First Designs on Terrorist Activities

Data Visualization
5 min readMar 28, 2021

--

Miro Board Structure

To support the exploration and development of terrorist activities visualizations, we decided to create a format to give some structure to our MIRO board. We did so considering the Diverge-Emerge-Converge process. As shown on the image below, the board illustrates the process we followed during the design session:

  1. Diverge. We first began with a brainstorming session, in which we looked for previously developed visualizations (some of them in the context of violence/terrorist activities). We felt it was critical to examine what others had done before, to guide us through the process of creating truly innovative and valuable insights into the terrorism data set through our visualizations. Furthermore, we looked for basic designs that could be combined to create more comprehensive layouts.
  2. Emerge. After the initial discussion of ideas, we came up with several sketches. At this point, we split the board into four columns, each of which attempts to gather illustrations under a given research question. Some of the designs created during this stage are explained hereunder.
  3. Converge. Notice that this stage has not been carried out yet. We still need to decide what sketches we will take forward and, potentially, develop using visualization tools.
Figure 1: The Illusionists’ Miro Board

Diverge and Emerge

During the initial stages of our investigation, we tried to capture as many ideas as possible by incorporating many of the features that were available. At this stage, we did not yet attempt to answer any specific questions, but rather attempted to see how we could depict the available information in a variety of different ways.

In our first designs, we discovered that we relied heavily on geographical and temporal features. These characteristics allowed us to intuitively describe and conceptualize the parameter space. The plots provide an easy-to-understand method for explaining political and historical contexts by revealing, where terrorist attacks are occurring and how they are changing.

  • Diverge:
Figure 2: World Map by Attack Type

Represented here we have perhaps one of the most intuitive plots to effectively summarize key features of the terrorist data. The plot shows a map of the world with each dot representing an event. At a glance, the colors and size allow for an intuitive interpretation of the information. To expand on this graph, we could change the shape of the dot to add another parameter to the visualization. The shape of the dots, for example, could be used to represent the target type, though this may make the map more cluttered and difficult to read. Overall, we believe that this is a very powerful and efficient way to summarize and display a large amount of information in a compact manner. The plot capitalizes on people’s familiarity with map data and allows them to derive immediate understanding from it.

Figure 3: Development of Attack Type over Time

In this visualization, we make use of the time dimension, as we attempt to show progression over the decades. To understand terrorist activities, we must first comprehend how they have evolved, exemplifying the changing dynamics of specific features. On the y-axis, we labeled the categorical variable attack type, which is further distinguished by color for readability. On the x-axis, time is labeled. Each dot represents the year by its position and the number of deaths by its size. To further assist the visualization of changes with time, animations can be utilized to show shifts in pattern, which we may incorporate in adapted versions of the plots we describe here. This plot, however, conveys the same meaning without animations and allows the reader to compare the different time points more easily.

Figure 4: Alluvial Diagram of Perpetrator, Target, and Weapon

Finally, in addition to temporal and geographic features, we tried to illustrate how other categorical variables are linked to one another. We discovered that alluvial diagrams are remarkably useful in explaining these relationships. The distinct groups of Al-Qaeda are depicted on the left and right, with the width representing the proportion of total frequency of terrorist activities for each group. The branches show how many attacks involved specific targets/ weapon types for each group in comparison. These variables are all linked together in such a way that the reader can easily understand how they are related. Although this graph appears cluttered here, using advanced graphing software, the connections of interest can be later highlighted, allowing for a more straightforward interpretation.

  • Emerge:
Figure 5: Flow Map

In this visualization we combined the flow dynamics seen in the alluvial diagram, adapting the interpretation of the branches to the plot type. Furthermore, these branches are plotted on a map for ease of interpretation. This visualization highlights the network of activities, depicting on a global scale how the various terrorist groups are linked and how they relate to their point of origin. The thickness of the branches describes the magnitude of the activity, and the colors show the related terrorist organisations. This plot provides a accessible interpretation to the terrorist networks.

Figure 6: Heat Map of Terrorist Attacks over Time

Combining some of the previously bespoken elements we obtain this graph. It depicts a heat map of Al-Qaeda attacks across months and years. For simplification purposes, only the period between 2003 and 2017 is shown. Each time period is assigned a color from the green-red scale shown on the right side. This way, green corresponds to those months when the number of attacks amounted, at most, to the first quartile of the attacks number distribution, and red, to the months when attacks exceeded the third quartile. The use of this scale allows to get a sense of the level of danger across months, in a similar way to the defense readiness condition (DEFCON) alert state used by the United States Armed Forces. Moreover, the color scale aids in identifying trends in the data. On the upper side of the map, a stacked bar plot was included to show the number of wounded and killed civilians across years.

These are only some of the many sketches that we have come up with. On our MIRO board you can find further visualizations to the terrorist data. In the next step we hope to concisely combine some of the ideas that we have collected here and improve the readability and interpretation of the plots by combining graphical elements.

--

--

Data Visualization

This blog was created for the class on Data Visualization at KU Leuven