 Go ahead and get started. Welcome everyone. Glad you could join us. I'm Cliff Lynch. I'm the director of the Coalition for networked information. And you've joined us for one of the project briefing sessions for week four of our fall 2020 CNI virtual member meeting. Just to remind you deals with emerging issues and responses to the current crises that we are confronting collectively. A few things just about this session. This is being recorded. The recording will be available publicly after the session. We do have closed captioning available please feel free to turn that on if it's helpful. There is chat available and feel free to use that during the session. There is a Q&A tool at the bottom of your screen after we hear from our presenters. Diane Goldenberg Hart from CNI will moderate a Q&A session and we will deal with as many of the questions as we can during that time. And with that, let me just say a few words to introduce our presenters and the session itself. We'll be hearing today from Shanae Yvette Moraine, who is the DPLA director of community engagement, and from Anne Hanlon. Anne is with the University of Wisconsin at Milwaukee, and is currently serving as DPLA network council chair. This is a very interesting presentation, which I am just delighted to have at CNI because it combines a number of really interesting and important issues. Part of the story here is about how we build collaborative collections. Part of the story here is about how we fill in neglected histories and neglected communities in our historical and cultural record, how we make those visible. Part of it is about how we appropriately talk about and describe these kinds of collections and how to do that again in a collaborative context. And this is an area that DPLA has really taken on, I think, in a very thoughtful way and produced some guidance that I think may be quite helpful for others who are working on these kinds of challenges. And with that, I don't want to steal any more of the story because I will leave it to Shanae and Anne to tell it. Welcome. Thank you for joining us and over to you. Thank you for such a wonderful introduction. This session is on dismantling racism and collaborative collections, and we are myself and my colleague Anne Hanlon who is the former chair of the network council and Penelope Schumacher who might join in later in later on during the Q&A. We'll be talking about how DPLA through our Black women's suffrage digital collection has been experimenting and thoughtfully creating policies and practices for our larger network around diversity, equity and inclusion. A bit of context, the digital public library of America empowers people to learn, grow and contribute to a diverse and better functioning society by maximizing access to our shared history, culture and knowledge. At DP.LA anyone with a mobile or internet connection can discover over 40 million images, text videos and sounds from across the United States. In 2019 DPLA launched a new strategy with an emphasis on uplifting stories from and working with institutions and projects and communities that have traditionally been excluded from projects like ours. Recognizing the legacies of racism and white supremacy and member institutions and how that has resulted in the failure to develop collections that are truly representative of the public. When we serve, the DPLA network council began drafting a document to guide the creation of more inclusive and diverse collections as a part of their 2019 priorities. The idea statement, which the acronym stands for inclusion, diversity, equity, access and social justice. It unifies the membership around a clear expression of intent to actively work on closing this content deficit and identify how a national digital library can and should respond. With the COVID-19 pandemic and the global racial reckoning of this summer, DPLA had begun to thoughtfully and critically examine what it means for our organization to support access and inclusion. We have seen over the past year that when we strive to put equity and inclusion at the center of our work, the work itself is enhanced. We are better able to serve our community and advance our mission and our brand is strengthened as a result and will speak to the ideas work and the crafting of the statement and The DPLA network council is the membership arm of DPLA and it's made up of representatives from each of the member hubs. The document that Shene mentioned to guide the creation of more inclusive and diverse collections is what became our idea statement. A document that unifies the membership around prioritizing diversity, equity, inclusion and social justice in our approach to building digital collections and as Shene said, and then how a national digital library responds. The impetus for the idea statement came out of strategic planning that the network council initiated alongside the DPLA's own strategic planning process in 2019. The network councils was aimed at the network itself, of course, and the goals of the contributing institutions. As part of an in-person meeting in Chicago in April 2019, the network council spent a day brainstorming and discussing the challenges that we faced, the opportunities that DPLA presented both as a platform and as a community, and how we should prioritize the laundry list of items that all of us saw as necessary and continuing to build and improve a collaborative, aggregated, open access, useful, discoverable, ethical, representative and truly national digital library. We identified a number of priorities as you can imagine things like metadata, write statements, SEO, curating collections, using IIIF, exploiting DPLA's API. But the most, excuse me, the most important and the clearest thread across the discussion that day and across all of the hubs was diversity, equity and inclusion. The task force grew out of the first action item in our top-level priorities document that we developed, and that's what you're seeing here. And for that task force, we brought together volunteers from the member hubs to help draft a statement that could guide our mission and priorities. We were able to pull together a great task force from across our hubs, and that task force included Shanae, who you've just heard from, as well as Penelope Schumacher from the Ohio Digital Network, who was also one of the co-chairs of the DPLA's metadata working group, and also contributed to