UC DLFx 2018: Defining and Sustaining Digital Collection and Scholarship Services

UC DLFx 2018

Defining and Sustaining Digital Collection and Scholarship Services

Zoe Borovsky UCLA), Mary Elings, Erik Mitchell (UCB), Laura Smart (UCI), Carl G. Stahmer (UCD), Stacey Reardon (UCB) 

Dialogic open space: Panelists will introduce. 

Framing questions

What current use cases demonstrate a need for DS?

Who are we missing? Demographics of folks we're serving?

How are digital outputs changing our collection and preservation strategies and what changes do we need to make in the future?

What additional or redeployed resources and labor will be required to provide necessary services?

Are current and imagined services sustainable compared to traditional library services?

https://ds.lib.ucdavis.edu/ucdlf has the questionnaire

No one definition of digital scholarship, but seeing working with different groups on campus, but not sure how working best with that group. Is this question of expertise or infrastructure that we're providing? Different views on best/most sustainable approaches.

They might not be different from each other; expertise in staffing is a form of infrastructure. DH has an external group that does DH and the library collaborates with other units like data lab and center for new media to frame DH and programming for DH. Library is most stable entity in that relationship (UCB), expected to be around awhile. Longevity. Other ways we provide infrastructure for scholarship so it may just be case of adapting to emerging digital scholarship.

UCLA - chosen to foreground expertise than infrastructure through program called DressUp - targeted graduate students and bring expertise form behind the scenes ot the foreground - flipping the library. The expertise is public service is liaison/subject librarians who engage with public. but expertise researchers and scholars need the expertise of the metadata/cataloging librarians. Bringing that to forefront has been mission, to connect people/researchers with metadata and digital library experts. In touch with how researchers use collections instead of going through liaison librarians to get that info. The advantage of grad grad students is that they leave, so need to make them as portable and independent as possible - open science framework, Github. Get them used to looking outside institution to larger collaborations for the support they will need. 

Tension between expertise and infrastructure but like chocolate and PB - go well together. UCI is slow to grow, little department growing (3 people, no one dedicated data curation, have an interim data curation librarian but took time and negotiation for resource reallocation). Can't build infrastructure without expertise to inform it, cant inform without infrastructure. Queen of no - do a lot of referrals to vendors for digitizing because don't handle VHS, or private data; they're like a boomerang. Slow success - motto is we can only do what we can do. Focused on management of hiring and staff, focused on in reach within the library making sure people who will be asked to do this kind of work are well equipped and cognizant that they are expected to do this work. At UCI everyone is a part of digital services. All liaisons do collection, instruction, etc. but when it comes to boots on ground, too busy with other thing, they don't want to be bad guy and tell faculty no. Success with training - adapted 23 things data training as asynchronous so they can do it as they have time--no excuses. For Dh works, use consultation requests to work with liaison librarians one on one - you have faculty member who wants to do X, can't do it unless you also partner and contribute with Y. 

What might entice people to form deep partnership with library and digital scholarship group. DH group at Berkeley entice by offering funding. Berkeley institute for data science provide funding and infrastructure because agreed on common set of tools (Jupiter) which helped university move to large scale. Libraries don't want to serve an individual community deeply at expense of other communities. What does it mean to serve and exclude others. No habit of providing funding to do work. Consultations and expertise provision was helpful because it was hard to predict what infrastructure is necessary. Consultation, expertise, are we caring about providing access to completed resources, what about digital scholarship?

Participating in collaborate research grants with Dh funded projects - library provides time and expertise, DH provides funds. Trying to figure out what do we want to collect, what can we collect, what can we maintain over time? Many projects quickly done using various tools, so long term is murky. Lots of expertise to train undergrads with DH group and faculty working with digital collections. Is it a departmental responsibility or library responsibility for that training? 

There was discussing about individual collections. Problematize: one case - when there is an infrastructure to support, then collection walk and end up somewhere else. try to make it an "also will help you to" the researchers get primarily involved with people in infrastructure to produce the research, hard to reel them back in to the Library. Figuring out a better way to mitigate that problem is crucial to our success. 

A cycle - once they walk away and work with it for ten years and lose funding, then they want to give it to the Library. 

Tried to shift notion of the project away from product and onto processes. if we look at the fact that the library should retain the process and workflow researchers are creating, is more bound in workflow as opposed to keeping final project and product. Grad students - we know they will leave and they don't want to invest in something the Library will have ot maintain for them. Having library be stewards is more productive in terms of measuring impact. Sharing more workflows and processes than just keeping product alive.

Faculty also doing digital projects and that  might mean they sunset earlier than desired if library can't provide infrastructure. Researchers want to gift to library to preserve in some way - do we have the bandwidth, do we know how to go about preserving project? Shakespeare project - campus film studio recorded on VHS of students doing Sjhakes[eare, licensed, went through distributor, used in courses throughout country. To get online, built Drupal site, which wasn't hosted by campus, left it, then had to sink funds to revitalize it. Now library is digitizing VS tapes and out in catalog to guarantee preservation. but what happens to website as it continues to deprecate? At some point project would sunset - faculty member terrified and retired so can't steward as was done in the past. Faculty desperately want to have their projects live on but also validation of scholarly content. Ended up web archiving the site before it went any further away. What is best way to preserve sites they want to give? Like monographs, they build research product and want us to take them on.

Web archiving platform - generic platform that a little expertise was used to develop documentation. Work done before moving to next Drupal, archived it using ArchiveIt. came up that researcher had lost feedback from other scholars, was able to pull it up. It works but need to scale it and publicize. 

