizing rights, and public domain
• Extent to which the collections are cataloged and what systems have been used
• Whether content management and collection management systems are in place or are desirable
• Extent to which collections circulate
We inquired about both analog and digital collections, specifically focusing on whether analog collections are being digitized or planned for digitization within the next three years, and the intended purpose of this process.
Task Force members and liaisons distributed the survey along with a cover letter to collection managers on their campus, starting October 30 with a deadline of November
Respondents participated in the survey online via WebSurveyor, and a summary of the findings has been compiled by Rosalie Lack from the California Digital Library Comparative summaries of various questions for the Task Force are included in Appendix II.
1 See http://www.cdlib.org:8081/libstaff/visualresourcesurvey
Scope of the Collections
What visual resource collections, including both moving and still images are now owned or managed by UC libraries?
All ten campuses' collection managers participated in the survey, revealing fifty-six notable visual resource collections (refer to Appendix I 2) As a result of these surveys, the estimated total of visual materials owned by UC libraries is approximately 11.4 million items.
Most collections primarily exist in analog format, with only a few available solely in digital form or as digital versions of their analog counterparts Currently, there are approximately 236,000 images across twenty-three digital collections Additionally, some of these digital collections have been prepared for collaborative programs at UC, including initiatives in California.
Heritage Project or the Online Archive of California
The fifty-six collections can be categorized into four main subject areas: 25 collections (45%) focus on the Arts, encompassing fine arts, architecture, and theater arts; 15 collections (27%) are dedicated to the Humanities, covering topics such as travel, history, and news coverage; 8 collections (14%) pertain to Science, including fields like medicine, astronomy, and biology; and the remaining 8 collections (14%) are centered around Maps.
The University of California system boasts a diverse range of image collections, from the extensive analog holdings at UCB’s Bancroft Library Pictorial Collections and UCLA’s Los Angeles Times Photo Archive, each containing approximately three million images, to smaller collections such as UCI’s 300 medical slides and unique art collections at UCLA In terms of digital resources, UCSD’s Visual Resources Collection leads with 90,000 images, followed by Santa Barbara's map collections with 58,900 images, and UCB’s Bancroft Library Pictorial Collections featuring 45,000 images, all of which are being digitized for ARTstor.
Of the forty-five collections responding to the survey, 75% are primarily used for research , while others are used for instruction or for other purposes such as recreation
The two largest instructional collections are the slide collections held at UCSD and UCSC, although nine others report some instructional use (Appendix II, Table 2)
Nearly half of the collections consist of photographic formats, while respondents highlighted the presence of various other media, including film and video Additionally, specialized formats like 35mm slide film often complement or serve as counterparts to the primary formats Notably, one collection at UCSF also features streaming media.
Respondents emphasized the uniqueness of their collections, highlighting the richness and depth of materials owned by UC libraries There is a strong desire to enhance accessibility to these unparalleled collections.
After the survey, we obtained information on several UC Davis collections due to the arrival of the new Head of Special Collections, as well as two late responses from UCI The totals from these collections are included in Appendix I to illustrate their potential contributions, although the detailed analysis of survey responses does not encompass this information, which is marked as “no survey completed” in Appendix I.
Users of the Collections
What users groups do these collections now serve?
Analog collection statistics (Appendix II, Table 4)
The survey results reveal an incomplete understanding of the use and users of visual resources, as many collections do not differentiate this data from other materials Additionally, certain questions yielded limited or conflicting responses, preventing us from drawing reliable conclusions.
Out of 45 surveyed collections, 33% (15) track usage statistics, although 7 cannot distinguish visual resource statistics from general data Additionally, 17% (8) do not collect any statistics, while 49% (22) did not provide a response Furthermore, 55% (25) of respondents stated that visual resource materials are limited to in-house use, with only 4% (2) reporting that their resources circulate, one to the public and another exclusively to campus affiliates.
In the 2001-02 reporting period, 49,573 patrons accessed analog visual resources collections, with 31% being faculty, 4% students, and 1.4% categorized as “Other.” When identifying their primary clientele, 37% of respondents indicated that "All Users" were their main audience, while 41% specifically noted Non-UC affiliates and 29% identified UC affiliates However, the open-ended responses do not provide clear insights into the specific subject disciplines that primarily utilize these analog visual resources.
Usage Statistics kept Yes: 15 (33%) of 45 but only 8
Circulate Yes: 2 (4 %) of 45 No: 25 (55%) inhouse use only No response: 18 (40%)
% of those identifying All Users 5 (29%) Non-UC 7 (41%)
Digital visual resources statistics (Appendix II, Table 12)
There was even less information reported concerning the use of digital visual resources
Among the respondents, 37% reported having digital visual resources, yet only 9% maintained web usage logs, and a mere 4% tracked user counts This suggests that the reported figures of 2,041 users and 224,000 hits may only represent a small portion of actual engagement Most digital collections are publicly accessible, although three are restricted to UC affiliates or researchers, and one is still undergoing digitization Additionally, UCSC and UCSD provided usage statistics for their digital reserve services from 1999 to Fall 2003, indicating that around 8,000 images were placed on reserve during the 2001-2002 period.
57 courses in a variety of departments UCSD experienced a 68% increase in hits to its Digital Image Reserves web site over 2000-2001.
Use of Standards
A recent survey examined the types of descriptive metadata and classification systems utilized for various collections Out of the 45 collections analyzed, more than half were either uncataloged (15) or did not provide sufficient information (11) Among those that did report their practices, nine respondents indicated they use MARC for descriptive metadata, with two of them noting that they have extended or supplemented this system Additionally, three respondents mentioned using Encoded Archival Description (EAD).
Archival Description (EAD) (or EAD plus Metadata Encoding and Transmission
The study reveals that while one collection adheres to the Metadata Encoding and Transmission Standard (METS), another follows the Federal Geographic Data Committee standard, and five respondents utilize a locally developed classification scheme In terms of classification systems, eight collections employ the Library of Congress system, one uses Fogg, and 16 respondents reported using various other systems.
The digital collections exhibit a diverse range of metadata standards, with six utilizing MARC (including two with extensions), two employing EAD, two using Dublin Core, and two adopting alternative systems Additionally, seven of these collections are organized according to the Library of Congress classification, while four utilize different classification systems, and four remain unclassified.
Larger collections typically adhere to established cataloging standards, while smaller, specialized collections often lack formal cataloging or employ unique, locally developed systems Additionally, many special collections may only be cataloged at a broader collection level, rather than individually.
Management Tools
To effectively manage and enhance accessibility to collections, libraries require both content management systems and collection management systems Content management systems facilitate the distributed creation and access of items, while collection management systems focus on inventory and accession functions These functionalities may be integrated within a single system or handled separately through different systems.
Content management: distributed creation and access
3 There are 17 survey responses with two additional collections identified by UCLA.
Out of 45 analog collections surveyed, only nine participants indicated having a management system in place, with just three planning to acquire one within the next three years—one of whom already possesses a system Eight collections utilize the library's integrated library system (ILS), although one respondent criticized it as merely a MARC catalog that falls short as a management tool Additionally, five respondents reported using alternative systems, including a locally developed 4D database, the Autonomous Archive, and Alexandria Digital Library’s proprietary software, while two did not specify their systems It is worth noting that some references to different systems may pertain to digital collections, as respondents indicated a lack of a content management system.
Digital collections show a positive trend, with 23 collections either containing digital objects or planning to digitize Among these, seven have implemented a content management system, and two intend to acquire one within the next three years Notably, three of these systems are integrated within library systems such as GLADIS, Harvest, and Roger, while the remaining four utilize specialized systems.
Collection management: inventory and accession
Seventeen analog collections currently possess collection management capabilities, with one additional collection planning to implement a system within the next three years Among these, five collections utilize their Integrated Library System (ILS), while eight rely on finding aids or inventory lists Additionally, six collections reported using a database for management purposes, with three of those employing finding aids also utilizing an Access database, one using Filemaker, and another using Excel.
