Geovisual Analytics and Storytelling Using HTML5

Abstract

The large and ever-increasing amounts of multivariate, multi-source, time-varying and geospatial digital information represent a major challenge for the analyst. The need to analyze and make decisions based on these information streams, often in time-critical situations, demands efficient, integrated and interactive visualization tools that aid the user to explore, present, collaborate and communicate visually large information spaces. This approach has been encapsulated in the idea of Geovisual Analytics, an interdisciplinary field that facilitates analytical reasoning through highly interactive visual interfaces and creative visualization of complex and dynamic data integrated with storytelling. Collaborative mapping is exemplified in this paper through telling stories about public statistics development over time that could shape, for example, economic growth and well-being. Discoveries are made that leave lasting impressions by stimulating the readers' curiosity making them want to learn more and convey a deeper meaning. In addition, the user can interactively participate in this web-based process which is important to the education and dissemination of public statistics. The storytelling mechanism assists the author to improve a reader's visual knowledge through reflections such as how life is lived by using a variety of demographics, such as healthcare, environment, and educational and economic indicators. Integrated snapshots can be captured at any time during the explorative data analysis process and thus become an important component of a storytelling reasoning process. The public can access Geovisual Analytics applications and explore statistical data relations on their own guided by the stories prepared by the experts. With the associated science of perception and cognition in relation to the use of multivariate spatialoral statistical data, this article contributes to the growing interest in visual storytelling engaging the public with new experiences.

Full text

Geovisual Analytics and Storytelling Using HTML5
Patrik Lundblad, Mikael Jern
Linköping University
{patrik.lundblad@liu.se, mikael.jern@liu.se}
Abstract
The large and ever-increasing amounts of
multivariate, multi-source, time-varying and geospatial
digital information represent a major challenge for the
analyst. The need to analyze and make decisions based
on these information streams, often in time-critical
situations, demands efficient, integrated and interactive
visualization tools that aid the user to explore, present,
collaborate and communicate visually large information
spaces. This approach has been encapsulated in the idea
of Geovisual Analytics, an interdisciplinary field that
facilitates analytical reasoning through highly
interactive visual interfaces and creative visualization of
complex and dynamic data integrated with storytelling.
Collaborative mapping is exemplified in this paper
through telling stories about public statistics
development over time that could shape, for example,
economic growth and well-being. Discoveries are made
that leave lasting impressions by stimulating the readers’
curiosity making them want to learn more and convey a
deeper meaning. In addition, the user can interactively
participate in this web-based process which is important
to the education and dissemination of public statistics.
The storytelling mechanism assists the author to improve
a reader’s visual knowledge through reflections such as
how life is lived by using a variety of demographics, such
as healthcare, environment, and educational and
economic indicators. Integrated snapshots can be
captured at any time during the explorative data analysis
process and thus become an important component of a
storytelling reasoning process. The public can access
Geovisual Analytics applications and explore statistical
data relations on their own guided by the stories
prepared by the experts. With the associated science of
perception and cognition in relation to the use of
multivariate spatial-temporal statistical data, this article
contributes to the growing interest in visual storytelling
engaging the public with new experiences.
Keywords--- Geovisual analytics, collaborative
mapping, storytelling
1. Introduction
Storytelling is one of the most impactful ways to
teach, learn, and persuade. As social creatures, we love
telling stories and sharing them with others. Analytics
visualization experts are challenged to elicit a strong
engagement and wonder in their stories stimulating the
readers’ curiosity making them want to learn more and
convey a deeper meaning. Statistical data may look
deceptively boring if you see them in Excel rows and
columns. But if these statistical numbers represented in
colourful, visual views and a user can interact with the
visualizations to find meaning – and, even better, if a
good story about the numbers is included, these numbers
can become most interesting and engaging [8, 9, 11]. An
engaging and compelling story should contribute to the
following features:
• Hold the audience’s attention with interesting
visualization views, snapshots and metadata;
• Leave a lasting impression by stimulating a
user’s curiosity and making him/her want to learn
more and express a deeper meaning;
• Contribute to collaboration by communicating
engaging interactive visual content [8]
In this context, we introduce Geovisual analytics [1]
software for HTML5/JavaScript platforms with
integrated storytelling applied to large spatial-temporal
and multivariate data by means of dynamic visual user
interfaces and its increasing role in worldwide
collaboration, communication and education.
