Author Archives: Abigail Shelton

IIIF at Notre Dame, or, the heart of MARBLE

Posted on by

By Abigail Shelton and Jon Hartzler 

What if you could compare a Rembrandt van Rijn print of David and Goliath held at the Snite Museum of Art in South Bend, IN, with the same print at the National Gallery of Art in Washington, D.C., online, at any time of day? 

Or, zoom in so close that you could see the paint texture on this Vincent Van Gogh self-portrait hanging across the country at the Harvard Art Museum? 

The answer is that you can!  With the help of an open-source image sharing system called IIIF or the International Image Interoperability Framework. 

This set of specifications, pronounced triple-eye-eff, provides a standardized way of storing and displaying images. This allows institutions using IIIF to easily share images. IIIF also has a suite of enhanced features that offers end-users a powerful research experience. 

IIIF is at the heart of the Museums, Archives, Rare Books and Library Exploration (MARBLE) platform in development at the University of Notre Dame.

What is IIIF? 

If you are new to IIIF, a helpful way to think about it might be the analogy of a travel adapter. A recent blog post written by museum developers describes IIIF technology in this way:

  • Picture your preparations for an international trip. 
  • You gather all of your travel essentials and realize that you don’t know anything about the power outlets in your destination country. 
  • So, you look it up and find that they use the same outlet as you do at home. 
  • What a relief! No need to be concerned with bulky adaptors. 
  • You just know it will work—”it’s one less thing to worry about,” as stated in the aforementioned post.

Similarly, when libraries and museums use IIIF compatible images, it’s easier for developers and audiences to work with these images, both locally and across institutions. Developers don’t need to build custom “adaptors” because IIIF standardizes the way images are delivered and displayed. Users have a seamless viewing experience. Every time.

And now for a more technical explanation. The IIIF specification defines two application programming interfaces (APIs); one for image delivery and one for presentation. An API is a software program that allows information to transfer between applications. 

The image delivery API provides a standardized way for web applications to pull images from where they are stored (server) to where they are displayed (website). This makes it easy to retrieve images in different sizes (thumbnail vs. full-size), different orientations (upside down), and different color scales (greyscale, black and white, color, etc.). 

The presentation API provides a structure for how images are displayed online. It sets rules for how images are ordered (e.g. pages in book), and what information is presented to the user about the images (e.g. descriptive metadata). The presentation API is what allows developers to display images in IIIF compatible viewers (like Universal Viewer or Mirador) on a webpage.  

The combination of these two API standards means that images can be easily used and reused in predictable and reliable ways. Developers no longer need to build highly customized applications for a diverse array of image formats. Or, in the travel analogy, no need for bulky adaptors for every new destination!  

How is the University of Notre Dame using IIIF?

IIIF is at the heart of the MARBLE project. Our IIIF server is the primary way that images are presented to users through the website. The process works a little like this: 

  • Library and museum colleagues deposit their images in a Google Drive directory, along with metadata that describes what the image is and how it should be displayed. 
  • Once the deposit is complete, our open-source IIIF Pipeline automatically detects the new images and information and creates a IIIF compatible document (IIIF manifest). This document, full of standardized image metadata, allows us to use the same image in multiple ways, across multiple platforms, without duplicating our efforts. 

For example, the IIIF document for Paul’s Wood’s Absolution Under Fire could be searchable in the MARBLE site, highlighted in a digital exhibit on the artist at Notre Dame, and featured in a specialized collection website highlighting all of the Civil War materials on campus.  With IIIF, this can all happen at the same time, without the need to re-enter any metadata or copy any images.

Priest ministering to soldiers on battlefield

Paul Henry Wood (American, 1872-1892), Absolution Under Fire, 1891, oil on canvas. Gift of the artist, 1976.057. Snite Museum of Art.

We’ve also embedded a IIIF image viewer called Mirador in the website to empower users to examine objects closely and compare them side-by-side. Ultimately, we’d also like to enable users to annotate images in the viewer and then save, search, and share those annotations for personal research or classroom uses. 

Several of our staff are also involved in the IIIF international community, helping to spread the word and develop new features for this open-source software. IIIF is an open-source application meaning that there is a group of volunteers from institutions around the world contributing to its development. Notre Dame colleagues have attended the annual conference, regional working group meetings, and participate in regular community conference calls. Our involvement in the IIIF community means that we’re learning from our peers and bringing ideas back to campus for how to showcase our collections in creative ways.