In the realm of digital collections, eight institutions utilize a collection management system Among these, three rely on their Integrated Library System (ILS), while four employ relational databases such as MS Access, ProCite, and Informix Additionally, one institution manages its inventory using an Excel spreadsheet.
The largest analog collections, such as UCB’s Bancroft Pictorial Collections and UCSB’s Map and Imagery collection, benefit from robust system support, utilizing various content and collection management systems Among the three largest digital collections derived from these analog sources, two have implemented content management systems, while only one possesses a collection management system Notably, all major analog collections feature automated catalogs, with the exception of the Bancroft Pictorial Collections, which remains partially automated and primarily cataloged at the collection level, alongside UCLA’s Los Angeles collections.
The Angeles Times Photographic Archives and the UCSB Map and Imagery Laboratory feature partially automated systems Among the smaller analog collections, ten have implemented automated catalogs, while one has reported partial automation However, five of these collections indicate that cataloging is conducted only at the collection level.
The libraries with the highest total number of images, including UCB, UCLA, and UCSD, present a varied landscape At UCSD and UCB, distinct management systems are employed for various types of collections, while UCLA boasts a singular, extensive collection, notably the Los Angeles Times.
The Photographic Archives holds approximately 3 million images out of a total of 3.2 million, currently lacking a formal management system aside from an unpublished finding aid While some smaller collections utilize finding aids backed by an MS Access database, a comprehensive organizational framework is still needed for the larger archive.
Readiness of Digital Collections to be Federated
Our analysis focused on the cataloging and searchability of existing collections, as well as their potential for follow-up This assessment aimed to determine if there are enough readily available digital images to initiate a union catalog or explore alternative federation methods.
A review of the UC collection responses indicates that 235,000 digital images are housed within 19 of the 45 collections, with only UCSB and UCSD reporting portions of their collections as "born digital." The majority of collections have analog counterparts, and four collections noted that over 10% of their analog materials are also available digitally Only seven collections plan to digitize in the next three years, primarily for preservation and increased access to visual materials, with two already participating in the OAC's California Cultures program Currently, these digital collections are accessible through local UC campus library sites, and all but one collection, UCSD's Geisel Library/Art & Architecture Library's Visual Resources Collection, are primarily utilized for research purposes.
To maximize the potential of UC digital image collections for research and instruction, it is essential to engage with collection managers to evaluate the content and readiness of existing image files for future integration into a UC union catalog/image service Understanding the commitment level of collection managers and staff toward processing and maintaining these collections is crucial, especially since only 9 out of 19 collections are fully cataloged There is a diverse range of encoding standards in use, including EAD, MARC, Dublin Core, and VRA Core Categories, alongside inconsistent image capture formats and sizes An in-depth assessment is needed to ensure images meet sufficient resolution and quality standards, as well as to confirm that full cataloging and appropriate metadata documentation are in place Additionally, compatibility of encoding standards with those of the UC union catalog/image service should be evaluated to facilitate effective mapping and crosswalking of metadata.
User Studies
The Task Force analyzed various user studies focused on the use of images to gain insights into content issues, user support needs, and system features The research highlighted differences between instructional and research requirements, as well as disciplinary variations While not a comprehensive literature review, we identified relevant recent studies, particularly those emphasizing the use of images in art for instructional support Additionally, we examined studies on web-based image search engines to understand user expectations in academic systems Research on how humanities scholars adapt their methods in the electronic environment was also reviewed, alongside an informative report from the LUCI project, which addresses challenges in creating a union catalog of images Lastly, we obtained a preliminary report from the Visual Image User Study, further contributing to our understanding of user needs.
The VIUS project at Pennsylvania State University, supported by the Mellon Foundation, is a comprehensive 26-month initiative focused on evaluating the needs for a digital image delivery system within the university's intricate structure.
We developed a comprehensive framework (Appendix III) addressing key issues identified in our studies, concentrating on three primary areas: content creation, content accessibility, and the utilization of content for instruction and research Within these areas, we classified the issues into specific categories for better clarity and understanding.
Content: which content and how much to provide, including federation of different types.
User Needs: what capabilities do users indicate they need to use image systems successfully, ranging from content to search features to image quality.
4 Marmor, Max and Barbara Rockenbach, “Image Matters: a Binocular View (Yale + Luna Imaging) of the Digital Marketplace”, presented at UCSD and UCI on March 27th and 28th, 2001.
5 Pitt, Sharon P, Christina B Updike, and Miriam E Guthrie, “Integrating Digital Images into the Art and Art History Curriculum”, Educause Quarterly, 25(2), 2002 http://www.educause.edu/ir/library/pdf/EQM0225.pdf
6 Hassan I.; Zhang J., “Image search engine feature analysis”, Online Information Review, v25, n2 (20 April 2001): 103-114.
7 Brockman, William S Laura Neumann, Carole L Palmer, Tonia J Tidline, Scholarly Work in the
Humanities and the Evolving Information Environment, Washington, D.C.: Council on Library and
8 Burns, Maureen, Library of UC Images—California Arts Project, Final Report to the University of California Institute for Research in the Arts, February, 2002
9 Pennsylvania State University, Visual Image User Study, Interim Report to the Andrew W Mellon Foundation, December 2, 2002 (not yet published).
Standards: what are the standards and other technical issues to be addressed.
Technical Infrastructure: what are the issues that involve the wider campus infrastructures for delivering image systems (networking, classroom support).
User Support: what assistance do users need from libraries or other organizations to use digital images.
Organizational: what are implications for how these issues affect our staffing, organizational structure, partnerships, etc.
The framework emphasizes the importance of defining roles in addressing issues related to library systems, highlighting that libraries play a crucial role due to their vested interest in user success However, it is often essential to collaborate with external academic parties to enhance effectiveness and ensure successful outcomes.
Content
User studies consistently emphasize that providing a substantial amount of relevant content is essential for attracting users to digital collections Ensuring the availability of the "right content" is crucial, and this may require the conversion of some existing materials to meet user interests effectively.
The transition of UC analog collections to digital format, alongside the licensing of commercial collections, empowers primarily faculty users to incorporate their personal collections while emphasizing teaching requirements Research indicates that unique and specialized materials hold significant value for both research and educational purposes; however, user studies highlight the necessity of curating the "right" content to enhance its effectiveness in instructional settings.
Researchers have access to a diverse range of large image databases, including fee-based products like RLG’s Cultural Materials, the AMICO Library, and the Bridgeman Art Library, which offer extensive online collections of high-quality images from various cultures and time periods These resources enable researchers to browse and identify original works while obtaining valuable contextual information However, Brockman notes that digital products tend to prioritize the "canonical and influential" over the "marginal and esoteric," leaving researchers in search of lesser-known primary sources To complement these commercial offerings, many non-fee-based image collections are also available online, such as the Library of
Congress American Memory project; several from museum collections: e.g., San
Francisco Fine Arts Museums, the Kyoto National Museum, and the Guildhall Art
Gallery, London offers a wealth of content and image-rich resources; however, it's important to recognize that many open access collections are available at 72 dpi, which may not be suitable for high-quality classroom applications Furthermore, smaller digitized collections from various UC special collections primarily focus on supporting research rather than meeting instructional requirements.
Faculty members anticipate that a digital image delivery system will be more beneficial for teaching than for research or outreach To meet academic instructional needs, digitized collections must include relevant content, with 300,000 images often cited as the ideal number for a teaching collection in the visual arts Currently, while UC houses extensive analog image and photograph collections, only a small fraction has been digitized The library-managed visual resources collections at UCSD contain 150,000 digital images, with an additional 77,000 identified in other UC VR collections Although some duplication exists, these collections are likely to encompass essential materials The Mellon Foundation’s ARTstor collection aims to create a core digital image repository based on teaching needs, potentially streamlining digitization efforts nationwide by incorporating 80% of UCSD’s VR collection along with specialized research materials.
The integration of core collections with licensed materials will significantly enhance UC libraries' ability to create a substantial and relevant teaching resource for visual arts For a detailed overview, refer to Appendix IV, which summarizes the UC, licensed, and free collections available.