Interactive data visualization supplements
storytelling with three important tasks:
comprehensibility, credibility and involvement. Building
a story incrementally, enhancing it with snapshots and
descriptive metadata, letting the user interrupt and
control the story could reduce the risk of a visual
presentation that’s poorly understood and thus improve
comprehensibility. The visualization’s credibility is
improved by letting the user interact at any time and
verify that the story represents what it claims. Allowing
users to interact with visualizations will transform them
from passive viewers to active participants and increase

Figure 1. Multiple views dashboard HTML5-based visualization with associated storytelling panel based on
statistical national data from the World Bank. A compelling story about the world ageing population problem
during 1960-2010; map (age 65+) with 4 map layers (Figure 3) including the pie chart layer (age 0-14 vs. 65+),
treemap showing total population (rectangle size) and age 65+ (colour) and bar chart (65+); The story
acknowledges the 4Ws concept “What-When-Where-Why” and includes 8 linked snapshots embedded into the
metadata. http://mitweb.itn.liu.se/GAV/dashboard/#story=data/Ageing Population in the World.xml&layout=[map,(treemap,barchart)]
involvement. Consequently, by adopting these three
tasks, a user could feel a greater and more personal sense
of engagement with the statistical data being presented.
A story is composed of texts and visual
representations together with snapshots, metadata,
hyperlinks and references providing answers to the
classical 4Ws questions (Figure 1): “What” is the data?
The data are shown for a given time step - “When”?
Moreover time series can be viewed interactively and
simultaneously in the in the linked graphs; “Where” is
mostly considered in regional comparisons. “Why” is the
most important aspect of the story and explained in the
metadata text section together with snapshots and
hyperlinks that enable users to further analyse aspects
related to topics and are helpful tools for both specialised
and non-specialised users.
Clearly, the need exists to consider how storytelling
facilitates a leap in understanding by the public enabling
them to appreciate how statistical indicators may
influence our society. The use of visualization tools helps
to communicate statistics to non-expert users; however
any specific language used in the context of
communication and interpretation of reality implies at
the same time both opportunities and risks. Many tools
offer the capability to make order in our experiences,
combining and reading data through its basic elements
(words and text for the natural language, images and
graphs for the visual one) but they usually prevent users
from considering other specific issues. For instance, it is
impossible to extract synthetic snapshot from texts in
natural languages and, on the contrary, to highlight
single events in, for example, a thematic map at a glance.
The paper introduces new HTML5 visualization
tools that also support the collaboration and publication
process facilitating storytelling aimed at producing, for
example, statistical news content in support of a semi-
automatic authoring process (Figure 6). Storytelling, in
our context, is about telling a story on the statistics data
and the related analytics reasoning processes on how
knowledge was obtained (Figure 2). Storytelling within
an interactive web context could more engagingly draw
the user into reflections and sometimes change a
perspective altogether. The story is placed in the hands
of those who need it, e.g. policy and decision makers,
teachers and also informed citizens. This dynamic visual
storytelling is a way of telling stories through interactive
web-enabled visualization with integrated snapshots
highlighting important discoveries. Stories enable a leap
in understanding by the user in order to grasp how
statistical indicators may influence society. The
conceptual approach and framework of the geovisual
analytics storytelling implementation is based around
three complementary characteristics: explore and gain
insight, tell-a-story and publishing:
• Authoring: data provider, explore and gain
insight, motion visual representations.
• Tell-a-story: snapshot mechanism that captures
interactive scenarios, metadata with hyperlinks,

story and chapters, edit, capture, save, and export
story.
