Prediction Of The Action Identification Levels Of Teachers Based On Organizational Commitment And Job Satisfaction By Using K-Nearest Neighbors Method

Abstract

In this paper, the data mining techniques, which are quite hot in educational environment, are used to predict the action identification levels of the teachers. To this end, the organizational commitment and the job satisfaction levels are used as input to the data mining techniques. The well-known k-nearest neighbors (k-NN) approaches are considered due to their simple and non-parametric nature. Six different k-NN methods namely; fine, medium, coarse, cosine, cubic and weighted k-NN are considered and the obtained results are evaluated based on the prediction accuracy score. A dataset, which covers both organizational commitment and the job satisfaction levels of the teachers, is collected from 126 teachers. Extensive experimental studies are carried out with 5-fold cross validation test in MATLAB environment and the obtained results are recorded accordingly. The obtained results show that the proposed scheme is quite successful in prediction of the action identification levels. Especially, for some of the action identification levels, the obtained accuracy scores are 88.1%, 89.7% and 93.6%, respectively which show the success of the proposed idea.