The University of California boasts significant complementary collections of maps and photographs, particularly focused on California To enhance accessibility, Shared Cataloging has launched two key initiatives aimed at cataloging major collections that are freely available: the David Rumsey historic map collection and the maps found in the Library of Congress American Memory collection.
UC offers a diverse range of digitized collections, primarily located in Special Collections, encompassing various disciplines These collections feature maps, satellite images, aerial photography, historical finding aids, photographs, medical slides, and 35mm slides, providing valuable resources for research and exploration.
The demand for teaching collections in health sciences is increasing, as highlighted by the Health Education Assets Library (HEAL), which offers approximately 2,000 diverse resources, including images, sounds, and videos These materials span various medical fields such as dermatology, obstetrics and gynecology, neuroanatomy, neurology, pathology, biochemistry, and cardiology To further enhance these resources, UCSF and UCLA are pursuing a grant from the National Library of Medicine (NLM).
To effectively localize the service, there is an exploration of complementary commercial collections; however, initial discussions suggest that faculty members may favor utilizing their own images or those developed by their colleagues.
At UC, we are privileged to possess both essential and impactful visual resources through our core teaching collections, alongside unique and specialized materials found in our special collections and faculty collections.
Academic research institutions, such as UC, typically maintain visual resource collections averaging 300,000 images, which are curated based on direct requests from faculty members.
12 See http://www.healcentral.org/index.htm
13 Persily, Gail, UCSF Library, interview conducted by Laine Farley, March, 2003.
User needs
The demand for an intuitive user interface is evident, as both experts and general users require effective search and browsing capabilities, with experts favoring direct searches for their tasks Variability in descriptive metadata further drives the need for robust browsing options Platforms like OAC, the CDL public website, and Luna’s Insight offer essential search and browse features Additionally, innovative methods that merge traditional concept-based retrieval with content-based algorithms—utilizing attributes like color, texture, shape, and spatial distribution—aim to resolve common challenges in image searching.
Image quality is crucial for research and classroom presentations, particularly in the arts, where faculty demand high-resolution, accurately color-managed digital images for lectures High-quality images facilitate formal analysis and detailed observation, ensuring compatibility with other teaching media While there are varying quality levels—ranging from born-digital items to scanned images—lower resolution may be acceptable in certain disciplines or for non-lecture purposes, such as browsing, visual identification, and creating supplementary materials like class websites Even art historians may accept lower quality for supplementary content.
Users increasingly seek efficient ways to view images through thumbnails, enabling them to select images for larger displays and side-by-side comparisons As these features advance, the demand for them is likely to grow James Mundy, director of Vassar’s Frances Loehman Loeb Art Center, notes that digital imaging provides art historians with enhanced control and detail, akin to the benefits microscopy offers to research scientists This technology allows users to notice details they might otherwise overlook Additionally, users desire comprehensive options to download, print, and email images and metadata for personal collections or presentations Public users may also wish to purchase access to high-resolution images or obtain permission for their use.
14 Enser, Peter, “Visual Image Retrieval: Seeking the Alliance of Concept-based and Content-based
The article titled "Paradigms" published in the Journal of Information Science, volume 26, issue 4 in 2000, spans pages 199-210 and can be accessed at http://pippo.ingentaselect.com/vl89277/cl2/nw=1/rpsv/ij/bks/01655515/v26n4/s1/p199 Additionally, the Princeton 3D Models Search Engine is available at http://shape.ps.princeton.edu/search.html.
15 Read, Brock, “Art History Without Slides”, Chronicle of Higher Education, v 49, n20 (24 Jan 2003) Available at http://chronicle.com/free/v49/i20/20a02901.htm
16 Hassan I., and J Zhang, “Image search engine feature analysis”, Online Information Review, v25, n2 (20 April 2001): 103-114 Available at http://www.emeraldinsight.com/rpsv/cgi-bin/linker? reqidx=/cw/mcb/14684527/v25n2/s4/p103.idx
Users also want to integrate images with other systems including those used for courses (learning management systems, web pages, reserves) and for users’ own tools
Faculty members are increasingly utilizing the web to share course-related information with students, incorporating both text and images At UCSC and UCSD, visual resource collections provide valuable image reserves Regardless of whether these image reserves evolve into an integrated image presentation service or remain distinct, it is essential for faculty to have the capability to curate a selection of images for student study.
User research reveals that both general users and specialists are keen to access additional contextual information, including artists' biographies, critical analyses, and insights into the broader cultural context Given that libraries often subscribe to this information from commercial sources, there is potential to offer access through a Subject Access Portal or a linking system like UC-eLinks at the point of discovery.
Standards
User studies highlight the need for developing effective standards and best practices for metadata that align with user requirements It is essential to evaluate whether these standards are suitable for the relevant subject matter and intended applications, as well as their compatibility with access and delivery systems Additionally, the scalability and sustainability of these standards must be considered to ensure long-term effectiveness.
Retrieving accurate images from the web remains a significant challenge for users According to Patricia Harpring, effective image retrieval relies on intelligent indexing, which can be viewed as the 'language' of retrieval; this, in turn, depends on proper methodology and suitable terminology For future large-scale projects like AMICO and ARTstor, the importance of data structure, cataloging rules, and vocabulary control cannot be overstated.
Expecting high metadata standards for existing faculty-owned analog slide collections and their digital materials is unrealistic Setting the bar too high may discourage contributions, so it's essential to allow for dynamic metadata that can evolve over time through user comments and technological advancements The challenges of creating metadata and reaching consensus on descriptive standards among academic communities indicate the need for user-generated metadata, which can include notes and comments While various metadata standards for images will exist, a flexible minimum standard should be implemented Over time, both users and collection managers can be incentivized to enhance metadata quality, leading to a progression from basic to more comprehensive standards.
17 Frost C.O., B Taylor, A Noakes, S Markel, D Torres, K.M Drabenstott, “Browse and Search Patterns in a Digital Image Database”, Information Retrieval, v1, n4 (January 2000): 287-313 Available at http://www.kluweronline.com/issn/1386-4564/contents
18 Baca, Murtha, ed Introduction to Art Image Access Issues, Tools, Standards, Strategies, (Los Angeles, CA: Getty Research Institute, 2002), p.20.
Technical Infrastructure
Campus infrastructure is often centered around two key themes: classroom readiness and a strong technical framework Faculty frequently voice concerns regarding the reliability of classroom projection equipment and network connectivity Essential elements of classroom readiness encompass access to suitable computers, high-quality projection systems, dual projection capabilities, equipment that meets the highest resolution standards, and adequate technical support.
High-resolution images and advanced projection technology can convince faculty that digital images are a viable alternative to traditional slides John Davis, chair of the art department at Smith College, noted his initial impression of digital projections, stating, "The images are just juicy on the screen."
Technical infrastructure is essential for effective real-time classroom instruction, requiring sufficient bandwidth, dependable network servers, and robust connectivity While off-campus access and round-the-clock system availability are crucial, faculty prioritize minimizing disruptions during valuable classroom time To address these needs, various campuses have established standards and specifications for the creation and maintenance of digital teaching environments.
User support
As the sharing and provision of instructional images transition into the digital realm, it is essential to address the diverse user needs that arise from complex campus-wide issues Libraries and other campus organizations must enhance their user support to effectively tackle these challenges While many of these issues are technical in nature, the most significant hurdles often stem from cultural factors that influence user engagement and resource utilization.
There is a growing interest in curating personal collections from central resources, which necessitates assistance in managing copyright, fair use, and permissions for access at different levels.
At Pennsylvania State University, faculty members emphasize the importance of creating personal collections but recognize the necessity for curatorial assistance in managing these collections Visual resources curators and instructional support staff will play a crucial role in helping faculty scan and prepare images and metadata However, incentives for faculty to contribute their collections to digital image catalogs are likely to develop gradually, primarily due to the challenges posed by technical issues.
