 ahead and get started. Welcome. 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 on the first day of week four of the CNI fall 2020 virtual member meeting. To remind you week four is focused on responses to the current set of crises that our community is trying to navigate and to help others navigate. This session is being recorded and the recording will be publicly available after the session. There is closed captioning available if you'd like to make use of that. We do have a chat and feel free to use that. And I'd also direct your attention to the Q&A tool at the bottom of your screen. You can use that to pose questions to the speakers at any point during the presentations. We'll go through both of the presentations first and then we will address as many questions as we have time for Diane Goldenberg Hart from CNI will moderate that Q&A session. And with that introduction, let me turn to introducing our speakers. We have with us today Mara Blake and Rena Murray from the libraries at Johns Hopkins University. And what you're going to hear today is what I think you will find to be a fascinating look at an activity that the libraries there got deeply involved in and that suddenly found the world was watching fundamentally. I've had a little bit of a preview of this project in a conversation with Winston Tab, the Dean at the Hopkins libraries and it's really just a wonderful project. So I'm delighted that Mara and Rena have joined us to tell everybody about this. And with that, I'll just welcome them both and turn it over to Rena. Great. Thank you very much, Cliff. And welcome everyone. Thank you for tuning in today. Welcome to our talk on a library's role in times of COVID lessons learned from the JHU Global COVID-19 map. Before we get started, some quick introductions. Like Cliff mentioned, Mara and I are both from Johns Hopkins University. I'm the Geospatial Data Curator and Applications Administrator. And Mara is my manager and we are both from the data services team. We want to emphasize before we start that we represent and are speaking on behalf of a larger team of dedicated individuals who have been working on this project since the beginning. You're going to hear us talk about Dr. Lauren Gardner and Ensheng Frank Dong from the Center for Systems and Science Engineering, also known as CSSE. But the core team consists of more members from CSSE, as well as others from the Sheridan Libraries, and also from the Applied Physics Laboratory, which is a not-for-profit research center that's affiliated with Johns Hopkins and that conducts both independent research and consulting services to a wide range of clients. And the names that you see listed here are really just the tip of the iceberg. You can see some of the core team members here during one of our many daily briefings and calls. This is just to represent to you that this dashboard has truly been a team effort and we're very excited to be talking on behalf of our colleagues today. So before we get started, we want to position our talk a little bit. We will be talking about this project from our perspective as people working in the library. And just as the pandemic is still ongoing, this project is still ongoing. And so our approach likewise continues to evolve as we respond to what's going on. And last, we talk in this presentation a lot about public projects as opposed to research projects. And we're not suggesting that research projects don't have a public-facing component, but we're making a distinction between sort of more traditional research projects and a project like this, which has just such a broad public audience and has a really strong public service-oriented goal at its core. So in this presentation, we're going to start by talking a little bit about us in the library. Then we'll move on to talking about the history of the dashboard and how it evolved over time before going into specifics about the library's role. And then we'll talk about some lessons learned, go to our conclusion, and then we'll open it up for questions. So with that, I'm going to pass it to Mara to talk a little bit more about us in the library. Thanks, Raina. So before we jump into telling you more about the dashboard, we wanted to give you a little context about who we are. And Raina and I are both members of the Sheridan Libraries, which is the largest collection of libraries and museums at Johns Hopkins University that primarily serve the main campus, but in particular, our group works across the JHU campuses in many, many regards. And we're part of a section of the library called the Data Management Directorate, which is made up of three groups. The first group is the Digital Research and Curation Center, which is a team of software engineers and systems administrators and systems engineers who develop new systems, new software. And many of our colleagues who also work on this project from the library are from the DRCC, as we call it. The next group is library applications, which is a team of administrators and engineers who oversee the core applications of the library, thinking like the library catalog and so on. And then our team, data services that Raina and I are both part of, which is the user facing component of this group. And our team and data services, we have nine people, and we focus on providing a suite of services that help researchers do research or teaching throughout sort of a research or teaching process, different aspects of data. So we help people find, use, visualize, manage and share data. And you can see some examples of some of the services