UC DLFx 2018: Combined Session E Notes
Launching the Digital Lifecycle Program at UCB
Lynne Grigsby (Head Library IT) and Eric Mitchell, AUL/DCS
Major strategic initiative through Library's strategic planning process. Last year and a half, developing current digitization strategy.
Digital Lifecycle Program Mission and Goals
To convert and publish on a massive scale the UCB collections. (To digitize content and shepherd all digitally converted assets of the library through their lifecycle to service the mission of the library.
Goal 1: Create and manage digitized assets
Goal 2: Have a formal, ongoing, and sustainable digitization program that focuses on the entire digital lifecycle
Goal 3: Provide effective access and widespread dissemination of content
How do you effect this change from ad hoc project based digitization to high throughput.
Convert:
- Respect and identify multiple digitization streams and make sure they don't conflict (wax cylinder digitization, vendor sourced 2D material digitization, AV content conversion, born digital content conversion, digital content QC)
- How to get form print object to digital published object quickly.
New Digital Asset Management System
- metadata enhancement and validation, publishing QC, content preservation
Critical path of Digital conversion - Publishing
- is metadata a required element, and to what extent? Metadata is a critical component, but how critical for different items?
- dissemination through UCB branded portal
Enable New Forms of Research
What is the DAMS?
Goal is for everything to go into DAMS to manage all digital files, but users to go through a website. Others have to have metadata and links to files
Infrastructure is incredibly critical but hard to sell that you have to stop and do infrastructure before can start with digitization projects.
Failed recruitment of a DLP manager
Prioritize leveraging in house scanning a lot higher
Launch of advisory group faster because of recruitment failing
Impact for IT
- more work
- digital collections website for UCB (have always sent to aggregators like Calisphere)
- DAMS - review and implementation (DMAS have many moving parts) - current is homegrown plus Sofia plus others
- Gathering everything - where is it? What is it?
- Storage - have 70TB of storage used for digital assets, constantly migrating. Only 70 bc biggest users of Merritt and delete own copy.
- producing more (90,000 images in January), but need to move more. How? Want to automate
- Submit book material to Hathi
- Designing new site - UI, UX,
- Need to answer questions: how to include forensic data (disk images), what goes into the DC? How to get things to preservation. finding restricted data, how do we preserve that, what are the options? What goes into the digital collection site? individual items? Collections?
Working on figuring out how to use the right vendors to get things done - digitization of microfilming from UCR rlf and submission to Hathi - Berkeley throughput.
Next thing: how do we get things in the queue? Books are easy because most are cataloged with metadata and barcodes as unique identifiers. But aerial photos - some have metadata some don't, some only have a catalog record.
DAMS: Most attractive is Tind (vendor hosted, Amazon?). Sensitive to time and investment in setting up a DAMS especially given the homegrown. IT is looking for a 3-5 year solution, can't think about 25 year solution. If we can find something to consolidate in 3-5 years, then we can at least export.
Q: How do you prioritize?
A: UL says don't prioritize, go from shelf 1 and do everything. Advisory group will help conduct library wide audit of collections that would be candidates. Does it have metadata, are the rights cleared, etc. It's currently ad hoc, librarian brings a collection they care about to Lynn. Because of the metadata bar, not that much is ready at that level aside from books which are easier.
Q: To what standard to we hold the metadata?
A: Is it at finding aid level, box level, photograph level, folders of content. Partly is what we need to define. The curators and Bancroft will have a higher bar. Also, get a lot of stuff in spreadsheets. Doing a lobby card collection with minimal metadata but enough that it's priority for a particular library. One of the models we've been exploring is as we higher new catalogers, some of work supports digital lifecycle program. 50% traditional cataloging, 50% metadata management for these digital projects.
Q: What's the mechanism for measuring usage of these collections?
A: We've hired a new assessment librarian with key role to determine that. We're not just interested in numbers of download. Recently we have worked out an agreement with a communications team and keep informed as we push out parts of digital projects. Publishing stories around substantial uses of the collections.
Library Carpentry: Data and Software Skills Training for Librarians
Tim Dennis UCLA
Used to part of social science - building data services out of the data archive. Data and software and library software instructor, also an instructor trainer.
What is library carpentry? Metaphor. Late 90s software carpentry: to teach early career scientists at national labs how to code, because needed to write software to do their research. 2011 - 2 day bootcamp teach R Python with Bash scripting and Git. Funding achieved. Separate entities - data carpentry is data analytics management and visualization. Software carpentry vs data carpentry, both address skill gaps of researchers. (Data carpentry is domain intensive)
Ecosystem
software-carpentry.org/blog/2016/03/venn-diagram.html
Code of Conduct for Software and Data carpentry
Building community. Difficult to learn and can be scary so warm, welcoming environment
It scales: 22,000 learners in 35 countries.
