Unverified50% confidenceFactExact time
The evaluation metrics of the ADReSS benchmark are accuracy and F1, and in recent years most systems' accuracy has clustered in the 75%–85% range.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/21/2026
First Seen
Valid until: 1/19/2027
Sources
Related Entities
Related Claims
UnverifiedThe framework achieved a precision of 0.82, a recall of 0.85, and an F1 score of 0.82.75% similarUnverifiedThe best F1 scores for sarcasm detection tasks on standard datasets (such as iSarcasm and the SemEval series) have long hovered in the 70%–80% range.73% similarUnverifiedBrainAid achieved 98% precision, a 98% F1 score and 98% accuracy on public datasets.72% similarUnverifiedIn the benchmark test, Logistic Regression achieved 86.7% overall accuracy, 100.0% impact recall, and 0.0% cadence false positives.70% similarUnverifiedThe OCR model achieves an F1 score of 87% on a real invoice dataset.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1078828API
curl https://kongchang.com/api/v1/knowledge/claims/1078828MCP
get_claim(id=1078828)