repositories for the rest of us: DigCCurr 2007: Concurrent session: Building Capabilities for Digital Curation

2007/04/20: DigCCurr 2007: Concurrent session: Building Capabilities for Digital Curation

I attended "Defining Capabilities" Speakers: Liz Bishoff, Nancy McGovern, Oya Rieger

Liz Bishoff: Digital Preservation Assessment: Readying Cultural Heritage Institutions for Digital Preservation

Benchmark paper 1996 "Preserving Digital Information" by ARL defined the issues and possible solutions to digital preservation. Cultural heritage includes scientific as well as arts.

The number of projects doing digital preservation is minuscule compared to the number of cultural heritage institutions that are digitizing, or having born-digital items. The publishing of the ARL white paper indicated that digitization can be preservation. More funding agencies have preservation solutions in their requirements. There are emerging state, national, and international initiatives for digital preservation. Are all these projects ready for managed digital preservation?

Bishoff surveyed institutions on their readiness. Also Anne Kenney at Cornell studied institutions participating in NEH funding 2003-5. Cornell study results avail RLG diginews. Cornell found that 90% still using CD/DVD for digital storage. That number is now about 70%. Only 50% had policies but only 30% have implemented those policies.

2005 NEDCC Survey - 66% of institutions had no one responsible for digital preservation, paralleled Kenney's findings. Also a fair number of institutions indicated that they had backed up only once (or not at all!!!)

Findings from Bishoff's survey: Issue of digital preservation is just now coming to the forefront of discussion and action. Many institutions are still at the the project stage and have not yet gotten to the ongoing program stage. Written policies and documented digital preservations practices are lacking. Preservation/conservation staff are generally NOT directly involved in many of the digital initiatives.

They also found: Few have coordinated institutional approach to their digital initiative especially in the areas of standards (imaging, metadata), quality control, access, promotion, digital preservation. A big lack of understanding of when institution has a "born digital" material. CD/DVD is the major storage media but moving to networked servers. Refreshing data on CD/DVD with lengthy periods between refreshing. Quality control of master images is inconsistent at best. Education is important before doing a digital preservation project. Ability to advocate for digital preservation is lacking at many institutions. Funding is primarily through local funds and grants.

Areas of policy which support digital content: mission and goals, collection development, emergency preparedness, exhibitions, preservation, strategic planning, public services, rights and licensing.

If the institution was out sourcing, do they follow the elements of a trusted digital repository (TRACK?). Financial viability of the company you choose is very important.

Types of recommendations: improved documentation (continuity planning, work flow processes, etc.), review digital preservation activities including refreshing schedules, quality control, etc., review system back-up procedures and implement off site storage.

So what does it mean? focus of long term preservation has been on the technology and standards, certification, etc. to build the infrastructure. To make it reality we now need to
*expand advocates for long term preservation
*expand the knowledge base of practitioners
*move from digital project to digital program
*integrate preservation into all aspects of digital life-cycle
*develop best practices
*make policy examples available

Education needs to be moved to the state and regional level. Also we need both professional AND continuing education. That needs to focus on technology and standards, policies and tactical strategies (development and implementation), work flow and documentation, business planning and all that it involves such as market research, financial analysis and planning.

Conclusion: progress is being made, need to increase awareness of importance, most institutions which are doing digitization, however, are not doing the the basic preservation activities.

Question - what is the definition of digital preservation? Response from ALCTS Preservation co-chair -- they are sending out a definition to various email lists in the next couple of weeks.

Nancy McGovern: Canary in a Coal Mine: A digital preservation response to technological change

How to deal with open-ended change? We've lacked specificity and scope about how to respond to that change. How to we detect things that might have adverse implication for digital preservation. How do you go about doing the assessment?

Outline: technology response requirements, common response, scope of interest, priorities for digital preservation, timing response to technology.

Technology response: the call for responding came in the 1996 seminal paper. The specification is most explicit in OAIS.

OAIS monitor technology: objective: track emerging technologies, information standards, computing, platforms. purpose: avoid obsolescence.

