African Port and Trade Clearance Dataset

This is a synthetic dataset generated from high-quality expert-labelled seed data. All records are algorithmically derived — statistical distributions, inter-field correlations, and annotation characteristics faithfully replicate real-world patterns from the source data, while ensuring no real individual, organisation, or transaction can be identified or reconstructed.

The African Port and Trade Clearance Dataset aggregates 2M+ import and export declaration records processed at major seaports and land border crossing points across Nigeria (Apapa, Tin Can Island), Ghana (Tema), Kenya (Mombasa), and South Africa (Durban, Cape Town). Each record captures the full customs clearance lifecycle: declaration lodgement, document verification, physical inspection decision, examination outcome, duty assessment, and release — with timestamps at each stage enabling end-to-end dwell time modelling.

Declarations are classified by HS code chapter (2-digit) and a 6-digit HS code where available, commodity description, trade flow (import/export), declared value, and country of origin or destination. Risk channel assignment (green, amber, red lane) and inspection outcomes (compliant, documentary discrepancy, undervaluation, prohibited goods) are included for the subset of declarations that were risk-selected for examination, enabling supervised customs compliance and fraud-detection model development.

The dataset is anonymised at the declarant level — importer/exporter names are replaced with sector and size-band codes — while preserving the statistical relationships between trader characteristics, commodity types, and compliance outcomes that are essential for risk-scoring model training. A companion port-level metadata file covers berth capacity, handling equipment, and seasonal throughput indices for port-operations forecasting tasks.