Looking for University experts and research collaborators? There's a map for that

Image
KMAP displays the University's expertise in the form of a map, along with researchers' past collaborations, research funding, citations and more. (Illustration by Eddie Canto/Research, Innovation and Impact)

KMAP displays the University's expertise in the form of a map, along with researchers' past collaborations, research funding, citations and more. (Illustration by Eddie Canto/Research, Innovation and Impact)

Image
The map presents the University's expertise as a continent, with different units as countries. Overlays provide information on grant funding and citations.

The map presents the University's expertise as a continent, with different units as countries. Overlays provide information on grant funding and citations.

Image
Stephen Kobourov, professor in the Department of Computer Science

Stephen Kobourov, professor in the Department of Computer Science

Image
Nirav Merchant, director of Data7, the Data Science Institute

Nirav Merchant, director of Data7, the Data Science Institute

Image
The map illustrates researchers' collaborations with lines. Pop-up boxes give details about each researcher's previous grants.

The map illustrates researchers' collaborations with lines. Pop-up boxes give details about each researcher's previous grants.

Image
Md Iqbal Hossain, data research scientist in Research, Innovation and Impact

Md Iqbal Hossain, data research scientist in Research, Innovation and Impact

The University of Arizona has a vast ecosystem of expertise and knowledge, stored in the brains of 16,000 faculty and staff members. Many might say it's one of the University's biggest strengths.

But these experts are spread across more than two dozen colleges, schools, programs and other units.

So, when one expert needs to find another expert for a research project and a simple Google search just won't cut it, where can they go?

Introducing the University's institutional knowledge map, or KMAP.

The site, which was officially launched last month and is accessible to anyone, organizes the University's experts into a map, allowing users to see who has collaborated with whom. KMAP also has overlays that show where grant funding has gone and how many citations researchers have received. A search feature allows users to type or paste in keywords, sentences or even entire articles and abstracts to see a list of colleagues with relevant expertise.

KMAP is supported and funded by Research, Innovation and Impact and the Data Science Institute, known as Data7.

"We are ranked among the top 20 U.S. public research institutions in research expenditures, highlighting our commitment to better understanding our world and improving the quality of life for the people in Arizona, the nation and around the globe through innovative, forward-looking solutions," said Elizabeth "Betsy" Cantwell, senior vice president for research and innovation. "Now, KMAP opens countless more opportunities for researchers from across the University to work together on society's toughest questions."

A question at a workshop

Stephen Kobourov, a professor in the Department of Computer Science, had the idea for KMAP in 2015 after a workshop with other researchers on creating opportunities for collaboration. There was one question that stood out to him: How can we identify gaps in our strategic areas of expertise?

"That led to five years' worth of research projects and papers," Kobourov said.

It turned out to be the driving question behind what KMAP has become.

About a year after the workshop, Data7 had just been launched and Kobourov had built a prototype for the website. The institute, where Kobourov is associate director, had undertaken KMAP as its first project.

"Our goal was to make something that allowed us to build teams quickly on topics that were coming up," said Data7 Director Nirav Merchant, adding that organizations that sponsor research were and still are focused on increasingly interdisciplinary proposals.

Kobourov hired Md Iqbal Hossain in 2016 as a postdoctoral researcher to build out the site using Kobourov's ideas. Hossain is now a data research scientist in RII and manages the development of KMAP.

Before he could start building it out, Hossain needed bio sketches, curricula vitae, lists of grants and publications and other information the site would organize and display. This involved gathering data from internal sources, such as University Analytics and Institutional Research and UAVitae. UAIR collects detailed data about the University's research spending from UAccess Research, which logs expenses related to sponsored research projects every day, said Ravneet Chadha, executive director of UAIR.

Hossain also turned to external sources such as ORCID iD, which tracks researchers and the impact of their scholarship, as well as Google Scholar and Grants.gov. When setting out to get information from these platforms, he began by simply asking the services to provide it. But in many cases, it required building computer programs – known as crawlers – that could comb the web and gather the information autonomously.

The internal University data is what sets KMAP apart from these external sites that offer similar services, Hossain said. For example, KMAP's researcher data includes information from rejected funding proposals, which can't be found elsewhere.

"Even if a proposal has not gotten funded, it still provides a lot of information about someone's interest in a collaboration and their skills," Hossain said.

Once Hossain had built the framework for the site, Data7 provided the infrastructure to make it available to the University community, Merchant said. The institute's web developers contributed code to the site, and its servers gave it a home.