19 Read, Brock, “Art History Without Slides”, Chronicle of Higher Education, v49, n2 (24 Jan 2003) Available at http://chronicle.com/free/v49/i20/20a02901.htm
20 See UCR Classroom Technology, Classroom Profiles at http://mediaresources.ucr.edu/classtech/profiles.html ; similar information at Dartmouth at http://www.dartmouth.edu/~insvcs/classrooms/standards.html
However, we should expect requests for these collections to become a part of the image catalogs in the future
Faculty members are seeking clarification from the university regarding copyright issues and the use of images in academic settings While guidelines for placing texts on course reserves have been established, many educators desire similar clarity for image usage They aim to make certain materials available for teaching purposes, ideally restricted to class members for the duration of the course Additionally, faculty may wish to contribute their own photographs or other images to a permanent collection, with varying levels of access restrictions These intellectual property concerns necessitate further examination and institutional guidance.
A Penn State study revealed a notable contradiction: faculty members believe a digital image system would greatly enhance teaching, yet they primarily utilize digital images for research purposes This trend aligns with other research indicating that analog collections will remain in use for the foreseeable future The findings suggest that the current support for integrating digital images into teaching is insufficient, particularly regarding the availability of content and the infrastructure needed for reliable delivery in classrooms Consequently, faculty find it easier to access and work with digital images tailored to their individual research needs.
To enhance digital skills across various levels, it is essential to provide training for faculty in utilizing image databases and presentation software, along with guidance on image preparation and management This support extends beyond the library's responsibilities and would be greatly enhanced through collaborations with professionals in visual resources, educational technology, and academic computing.
Research indicates that engaging enthusiastic early adopters among faculty is essential for integrating images into the classroom effectively While some faculty may need extensive support to gain confidence in using technology, others may remain hesitant to embrace digital content It may take a new generation of educators, currently in graduate school, to fully normalize the use of digital resources in teaching Therefore, our focus should be on preparing and supporting those who are entering the higher education field.
It is essential to recognize the needs of "cultural tourists," who represent a significant group of potential users for our publicly available image collections Research on web-based image search engines indicates that their requirements may align with those of general users, yet they often come with an expectation for purchasing options The newly launched California Digital Library (http://www.californiadigitallibrary.org) serves as a platform to showcase various digital collections and offers valuable insights into how public users prefer to engage with these resources.
Organizational
The organizational challenges derive directly from user needs and user support issues
To ensure compliance with copyright and fair use, it is essential to establish clear policies and procedures that guide faculty on the appropriate use of materials and provide reassurance regarding the contribution of their own collections Additionally, we must explicitly outline these policies as they relate to the images we own.
Supporting both analog and digital access for teaching collections necessitates adjustments in organization and staffing Additionally, the availability of digital images may alter usage patterns for physical collections in special collections.
To enhance user experience, it is essential to establish partnerships with various campus organizations Collaborating with instructional and educational technology teams can facilitate the integration of learning management systems and improve digital skills support Engaging with visual resources curators provides valuable curatorial assistance, guidance on building or licensing collections, and general instructional support Additionally, partnering with museum professionals can offer insights into supporting cultural tourism Lastly, working with information technology is crucial to ensure robust network and classroom capabilities.
Libraries cannot address these challenges alone; they require campus-wide initiatives, adequate funding, and a strong commitment to establish the necessary technical framework and encourage a cultural shift that supports and enhances the digital classroom experience.
Organizational and Operational Infrastructure
For image collections managed by the libraries, some of the organizational and operational issues to consider for providing digital images include:
capacity to create and manage images, including access to systems support;
agreement on standards and best practices; agreement on minimum standards and a process to move toward ideal standards
service capabilities: ability to provide access directly and/or federate with other collections; ability to support instructional, research and public needs
partnerships with information and instructional technology units or other professionals such as visual resources curators
OAC Manager Robin Chandler identified organizational requirements for data creation, data ingest, and data dissemination based on the OAC experiences For data creation, libraries need staff to:
analyze/determine status of metadata
assess how to encode metadata (inhouse or outsource)
develop manuals or training for staff
For data ingest or submitting the data to another agency for federation, the requirements might include:
provision for storing the digital content—who is responsible, what equipment is required
ability to handle version control to revise/submit metadata and content
ability to meet submission requirements
ability to automate submission procedures
As described in the previous section, many user studies illustrate the service impact on organizations of providing digital images including the following:
support for including users’ images—extension of reserve model
interpreting copyright/fair use policies; developing procedures
support for ad hoc requests for new images
support for faculty in gathering and managing images
Libraries are implementing various policies and tools to effectively manage permissions and authorized access to images While the survey did not specifically inquire about the organizational structure of libraries in relation to digital activities, insights from multiple campuses revealed their existing capabilities in this area.
The UCB Library has established the Digital Publishing Group (DPG) to serve as a model for libraries aiming to digitize and manage image collections for inclusion in union catalogs DPG is actively supporting other UC campuses through the California Cultures project and enhances digital library initiatives by offering managerial and technical analysis, scheduling scanning processes, developing digital content programs, providing digital publishing tools, and training staff.
UCLA has established a digital storage and retrieval system based on Making of America 2 (MoA2) principles, which supports Dublin Core metadata and can be adapted for various metadata types Meanwhile, the UCSF Library boasts significant expertise in the creation and management of digital texts, particularly through its projects involving digital images of tobacco industry documents.
The UC Davis General Library has established a Digital Library Collections Task Force to enhance digital initiatives This task force will soon present a report detailing current and potential digital projects, along with the necessary procedural, technological, and funding requirements for a UC Davis Digital Library Repository Key recommendations include creating a Digital Initiatives Program to serve as a hub for standards and best practices, as well as managing a digital laboratory to preserve UC Davis digital content and foster collaboration with other UC campuses and the broader University of California Digital Library.
This section examines the issues above in relation to how our image collections are organized and function within the libraries.
Organizational Characteristics of UC Library Collections
A recent survey indicates that all campus libraries oversee image collections, which can be categorized into four primary organizational units Despite variations in size, complexity, and technical capabilities, these collections often share common standards and service challenges.
Over half of the images managed by UC libraries, totaling around 6.5 million, are housed in special collections, with roughly 10% available in digital format These collections are characterized by their uniqueness, research focus, and limited circulation, with some not yet accessible When digitized, they may align well with the model of other Online Archive of California (OAC) collections.
Many collections lack reported content and collection management capabilities Nevertheless, the initiatives by OAC and its collaborations, such as the Museums in the Online Archive of California and the LSTA funded sustainability project with public libraries, have produced valuable resources These include metadata standards, training workshops, best practices, asset management system information, and documentation to assist departments in developing EAD finding aids and digital images.
Federating special collections with other images can lead to service discrepancies, particularly regarding access tools Typically, finding aids serve as the primary means to access images; however, if image access is disconnected from the finding aid, it may result in limited descriptive metadata and a loss of contextual information The OAC interface and CDL’s public site currently allow for separate image searches while maintaining a link back to the corresponding finding aid.
The map libraries at UCB, UCSB, UCSC, and UCSD form a significant cluster, housing an estimated 4.5 million items, including approximately 75,000 digital images The Alexandria Digital Library (ADL) stands out as a key service with ties to the research community and collaborations with the San Diego Supercomputer Center (SDSC) for operational support Challenges in service provision may arise from a cost recovery model for public access to images There is a recognized need to federate collections within the University of California system, particularly for resources like the Sanborn Fire Insurance maps, and to collaborate with government agencies While collection management varies, most collections are cataloged to some degree, typically utilizing MARC and Library of Congress standards.
The second largest group of art-related resources comprises approximately 4.5 million images, including around 152,000 digital images This group encompasses instructional visual resource collections at UCSD and UCSC, along with special collections at various campuses, such as UCLA and UCSB, which house significant art-related materials Additionally, many campuses feature specialized collections focused on art The visual resources collections are enhancing their capabilities by adopting content and collection management systems, while also collaborating with visual resource curators from other institutions Their involvement in the LUCI project 23 aims to develop a comprehensive image catalog, highlighting their commitment to providing extensive services related to image collections The art libraries and special collections exhibit similarities to the special collections in OAC.
