Using cues to forage for information on the Web
| Published date | 04 November 2014 |
| Date | 04 November 2014 |
| DOI | https://doi.org/10.1108/JSIT-01-2014-0003 |
| Pages | 296-312 |
| Author | Stella D. Tomasi |
| Subject Matter | Information & knowledge management,Information systems |
Using cues to forage for
information on the Web
Stella D. Tomasi
Department of e-Business and Technology Management,
Towson University, Towson, Maryland, USA
Abstract
Purpose – The purpose of this paper was to study users’ behaviour when using different search engine
results pages (SERPs) to identify what types of scents (cues) were the most useful to nd relevant
information to complete tasks on the Web based on information foraging theory.
Design/methodology/approach – This study has designed three interface prototypes and
conducted a qualitative study using the protocol analysis methodology. The subjects were recorded and
videotaped to identify patterns of searching behaviours on visualization interfaces of SERPs.
Findings – The study found that users found titles of categories or websites, keywords of categories,
orientation of results and animation are strong scents that users follow to help nd information on
SERPs. If certain scents are not used followed on an interface, then their strength will diminish.
Furthermore, the study showed that simple scent trails are more important to users than complicated
trails.
Originality/value – This study uses a qualitative approach to explore how users behave with
different SERP formats, particularity a visualization format, and identify which scents on the interface
are important for users to follow to successfully complete tasks on the Web.
Keywords Visualization, Search engines, Web browsing, Searching, Interface design, Information
overload, Information foraging theory, Cues
Paper type Research paper
Introduction
As websites continue to grow on the Web on a daily basis, the amount of information
that one has to navigate can be complicated. How to locate relevant and valuable
information is one of the biggest challenges for users. Current search engines use a
textual list, where websites that match the query are ranked and ordered, to provide
users with results to browse for their specic tasks. However, many users are not able to
formulate accurate queries to retrieve the proper results nor are they able to navigate
through all the available information to gain insight (Baeza-Yates and Ribeiro-Neto,
2011). Research has also shown that when completing tasks on the Internet, individuals
stop browsing after 20-30 results from a search engine, whether or not they have
completed their search (Roussinov and Chen, 2001). Another way to explore massive
amounts of information is through information visualization. Information visualization
is a collection of techniques for combining cognitive senses with visual cues that allow
for better understanding of information (Turetken and Sharda, 2004). Visualization
takes large data sets and turns it into a visual map of data so that relationships within
the data set can be recognized. In addition, it allows users to see the entire data set in a
holistic manner rather than individually. Individuals are able to more easily understand
information when it is presented (Dork et al., 2012) in a visual format.
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1328-7265.htm
JSIT
16,4
296
Journal of Systems and Information
Technology
Vol. 16 No. 4, 2014
pp. 296-312
© Emerald Group Publishing Limited
1328-7265
DOI 10.1108/JSIT-01-2014-0003
However, even with visualization, nding relevant information is a difcult task
(Dumais et al., 2001;Roussinov and Chen, 2001;Turetken and Sharda, 2005) because
there is a lack of appropriate cues available to help users navigate through the large
information space that a search creates. Based on information foraging theory (IFT),
scents or cues that are available in the searching environment can lead individuals to
decide whether to exert time and energy navigating through the information or to make
better use of resources elsewhere in another set of results (Pirolli and Card, 1999).
Information scent characterizes the individual’s use of environmental cues in judging
which information sources are important while navigating through results in a search
environment (Pirolli, 2007). It is important to understand users’ behaviours when
completing tasks on search engine results pages (SERPs) to identify which scent trails
allow users to be successful. According to Dumais et al. (2001), there is little research that
identies which cues are the most desirable in the visual presentation of search results.
Card et al. (1999) suggest that other features, such as animation, have been underutilized
in visualization. In an analysis of Web space visualization, Turetken and Sharda (2007,
p. 77) state “a great majority of the systems surveyed […] do not use animations in spite
of the technical feasibility of these visual aids”. Therefore, in this study, we investigate
how users interact with visualization formats to complete Web tasks to identify which
cues help nd relevant information. This is a rst attempt in a study that draws on IFT
to understand which scent trails are more useful than others to help users retrieve
information using visualization formats.
Previous literature
A number of studies report that insight and problem-solving performance can be
improved with appropriate visualizations (Dork et al., 2012;Clarkson et al., 2009;Crapo
et al., 2000).Wise et al. (1995, p. 51) developed an application that transformed text
content into a spatial representation that enabled enhanced visual browsing and
analysis. They stated that the application enabled users to avoid language processing,
thereby reducing mental workload. Their contention was that this creates “an
interaction with text that more nearly resembles perception and action with the natural
world than with the abstractions of written language”.
Ahn and Brusilovsky (2013) introduced an interactive visualization information
retrieval system called Adaptive VIBE. They tested the adaptive visualization against a
baseline system. The adaptive VIBE visualizes documents based on similarities to
keywords. Users can manipulate and interact with the visualization. The authors found
that the visualization system had a better quality performance than the baseline to
visualize the results and nd quality documents.
Wu and Hsu (2013) proposed a novel three-dimensional (3D) information
visualization method that allowed users to control the details with how they viewed the
data through their design of the Magic Eye View. They conducted effectiveness of the
tool through simulations and found this approach to work better than existing
approaches. Hou and Pai (2009) developed a knowledge visualization methodology that
converts knowledge into a visualization format. The authors were able to evaluate the
effectiveness of the visualization format through a case study.
Wise et al. (1995) developed two different visualization approaches to the same
information in a document repository. The Galaxies visualization clusters similar
documents together into a two-dimensional (2D) scatterplot of “docupoints”. Users were
297
Using cues to
forage for
information
To continue reading
Request your trial