Introduction - Collaborative Web Search: Who, What, Where, When, and Why - page 6

cilitates suspension and resumption of search tasks; we explore the S 3 system ( Morris and Horvitz ,
2007a ) as an in-depth example of this approach.
As an example of synchronous collaborative search, a group of coworkers may gather around a
computer to research a topic of interest, with one person controlling the mouse and keyboard and the
others offering comments and suggestions, as was done by Martha and George during their search for
common asthma irritants. Many systems have been designed to support synchronous collaborative
search within specialized domains or with specialized devices (e.g., TeamSearch ( Morris et al. , 2006 ),
C-TORI ( Hoppe and Zhao , 1994 ), and MUSE ( Krishnappa, R. , 2005 )). Other systems that support
synchronous collaborative search are more general, such as SearchTogether ( Morris and Horvitz ,
2007b ). The different temporal configurations impact the options collaboration tools should offer to
searchers. Studies of the CoSense system ( Paul and Morris , 2009 ), during both asynchronous and
synchronous collaborative search scenarios, have found that these two collaboration styles benefit
from different types of sensemaking support.
1.3.5 WHY?
Research can also give us insight into why people work with the group members they do during
collaborative search activities. We begin this chapter by discussing how people are brought together
because of a shared interest in a common topic. For examples, Martha's asthma diagnosis makes
the topic of interest to her family members, and it inspires the group to work together to learn
more about the disease. Tools like question-answering systems (e.g., Ackerman and Malone ( 1990 ))
and expertise-finding systems (e.g., Bernstein et al. ( 2009 )) help bring together people with shared
interests.
But people also collaborate during search for social reasons. Beth may help her sister find
information about her asthma not just because Beth is interested in learning more about asthma, but
also because it helps bring her closer to her sister, makes her feel good about herself, or even makes
it more likely that Martha will help her in the future with some task that interests her. Systems can
support the social aspects of collaborative search by providing awareness for all group members of
how individuals contribute to the shared goal, much as is done on Q&A websites ( Raban and Harper ,
2008 ).
Although we focus in this lecture on intentional collaborations, there are many examples
where people implicitly collaborate with others on search tasks. For example, the links people
click ( Joachims, T. , 2002 ) and the queries people issue ( Kleinberg, J. , 1999 ) feed back into many
commercial search systems to improve the experience of future people searching on the same topic.
Understanding when implicit groups stand to benefit from collaborative interventions provides
opportunities for systems to suggest serendipitous group formation, creating the opportunity for
implicit collaborations to transform into explicit ones.
In the following chapters, we dive more deeply into the who , what , where , when, and why of
collaborative search. It is our hope that in doing so, we give insight into how emerging solutions
might best address collaborators' future needs.