• Publisher (Vislet): import stories and create
HTML5 code for embeddable interactive motion
visual representations for publishing on a web site or
blog using Flash or HTML5.
Figure 2. Reiterate analytical storytelling loop -
Statisticians with diverse backgrounds and expertise
participate in creative discovery processes that
transform statistical data into knowledge. Storytelling
tools integrate this geovisual analytics process using
collaboration. This knowledge exchange process
develops a shared understanding with other
statisticians and, after consensus has been reached, it
can be placed in the public domain. The snapshot
mechanism helps the author of a story to highlight
data views of particular interest, and subsequently
guide others to important visual discoveries.
2. Related Work
Tools for explorative data analysis are many. In this
section we divide them into two categories: tools
supporting for visualization and tools augmented with
storytelling capabilities.
Storytelling has been a topic of many research areas.
For example, Chris Crawford [4], from a game
designer’s point of view, presents the important points of
interactive storytelling in which the story is decided
based on the interaction between the player(s) and non-
player characters in the stories. He also discusses
challenges in development of an interactive story engine
and presents implementation strategies for such an
engine. Mateas [13] presents a survey on the aspects of
narrative intelligence in the artificial intelligence area.
Many research efforts were also made to bring
storytelling into visualization areas such as information
visualization, geovisualization and scientific
visualization. Gershons [5] presents a good example of
using animation, voice narration, and visual annotations
to tell a story about analysis. Cartwright [3] develops
prototypes using a ‘games’ interface for accessing
geographical information and confirm the effectiveness
of the approach.
MacEachren [12] has emphasized the importance of
a capacity to snapshot explorative data analysis processes
and then reuse them for presentation and evaluation
within the same environment. CCMaps [2] presents a
mapping tool that allows users to save snapshot events
and reuse them for presentation purposes. Robinson [14]
describes a method they call “Re-Visualization” and a
related tool ReVise that captures and reuses analysis
sessions. Keel [10] describes a visual analytics system of
computational agents that supports the exchange of task-
relevant information and incremental discoveries of
relationships and knowledge among team members
commonly referred to as sense-making. Many Eyes [15]
is an interesting storytelling approach implemented for a
public web site where novice users can upload their own
data, create dynamic visualizations and participate in
discussions.
To our best knowledge, there is no publication of a
web-based visualization toolkit/system which can bring
interactive visualizations to the public through various
platforms and devices including mobile devices and at
the same time supports various features of a geovisual
analytics system such as filtering, brushing, highlighting,
multiple-linked views, focus+context, a choropleth map
with dynamic layers and especially storytelling
capabilities to support non-expert users with better
understanding of statistics data.
3. GAV HTML5 Toolkit and Framework
The GAV (GeoAnalyticsVisualization) toolkit [6],
programmed in Adobe’s ActionScript, was introduced in
2009 [7] and represents the basis for many of today’s
statistics visualization applications such as “OECD
Regional eXplorer” [16]. The statistical community
required visualization tools that were accessible to
almost 100% of the web users, a solution that was
available through Adobe Flash. With the emerging
HTML5 standard which is promoted as the future web
content and supported by most devices including mobile
devices, the focus has now transferred to include support
for also tablets and smartphones. Using HTML5 canvas
and JavaScript also facilitates the integration of the
framework into other web-based systems.
Figure 3. GAV HTML5 framework with a large
selection of innovative interactive visualization
components – visualization methods can be evaluated
in this dashboard application:
http://mitweb.itn.liu.se/GAV/dashboard/
Development of a new GAV framework based on
HTML5 and JavaScript faces new challenges mostly
coming from the nature of this programming language.

JavaScript is a prototype-based programming language
and it is totally different from conventional object-
oriented programming languages such as C++ or Java. It
does not have the concepts of class and class inheritance
which facilitate the development of large scale systems.