that we offer out of our team right here. One of the most relevant service areas that we offer to this project is our services around geographic information systems or GIS. So like many university libraries, we help our users working with geospatial data either finding that data or working with that data conceptually or in an applied way in a particular GIS software or tool. So we provide individual consultations around those type of topics, open workshops or curriculum integrated instruction on those types of topics and, you know, a collection of resources, guides, collections of geospatial data to support those efforts on campus, very similar to core library services. Another aspect of those services that we provide out of our group deals with Esri support. So Esri, if you're not familiar, is the largest supplier of proprietary GIS software and tools. And they have an educational site license program and different institutions manage and maintain that site license different places. It might be an IT, it might be in an academic department like geography. But in the case of Johns Hopkins, we maintain that educational site license out of data services. Some things that go along with maintaining that educational site license. And so it's the site license that provides access to this large collection of tools provided by Esri to all members of the JGU campus. We administer our organizational instance of ArcGIS online, which is a web based software as a service and our lead administrator for that is Reina here today. And it's a platform, if you're not familiar with it, where, you know, our users can sign on with their single sign on university credentials and access tools for working with geospatial data, publishing it, sharing it, that type of thing. So, you know, the other in addition to ArcGIS online, we also help with technical troubleshooting and access support for all of the Esri products that are provided to campus for through this educational site license. So, Reina's going to tell you more about the coronavirus dashboard project, but you know, our involvement is a little outside of the scope of the normal services we offer, but we have consistently felt like it connects well and fits with the values and mission that we have in the library to provide access to information and for our team in particular to help people really understand and know how to work with data and have access to that information. And so, Reina's going to tell you more about the dashboard now. Thanks, Mara. So, the origins of the dashboard involve two individuals, Dr. Lauren Gardner, who is an associate professor of civil and systems engineering at Hopkins, and Cheng Frank Dong, who is a first-year PhD student in her program. And in late January, they got to talking about the coronavirus outbreak, and Frank mentioned that it'd be really interesting to try to track it. So, Lauren encouraged Frank to, you know, put something together just as a fun little side project, and he went home and basically put the dashboard, the first iteration of the dashboard together and shared it with Lauren. She tweeted about it the next day, and then she immediately boarded a plane for a business trip. And by the time she landed, her phone and her email were just pinging nonstop, and really the interest in the dashboard has just grown exponentially from there. So, like I was saying, the interest in the dashboard grew rapidly once it was shared. Early on, we were averaging about 30,000 hits per second. And as the virus spread, we just saw those numbers continue to tick upwards. The graph on the left shows the total dashboard views for the first two and a half months of its existence in millions. And the graph on the right shows the total number of feature layer requests in billions over the same time period. And feature layer requests, this basically gives you a sense of user interaction with the data. Feature layers are what the underlying data is structured as on our dashboard. And what a request is, is anytime someone clicks on something in the dashboard and it results in a query and something changes on the map or in the tables, that's sort of that's demonstrating a request. So this gives you a sense of just how much, how many, how many user or sorry, how, how many times users are sort of fiddling around and interacting with our data and exploring it more. As you can imagine, this was an entirely new experience for us. And as COVID-19 spread across the globe, we just continued to see this interest and demand grow. And so what started as the small side project quickly became a top priority. In January, APL, the library and a team from Esri all joined Lauren and Frank at CSSE to help provide some support for this project. And the first two and a half months of our work coincided with major and significant global events as COVID-19 spread. So this includes some events like what's showing on that graphic right now, as well as things like the WHO declaring the outbreak of pandemic. This also coincided with major changes here at Hopkins, as I'm sure at your institutions where we were adapting to the reality of the virus, you know, reaching our backyard. And during all this time, the interest in the dashboard just continued to grow. So we had to adapt and continue to find ways to accommodate this growing interest and demand. So as you can imagine, our team also had to expand and evolve. The team members that you saw listed at the beginning of the presentation make up the core part of the team that provides technical support. So this is stuff like updating and