A smaller cluster of campuses, including UCSF, UCI, and UCLA, features medical schools that collectively reported approximately 21,000 images, with 1,000 available in digital format Notably, there is a distinction between special collections at UCLA and UCSF.
21 See OAC Especially for Archivists, http://www.oac.cdlib.org/about/forarchivists.html
22 See OAC’s image search at http://www.oac.cdlib.org/search.image.html and CDL’s public site at http://www.californiadigitallibrary.org
The survey indicates a lack of clarity regarding the experience and expertise in creating or delivering digital images among certain educational units, particularly at UCSF and UCI, which focus more on teaching However, UCLA stands out with its strong expertise in digital resources within its library system Additionally, UCSF has experience with various digital formats and manages the Center for Instructional Technology, showcasing its capability to support instructional and classroom needs, especially as the demand for digital image support continues to rise.
Collection Management Capabilities
Survey responses indicate a significant variation in the capabilities of collections regarding accessibility, contributions to a union catalog, and the development of digital collections A systematic evaluation of content and collection management needs across campus libraries could yield collective benefits, especially in preparation for transitioning to digital images The OAC has explored digital object management tools, and many partners possess relevant experience If libraries opt for different systems, it is crucial to establish interoperability requirements aligned with the union catalog model and architectural model outlined in the appendices.
Core Collections That Are Not Library Managed
The Task Force focused on library-managed collections, yet numerous valuable image collections exist across campuses that are essential for research and teaching These collections often provide critical context for subjects only partially covered by library resources While many intriguing collections are found in various campus departments, certain groups manage image collections with missions aligned with libraries, supporting research and instruction Additionally, these groups possess specialized expertise in their subject areas and offer unique insights into image-related service issues.
Partnerships with some of these groups have already been established, notably with the museums that are part of MOAC, and with the Visual Resource curators (aka the
The LUCI project facilitates collaboration among various groups, including those with a history of cross-campus interactions and similar interests in other universities The ongoing evaluation of the MOAC partnership, part of the OAC initiative, is complemented by CDL’s Image Demonstrator project, which also serves as a new opportunity for collaboration with the Sliders Future collaboration may present anticipated challenges that need to be addressed.
24 See Digital Object Management Tools, http://dynaweb.oac.cdlib.org/dig/onres_dig_domgmttools.html
Standards: lack of awareness or adherence to different standards
Collaboration can be challenging, especially when balancing the desire for independence with the advantages of working together, as demonstrated by the Sliders' experience with LUCI Despite being a well-organized group, they faced difficulties in maintaining focus on their collaborative goals It is essential to designate a responsible party to ensure that these objectives remain on track and that the benefits of cooperation are fully realized.
Different user needs for system functions and support services
Forging alliances with various groups requires significant effort, but the OAC/MOAC experiments have proven successful CDL is committed to investing resources in facilitating promising collaborations and is starting to document these efforts effectively.
Libraries can enhance their activities by adopting collaborative structures and strategies that have proven effective for other groups Although it may be too early to pursue new partnerships, libraries are in a strong position to capitalize on emerging opportunities as their focus on digital image collections gains visibility.
Support for Faculty
As digital images increasingly play a crucial role in research and teaching, faculty will require varying levels of support Library visual resource collections are already assisting faculty by offering services such as scanning, creating digital reserves, and integrating diverse materials into course-related sites They also provide guidance on curating personal collections, conducting picture research, and locating visual materials for publications, along with advice on copyright and intellectual property matters Similar support is available from visual resource curators at other campuses.
Technical support for integrating digital images in the classroom is typically offered by campus instructional or information technology units Many campuses allocate funds through departmental budgets, instructional improvement grants, or teaching committees to enhance pedagogical development, often focusing on faculty website creation and media production for specific courses Collaboration between libraries, visual resource curators, and educational technologists is essential to support faculty in developing or redesigning courses with digital resources and to promote innovative technology use in teaching Given the complexity of presentation software, faculty will require in-depth training to effectively leverage its capabilities It is important to explore models that utilize existing campus resources to support faculty in these endeavors.
Copyright and Fair Use
The complexities of intellectual property rights in digital image creation and usage largely hinge on ownership and client circumstances For example, UC Libraries holds significant image content, particularly within its special collections of historic and primary source materials, and generates revenue by providing access to these images Conversely, maps produced by the U.S government typically face no copyright restrictions Additionally, the creation of slides or digital derivatives from copyrighted materials for educational purposes heavily relies on the principles of "fair use."
To comprehend copyright issues related to digital image usage in U.S educational non-profits, it's essential to explore the background of acquiring and creating slides as teaching materials.
Visual resource collections often encounter copyright challenges when creating slides from print materials to fulfill instructional needs While a limited selection of high-quality images is available from commercial vendors, many collections rely on faculty donations of slides taken during travels or hire photographers for original onsite captures, particularly of architectural and sculptural works Nonetheless, the majority of slides in these collections are produced by photographing images from books and periodicals to meet the demands of instructors To facilitate this, visual resource curators either employ photographers or train on-site staff to perform copy stand photography.
Visual resource collections typically establish policies for acquiring and producing slides, balancing the needs of instructors with the rights of image providers, such as publishers and artists Many collections interpret the "fair use" provision of U.S Copyright Law to allow the creation of slides from print sources for specific educational purposes To ensure compliance, these collections often limit the percentage of images sourced from any single provider and restrict slide use to institutional-specific teaching, prohibiting loans to other institutions.
In the 1990s, the rapid development of digitization technologies and the internet transformed the distribution of educational content, leading visual resource collections to adopt e-reserves This involved digitizing images for web access or using CD-ROMs for specific courses, effectively replacing traditional slide carousels due to student demand Digital image reserves allowed students to study anytime and anywhere, with access typically secured through passwords or IP restrictions However, this access was often temporary, limited to the course duration, reflecting a copyright philosophy that prioritized short-term availability over the creation of permanent collections.
The visual resources community has initiated the development of image databases aimed at enhancing discovery and display, with the potential to support digital image projection in classrooms At UC, SPIRO 25 (UCB-Architecture) was the pioneering effort to create a digital image database while addressing fair use concerns Additionally, the LUCI project investigated the concept of a union catalog of images and shared cataloging data, starting with contributions from UC faculty and staff, while restricting access to high-resolution images.
In 2002, the UCSD Libraries obtained a grant from the Mellon Foundation to digitize around 80% of its slide collection, totaling approximately 220,000 images Additionally, a related grant from the Mellon Foundation supported the Union Catalog of Art Images (UCAI) project, which highlighted the importance of this digitization effort.
Visual metadata, particularly in the form of thumbnail digital images, plays a crucial role in enhancing descriptive image records The inclusion of digital images significantly benefits the UCAI project by complementing the cataloging data for a large portion of the UCSD slide collection Additionally, the digitization of UCSD's slide collection aligns with Mellon’s goal of incorporating this essential set of instructional images into ARTstor However, it is important to highlight that slides obtained by UCSD from commercial vendors will not be digitized under the Mellon grant.
When digitizing special collections, it is essential to examine the donation terms and potentially obtain explicit permission from the donor or their heirs for digital content access, whether restricted or unrestricted This process ensures compliance with copyright regulations and respects the rights of the original contributors.
Statement and Statement of Use reflect the issues addressed by curators as well as notification to users about acceptable use 27
The Alexandria Digital Library also posts a statement about access and use which limits use to the UC community:
• Access constraints Access to full images by University of California students, faculty, and staff, for use on UC projects only
When utilizing materials copyrighted by CNES/Spot Image Corporation, it is permitted to create an unlimited number of films or prints for use in UC projects However, all copies, whether in hardcopy or digital format, must include the specified copyright notice.
25 See SPIRO at http://shanana.berkeley.edu/spiro/
26 See LUCI’s Copyright statement at http://vrc.ucr.edu/luci/luci.html
The content is licensed under CNES and managed by SPOT Image Corporation, with copyright restrictions outlined in the SPOT-UC agreement For detailed information on copyright and usage conditions, please refer to the Copyright Statement available at http://oac.cdlib.org/about/copyright.html.