Metrics for success. How scalable for digital scholarship have to be for us to consider successful effort for the library? Count number of grad students touched in a quarter? What is evaluation matrix? Can we come up with one now, are we even ready for that or nowhere near and live in morass of thinking that we're good because we're busy and working on interesting things? Right now measures, because no bar on which to compare other than qualitative feedback from community members. Not a good enough measure, but how do we come up with others? What would be the measures we could be shooting for vs our subject liaisons. 

Our research IT group did an impact IT study to see how they were impacting research with their services, insert library and it could have applied to libraries as well. How do you measure impact is the real unknown. 

Really difficult problem. During job interview, asked "what would constitute success?" - we're nationally known for Something. Heh. Think back to traditional library school for measures and needs analysis--difficult to get people to go our and audit what is happening local environment. Where is there a gap, and did we fill it?

Have to measure the community of practice that you're generating. Difficult to assess because used to info lit and how many seats you fill in a workshop. Need to measure depth and quality of service instead of seats. Assessment should be done from librarian - how much did you learn? How did this impact your practice. What's the value to the Library? Assess internal instead of constant focus on outside. We measure success from amount of  grant proposals we bring in. Value of Dh was in its status of cash cow - best thing to see is the move into the library in that we can support tools and methods beyond DH curriculum and projects that get funding. Idea of trickle down and if engaged in high level projects those tools and practices would trickle down to students on campus, but that didn't happen. Here's a gap and a place librarians should be. take tools and practices from grant proposals and take to students and researchers who cant attract those grants and have access to that. Should be library's mission and not just measured by dollars generated.

Q: Do you feel programs or successful, or success along the way or feel like could be doing more?

A: UCB doesn't have a program as multiple threads of direction and effort. Stacey is leading digital scholarship working group for librarians to talk about librarians' projects. Exhibits on GIS and maps as outreach. Insane amount of time spent on objects but little trickle down. UCB only in control of very small number of things. Group of librarians wrote compelling letter to leadership laying gout this issue - stumped. 

A: Not sure more I can do - doing too much. As an institution we could invest more in needed resources to scale up and provide services to more types of people. Current focus on graduate students, but UCB could focus more on undergrads. This summer: DH minor over two successive summers. Need faculty with expertise in this area, new budget lines, support structure. best place to support undergrads is the library. Tinkering with soft launches of pilot projects - data and digital research support - everyone is doing something in digital scholarship and we can leverage combined expertise of librarians. Also undergrad library fellows with specializations and then share out what they learn with peer. Library also contributing collections to those efforts. Contribute in as many ways as we can. Successful? Depends on the day. Overall yes, we can't expect success to be instantaneous and will go far more slowly than we'd expect or like. They've become a full department, They've added staff. Training for librarians to feel comfortable doing initial consultations. Library Carpentry as option to help folks understand. Need a thick skin to be in digital scholarship, keep eye on very long term prize. Think back to first job in 1996 at San Diego Computing Supercenter - model of San Diego oil spills - map - each pixel represented a piece of data, Maintained bibliography of everything related--wouldn't be awesome to tie together as a complex object and wouldn't it be cool where people could se and change for their own use?

Q: from Panel: How do you gauge success? 

Q: People come with interesting ideas but no experience with coding or infrastructure and no money or support but they want us to help build it. Challenge is how do we build some of these things if it takes time to build the infrastructure? But person we want to help is also on a timeline. Feels like putting on a play - need to make it work, so pressure to build a thing over time and does it work?

A: We struggle bc loose group at Berkeley with expertise but no time to be committed to someone else's project. There's a formal group that shares expertise, but it's just advising. Expertise vs infrastructure issue. 

We try to focus on prototyping, we don't do projects for them, so we bring in and train grad students or teach faculty themselves; view ourselves as instructors rather than doing it. We spend most of our time helping people scope. Early career researchers we can help them plan things out in phases. 

UCD does sling code. Rolled out services like exploratory data mining and roll them out to standardized report - they have to do megacleaning but will advise on that, then library runs it for researchers to see fi there's a 'there' there. Useful land likely scalable, run in same R library. Scalable and reproducible so we can dee

Research data - not identified as space library wants to serve, not central to library mission. UCB is to form consultation network so don't take ownership over that. Web archiving is a good choice, but IT deals with storage more challenging fit due to cost. Does library have role in persisting disk storage or preservation storage for research projects? [Silence]

Instruction is key which is always a service we've provided, this is just new topics. Scholarly resource procurement and showing folks how to get it. Are we doing hands-on help they need? More of a maybe. Collecting digital projects for the collection? Maybe. Assessment question is one we can start to work on - how to measure, how to show we are being successful?

Provide temp and long term storage - Cumulo solid state drive for temp storage, library doesn't want them storing in places that can be lost. Then also work to get data into the DAMS so it has metadata and object modeling expertise librarians provide. DAMS is them preserved in Kronopolis (sp?). Some of these research products are akin to these and dissertations, not a question of should we preserve them (ye, they're the products of our research enterprise and knowledge production we have as universities  - we would never argue against preserving theses, dissertations, monographs!). Maybe way we're looking at it wrong, maybe we should look at it as content that is software dependent (web projects dependent on the browser, which makes it hard to access from older browsers). Emulation could be scalable could be preserving without having to upgrade them all the time.

Comparison to these and dissertations comforting. UCB digitizes through Google projects. People who do textile degrees insert samples, etc. So interesting to see info container of monograph and finding ways to include nontraditional material. How to make an emulation or capsule to capture that - squarely in library's preservation mission.