maintaining the workflow for the dashboard itself. So data collection, data cleaning. It also involves things like building the back end infrastructure to support the high demand and usage that we see. This technical support work centers around Lauren, Frank and others at CSSE. But it also includes a team from Esri, like I mentioned before, as well as from the Applied Physics Laboratory and from the library. We also had a growing support team that provided a lot of necessary project support from other parts of the university. So this isn't things like communications outreach. And we've just had a lot of people from various members of the JHU community who have worked tirelessly on these types of aspects of the project, as well as also a slew of external partners who have pitched in to help us out. And the names that you see listed here is not, you know, there's just so many more who have been involved. So if you've been following our dashboard, you'll notice that it's changed a lot over time. And the changes to its interface really reflect the changes that have been happening behind the scenes to not only the workflow, but also to the type of data that's available that we're able to gather and condense. You know, over time, we've been able to provide more information on not only cases, but also things like a case fatality ratio and testing ratio. And you can expect to continue to see more changes in the future. So you'll also notice or you might have noticed that the changes in the dashboard are also accompanied by changes and how users can access our data. Like I mentioned before, our dashboard lives on ArcGIS online. So the underlying feature layers are available on Esri's living Atlas platform. We also have a GitHub repository where users can go to grab our latest updates as well as time series data. And the interest in the dashboard also culminated in what you see here on this video, the JHU Coronavirus Resource Center, which is sort of like our one stop shop. So it includes the global dashboard, but it also includes a lot of other applications that are directly pulling stats from our global COVID-19 dashboard and data. And it's just a great resource that's available for people to use. So to date, the dashboard, as of a few days ago, we surpassed three billion views on the dashboard. The number of feature layer requests are at about 203 billion. So again, this sort of shows user interaction with our data. And our GitHub repo has been starred about 25,000 times and forked 16,000 times. So with that, I'm going to pass it back over to Mara to talk more specifically about the library's role in this project. Thanks so much, Rena. So we just want to highlight for you of this big project with many parts what role the library played throughout the project. And we'll start with showing you the general workflow of the overall project, just so you can get a sense of how things work to make this dashboard happen. So there's a bunch of data sources out there available. And there is data processor that lives on a server that collects those data sources, runs the processing that's needed, and takes the data, puts it in a GitHub repository, which is aggregated by the Esri feed aggregator, and pushed into the dashboard, which lives in our institutional instance of ArcGIS online, which shows you the dashboard that you see if you visit it online. And then that is part of, you know, one of the many resources collected in the Coronavirus Resource Center that the university provides access to the suite of resources that Rena was just showing you. So one of the areas where we have a lot of involvement and have had a lot of involvement with the dashboard is around the Esri piece, you know, I mentioned we oversee the educational site license and the dashboard itself lives in our institutional instance of ArcGIS online. Given the nature of this project being a little atypical for a standard project that would happen at a university, there's been lots of licensing questions that we've worked through collaboration with the Esri staff around that. And then also on the infrastructure side, ArcGIS online is hosted on the back end by Esri, but it's certainly not designed to handle the traffic that Rena was just sharing with you in terms of volume of views. And we've had to do a lot of work with Esri to ensure that the back end infrastructure is upgraded. We've had a couple of times where our entire ArcGIS online sort of failed. So that's not something we want to happen to the dashboard or to our organizational platform that we use for other things as well. So we've also consulted on the project about, you know, using Esri tools, how best to use them and done some usage reporting like, like the information that Rena just shared with you in terms of views and feature layer requests. That part of the workflow with the server that does the data processing is another area where we've had a lot of involvement at the library. So when the project first started, the data process was happening on someone's local machine, which, you know, was not the right place for a project of the scale as things grew. So the library stepped in to set up a server to host the data processing workflow and did that early in the spring and then contributed to the monitoring of that. So certainly there's monitoring that happened programmatically, but it required a lot of human babysitting. And so we contributed to a workflow that was almost 24 seven for several months this spring. Obviously, this is a little outside of our typical services, but something we offer for this