The Regents of the University of California disclaim all implied warranties, including those of merchantability and fitness for a specific purpose, regarding the data provided This data is offered on an 'as is' basis without any accompanying warranty Furthermore, the Regents will not be liable for any direct, indirect, special, incidental, or consequential damages resulting from the use of this data or its availability, even if they have been notified of the potential for such damages.
Key questions related to the provision of digital images include:
1 If in creating a slide collection “fair use” principles have been followed, does this imply that providing digital versions of those images is allowed?
2 Can digital images be shared legally between multiple institutions?
3 Are there different legal considerations when providing digital image reserves versus providing permanent digital image collections?
4 Should there be limits to public access of digital images because of copyright issues?
To effectively navigate copyright and fair use challenges, UC libraries should leverage insights from visual resource collections, special collections, and Alexandria By developing tailored policies and procedures, these libraries can enhance their operational practices and provide valuable guidance to faculty.
Technical Architecture and Infrastructure
UC libraries manage diverse image collections that cater to various constituencies and exist within different organizational units This necessitates a distributed network to integrate with other collections, whether within UC or from commercial and free sources Management functions are likely to be located within collection units or their parent libraries, and access mechanisms should offer multiple options, either centrally or locally To leverage specialized search and presentation capabilities, central hosting of metadata may be essential, while persistent access to images could also benefit from a centralized approach For effective classroom delivery, distributing image files may be necessary, leading to potential duplication of metadata and images The CDL is advancing towards a layered service model that allows libraries to utilize components of centrally supported services Achieving this flexibility requires establishing minimum standards for metadata, export capabilities, and possibly developing tools for transferring metadata or images based on user needs.
Standards and Best Practices
The goals of standards as articulated by the OAC are relevant when considering digital images:
Balancing multiple goals can create tension, necessitating practical short-term solutions that pave the way for achieving more comprehensive standards in the long run.
Standards for digital images apply to metadata, interoperability and even to media- capable classrooms.
Libraries utilize diverse encoding standards for descriptive metadata, influenced by the subject matter, collection purpose, and available cataloging resources This results in a combination of various formats, sizes, and both structural and administrative metadata.
Building on the CDL Digital Object Standard 28 is a logical choice due to its comprehensive approach and established precedent, making it easily understandable for all stakeholders Additionally, its integration within METS allows for enhanced functionality and versatility in digital object management.
“genre” statement or “materials designator” 29
If a distributed model for collections is adopted, mapping existing descriptive metadata to Dublin Core is the most effective approach There is currently no universally accepted standard, and the likelihood of one emerging to meet all needs is low Although MARC has been utilized, it often necessitates the expansion of fields for digital objects While VRA Core serves as a strong standard for art history teaching collections, its applicability may be limited when considering the broader spectrum of visual resources, particularly for certain special collections relevant to art research.
Dublin Core offers an accessible entry point for new collections, making it applicable beyond the humanities Many existing metadata models already integrate elements of Dublin Core Additionally, it is evolving to establish best practices and standards for categorizing various formats and types of media, including video The Dublin Core Metadata Initiative group aims to finalize recommendations within two years to address the growing demand for video metadata.
Users increasingly seek comprehensive metadata to improve the discovery and utilization of instructional materials However, the necessary infrastructure for efficient metadata creation is just beginning to take shape, with the introduction of authority lists and thesauri such as the Union List of Artist Names and the Thesaurus of Geographic Names.
The creation of effective metadata in architecture faces challenges such as high costs and differing academic standards To enhance the cataloging process, it is essential to empower users to contribute metadata and facilitate dynamic storage of their notes and comments This approach allows for the continuous enrichment of content as technology evolves and user interactions are documented, ultimately leading to a more comprehensive and engaging resource.
Preservation requirements may necessitate elevated standards, yet it is essential to maintain flexibility to support various purposes and long-term objectives If Dublin Core serves as the baseline standard for discovery and integration with diverse collections, additional standards can be implemented based on the specific goals of collection sponsors Acknowledging the necessity for both minimum and ideal standards, as well as best practices, is crucial for effective accommodation in this field.
The CDL Digital Object Standard outlines essential guidelines for metadata, content, and encoding, as detailed in the document published on May 18, 2001 For further insights into digital image formats, refer to the Digital Image Format Standards, also available through the California Digital Library.
29 See METS Official Web Site at http://www.loc.gov/standards/mets/
30 See Getty Research Tools/Vocabulary Databases at http://www.getty.edu/research/tools/vocabulary/
31 See Cataloguing Cultural Objects: A Guide to Describing Cultural Objects and their Images: A Project of the Visual Resources Association at http://www.vraweb.org/projects.html
32 See Union Catalog of Art Images at http://gort.ucsd.edu/ucai/project.html
Architecture Implications for Existing Systems
To address user needs, we developed a flexible architecture that leverages existing and potential systems for users and collection managers Our model envisions a distributed system employing various methods to federate collections, with digital object creation and management handled by campus libraries or their units Metadata can be harvested into a central repository, or ingested for specific collections, such as those in OAC, facilitating easier access to metadata through various discovery tools.
For effective interoperability, it's essential that images and their associated information can be accessed across various environments To achieve this, multiple methods for harvesting and delivering image data are necessary, even if the images are stored in a single repository This flexibility can be realized by enhancing existing systems, allowing for diverse access options that meet our objectives.
An effective image repository should integrate CDL Digital Object Standards with Open Archives Initiative (OAI) harvesting This involves mapping existing collections to Dublin Core, a straightforward task for technical staff across campuses Proven harvesting techniques will be utilized to create a Dublin Core-based service, with the harvester, likely developed by CDL, collecting and merging records at regular intervals UCLA's involvement in the Sheet Music Consortium project has provided valuable experience with this model.
Results obtained from this service can be referenced back into the source collection on a
The UC campus and repositories like OAC can extract images from EAD documents as separate objects, which can be delivered to specialized tools such as Luna Insight or the Alexandria Digital Library This process not only identifies the original collection but also provides users access to the intellectual context if needed Additionally, when distributed collection metadata is harvested, users can navigate back to the host server for further metadata This approach offers the potential for enhanced functionality and harvesting capabilities, allowing for innovations like "topic maps" that enable users to create personalized virtual collections and store their results locally.
The next section further explores this model in the context of evaluating different methods of providing union catalog functions and meeting user needs.
Union Catalog Issues
The Sheet Music Consortium, which includes partners such as UCLA, Johns Hopkins, Indiana University, Duke, and the Library of Congress, has successfully aggregated approximately 500,000 digital sheet music pieces in a brief period For more information, visit [UCLA's Digital Sheet Music](http://digital.library.ucla.edu/sheetmusic/).
Reasons for Creating a Union Catalog
Given UC's extensive experience with the Melvyl Union Catalog for bibliographic information, establishing a union catalog for images appears to be a logical progression The objectives and advantages of an image union catalog closely align with those of a bibliographic database, offering enhanced access and organization of visual resources.
Brings together unique but complementary collections within UC.
Exposes our collections for researchers at UC and elsewhere.
Creates an incentive for defining standards and best practices.
Provides a focus for units, especially small ones, to work toward achieving standards and best practices.
Reveals duplication and provides an opportunity to assess how to avoid unneeded duplication.
Provides an opportunity for sharing costs of tools for managing collections, either licensed or homegrown.
Provides an opportunity for sharing costs of specialized tools such as sophisticated presentation and manipulation tools.
Takes a step toward managing persistent access to and preservation of UC’s digital images.
Union Catalog Models and Access Integration
We analyzed various union catalog models to identify their goals, access levels, and potential redundancies in effort Some of our image collections are represented at the collection level in Melvyl and OAC, while a few have item-level representation in OAC Additionally, map collections from UCB and UCSB have item-level visibility in Melvyl, and UCSD’s Visual Resources collection is listed in Roger at the item level, though these records are not included in Melvyl Given the significance of item-level access for most image collections, there currently exists no cohesive system to support this for all image types.
