25-08-2017, 09:32 PM
SparkClouds: Visualizing Trends in Tag Clouds
INTRODUCTION
Tag clouds are a text-based visual depiction of tags (or words),
typically used to display the relative tag frequency, popularity, or
importance by font size. They can also serve as a visual summary of
document content. In the last decade, tag clouds have proliferated
over the web. They are now a common visualization in news sites for
displaying the most active news story themes [1], photo sharing sites
for conveying the distribution of image content [15], and social
bookmarking sites for showing popular tags [6].
RELATED WORK
The origin of tag clouds goes back to 1976 when an experiment was
carried out by Stanley Milgram [14]. A collective “mental map” of
Paris was created using font size to show how often each place was
mentioned as a landmark in the city. In 1997, Search Referral
Zeitgeist was created by Jim Flanagan as a way to visualize the
number of times a term was used to find a given website by font size.
Among high-profile websites, Flickr [15] used tag clouds first,
followed by other Web 2.0 sites (e.g., Del.icio.us [6]) [3]. For more
details about the history of tag clouds, see [25].
DESIGNING SPARKCLOUDS
The basic idea behind SparkClouds is to retain the advantages of tag
clouds while incorporating minimal but sufficient indication of
trends for a reasonable number of related tag clouds. In particular,
we focused on these advantages of tag clouds: