 Good afternoon, everyone. My name is Barry Hayes, and I'm the bioinformatics research data librarian at the PostSciences Library at UNC. Hi, everyone. My name is Amanda Henley. I'm the head of digital research services at the University Libraries at the University of North Carolina at Chapel Hill. We are pleased to speak with you today about the services the University Libraries have been developing to support faculty members, research staff, and campus units and demonstrating impact and improving efficiency through data-driven decision-making. Like other R1 institutions, UNC at Chapel Hill has been working to find efficiencies and demonstrate the impact of our teaching, research, and service. Our Provost Bob Lohan was quoted recently in the University Gazette saying that we have to take better advantage of big data analytics, predictive analytics, in order to accelerate and support the work of our faculty, staff, and students, and that we should be calling on people from across campus who have expertise and interest in these issues in order to improve and optimize the way the University operates. The libraries have seen an increased interest in communicating impact and improving efficiency from many areas and campus in recent years. We've received support requests from faculty members, research staff, and campus units that want to communicate their impact, identify areas of strength, demonstrate collaboration, patterns, and growth, and inform efforts to reduce costs and maintain quality. To accommodate this demand, we've made organizational changes and built expertise. Given the labor-intensive nature of this work, we're building these services with a focus on building sustainability. We're growing services in these areas and our research hub, as much of this work that is being done, uses methods and tools and technology supported by these units. The University Libraries Research Hub resulted from the library's five-year plan that began in 2013. The goal of the research hub is to foster research success by providing active support throughout the research lifecycle. We decided on a distributed model for our research hub because we recognize that in-depth research support for new modes of scholarship is needed across our library system. The research hub sites are the health sciences library, Kenan Science Library, and Davis Library. Three sites have some commonalities, including technology, spaces, expertise, campus partners, and research data management services. But each site also has special services tailored to the users of each library. To adapt our services, we've made organizational changes and built expertise while, like I said, maintaining a focus on sustainability. Two main organizational changes remain. In 2014, the Digital Research Services Department was created to support the Davis Library Research Hub. The department was composed of existing staff from multiple departments, including e-resources and serials management, library information and technology, and research and instruction. One new position was added in 2016 and that was the Data Visualization Services Librarian. In early 2016, the Health Sciences Library began exploring citation network analysis and visualization tools to incorporate alongside building metric tools and methods in its work with the UNC Clinical and Translational Sciences Award Unit. With promotion via the research hub, the Health Sciences Library saw a steady increase in requests for these services from other units and individuals. So in 2017, to build capacity to support the growing services demands and to take on larger institutional projects, the Health Sciences Library created a cross-departmental impact measurement and visualization team from the Public Services and Health Technology Informatics Unit. To build expertise, we added new positions, updated roles, provided staff training and the time needed to participate on project-based learning. Existing staff and Digital Research Services advanced skills to provide a base level of visualization support until we were able to hire our Data Visualization Librarian in 2016. And since then, our Data Visualization Services Librarian has taken the lead on large visualization projects. She also offers short courses on visualization principles in Tableau, which are streamed to the Health Sciences Research Hub site. The short courses are open to everyone and the library staff are encouraged to attend as well. Staff members in the Health Sciences Library built on existing bibliometrics expertise and trained in citation network analysis and visualization methods. And in 2017, the Health Sciences Library added a new staff position of Health Technology and Informatics Librarian and revamped an applications analyst position to build the Impact Measurement and Visualization or IMV team. The IMV team members all participated in the bibliometrics and scientific for research evaluation course and other trainings covering citation analysis and bibliometric visualization tools and methods. Another way that library staff members are building skills is to project work to build skills and text analysis, which is another growth area for us. Staff members have taken a leadership role in establishing a Carolina seminar called Transforming and Query through Digital Text Analysis. This is a cross campus group that works on text analysis projects. Our staff members have also been involved in training librarians at UNC and across the country as part of the IMLS funded project, digging deeper, reaching further libraries and powering users to mine the Hottie Trust Digital Library resources. As service requests and project partnerships grow, the libraries are leveraging a tier engagement model and tools like project charters to help efficiently and effectively manage and scale services. The research hub sites offer training and tools and methods useful for measuring and communicating impact. For lower level inquiries, patrons are assisted at walk-in service points at the Davis and Keenan Science Research Hub locations. More intensive inquiries are addressed through scheduled consultations. To manage the work of collaborator level projects, the Davis Library Research Hub is piloting