this presentation. The statement came together as comprised of three main parts plus a preamble to provide context, and those parts are collections, membership and partners, and a good faith statement. We built the statement this way because this represents the areas where we can take action and where our deficits are most apparent. So the preamble acknowledges the unique position that a collaborative national digital library finds itself in terms of potentially being able to take a bird's eye view of the historical homogeneity of our collections. Though it's hardly a trivial task, the potential is there to tap into the supposed breadth of our collections to better see what's missing and to look more critically at what is there. Fundamentally, though, the preamble sets up the statement as a call to reorient our collective work around a quote, collective commitment to an authentic reckoning with our past and an active stance toward doing better for the future. The collections aggregated in DPLA are an uncurated aggregation of the digitized portion of the collections of the collections our institutions have collected over the decades. So in many ways they provide a clear window into where our collecting has been focused and also to what we've prioritized for digitization of course. But the statement itself is forward facing the active stance toward the future aspect of the statement and addresses some of the key components of the digital library itself, the content, the coverage and the descriptive language that enables discovery and use. The cultural heritage organizations that make up the DPLA hub network must be representative and diverse to increase the diversity of our collections, but also to provide other perspectives on how we do the work of digital libraries. This means that as we take on new partners, we need to be open to new ideas about how our digital libraries are built and what they include, how we provide access and certainly how we describe and provide context, and also how our collections to themselves are used. This is again part of our active stance and is meant to signal an openness to new partnerships and to how those partnerships are negotiated. And finally, the good faith statement was key because we felt that the statement needed to go beyond identifying the material aspects of the strategy and where we needed to essentially grow. The good faith portion unifies us in our approach and understanding of what got us here. We're all inheritors of biased collections that we learned how to collect and prioritize based on principles that favor institutional actors, the already powerful and prominent, and that the perpetuation of that more visible historical evidence served to perpetuate authority as white and male and upper class, rather than a record of the diversity that makes up our rich cultural heritage. As a member of the DPLA network, your hub and your institution acknowledges these deficits and acknowledges the importance and priority of correcting the historical record, and does so explicitly. The ideas adoption process was also part of a larger effort during the 2019 20 member year to overhaul network communications to ensure that voices were being heard and members had adequate opportunity to express the needs motivations questions and comments of the hubs that they represented. In October 2019, we introduced the ideas task force concept and called for members in January of 2020 we kicked off the task force, and in April 2020 we would have met in person. But instead we met virtually and debuted the draft statement there and asked representatives to take the draft back to share with hubs for feedback and created an open document to collect feedback with a deadline of June 26. The final version of the statement was largely unchanged from the draft submitted for review in April. Nearly all of the feedback stated a strong desire to ensure timely action to begin work that reflects the values articulated in the idea statement. In June we opened the vote for adoption to the network Council representatives and the adoption vote passed handily, and the idea statement was officially adopted by the network Council in August. And now speaking of action. I'm going to toss it back to Shanae to provide some details on a couple of projects that were launched alongside the idea statement and to speak to how DPLA will continue to grow with the idea statement as our guide going forward. Thank you and as and just teed up further embodying our diversity equity and inclusion work. DPLA has been intentionally spotlighting the stories of African American women in voting since this year is the centennial of the 19th amendment. We've been exploring how black women's history of activism laid the groundwork for some of the activism we're seeing make an impact today. The black women suffer digital collection which launched in October is a new dedicated site featuring original contextual content. Key suffrage bios timelines for users who might be unfamiliar with this history. The site is powered by funding from pivotal ventures which is an investment in incubation company created by Melinda Gates who has a focus on women storytelling. The project provides digital access to materials documenting the roles and experiences of black women in the women's suffrage movement and more broadly, women's rights voting rights and civic activism between the 1850s and 1960. The black women suffers digital collection is an exciting opportunity for DPLA to work with cultural heritage organizations outside of our network, which is important to us. Our partners at the Atlanta Center, Atlanta University Center Robert W. Woodruff library every research center for African American history and culture in Charleston, South Carolina. Tulane University, which is in Alabama and does not feed into a DPLA hub. The Amistad Research Center at Tulane University which is in the end does not feed up to a DPLA hub and the Southern California library each received sub awards up to $25,000 to support new digitization, data creation and remediation for these project partners. The content featured in black women suffrage explores the linkages between women's suffrage and other social causes of the 19th and early 20th centuries, anti slavery anti lynching immigration reform and civil rights, as well as racism within the suffrage movement and by extending to the mid 20th century the