This project explores a social media advertisement performance dataset using SQL in JupyterLab and visualizes the data through dashboards in Power BI and Tableau.
This project explores a social media advertisement performance database obtained from Kaggle. The dataset consists of four synthetic tables containing user information, campaign details, ad specifications, and recorded ad events. The data was already cleaned and ready for analysis, eliminating the need for additional preparation. I queried the data in JupyterLab using SQL, responding to targeted questions from ChatGPT by writing and executing SQL queries. This allowed for focused exploration of relationships and patterns across the interconnected tables.
After completing the SQL queries, all four tables were exported as CSV files. These files were then used to create dashboards in Power BI and Tableau, each built with unique visualizations to highlight different aspects of the data.
Analysis was done with Python in JupyterLab. The following key packages were utilized:
Kaggle – access datasets from Kaggle
sqlite3 – work with SQLite databases
pandas - data manipulation and analysis
sqlalchemy - database connection management
Dashboards were created using Power BI and Tableau Public.
In this repository, I've included the notebook used in JupyterLab for analysis, named project_code.ipynb. This notebook contains the questions and the SQL queries used to answer them. I've also included the database ad_campaign_db.sqlite and the four data tables in CSV format, located in the datasets folder.
The two dashboards, created in Power BI and Tableau Public, are uploaded as PDF files: Power BI Report.pdf and Tableau Dashboard.pdf. You can view the Tableau dashboard online here.
I'm a physics graduate with a focus on data analysis and Python programming. I work with Python, data visualization, and building tools that simplify complex tasks.