What can I do with IIIF as a student, educator, scholar, practitioner, ________?

So glad you asked! When the MARBLE portal goes live in 2021, you’ll be able to use Notre Dame’s unique collections in a variety of ways, including: 

close up of family photo album

Detail of Le Rossignol Family Photo Album, Rare Books and Special Collections Department, Hesburgh Libraries, University of Notre Dame.

  • You’ll be able to examine artworks and textual materials far closer than you could ever safely do in person. 
  • As an educator, you could project these images in your classroom to help students see details they may have missed during their museum or library visits. 
  • Students will be able to zoom into images while working on course assignments outside of open hours for the archives or museum. 
  • You will also be able to display multiple images side-by-side, either from the Notre Dame collections or from any other institution with IIIF compliant collections. This could allow you to compare details of multiple paintings by the same artist or read different versions of a text together. 
  • Using the same feature, you could digitally reunite collections of objects or even pages from libraries or museums around the world.
a photo album next to a handwritten letter

Le Rossignol Family Photo Album (leftside) next to a Letter to Ethel Le Rossignol (rightside) in Mirador viewer. Both from Rare Books and Special Collections Department, Hesburgh Libraries, University of Notre Dame.

Here’s an example project where four European libraries combined their early Qurʾānic fragments to recreate original codices using IIIF. 

While IIIF opens up a myriad of possibilities for international collaboration, it will also allow us to reunite collections on our campus. For instance, the Snite Museum of Art’s current exhibition of Irish art features a number of broadsides, prints, and volumes from the Rare Books and Special Collections department alongside paintings and photographs from the museum collection. 

The only way for audiences to see these objects together is to see the exhibition in person before it closes in December (strongly encouraged!). But what if these connections could be preserved? By making these materials available online in a IIIF compatible way, curators can maintain and highlight relationships between collections beyond the close of a special exhibition.  

IIIF also powers fun! From art puzzles to Google Chrome plug-ins to mindfulness apps, the possibilities are endless. We’re excited to see how our users experiment with IIIF and create new ways of seeing and experiencing Notre Dame’s unique collections. 

 

Teamwork Makes the Dream Work

Posted on by

By Abigail Shelton 

At universities, there is often the desire for project partnerships, but differing organizational priorities sometimes inhibit true, long-term collaboration. The question is, how can two independent campus departments work effectively together toward sustainable outcomes?

At Notre Dame, the leadership for Hesburgh Libraries and the Snite Museum of Art started by identifying a project with shared goals that aligned with institutional priorities. 

Securing funding was the next step. The partners jointly submitted a grant proposal for a three-year project, the success of which depended on multiple stakeholders working together to develop collaborative solutions over time. In December 2017, The Andrew W. Mellon Foundation funded Notre Dame’s project to create a unified discovery and access platform for library and museum collections. 

It was clear early on that the project’s complexities demanded that technical and subject matter experts from both organizations had appropriate representation and a framework to sustain what would become a new, shared campus service. Ultimately, strong project management, persistent outreach, ongoing communication, and cross-departmental teams became important ingredients to collaboration. 

Early Planning and Outreach

Project organizers set some basic guidelines for collaboration: each unit would receive one grant-funded position; team members would regularly work in each other’s spaces; the project team would meet every quarter for an update; and staff would tour and gain familiarity with each others’ collections.

These structures worked well for those directly involved in the project’s work, but there was still a question of how to ensure the diverse community of experts from the museum and library were also involved and invested over the long-term. 

The first strategy was targeted outreach by the Outreach Specialist who met one-on-one with library and museum curators, archivists, subject specialists, metadata librarians, and other stakeholders. Project staff also held quarterly meetings open to all library and museum personnel to hear project updates, ask questions, and learn about next steps. The outreach specialist and project manager also began working towards a website and blog to be a communication resource for staff as well as the wider world.

The Content Team Forms 

As 2019 approached, the core team realized that they would also need needed investment from collections experts to make decisions about digital content. They invited curators from the Snite Museum of Art, archivists from University Archives, curators from Rare Books and Special Collections, and librarians from Hesburgh Libraries to form the Content Team. Led by a curator from Rare Books and Special Collections, this group began to meet regularly to select representative content from across collections for inclusion in the platform. 

