FACULTY OF COMPUTER & MATHEMATICAL SCIENCES

UITM SHAH ALAM

BACHELOR OF SCIENCE (HONS) MATHEMATICS (CS249)

DSC 651: DATA REPRESENTATION AND REPORTING TECHNIQUES

ASSIGNMENT 2: SPATIAL TEMPORAL VISUALIZATION

CS2496C

NAME : FAZLIYANA BINTI GHAZALI

MATRIX NO : 2017792083

PREPARED FOR: DR. ZAINURA BINTI IDRUS

1) What is spatial temporal visualization?

Spatial temporal visualization concerns changes in information in space and time. Spatial
refers to space and temporal refers to time. It has the natural advantage of revealing global
trends and patterns of movement. Spatial temporal data visualization and analysis can be
difficult because the space has infinite directions, such as up, down, sideways, north, south,
east, west while time can only travel forward. It can be difficult to integrate and analyze the
two dimensions of space and time.

The data can produce different results depending on how space is defined, for example, using
a zip code, a census tract or a state. Time can also provide conflicting answers depending on
whether it is measured in seconds, minutes, hours, days or years.

It makes the notion of time accessible to non-expert users compared to traditional visual
representations, and thus constitutes an important decision-making tool used in many
application scenarios. Substantial progress has been made in various fields as an
interdisciplinary approach, such as geographic information science, visualization, or visual
analytics, but there is still plenty of room for further progress.
2) 3 case study that make use of spatio temporal visualization

 Spatio-Temporal Partitioning of Large Urban Networks for Travel Time Prediction

a) Describe briefly what the case study about.

The graph represents the road network of the Stockholm, Sweden. The objective is
to identify the impacts of network partitioning on the computational time and
overall accuracy of travel time prediction. Spatio-temporal partitioning is used in
the clustering phase to represent the temporal effects of the connection speeds along
with spatial locations. Functional classes up to 5 are included here; for classes
higher than 6 (minor service or side streets) over 80 percent of the contacts are not
observed throughout the day and intervals. The case study road network consists of
11,340 links.
 b) How spatio temporal visualization support the story/knowledge behind the graph.

The dashed line represents the north-south axis of Stockholm city. While the line or
link with different colors represent the functional class of the road such as
freeways/motorways, arterials and urban streets. The functional classes 0 and 1 are
considered as the motorways and rest as the urban road network.

c) The basic visual elements used to explain the knowledge behind the graph.

The basic visual element used in this graph is a line. In this graph implied line is
used as connections between one or more points. This line refers to the path that the
viewer’s eye takes as it follows colors and forms along any given path.

Different kinds of colors also used in this graph to represent the functional class of
the road. The used of colors is important as it can help the viewers to identify each
class better, hence leads to the understanding of the graph.

 Spatio-temporal dynamics of global H5N1 outbreaks match bird migration patterns

a) Describe briefly what the case study about.

This graph shows the spread of H5N1 outbreaks worldwide from December 2003 to
December 2006, shown by year (a) and population type (poultry or wild birds) (b).
The space-time permutation model used in scan statistics was used to check and
classify their approximate location and timing for spatio-temporal clusters.

Figure (a) reveals that during 2003 and 2004 all of the outbreaks were concentrated
in South and Southeast Asia. Outbreaks continued to occur in Asia in 2005, but also
in Western Russia and Europe began to occur. The disease became a prevalent in
2006 around the Black Sea region, the Mediterranean region, Western Europe, and
East Africa.

While figure (b) shows the concentration of wild bird outbreaks in Western Europe,
intermittent outbreaks in Southern Europe and Central Asia are observed.
Nevertheless, poultry outbreaks were concentrated in Southeast Asia (along with
wild bird outbreaks and some mixed outbreaks), West Russia, the Black Sea region,
Arabia, and Africa.

b) How spatio temporal visualization support the story/knowledge behind the graph.

The space-time permutation model analyzes both spatially and temporarily clusters
of H5N1 outbreaks by checking whether outbreaks close to space are also close in
time. Space-time clusters occur when an excess number of H5N1 outbreaks occur
within a spatial and temporal range identified by the consumer.

The transmission of disease is strongly influenced by the at-risk population's spatial

and temporal behavior. Numerous species are worried about the H5N1 outbreaks,
which do not always have a clearly defined range, such as wild birds. A dot with
different colors in (a) is the year of H5N1 outbreaks, whereas the dots in (b) are the
bird population type.

c) The basic visual elements used to explain the knowledge behind the graph.

