K-Fold Cross-Validation through Identification of the Opinion Classification Algorithm for the Satisfaction of University Students

When using machine-learning techniques to determine algorithms or ranking models that identify student satisfaction, are often trained and tested on a single data set, leading bias in their performance metrics. This article aims the best algorithm classify satisfaction of university students applying K-fold cross-validation technique, comparing error rates metrics before after its application. The method used began with collection opinions teaching social network Twitter during an academic semester. Then, sentiment analysis was for processing, through which it possible categorize into “satisfied” “dissatisfied.” results showed lowest rate metric support vector machine (SVM). In addition, identified classification probability reached accuracy 91.76%. It is concluded SVM will contribute determining factors associated teacher’s didactic strategies should be improved each class session, since traditional surveying have shortcomings.