the use of project charters for both short term and long term projects. These charters outline the goals, scope, responsibilities and sunsetting of projects and ensure proper recognition for the university libraries. For units requesting dedicated, ongoing staff expertise, the libraries have arranged with those units to provide support for the expertise. One current example is the Health Sciences Library's arrangement with the North Carolina Clinical and Translational Sciences Institute, which supports a portion of a position in the Health Sciences Library's impact measurement and visualization team to serve on their evaluation group and provide ongoing bibliometric and citation network analysis for their reporting to the NIH and other stakeholders. To illustrate the kinds of work being done, we have a few project examples to share. The first example is this cord diagram that was created for staff members in the university's Career Services Department. This animation shows how the visualization works. The nodes on one side of the circle are associated with schools at UNC and the nodes on the other side are associated with career fields. The width of the arcs connecting them represent the number of UNC graduates choosing the different fields. This visualization helps staff understand and communicate career opportunities for students. And the data used to create the visualization are from the first destination survey, which surveys may graduates receiving their bachelor's degrees. The next two examples communicate geographic impact. UNC's blueprint for next is our strategic framework meant to guide the university during the next decade. Over the strategic, one of the strategic initiatives in the blueprint is scaling global, which is intended to scale up Carolina's best ideas and initiatives to take on interrelated local and global issues. And they have our impact both at home and in other countries where we can make a difference. This initiative has resulted in the need for campus units to demonstrate their global reach. The first example that I have here is a story map created through a collaboration with the GIS librarian from digital research services and staff members at the Gillings School of Public Health. The map provides geographically situated stories about the school's global practica program. Our collaborator on this project shared feedback on our collaboration stating that public health is the most powerful when we work across boundaries. In this case, we're grateful that you and your colleagues on Davis Library are partners and you'll see that your reach extends from the globe. In this example, the Carolina Asia Center demonstrates their broad engagement in Asia. This map is embedded in their website and provides filters for program type, department, institution, and geography. In this example highlights a project with the Eshelman School of Pharmacy to identify overlap across graduate courses. Faculty and students tagged each syllabus with a controlled vocabulary of topics and skills covered in the program, and the research hub data analysts transformed topics into overlap reports. This project is another example of how the libraries are assisting campus stakeholders to be more efficient and make better data informed decisions. At this point, I'm going to turn it over to Barry to talk about some product examples from the Health Sciences Library, Future Plans, and then conclude the presentation. Thank you, Amanda. The syllabus analysis project that Amanda just described was a key component of the Eshelman School of Pharmacy's ongoing curriculum redesign initiative. The Pharmacy School engaged the Health Sciences Library in this initiative as well. As a result of the curriculum committee, at the request of the curriculum committee, AGSEL's impact measurement and visualization team, or IMB team, as Amanda mentioned earlier, analyzed five years of pharmacy faculty publications and produced citation network visualizations to illustrate faculty collaboration patterns and to identify areas of research strength, as well as overlap across five pharmacy school divisions. The team conducted these analyses for each of the five divisions as well as from the school as a whole. These are two examples of the faculty co-authorship networks we produced to demonstrate faculty collaboration. The first image, the ESOP faculty collaboration network, on the left, illustrates co-authorship collaborations for 62 pharmacy faculty, authoring one or more of 1,858 publications published between 2012 and 17. This network also includes 1,150 UNC affiliated collaborators outside the School of Pharmacy. In these citation network images, each node or bubble represents a pharmacy faculty author or one of their co-authors. The size of the bubble represents the relative number of articles and the citation data set produced by a given author. Larger nodes represent more productive authors. Network clusters shown as different colors are based on co-authorship links and are assigned by application algorithms and the proximity of the nodes reflect the relational connectivity of any two authors. In general, the closer two authors are to one another, the stronger their relatedness. Lines in the image also represent relatedness, in this case, the co-authorship relationship. The image on the right shows co-authorship collaborations for 22 faculty members in the DPEC division and their 577 internal and external collaborators. This is a time overlay network visualization. It displays a co-range based on the average publication date of the associated article citations for each author. In this image, dark blue to blue-green in items indicate authors predominantly publishing at the beginning of the publication date range, while light green to yellow items indicate authors with more of these publications. In addition to understanding faculty collaboration patterns, the pharmacy school was interested in identifying areas of research specialization and overlap across their five divisions. Similar to the syllabus overlap analysis performed by the Davis Research Hub, the HSL-IMV team produced research topic maps and used color to indicate topic areas where different divisions