The Carpentries
Communities around lessons. Through GitHub and tools, collaborate, iterate, improve over time.
Library Carpentry - by James Baker, archivist at British. Helps librarians to:
automate repetitive tasks
work effectively with iT and systems colleagues - developing common vocabulary between Library, IT, researchers
create, maintain, and analyze sustainable and reusable data
better understand and support the use of data/software in research
more: continuing communities through meet ups after the actual sessions
Little Women as a Bash exercise. Use authentic experiences. Please contribute datasets. Goal: better align skills to what is happening in research.
Library carpentry:
By librarians for librarians
Relies heavily on pedagogy, structures, norms of the carpentries
welcoming and safe environment
Collaborative environment for building and improving training modules
Core: about volunteer network--all the teachers are volunteers, small staff is through grant funding and partnership. You can buy into membership, you get seats in a cohort of instructors.
They teach the tools, not monolithic pieces of software. Small pieces loosely joined (Unix). Has impact for reproducible research and open philosophies.
Lesson development: James started it in 2015
- with four other collaborators.
- Lessons on Data intro, Shell (how to compose commands), Git, OpenRefine. Do count of names in a book to show how fast Shell is and basics of computing.
- Designed to be a one-off over 4 weeks. Someone in Australia started a community there, hooked up with Mozilla science groups and they did global sprints, so they wanted to develop lessons. Added Python, SQL, web scraping, Gitter group chat
2 day sprint event (like a hackathon) to create new lessons, refine existing ones. 16 sites (NZ, Australia, UK, US, etc.) Over 800 contributions to lessons last year. Teach, get feedback, discuss, revise lesson.
UC3 - want ot build capacity and have librarians who have been certified (pedagogy)
Workshops - 44 total since 2015
Bottleneck: need to hire a coordinator
30 cities in 10 countries
Next; Challenges
- Need infrastructure and staff to grow and respond to burgeoning demand
- Need a steering community that can engage with funders
- Need ot establish ground rules about what kind of community we are
- Need a formal assessment program (current one is ad-hoc)
- Need more library professionals to join us as teachers and lesson contributors
Recent
- IMLS grant won by CDL's UC3 to fund Library Carpentry coordinator role
- interim committee to...
Chat web twitter: Lib carpentry
Q: Available online?
A: Open and available on GitHub, can use as self-directed
Q: Unanticipated benefit - lowers barrier to entry for professional development.
A: Experts of these tools are not usually best teachers, it's people 3 weeks ahead of the class.
Lynne Grigsby (Head Library IT) and Eric Mitchell, AUL/DCS
Major strategic initiative through Library's strategic planning process. Last year and a half, developing current digitization strategy.
Digital Lifecycle Program Mission and Goals
To convert and publish on a massive scale the UCB collections. (To digitize content and shepherd all digitally converted assets of the library through their lifecycle to service the mission of the library.
Goal 1: Create and manage digitized assets
Goal 2: Have a formal, ongoing, and sustainable digitization program that focuses on the entire digital lifecycle
Goal 3: Provide effective access and widespread dissemination of content
How do you effect this change from ad hoc project based digitization to high throughput.
Convert:
- Respect and identify multiple digitization streams and make sure they don't conflict (wax cylinder digitization, vendor sourced 2D material digitization, AV content conversion, born digital content conversion, digital content QC)
- How to get form print object to digital published object quickly.
New Digital Asset Management System
- metadata enhancement and validation, publishing QC, content preservation
Critical path of Digital conversion - Publishing
- is metadata a required element, and to what extent? Metadata is a critical component, but how critical for different items?
- dissemination through UCB branded portal
Enable New Forms of Research
What is the DAMS?
Goal is for everything to go into DAMS to manage all digital files, but users to go through a website. Others have to have metadata and links to files
Infrastructure is incredibly critical but hard to sell that you have to stop and do infrastructure before can start with digitization projects.
Failed recruitment of a DLP manager
Prioritize leveraging in house scanning a lot higher
Launch of advisory group faster because of recruitment failing
Impact for IT
- more work
- digital collections website for UCB (have always sent to aggregators like Calisphere)
- DAMS - review and implementation (DMAS have many moving parts) - current is homegrown plus Sofia plus others
- Gathering everything - where is it? What is it?
- Storage - have 70TB of storage used for digital assets, constantly migrating. Only 70 bc biggest users of Merritt and delete own copy.
- producing more (90,000 images in January), but need to move more. How? Want to automate
- Submit book material to Hathi
- Designing new site - UI, UX,
- Need to answer questions: how to include forensic data (disk images), what goes into the DC? How to get things to preservation. finding restricted data, how do we preserve that, what are the options? What goes into the digital collection site? individual items? Collections?
Working on figuring out how to use the right vendors to get things done - digitization of microfilming from UCR rlf and submission to Hathi - Berkeley throughput.
