Taggram: Exploring Geo-Data on Maps through a Tag Cloud-based Visualization
Dinh-Quyen Nguyen, Heidrun Schumann
Institute of Computer Science, University of Rostock
{nguyen, schumann}@informatik.uni-rostock.de
Abstract
Searching and exploring on digital maps are
normally performed through simple text boxes and zoom-
and-pan interfaces. In this paper, however, we present a
novel technique, namely Taggram, which combines texts
onto maps to support geo-tagged data exploration. It is
designed to show geo-tagged data in form of size-varied
and colorized tags, benefiting by the concepts of Tag
Clouds, and to support exploring those data interactively
through a fisheye menu adaptation. The technique was
demonstrated for geo-tagged data exploration; however,
as other thematic geo-data can be partially represented in
abstract form of texts, Taggram can be a method for the
presentation and exploration of such geo-data in other
application scenarios.
Keywords--- Text Visualization, Tag Clouds, Geo-
Tagging, Map Labeling, Cartography.
1. Introduction
Text, the special kind of visual artifacts commonly
used for social communication [1], and geospatially
referenced data, which exist in almost all our everyday
life topics [2], have been the subjects for many studies in
information science as well as information visualization
(InfoVis). Nevertheless, the combination of the two is not
commonly found in InfoVis due to the different
characteristics of text visualizations and geo-
visualizations, except, for instance, traditional techniques
on label placement in digital cartography [3].
In contemporary geovisualizations, maps with
conventional zoom-and-pan interfaces play the central
role, where the popular way of representing aspatial (aka
thematic or attribute) data is using cartograms and other
cartographic techniques, leveraging by visual metaphors
for the encoding of the data (i.e. using choropleth color
and other Bertin’s visual encodes [4]). However,
according to Jo and Ryu [5], it is still sometimes
insufficient for the users to explore geospatial data just
through zoom-and-pan interfaces. Consequently, we
exploit texts as a special metaphor for a supplemental
technique for geo-data exploration. The idea is based on
the 1976 Milgram’s “collective mental maps of Paris” as
he showed a collection of tags in various font sizes
representing how often landmarks in Paris were
mentioned in a survey (Fig. 1) [6]. Milgram’s work can be
seen as the first example of Tag Clouds, the visualization
which is nowadays very popular in social tagging
software [7], although in those applications, the idea of
Milgram has not been well reflected, as the tagged data
are very often not connected to a geographical context.
In this paper, we introduce Taggram (“tags” +
“gramme”) as a visualization technique adopting exactly
the original idea of Milgram for the presentation of geo-
tagged information. The terminology of gramme can be
understood as “thing written”, as used in cartogram
(where carte means “card” or “map”). In other words,
maps act as the context, while aspatial data are presented
and explored on maps through the tags.
In doing so, we analyze general issues and strategies
for the ideas of placing and handling various labels on
arbitrary-shaped geographical regions, following the
viewpoints of InfoVis. We did the same as Jo and Ryu [5]
in combining texts with other visual metaphors for the
exploration of information on maps, instead of using just
conventional cartographic approaches.
Figure 1. Stanley Milgram’s collective mental map
of Paris [6]
2010 14th International Conference Information Visualisation
1550-6037/10 $26.00 © 2010 IEEE
DOI 10.1109/IV.2010.52
316
Information Visualisation
22