To overcome this problem, we have developed a core
library to simulate these concepts in JavaScript. As a
result, GAV HTML5 is based on the principle of
modularity allowing application developers to select
methods from a wide range of geo- and information
visualization techniques (Figure 3), data providers and
data transforms and combine them in any way they want
to assemble dashboard applications.
The GAV framework includes a wide range of
geovisualization and information visualization
techniques from common ones such as choropleth map,
scatter plot, bar chart, time graph, and table lens to
advanced ones such as parallel coordinates plot, scatter
matrix, distribution plot and treemap (Figure 3), which
can be applied in dashboard applications. It also supports
various interactive features to facilitate the process of
visual exploration and analysis of data. Interesting
subsets of the data can be interactively selected and
brushed, and the relations can be investigated in all the
linked views. Data-of-interest methods that visually
separate important discoveries or outliers from the rest of
the data are integrated. Innovative “focus+context”
visualization with fish eye lens are available in bar charts
and dynamic queries extending the analysis capabilities
to scenarios with larger data sets. Other interactive
features are tooltips, brushing, highlight, visual inquiry,
conditioned filter mechanisms and methods supporting
time-linked multiple views. The framework also supports
data analysis algorithms, connects the components to
each other, and supports data providers that can load data
from various sources (Figure 3).
Figure 4. The layer architecture of the choropleth
map component.
The heart of our choropleth map is a map base class
which contains no visual parts at all, but controls
everything needed to display content within its context. It
creates transformations to deal with projections and
keeps track of the user's input as they zooms or pans. The
visuals are instead created in layers, for example a
polygon layer for coloured regions and a glyph layer for
region data (Figure 4). By making the map independent
of its layers, we can combine them in any way we
choose, or create new ones as long as they adhere to the
basic principles of the map.
To support users working with geographic regions
easier (e.g. identifying the geographical positions of cells
in a mobile phone network) as well as exploiting the
advantages of online maps, we built a background map
layer which allows the usage of WMS maps such as
Google map or Bing map as a background map. The
users can switch between these online maps or turn them
off. Figure 4 shows a number of typical layers of our
choropleth map: background map layer (Google map in
this case), border layer, coloured region layer, glyph
layer (a pie chart glyph layer in this case).
4. Storytelling with Integrated Snapshot
Mechanism
GAV HTML5 is not only an exploration and
discovery tool, but also a means of incrementally
collecting snapshots and writing associated explanation
to support a social style in the use of geovisual analytics.
Storytelling is achieved through the Snapshot Manager
that supports the storage of interactive events in an
analytical reasoning process through “memorized
interactive visualization views” or “snapshots” that can
be captured at any time during an explorative data
analysis process.
Traditional passive storytelling prohibits interaction
on the reader’s part. The author fully controls all
domains. Our storytelling is based on interactive
approval that allows passive storytelling to pause at any
points and lets readers take control, for example, change
the visualization’s view, representation, and even
content. Once they are satisfied with this interactive
exploration, storytelling continues with another snapshot.
Readers can completely detach from the story and
engage in interactive visualization with total freedom.
When exploring and making sense of comprehensive
statistical data there needs to be a way for the author of
storing discoveries in a coherent and cognitive
workspace, which can be organized, navigated and
maintained within the application so that discoveries may
later be loaded. The snapshot service includes such
means by giving the analyst the option of capturing,
saving and packaging the results of an exploration “gain
insight” process. The insights are captured in a series of
“snapshots” that could help the analysts to highlight
views of particular interest and subsequently guide
readers to follow important discoveries. Snapshots are
created as a series of visualization captures during the
exploration process and form the foundation for a story.
The analyst is immersed in the data domain and
assembles pieces of the story while gaining more and
more knowledge about the data locating interesting
views and specifies snapshots as key frames until the
exploration is complete and the resulting story is
satisfactory.