project helps to spell the notion that all women successfully secure their voting rights following the adoption of the 19th amendment. And as a result of us expanding that timeline and including all of those topics we had to create a process to surface those materials. As Anne said, DPLA collects and makes freely available materials from thousands of libraries and cultural heritage organizations. And each institution or contributed shares materials with DPLA according to its specific policies and objectives, which means we get all kinds of metadata of various, you know, various quality and various description. As part of what we call the noise cancellation party DPLA staff dedicated time to gather together and to search and collect subject terms or keywords related to black women suffrage from and also look through those results to further window them to the relevance of the site. We developed a methodology for distinguishing content about people we wanted. And we call this noise cancellation. And I came up with that that term thinking of these quiet dance parties that started to to become popular last year where everyone has their own set of headphones and they're listening to their own, you know, their own music and I thought about all of us getting on one channel. This noise cancellation included the intellectual and contextual work of searching for women's names and control of that vocabulary and looking for people places and organizations including terms no longer use, including terms that are considered today to be racist, misogynistic. We took notes about characters and relationship information for context we use the Wikipedia pages, and also created content through our with immediate comments project to provide more information about who these women are and why they're important to the collection. All of this with a goal of identifying reliable information about under researched women. So the work of dismantling racism and collections requires deep intellectual labor subject specialty. Many of the staff did not have. We didn't have precise expertise in black women suffrage history so we read a lot of secondary sources. This required dedicated time. With respect to our noise cancellation party started in April, and we met regularly to refine our larger query and all of our various spreadsheets that informed that query through the end of June with the launch of the physical site in October. The DLA developer Audrey Altman led myself and Kat Williams who's our communications director through applying for an API and looking through the query through GitHub. But it was mostly myself and Kat providing insight for Audrey to test and tinker and build our query is updated every Friday and new records that are added are pulled from the larger DLA aggregation. So we applied a black women's suffrage tag explicitly and then later refined and we also applied that tag to our partners materials that were coming in for digitization. We tagged collections via subject headings for early for an early query as an effort to put in what we expected to find about the people places and things that we already knew. In this spreadsheet now in hindsight, we can see where a lot of the noise came in. Much of it was from a relevant government records like the records of the US food administration cemetery records Bureau of Ships. These collections contain some topical relevance but they needed Boolean descriptors and some cases multiple phrase Boolean descriptors to become more relevant. So we put together some spreadsheets with candidate terms using initial organization titles and we expanded the results pulled in from the DPLA larger aggregation. Then we searched and collect subject terms and keywords from looking through those results to further window the relevance. In our next iteration of this candidate term spreadsheet, you'll see these super broad terms that we got a huge dump with all of these results to a cold through some examples of that are universal suffrage abolition to which we added names, and geographic limiters in some cases language limiters based on our knowledge of the history of black women's suffrage. And this was a real learning process. In some unexpected areas of expertise came up as a result. For example, Jim Crow. We know Jim Crow by definition is the practice that limited the free movement of black people during the segregation period. Through this process we also learned Jim Crow the character we discovered that the broad term pulled in results related to Tom Rice who was the white actor, who was famous for playing Jim Crow in blackface. We also learned we decided to keep some of the origins of Jim Crow is a menstrual failure because they be relevant to the collection, having to do with civil rights and segregation, but we also found that Jim Crow was a really popular name for horses. So we added Jim Crow and horses to remove all of those photos of Jim Crow horses. We, as part of our work. Also, if you're thinking about the noise and the sound metaphor, we had to determine what loud noise was an example of that is the term white voter which we found in over 6000 records from the State Library and Archives of Florida. And these were just voter roles that identified voters by name and race. So we pulled those out. Continuing on with our noise cancellation work through the summer, we started to use fetch back results to randomize our noise cancellation. And what we were doing with this is essentially putting the playlist on shuffle and testing our work. One error we found was creating what we began to call historically inaccurate Boolean surges, like Susan B Anthony and racism, racism is not a term that was available when these primary sources were created to describe Anthony's exclusion of black women in the movement. Racism would be used more secondary sources. So we tested that and found zero results. So we knew racism is not a term that we can use, prior to a certain date. We're also still finding many records related to suffrage but that were focused on white women so we were thinking about how to parse those. For example, the temperance movement. We added some geographic limiters and for voter registration we added and civil rights and African American. So one of some of the, the variation in noise cancellation work we did for names was with Shirley Chisholm. This is a before and after screenshot showing all the name variants for