Ireoluwa007/Nigeria-Insecurity-Dashboard
Domaine:
peace and securitygeospatialType de record:
dataset
Créateur:
Ire
Hôte:
A Spatiotemporal Analysis (1997-2025) and Forecast
# Nigeria Insecurity Dashboard Report: A Spatiotemporal Analysis (1997-2025) and Forecast
## Executive Summary
This project provides a comprehensive spatiotemporal analysis of Nigeria's security landscape spanning 1997 through 2025, alongside a predictive risk forecast extending through December 2026. Synthesizing over 43,000 incident records and 129,000 total fatalities, the analysis tracks the evolving dynamics of conflict—from Boko Haram/ISWAP ideological insurgencies in the North East to high-frequency banditry and tactical kidnappings across the North West, North Central and South West regions.
**Page 1: Historical Insecurity Landscape: Spatial Density and Hotspots**
**Page 2: Conflict Anatomy: Regional Breakdown and Tactical Event Profiles**
**Page 3: Predictive Risk Modeling: Regional Volatility Forecast**
## Data Architecture, ETL Pipeline & Modeling Workflow
### Step 1: Raw Data Ingestion & Column Pruning (Excel)
Data Source: Extracted raw spatiotemporal conflict records (1997–2025) covering Nigeria from the Armed Conflict Location & Event Data Project (ACLED).
Column Pruning Logic: Dropped non-essential metadata columns (e.g., notes, source, source_scale, geo_precision, civilian targeting, associated_actor_1/2) to streamline memory usage and speed up Power BI visual processing. Retained 17 core operational variables and renamed some for clarity: event_id_cnty, event_date, year, disorder_type, event_type, sub_event_type, actor1, actor 2, interaction, time_precision, admin1 (state), admin2 (LGA), location (town), latitude, longitude, geo_precision and fatalities.
### Step 2: Data Transformation & DAX Modeling (Power BI)
Power Query Cleaning: Imported the pruned dataset into Power BI, standardized LGA and State key strings to reconcile spelling inconsistencies across years, addressed missing geocoded coordinates, and built a dedicated Dim_Date dimension table linked via a 1:Many relationship to enable accurate time-series modeling.
DAX Im …