Global-Terrorism

Photo Source: Google

Overview

Between 1970 and 2020, the world has faced a serious problem called global terrorism. Different groups and individuals used violence to pursue their goals, whether political, religious, or social. Some groups wanted to change their countries, while others sought to spread their beliefs worldwide. However, countries and organizations came together to fight against terrorism using different methods including technology to analyze data and track threats. Even today, the fight continues against terrorism, but we hope for a safer world in the future.

Photo Source: Google

Problem Statement

The primary problem this project aims to address is the need to analyze and make sense of the extensive global terrorism dataset from 1970 to 2020. This dataset comprises detailed information on thousands of terrorist incidents, including data on perpetrators, targets, methods used, locations, and the socio-political context surrounding each event. The challenge lies in efficiently processing, analyzing, and extracting relevant information from this vast dataset to derive actionable insights. This analysis aims to answer important questions regarding global terrorism which are:

1. Global trends and patterns: What are the trends in terrorism over the past five decades?
2. Evolution of terrorism: How has terrorism evolved in tactics, targets, and geographical spread?
3. Responsible groups: Which groups are mostly responsible for terrorist attacks?
4. Temporal and geographical clusters: Are there significant clusters of terrorist activity?
5. Success rate and duration: What is the success rate of attacks and their durations?
6. Impact on countries: Which countries have been most affected by terrorism?
7. Weapon types: What are the commonly used weapons in these attacks?
8. Claimed responsibility: Which groups have claimed responsibility for attacks?
9. Perpetrators, casualties, and weapons: How many perpetrators, casualties, and what weapon types are involved in attacks?

Addressing these questions will not only contribute to academic research but will also provide practical insights for counter-terrorism efforts, policy formulation, and security strategies. The outcomes of this project will help stakeholders gain a comprehensive understanding of the nature of global terrorism, enabling the development of effective preventive measures and interventions to mitigate the threat and promote global stability.

Photo Source: Google

Data Sourcing:

The dataset was downloaded from Kaggle. Kindly find the link to the original dataset here

About the Dataset

The Global Terrorism Database (GTD) dataset contains a comprehensive record of terrorist incidents worldwide from 1970 to 2020. The dataset includes numerous columns describing different aspects of each incident. It contains 135 columns and 209707 rows.

Tool Used: Power BI

Data Preprocessing

After downloading the dataset onto my local machine, I proceeded to load it into Power BI desktop for analysis. During the initial examination, I observed several noteworthy aspects of the dataset:

• Null Values: There were instances of missing data points within the dataset, represented as null values. These missing values can potentially affect the accuracy and reliability of the analysis, and strategies for handling them will need to be implemented.

• Incorrect Formatting: I noticed instances where the data was not formatted correctly according to the expected structure or data type. This can lead to issues in data manipulation and visualization, necessitating appropriate formatting adjustments.

• Spelling Errors: I identified instances of misspelled or incorrect spellings within the dataset. These errors can impact data integrity and consistency, requiring careful attention and potential corrections to ensure accurate analysis.

• Presence of Placeholders: I observed the presence of placeholders, such as the value '-99', indicating missing or unknown information. These placeholders can introduce biases or inaccuracies in the analysis if not handled appropriately. Appropriate treatment or imputation methods will be necessary to address these placeholders effectively.

By recognizing these issues within the dataset, it is essential to take corrective actions to ensure the reliability and accuracy of the subsequent data analysis process. Cleaning, formatting, and imputation techniques will be employed to address these challenges, ultimately leading to a more robust and trustworthy analysis of the global terrorism data.

Data Cleaning

• After successfully loading the dataset into Power BI desktop, I previewed it and identified the need for transformations to prepare it for further analysis. To begin the transformation process, I clicked on the transform button, which opened the Power Query Editor. Loading the datset into PowerBI desktop before proceeding to power query editor

• In the Power Query Editor, I took the precaution of duplicating the dataset (although not a standard procedure) and disabled one of the duplicates from loading and refreshing. I then proceeded to rename the duplicated dataset, ensuring clarity in the subsequent steps of the transformation process.

