Searching for Better Results: The IISLab at SILS

December 15, 2018

At SILS’ Interactive Information Science Laboratory, researchers study how people search for information and interpret results, and develop new tools and approaches to help people find what they need faster.

If you pass through the Interactive Information Science Laboratory (IISL) on the ground floor of Manning Hall, you might hear someone mention “the matrix.” The phrase is not a reference to the Keanu Reeves’ sci-fi movies, but rather a nickname for a new tool under development that will allow people to drag and drop information from various websites into a grid for comparison across dimensions. For instance, if you were shopping for a car, you might compare gas mileage, price, cargo space, and other features.

Jaime Arguello

“The matrix” is one of several ongoing projects at the IISL aimed at learning more about how people search for information and improving that process. UNC School of Information and Library Science (SILS) Associate Professors Jaime Arguello and Rob Capra lead the lab, which includes nine doctoral students, as well as some master’s and undergraduate students assisting with research. Arguello and Capra have each received National Science Foundation (NSF) CAREER Awards to support their research, as well as a $500,000 NSF grant in 2017 for a project focused on search assistance. In 2018, Arguello and Capra were named Francis Carroll McColl Term Associate Professors at SILS. The two-year professorships recognize faculty excellence and provide salary supplements and funds for research and travel.

Arguello’s CAREER-funded project focuses on aggregated search systems, those that pull results from a variety of other systems and package them together. Previous research has primarily examined which results to present, but Arguello is investigating how results are presented and the impact that can have.

“We’ve been looking at different cognitive abilities that people have and trying to understand how they influence the way they search and to see if certain layouts are better for people with certain abilities or certain skills,” he said. “For example, if you have low perceptual speed that limits how quickly you can scan a visual display, you might benefit more from a layout that’s blocked with clear separations.”

Rob Capra

Arguello and Bogeum Choi, an MSIS 2018 graduate who started the PhD program in August, conducted a study in the fall of 2017 to test some of these scenarios. Arguello analyzed the data in the spring and they are now working to publish an article with the findings. Arguello’s CAREER grant has also been facilitating research by PhD student Sandeep Avula, who is investigating how integrating search engines into the Slack messaging platform affects how people search and evaluate results. 

As part of his CAREER project, Capra has been looking at how people structure information when they attempt a complicated search. He conducted a study that asked people to take notes on paper while they were searching for information about a topic, first for themselves and then for another person, to see how the organization might differ depending on the intended audience. Among other findings, the analysis revealed that people generally use relatively simple structures to organize information, relying on lists and groups rather than complicated hierarchies or concept maps. Capra collaborated with doctoral students Anita Crescenzi, Yuan Li, and Yinglong Zhang on the study, and the team is now working to publish the results.

Another thread of Capra’s CAREER project looks at how structures and organization can help people as they search. One structure that Arguello has been helping to develop is “the matrix,” a tool that supports comparative tasks by letting you drag and drop information into a grid, bookmarking the information’s original source, and providing you with a visual representation of you progress. PhD student Yuan Li is planning to look at how the tool might help with task resumption, and Austin Ward and Kelsey Urgo are contributing to the project.

“One aspect of this is that it can help people keep track of what they’re doing,” Capra said. “Another aspect that would be a little more challenging is if you filled in part of the matrix, could the search system use what you’ve added to fill in the rest. Could it find data that you were having trouble locating on your own?”

For their joint NSF project that launched last year, Arguello and Capra are examining various search assistance tools, including “search trails,” guides that can assist in a current search by displaying the steps someone else took in a previous search. They are also looking at how people use different types of information for different tasks. In a recent study, they gave people simple fact-finding tasks and complex tasks in which they had to create a completely new solution to a problem. A tool developed by PhD student Austin Ward offered participants four different types of information – facts, opinions, insights, and concepts. The team is still analyzing the results, but the outcome may ultimately enable systems to offer more appropriate results depending on the type of task someone is attempting to complete.

“To me all of this falls into the same category,” Capra said. “We’re trying to take information that has come out of someone else’s search and see if it can be helpful to someone in a future search.”


  • Sandeep Avula
    Information retrieval, human-computer interactions, collaborative search, and conversational systems.
  • Bogeum Choi
    Information seeking behavior, task-based information retrieval, human factors (cognitive and affective factors in HCI), search as learning.
  • Anita Crescenzi
    Human-computer interactions, interactive information retrieval, collaboration and information seeking and behavior, especially with regard to health.
  • Heejun Kim
    Credibility assessment of health information in health-related online communities, dynamics of social and information networks, data analytics, text mining, and machine learning.
  • Yuan Li
    Human-computer interaction, interactive information behavior, collaborative search, information/knowledge organization, and knowledge re-use.
  • Kelsey Urgo
    Human-computer interaction and interactive information retrieval, particularly in relation to conversational search systems and shared control.
  • Austin Ward
    Human-computer interaction, information architecture, and information in virtual and augmented reality.
  • Shenmeng Xu
    Scientometrics, scholarly communications, altmetrics and related areas.
  • Yinglong Zhang
    Human-computer interaction, collaborative search, data science, eye-tracking.

Students at an IISL lab meeting in November. From left: Yu Yuan (master’s student), Hanlin Zhang, Yuyu Yang, Austin Ward, Kelsey Urgo, Yuan Li, Sandeep Avula, Gaby Matalon (undergraduate student), Yinglong Zhang, Heejun Kim, and Cassie Liu (master’s student).