Information Technology Reference
In-Depth Information
that enabled yoked browsing and telepointers; collaborators could also associate comments with
jointly-viewed pages.
A variety of social media can also be used to support remote collaboration on a search task.
Social media enables users to share links or thoughts about online content with others; like co-
browsers, these systems are not specifically targeted at supporting Web search, but can be used to
support some related functionality such as link-sharing or commenting. For example, the Sociable
Web ( Donath and Robertson , 1994 ) and Community Bar ( http://www.communitybar.net ) are
systems that allow a user to know when others are currently viewing the same webpage; they can
then initiate a chat with those users about that page's content. Systems for sharing bookmarks
include WebTagger ( Keller et al. , 1997 ), Wittenburg et al.'s system ( Wittenburg et al. , 1995 ), and
commercial sites such as del.icio.us ( http://del.icio.us ) .
Users can also participate in remotely-guided search experiences, such as receiving
help from a reference librarian ( Taylor, R. , 1968 ); online services such as QuestionPoint
( http://www.oclc.org/questionpoint ) enable this experience even when the librarian is re-
mote. Some search engines, such as Cha Cha ( http://www.chacha.com ) route users' queries to
human beings who help them find answers, although such systems are often designed as cost-effective
ways to simulate natural language processing rather than as true collaborative tools.
There are also systems designed specifically to support remote collaboration on search tasks.
Such systems generally provide some support both for working together on the process of searching
(e.g., selecting query terms, selecting search sources) as well as for collaborating on the products of
searching (e.g., identifying content that is most relevant).
Some collaborative search tools are designed for special-purpose databases. For example,
C-TORI ( Hoppe and Zhao , 1994 ) allows a group to collaborate when searching over a relational
database, and MUSE [Krishnappa, 2005] and MUSE ( Krishnappa, R. , 2005 ) allows pairs of users
to search over a medical database.
Systems have also begun to emerge to support more general purpose collaborative search over
the Web. SearchTogether ( Morris and Horvitz , 2007b ), discussed in more detail below, is one such
system. HeyStaks ( Smythetal. , 2009 ) (Figure 4.1) is another example; HeyStaks allows users to
share “staks” of relevant content with collaborators with similar interests, and the results they click im-
pact the ranking of results when their collaborators execute similar queries. S 3 ( Morris and Horvitz ,
2007a ) allows users to asynchronously collaborate on search tasks by treating a search session (in-
cluding keywords, results, and comments on found pages) as a file that can be saved and loaded
using the special S 3 browser; these files can be emailed among collaborators in a manner analogous
to other types of office documents. CIRE ( Romano et al. , 1999 ) (Figure 4.2) allows users to search
singly, but enables them to attach comments to the pages they have found, which are then visible to
other group members who encounter those pages later.
Implicit information about collaborators' searches, including clicks, queries, and other actions,
can also be shared to group members to provide increased awareness of other's activities. For example,
with group hit highlighting ( Morris et al. , 2008 ), not only are an individual's query terms highlighted
Search WWH ::




Custom Search