• To start cleaning the data, I meticulously reviewed each column in the dataset. I performed various cleaning operations, such as removing null values, correcting formatting errors, and addressing misspellings. I made sure to format each column with the appropriate data type and renamed them accordingly to maintain consistency and clarity.

• To ensure consistency and clarity, I made replacements in the 'extended,' 'suicide,' and 'success' columns. Specifically, I substituted '1' with 'Yes' and '0' with 'No' to accurately represent the corresponding values in these columns. This modification enhances readability and facilitates a better understanding of the dataset. Using replace values command to replace values

• One particular issue I noticed was that the day, month, and year information were separated into individual columns and not properly formatted. To rectify this, I merged these columns to form a single date column and ensured the correct formatting.

• Additionally, I extracted values for the year,month name and number from the date column since I required this information for my analysis. Instead of creating a separate calendar table, I utilized the existing date column for this purpose. Extracting values for year,month name and month number from the date column

• To further enhance the analysis, I created a date range column using the conditional column command. This allowed me to group the dates into meaningful intervals such as '1970-1980,' '1981-1990,' '1991-2000,' '2001-2010,' and '2011-2020,' facilitating temporal analysis. Creating a year range column using conditional column

• Regarding the "number of perpetrators" column, I noticed the presence of a placeholder value (-99). To ensure accurate analysis, I applied a filter to exclude this placeholder value when examining the number of perpetrators involved in each incident.

• In the columns for "nkill" (number of people killed) and "nwound" (number of people wounded), I encountered null values. To address this, I replaced the null values with a placeholder (-99) and excluded them from the analysis when assessing the casualties.

• Finally, I dropped unnecessary columns from the dataset that were not pertinent to my analysis, streamlining the dataset for further exploration and insights.

By performing these transformations and cleaning operations, I have prepared the dataset to be more suitable and reliable for in-depth analysis in Power BI desktop.

As mentioned earlier, the original dataset contained 135 columns and 209,707 rows. However, after undergoing the cleaning and transformation process, the dataset has been refined, resulting in a reduced number of columns to 26 while retaining the same number of rows, 209,707. The descriptions of the remaining columns are as follows:

• eventid: Unique identifier for the incident.

• date: date of the incident.

• extended: Indicates whether the duration of the incident extended beyond 24 hours (0 = No, 1 = Yes)

• country_txt: Name of the country where the incident took place

• success: Indicates whether the attack was successful (0 = No, 1 = Yes).

• attacktype: Description of the primary type of attack.

• targettype: Description of the primary target of attack.

• group responsible: Name of the group claiming responsibility for the attack.

• number of perpetrators: Number of perpetrators of the attack

• weaptype: Description of the primary weapon type.

• number killed: Total number of individuals killed in the incident.

• number wounded: Total number of individuals wounded in the incident.

• year: year the incident occurred

• Month: Month of the incident.

• Month Number: Month Number of the incident.

• suicide: Indicates whether the attack was a suicide attack (0 = No, 1 = Yes).

• nationality of the target: Description of the nationality of the target.

• property: Indicates whether property was damaged or destroyed in the incident (0 = No, 1 = Yes).

• year range: Range of years of attack

• region_txt: Name of the region where the incident took place.

• city: City where the incident took place.

• multiple: Indicates whether the attack was multiple or not

• guncertain: Indicates whether there was a use of gun for each attack

• individual: Indicates if the attack was an individual attack or not

• claimed: Indicates whether any group claimed responsibility for that attack

• claimedtxt: Indicates the means used by the group that claimed responsibility for that attack

ANALYSIS

• The dataset covers a total of 204 countries and 45,000 cities, with recorded attacks reaching 210,000. Out of these attacks, 185,000 were successful, resulting in 479,000 deaths and 586,000 injuries. The dataset includes information on 4,000 known terrorist groups.

