The Visualization of Researcher Impact Metrics in a Variety of Display Environments
By: William H. Mischo, Mary C. Schlembach, and Saajan Dehury
The UIUC Library has a significant body of experience in creating services that support large-scale visual and immersive environments for scholarship and instruction. In 2015, the UIUC Library remodeled approximately 9,000 square feet of space in the Grainger Engineering Library Information Center (GELIC) to create the Grainger Engineering Library IDEA (Innovation, Discovery, DEsign, and DAta) Laboratory. The IDEA Lab provides facilities and services for faculty/student collaboration, innovation, and entrepreneurship activities and supports a wide range of digital scholarship, design learning, and visualization functions. The Lab serves as a demonstration and prototype site for exploring the interface between informatics, design thinking, visualization, and data analytics. It also serves as a node in a multi-facility campus design learning network and will complement services offered in the new Siebel Center for Design (SC4D) scheduled to open in 2019.
The key IDEA Lab visualization components are: a 13.5’ by 7.5’ visualization wall with a 4x4 array of 46” monitors with an aggregate 7680 x 4320 (33 million pixels) high-definition display at a true 8K; two informatics laboratories with a total of 10 movable 58” or 65” large screen monitors with attached workstations; two presentation areas with paired 84” monitors; and a Virtual and Augmented Reality Room with HTC VIVE, Oculus, and HoloLens systems. Currently under construction is a smaller 4x4 visualization wall with 55” monitors that will utilize the same software and development platforms as the IDEA Lab’s large wall. We envision these smaller walls as a type of “commodity wall” that can be placed in various areas of the University of Illinois Library system.
IDEA lab staff have been collaborating with various University of Illinois early adaptor groups on project activities. These include: the Technology Entrepreneur Center Fellows and students; the iVenture Accelerator interdisciplinary student design groups; students and instructors in the Mechanical Science and Engineering and Electrical and Computer Engineering capstone senior design classes; Department of Computer Science interface design classes; and two additional for-credit classes: Physics 498: Where Art Meets Science, and an iSchool at Illinois class Entrepreneurial IT Design. Other student/faculty groups have made use of the Lab and the IDEA Lab staff have developed a number of supporting software modules for these groups.
University faculty and administrators are increasingly interested in (some would say preoccupied with) research impact measures and assessment. Research impact metrics play an integral role in the scholarly communication environment and are an important element in institutional budgeting, promotion and tenure decisions, grant applications, publicity efforts, and many other activities. The individual scholar uses impact metrics to secure grant funding and to justify the budget and resource needs of their research group. These metrics include publication totals, citation counts, article and paper download numbers, the number and total budget of grants received, co-author totals, Relative Citation Ratio, and H-Index and other productivity measures. In addition, a number of altmetric measures are typically being collected and recorded at the publication and author level.
Libraries and librarians have an important role to play in collecting and disseminating researcher metrics. Libraries are utilizing a variety of available information resource services to produce tools that provide access to research impact data and also display impact metrics information. In particular, libraries are developing visualization tools and dashboards that process and interpret impact metrics data within high-definition and specialized displays.
For the library, these visualizations are important for gathering data on scholarly information patterns, facilitating faculty liaison work, assisting with requests for promotion and tenure support data, and working with faculty groups on grant applications. The University of Illinois at Urbana-Champaign Library has developed custom web-based visualization tools that provide contextual and comparative researcher productivity data. This is being tested on several faculty groups within the new Carle Illinois College of Medicine. We have developed a visualization and dashboard system that displays research impact data for the Illinois Cancer Center (ICC) cohort of 73 researchers. A screen shot of the ICC dashboard visualization is shown in Figure 1.
This particular application has been developed to support a National Institutes of Health (NIH) grant application that requires cohort publication and co-author collaboration information be included in the proposal.
The visualization is built from a single relational database table comprised of two columns containing researcher names and Scopus ID numbers. Using the Scopus API, a series of scripts are executed to compile a supporting database with publication, citation, and co-author tables used for the subsequent visualizations.
The result is a dashboard display of clickable faculty name bubbles and number labels that provide publication numbers over the last five years, citation information (a cited by number), NSF and NIH grant totals, and the number of co-authors within the cohort. The bubbles are scaled to the number value range within each of the four reported metrics. Clicking on the publication bubble produces a display of matching article and paper matches with metadata extracted from Scopus. The article metadata includes links to the full-text (via the DOI), the Scopus short record page, and a citing articles display in Scopus.
The citations bubble displays currently links to the Scopus author information page which contains the author’s H-Index, the number of times the author has been cited, a link to all articles citing the author, and a link to an author analysis display.
The grants link displays matches from a locally generated database of NSF and NIH grants awarded to the University of Illinois. The grant display record provides a link out to the NSF and NIH sites grant description page.
The co-authors display bubble links to another visualization which shows co-authors within the group and a clickable label with the number of co-authored papers between each two authors. An example visualization is shown in Figure 3.
To maximize the usage of these impact metric dashboards, it is important that this visualization be available in multiple screen display configurations and resolutions. It must be available for standard screen displays on workstations, laptops, and tablets and also on large visualization walls. The work we are performing within the Immersive Scholar grant program is seeking to address this issue of adapting web site displays to multiple screen resolutions and sizes. We are developing mechanisms to easily switch between standard display configurations and configurations that allow optimal viewing in a high-definition visualization wall environment. The approach we are currently investigating employs a server-side scripting solution where the target URL contains a Querystring value pair (e.g. view=1920 or view=7630) that indicates the horizontal pixel value of the display monitor. The server-side code then constructs a display that optimizes the screen layout and presents the user with a dashboard that fills the display area. We are also investigating the use of the Viewport parameter within the Scalar Vector Graphics (SVG) displays that we are using in these dashboard applications. The Viewport values and the SVG size and coordinate values can be modified on-the-fly in conditional statements in the server-side code. This approach is very promising and we expect to be able to test this in a variety of Immersive Scholar cohort sites in the near future.