Melvyl already has item level material and could accommodate creating a separate
To effectively tackle potential issues related to varying granularity among users, it is essential to consider the implications of adding more item-level image records However, the necessity of item-level MARC cataloging for all image collections remains debatable, and certain collections, such as Alexandria, may require enhancements to the MARC format.
The majority of image collections are housed in Special Collections, and while the Online Archive of California (OAC) has been expanded to include museums through the MOAC project, it is unlikely to encompass other image types like maps or Visual Resource collections due to the absence of a finding aid equivalent for these types The finding aid concept has proven effective for museums, as it allows for the independent discovery of associated images, which hold individual value User studies indicate a strong interest in “freeing the images” linked to special collections, leading to the implementation of an image search and browsing feature in OAC Although OAC remains beneficial for special collections by providing contextual finding aids, it does not offer a comprehensive solution for collocating all image types.
Luna Insight is a specialized system designed for item-level access to images, making it an attractive option for creating an image union catalog It effectively manages three main categories of images in collections: visual resources, special collections, and maps, although it lacks geospatial capabilities for maps The system allows for federated image searching through cross-collection queries or by integrating images into a single collection, provided the original metadata can be appropriately mapped It also accommodates complex objects, such as multiple views and hierarchical relationships found in artists' books, and is developing features for multimedia formats While these capabilities are crucial for many image collections, it remains uncertain if this system is suitable for all The CDL Image Demonstrator project may highlight key considerations for evaluating this and similar solutions.
Alexandria serves as a versatile tool that enhances search capabilities by integrating geo-referenced data, making it particularly useful for map images Its functionality extends to searching distributed collections and metadata, thereby enriching the process of federating various geo-referenced materials.
A new UC Image OPAC could be developed if basic search and presentation capabilities are adequate, similar to the existing Melvyl model but with enhanced metadata integration Libraries could manage their own data creation, with the option to store images either locally or centrally, akin to the OAC model The CDL could gather metadata via the OAI protocol into a centralized repository and create a fundamental search interface for this data Examples include the image search and browse features in OAC and the public website, both utilizing the same structural foundation A UC version of the public site could serve as a generic discovery service for digital objects, such as images and texts, linking to more advanced presentation and discovery tools However, while this approach is logical, it may not fully meet current needs.
The David Rumsey map collection, accessible through Insight, features a GIS browser developed by Telemorphic, offering distinct advantages that fulfill specific needs not adequately addressed by alternative solutions.
Creating a new Image OPAC alone is insufficient for meeting the diverse research and teaching needs of our users, as our analysis shows that a union image catalog limited to UC collections falls short It is essential to connect our library collections with other image collections in multiple ways In fact, enabling federation of our collections may be equally or more crucial than developing a separate access system solely for library-managed image collections.
The Access Integration Model (AIM) serves as a framework for exploring the federation of image collections, which can be categorized as a Material Type Portal There is a documented demand from arts bibliographers for integration with both commercial and publicly accessible image sources Similarly, medical images are expected to follow this trend, while maps are closely linked to free government resources, and photographic collections can enhance various subject areas The concept of a Subject Access Portal and a Global Access Portal effectively addresses these requirements Therefore, consolidating our collections to facilitate their delivery into these diverse portal environments is a valuable objective.
To enhance the delivery of collections, establishing minimum metadata standards is essential Additionally, enabling harvesting through OAI, despite the need to simplify metadata to align with Dublin Core, provides a low-barrier solution for discovery, particularly beneficial for a Global Access Portal.
A viable solution for accessing diverse images from various systems is to implement an access portal, like SearchLight2, which can query platforms such as Melvyl and Alexandria for maps, or OAC for special collections This portal can effectively harvest and store metadata without requiring a separate user interface By utilizing OAI, campus library collections can be centralized and easily queried through the SearchLight interface, with OAI sets created as necessary for selective harvesting This method facilitates the federation of collections, including both free and licensed sources like the AP Photo Archive.
The Joint Steering Committee on Shared Collections conducted a survey in 2002 focusing on UC Subject Selectors, which includes insights on Visual Resources and Art and Art History For detailed findings, refer to the survey results available at the California Digital Library's website, specifically the Visual Resources report at http://www.cdlib.org/libstaff/sharedcoll/protected/jscsurveys/survey02/visres02.rtf and the Art and Art History report at http://www.cdlib.org/libstaff/sharedcoll/protected/jscsurveys/survey02/art02.rtf.
It is possible to create a subset for searching only UC library images or potentially UC-wide images in the future Users can also search for images alongside other material types relevant to specific subjects, such as books, article databases, and architecture While this approach may compromise the consistent presentation of images, it can effectively achieve the discovery objectives of a union catalog without duplicating data from specialized systems that cater to the unique requirements of each collection type.
Beyond the union catalog
User studies reveal that effective presentation features, such as thumbnail browsing, varying image sizes, and image comparison, are crucial for discovery tools Simply having a discovery tool without these capabilities is inadequate, especially in fields like the arts where advanced features are essential The CDL Image Demonstrator Project exemplifies this approach by developing both a Material Type Portal and a Subject Access Portal tailored for arts images, specifically designed for classroom use Utilizing Luna’s Insight software, the project integrates commercial collections, including Saskia, to enhance accessibility and usability.
The integration of VR slide collections and AMICO museum images with UC museum collections from MOAC, including select Bancroft Library collections, enhances accessibility to a variety of freely available resources from other Insight customers Since MOAC images are encoded in EAD and wrapped in METS, they can be delivered with enriched metadata Many other collections utilize VRA Core, providing a more comprehensive metadata set While it's not essential to deliver all image collections in this manner, converting them into METS or creating them in METS with appropriate metadata schemes would allow for streamlined delivery to specialized tools as a unified input source.
Alexandria is a specialized search system designed for efficient retrieval and manipulation of subject-specific materials It enables users to locate precise images, maps, or documents linked to specific geographic locations, rather than just accessing broader parent records Additionally, Alexandria features a metadata mapping tool that facilitates the integration of both collection-level and item-level metadata, along with geospatial search capabilities utilizing place names from the ADL Gazetteer or specific point coordinates.
Visual resource collections and art collections play a crucial role in educational instruction, whether accessed through specialized presentation systems, learning management systems, or personal presentation methods The ability to capture images from these resources for classroom use is essential, highlighting the necessity for integrated capture tools within and between systems.
36 See ADL Collection Metadata Insert at http://alexandria.ucsb.edu/adl/docs_public/ADLMetadata_ingest/ADL_CLM_insert.php
A significant challenge in union catalogs is enabling users to submit their own images, either temporarily or as permanent contributions This feature, traditionally linked to educational purposes, is becoming increasingly important in the realm of eScholarship, where scholars may want to include images that complement their works and share them more broadly when feasible.
The discovery system must include a tool that enables users to upload their images to a central repository for efficient management and distribution to specialized discovery and delivery tools, as well as learning management systems Both the "capture" and "submit" tools play a crucial role in facilitating the connection between these systems.
In conclusion, it is advisable to focus on delivering images through various methods by establishing metadata standards and harvesting protocols We can utilize a SearchLight2 system to federate UC images with other collections and materials based on subject OAC and Alexandria will continue to offer specialized access to these collections while we assess Luna Insight as a potential tool for federating diverse collections that require unique presentation Additionally, we can centralize metadata harvesting from local collections to enhance delivery across these systems The creation of a dedicated search and presentation system for UC images should be postponed until we gather more insights from these initiatives.
Recommendations
Digital images add complexity to library services, necessitating closer alignment with user habits and collaboration with campus instructional units There isn't a single plan for implementation; instead, a series of simultaneous activities will help clarify uncertainties The recommendations focus on both planning and action, a strategy that has proven effective in tackling challenging initiatives Our extensive collections provide a strong foundation for these efforts, highlighting the importance of taking initial steps to establish a solid base for future development.
The recommendations are organized by their likelihood and logical sequence, with three key recommendations emphasized as essential for implementing the others The accompanying chart indicates which groups could take the lead on these recommendations under SOPAG's guidance Additionally, SOPAG may opt to establish a new steering committee to directly oversee these initiatives.