This graph used different hues of colors to represent each year and each
population. In (a), year 2003 represented by black, year 2004 represented by red
color and so on. While in (b), green color represent wild birds, blue color
represent poultry and wild bird and etc. Thus, it will be easier for the viewer to
 differentiate and analyze the graph as the colors used in this graph not exceed 7
type of colors that can confuse the viewer.

 Spatio temporal characteristics of elderly population’s traffic accidents in Seoul

using space-time cube and space-time kernel density estimation

a) Describe briefly what the case study about.

The space-time cube structure is shown in the above graph where (a) represents 3D
space-time bins, while (b) represents 2D bins developed for emerging hotspot
analysis. The purpose of this study is to examine how the spatial temporal
characteristics of traffic accidents involving the elderly in Seoul change over time.

In order to evaluate the space-time characteristics, space-time cube, emerging

hotspot, and space-time kernel density estimation analysis are applied. Monthly
analysis showed that during winter and summer the hotspots are small, while during
spring and fall they are high in the hiking and climbing areas of Seoul. In contrast,
the hotspots of older victims are more intermittent than the hotspots of older
drivers.

b) How spatio temporal visualization support the story/knowledge behind the graph.

Space-time cube analysis is a technique of 3D geovisualization which maps space-

time data in a cube and is useful in finding space-time patterns. A 3D cube consists
of space-time bins representing the x and y dimensions of space and the t dimension
of time. Every bin has a fixed position in space (x, y) and in time (z). In this study,
 the value of space (x, y) was 430m x 430m, which was calculated by dividing the
longest side of the study area (42750m) by 100.

Two different approaches have been used to analyze the changing trends of traffic
accidents over time. Hourly trends of traffic accidents have been studied in the first
method. Since the regions where traffic accidents are concentrated tend to change
regularly depending on a day's traffic flow, the locations of senior traffic accidents
are analyzed in 24-hour units. To this end, information on the year, month and day
of traffic accidents have been omitted so that only the date of occurrence has been
taken into account. The temporal and spatial patterns of traffic accidents were then
classified as 24-hour units and analyzed.

The second method is to evaluate the trend in traffic accidents that happened from
time to time in 2013. In this method, year, month, and day, as well as accident time,
are put in monthly bins of the same spatial size (430 m x 430 m).

c) The basic visual elements used to explain the knowledge behind the graph.

The basic visual element used in this graph is the shape. A space-time was
represented by a cube and was generated by using 24 bins in the z-axis (time), and
each bin represents one-hour units. Then, hotspot analysis was performed for each
period to determine its related characteristics. Within each bin, the points were
counted and their specified attributes were aggregated. For all bin locations, the
trend for counts and summary field values were evaluated and emerging hotspot
analysis was performed to analyze the trend by expressing the generated bins in 2D.

This graph also used colors to represent bin time series, time slice and bin. The
usage of these colors is good to make sure the viewer can identify and analyze the
difference between each bin. It will be easier for the user to analyze this graph as
this graph did not used various kind of colors that will distract the users.
3) Strength and weakness of spatial temporal visualization

STRENGTH WEAKNESS
Coordinated  Plain 2D views  Split attention effect
views  Connected by linking and  High visual work
brushing  No spatiotemporal integration
Animation/  Salient detection of change  Memory overload
slideshows  Small screen occupancy  High interaction costs
 “Quasi-integration” in working  Interpolation of animation can
memory create data artifacts
 Slideshow can reduced
continuous temporal data to an
ordinal variable
Layer  Spatiotemporally integrated and  Visual clutter
superimposition aggregated view  Design too flat, generates little
 Small screen occupancy attraction
 Time needs to be mapped to an
additional visual cue (like color)
Layer  Parallel visualization  Large occupancy of display
juxtaposition  High visual literacy to be space
expected  Temporal data is reduced to an
ordinal variable
 Split attention effect
 High visual work
Space-time cube  Spatiotemporally integrated  Visual clutter
view  High interaction costs
 Metashape  Medium occupancy of display
 Attraction power space
 Cognitive and conceptual
mediator

REFERENCES

1. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0196845

2.https://pdfs.semanticscholar.org/bd66/d55ba7b756ab1e00f0ccc86a1989ef86cb65.pdf?_ga=2.84
854750.656059128.1576511385-965475125.1576511385