This team organized the first batch of content around a theme that could encompass materials from the archives, library, and museum. In choosing performance as the opening topic, curators drew from the library’s rich sports collections, Irish music broadsides, and religious materials. Museum specialists selected objects from the Meso-American collections, prints depicting historic theater performances, and a portrait of one of the greatest nineteenth-century French actors, among others. Archives staff selected Notre Dame student jazz festival programs, historic commencement programs, and speeches from the University’s beloved past president, Rev. Theodore Hesburgh.

PPT slide with digitzed books, photographs, and archival materials

Sample of selected content

The team was sure to choose diverse formats — from photographs and print materials to three-dimensional objects and paintings — in order to challenge the next two teams in the chain: Workflow Team and Metadata Team.

From Content to Workflow and Metadata

After forming the Content Team, the core project team created the Workflow Team to empower those who would ultimately implement the project workflows. Once the Content Team handed off the first test batch, the Workflow Team set out to take the raw material and make it accessible online.

The first challenge was to identify the source systems for the digital assets and associated information for the pilot collections. Hesburgh Libraries and the Snite Museum of Art have been digitizing their rare materials for conservation, research, and teaching purposes over the course of several decades. As a result, digital images are scattered across a range of storage locations including file folders, hard drives, and cloud-based storage systems. 

The associated metadata is also dispersed across several locations, such as an integrated library system (ILS), proprietary museum database, ArchivesSpace (finding aids for archival materials), standalone websites, spreadsheets, our institutional repository, printed catalogs, digital exhibits, museum object files, and most importantly, the minds of our subject specialists. 

Once the Workflow Team identified the source systems, they began talking to the gatekeepers about formats and access. The team is currently working towards moving digital assets and metadata from their current locations into a directory structure, through a IIIF pipeline, into the search index, and ultimately to the user interface. 

Image of group working at desks and white board with diagrams

Workflow Team meeting and diagrams

A Metadata Team was also started to engage experts from each organization to make collaborative decisions about how to manage metadata across the platform. They work in tandem with the Workflow and Content Teams to plan for metadata remediation and mapping. 

One of their greatest challenges is devising a solution that enables a robust search across archives, museum, and library collections when each project partner uses different data standards and formats. 

The team has been working on better understanding how each partner is cataloging their items and storing their metadata. In the next phase of the project, the team plans to make recommendations for how to reconcile subject term differences and develop basic metadata profiles. They will also document how to map museum and archival metadata into a unified search index so that users can search across the repositories’ holdings. 

Screenshot of website with painting on leftside and metadata on rightside

Metadata display for painting from Snite Museum of Art

The Foundation of Collaboration

At the onset, the core team understood that the software would only as successful as the collaboration and the long-term commitment of the people behind it. Opportunities for staff in the museum, library, and archives to better understand each others’ collections, professional best practices, and workflows were intentionally designed to strengthen the partnership between the two campus departments.

Already, members of these teams have begun to learn the best practices of others. Walk into a Workflow Team meeting and you might hear a metadata librarian talking about museum software. Sit down with the Metadata Team and the museum database coordinator might tell you all about archival finding aids. 

The Content, Metadata, and Workflow Teams have limited project charters — their mandate is to tackle specific use-cases and questions within the three-year scope of the grant. However, the investment in developing shared understanding and vision through teamwork has cemented working relationships that will thrive after the grant period ends. This will be critical to ensuring the success of the platform well beyond the initial development and launch.

The Origin Story

Posted on by

By Abigail Shelton

What and Why?

Like many academic institutions, the University of Notre Dame’s campus libraries and art museum have been digitizing collections for over a decade. Hesburgh Libraries and the Snite Museum of Art have shared these surrogates with researchers and faculty via email, focused digital exhibits, and collection-specific websites. However, this disjointed digital presence has reduced the utility and accessibility of these digital objects for scholars and students who are interested in searching digital collections and librarians or museum staff who are called upon to provide on-demand digital surrogates on behalf of users.

Enter the University’s 2017 grant from The Andrew W. Mellon Foundation.

This three-year grant project aims to bring those disparate digital collections into one access platform by creating multiple pipelines to a unified search index and user interface. The University is taking a cloud-first approach and the resulting application will be developed using Amazon’s cloud computing services.