KMAP also has NetID integration, allowing researchers to log into the site, see how it characterizes their work and even submit additional information, such as research profiles, CVs and bio sketches.

Intuitive as Google Maps, better than a Google search

From the homepage of the KMAP site, a menu bar midway down the page takes users to its main tools – the map and the expert-matching search.

The map resembles a continent, with the countries being the various schools, academic departments and other University units, which are called nodes. These nodes share borders based on how many collaborations within the University their faculty have had. The departments of mathematics and computer science, for example, share a border because their faculty members frequently collaborate on research.

Zooming into the map, users can see individual people in each unit. The larger the person's name, the more collaborations they have had over time. Clicking a person's name will show lines connecting them to other names on the map, illustrating their collaborations. Clicking the small pop-up window above the name, which displays the person's title and photo, will bring up a list of their recent grants and publications on the left side of the page.

The map currently includes all faculty, as well as anyone else – such as doctoral students and postdoctoral researchers – with a connection to someone else at the University through a paper, grant or proposal. Those wishing to be included on the map can click "NetID Log In" at the top of the homepage. After logging in, they will be taken to a page with instructions for submitting a request.

A map was the best way to visually communicate the data, Kobourov said, since maps are so widely used and understood. Roads connect cities, and cities that are closer together are often similar. Adjacent cities form a county, counties that share borders form states, and states form countries.

"This concept also works well when you're visualizing social networks or collaborations," Kobourov said. "People understand that if you have a bunch of people in a region, these people are similar in some way."

A key appears at the bottom right corner, along with three overlay options: a visualization of grant funding, a heat map of researchers' citations, and experts' h-index – a metric that measures authors' scientific output and impact.

The homepage menu bar also leads users to the site's expert-matching search tool. The search box allows users to type in keywords, research topics, sentences and even entire articles to find an expert in a particular area.

KMAP's builders point to several potential uses of this feature:

  • Event organizers planning workshops or lectures can paste in the event's abstract to get a list of colleagues who, based on their work, may find the topic interesting. They will also see those colleagues' email addresses so they can contact them.
  • Pasting in a news article about a specific topic can help narrow a search for University faculty members and researchers who are experts in that area.
  • A video on the site demonstrates how researchers can use the tool to find potential collaborators for a funding proposal by pasting in the text from a call for proposals and seeing University colleagues who would be a good match.

Using data to create opportunities

David Glickenstein, professor in the Department of Mathematics and associate head of its graduate program, heard about the KMAP in passing from Kobourov and began using it about a year ago. He said the ability to see expertise on a map is what makes KMAP unique.

The best research is achieved through finding new researchers to work with, Glickenstein said – "connecting in new ways that you haven't before."

"The best way to do that is to find people that have been successful, (through) those small collaborations, and enhance them, use them as jumping off points," he added. "As a researcher and just as somebody trying to promote research at the University, this is a great tool to help us create mechanisms to collaborate."

Glickenstein said he's spent considerable time looking over the map and using the overlays to see who has received significant grants and the most citations. But only focusing on the people with the largest names on the map as potential collaborators is "not necessarily the right way to go," he acknowledged.

"You want to create opportunity to find people and groups that might be interested" in a particular project, he said. "I would hope KMAP could serve as a starting point and then researchers can expand their views from there to promote inclusivity and encourage creativity."

Promising feedback and endless potential

With the site officially live for only a month now, Hossain is in the process of collecting feedback from users and making improvements.

He urges anyone with recommendations to reach out using the site's contact page. Those interested in KMAP-related news and updates can also subscribe to kmap@list.arizona.edu.

He's already received positive feedback from beyond the University. Researchers from Arizona State University, Georgia Tech and San Diego State University have reached out to ask Hossain how their universities could build something similar.

Seeing the potential for others to benefit from the technology, Hossain and the KMAP team are working with Tech Launch Arizona, the University office that commercializes inventions stemming from University research and innovation, on strategies to license the technology to other universities to help them understand their own unique research strengths and weaknesses.

Merchant said he hopes to one day see KMAP integrate the student experience, perhaps by including details of student contributions to research projects. He also envisions a future iteration of KMAP that helps students who are unsure about what they want to study by enabling them to more efficiently find courses that appeal to them.

"What KMAP has done is given us a new way to look at our most critical institutional asset, which is our faculty," Merchant said. "What's missing now is our students. If you think of our faculty as anchors and cities on these maps that together build our University, navigating what they do and introducing that to our students, and then charting their journey through the courses – KMAP allows you to do that."

Resources for the Media