Introduction - Collaborative Web Search: Who, What, Where, When, and Why - page 1

CHAPTER 1

Introduction

Web search has become one of the prominent information behaviors of the new millen-
nium ( Fallows, D. , 2008 ). Web search refers not only to the process of entering keywords into a
search engine site, but it also includes a related ecology of online information-seeking activities,
such as browsing to specific URLs, making sense of found content, iteratively revising a query, and
disseminating results. Web searches may be accomplished with a single query, or they may span
multiple work sessions. For example, exploratory search tasks ( White et al. , 2006 ; White and Roth ,
2009 ) can potentially last days, months, or even longer.
Traditionally, the tools that facilitate Web search have been designed for solitary use. Web
search engines, for example, are designed for a single person to enter a few keywords related to what
they are looking for, and then sift individually through a list of the results the search engine returns.
More generally, when people search for information online, they typically use Web search engines or
other sites merely as starting-off points to orienteer to an information target ( Teevan et al. , 2004 ),
but even such broader considerations of the use of Web browsers and link following for information
seeking are primarily studied as and designed for the single user scenario.
Despite Web search's focus on the individual user, there are many situations in which people
choose to collaborate on a shared search task. For example, students may need to jointly research a
group homework assignment, family members might seek information about a loved one's medical
condition, or friends might plan vacation travel together ( Morris, M.R. , 2008 ). The process of more
than one person searching in collaboration with others for a shared goal is called collaborative search .
When groups of people searching with a shared goal are able to effectively collaborate on Web
search tasks, they can enjoy several advantages over solo searching, such as increased coverage of the
relevant information space, higher confidence in the quality of their findings, and greater productivity
due to a reduction in unnecessary redundant work ( Morris, M.R. , 2007 ). Tools designed to support
collaboration as a first-class aspect of Web search have the potential to offer substantial benefits over
status quo tools.
However, a one-size-fits-all solution to collaborative search seems unlikely. The term “col-
laboration” encompasses a variety of work configurations that vary along several dimensions, such
as who comprises the collaborative group and how they relate to each other (symmetric or asym-
metric), what aspects of the search they need to collaborate on (the process or the product), where
the collaborators are located (co-located or remote), when the collaboration occurs (synchronously
or asynchronously), or why the group has been formed (explicitly or implicitly). In this lecture, we
review the body of work done to understand the needs for each of these work styles and to design
and evaluate collaborative search systems for each scenario.