Verification of ELECAM's latest overseas registration data (34,411) against the previous snapshot (34,296).
Old Data Total (Dataset 1)
Verified sum from previous polling stations list.
New Official Total (Dataset 2)
Figure confirmed from Dataset 2 Grand Total.
Net Change (Discrepancy)
Minimal net increase between the two snapshots.
Our audit uses a data pipeline incorporating Optical Character Recognition (OCR), Artificial Intelligence (AI), and Machine Learning (ML) models to extract, clean, and verify data from the official ELECAM images.
The core of the verification process is a Python script used for automated consistency checks:
# Illustrative Python snippet for data extraction & verification
from ocr_tool import process_image
# 1. Image preprocessing and OCR for raw data capture
data_set_raw = process_image("diaspora_list_new.jpg")
# 2. AI/ML-enhanced Data Cleaning & Structuring
df_diaspora = ml_model.structure_data(data_set_raw)
# 3. Mathematical Consistency Check (e.g., Africa Total)
computed_africa = df_diaspora[df_diaspora['Region'] == 'Africa']['Women'].sum() + df_diaspora[df_diaspora['Region'] == 'Africa']['Men'].sum()
print(f"Computed Africa Total: {computed_africa}")
ELECAM Facebook Earlier Diaspora Post (External Verification):
October 5th Post (Data 1)Raw Source Images (Internal Links):
Click on images below to view the original documents used for OCR:
ELECAM Facebook Later Diaspora Post (External Verification):
October 9th Post (Data 2)Raw Source Images (Internal Links):
This audit checks the mathematical consistency of the aggregated figures in the new data (Dataset 2), revealing two critical arithmetic errors in the source document itself.
The "AFRIQUE / AFRICA" row states a total of 19,983 voters, but the sum of the gender breakdown is inconsistent:
Women + Men (5,783 + 13,200) = 18,983
Stated Total = 19,983
Arithmetic Discrepancy: 1,000 Voters
The Grand Total for "Persons with Disability" is mathematically inconsistent with the sum of its regional components:
Sum of Regional Totals (20+4+9+3) = 36
Stated Grand Total = 27
Arithmetic Discrepancy: 9 Persons
This documents an earlier audit of the polling station-level data (older source images) against the summary total they were meant to support, revealing a significant lack of historical consistency.
This audit compares the total registered voters per country from Dataset 1 (Old: 34,296 total) against Dataset 2 (New: 34,411 total).
| Country | Old Reg. (Dataset 1) | New Reg. (Dataset 2) | Net Change | Discrepancy % |
|---|---|---|---|---|
| COTE D'IVOIRE | 736 | 969 | +233 | +31.66% |
| REP. CENTRAFRICAINE | 652 | 853 | +201 | +30.83% |
| CANADA | 1,186 | 1,505 | +319 | +26.89% |
| CONGO | 791 | 890 | +99 | +12.52% |
| EGYPTE | 269 | 205 | -64 | -23.79% |
| GABON | 6,267 | 6,267 | +0 | 0.00% |
| NIGERIA | 2,827 | 2,827 | +0 | 0.00% |
| GUINEE EQUATORIALE | 2,846 | 2,846 | +0 | 0.00% |
| ALLEMAGNE | 1,907 | 1,907 | +0 | 0.00% |
| FRANCE | 4,509 | 4,509 | +0 | 0.00% |
| ARABIE SAOUDITE | 779 | 779 | +0 | 0.00% |
| ITALIE | 1,015 | 1,015 | +0 | 0.00% |
| MAROC | 609 | 609 | +0 | 0.00% |
| AFRIQUE DU SUD | 988 | 988 | +0 | 0.00% |
| SENEGAL | 1,040 | 1,040 | +0 | 0.00% |
| GRAND TOTAL | 34,296 | 34,411 | +115 | +0.34% |
0.43%
Diaspora share of the National Electorate (7.9M+)
65.5%
Percentage of Diaspora Voters Who Are Men (22,549)
France (4,509)
Represents 13.1% of the Entire Global Diaspora Electorate
The analysis confirms that while the net change of +115 voters is minimal, the audit reveals systemic issues in data integrity across all phases of reporting: historical inconsistencies, internal arithmetic errors, and significant country-level adjustments (e.g., +31.66% in Côte d'Ivoire).
Learn the Data Skills for Free on Our YouTube Channel: Learn IT Free!