Attacks by success status

• Based on the analysis of attack success rates, it was found that 88.38% of attacks in the dataset were classified as successful, while 11.64% of attacks were categorized as not successful. These findings highlight the significant proportion of successful attacks, indicating the effectiveness of terrorist operations in achieving their objectives.

Attacks by duration

• Furthermore, when examining attack durations, it was observed that 5.15% of attacks extended beyond 24 hours, while the majority of attacks, accounting for 94.85%, did not exceed the 24-hour mark. This analysis emphasizes that the majority of attacks were relatively short-lived, with a vast majority resolved within 24 hours.

Top 5 countries by attack and year

• Analyzing the top 5 countries based on attack frequency within different time periods, Afghanistan had 4 attacks between 1970 and 1980, while Colombia recorded 560 attacks, India recorded 34 attacks, Iraq recorded 12 attacks, and Pakistan recorded 18 attacks. The numbers increased in subsequent decades, with Afghanistan, Colombia, India, Iraq, and Pakistan experiencing varying levels of attacks.

Attacks by target

• The most targeted groups based on the number of attacks were private citizens and property (51,985 attacks), followed by the military (34,131 attacks), police (28,568 attacks), government (23,828 attacks), and businesses (22,169 attacks).

Number killed and injured by top 5 terrorist groups

• Examining the top 5 groups claiming responsibility for attacks, the group categorized as "unknown" caused the highest number of casualties, with 111,570 deaths and 220,321 injuries. Other notable groups include the Taliban, Islamic State of Iraq and Levant (ISIL), Boko Haram, and Shining Path.

Attacks by year

• Attacks varied by year, with 1970 seeing 651 attacks, 1980 witnessing 2,661 attacks, and 2014 reaching the highest point with 16,960 attacks. The lowest number of attacks occurred in 1971, with 471 incidents.

Number killed and injured by attack type

• Bombing/Explosion was the most deadly attack type, resulting in 157,418 deaths and 330,713 injuries. Unarmed Assault had the lowest death toll (849). Bombing/Explosion accounted for 38.73% of the total number of deaths.

Number killed and injured by top 5 weapon type

• Weapon type analysis showed that explosive-based attacks caused the highest number of fatalities (171,453) and injuries (417,940). Firearms and unknown weapons were also responsible for significant casualties.

Attack type by number of perpetrators

• Armed Assault had the highest number of perpetrators (358,320), followed by Facility/Infrastructure Attack (329,638). Hijacking had the fewest perpetrators (3,156).

Successful attacks by attack type

• Bombing/Explosion had the highest count of successful attacks (85,848), while Hijacking had the lowest count (676). Bombing/Explosion accounted for 46.33% of the total count of successful attacks.

These insights provide a glimpse into the global terrorism landscape, highlighting the countries, groups, attack types, and their impacts.

Visualization

Below is the dashboard showing the analysis of global terrorism from 1970-2020. Kindly click here to interact with the dashboard

Dashboard of the global terrorism attack for year 1970-2020

Recommendation

According to estimates, the world population in 1969 was around 3.6 billion people, while as of 2021, it is estimated to be around 7.9 billion people. Please note that these figures are approximated and may vary based on different sources and methodologies but you can find a reference for the world population here

Global terrorism has undoubtedly affected the population and lives of humans in various ways. The insights from the global terrorism dataset such as the high number of attacks, casualties, and the impact on different groups and regions, highlight the significant consequences of terrorism on human lives and societies. The loss of lives, injuries, and psychological trauma resulting from terrorist attacks have a profound and lasting impact on individuals, families, and communities.

The effect of global terrorism extends beyond the immediate human toll. It disrupts social cohesion, erodes trust, hampers economic development, and destabilizes nations. Terrorism can lead to displacement, migration, and the displacement of communities, further exacerbating social and humanitarian challenges. Moreover, the fear and insecurity generated by terrorism can impede progress in education, healthcare, and other areas of human development.

Based on the analysis of the global terrorism dataset, the following recommendations can be made:

• Strengthen international collaboration: Given the global nature of terrorism, it is crucial for countries to enhance collaboration and information sharing to effectively combat this threat. Sharing intelligence, best practices, and coordinating efforts can lead to better prevention and response strategies.

