: Vocabulario de PLN: recall

NLP resources and evaluation > measure > confusion matrix > recall

recall

A performance metric representing the percentage of correct answers successfully retrieved in a given dataset.

Contrarily a true positive rate of 62.6% indicates that participants were better at identifying real news articles as human-written. (Almasi & Schi\\onning, 2023)
For evaluation metric we calculate recall on the instances in the held-out set. (Bing, Chaudhari, Wang & Cohen, 2015)
The recall is lowest for V-V derivations suggesting that the derivational phenomena for this POS combination are not yet covered satisfactorily. (Zeller, Šnajder & Padó, 2013)
Through a detailed analysis we conjecture that this sensitivity arises when LLMs are uncertain about the prediction between the top-2/3 choices and specific options placements may favor certain prediction between those top choices depending on the question caused by positional bias. (Pezeshkpour & Hruschka, 2024)
To evaluate with AUC score we first plot the ROC curve through the True Positive Rate (TPR) and False Positive Rate(FPR). (Ye, Hu, Li, Wang, Chen & Zhao, 2024)

http://data.loterre.fr/ark:/67375/8LP-MFTWN0D1-0

RDF/XML TURTLE JSON-LD última modificación 26/6/24

Loterre