Shirley Chisholm as an example. Query expansion is really an exercise to get to both sides of the same coin to ensure we're not pulling in junk but expanding the query to make sure we're casting a wide enough net so that's adding forward and backwards, adding geographic references, adding dates, adding buildings and movements that also are associated with her name and her work. And that's really important in the context of recovering underrepresented people and, you know, all the information that that's key to the collection. In the next slide you'll see a screenshot of a code change where we expanded the tags to keep out noisy results. When we did batch results and then reviewed I did an individual identifiers to test how well we did. We saw that there was still room for improvement and more Boolean operators. So thinking about, you know, all that we all that we uncovered about language and DPLA serving as an aggregator. We were thinking about assessing our work and presenting this collection for the general public and as part of our work we have these these working groups like the metadata working group, who ideates and solve problems in the digital space that we Audrey Altman the developer previously mentioned, wrote the first draft of a harmful language description and I work with the metadata working group to refine the statement to be sure that it align more with the black women suffers digital collections content and the project needs. The audience for this work is the general the general public anyone who's visiting the site so it's not a content warning. It's more so an education tool for students and scholars and others who don't necessarily understand descriptive or archival practices. The metadata working group, in addition to writing this statement brought brought a set of FAQs, thinking about an audience of middle school students to educators as well as a general public. We wanted to work to avoid using library jarion to meet this really, you know, approachable and understandable. And the group wanted to explain why potentially harmful content might be found in this collection, and also why that content is important to history you know we wanted it to remain accessible. In addition, as a group, we wanted to address the potentially harmful metadata and descriptive terms that surface those results we wanted to acknowledge that some librarians and archivists are working to solve the problem and the hint and the scope of the issue and why it may still be harmful in terms of metadata. We also talked about reuse terms and outdated terms used and at the same time we wanted to acknowledge that the problem should be the scope was large and set a process for folks to report things through our info.dp.la email address and then create a process within our hubs to either take those down to dp.la or at the home institution and that work continues with the approval of a larger harmful language description statement for the entire dp.la project. The black women suffrage site and the harmful language product of that implementation I've already been used as models for other libraries who are doing this work, Florida University and University of Milwaukee where Anne is. And we're in just last week, the metadata working group shared the updated version of this statement with our advisory council and our councils going into 2021 will be thinking about how to implement this and what it means to set this forth as a value and how to ensure that dismantling racism through the idea statement and through the harmful language description are not just words that we put on a page but they are actions for the benefit of those who use our collections. Thank you all for your attention and Anne Penelope and I are happy now to take your questions and comments. Thank you so much, Shanae and Anne for that really interesting overview of this fascinating site, this wonderful project and the work leading up to it. I've been checking out the website a little bit as you've been talking and it's just fascinating I dropped the link, a few links into our chat I hope folks will check that out. I said, the floor is now open for questions so I hope our attendees will share their questions in our Q&A. And while we are waiting for folks to sort of chew on what we just heard. I'm curious to know if you have gotten any feedback. In particular, I mean about the site because I realized you just launched this summer right about the site but in particular this part that you were just describing Shanae on language that may be harmful. Have you gotten any feedback on that? We have not received yet any requests for takedowns which indicates to me that the statement does what it is intended to do and it contextualizes why these things are available on the site. Since the site launched in October, we've had a number of public programs at DPLA and also outside of the library world like an online book festival called Well Red Black Girl and we presented on the site. And after that I looked at our Google analytics for the site and was really curious about where people were landing because we did all this deep contextual work, providing a timeline for women's suffrage and also looking at some of the key figures. And it was really interesting that those were not the most used pages. The search were, you know, the advanced search and the basic search were the most used pages which was really encouraging because people are just jumping into it. So for me those analytics are really great feedback. And they also make me think more about the user experience and maybe there's a way to direct people more explicitly to the key figures in the timeline, but it seems like for now they're just open to exploring. So interesting. Well, thank you very much for that and I see that we're just a little bit past time here so I want to be respectful of folks time. I think I'll go ahead and turn off the recording here and invite any attendees who wish to stick around and join the conversation. Our panelists have graciously agreed to stick around a little bit longer and chat with us. If you're interested in joining just raise your hand I'll be happy to turn on your microphone. And with thanks again to our speakers and Shanae Penelope. Thank you for being here as well. And to our attendees for spending some time with us here at CNI. Thank you so much. Have a great day. Hope to see you back at CNI soon. Bye bye.