We don’t promise to solve all of the challenges posed by legacy metadata, varying image formats, copyright policies, and proprietary or varied data stores. But we do plan to build intuitive, visually engaging user interfaces and a middle-manager data pipeline to begin tackling these discovery and access issues within our own institutional ecosystem.

Our goal of the Updates section is to keep you informed of our progress. We’ll be documenting our process from a technical perspective and a perspective on the project microsite, this update feed, our GitHub pages, and in an Open Science Framework portal. The ultimate hope is that our documentation and our code base will help not only Notre Dame, but also other institutions looking to provide greater online access to their own library and museum collections.

The First Year

In the first year or so of the grant period, we’ve focused on these foundational areas:

  • Hiring grant-funded staff (including yours truly),
  • Developing stakeholder relationships,
  • Building an Alpha prototype,
  • Forming subteams to empower subject and metadata experts to make decisions about content and information, and
  • Implementing an IIIF (International Image Interoperability Framework) pipeline.

Developing Stakeholder Relationships

As the Outreach Specialist, I set out immediately to meet with as many library and museum faculty and staff as possible to find out what their requirements might be for this new platform. In addition, I’ve met with teaching faculty and students from a diverse range of campus departments to find out how they use digital resources and how they might like to use unique digitized items from the Hesburgh and Snite collections.

Unsurprisingly, the number one need for these users was the ability to quickly search the collections by keyword, followed by the need to narrow those results using a number of facets and filters. Secondary needs include the ability to conduct a known-item search, manipulate images (zoom, download, print, etc.), and create their own custom groupings.

A network of outreach contacts

Our network of outreach contacts

Building an Alpha Prototype

We used these ideas and feedback to work on wireframes and, eventually, designed mockups to refine the features and look of the site. Our web developers have been implementing this vision in an alpha prototype website — internal users can now search and browse for a limited number of collection items to test early features and functionality. In the coming year, we’ll begin to test these designs with our users: teaching and library faculty, external researchers, undergraduate and graduate students, university staff, museum and library curators.

And as we’re refining the user experience, the developers will be building the pipelines that will connect our data and image stores to the search index and user interface. Our intent is not to force metadata or digital assets into one format or repository, but to connect an ecosystem of existing systems. This will require bridging open-source and vended solutions and finding a way to accommodate differing metadata schemas.

Alpha prototype homepage

Marble Alpha prototype homepage

Forming Subteams

We have also formed three working groups: Content Team, Metadata Team, and Workflow Team. Future posts will highlight the work of these groups in more detail, but, at the highest level, these teams have been a key vehicle for empowering experts to make project decisions related to their fields. For instance, the Content Team is composed of library and museum curators, subject librarians, and digital project specialists. Their mandate is to select digitized collections to include in the unified platform for the Beta launch.

As subject matter experts, these team members know the University’s collections best, and we knew they needed to be involved in the decisions surrounding online content throughout the grant process. In addition, as scholars in their own right, they can provide feedback on the usability of the platform from a researcher perspective. Similarly, the Metadata Team and Workflow Team include library and museum staff with experience and knowledge related to those areas of the project.

A listing of team members, organized by team and color-coded by institutional affiliation

Project teams, color-coded by institutional partner and area of expertise

Implementing a IIIF Pipeline

Deeper below the surface, the library’s developers have been building and implementing a pipeline that accepts common image file formats and transforms them into IIIF (International Image Interoperability Framework) compliant images. The pipeline also serves these images up in a IIIF viewer, which allows the user to zoom deeply, rotate, scroll through alternate images, download, and print the image from the viewer.

One of the benefits of using IIIF compliant images is that they can be used alongside any IIIF compliant images of other institutions. For example, scholars will be able to compare medieval manuscripts from the University of Notre Dame with manuscripts from Oxford University, Stanford University, Bibliothèque Nationale de France, or the Vatican Library side-by-side in the same viewer.

Image viewer showing manuscript notebook

Image viewer containing Nathaniel Rogers Sermon Notebook (MSN/COL 9405, Rare Books and Special Collections, Hesburgh Libraries, University of Notre Dame )

Looking Ahead

In future posts, we’ll dive into feature development, our sub-teams and their work, the evolution of our technical architecture, special digital collections, metadata mapping, and any other facets of our work that we think would be useful to share.

If you’d like to follow along, bookmark this site or contact Abby Shelton (Outreach Specialist) at ashelto3@nd.edu.