have overlapping expertise. In another view of the same citation topic data, the IMV data analysts created a Tableau dashboard to illustrate the top offer and index keyword terms by division. These data, along with the syllabus overlap data, are helping the pharmacy school make decisions about areas for critical inefficiencies across their divisions. Looking quickly at these examples, the top left image illustrates the topic overlap among three divisions in orange, DPEC, DPEC, and DPEC. It also shows topic overlap occurring between only two of these three divisions in pink, CBMC and DPEC only. These divisions overlap on topics related to the basic science aspects of pharmacy. The top right image shows the topic overlap among a different combination of divisions in green, D-POP, PACE, and DPEC. These divisions overlap on research topics related to pharmacy policy and practice. As you can see, DPEC topic areas overlap with both groups. At the bottom is a snapshot of the interactive dashboard I mentioned. This view shows the top 25 author and index keyword terms by division. At the conclusion of our work on this project, the HSO director and IMB team staff were invited to present these findings in a faculty retreat focused on the curriculum that is on them. In advance of the retreat, the chair of the committee leading it shared this feedback on our collaboration. The information you've shared was extremely helpful. Based on our discussions and a review of the data I believe we have an excellent plan for the breakout groups for upcoming faculty retreat. We sincerely appreciate all the time and effort you're investing in this project, which is absolutely critical to our project program transformation. The next several slides highlight a few examples of bibliometric analysis and citation network visualizations that IMB staff produce as part of HSO's ongoing collaboration with the North Carolina Translational and Clinical Sciences Institute for NC Tracks. This institute is UNC Chapel Coast Clinical Translational Science Award unit, which leverages NIH funding to support translational research by providing pilot funding, research infrastructure and support services to UNC researchers. HSO began partnering with NC Tracks to perform bibliometric analyses on NC Tracks supported publications in 2010 and in 2016 NC Tracks asked if HSO could also provide citation network analyses expertise to help them expand and enhance the assessment of the influence reach and collaboration patterns of UNC CTSA supported research. Since that time one of HSO's IMB team members have provided this expertise and served on the NC Tracks evaluation group. As Amanda mentioned earlier, NC Tracks now also supports a portion of this position to provide dedicated ongoing expertise in these areas for their assessment and recording activities. This staff person produced this figure in the next several visualizations for recent analysis in NC Tracks supported publications. This particular figure shows how the citation impact in NC Tracks supported publications compares with the citation impact of the average NIH funded papers in the same field and year. Relative citation ratio scores generated from NIH's eyesight tool or the citation based measure here used to compare how the influence of the NC Tracks portfolio articles compares to the remaining articles that came out of grants funded by the NIH and is in this way gauges the research influence of the NC Tracks portfolio. This citation network visualization illustrates co-author collaboration patterns among UNC Chapel Hill health affairs and other units on NC Tracks supported research projects. Similarly, this visualization illustrates international collaboration patterns on these projects. This topic network demonstrates the predominant topic keywords occurring at NC Tracks supported research article citations. The term clusters produced by the analysis software are terms that occur together in article citations specifically in the text of titles or abstracts. The NC Tracks evaluation group created a legend which you see here to describe the type of topic terms that occur in each cluster shown in different colors. Across the clusters the terms represent NC CTSA supported research areas across the spectrum of translational search from basic biomedical research to clinical research practice. The example projects we've shared today illustrate some of the ways the university libraries are engaging with faculty members, research staff who can't choose and helping them demonstrate impact to improve efficiency for data informed decision-making. As we continue to adapt and sustain these services we're exploring new methods of tools and extending trading to a broader base of library staff. Some of the methods and tools we're exploring for incorporation of our work include analyzing other research data types beyond publications like rep proposals and patents using programmatic and machine learning methods for automating labor processes like record duplication and screening. Continuing to build the text analysis skills as Amanda mentioned earlier including mining of article text to enhance some of the analyses we're currently doing with citations. Finally to build a broader base of expertise and sustain services that will support the growing service demand going forward the libraries are extending training opportunities to all interested library staff. This year the library began explicitly promoting research hub workshops not only to library constituents but to all library staff for the staff development committee. Libraries administrations sponsored a library carpentry training the 30 staff will express interest in their supervisors. And as a follow-up and expansion of that library carpentry workshop each hub staff are organizing a community of practice to facilitate ongoing development of staff knowledge and skills. So finally thank you for your interest in our presentation about the services of UNC Chapel Hill University Libraries that we've been developing to support campus units researchers and just demonstrating impact and decision-making and efficiencies. Please do not hesitate to get in touch just for your questions or comments for us.