Web users re-visit in steps

By Kimberly Patch, Technology Research News

Half the battle of finding information on the Web is getting back to a page you've already seen.

The Web has long spurred researchers to study how people initially find information, but the tactics people use to get back to previously discovered information remain less understood.

Researchers at Virginia Polytechnic Institute and State University are examining how people relocate information rather than how they find it the first time. "There is evidence that users usually can re-find a page they're looking for, but we hope to guide the development of computer tools that can make the re-finding process easier, especially if the user is mobile," said Robert Capra, a researcher at Virginia Tech.

The researchers' study showed that people tend to use a two-stage process to find information they have seen before, that they use domain information and context to move closer to a goal, and that annotations make things easier.

The results could lead to tools that would help users re-access Web pages more quickly and easily, including finding the same information using devices ranging from desktop computers to mobile phones. "Imagine that you were preparing to go on a trip and had browsed a number of restaurants, events and attractions that you were interested in," said Capra. "Once you're on the trip, such a system can help you re-find that information on your PDA."

To study how people re-find information, the researchers asked six subjects who were familiar with Web browsing to carry out a set of tasks using the Web, including finding movie theater show times, restaurant phone numbers and addresses, and tourist event information. The subjects also did a freeform search for information. The subjects took notes during the searches using an annotation tool.

In a second session about a week later, the six subjects were each paired with six additional users, or retrievers, with instructions to direct the re-finding process by phone. The retrievers had access to the original subjects' annotations and browser history logs of the searches completed a week earlier, but the original user did not.

The users' naturally broke the re-finding process into steps, and the process often consisted of two stages, said Capra. In the first stage, users tried to get back to a particular page they remembered or thought would be useful. In the second stage, the subjects asked specific questions about the information they were trying to re-find. "At times, [the subject] provided very partial information, allowing the retriever to reach a specific location before making the next request," said Capra.

The subjects used information they remembered to move incrementally closer to the information being sought, said Capra. In more than three-quarters of the re-finding tasks, users tapped waypoints, or Web pages they remembered from along the paths they originally took to find information, he said. Sometimes subjects remembered an exact URL, but other times used a title or just a description.

The research could be applied to Web tools to more easily help users find what they have previously seen on the Web, said Capra. One possibility is software tools that support the approach of making incremental progress toward a goal, he said.

The results are not particularly surprising, but the work helps augment the body of science surrounding the issue of finding information on the Web, and the use of the telephone is interesting, said Marti Hearst, an associate professor of information management and systems at the University of California at Berkeley. "The results are useful in that the study provides another empirical data point to support the assumption that search tasks are done progressively and incrementally, and often involve first locating a source and then navigating that source," she said.

There are two design and analysis flaws in the study, however, said Hearst. The users' are explicitly told that they will be asked to find information again, and the study assumes that people will instruct others to find something in the same way that they would on their own, she said. "It could well be the case that people... break it down into pieces to better help the other person keep track of the different aspects of a task."

The researchers' next step is to run a study that looks at the difficulty of tasks and whether the users' familiarity with a task affects the approach a user takes to re-find information.

They're also planning to build a prototype Web browser add-on aimed at helping users re-find information, said Capra. The prototype could be ready next year, he said.

Capra's research colleague was Manuel A. Pérez-Quiñones. The research was funded by the National Science Foundation (NSF) and IBM.

Timeline:   < 2 years
Funding:   Corporate; Government
TRN Categories:  Databases and Information Retrieval; Internet
Story Type:   News
Related Elements:  Technical paper, "Re-Finding Found Things: an Exploratory Study of How Users Re-Find Information," posted on the Computing Research Repository (CoRR) at http://arxiv.org/abs/cs.HC/0310011


February 11/18, 2004

Page One

Light-storing chip charted

Coincidences set up mental error

Noise boosts nanotube antennas

Web users re-visit in steps

All-plastic display demoed
DNA sorts nanotubes
Electricity teleportation devised
Mechanical storage goes low power
Scientists brew tree-shaped DNA
Magnets tune photonic crystal


Research News Roundup
Research Watch blog

View from the High Ground Q&A
How It Works

RSS Feeds:
News  | Blog  | Books 

Ad links:
Buy an ad link


Ad links: Clear History

Buy an ad link

Home     Archive     Resources    Feeds     Offline Publications     Glossary
TRN Finder     Research Dir.    Events Dir.      Researchers     Bookshelf
   Contribute      Under Development     T-shirts etc.     Classifieds
Forum    Comments    Feedback     About TRN

© Copyright Technology Research News, LLC 2000-2006. All rights reserved.