Needs Assessment
Several extensive user studies are currently underway or planned, making additional studies unnecessary at this time Key studies to monitor include the Pennsylvania State University Visual Image User Study, expected to be completed in early 2004, the CDL Image Demonstrator assessment with a report due in December 2003, ARTstor’s testing in Fall 2003, and a proposed needs assessment by HEAL It is crucial to analyze these studies for insights into user needs regarding functions and support, as well as user expectations and concerns, which may impact library services and collaborations with other professionals Preliminary findings indicate that the distinction between image use for research and instruction may be diminishing as barriers to instructional use are reduced The primary focus of analysis should be on user needs, irrespective of their classification, as library management priorities will likely align with how effectively we address these needs.
Architecture model – key recommendation
Collection managers, digital library managers, and technology staff should evaluate the proposed architecture model for feasibility and suggest improvements, while public services staff must assess its potential to meet user needs A final endorsement from SOPAG and the University Librarians will guide all libraries in making informed decisions regarding standards, systems, and integration options Additionally, the CDL Image Demonstrator offers a valuable opportunity to test various components of this model.
Metadata and digital object standards – key recommendation
The initial goal should be to define minimum standards which work for all communities and intended uses, including user contributions and which support aggregation of our image
Artstor has transitioned to JSTOR, and the previous Artstor page is no longer accessible Collection managers must be informed about the standards for digitizing and creating metadata, which should align with the architecture model The long-term objective is to establish ideal standards and best practices based on the OAC model, recognizing that these may vary across different communities and applications.
Collection/content management
A comprehensive review of existing collection and content management systems is essential to assess their compatibility with the architecture model Libraries are likely to continue using diverse systems, prompting the need to establish minimum requirements for metadata provision While the task force did not investigate broader collection management for various digital assets, understanding the general capabilities on each campus is crucial It remains uncertain if dedicated systems for images are necessary or if they can be integrated with other digital objects Furthermore, the potential benefits of developing a systemwide or shared infrastructure warrant consideration Although outside the task force's scope, SOPAG may benefit from examining current capabilities across campus libraries to explore opportunities for shared systems Focusing on the architecture model and metadata standards could facilitate the identification of common goals for collaborative management systems, with a noted preference for a unified system for specialized image delivery.
Federation of digital collections
Conduct a more thorough analysis of the collections identified in the survey, following up on
“readiness” questions identified in section 2.5 to determine which collections could be federated.
To effectively assess image "readiness" for catalog inclusion, it is essential to establish minimal standards The OAC has already initiated inquiries with collection managers through its MOAC and California Cultures programs, as well as the CDL-appointed Task Force on the Image Service Demonstrator Project By learning from these initiatives, we can enhance future sessions with collection managers regarding visual collections Once the architectural model is in place, we must also consider the library's capacity to provide necessary support, as the OAC and the Image Demonstrator offer various methods for federating collections.
Interaction among existing systems
Once the architectural model is implemented, preliminary experiments can be conducted to explore the integration of systems such as OAC, SearchLight2, Luna Insight, and local systems For instance, CDL could create an OAI harvesting tool to identify and harvest digital collections from libraries that are or could be OAI compliant, adding these resources to the METS repository and/or Insight Additionally, CDL could collaborate with vendors to test innovative solutions for system integration.
“Capture it” tools that could move images from discovery sources to presentation or learning management systems.
Specialized software development
To effectively meet user needs in the development of specialized image presentation software, it is essential to influence the design of these systems Utilizing the model from the CDL Tools and Services Working Group, functional requirements for image delivery systems should be established based on user needs assessments, usability studies, and practical experience The CDL Image Demonstrator assessment presents a valuable opportunity to initiate this process, while the Penn State VIUS may also provide useful insights This approach aligns with Recommendations 7.6 and 7.12.
Federation of UC images with other free or commercial content
Experiments in federating content could happen at different levels The CDL Image
Demonstrator is already exploring this issue by combining MOAC with commercial collections and freely available collections from Insight CDL is also exploring a replacement for
SearchLight which could provide other opportunities (e.g Google images, American Memory,
AP Photo Archive) These effort should analyze which standards and protocols are most successful, including harvesting, web services, Z39.50, etc.
Copyright and fair use policies and guidelines - key recommendation
To enhance the sharing of image resources across UC libraries and ensure public access, it is essential to establish systemwide policies and procedures informed by current practices Collaboration with faculty groups will help identify key policy issues related to faculty-created materials and the use of other digital images Although these challenges are complex, UC libraries have established precedents for resolving them This is a pivotal moment for crafting a unified message that supports both library staff and users in navigating these policies effectively.
Digitizing priorities or guidelines
Task Force members recognized that digitization efforts would vary by campus based on specific priorities and opportunities However, they emphasized the potential for collective strategies that all libraries could explore A thorough review of collections is necessary to identify logical areas for targeted digitization Libraries with digital image reserves can utilize usage data to prioritize digitization efforts aligned with instructional needs Furthermore, making metadata accessible to discovery tools for analog collections lacking digital versions is feasible within the current architecture Utilizing usage data or establishing a user-requested digitization service can also help set priorities effectively.
Cataloging/metadata sources and authorities
The Union Catalog of Art Images project, led by UCSD and funded by the Mellon Foundation, is a significant initiative aimed at testing the feasibility of a cataloging utility for art images If successful, it could enhance the efficiency of cataloging core materials in the art field Additionally, UC's involvement with three members on the advisory committee for the Cataloging Cultural Objects project presents further opportunities to shape standards and guidelines While special collections may still require original cataloging, many could benefit from the practices and authorities established through these collaborative efforts.
The Visual Resources Association (VRA) is a multidisciplinary organization committed to advancing research and education in image management across educational, cultural heritage, and commercial sectors Key members of the advisory committee include Mary Elings from the Bancroft Library, Layna White from the UCLA Armand Hammer Museum of Art, and Jan Eklund from the University of California, Berkeley For more information, visit their official website or contact them via email.
Integration with learning management systems
The central repository in the architecture model must be capable of delivering images to learning or course management software, while commercial presentation software should seamlessly integrate with learning management systems Libraries need to stay informed about the standards for these systems and seek opportunities to engage in experiments that facilitate the delivery of digital objects to enhance educational experiences.
Campus network infrastructure and classroom capabilities
Libraries play a crucial role in shaping campus and national media classroom standards, particularly concerning digital images They should advocate for the significance of these standards to Instructional Technology, Media Services, and Information Technology units Collaborating on a "toolkit" to support this initiative can be beneficial, but each campus library must create its own tailored strategy to effectively convey this message.
Support for faculty to create and use image collections
Current campus models for instructional support are being investigated to assess the potential role of libraries in this area The CDL Image Demonstrator project is evaluating the "personal collections" feature of Luna Insight, while VIUS is also examining this aspect Building partnerships with other campus professionals will be essential for success in this initiative.
Applied research agenda
Current research focuses on two key areas: enhancing advanced search and ranking tools that operate with limited metadata, and integrating content-based search systems with traditional concept-based approaches for practical applications Engaging our research community partners in these discussions could yield valuable insights.
7.1 Needs assessment: Image demonstrator project, Penn
State VIUS, HEAL proposed study
User Needs Research activity; in process for Image Demonstrator
7.2 Architecture model Technical Key recommendation Review, establish policy 7.3 Metadata and digital object standards: Develop minimum, ideal/best practices
Standards Key recommendation Establish policy, promote, experiment
7.4 Collection/content management Technical Need more information, e.g., from LTAG
7.5 Collections ready to federate Content
Standards Depends on partnerships Evaluate
7.6 Interaction among existing systems: e.g., OAI harvesting Technical Need more information; depends on architecture model
7.7 Specialized software development User needs Depends on partnerships; in progress with Image Demonstrator
7.8 Federation with free & commercial content
Depends on partnerships; in progress with replacement for SearchLight
7.9 Copyright and fair use User needs Key recommendation Establish policy, promote7.10 Digitization priorities Content, Need more Establish policy,