In a typical scenario the analyst selects relevant
attributes for a snapshot, e.g., time step, highlighted
regions for comparisons, class values for colour legend,

filters inquiry conditions for a reduced dataset and finally
highlights the “discoveries” in the map view from a
certain angle. When the analyst presses the capture
button, all the components within the application store
their properties, thus creating a state, which can later on
be recreated (Figure 5). The snapshot is then inserted in
the story by creating a hyperlink in the text that is
highlighted. When the reader follows the story the reader
can click the highlighted link and the state of the
visualization when captured is recreated so the reader has
the same view as the author. The reader may then also
make changes to the application such as changing filter
values or highlighting other areas of interest. This new
discovery may then be stored as a new snapshot and be
inserted into the story.
Figure 5. The snapshot system scans through all
active views to gather and apply states.
Incrementally building a story, enhancing it with
snapshots and annotations, and letting the readers
interrupt and control reduces the risk of presenting an
overloaded visualization that’s poorly understood and
thus improve comprehensibility. The author creates a
single or discrete series of captures during the
explorative process by selecting relevant indicators,
regions-of-interest, colour schema, filter conditions
focusing on the data-of-interest, or a time step for
temporal statistics. Associated explanatory text provides
a richer functionality than simple metatext by allowing
the reader to click on key words and learn about topics in
the story (Figure 6). A story hyperlink is a reference in
the story metatext that links to an external URL web site
or a captured snapshot. To insert a hyperlink in the
metatext the text is selected and a button “Link” is made
visible and two options appear: a) new capture
(snapshot) or b) link to an external URL.
Before the actual capture is done, the author
navigates to a particular region and selects which
indicator to use for the map, selects indicators for the
scatter plot, and selects a time period. A “Capture” is
made and all preferred states are saved. When the story is
read subsequently, hyperlinks can be initiated and the
dashboard application will display the state-of-the-
snapshots. Hyperlinks that instantiate a state in the
dashboard application are a central feature of the
storytelling mechanism, with associated descriptive text
that can guide the reader. A few words and a snapshot
can provide the difference between a pretty picture and
understanding.
Figure 6. Example of a snapshot linked to the
metadata in the story text - A few associated words
and a snapshot can provide the difference between a
pretty picture and understanding.
5. The Publishing Process
The story is created and understanding is shared
with other analysts and can now be placed in the public
domain (Figure 1). Statistics Publisher is a server-based
tool that imports a story and generates the HTML code
that represents the Vislet and metadata. A Vislet is a
simple HTML5 dashboard (widget) assembled from
GAV HTML5 components. A Vislet facilitates the
translation of selected detailed statistics data into
heterogeneous and communicative sense-making news.
It also integrates snapshots, metadata and dynamic
embedded animated visualizations that can engage the
user.
First, the user selects an appropriate layout for the
visual representation for the Vislet. Depending on where
the Vislet is to be published either one or more
visualizations can be used and the order and orientation
of the visualizations can be adapted to the Vislets
environment. After this choice the user then selects the
appropriate visualizations to be used e.g. map, scatter
plot, parallel axes, table lens, or time graph. Then the
size of the Vislet window with metadata is set together
with other available options and Publisher generates the
HTML code.
This code is then manually copied and embedded
into a web page. The Vislet can then be opened in the
user’s web browser and dynamically communicate the
story (Figure 7). A Publisher server maintains the Vislet
files with a story repository, statistical data and, if
required, regional shape maps. The Vislets run locally in
the client’s HTML5 web browsers and can therefore
achieve dynamic interactive performance, unlike static
images posted normally used to publish statistical data.
Interactive features in a Vislet are available to all linked
visualizations including tooltips, brushing, highlight, and
filters that can discover outliers and dynamic multiple-
linked views.

Figure 7. Storytelling facilitates three complementary characteristics: explore and gain insight, tell-a-story and
publish (Vislet) which finally can be embedded in a web page/blog using the HTML code.
6. Case Studies
In this section we introduce three case studies on
how visualization and storytelling apply in practice.
