Overview
We present a redesign project where we replace browser tabs with search trails to support open-ended search tasks.
Motivation
Browsers are great. They let us navigate to websites and find all sorts of information, from todays’ weather to the history of physics. Browsers are very good at taking us to where we would like to go online, and at keeping a faithful record of where we have been to. However, the classic browsers are not designed to capture the thinking process behind a set of related browsing records.
Sometimes, we use the browser to perform quick, one-off search or navigation tasks, like reading tweets, checking the weather, or looking for an address of a restaurant. There is not a lot of thinking process worth capturing behind these kind of tasks. Other times, we use the browser to perform more open-ended tasks, like collecting background information for a blog post, planning a trip, or searching for gift ideas for the next Christmas. This kind of tasks often involve a complex thinking process where we keep refining and coming up with ideas that we would like to explore and collecting relevant information for using the browser. The route one has taken to arrive at a destination may be equally or even more valuable than the destination: it reflects the facts and options we have considered to arrive at the final decision and contains leads to alternatives that we have not yet explored. Indeed, previous resesarch[1] has shown that search trails could provide more topic coverage and diversity than the search destination alone. The guiding hypothesis behind this redesign project is that a redesigned browser where search trails are used in places of tabs can help us become better at open-ended sesarch tasks.
Design Overview
When redesigning the browser with search trails, we followed the following design principles:
1. The search trails in the redesigned browser should support any operation that can be done with tabs in the classic browser.
2. The search trails should arrange the sites opened by the user in a logical sequence that reflects the route one has taken during the search process.
We explored two alternative designs with different metaphors: a time-line design and a file-hierarchy design.
Research Questions
We would like to address two overarching research questions with the designs presented here:
1) Can the redesigned browser with search trails help users complete open-ended search tasks faster and with less cognitive effort?
2) Is one of the two alternative designs we explored better than the other?
Design Features
1. Brownser tabs are replaced by nodes representing webpages. Nodes are arranged according to the search trail structure.
2. Search trials are automatically expanded when users navigate to new webpages. When a new webpage is opened, a node representing the new webpage will be appended to the webpage that was active before the opening of the new webpage.
3. Users can “archive” a webpage that is not immediately useful but may still worth revisiting later during the search session. The "archive" action functions as a temporary bookmark that is specific to a search session.
4. Users can turn a sub-trail into an individual trail.
Concept Sketches
Time-line design
File-hierarchy design
In both designs, the user can click on a node to show the webpage it represents. Only the name for the active webpage is shown by default, but names for other webpages will be shown when the corresponding node is moused over.
Example Scenario: plan a trip to Vermont
Finally, we show below a set of mockups demonstrating the search trails in action with a hypothetic scenario of me planning a trip to Vermont. If the example is not displayed properly, click here to open it.
References
[1] White, Ryen W., and Jeff Huang. "Assessing the scenic route: measuring the value of search trails in web logs." Proceedings of the 33rd international ACM SIGIR conference on Research and development in information retrieval. ACM, 2010.