project. And then as the project grew, we worked with our colleagues at the applied physics laboratory to migrate that process to a new server that an infrastructure that they're running. And then we keep ours as a backup in case the first one fails. And throughout this process, we've contributed to overall cybersecurity reviews for the project. And so as Raina kind of mentioned, this really led to a balancing act for us with the COVID-19 global map and dashboard on one side and the work associated with that. Well, at the same time, like all of us were dealing with transitioning our services and our campus to a virtual and remote setting. It's a really big change. And so on this one hand, we had supporting the dashboard with our existing staff and repurposing our work hours to support this project in sort of an emergent situation and prioritizing this. With on the other hand, dealing with our regular services and modifying them as needed for the new environment, switching our workshops to a webinar format, handling our consultations via Zoom instead of in person, offering online chat services, helping our campus community continue to get access to GIS software, much of which is typically deployed in a desktop environment, which is much more challenging in a remote setting. And throughout this, you know, we really have seen an increased interest across campus beyond this dashboard in dashboards, definitely, but also in GIS overall. So still seeing even in this strange time and increased interest for our regular services and having this special project and navigating that challenge. So Raina is going to share with you some of the things we've learned so far from from doing this. Thanks, Mara. So like Mara mentioned, the COVID-19 dashboard led to, I think, generally overall, a lot of increased visibility and awareness globally about how geospatial technology can be used in a situation like this. And locally, we've just seen a lot of growing interest among our researchers, faculty and staff who also want to learn how to use and apply products like this to their own projects. And as we've seen this shift from research projects to also more public facing, public service oriented projects, this has required us to sort of change our approach and take on more of like a product development mindset. Scaling up has required, you know, developing things like a formal intake process, making sure we have clear delineation of roles and responsibilities, clear scoping of our services of the services that we can provide and the level of support we can provide. Another big one was making sure that we were very aware and compliant with any campus policies and procedures that we're developing around products, projects of this scale and scope. And as you can imagine, those policies and procedures are also changing as they get more informed by the progress of a project like this COVID-19 dashboard. So we have to kind of stay on top of that and make sure we remain aware and compliant with any changes in those policies. And so this resulted in us needing to do things like create agreement forms and policies that we can share with any interested users who are who want to create similar dashboards or projects using our geospatial technologies. And you can see a screenshot of some of them to the left there. And if you've gone to our JHU Coronavirus Resource Center, you probably encountered some of those applications there. So to sum up, you know, for us, this meant that we had to make sure we were connecting and collaborating with the with our partners here on campus or in the JHU community. And we had to make sure we were remaining aware and compliant with our university's procedures and policies. As this interest grew, we also likewise identified a lot of areas of opportunity for increased education and training to foster this type of skills needed for this type of project. And so these are in hard skills like WebGIS and ArcGIS online administration, as well as in tools for automating and scripting work. In the case of the COVID-19 dashboard, mostly in Python, but also in other languages. But it also includes a lot of soft skills that are particularly important for such a public facing project around things like project management and communications and outreach. As we worked on not only the COVID-19 project, but also began to get a filter and request from other researchers who were interested in doing similar projects, we had to make sure we decided and figured out what we were capable of supporting and to what level and to make sure that we communicated that clearly. I mentioned earlier about like delineating rules and making sure you have a clear scope of what service areas you'll provide and sort of the scan. Those all became very important to make sure that everyone involved in a project was aware of their responsibilities and areas to focus on. So with that, I'm going to pass it back to Mara to conclude to provide our concluding thoughts. Just have a few thoughts to leave you with before we open it up for questions. So the first one is more of a question. So Raina mentioned these areas for development. And so this is the question we're grappling with, which is of the skills that we've identified are important to develop in our campus community and our researchers who want to pursue projects like this. What will we take on as a library and as a data services team? So this is something we're working through and we're kind of starting with the ones we're more certain we might support moving on to ones that we might look to other