6.1 Analysis of Multivariate Data
Figure 8: Example of a Vislet aimed at a comparative
visualization using linked Scatter Matrix and Scatter
Plot to analyse national correlation between 6
indicators during 1960-2010 from the World
Databank. In the Scatter Matrix the correlation
between the indicators is shown both visually as
smaller scatter plots, but also with highlight that
shows positive or negative correlation.
http://mitweb.itn.liu.se/GAV/dashboard/#layout=[scattermatrix,sca
tterplot]&story=data/Ageing Population in the World.xml
Statistics data is most often multivariate to its nature
and using visualizations made for presentation of data it
is not easy to find correlation within the data. Therefore
to analyse the data it is important to give the analysts
expert tools that they can use for the exploration. The
scatter matrix, for example, gives the analyst a good
overview of all correlations between the selected
indicators. The analyst can here use the scatter matrix as
an overview and then steer the scatterplot for interesting
detailed combinations over time (Figure 8). This
focus+context visualization technique is popular within
the information visualization community.
Other popular multivariate techniques are the radar
and parallel coordinates plot (PCP) (Figure 9) which also
give the analyst the ability to compare the profiles of two
or more selected regions. The radar plot represents each
region with a star-like appearance and the overlap
between the stars expresses similarity. This visualization
is good to compare a few regions to each other, but does
not show how the selected regions compare to all the
other regions. The PCP can display an unlimited number
of regions in the same graph where the number of
indicators is only limited by screen space. The profiles of
selected regions are highlighted on top of all other
regions represented by their faded lines in the
background. A user can easily identify if the regions
have similar statistics profiles. Experts have different
opinions and preferences regarding these two
multivariate methods.
Figure 9: Example of a Vislet aimed at a comparative
visualization using linked Radar Plot and Parallel
Coordinates Plot to analyse national correlation
between 6 indicators during 1960-2010 from the
World Databank. Profiles for two selected countries
Germany and Niger are highlighted. The Niger
profiles shows high values for age 0-14, low for 15-64
and 65+, lower life expectancy and high fertility rate.
Germany shows an almost reversed profile.
http://mitweb.itn.liu.se/GAV/dashboard/#layout=(radarchart,pcp)
&story=data/Ageing Population in the World.xml

6.2 Storytelling Applied to European NUTS2
Regions in Dashboard
Figure 10. The Distribution Plot is a functional
visualization technology for categorical classification
(countries) and here applied to EU NUTS2 regions
age 65+. http://epp.eurostat.ec.europa.eu/cache/RSI/
Statistics referring to ageing population, fertility and
gross national product, health and education levels tend
to be measured and debated in national terms (Figure 1).
However, economic performance and social indicators
can vary within countries as much as between countries.
In recent years regional (sub-national) differences in
economic output, labour productivity, job creation,
labour force participation and education within countries
have been at least twice as great as those among
countries (Figure 11). Understanding the variety in
regional economic structures and performance is
essential knowledge for initiating developments which
could improve regional competitiveness and in turn
increase national growth. Interactive European region-
by-region visual comparisons in the form of motion
graphs could lead to better identification of areas that are
outperforming or lagging behind. Patterns of growth and
the persistence of inequalities are analysed over time,
highlighting the factors responsible for them. For
example, the distribution plot (Figure 10) represents a
special visualization technology that displays the
variation within individual European countries.
This significant knowledge about these statistical
facts can now be brought together in engaging stories
and published to analysts and citizens. The HTML5
dashboard application (Figure 11) is optimized to sustain
interactive performance for a larger number of sub-
national European regions (NUTS2 and NUTS3) with
associated indicators and many time steps.
6.3. The Motala river story
In this case study we demonstrate the ability to take
other types of geographical mapping and how the
introduction of a new layer in the choropleth map, Figure
12, can enhance the functionality of the visualization,
without changing the basics of this component. The story
and data comes from the Swedish Meteorological and
Hydrological Institute (SMHI) who have modelled the
water flow and the emission of nitrogen and phosphor
into the water. In this example the data has been
aggregated from hourly level to a monthly to show the
bigger picture, but the data can be on any timescale.
