Shark Attack Analysis Dashboard

Understanding Shark Attack Trends: A Data-Driven Approach

This Power BI dashboard provides an in-depth analysis of reported shark attacks over the past 100 years. It includes various details such as location, activity during the attack, victim information (name, gender, age), and shark species involved. My aim was to transform raw, messy data into clear, interactive visualizations to uncover trends and patterns in shark attack incidents.

Project Overview & Analytical Goals

The primary objective of this project was to explore historical shark attack data to identify key trends and insights. I focused on addressing questions such as:

• What are the annual trends in shark attacks since 1900?
• Which countries report the most shark attacks? Within those countries, which areas and locations seem to be the most dangerous?
• What body parts are most often injured?
• Are shark attacks more common during certain parts of the day?
• Which species of sharks are most often involved in attacks?

Data Source

The dataset used for this analysis comprises reported shark attacks over the past 100 years. This raw data required significant cleaning and transformation before analysis could be performed.

Methodology

1. Data Cleaning and Transformation: The raw data for shark attacks was highly uncleaned and required extensive preparation in Power Query. This involved loading the data, promoting headers, and performing initial data type conversions. Significant effort was put into handling missing values by replacing them with "No data" for text fields and "0" or a default age for numerical fields. Dates were standardized and cleaned, and new columns were derived to categorize data points like "Fatal (Yes/No)," "Specie name" (grouping similar shark species), "Body part" (from injury descriptions), and "Time of day" (from attack times). Redundant columns were removed, and rows identified as hoaxes or containing invalid year/date data were filtered out to ensure the dataset's accuracy and readiness for analysis.
2. DAX Calculations: Key measures were created to quantify total cases, fatalities, and survivors.
3. Interactive Dashboard Design: I designed a Power BI dashboard featuring:
   • Total Cases: 5565
   • Total Fatalities: 1370
   • Total Survivors: 4195
   • A trend graph showing shark attacks annually since 1900.
   • A breakdown of cases by time of day, showing Afternoon with the highest number of cases (around 1000).
   • A chart indicating cases and fatality by city, with Florida having the highest cases at 978.
   • Analysis of body parts most often injured, with "Lower limb" being the most frequent.
   • Identification of shark species with the most attacks, with "White shark" leading (around 400 cases).
   • A map highlighting countries with the most shark attacks.

Key Findings & Insights

• The trend of shark attacks shows fluctuations over the years, with a notable increase in recent decades.
• The Afternoon period has the highest incidence of shark attacks.
• Geographically, Florida and New South Wales in Australia are identified as regions with a high number of shark attacks.
• The Lower limb is the most commonly injured body part during shark encounters.
• The White shark is implicated in the highest number of attacks among identified species.

Project Impact & Value Proposition

This project showcases my data cleaning and visualization skills, especially when dealing with raw and unstructured data. It provides valuable insights for:

• Public Safety: Informing swimmers and surfers about higher-risk times and locations.
• Marine Conservation: Understanding species behavior in relation to human encounters.
• Research: Providing a foundational analysis for further academic or ecological studies on shark attacks.