Examples of technology watch: DPC, DCC, DigiCult, LITA, PRONOM

Characteristics: range in services provided by technology watch services reflects absence of definition. Providers select topics not community, lack access to accumulated data, defined levels of service is rare (detailed synopsis? headline?).

Community formalization: Digital preservation for museums CHIN 2004 service requirements, LIFE project UCL/BL, 2006. Strategic priorities of SAA 2006-7 calls for leadership and training on how to respond to technological change.

Scope of interest: macro taxonomy
Object - file formates, media metadata
Collection - relationships, metadata
Repository - software, tools, modules
Platform - protocols, security, software, hardware

Scope: micro taxonomy
35 technology types enable OAIS
Examples: communication (ability to convey message), logs, policy enforcement

Priorities for digital preservation: Contact, interaction, exploitation, risk management, automation.
Contact: requires direct contact with digital content
Interaction: must respond to, not just be made aware of, changes in digital content
Exploitation: potential to contribute to digital preservation strategies by exploiting opportunities
Risk management: participates in the avoidance of risks to integrity, longevity, or authenticity
Automation: potential to perform more effectively

Timing response is important
Identify potential new technology, monitor, assess, respond, act to avoid obsolescence of existing technologies.

Technology responsiveness - Community objectives:
*accumulate current and historical information
*develop competencies and tools
*incorporate community developments
*build a network of contributors and users
*ensure sustainability

Question: is there a need to have some kind of peer review process/consensus building on deciding whether a technology matters or not or how to go about implementing collaborative technology watch?
Answer: Assessment is key. Organizations should be able to pick the right size thing for them which fits their requirements. People must be able to picture themselves in the results.

Digital preservation is research and development although we do it in a production environment we need to keep questioning and monitoring and assessing. William Gibson "The future is already here, it's just not evenly distributed yet."

Oya Reiger: Select for Success: Key principles in assessing Repository Models

Within a life-cycle framework, digital curation involves a series of technical, intellectual, and managerial activities in support of stewardship for digitized or born-digital information assets.

What is a repository system? A system to capture, store, index, manage, preserve, and deliver digital objects.

Factors in choosing a repository model
*development characteristics
*financial sustainability
*digital library infrastructure
*interoperability and support for standards
*institutional policies and practices
*support of archival business requirements
*content type characteristics
*preservation functionality
*usability (staff and end-user)
*search, browse, access features

See Art Libraries Society of North America "Digital Image Database Standards Checklist"

RLG/OCLC " Trustworthy repository certification"

Key principles in selecting a repository
1. Identify key stakeholders (users, programmers, subject experts, etc.). It builds awareness and trust, gathers feedback, build trust, get support, expand resources, understand risks
2. Conduct needs assessment to characterize your environment. Include documents (document type, condition, metadata attributes, selection criteria, usage restrictions, relation to other collections), users, and resources (available staff, money)
3. Explore resource requirements. Institutional repositories SPEC KIT- start up range from $8,00-$1,800,00 (mean=$182,550) and an average ongoing operating cost of $113, 500 . There are many hidden costs. No common metrics yet to determine what information points to include
4. Understanding the existing and evolving human landscape. Work culture and practices, relevant social groups, interpretive flexibility, appropriation (how technology fits into the workplace and how it supports your culture).

Quoted Judson King's report from Institution for Studies in Higher Education, 2006 "Scholarly communication: academic values and sustainable models" -- "Approaches that try to move faculty and their deeply embedded value systems directly toward new forms of archival systems are destined to fail"

Conclusions:
*flexible and scalable repositories - Choudary and Martino 2005 "At Johns Hopkins, we are promoting the idea that applications should access repositories through an abstract, repository agnostic layer, rather than through custom application to repository integrations" see Cornell, as pioneers they ended up with many different repository systems (Greenstone,
*web services/service oriented architecture models - ex. file format migration, file obsolescence service, social tagging, citation analysis, text annotation, plagiarism detection (added to arxive recently)
*repurposing - ex. Cornell making their digitized books available via Amazon print-on-demand
*new information chain see Van de Soemple DLIB article on how the information chain is expanding

Labels: DigCCurr 2007