Qwodjo/Ghana-mobile-money-insights

An end-to-end data engineering project focused on Mobile Money usage trends in Ghana. It collects raw MoMo datasets, performs data cleaning and transformations, builds automated pipelines, and produces analytics-ready tables and visual dashboards to reveal financial behaviour patterns across regions and demograph # Ghana-mobile-money-insights An end-to-end data engineering and analytics project focused on Mobile Money (MoMo) usage trends in Ghana. The project starts from a raw transactional dataset and performs exploratory data analysis (EDA) to reveal patterns in financial behaviour across regions, age groups, gender and time. --- ## 1. Project Overview **Goal:** - Understand how Ghanaians use Mobile Money across regions and demographics. - Simulate a realistic Ghana-style MoMo dataset and use it to generate insights that could inform policy, product design and financial inclusion strategies. **Key Questions:** - How is transaction activity distributed across regions in Ghana? - Which age groups are most active on Mobile Money? - Are there noticeable differences between male and female usage? - At what times of day do people transact the most? - How are transaction amounts distributed (small vs. large values)? The project is organised into: - `notebooks/analysis.ipynb` – data cleaning and enrichment. - `notebooks/eda.ipynb` – exploratory data analysis and visualisations. - `data/` – raw and cleaned datasets (large CSVs are **not** stored in GitHub; see Data Access note below). --- ## 2. Data and Preparation The dataset contains individual Mobile Money transactions with the following key fields: - Transaction amount - Timestamp (date and hour of the transaction) - Customer region (16 administrative regions of Ghana) - Customer age group (18–24, 25–34, 35–44, 45–54, 55–64, 65+) - Customer gender (male/female) Before analysis, the data is checked and cleaned: - Duplicate records are removed. - Rows with missing key fields (amount, timestamp, region, age, gender) are dropped. - Technical index columns that do not carry business meaning are excluded. After cleaning: - There are no missing values in the analytical dataset. - The remaining columns are consistent and ready for descriptive and visual analysis. This ensures that the patterns observed in the analysis …