Figure 11. HTML5 dashboard visualization with
three linked and coordinated views showing the
problem related to ageing population in several
regions of Europe. Highlighted Liguria and Toscana
in Italy and Chemnitz in Germany are examples of
regions with a dramatic increase in age group 65+
while the younger generation 0-14 at the same time
diminish.
http://mitweb.itn.liu.se/GAV/dashboard/#layout=[map,(barchart,di
stributionchart)]&story=data/nuts-regions-ageing-population-in-
europe-2010.xml
Figure 12. The map layers used in the story about the
Motala river.
The map has been divided into around 700
hydrological areas which are used for modelling the
river. In figure 13a we see how the polygon layer is
coloured according to the local addition of water to the
river areas, where blue is high amount and red is low,
while the total amount of water in the river areas is
visualized by the thickness of the stream layer on top.
This extra layer has been developed specifically for this
scenario and has been limited to showing the width
animated over time, for other types of data and scenarios
extra map layers can be implemented to highlight
important aspects of the data.
In addition to the modelled water data there is also
data on local addition of nitrogen and phosphor as well
as the total output of these nutrition’s. In figure 13b the
polygon layer is displaying the concentration of nitrogen
in the water that is added to the river and the stream layer
shows the total concentration. As can be seen here the
highest input of nitrogen is in the centre of the map,
which is an area used for agriculture.

Figure 13. (a) Shows the local and total water flow.
(b) Shows the nitrogen concentration.
For this scenario the storytelling system is used to
tell the story of the waters path from the source to ocean.
This is exemplified by the lake Roxen where three big
rivers merge together as can be seen in figure 14. In the
time bar chart the amount of phosphor leaving each area
is plotted over time, where each time step has four bars.
The first three bars from the left display the phosphor
from the three areas going into the lake, while the fourth
bar shows the total amount of phosphor coming out of
the lake. As can be seen the amount differs depending on
where the water is coming from as well as the time of the
year. One can also see that during parts of the year the
total amount going in is higher than the amount going in,
which shows that some phosphor is being left in the lake.
The user can also change and look at the concentration of
phosphor to see how it differs from the areas, as well as
follow the phosphor emission back to the areas with
highest emission.
Figure 14. The time bar chart shows the calculated
total value of phosphor going into the lake Roxen in
the first three bars for each time step. The fourth bar
shows the total amount of phosphor leaving the lake.
The map shows the geographical overview of the area
in the background and the river segments coloured in
blue for the stream layer.
7. Conclusions and Future Development
This paper tries to consider how the integration
between visualization and storytelling can make
statistical findings more comprehensible and accessible
to the general public. Traditional practices for geovisual
analytics methods applied to official statistics have
required tools for greater user participation and an
improved user experience. The approach set out in this
paper allows the analyst (author) to communicate with
interested readers through visual discoveries captured in
snapshots with descriptive text. Selected indicators and
visual representations can be published with their
metadata, thus facilitating the comprehension of
statistical information by non-expert readers. At the same
time, the Vislet technique applied to statistical
visualization can help in developing immediate on-line
publications, which draw attention to recent trends in
statistical data. This advanced storytelling technology
could also become very useful for educational purposes.
The integrated statistics exploration, collaboration
and publication process addresses editorial storytelling
aimed at producing statistical news content supported by
a semi-automatic authoring process. HTML code
representing insight and knowledge is embedded into a
web page or blog and communicates statistical data using
interactive visualization. Open Data and geovisual
analytics storytelling tools opens up for opportunities for
journalists and can quickly become a key technology for
tomorrow's editorial. By simply accessing open
statistical data can journalists be the narrator describing
what a particular statistical development can actually
mean to the world. So far, the huge manual effort
required and only a few pioneering journalists have
shown how data sources can be drilled down to create
deeper insights into the dynamics of complex situations.
Furthermore, an increasing number small businesses,
freelancers, think tanks and public institutions expressed
interest. The increasing number of mobile tablets and
smartphones is driving the interest for more visual
statistical analysis and related informative knowledge.