• Focus on high-impact countries: Countries such as Afghanistan, Colombia, India, Iraq, and Pakistan have experienced a significant number of attacks over the years. Allocating resources and implementing targeted counterterrorism measures in these regions can help mitigate the impact of terrorism and protect the population.

• Enhance security measures for high-risk targets: Private citizens, military personnel, police, government institutions, and businesses have been frequent targets of terrorist attacks. Strengthening security measures, implementing surveillance systems, and providing specialized training for personnel working in these sectors can help prevent and respond to attacks more effectively.

• Improve intelligence gathering and analysis: Investing in advanced technologies and intelligence capabilities can enhance the collection, analysis, and interpretation of terrorism-related data. By identifying patterns, trends, and emerging threats, governments and security agencies can proactively address potential risks.

• Counter radicalization efforts: Understanding the ideologies and factors that drive individuals to join terrorist groups is crucial. Implementing comprehensive counter-radicalization programs that address social, economic, and political grievances can help prevent individuals from being influenced by extremist ideologies.

• Strengthen border security: Enhancing border control measures and cooperation between countries can help curb the movement of terrorists, weapons, and illicit funds. Robust border security measures, including advanced screening technologies and intelligence sharing, are essential for preventing cross-border terrorism.

• Promote community engagement and resilience: Building strong community relationships and fostering trust between law enforcement agencies and local communities can aid in identifying and preventing potential threats. Encouraging community engagement programs, promoting inclusive policies, and addressing social divisions can help create a resilient society less susceptible to extremist ideologies.

• Support victims and survivors: Providing comprehensive support and rehabilitation services for victims and survivors of terrorist attacks is crucial. This includes medical care, psychological support, financial assistance, and legal aid to help rebuild their lives and restore their sense of security.

• Continuously update and refine the dataset: Regularly updating and expanding the dataset with new information, including emerging terrorist groups, evolving attack tactics, and emerging trends, can provide more accurate and comprehensive insights for future analysis and decision-making.

• Foster international dialogue and cooperation: Convening international forums, conferences, and workshops focused on counterterrorism efforts can facilitate knowledge sharing, foster collaboration, and encourage the exchange of best practices among governments, security agencies, and relevant stakeholders.

Conclusion

The analysis of the global terrorism dataset spanning from 1970 to 2020 has yielded valuable insights into the dynamics, trends concerning terrorism worldwide. Through meticulous data cleaning, transformation, and exploration, we have uncovered significant findings that deepen our understanding of this global menace.

The refined dataset, comprising 26 columns and 209,707 rows, ensures the integrity and accuracy of the data for further analysis. By examining the dataset, we have identified patterns and trends in terrorism over the past five decades, including targets, weapons used and geographical spread. Moreover, our analysis has shed light on the evolution of terrorist groups, providing crucial information for policymakers, security agencies, and researchers.

These findings can inform decision-making processes and facilitate informed discussions on global terrorism, ultimately guiding the development of effective preventive measures and interventions to mitigate the threat and promote global stability.

The implications of our analysis are particularly significant for counter-terrorism efforts and policy formulation. They offer actionable insights for policymakers to design strategies that address the ever-changing nature of terrorism. Moreover, the analysis underscores the necessity for ongoing research and collaboration to stay ahead of emerging threats.

In conclusion, the analysis of the global terrorism dataset has furnished comprehensive insights, revealing patterns and trends that contribute to the collective global efforts in combating terrorism. This analysis serves as a foundation for further research, policy development, and security strategies aimed at fostering global peace and security. Continual analysis and monitoring of global terrorism are vital in order to adapt and respond effectively to this evolving threat.

By implementing these recommendations, governments, security agencies, and policymakers can strengthen their counterterrorism efforts, safeguard communities, and work towards establishing a safer and more secure global environment.

Thank you for taking your time to go this this project. Your Comments and recommendations will be highly appreciated

Kindly connect with me on LinkedIn and Twitter

See you next time 😄🤗