Assistant Professor Jacek Gwizdka (left) and doctoral student Michael Cole will use eye-tracking technology, galvanic skin response, and electroencephalography to help detect what search engine users find relevant.
Jacek Gwizdka, assistant professor of library and information science, has received more than $70,000 from search engine giant Google to conduct studies measuring physiological changes in searchers to detect when they find relevant search results.
The number of people who search for information on-line and the speed with which they expect results have both grown astoundingly in the past couple of decades.
Before the era of Google, information searching was, by and large, limited to specialists, and was much less speedy.
“Back then, if you worked in a research institution, you would tell your librarian what you are looking for,” Gwizdka said. “This person would do searches in databases, may contact you to clarify something, do more searches, and possibly after some days give you a list of articles that may contain the information you wanted.”
Research methods to study searching have changed as well – in the past, information scientists relied on self-reports from users as to what they found most relevant in search results. But Gwizdka said those results are often not reliable because they are subjective and self-edited.
“My interest is in trying to learn about searchers using implicit techniques,” Gwizdka said. “I want to learn what they are doing without asking them.”
Gwizdka and doctoral student Michael Cole will use eye-tracking technology, galvanic skin response (GSR), and electroencephalography (EEG) to measure pupil dilation, hand sweat, and brain activity, respectively. Their project is titled “Implicit Detection of Relevance Decisions and Affect in Web Search.”
The use of EEG is new to the field of information science, Gwizdka said. Neuromarketing and neuroeconomics researchers have used EEG for several years to study consumer behaviors and investor decisions. But as EEG headsets have become commercially available to a greater number of people – they are increasingly being used in video gaming – it is easier for researchers to incorporate this technology into their studies.
Using EEG, eye tracking, and GSR technology gives researchers additional signals that indicate how people respond to the search process while they are in the flow of search. “This is a new frontier of data for information science researchers,” said Gwizdka. That knowledge can be used to build search systems that can adapt automatically to users while they are searching. “We’re looking to find out why people click – whether they click because it is something interesting to them or because they want to confirm it is not what they want. That's the point of the project,” Gwizdka said.
"This Google funded project paves the way for our quest for additional external funding for this innovative line of research," he added.