Next thing: how do we get things in the queue? Books are easy because most are cataloged with metadata and barcodes as unique identifiers. But aerial photos - some have metadata some don't, some only have a catalog record.
DAMS: Most attractive is Tind (vendor hosted, Amazon?). Sensitive to time and investment in setting up a DAMS especially given the homegrown. IT is looking for a 3-5 year solution, can't think about 25 year solution. If we can find something to consolidate in 3-5 years, then we can at least export.
Q: How do you prioritize?
A: UL says don't prioritize, go from shelf 1 and do everything. Advisory group will help conduct library wide audit of collections that would be candidates. Does it have metadata, are the rights cleared, etc. It's currently ad hoc, librarian brings a collection they care about to Lynn. Because of the metadata bar, not that much is ready at that level aside from books which are easier.
Q: To what standard to we hold the metadata?
A: Is it at finding aid level, box level, photograph level, folders of content. Partly is what we need to define. The curators and Bancroft will have a higher bar. Also, get a lot of stuff in spreadsheets. Doing a lobby card collection with minimal metadata but enough that it's priority for a particular library. One of the models we've been exploring is as we higher new catalogers, some of work supports digital lifecycle program. 50% traditional cataloging, 50% metadata management for these digital projects.
Q: What's the mechanism for measuring usage of these collections?
A: We've hired a new assessment librarian with key role to determine that. We're not just interested in numbers of download. Recently we have worked out an agreement with a communications team and keep informed as we push out parts of digital projects. Publishing stories around substantial uses of the collections.
Library Carpentry: Data and Software Skills Training for Librarians
Tim Dennis UCLA
Used to part of social science - building data services out of the data archive. Data and software and library software instructor, also an instructor trainer.
What is library carpentry? Metaphor. Late 90s software carpentry: to teach early career scientists at national labs how to code, because needed to write software to do their research. 2011 - 2 day bootcamp teach R Python with Bash scripting and Git. Funding achieved. Separate entities - data carpentry is data analytics management and visualization. Software carpentry vs data carpentry, both address skill gaps of researchers. (Data carpentry is domain intensive)
Ecosystem
software-carpentry.org/blog/2016/03/venn-diagram.html
Code of Conduct for Software and Data carpentry
Building community. Difficult to learn and can be scary so warm, welcoming environment
It scales: 22,000 learners in 35 countries.
The Carpentries
Communities around lessons. Through GitHub and tools, collaborate, iterate, improve over time.
Library Carpentry - by James Baker, archivist at British. Helps librarians to:
automate repetitive tasks
work effectively with iT and systems colleagues - developing common vocabulary between Library, IT, researchers
create, maintain, and analyze sustainable and reusable data
better understand and support the use of data/software in research
more: continuing communities through meet ups after the actual sessions
Little Women as a Bash exercise. Use authentic experiences. Please contribute datasets. Goal: better align skills to what is happening in research.
Library carpentry:
By librarians for librarians
Relies heavily on pedagogy, structures, norms of the carpentries
welcoming and safe environment
Collaborative environment for building and improving training modules
Core: about volunteer network--all the teachers are volunteers, small staff is through grant funding and partnership. You can buy into membership, you get seats in a cohort of instructors.
They teach the tools, not monolithic pieces of software. Small pieces loosely joined (Unix). Has impact for reproducible research and open philosophies.
Lesson development: James started it in 2015
- with four other collaborators.
- Lessons on Data intro, Shell (how to compose commands), Git, OpenRefine. Do count of names in a book to show how fast Shell is and basics of computing.
- Designed to be a one-off over 4 weeks. Someone in Australia started a community there, hooked up with Mozilla science groups and they did global sprints, so they wanted to develop lessons. Added Python, SQL, web scraping, Gitter group chat
2 day sprint event (like a hackathon) to create new lessons, refine existing ones. 16 sites (NZ, Australia, UK, US, etc.) Over 800 contributions to lessons last year. Teach, get feedback, discuss, revise lesson.
UC3 - want ot build capacity and have librarians who have been certified (pedagogy)
Workshops - 44 total since 2015
Bottleneck: need to hire a coordinator
30 cities in 10 countries
Next; Challenges
- Need infrastructure and staff to grow and respond to burgeoning demand
- Need a steering community that can engage with funders
- Need ot establish ground rules about what kind of community we are
- Need a formal assessment program (current one is ad-hoc)
- Need more library professionals to join us as teachers and lesson contributors
Recent
- IMLS grant won by CDL's UC3 to fund Library Carpentry coordinator role
- interim committee to...
Chat web twitter: Lib carpentry
Q: Available online?
A: Open and available on GitHub, can use as self-directed
Q: Unanticipated benefit - lowers barrier to entry for professional development.
A: Experts of these tools are not usually best teachers, it's people 3 weeks ahead of the class.
Comments