Human errors in numerical data entry can lead to serious consequence but it is difficult to predict thoseerrors because mechanisms of human errors vary and no contextual clues are available. This study suggestsintegrating human behavior modeling and data mining as an advanced method to predict human errors.Human behavior modeling utilized top-down inference to transform interactions between taskcharacteristics and conditions into general inclination of an average operator to make errors, while datamining parsed psychophysiological measurements into individual’s likeliness of making errors on a trialby-trial basis through bottom-up analysis. Specifically, an enhanced Queuing Network-Model HumanProcessor (QN-MHP) generated modeling features to be combined with real-time EEG features that werecollected in a realistic numerical typing experiment, and potential errors were predicted by detecting errorassociatedfeatures by linear discriminant analysis (LDA) classifiers before responses. The detection couldbe made as early as 300 milliseconds beforehand, and the results showed that integration improved theLDA classifiers’ performance by 31.7% in keenness (d') and by 12.5 % in area under ROC curve (AUC)from that of using EEG only. The integration may help implement future adaptive augmented system toprevent cognitive breakdown by determining appropriate automation/augmentation levels.
展开▼