collaborators to support more fully. So we're really seeing a growing trend towards projects like these. We don't anticipate there will be any completely analogous to the COVID-19 global dashboard. It's pretty unusual as a project, but as Raina described, we are seeing lots of interest at our campus and that I think we're starting to hear that this is these public research projects are of interest on other campuses as well. And so from our perspective, you know, it's really thinking about that balance between the core and existing services that we're already supporting full-time and then sort of the balance and even tension in some cases with these other projects and how we can contribute to them. But even with that, I think we have really found that there is an important role for the library as our institutions pursue these public facing public oriented research projects and that we're really excited to talk to you all about how to prepare ourselves as a field for that role that we will continue to play with these projects. So with that, I think we are happy to talk to you more with questions and any discussion that you have. And thank you so much for joining us today. We're really grateful, grateful for your time and looking forward to hearing more from you. Thank you, Mara. And thank you, Raina. That was a terrific talk. Just an amazing tool that you've built at Hopkins and we're so delighted that you've come to CNI. To share some information about it with us. And thank you to our attendees for being here with us today as well. As Mara mentioned, the floor is now open for questions, so please type your questions in the Q&A box and Mara and Raina will be happy to address those now. While we're waiting for folks to share their questions with us. First off, I was just fascinated to hear that origin story that all this started with a faculty member in the School of Engineering and a graduate student. And I was wondering if you could elaborate a little bit more on what happened after that? I mean, how did the library end up getting involved in this project? I can take that question. So, Frank, or Ensheng Frank Dong, who is the first year graduate student who produced this dashboard, he built it, as Mara and I mentioned, on our ArcGIS online platform. And so within about a week of it being out, it was getting so much traction. And it was we were sort of we're able to monitor our usage and we were seeing these crazy spikes. And also he reached out to us saying this is I've been trying to manage this, but I need some help. Can you help me? And so we kind of within about it wasn't even a week, but within a few days, we were sort of aware of the project and unable to jump in and work with him. And also by then, Esri was also aware because they had likewise never had something on their platforms just explode like this. So we were all just sort of. We were made more in some ways. We got to sort of see the outbreak from a very or I think from an earlier point than maybe some of our colleagues in the US were aware of just because we were working on this. Yeah, I thought that was also really interesting. I didn't I didn't realize that the dashboard had had gotten started so early. So yeah, and those numbers were eye popping to be sure. OK, thank you. I appreciate you're addressing my question. Now we have a question from Nathan Geirth and Nathan says, so this project started by accident. Could you talk a bit more about how you hope to be more intentional about public research projects in the future? That's an interesting question. You can take that one. Raina, that sounds good. Thanks for your question, Nathan. So certainly. That's a really great point. And Raina described some of the policies and procedures we're putting in place from our team about the use of the resources we support. To be more intentional going forward, obviously it was really. Seat of our pants kind of situation with the pandemic and the dashboard both taking off for the same reason at the same time. But going forward, we have a different approach where we have more requirements for teams wanting to use that infrastructure and policies that we share with them up front about describing what we do, what we're expecting them to do, making them aware of campus policies that we've been aware of now through this project to make sure everything's in line. And, you know, we just couldn't scale the support that we gave and continue to give for this project to to every project. So finding a way to strike that balance and identifying like we can trainings, things like that that can prepare people for this because we can't give that kind of dedicated support across the board. And if if you're interested, we're happy to share more about, you know, the nitty gritty of how how we set things up going forward, if it's relevant to your work. Great. Thank you, Mara. And thank you, Nathan, for that question. I see that we are actually a little bit past time here. So I think I'm going to go ahead and close the public portion of the presentation. I'll start to stop the recording at this point. But I'll just invite anyone who would like to join the conversation after we've turned off the recording, please stick around, raise your hand, I'll turn on your microphone and you can just have a chat with us here afterwards. So with that and a final thank you to Raina and Mara for sharing this fascinating project with us. Thank you so much and to our attendees for making time out of your day to join us. Thank you, everyone. Be well. We hope to see you back and see you nice soon. Bye bye.