Análisis y comparación del índice kappa para la identificación de patrones en personas ancianas

Abstract

In the present work, a methodology is proposed for the creation of a corpus of videos of Alzheimer’s in which emotions are recognized, whose identification is validated by statistical methods that ensure the consistency and reliability of identification. For the construction of the corpus we used videos used for the analysis of the emotions experienced by the patients when they were interviewed, each video was analyzed by two scorers responsible for labeling the emotions manually and independently, using in both cases a window Time of twenty seconds. To validate the emotions detected by the observers we use the kappa and weighted kappa index. The kappa index has an average score of 0.60 which indicates a moderate agreement among observers and the weighted kappa index has an average score of 0.67 indicating good agreement among observers.