Read more at: http://ncva.itn.liu.se/
References
[1] Andrienko, G., Andrienko, N., Demsar, U.,Dransch, D.,
Dykes, J., Fabrikant, S.I., Jern, M., Kraak, M.J.,
Schumann, H.,Tominski, C.:Space, time and
visualanalytics. International Journal of Geographical
Information Science 2010; 24(10):1577-1600.
[2] Carr, D.B., White, D., MacEachren A.M: Conditioned
choropleth maps and hypothesis generation. Annals of
the Association of American Geographers 2005; 95(1):
32-53.
[3] Cartwright, W.: “Using the Web for Focussed
Geographical Storytelling via Gameplay,” presented at
ICA UPI Map 2004, Tokyo, 2004.
[4] Crawford C.: “Chris Crawford on Interactive
Storytelling,” New Riders, 2005.
[5] Gershon, N., Page, W.: "What Storytelling Can Do for
Information Visualization," Communications of The
ACM, vol. 44, pp. 31-37, 2001.
[6] Ho, Q., Lundblad, P., Åström, T., Jern, M.: “A web-
enabled visualization toolkit for geovisual analytics,”
Information Visualization January 2012 Volume 11, No.
1, Pages 22-42, DOI: 10.1177/1473871611425870.
[7] Jern, M., Thygesen, L., Brezzi,M.: A web-enabled
Geovisual Analytics tool applied to OECD Regional
Data. Reviewed Proceedings in Eurographics 2009,
Munchen, March 2009.

[8] Jern, M.: Educating students in official statistics using
embedded geovisual analytics storytelling
methods.Reviewed Proceedings in Eurographics 2010,
Norrköping, May 2010.
[9] Jern, M., Thygesen, L., Brezzi, M.: Storytelling – How to
visualize statistics. Reviewed paper, NTTS 2011 (New
Techniques and Technologies for Statistics), Brussels,
February 2011. http://www.cros-portal.eu/page/day-3.
[10] Keel, P.: Collaborative visual analytics: Inferring from
the spatial organisation and collaborative use of
information. In: Proceedings of the IEEE Symposium on
Visual Analytics Science and Technology (VAST); 2006
Oct 31 - Nov 2; Baltimore, Maryland, USA. Washington
DC: IEEE Computer Society; pp. 137-144.
[11] Lundblad, P., Jern,M.:Visual Storytelling in Education
applied to Spatial-temporal Multivariate Statistics Data.
Expanding the Frontiers of Visual Analytics and
Visualization, Editors Dill J, Earnshaw R, Kasik D,
Vince J, Wong P.C. (Eds.), May 2012 Springer
2012,SBN 978-1-4471-2803
[12] MacEachren, A. M., Brewer, I.,
Steiner,E.:Geovisualization to mediate collaborative
work: Tools to support different-place knowledge
construction and decision-making. In: Proceedings of the
20th International Cartographic Conference; 2001 Aug 6-
10; Beijing, China. Beijing: Chinese Society of Geodesy
Photogrammetry and Cartography; pp. 6-10.
[13] Mateas, M.,Sengers, P.: "Narrative Intelligence," in
Narrative Intelligence, M. Mateas and P. Sengers, Eds.:
John Benjamins Pub Co, 2003, pp. 340.
[14] Robinson, A.: Re-visualization: Interactive visualization
of the progress of visual analysis. In: Proceedings of
GIScience Workshop Visual Analytics & Spatial
Decision Support; 2006 Sep 20; Muenster, Germany; pp.
1-21.
[15] Viégas, F.B., Wattenberg, M., Ham, F.V., Kriss, J.,
McKeon, M.: Many Eyes: a Site for Visualization at
Internet Scale. IEEE Transactions on Visualization and
Computer Graphics 2007, 13(6): 1121-1128.
[16] OECD eXplorer:
http://stats.oecd.org/OECDregionalstatistics