Search engines are designed to help people get things done: find a local business, plan a vacation, or understand an unfamiliar concept. This focus is demonstrated by how search businesses measure their own performance–by how quickly a user find the page they were looking for. It’s considered bad if someone clicks the back button to return to the search results.
But this attitude ignores a growing portion of searches: those performed when people have a few moments to kill and want to discover something entertaining or amusing–for example, when a user searches for “funny pictures” or “interesting new documentaries.”
New research suggests that many people use search engines this way, and their behavior is fundamentally different from other searchers’ behavior. Understanding this difference might provide new ways for search engines to attract more users and drive more traffic.
Researchers from Swansea University in Wales and the University of Erlangen in Germany presented the results of this research last week at the Human Computer Interaction and Information Retrieval (HCIR) workshop in New Brunswick, New Jersey.
Max Wilson, a lecturer in the Future Interaction Technology Lab at Swansea University, says that around 90 percent of research into Internet search focuses on improving goal-directed searches–in other words, the kind of search you might do at work. He expects to see more people searching the Internet for a quick moment’s entertainment, especially as the number of Internet-connected devices increases. To best serve those people, Wilson says, companies and researchers should look for more ways to provide welcome serendipity through search.
To understand how people use search engines to entertain themselves, Wilson and colleagues carried out two studies. The first involved asking users to keep a diary of how they searched while they watched television–the assumption being that those searches would likely be more entertainment-focused. For the second, the researchers mined Twitter for tweets containing words such as “browse” or “explore” and narrowed in on those related to “casual searching” (2.4 million unique posts over the course of five months).
They found that when searching casually, users were less interested in getting away from search results, explains David Elsweiler, a Humboldt Research Fellow in the department of computer science at the University of Erlangen. Elsweiler was involved with the two studies. He says people seemed intent on producing a certain mood: They used such terms as “interesting,” “entertaining,” “distracting,” or “challenging” to describe what they sought. “These are very subjective descriptions, and search engines are not good at dealing with this kind of task,” he says.
Elsweiler says this type of behavior explains the appeal of sites such as StumbleUpon, which recommends new websites. He thinks that if search engines adjusted for this type of search, they could be easier to use, particularly for people who don’t have much experience with the sort of goal-oriented searching that’s often done at work.
Robert Stebbins, who researches the way people spend their free time, notes that historically, information scientists have neglected to design tools for people who search just for fun. Stebbins says people look for several sorts of things when searching the Web this way: sensory stimulation, interaction with others, and active entertainment–such as games and videos. Theoretically, he says, search tools could use these needs as a framework for classifying links.
Search engine research has traditionally focused elsewhere, says Daniel Tunkelang, an engineer at Google who is an expert on information retrieval. “It would be easy to dismiss [casual search behavior], except for the fact that people are doing it,” he says.
Some sites, such as YouTube and Amazon, already help users find completely new content, Tunkelang adds. Both sites have strong search tools for those who know what they want, but for those interested in meandering from one thing to another, they offer recommendations based on what other users liked, links to content made by the same people, or links to similar content.
Wilson says it may be possible to incorporate insights about casual searching into the design of traditional search tools. “To some extent, it’s up to systems to recognize and facilitate people who are putting themselves out there with different motivations, like to have fun,” he says. “It may be feasible, as systems like Google do already with types of target content, to track ‘fun’ queries and assume a little more about casual intentions.”
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Driving companywide efficiencies with AI
Advanced AI and ML capabilities revolutionize how administrative and operations tasks are done.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.