NUFORC Dramatic
and Unusual UAP
Sightings Classifier

Describe your sighting and see how dramatically it reads compared to 20,000+ real NUFORC reports.

What does "dramatic" mean? It refers to language patterns — not whether your experience was real or significant. NUFORC editors flag reports with unusual vocabulary, vivid technical phrasing, or descriptions that deviate from typical misidentification reports. A plain account of an extraordinary event can still score low.

Sighting Description

Details

Shape

State / Province

Country

Hour of Sighting (24h) 21

Permutations

Words to highlight 12

Your Score

—%

of NUFORC reports

TypicalUnusual

Describe your sighting on the left and click Classify Sighting to see where your report ranks.

Score: — · — percentile · Threshold: —

Model Architecture & Performance

Model: CatBoost Feats + Text (Orig) · Outcome: Dramatic UAP Sighting Classification · Data: NUFORC 2022–2026 (~20,000 reports)

Performance Metrics

Metric	Value
ROC-AUC	0.8140
Average Precision (AP)	0.3860
F1 — Bootstrap 2.5th pct	0.0850
F1 — Bootstrap 50th pct	0.3430
F1 — Bootstrap 97.5th pct	0.5060
F1 — Bootstrap Max	0.8770
Model Threshold	0.160

Hyperparameters

Parameter	Value
n_estimators	1,140
depth	8
learning_rate	0.03
l2_leaf_reg	10
min_data_in_leaf	10
subsample	1.0
leaf_estimation_method	Newton
early_stopping_rounds	150
text tokenizer	Space (Word tokens)
dictionary max size	50,000
text feature calcers	BoW, BM25

NUFORC Dramaticand Unusual UAPSightings Classifier

Model Architecture & Performance

Performance Metrics

Hyperparameters

Notes

NUFORC Dramatic
and Unusual UAP
Sightings Classifier